|USFDC Home | USF Electronic Theses and Dissertations||| RSS|
This item is only available as the following downloads:
Generation of Carbon Dioxide and M obilization of Antimony Trioxide by Fungal Decomposition of Building Materials by John D. Krause A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Environmenta l and Occupational Health College of Public Health University of South Florida Major Professor: Yehia Y. Hammad, Sc.D. Noreen D. Poor, Ph.D. Ann C. Debaldo, Ph.D. Diane Te Strake, Ph.D. Date of Approval: March 25, 2005 Keywords: mold, mould, carbon dioxide, antimony trioxide, flame retardant Copyright 2005, John D. Krause
Dedication For their love, support, patience and unde rstanding throughout this endeavor, I dedicate this work to my family, daught er, and most of all, my loving wife.
Acknowledgements I would like to acknowledge the following individuals and companies for their assistance in this research. Daniel Webster and Cliff Wolcott of Quest Technologies for supplying equipment and supplies to monitor carbon dioxide. Dr. Morten Reeslev, and Dr. Morten Miller of MycoMeter, Aps in Copenhagen, Denmark for supplying fungal isolates and advice for use of the MycoMeter test method. Dr. Lori Streit and Dr. Robert Scarry for their advice on analytical methods Dr. Lauren Ball for her motivation, advice, and support in the completion of my coursework and research. I also acknowledge my fellow graduate stude nts at the University of South Florida College of Public Health. One of the greates t resources I found were those students with diverse backgrounds and experi ences who often helped me to see things in a very different light. I am convinced that half of what I learned in graduate school was through discussions with fello w graduate students.
i Table of Contents List of Tables iii List of Figures v Abstract vii Introduction 1 The Paradox of Mold Exposure and Health Effects 5 Literature Review Health Effects of Exposure to Fungi 8 Prevalence of Mold Growth in Buildings and on Building Materials 18 Quantifying Mold Growth Using -Nacetylhexosaminidase Activity 19 Finding Hidden Mold Growth 22 Use of Carbon Dioxide to Detect Metabolic Ac tivity of Mold Growth 23 Studies of Carbon Dioxide Production from Mold Growth 25 Mold Growth on Gypsum Wallboard 27 Mold Growth on Fiberglass Ductboard 29 Mechanisms of Fungal Growth and Decomposition of Materials 31 History of Scheeles Green 35 Biomethylation of Metalloids (As, Sb, and Bi) 36 Prevalence of Antimony in Buildin g Products 42 Health Effects of Exposure to Antimony Compounds 46 Trends in Antimony Use and Asthma Prevalence 50 Focus of Research 55 Carbon Dioxide as a Screening Test for Hidden Mold Growth 55 CO 2 Generation Rates During the Life Cycles of Two Mold Species 56 Potential for Antimony Release from Sb 2 O 3 Flame Retardant as Trimethylstibine 57 Measuring Antimony Release from Mold Growth on Sb 2 O 3 Flame Retardant-Treated Building Materials 58
ii Materials and Methods 61 CO 2 Generation from Fungal Grow th on Gypsum Wallboard and Fiberglass Ductboard 61 Mobilization of Antimony Trioxide Due to Fungal Growth on Fiberglass Ductboard 73 Measurement of Fungal Biom ass on Gypsum Wallboard and Fiberglass Ductboard 77 Statistical Analysis of Measurem ents 78 Results CO 2 Production Rates on Gypsum Wallboa rd 79 CO 2 Production Rates on Fiberglass Ductboa rd 81 Modeled Concentration of CO 2 in Wall Cavities 85 Modeled Concentration of CO 2 in HVAC ducts 87 Discussion 88 Measured Emission Rates of Stibine as Total Antimony 90 Measured Antimony Levels in Test Materials with Fungal Growth 91 Estimated Antimony Emission Rates 94 Modeled Concentration of Trimethylstibin e Oxides as Antimony 96 Discussion 99 Concluding Remarks Public Health Significance of Research 103 Practical Applications of This Research 106 Study Limitations 106 Recommendations for Future Research 108 References 110 Appendix 1: Carbon Dioxide Quality Cont rol Data 121 Appendix 2: Antimony Quality Control Data 126 About the Author End Page
iii List of Tables Table 1 Products containing antim ony trioxide fire retardant 45 Table 2 ICP analysis results from preliminary study to identify materials containing Sb 2 O 3 fire retardant 59 Table 3 Carbon dioxide production rates for S. chartarum on gypsum wallboard 80 Table 4 Carbon dioxide production rates for A. versicolor on ductboard surface coating 81 Table 5 Carbon dioxide production rates for A. versicolor on ductboard FSK exterior wrapping 84 Table 6 Modeled concentrations of CO 2 inside a wall cavity colonized with (S. chartarum) 86 Table 7 Modeled concentrations of CO 2 inside a static supply duct ( A. versicolor ) 87 Table 8 Antimony in fiberglass ductboard surface coating test material 91 Table 9 Antimony in fiberglass ductboard FS K exterior wrapping test material 93 Table 10 Modeled concen tration of antimony as trimethylstibine oxides 98 Table 11 Empty chamber #4 on 5/29/2004 121 Table 12 Gypsum wallboard control CO 2 test on 5/22/2004 122 Table 13 Ductboard surface coating control CO 2 test on 3/29/2004 122 Table 14 Ductboard surface coating control CO 2 test on 4/49/2004 123 Table 15 Ductboard surface coating control CO 2 test on 4/11/2004 124 Table 16 Ductboard surface coating control CO 2 test on 4/25/2004 124
iv Table 17 Ductboard FSK ex terior wrap control CO 2 test on 5/2/2004 125 Table 18 Antimony analysis results fo r control test pieces on 3/28/2004 126 Table 19 Antimony analysis results fo r control test pieces on 5/30/2004 126 Table 20 Antimony calibration data 128
v List of Figures Figure 1 Photographs of mold growth on water damaged gypsum wallboard 28 Figure 2 Photographs of mold growth on fiberglass duct liner 30 Figure 3 Periodic table sect ion showing metalloids 37 Figure 4 Typical reactions of the Challenger mechanism 38 Figure 5 Challenger mechanism for the conver sion of arsenate to trimethylarsine 39 Figure 6 Antimony oxide imports and asthma prevalence in the U.S. 52 Figure 7 Antimony oxide imports vs. asthma prevalence in the U.S. 52 Figure 8 Photograph of test ri g for gypsum wallboard 63 Figure 9 Photograph of test chamber, acry lic lid, inlet and exhaust fittings 66 Figure 10 Diagram of test equipment 68 Figure 11 CO 2 generation and biomass density on gypsum wallboard 80 Figure 12 CO 2 generation and biomass density on ductboard surface coating test #1 82 Figure 13 CO 2 generation and biomass density on ductboard surface coating test #2 82 Figure 14 CO 2 generation and biomass density on FSK exterior wrapping 84 Figure 15 Sb concentration and biomass de nsity on fiberglass ductboard test #1 92 Figure 16 Sb concentration and biomass de nsity on fiberglass ductboard test #2 92 Figure 17 Sb concentration and bioma ss density on FSK exterior wrapping 93 Figure 18 Modeled Sb concentrations in a home with all ducts supporting mold growth 98
vi Figure 19 Modeled Sb concentration in a home with one duct supporting mold growth. 99 Figure 20 Monitoring data from an empty chamber on 5/29/2004 121 Figure 21 Monitoring data from control chamber on 4/49/2004 123 Figure 22 Monitoring data from a control chamber on 5/2/2004 125 Figure 23 Antimony calibration 3-28-04 129 Figure 24 Antimony calibra tion curve 4-4-04 130 Figure 25 Antimony calibra tion curve 4-11-04 131 Figure 26 Antimony calibra tion curve 4-24-04 132 Figure 27 Antimony calibra tion curve 5-2-04 133 Figure 28 Antimony calibra tion curve 5-16-04 134 Figure 29 Antimony calibra tion curve 5-30-04 135 Figure 30 Antimony calibra tion curve 6-27-04 136 Figure 31 Antimony calibra tion curve 7-31-04 137
vii Generation of Carbon Dioxide and M obilization of Antimony Trioxide by Fungal Decomposition of Building Materials John D. Krause Abstract Fungal contamination of buildings po ses numerous challenges to researchers, building owners and occupants. Public health agencies promote prevention and remediation of mold and water damage, but la ck sensitive methods to detect hidden mold growth and a complete understa nding of the biological mechan isms that make occupying moldy buildings a hazard. The wide spread use of the fire reta rdant antimony trioxide (Sb 2 O 3 ) on building materials and furnishings makes it inevitable that mold growth on treated materials will occur in some buildings with water damage. Several authors have speculation that microbial growth on materi als treated with antimony trioxide could mobilize antimony through a volatile in termediate, trimethylstibine. The purpose of this study was to dete rmine if fungal growth on a commonly used building material that contains antimony trio xide, fiberglass ductboard, results in the mobilization and release of antimony compounds. Additionally, CO 2 generation rates
viii were determined during fungal growth on fi berglass ductboard and gypsum wallboard. Results demonstrated a significant reduction of antimony concentration in fiberglass ductboard after fungal growth had occurred. Antimony emission rates and resulting concentrations of antimony oxide aerosols were estimated using an indoor mass balance mathematical model. Concentrations of CO 2 were also modeled within a wall cavity and static HVAC ducts to determine if fungal growth could elevate CO 2 levels above ambient concentrations. Although volatile phase antimony was not detected in chamber experiments, probably due to rapid oxidation and high humidity, mobilization of antimony trioxide from fiberglass ductboard components was demonstrated in several experiments. Indoor Air modeling of a residence suggest that concentrations of antimony could, under worst case conditions, exceed the reference concentration (RfC) of antimony trioxide by 10 to 1,000 times. These results suggest that biomethylation and mobilization of antimony by mold growth on building materials could re sult in elevated occupant exposures to antimony compounds. Antimony is a known respirat ory irritant that can be similar to arsenic in its toxicity. Modeling results also sugge st that elevated carbon dioxide concentrations due to fungal metabolic respiration are variable and dependent on environmental conditions. Measuring elevated carbon dioxide concentrations to detect hidden fungal growth was determined to not be a predictive assessment tool.
1 INTRODUCTION Indoor mold contamination has been the subj ect of recent interest from the news media, legislators, insurance companies and homeowners. Mold exposures have been a subject of concern for U.S. public health agencies, including the Centers for Disease Control and Prevention (CDC), the National In stitute for Occupational Safety and Health (NIOSH) and the U.S. Environmental Protec tion Agency (EPA) si nce the early 1990s (EPA, 1992; Macher, 1999). The Bioaerosols Committee of the American Conference of Governmental Industrial Hygi enists (ACGIH) defines biological contamination in buildings as the presence of (a) biologically derived aerosols, ga ses, and vapors of a kind and concentration likely to cause disease or predispose persons to adverse health effects, (b) ina ppropriate concentrations of outdoor bioaerosols, especially in buildings designed to prevent their entry, or (c) indoor biological growth and remnants of gr owth that may become airborne and to which people may be exposed (Macher, 1999 p.1-1). Due to the dearth of information on exposur e and disease associated with living and working in indoor environments with fungal growth, regulatory action has been slow and inconsistent (IOM, 2004).
2 Fungi colonize wet building materials in the indoor environment causing material degradation, volatile organic compound (VOC ) production, and release spores and hyphal fragments containing potent a llergens and respiratory irrita nts (EPA, 2001; Gorny et al., 2002; Macher, 1999). Some physicians and bu ilding occupants claim that exposure to indoor fungi results in a myriad of sy mptoms and diseases (Fung & Hughson, 2003; Hodgson et al., 1998; Johanning et al., 1996). However, exposure to ubiquitous fungal spores and hyphae in the outdoor environmen t is unavoidable (Burge, 1995; EPA, 1992). Many people inhabit buildings with extens ive mold growth and do not experience recognizable adverse health effects, while ot hers seem to react severely. The specific allergens, irritants or toxins responsible for occupant react ions to fungi are under study, but no consensus has been reached on whic h agents are responsible for the reported symptoms. Outdoor air spore concentra tions typically exceed indoor building concentrations with and without a history of water damage (Horner et al., 2004; Shelton, et al., 2002). Median outdoor f ungal concentrations were 6 to 7 times higher than indoor levels in an analysis of over 12,000 air sa mples collected throughout the United States (Shelton, et al., 2002). Some emphasis has b een made on the hazard that certain fungal species pose when growing indoors. Ho wever, no consensus has been reached on whether certain species pose a special hazard. Public health professionals examining the subject have advocated that i ndoor mold growth is a public he alth risk, irrespective of the species present (EPA, 2001; IOM, 2004; M acher, 1999; NYC DOH, 2000; OSHA, 2003). This conservative guidance attempts to pr otect all building occ upants, but does not resolve the biological mechanisms c ontributing to occupant illness.
3 Modern buildings can suffer from water da mage resulting in mold growth. Much of this damage occurs within wall cavities and hidden spaces such as heating, ventilation, and air conditioning (HVAC) ducts. Fungal bioaer osols generated within interstitial wall cavities can migrate into occupied areas of a building (Macher, 1999). HVAC ducts can spread airborne contaminants to the occupied areas exposing occupants to fungal bioaerosols generated in these hidden sp aces (EPA, 1995). Currently there are no measurement methods to detect fungal growth present within hidde n cavities that have been validated. Detecting hidden mold growth early could provide an important tool for protecting public health. Researchers have sp eculated that volatile emissions from mold growth may cause occupant illness. Gao et al. have conducted research on the volatile organic compounds produced by fungal growth a nd determined the concentrations within a building to be small, but possibly useful in detecting hidden mold growth (Gao et al., 2002; Macher, 1999). However, little informa tion is available on volatile emissions of metalloids due to mold growth on building materi als. The hazard from fungal growth on building materials known to contain toxic meta lloids has been essentially ignored since the early 1900s. Mold growth on building materials treat ed with the flame retardant antimony trioxide (CAS 1309-64-4; Sb 2 O 3 ), such as fiberglass ductboard, is inevitable in modern buildings. The increased use of synthetic ma terials to insulate and construct buildings has resulted in the greater use of fire reta rdants. Research on the biomethylation of antimony (Sb) by microorganisms has demonstrat ed that fungal and ba cterial growth can mobilize and release antimony as a toxic, volatile gas called trimethyl stibine (Grleyk et
4 al., 1997; Jenkins et al., 1998a). These findings raise the con cern that mold growth on flame retardant-treated building materials may release antimony as a toxic gas, which rapidly oxidizes into less toxic, but still hazardous, particulate matter. Occupant exposure to antimony resulting from fungal d ecomposition of fire retardant treated building materials has not been previously st udied. One purpose of this research study was to examine the possibility of antimony mobilization from a commonly used building material that often supports fungal growth when water damaged.
5 THE PARADOX OF MOLD EXPO SURE AND HEALTH EFFECTS A paradox is defined as an apparent cont radiction with common sense. Public health organizations recognize exposure to fungi, spores, volatile organic compounds and fungal products as a human health risk (EPA, 1992; IOM, 2004; Macher, 1999). The ubiquitous nature of fungi in the outdoor and indoor environments, in food, and industry highlights the paradoxical nature of this c oncept. Fungi constitute up to 25% of the biomass on this planet (McNeel and Kreutzer 1996). Fungi are the primary decomposers in the environment, breaking down leaves, gr ass cuttings, garden compost, trees, and organic material in landfills (Burge, 1997). Comparatively, fungi contribute more biomass to soil than all other microorga nisms. Many fungi ar e saprophytes, obtaining nutrition from dead plants and animals, c ontributing to decompos ition and recycling of nutrients (Deacon, 1984; Kendrick, 1985). The ea rth would be a differe nt place if fungi did not exist. Fungi are used to produce cer tain chemicals and flavoring agents used in household products and foods. Foods like chees e, beer and wine are dependent on fungi and yeast for their production and unique flavors (Burge, 1997; EPA, 1992; Kendrick, 1985). When fungal growth is allowed to go unchecked, and certain environmental conditions exist, it can become a hazard to plants, animals, humans and buildings. Many common and exotic plant pathogens are fungi. St ored grains typically contain some small
6 amount of fungal components, and can be spo iled if fungal growth takes place. Some of the most common infections in animals and hum ans are due to fungi, such as "athlete's foot" and the dermal disease ringworm or Tinea Corporis. When the immune systems of animals and humans are compromised they can succumb to debi litating or lethal fungal infections. Certain fungal spores can cause opportunistic infections in immunocompromised individuals (Fung a nd Hughson, 2003). Ingestion of toxic metabolites produced by some fungi growing on grains and feeds, by animals or humans can result in a variety of diseases. Epidemics of ergot ism, known also as St. Anthonys fire, experienced by Europeans in the Middle Ag es were due to inge stion of grains and bread containing mycotoxins. Ergotism results in the loss of peripheral circulation, gangrene, and death (EPA, 1992). Widespread equine epidemics in Poland, Hungary and Russia in the early 1900s killed thousands of horses from i ngestion of hay contaminated with Stachybotrys chartarum This fungus was found to prod uce a series of potent toxins called trichothocenes that resulted in the death of horses that fed on the hay (Lacey, 1985). Many fungi can produce colonies that are visible to the unaided eye, often making their growth on water damaged materials obviou s. The health risk from indoor fungal growth was recognized in ancient times and discussed in the Bibl e (Leviticus Chapter 14, 33-48), where a reference is made to mold gr owth in a home and measures to clean or discard all affected materi als are described. Not visi ble to unaided eye are the bioaerosols they produce, but the odors th ey sometimes emit can be perceived.
7 Accurately measuring fungal bioaerosols and determining the biologically active components have been evasive goals for resear chers and public health professionals. So far researchers have been unable to reach a consensus on useful biomarkers of exposure to environmental fungi in humans. Our world would not be the same without fungi, but a variety of diseases and adverse health outcomes can be attributed to fungi. It is apparent that when fungi are provided an environment they can exploit, they can grow out of control and harm plants, animals, people, and buildings. The environm ental conditions that enable fungi to grow are well documented, but the specific etiologic agents that may cause allergies, irritation and numerous non-specific symptoms in people have not been elucidated.
8 LITERATURE REVIEW Health Effects of Exposure to Fungi Evidence suggests that exposure to many fungal products can result in adverse health effects and decreased quality of life. In general, fungal aerosols and other bioaerosols found in damp environments, can cause allergy, inf ection, irritation, and toxicity. Determining the speci fic cause of these health e ffects has been impossible due to the presence of numerous microorga nisms and bioaerosols in damp indoor environments. Chronic and/or repeated e xposure to fungi has been associated with allergy, hypersensitivity pneumonitis, asthma symptoms in sensitized asthmatics, upper respiratory tract symptoms, wheeze and cough. Allergy and hypersensitivity disease are IgE-mediated immune responses to sp ecific allergens, while hypersensitivity pneumonitits (HP) is an IgG and T cell-mediated response (EPA, 1992; IOM, 2004). Allergies in genetically predisposed people can result from exposure to the antigenic components of fungi. From a study of workers in U.S. office buildings, the overall self-reported prevalence of physiciandiagnosed allergy to mold was 22%, with 18% in male respondents and 24% in female respondents (n = 2,435) (Crandall & Sieber, 1996b).
9 Asthma attacks can be initiated in people with allergic asthma and sensitivity to fungi (Fung and Hughson, 2003). The increased prevalence of asthma in the United States, and other industrialized countries, (CDC, 1998) has been speculated to be connected with indoor air pollution. Data clea rly indicate that expos ure to fungi plays a role in asthma. Several studies have doc umented the sensitizing potential of fungal allergens and relate th e existence of asthma to the pa rticipants sensitization. Although the number of studies is too few, there are data that support a rela tionship between fungal allergen sensitization and sy mptoms of asthma. In 2000 the Institute of Medicine Committee on the Assessment of Asthma and Indoor Air concluded there was sufficient evidence to demonstrate that exposure to f ungi could exacerbate symptoms in sensitized asthmatics. However, sufficient evidence was not available to determine if exposure to fungi caused the development of asthma (IOM, 2000). As more data becomes available about damp indoor environments, exposure to the microorganisms and the chemicals present, a clearer understanding should evolve. The underlying causes for the drastic incr ease in asthma prevalence throughout the United States are still unknown. Self-repor ted prevalence of as thma increased 75% from 1980 to 1995 for the US population as a whole. This equates to a 5% increase per year, with the greatest increase among children age 0-4 ye ars. This age group increased 160%, from 22.2 per 1,000 to 57.8 per 1,000 (CDC, 1998). The underlying cause of this drastic increase is unknown, but many have spec ulated it to be related to degraded indoor air quality. Some researchers have speculated the increased asthma prevalence is due to fungal contamination of indoor environments while many others point to the overall
10 microbial burden present in damp indoor envi ronments. Regardless of the speculation so far no consensus has been reached in light of the inconclusive and contradictory studies (IOM, 2000; IOM, 2004). Speculation has been made about the ro le of mycotoxins produced by fungi growing indoors. Many fungi can produce seco ndary metabolites that are classified as mycotoxins. It is estimated that the number of fungal species capable of producing mycotoxins range from 100-150. Over 200 diffe rent mycotoxins have been described (EPA, 1992). Toxigenic fungi and their spor es are ubiquitous, but documented cases of mycotoxicosis have been primarily due to ex tremely wet conditions and mold growth on specific materials (EPA, 1992; Sharma a nd Salunkhe, 1991). Mycotoxins are found on and in the spores of fungi that produce them, on the hyphae and in the dust from substrates where they have grown. Though not considered volatile, some mycotoxins have been shown to migrat e into the substrates that mold grow on (IOM, 2004). Mycotoxins range in their toxicity and eff ects, from mild acute toxicity to potent carcinogenicity. Most mycotoxins are cytoto xic, resulting in ce ll death (EPA, 1992). Concerns over the role mycotoxins may have pl ayed in a cluster of infant deaths due to acute idiopathic pulmonary hemorrhage (AIPH) arose after reports of a CDC investigation. Investigators in itially concluded that an asso ciation existed between infant AIPH and exposure to the toxigenic fungus Stachybotrys chartarum but later review by an external task force dismissed the associati on due to a lack of substantiating data (CDC, 1994; CDC, 2000; Montana et al., 1997).
11 Associations between exposure to mycot oxin-containing fungi and occurrence of disease in building occupants ha ve been rare. The vast majo rity of reported illnesses due to mycotoxins have been due to ingestion of contaminated food or in agricultural settings when grain handlers have been exposed to dense clouds of dust from contaminated grains. A critical review of the literature by Fung and Hughson concluded that mycotoxin levels in most mold-contaminated buildings are not likel y to result in a dose sufficient to cause measurable health effects (Fung & Hughson, 2003; Robbins et al., 2000). Researchers have been unable to de monstrate a dose response relationship between toxin-containing spores and occupant illness, in part due to the inadequacy of measurement and analysis methods for e xposures to fungal aerosols (i.e. spores, allergens, irritants, toxins and volatile meta bolites) (Canadian Public Health Association, 1987; Cohen & Hering, 1995; Dales et al., 1997 ; EPA, 1992;). The ideal measure of exposure to fungal aerosols is personal sampling of the breat hing zone, but until recently, personal samplers were not available to collect bioaerosols. Many drawbacks have been identified with area samples used to measur e fungal spore exposures. In addition to the bias introduced by stresses pl aced on the spores during cap ture, desiccation and culture onto agar media, contribute to underestim ating fungal spore concentrations (Macher, 1999). Studies have also demonstrated th e presence of a personal cloud created by building occupants. Air concentrat ions of particles smaller than 10 m (PM 10 ) measured in the breathing zones of subjects wearing personal particle monitors were compared to concurrent results from area samples. Resu lts indicated that area samples underestimated
12 personal exposures to PM 10 by 50%. Most fungal spores are in cluded in this size range of particles (EPA, 1992; IOM, 2000; Ozaynak et al., 1996). The etiologic agents that may cause the myriad of symptoms described by various physicians, researchers and building occupants in water damaged buildings have not been determined. Some of the unsubstantiated symptoms claimed to be due to mold exposures including dyspnea (shortness of breath), airflow obstruction, mucous membrane irritation, inhalation fevers, lower resp iratory illness, rheumatologic immune diseases, acute pulmonary hemorrhage in infants, skin symptoms, development of asthma, gastrointestinal disorders, fa tigue, neuropsychiatric symptoms and cancer. Despite case studies, anecdotal reports and media hype, rese archers and public health agencies have not found a link between these symptoms a nd exposure to moldy indoor environments (IOM, 2004). Despite the lack of studies and conclusive data, there is sufficient evidence in the literature that upper respirat ory tract symptoms, exacerba tion of asthma, wheeze, and cough are associated with both exposure to f ungi in damp environments and exposure to damp environments in general. Exposure to damp environments, with or without the presence of visibly detectable mold, appears to carry similar h ealth risks. Increased odds ratios for asthma-related health effects associ ated with dampness indicators were reported in at least thirteen studies published betw een 1993 and 1999. The symptoms described as asthma-related included wheeze, persistent cough, bronchial obstruction and asthma. Despite finding a positive asso ciation between these sympto ms and exposure to damp
13 environments the committee recognized the likelihood of multiple exposures to bacteria, fungi and dust mite allergens (IOM, 2000). Until studies with greater power are performed, that are capable of quantifying occupant exposure to the causative agents of illness, an association will be difficult to determine, if one truly exists. Adding to the difficulty in differentiating between the effects of multiple exposures to bioaerosols in damp environments, researchers must account for exposure to ubiquitous fungal spores present in the outdoor environment. Indoor spore concentrations are typically lo wer than outdoors, even in some buildings with visible mold growth (Horner et. al ., 2004; Macher, 1999; Sh elton et al., 2002). Humans have been exposed to fungi throughout history and are constantly ingesting and inhaling mold and their spores without detect able adverse reaction. Is it possible that exposure to fungal bioaerosols produced by mold colonies growing indoors poses a greater risk than outdoor fungal bioaerosols? The guidance currently provided by public health agencies including the CDC, US EPA, ACGIH, AIHA, OSHA, NIOSH a nd numerous state and city health departments is that indoor mold growth pos es an unacceptable health risk and should be remediated immediately, regardless of th e species present (AIHA, 2001; EPA, 2001; IOM, 2004; NYC DOH, 2000; OSHA, 2003). Guid ance for precautions and the extent of remediation necessary to protect building o ccupants and remediation personnel is only now under development (NIOSH, 2002). The methods of remediation and post-
14 remediation conditions that constitute a succe ssful and effective remediation are still under debate. Studies of human populations living a nd working in buildings with fungal contamination (i.e. growth) have not dem onstrated consistent correlations between measures of airborne spores or amounts of fungal growth, and the frequency or severity of occupant illness (IOM, 2004; Macher, 1999). The inability to detect correlations between airborne spore concentr ations and health outcomes or occupant complaints is in part due to the inadequacy of air sampling instruments and analytical methods. The inconsistency of findings and the inability to determine a dose-response curve or threshold level of exposure has frustrated researchers and public health agencies. Consensus is that human exposure to i ndoor fungi is predominantly from fungal spores via the inhalation route. The outdoor concentrations of these spores can vary by four orders of magnitude on a daily basis and indoor envi ronments often have lower spore concentration measurements than outdoors, even in buildings with extensive fungal growth (European Collaborate Action, 1993; Horn er et al., 2004; She lton et al., 2002). For example, extensive sampling (n = 476) of an office building over 14 months did not reveal elevated fungal levels even though fungal contaminati on was found in the air ducts (Burge et al, 2000). If exposure to fungal spor es were the sole cause of adverse health effects, and spore concentrations indoors are lower than outdoors, why should people suffer ill effect?
15 A guide post frequently described in the li terature is comparison of indoor versus outdoor air concentrations and the rank order of species. This comparison initially seems relevant because fungi are naturally occurri ng organisms, most lik ely originating from outdoor sources that enter a building via natu ral or mechanical ventilation (Baughman & Arens, 1996; EPA, 1992). The assumption made is that if indoor spore concentrations are higher than outdoor concen trations of the same species or if the biodiversity of species is shifted in th e indoor environment, then an indoor reservoir is likely present. While this approach can sometimes help an investigator identify an area with mold damage, it cannot be used to demonstrate th e absence of an indoor reservoir of fungal growth. Because of the temporal variability and infrequency of spor e release from indoor reservoirs, air sampling from indoor environm ents can result in false negatives. The probability of air sampling methods not detecting the presence of fungal growth when it truly exists, also known as the method speci ficity, has not been reported for existing methods (Dillon et al., 1996). The indoor to outdoor ratio approach is ba sed on an assumption that exposure to aerosols from indoor fungal growth poses the same risk as exposure to aerosols from outdoor fungal growth. That is fungal grow th occurring outdoors, in soil and on decaying leaves and trees, releases spor es with the same composition, a llergenicity and toxicity as fungal growth occurring on synthetic, treated and preserved man-made materials. If this assumption does not prove to be true a nd exposure to spores, hyphal fragments and volatile organic compounds (VOCs) from i ndoor fungal growth, poses a different risk
16 than exposure to outdoor fungal aerosols, then researchers may need to reconsider the relevance of the indoor to outdoor comparison. Research at the University of Cincinnati has revealed that in addition to fungal spores, sub-micron size hyphal fragments can be released from colonies of fungal growth. These fragments are typically smaller than 1 m in diameter and are probably not viable, meaning they cannot form a colony (Gorny et al., 2002). These two properties have significance in that microscopic analysis of spore-trap samples, taken to measure total spore concentrations indoors, cannot resolve or identify particles less than 1 m, essentially making these hyphal fragments undetectable by conven tional spore trap analysis methods. Because these fragments ar e not whole cells they are not likely to be viable. Commonly used impaction plate met hods used to capture and culture viable spores cannot detect the presence of non-viable fragments. Until this report by Gorny, et al. the existence and magnitude of this fungal aerosol component was unrecognized by researchers, scientists and public health professionals performing building assessments. This initial research on hyphal fragments al so examined the allergenicity of these sub-micron hyphal fragments. The research ers concluded that fungal fragments are released independently of spores, with th e number of fungal fragments exceeding spores by 2-3 orders of magnitude. These fungal fragments exhibited 2 to 5 times greater immunological reactivity than spores in monoc lonal antibody assays. These data suggest fungal fragments may have grea ter allergenicity than spores and their contribution to occupant allergic reactions may be greater (Gorny et al., 2002). This research also
17 suggests that much of the surrogate exposure data (i.e. tota l and viable spore concentrations in air) may be missing a significant component of fungal bioaerosols. The overall confusion and complexity of the exposure-dose-response paradigm may be due more to a lack of resolution than inconsistency of data. When the overall subject of Damp Indoor Spaces and Health was examined in a r ecently published report by the National Academy of Sciences Institute of Medicine, a very clear association was recognized between exposure to damp indoor spaces and increased health risks. However, because of inconsistent research methods and measures of damp environments and fungal exposures, a causal e ffect could not be ascertained from the numerous studies reviewed. The committee di d find sufficient evidence in the literature that exposure to damp indoor environments was associated with upper respiratory tract symptoms, cough, wheeze and asthma symptoms in sensitized asthmatic persons. A similar association with visible mold was found, but it was re cognized that other organisms and chemical agents could be contributing to the symptoms and illnesses reported (IOM, 2004). Using a broad base of numerous epidemiological studies, the IOM committee recognized an association be tween the presence of fungal growth and certain health effects, but due to the lack of exposure da ta, quantifiable biomarkers and consistent findings, a causal agent was not identified. Ironically, visible signs of indoor mold growth were more predictive of adverse h ealth outcomes than measured indices of exposure such as airborne spore concentrations. This sugge sts two possibilities that are
18 not mutually exclusive. One is that other organisms that thrive in damp environments and chemicals released when materials become wet may be contributing to or causing the reported symptoms and illnesses. The other pos sibility is that exposure to other fungal aerosols, besides spores, may be the causative agent. In light of the recently described fungal hyphal fragments (Gorny et al., 2002), the possibility of one or more un-measured fungal aerosol contributing to adverse health effects seems to exist. Prevalence of Mold Growth in Bu ildings and on Building Materials The prevalence of mold growth in U.S. buildings has not been fully reported. How much mold growth constitutes a health hazard is also unknown. In the 1970's and 1980's microbial contamination wa s identified as the primary cause for poor air quality in only 5% of 529 indoor air quality (IAQ) i nvestigations conducted by the National Institute for Occupational Safety and Hea lth (NIOSH); The remaining 95% were attributed to inadequate ventilation, entrainment of outdoor air contaminants, contaminants in building fabric and unknown sources (Crandall & Sieber, 1996a). Assessment of 104 building evaluations in 1993 revealed many conditions that can support mold growth were identified such as poor condensate pan drainage (16%), dirty HVAC system components (36.5%), and water d amaged duct liner (1%). The findings of this NIOSH study suggested that poor maintenance of HVAC systems related to building occupant complaints. Poor maintenance of HVAC systems can lead to proliferation of microbial contamination resulting in the HVAC systems becoming a contaminant source for occupied spaces of the building (Crandall & Sieber, 1996a). However, in the last 10 years, microorganisms were determined to be the primary source
19 of indoor air contamination in as many as 35-50% of IAQ cases (Lewis, 1994). Current trends in occupant complaints and pub lic awareness suggests the problem may be widespread, especially afte r seasonal storms, floods a nd hurricanes that cause both structural damage to buildings and introduce the necessary water for mold spores to germinate and grow. There is currently no monitoring program to track either the number of homes, schools or workplaces w ith mold growth or complaints. Mold growth on two of the major build ing materials currently used, gypsum wallboard and fiberglass ductboard, has been well documented (Price et al., 1994; Samimi and Ross, 2003; Van Loo et al., 2004; Reeslev et al., 2003; Hodgson et al., 1998; Doll, 2002; Flappan et al., 1999; Johanning et al., 1996). Studie s have shown that soiled building materials are more susceptible to mold growth at lower moisture levels than new materials (Chang et al., 1995; Gravesen et al., 1999; Samimi et al., 2003). The key determinant of fungal growth on building materials is water availability, although nutrients, pH, oxygen, and temperature can al so effect fungal growth (Doll, 2002). Quantifying Mold Growth Using -Nacetylhexosaminidase Activity Quantifying mold growth on building material s is impractical using most standard mycological methods. One variable is the surface area of affected material, but the density of mold growth can also vary drastica lly. An ideal indicator of mold growth is biomass density. Using the total surface area of growth and mold biomass density, the amount of mold biomass present in a building could be estimated.
20 Measuring the dry weight (i.e., mass) of laboratory cultures is sometimes performed, but samples must be handled care fully. Measuring the mass of samples taken from building materials cannot differentiate between mold biomass and other non-mold constituents. Instead, measurements of ch emical indicators such as ergosterol or -N acetylhexosaminidase enzyme activity have be come recognized as useful surrogates of mold biomass. Both of these markers are believed to be common to all known filamentous fungi, but some inter-species va riability exists (Nie lsen and Madson, 2000; Reeslev, et al., 2003; Schnurer, 1993). Ergosterol has been used to monitor and es timate mold biomass in seeds, grains, decaying wood and building materials. The analysis of samples for ergosterol requires extensive sample extraction and clean-up along with laboratory analysis by gas chromatography/mass spectrometry (GC/MS), high performance liquid chromatography (HPLC) or spectrophotometry (Nielsen and Madson, 2000; Schnurer, 1993). In general, there has been demonstrated a good correlation between ergosterol content, the degree of infest ation assessed by visual inspection, insp ection under a stereo microscope, and the measurement of -N -acetylhexosaminidase enzyme activity. A correlation between -N -acetylhexosaminidase enzyme activity and ergosterol measurements was reported by Nielsen and Madson, with r 2 = 0.75 (n = 54). Both methods were able to detect mold growth on samples before gr oss visible signs of colonization were observed ( 2000). Other studies have re ported that mold biomass density correlated well with both ergosterol content ( r 2 = 0.968; P<0.001) and-N-
21 acetylhexosaminidase enzyme activity ( r 2 = 0.968; P<0.001) (Miller, et al 1998; Reeslev et al., 2003). A separate study compared -Nacetylhexosaminidase enzyme activity with microscopic analysis results on both new a nd contaminated building materials. The ability of enzyme activity to detect fungal growth was compared with results of microscopic analysis using paired samples. The enzyme activity method was estimated to have a sensitivity of ~89%, a specific ity of ~100%, a positive predictive value of ~100% and a negative predictive value of ~95% (Krause et al., 2003). As a quantitative measure of fungal biomass, enzyme activity demonstrated a similar capability to ergosterol measurement. For specific strains of Stachybotrys chartarum (IBT 9695) and Aspergillus versicolor (IBT 16000) conversion factors were determined from experiments performed on agar medium. The study found a linear correlation between -Nacetylhexosaminidase activity and the actual biomass density measured by weighing the mold growing on agar plates covered with cellophane. The conversion factors (CFs) were used to estim ate the biomass density of molds grown on gypsum wallboard. The biomass densities estimated from ergosterol content and -Nacetylhexosaminidase activity data gave similar results. Not surprisingly mold growth on gypsum wallboard showed significantly slow er growth and lower stationary phase biomass density than on agar. The CF for Stachybotrys chartarum (IBT 9695) was reported as 8,275 fluorescence units per mg of biomass per cm 2 The CF for Aspergillus versicolor (IBT 16000) was reported as 12,370 fluorescence units per mg of biomass per cm 2 Because the conversion factors were de termined from colonies grown on agar medium, the biomass may not be representativ e of fungi grown on other substrates such
22 as gypsum wallboard or fiberglass ductboard, limiting the accuracy of biomass density estimates. Despite the recognized limitati on in very accurately estimating biomass density, enzyme measurements yielded the same information on biomass as ergosterol measurements over the time course of funga l growth in a study of the two species described above (Reeslev et al., 2003). The existence of enzyme activity to biomass density conversion factors for Stachybotrys chartarum (IBT 9695) and Aspergillus versicolor (IBT 16000) was one of the factors for choosing these two fungi. By using isolates from these two species the conversion factors to estimate fungal bioma ss would have greater applicability. The other factor for choosing these two fungal spec ies was that they have both been described in reports as growing on water damaged buildin g materials and as po ssible contributors to adverse health effects (Ezeonu et al., 1994; Flappan et al., 1999;Gao et al., 2002; Hodgson et al., 1998; Jo hanning et al., 1996). Finding Hidden Mold Growth Detecting the presence of hidden mold growth located within wall cavities, plumbing chases, attics, crawlspaces, and HVAC ducts is a daunting challenge for researchers, inspectors and building occupants. Locating hidden mold growth is critical to a building investigation because fungal spores and microbial volatile organic compounds (MVOCs) migrate from interstitia l cavities and HVAC ducts into the occupied areas where exposure can occur (Macher, 1999). Air samples taken from occupied areas may sometimes detect hidden ar eas of mold growth, but the sensitivity of
23 such methods for detecting hidden mold growth is not reported. Even if air samples do indicate a hidden source is present, dete rmining the exact location is beyond the capabilities of such methods. Spore trap methods modified to sample fr om wall cavities have been used, but the interpretation of sample results is not we ll defined. Because internal insulation may interfere with spore collection and constr uction debris can obscure the sample, the method has many drawbacks. A search of the peer-reviewed literatur e did not reveal any studies that evaluated the ability of spore trap methods to de tect mold growth within wall cavities. Without opening wall cavities, vi sually inspecting the area, and sampling surfaces for sources of mold growth, none of the methods currently available have been demonstrated to predictably find mold growth within the interstitial spaces of a building. Use of Carbon Dioxide to Detect Metabolic Activity of Mold Growth Fungi are known to generate a variet y of gaseous by-products commonly known as microbial volatile organic compounds (MVOC s). Along with aerobic bacteria, fungi metabolize carbon-based nutrient sources and oxygen for energy, releasing carbon dioxide (CO 2 ), water and a variety of other compounds (Deacon, 1984; EPA, 1992; Foster, 1949). Much emphasis has been placed on measuring signature MVOCs produced by mold growth on build ing materials. While many researchers have identified several compounds typically produced by mold growth, none have been shown to be unique or predictive of species. Most of the compounds tentatively identifies as indicator MVOCs are also emitted from bu ilding products without mold growth (Gao
24 et al., 2002). No consensus has been reached on relevant MVOCs and interpretation of sample results are subjective. The role of MVOCs as possible respiratory irritants is also not yet clear (Pasanen et al., 1998). On the other hand carbon di oxide is produced by all fungi when metabolically active. Production of CO 2 is often used in laboratory expe riments as a direct indicator of metabolic activity. The absence, or reduction, of CO 2 is also used to demonstrate the lack of metabolic activity i ndicative of growth (Foster, 1949; Korpi et al., 1997). By drawing air from inside a wall cavity through a small diameter probe, CO 2 concentrations can be measured inside a wall cavity. Research has demonstrated the usefulne ss of carbon dioxide monitoring for grain silos as an early detection system for spoila ge from mold growth (Bhat et al., 2003). While not a specific indicator of mold gr owth, carbon dioxide can be used as a nonspecific indicator of microbial metabolism, n ecessary for growth. To be useful as an indicator of fungal growth within building cavities an estimate of carbon dioxide production rates from fungal growth is necessary. Carbon dioxide has been used as an indicator of ventilation rates in buildings when CO 2 in the building is at equilibrium. This indicator is based on assumptions that include a constant CO 2 production rate from human o ccupants, constant occupancy, constant air exchange rate, and no other sign ificant sources of ca rbon dioxide production. In residential and commercial buildings, sources of CO 2 besides building occupants,
25 include combustion sources, dry-ice use, carbon ated beverage stations and animals. Carbon dioxide emission rates from build ing materials such as gypsum wallboard, lumber, fiberglass insulation or commonly us ed finishing materials have not been reported in the published liter ature. One study of fungal growth on gypsum wallboard designed to determine MVOCs did re port that without supply air CO 2 concentrations could reach 10% (100,000 ppm) (Gao et al., 2002). Because building products are not a source of CO 2 it can be used as an indicator of building ventilation (ASHRAE, 1990; NIST, 1994). Studies of Carbon Dioxide Pr oduction from Mold Growth Korpi reported that carbon dioxide from fungal growth in house dust revealed a rapid increase in CO 2 levels after the third day at 96%-98% relative humidity (RH). Carbon dioxide concentration in a sealed container reached 11% after 25 days. The mass of mold present in the house dust was not determined, but the microbial concentration was estimated by serial dilution plating of the dust onto agar and colony formation. Unlike previous reported studies, the carbon di oxide production rate did not correlate with MVOC production in samples tested under different conditions. Samples from fungi in house dust at 96-98% RH emitted lower MVOC levels than samples measured under 84-86% RH. These results we re inversely proportional to CO 2 generation rates. This finding was attributed to a lower ad sorption of MVOCs by th e Tenax TA sampling media at the higher RH. Reduced collection e fficiency was hypothesized to be the reason for these paradoxical results (Korpi et al., 1997).
26 Carbon dioxide production rates from house dust at 96-98% RH were reported to initially increase, after bei ng placed in an environment capable of supporting spore germination and colony growth, but declined over time. Seven days after introduction to the chamber carbon dioxide production was ~560 g CO 2 per gram of dust per hr; at 14 days 480 g CO 2 / per gram of dust per hr; and at 21 days ~420 g CO 2 per gram of dust per hr (Korpi et al., 1997). Carbon dioxide generation rates from funga l growth on grains have been reported for different species. A positive correlat ion was found between volatile metabolites and fungal biomass, as measured by ergosterol content, but no correlation was found between colony forming units (CFU) and volatile metabolites. The correlation between CO 2 and ergosterol was 0.65 (n = 72, P < 0.05). However, for each species of fungi the correlation was generally higher. Little difference was detected in carbon dioxide from growth on wheat versus oats (Borjesson et al., 1992). These studies, while not directly appl icable to fungal growth on building materials, support the conclusi on that carbon dioxide is prod uced by fungi during growth and is positively correlated with fungal bi omass at comparative stages of growth. Variables that influence CO 2 production included moisture availability (i.e. relative humidity), temperature, fungal species, and the nutrient composition of the substrate (Borjesson et al., 1992; Korp i et al., 1997; Vice, 2000).
27 Mold Growth on Gypsum Wallboard Modern buildings are often constructed us ing synthetic materials not previously used. Gypsum drywall was invented around 1917, but did not become a predominant building material until after World War II (US Gypsum, 2005). Gypsum drywall or wallboard essentially replaced plaster on lath e for interior wall surfacing. The calcium sulfate gypsum core, along with a variety of binders and st rengthening materials, is sandwiched between two cellulose coatings. Gypsum drywall is very susceptible to mold growth once it is made wet by water damage or long-term exposure to high humidity. US Gypsum, a major producer of gypsum wall board, has begun adding the antimicrobial agent sodium pyrithione to inhibit mold growth in a new product called Humitek TM (US Gypsum, 2003a; US Gypsum, 2003b). Construction practices, tr ansportation and building maintenance can expose gypsum drywall to wet and humid conditions th at encourage mold growth. Due to the porous nature of gypsum drywall, cleaning th e fungal growth from affected surfaces is neither practical nor effective.
Figure 1: Photographs of mold growth on water damaged gypsum wallboard The shift from plaster to gypsum drywall has resulted in buildings with a greater susceptibility to mold growth when they become wet. The extent of interstitial wall cavities that can harbor hidden mold growth encompasses the perimeter and demising walls of most homes, schools and office buildings. There are currently no effective methods to ascertain if a wall cavity is harboring mold growth or is free of mold growth. Gypsum drywall is often part of the exterior building envelope and half of its surfaces are hidden from view. These interstitial wall cavities are susceptible to water damage and mold growth if a leak occurs. Since the interior spaces are difficult to access for visual inspection, extensive mold growth can occur before any indicators are observed 28
29 in the building. If a water damage incident occurs, such as a plumbing leak or flood, these interstitial cavities can be difficult to dry, creating hidden harbors for mold growth. Mold Growth on Fiberglass Ductboard Air conditioning ducts were initially constr ucted of metal inte rior surfaces with external insulation. Internal ly insulating HVAC ducts with fiberglass insulation was begun to remedy two problems; (1) it prevents sweating of ducts due to cold exterior surfaces; and (2) it reduces noise transmission from turbulent air flow and fan operation. Predictably, the internal fiberglass surfaces also capture dust and absorb water carried over from the air handling unit. Fiberglass insulation was eventually made more rigid and coated with a thin foil exterior sheat hing, creating ductboard. Ductboard is now produced by several manufacturers and is used extensively for construction of HVAC ducts in residential, commercial and school build ings. The material is faster to install, requiring only a razor knife and foil tape, wh ere metal ducts require metal shears, saws, and screws. Air conditioning ducts construc ted of fiberglass ductboard create pathways for air that can serve as a home to mold growt h, but are inaccessible for inspection. HVAC systems with leaky ducts can overwhelm their ability to control moisture, resulting in conditions favorable to mold growth (Cum mings et al., 1996). Dust and debris introduced during construction and system opera tion ensure that adequate inoculum and nutrients are present to support spore germination and fungal growth. Growth of fungal colonies on fiberglass ductboard and other forms of internal fiberglass insulation are
common (Ahearn et al., 1992; Chang et al., 1995; Morey and Williams, 1990; 1991; Price et al., 1994; Samimi and Ross, 2003; Van Loo et al., 2004). Metal duct surfaces can resist fungal growth even under conditions that allow fungal growth on fiberglass liner (Chang et al., 1996). The fiberglass liner not only captures nutrient dust and water, the binders present in the fiberglass can serve as a nutrient source once growth is established. The addition of antifungal treatments to fiberglass duct materials in recent years highlights the susceptibility of this material to fungal growth, but studies have demonstrated their limited effectiveness (Samimi and Ross, 2003). Figure 2: Photographs of Mold Growth on Fiberglass Duct Liner Adoption of fiberglass ductboard has been less widespread than the shift from plaster on lathe to gypsum wallboard, but is common in residential construction. The use of fiberglass ductboard in schools and office buildings is less than seen in residential buildings. However, internal fiberglass liner in metal ducts is commonly used in 30
31 commercial buildings, schools and hospita ls, creating a similar condition. The widespread use of this material has resu lted in many buildings with an HVAC duct system prone to mold growth. Here the mo ld spores are undisturb ed by cleaning efforts and nutrient-laden dust is constantly bei ng deposited. Moisture can be supplied by carryover from improperly operating or installed air handling units. The internal surface area of fiberglass ducts serving a 325 m 2 (3,600 ft 2 ) home, delivering 1,500 cubic feet per minute (CFM) of conditioned air could exceed 58 m 2 (624 ft 2 ). In extreme cases mold growth can be present thr oughout such a duct system. A standardized example duct system, fabricated of fiberglass ductboard, was used to illustrate the amount of internal surface area potentially availa ble for fungal growth in a residence (ACCA, 1995). This example system was used for indoor contaminan t modeling in this research because it is an ideal system designed to all industry sta ndards and depicts a standard system that could be used in a typical residence. These two major changes in building materials have increased the susceptibility of modern buildings to microbial contamina tion by introducing thousands of square feet of materials that readily support mold growth when water damage occurs. Mechanisms of Fungal Growth and Decomposition of Materials Gypsum Wallboard The cellulose coating of manufactured duc tboard provides an ideal substrate for most fungi. Fungi that can break down cellulose, by producing th e enzyme cellulase, readily colonize and decompose water damaged gypsum wallboard. The gypsum core
32 serves as an effective sponge, making the mo isture it stores, read ily available to fungi colonizing its surface. Growth appears to in itiate on either the front face or rear backing, but can eventually spread to the gypsum core (Doll, 2002). Colonization of water damaged gypsum wa llboard by mesophilic fungi such as Stachybotrys, Chaetomium, Fusarium and Aspergillus versicolor is well documented (Ezeonu et al., 1994; Flappa n et al., 1999; Gao et al., 2002; Hodgson et al., 1998; Johanning et al., 1996). These fungi typically colonize after water intrusion, floods or prolonged water leaks. Fungi that require less water can grow on gypsum wallboard when it is exposed to high relative humidity (>85% RH) for prolonged periods of time (Baughman & Arens, 1996). Studies documenting the speci fic changes that occur in the various components that make up gypsum wallboard have not been published, but it has been observed that the structural integrity can be lost, even if the material is dried completely. How the growth and decomposition by fungi affects the fire rating and thermal insulation properties are not well descri bed by product manufactures. Fiberglass Ductboard Molds produce and excrete acids during digestion of their nu trient substrate (Deacon, 1984; Foster, 1949; Jentschke et al ., 2001; Roos and Luckner, 1984). The acidification of the substrate can help to break down organic and inorganic materials, including cellulose, adhesives, binders and even steel. Al ong with production of organic
33 and inorganic acids such as formic acid, citric acid and acetic acid, some fungi excrete protons (H + ions) as part of a plasma membrane transport system. In the presence of ammonium (NH 4 + ) extrusion of protons (H + ) is coupled with its cellular uptake. By exchanging H + for NH 4 + the fungi maintains internal electro-neutrality, but causes high concentrations of H + in the substrate, effectively lowering the pH. Cultures of Penicillium cyclopium were found to reduce the pH of th e substrate to below 2.0 before exceeding their capacity to maintain intern al pH (Deacon, 1984; Little and Staehle, 2001 Roos and Luckner, 1984). Numerous other fungi have been found to exhibit this linkage between NH 4 + uptake and H + excretion, resulting in the acidification of soil and other growth substrates (Amrane et al ., 1999; Jentschke et al., 2001). Fungal colonization of fiberglass has been reported in numerous case reports and comprehensively described in many contro lled laboratory studies (Ahearn, 1992; Chang, 1996; Morey and Williams, 1990; 1991; Price et al., 1994; Samimi, 2003; Van Loo et al., 2004). Fiberglass has been reported to cont ain fibrous glass, phenol, cured reaction products of hexamethylenetetramine and form aldehyde (HCHO) which is an established inhibitor of many microorganisms (Chang et al ., 1995). However certain species of fungi have demonstrated an ability to utiliz e formaldehyde as a sole carbon source at concentrations up to 100 ppm. The breakdow n product of urea, also found in fiberglass insulation, is ammonia. Both Aspergillus fumigatus and Aspergillus versicolor were found to use urea as a nitrogen source (Van Loo et al., 2004).
34 There are inherent characteristics of fi berglass that make fungal colonization possible. Fiberglass is hydroscopic and tends to absorb both water and nutrients. Thus, fiberglass insulation within HVAC systems can retain sufficient moisture to support spore germination and fungal growth. The ur ea-based resins commonly used to bind the glass fibers, may be broken down by extra-ce llular acids to ammoni a, a useful nutrient source for many fungi (Roos and Lu ckner, 1984; Van Loo et al., 2004). A recent study was performed on the resistan ce of fiberglass ductboard to mold growth that contained a biocide. Results indicated that when the material surface was coated with a nutrient source, such as cons truction dust, and exposed to high humidity growth was inhibited. However once water ac cumulated on the material no inhibition was detectable (Samimi and Ross, 2003). In a 1992 U.S. EPA document on Indoor Biological Pollutants, a brief mention was made of the ability of fungi to metaboli ze toxic solids into a gaseous state. The authors referred to the convers ion of arsenic to trimethylar sine, causing arsenic poisoning (EPA, 1992). Still no considera tion was made of the effect that fungal growth would have on the flame retardant applied to building materials. In a 1995 US EPA research and development report on HVAC Systems as Emission Sources Affecting Indoor Air Qual ity: A Critical Review, it was recognized that microbial agents metabolize metals, and can produce metal-containing gasses and aerosols. The report, prepared for the Office of Environmental Engineering and
35 Technology Demonstration, also re cognized that the area of bi omethylation of metals was poorly studied. However, the report di d not consider the ef fect of microbial degradation of antimony-based flame retardants (US EPA, 1 995). This is not surprising as the biomethylation of antimony under aer obic conditions was not reported in the scientific literature until 1998 (Andrewes et al., 1998). History of Scheeles Green A well-documented historical phenome non that caused numerous deaths in Europe in the late 19 th century may have modern day relevance. Arsenic poisoning due to fungal volatilization of arsenic-containing pigments during the late 1800s in Europe has been documented (EPA, 1992; Foster, 1949). During the in tervening century research has revealed that biomethylation of arsenic and other metalloids can occur by many fungi and bacteria. Despite the well rec ognized phenomena, little consideration has been made for the potential impact that toxic metalloid volatiles may have on building occupants due to fungal growth on building materials treated w ith metalloids (EPA, 1983; EPA, 1992; EPA, 1995; NAS, 2000). While arsenic was extensively used in pressure treated lumber until a recent change to copper-based preservatives, the use of antimony as a fire retardant is pervasive in many building materials, textiles and adhesives (USGS, 2004). Research into the biomethylation of me talloids can be traced back to numerous cases of arsenic poisoning in Europe in th e 1800s. In Germany cases of poisoning were ascribed to wallpaper and tapestries printe d with arsenical pigments (EPA, 1992; Foster,
36 1949). The predominant pigment used was known as Scheeles Green (copper arsenate, CuHAsO 3 ), but other arsenic containing pigments were also implicated, Schweinfurth green, Paris green, Vienna green and emeral d green. According to contemporaneous accounts, the arsenical papers were extensively used, especi ally in bedrooms, throughout Europe from the palace down to the navvy s hut. In 1897 it was finally recognized by an Italian physician Bartolomeo Gosio that when wallpaper containing Scheeles green pigment became damp and moldy, a volatile form of arsenic was released. The volatile gas, later determined to be trimethylarsi ne, was recognized by its garlic-like odor. Initially, the fungus Scopulariopsis brevicaulis was recognized as the microbial culprit of the pairing, but other species we re later found to be able to biomethylate arsenic (Foster, 1949). Biomethylation of Metalloids (As, Sb, and Bi) Further research into the mechanism of microbial biomethylation of arsenic and other metalloids by fungi was carried out by Frederick Challenger in the 1930s at the Leeds School in England. He proposed a mechanism for trimethylarsine formation by aerobic fungal metabolism. Further research by others has expanded this mechanism to other elements (i.e. antimony, bismuth a nd tin) and microorganisms (Bentley & Chasteen, 2002; Lovley, 2000; Michalke et al., 2000). The reduction and subsequent methylation of metals and metalloids is beli eved to essentially follow the same pathway described by Challenger.
Figure 3: Periodic table section showing metalloids IIIB IVB VB VIB VIIB B C N O F Al Si P S Cl Ga Ge As Se Br In Sn Sb Te I Tl Pb Bi Po At Metals Metalloids Non-metals Liquid at 20 o C 37
Figure 4 : Typical reactions of the Challenger mechanism 38 The top line indicates mechanism for the reduction of As (V) to As (III). Structures are as follows: R 1 = R 2 = OH, arsenate; R 1 = CH 3 R 2 = OH, methylarsonate; R 1 = R 2 = CH 3 dimethylarsinate. For reduction of trimethylarsine oxide to trimethylarsine, the process is a little different. Following proton addition, the structure H-O-As + (CH 3 ) 3 reacts with hydride ion leading to elimination of H 2 O. The bottom line indicates the methylation of an As (III) structure with S-adenosylmethionine (SAM) [shown in abbreviated form as CH 3 -S + -(C) 2 ]. A proton is released and SAM is converted to S-adenosylhomocysteine [abbreviated form, S-(C) 2 ] (Bentley & Chasteen, 2002). H H 2 O R 1 H As + H + O O H: R 2 As O R 1 As HO R 1 R 2 CH 3 S C C + H + As + O R 1 CH 3 R 2 + S C C H H 2 O H H: R 1 As + H + O O As HO R 1 R 2 R 2 C As O R 2 CH 3 S C + H + As + O R 1 CH 3 C R 2 + S C
Figure 5: Challenger mechanism for the conversion of arsenate to trimethylarsine 39 As + OH CH 3 As + HO OH CH 3 As + CH 3 CH 3 CH 3 As + HO OH OH HO As + O OH OH OH As + O OH CH 3 OH As + O OH CH 3 CH 3 As + O CH 3 CH 3 CH 3 Arsenate to arsenite to methylarsonate to methylarsonite to dimethylarsinate to dimethylarsinite to trimethylarsine oxide to trimethylarsine. The top line of structures shows the As (V) intermediates. The vertical arrows indicate reduction reactions to the As (III) intermediates (bottom line), and the diagonal arrows indicate the methylation steps by SAM (see Fig 2 for details of the reduction and methylation processes) (Bentley & Chasteen, 2002). Challenger observed that Aspergillus versicolor formed trimethylarsine to a limited extent in 1948. In 1984 Cullen et al. reported the yeast Candida humiculus converted chromated copper arsenate to trimethylarsine. This report has relevance in that chromated copper arsenate, used extensively as a preservative in pressure treated lumber, is very similar to the wallpaper pigments that prompted Gosios work. Additionally, the
40 wood rot fungus P. schweinitzi has been shown to produce trimethylarsine and trimethylantimony when grown in the pr esence of arsine and antimony (Bentley & Chasteen, 2002). Prior to the mid-1990s there was little evidence for microbial methylation of antimony. Challenger described a case of chro nic antimony poisoning in Vienna, Austria at the beginning of the 20 th century. Despite his susp icion that a volatile antimony compound was being released from silk curtains used to fix colors to the fabric with an antimony compound, he could not identify any using the limited analytical methods available at the time. Finally biomethylation of antimony by the bacteria P. fluorescens was confirmed in 1996 at Sam Houston State University in Huntsville, Texas (Grleyk et al., 1997). In 1998 two studies reported the biomethylation of antimony by the fungus Scopulariopsis brevicaulis (Andrewes et al., 1998; Jenkins et al., 1998a). The environmental fate of this volat ile antimony compound (trimethylantimony, trimethylstibine, or Sb(CH 3 ) 3 ) in the atmosphere has been reported. The oxidative products of trimethylstibine were described as a range of cyclic and linear oligomers containing stibine oxide units. Unders tanding the atmospheric chemistry of trimethylstibine in the indoor and outdoor environments is im portant in understanding its mobilization and potential for human exposure (Bentley & Chasteen, 2002). The most difficult aspect of detecting trimethylstibine is its fast oxidation in gas phase. Studies have reported the gas phase rate constant s for the oxidation of trimethylstibine and trimethylarsine to be 10 3 and 10 -6 M -1 s -1 respectively (Parris and Brinckman, 1976).
41 According to estimates by Jenkins et al. the half-life of trimethylstibine in air with oxygen is around 50 milliseconds (Je nkins, 1998; Parris & Brinkman, 1976).Trimethylstibine is also much less volat ile compared to trimethylarsine. The vapor pressure of trimethylstibine is 103 torr and trimethylarsine is 322 torr at 298 K. This means that trimethylstibine would have a highe r tendency than trimethylarsine to remain in solution if they were to exist in the sa me solution. These two facts point to the major downfall of previous experiment s in which the evolved gases were aspirated with sterile air into a solution to give a precipitate for further analysis. With such a fast oxidation rate, it is very likely that trimethylstibin e is oxidized to trimethylstibine oxide [(CH 3 ) 3 SbO], before it reached the sample colle ction media. Attempts to reduce the atmospheric oxidation of trimethylstibine have included decreasing the amount of O 2 in the aspiration air to 8% by adding N 2 However, the low oxyge n content obstructs the growth of the fungi, inhibiting the metabol ic mechanism causing the release of the antimony (Grleyk, 1997). Until the introduction of the flame retard ant, antimony trioxide after World War II, the potential for widespread public e xposure to antimony was low. Antimony-based flame retardants are now commonly used for indoor building materials, textiles, and furnishings (ATSDR, 1992; USGS 2004). However, the US EPA currently does not believe the use of these flame retardants in plastics or textiles will result in significant exposure to consumers. The belief stated in the review was not based on any cited research, but rather a cursory understanding of the physical properties of antimony trioxide, namely its low volatility, low water solubility and assumption that the flame
42 retardants are tightly bound to the matrix (EPA, 1983). Howe ver, due to the findings of preliminary toxicity studies that chronic exposure to anti mony compounds resulted in damage to the heart, kidneys, liver and lungs, the EPA recommende d further studies to evaluate chronic toxicity be conducted (EPA, 1983). Despite recognition of the biotransformation that can occur to antimony in the natural environment, no consideration was made of this possi ble mechanism occurring within indoor environments (EPA, 1983). Prevalence of Antimony in Building Products Antimony trioxide is a flame retardant fr equently used in building materials common to homes, commercial buildings and schools. Approximately 60% to 65% of all antimony used in the United States is in the form of antimony trioxide (Sb 2 O 3 ) as a flame retardant (USGS Mineral Comm odity Profiles: Antimony 2004). Antimony trioxide is also used in a wide variety of plastics which include flexib le polyvinyl chloride (PVC), polyolefins, polystyrene, polye thylene terephthalate (PET), acrylonitrile-butadienestyrene (ABS), and polyurethanes. Antimony trioxide is also used as a stabiliz er in plastics, as a pigment in enamels, paints, and rubber, as an antisolarant, decolori zer and a fining agent in glass. It is added, along with halogenated hydrocarbons, as a fl ame retardant synerg ist in adhesives, plastics, rubber and te xtiles (USGS, 2004).
43 Use of antimony trioxide in the U.S. has grown since the 1960s. In 1999 nearly 23,000 metric tons of antimony-based flame re tardants were used, accounting for 57% of primary antimony production (USGS, 2004). Antimony trioxide flame retardants are specified for use in adhesives & coatings, furniture, insulation, mattresses, roofing materials, textiles, wall and floor coverings, circuit boards, electrica l connectors, relays and switches. Applications are also mark eted and described for consumer products, including appliance housings, battery casings business machines, consumer electronics, TV housings and all types of wire and cable sheathing (Great Lakes Chemical Corporation, 2003). Antimony trioxides use as a pigment has been surpassed by titanium dioxide in the United States, with the exception of some exterior oil-based paints and enamels. The main advantage of antimony trioxide as a paint pigment is its resistance to chalking and UV degradation. These propertie s have lead to its use in yellow paints for school busses and yellow striping applied to road pavements. Pigments made from antimony trisulfide and antimony pentasulfide are used for colo ring rubber black and shades of yellow, orange and red. However interi or paint products containing Timonox Flame Retardant were available in the United Kingdom as recently as 2000 (MSDS,1999). Based on analysis of 400 paint samples, house paints wi th >2% antimony were reported to be rare, constituting less than 0.5% of old paint films (van Alphen, 1998). Other uses for antimony compounds include:
44 Fluid lubricants for increased chemical stability. A phosphor in fluorescent lamps. Antimony pentasulfide is used as a vulcanization agent for red rubber. Antimony trisulfide is used in primers for ammunition, high explosives and in fireworks. Tartar emetic (hydrated potassium antimonyl tartar) has been used as a pesticide, for stomach disorders, to treat the parasitic disease leishmaniasis, and a mordant for some acid textile dyes. Consumption of primary antimony in the Unite d States has grown at a rate of ~5.6% per year since the early 1980s. The two driv ing uses for primary antimony appear to be as a flame retardant in plastics a nd as a catalyst for PET (USGS, 2004). After recognizing the many types of produc ts that contain antimony in the indoor environment a search was made for buildi ng materials, including HVAC related products, coatings and paints that are specified for i ndoor use. Table 1 shows products that were found to contain antimony oxides. Initial estimations of an timony trioxide content were based on the Material Safety Data Sheets (MSDS) provided by product manufacturers. However, the amount (i.e. mass) of antimony trioxide could not be determined from product literature because product density a nd coverage rates were not specified. Experimentally derived data was necessary to determine the potential contribution that various products could have to the amount of antimony in the indoor environment.
45 The materials initially evaluated for antim ony content were chosen after a search of material safety data sheets (MSDS) identif ied a variety of materials and coatings that reportedly contain antimony. Most products did not specifically list the concentration of antimony, typically indicating that less than a certain amount by weight is present in the product. In order to asse ss the tentatively id entified materials for antimony content those with the highest reported amount s were purchased and tested. Reviewing the material safety data sheets and product literature revealed several brands of fiberglass ductboard that contai n up to 3% antimony tr ioxide by weight. Because ductboard is coated with antimony tri oxide in a factory setting, with presumably small tolerances, the variability of this manufactured material was expected to be low. The same may not be true with paintings a nd coatings applied to walls during or after construction. To determine the range of antim ony levels on materials after painting with a flame retardant coating experimental data w ould be necessary. Fiberglass ductboard was chosen as the test material to evaluate th e potential for mold gr owth to release antimony from building materials. Table 1: Products containing an timony trioxide fire retardant* Product Type Trade Name Manufacturer Antimony trioxide content Adhesive Pyralux DuPont Not Disclosed Paint Pigment Sicotan BASF 5% 17% PET Resins: Polyester Glass Reinforced Resin 115FR BK112 BASF 1% 5% by wt Interior Wall & Ceiling Paint Timonox Paint Akzo Nobel [2.5% 10%] Paint, Yellow Corn color A-100 Exterior Satin Latex Sherwin-Williams 1% by wt
46 Table 1: Products containing antimony trioxide fire retardant* (continued) Product Type Trade Name Manufacturer Antimony trioxide content Fire Retardant Varnish Ultra-hide Insul-blaze coating Glidden Paint, ICI Paint 1% 5% by wt Aerosol Lubricant -G75, 27A Aerosol Sandstrom Products 1% 5% by wt Fiberglass Insulation CM-26 (coated) OEM Diffuser board, TufSkin Rx Johns Manville >0.1% by wt in tape, facing, adhesives and/or coating All Service Jacketing Roofing Insulation Product ASJ 3035; ASJ 4535 Johns Manville 1.6% 4.0% Kraft paper backing laminated w/ flame retardant adhesive FSK-25 Johns Manville <2% Flame retardant adhesive LAWX-235D Johns Manville 7% 12% Fiberglass thermal insulation w/ Kraft paper backing AcoustaTherm Batts CertainTeed 3% by wt Fiberglass Ductboard ToughGard CertainTeed 0.9% to 3.0% by wt *Information compiled from MSDS provided by product manufacturers. Products with less than 1% Antimony by weight do not have to list Antimony on the MSDS. Therefore this listing cannot be considered comprehensive. Health Effects of Exposure to Antimony Compounds Little is known about the toxicity of antimony in comparison with knowledge about other metals such as lead, cadmium, mercury, or arseni c. The chemical form of antimony contributes greatly to its toxicity and a genera l Antimony Toxicity is not known to exist. Each form of the metal must be considered as a separate toxicological entity. The mechanism of toxicity for an timony compounds is unclear; but is probably related to antimonys high affinity for sulfhydryl (-SH) groups, which are essential for the structure and function of proteins. Trivalent antimony concentrates in the red blood cells, while the pentavalent form is found in the plasma. Both forms are excret ed in the urine and feces, but urine contains more of the Sb +3 and feces contain more of the Sb +5 Reported reference values of
47 antimony in the blood or serum range from 2.5 28.7 nmol/L (0.3 3.5 g/L), but less than 80 nmol/L is also reported as a norm al value. Reported reference values of antimony in the urine range from 1.7 18 nmol/L (0.2 2.2 g/L), but less than 160 nmol/L is also reported as normal (Bal dwin and Marshal, 1999; medtox.org, 2003). Organic trivalent antimony has a greater affinity for the liver and red blood cells and is excreted more slowly (Beliles, 1994). In general, trivalent an timony compounds exhibit 10 times the toxicity of pentavalent fo rms. Prolonged exposure to many antimonycontaining compounds is known to cause respir atory irritation (Krachler et al., 2001). Human poisoning due to chronic exposure to stibine or trimethylstibine has not been reported, although numerous cases of acute toxicity ha ve been described (Beliles, 1994; Parish et al., 1979). Prior to its us e as a flame retardant, human exposure to antimony was believed to be limited to foundr ies and manufacturing operations where the gas trimethylstibine is applied as a dopant to semiconduc tors (Trimethylstibine, 1991). From what is known about trimethylstibine, ch ronic exposure is unlikely due to its rapid oxidation rate to less volatile forms of antimony oxides (Paris & Brinkman, 1976). Very little information is available on health effects from acute or chronic exposure to trimethylstibine. Much of th e toxicity and health endpoints must be extrapolated from data available from studies on antimony (Sb), antimony trioxide (Sb 2 O 3 ) and stibine (SbH 3 ). Exposure to the methylated form of stibine, trimethylstibine (Sb(CH 3 ) 3 ) by inhalation may cause bleeding gums, metallic taste, nausea, laryngitis, anemia, skin eruptions, liver and kidney damage (MSDS Trimethylstibine, 2003).
48 Stibine is reported to equal or surpass arsine in toxicity, and ca uses specific toxic effects that closely resemble those of arsine. Based on a 1903 German publication On Stibine and Yellow Antimony one of the toxic mechanisms of stibine is reaction with hemoglobin in red blood cells, leading to th eir destruction. Antimony is reported to increase the activity of heme oxygenase (Beliles, 1994; Thayer 1984; Thayer, 1995). Also described are direct effects on brain tissue cells leading to various degrees of degeneration. Workers exposed to stibine also demons trated marked weakness, headache, nausea, severe abdominal and lower back pain, and blood in the urine. The authors of the article state th at further research on the chronic and acute effects of exposure to arsine and stibine is needed (Blackwell & Robbins, 1979). Acute exposure to stibine is reported to cause hemoly tic anemia and acute renal failure (DeWolff, 1995). Pneumoconiosis, l ung irritation and othe r respiratory effects have been reported in workers (ACGIH, 1986; Beliles, 1994; DeWolff, 1995). One of the early signs of overexposure to stibin e in humans may be hemoglobinuria. It remains unclear as to the mechan ism that is responsible for antimonys genotoxicity (Krachler et al., 2001). A r ecent report on plasmid DNA damage, caused by stibine and trimethylstibine, points to a possible mechanis m for the carcinogenicity of antimony. Trimethylstibine and stibine were fo und to be equipotent with trimethylarsine when evaluated using a plasmid DNA-nick ing assay. Growing evidence suggests that DNA damage by arsenic and antimony is due to the production of reactive oxygen
49 species (ROS) and resulting oxidative stress. However, th is study was only able to demonstrate DNA damage at a concentra tion 400 times the OS HA PEL of 0.5 mg/m 3 (Andrewes et al., 2004). The decisions made by the US EPA and the CPSC to allow the use of antimony trioxide on indoor building products never c onsidered the possibility of material decomposition due to fungal growth and the subsequent biomethylation and release of trimethylstibine (EPA, 1983; NAS, 2000). The U.S. EPA listed antimony as a prior ity pollutant in 1986. In 2000 the U.S. Consumer Product Safety Commission (CPSC) released a report on the Toxicological Risks of Selected Flame Retardants. This risk assessment was performed because the CPSC was considering promulgating standard s that would require most residential upholstery fabric to be treated with flame retardant chemicals. Because many flame retardant chemicals, including antimony trioxide, exhibit toxic properties an assessment was made of the risk to building occupants. Despite no consideration for the potential of microbial biomethylation and mobilization of antimony in a gaseous form, the report did conclude that worst-case exposure to pa rticle phase degrada tion products via the inhalation route could pose a non-cancer risk of respirat ory irritation, quantified by a Hazard Index of 1.2. Additionally, the repor t concluded the unit risk of lung cancer (cancer potency factor) is 7.1 x 10 -4 / g antimony trioxide/m 3 This translated into a lifetime excess cancer risk estimate of 1.7 x 10 -4 due to exposure to antimony trioxide
50 particles, or 1.7 additional lung cancers per 10,000 people exposed for 70 years (NAS, 2000). Based upon the findings of both an inhalati on hazard index of greater than one and a potential cancer risk, the committee re commended that the potential for particle release from treated fabrics be investigated (NAS, 2000). Trends in Antimony Use and Asthma Prevalence One way of examining associations is to compare historical indicators of exposure with disease prevalence. Asthma prevalence data is available for a 15 year period from 1980 to 1995 (CDC, 1998). Because the primar y use of antimony trioxide is fire retardants, and most of it has been impor ted to the United States, imports of Sb 2 O 3 can serve as a useful surrogate indicator of the populations exposure to th is product additive. Antimony oxide import data is available from 1950 through 1999. Examining overlapping data (1980 1995) for antimony oxide imports and asthma prevalence revealed a positive correlation (r 2 = 0.9343; P<0.01) (CDC, 1998; USGS, 2004). Figure 6 shows the available published data for both asthma prevalence and antimony oxide imports (CDC, 1998; USGS, 2004). Estimates of antimony oxide imports were interpolated from the charted data and comp ared with published asthma prevalence rates for 6 time periods. Figure 7 shows the scatte r plot and linear regression of the data (USGS, 2004).
51 After 1995 antimony imports continued to rise, but peak in 1997. Falling levels may be a result of an overall shift in manuf acturing of textiles a nd building materials to other countries. It is likely that fire retardant usage conti nues to increase, but that much of it is applied in the overseas factorie s where the products are manufactured. The reduction of antimony trioxide imports may also be effected by manuf acturers shifting to other fire retardants. For the time period between 1980 and 1995 asthma prevalence rose 75% and antimony oxide imports rose 82%. Asthma prevalence increased 5% per year while antimony oxide imports increased 5.4% per ye ar. While this comparison and positive correlation does not offer conclusive data for a causal link, it does suggest a possible association between exposure to antimony fire retardants and an incr eased prevalence of asthma.
Figure 6: Antimony oxide imports and asthma prevalence in the U.S. Antimony (Sb) Oxide Imports vs.Asthma Prevalence in U.S.05,00010,00015,00020,00025,00030,00019501955196019651970197519801985199019952000Year Sb Oxide Imports 30.734.653.846.642.937.6-30-60-45-15-75-90 Figure 7: Antimony oxide imports vs. asthma prevalence in the U.S. 52 Scatter Plot of Sb Oxide Imports vs. Asthma Prevalencey = 0.0033x + 3.2359R2 = 0.934325303540455055608,00010,00012,00014,00016,00018,000Tons of Sb Oxides Imported to USPrev ce Rate of Asthma in US per 1,000 alen
53 Using a common public health paradigm to evaluate causal links, the BradfordHill criteria, the following items should be examined before a causal link is presumed from any statistical association. 1. Consistent and unbiased fi ndings. Relationships that are demonstrated across a number of studies, different populations, different ci rcumstances, and different study designs. 2. Strength of association. Str ong associations are less likely to be caused by bias. 3. Temporal sequence. Exposure must preced e the disease, and the latency period. 4. Biological gradient (i.e. dose -response). Changes in exposure relate to a change in relative risk, quantitative relationships be tween the risk factor (exposure) and the outcome, intensity or duration of exposure may be measured. 5. Specificity and biological plausibility. If an exposure leads to a single or characteristic effect, or affects people with a specific susceptibility. Causal mechanism proposed must not contradict what is known about natural history and biology of disease, must fit with know facts. Proposed causal mechanism should be biologically plausible. When rem oved from the offending agent the health complaints disappear or resolve 6. Experimental evidence. Controlled expe riments should be able to demonstrate onset of the disease in animal models. Most of the criteria set forth by Sir Au stin Bradford-Hill have not been fully examined, much less met. However, two compelling points encourage further examination of this possible link between asth ma and antimony trioxide fire retardant.
54 The first is an association between a surroga te indicator of exposure, namely antimony oxide imports, with the asthma prevalence in the U.S. Recognizing the multitude of indoor products and materials that are treated with antimony trioxide, should encourage researchers and federal regulat ory agencies, such as the US EPA and CPSC, to conduct a comprehensive assessment of antimony exposures in all types of indoor environments. The second point to consider is the mech anism of toxicity attributed to antimony trioxide particulate inhalation. The respiratory irritancy of these antimony compounds is well known when acute exposure occurs even at very low concentrations. For asthmatics, exposure to any respiratory irrita nts may result in exacerbation of symptoms. No studies have been identified that examin ed respiratory irritation and the development of chronic bronchial hyper-reactivity (i.e. asth ma) with chronic exposure to any form of antimony. The potential for adverse health effects from chronic exposure has been recognized for years, but not yet studied in human populations. Due to the widespread use of antimony-base d fire retardants a large portion of the U.S. population is potentially exposed to an timony on a daily basis. When microbial degradation of treated building materials occurs, there is the possibility of antimony fire retardant mobilization and the ge neration of aerosols. Based on what is known about the toxicity of other antimony compounds, chroni c exposure to moldy indoor environments may present a previously unrec ognized public health hazard.
55 FOCUS OF RESEARCH Fungal growth in buildings poses numer ous challenges to researchers, building owners and occupants. Public health agen cies promote prevention and remediation of mold and water damage, but limitations in current knowledge hamper meaningful research. The first problem is that a sensitiv e, low-cost tool that can detect hidden mold growth does not exist. Mold often begins to grow within wall cavities and spaces not accessible for inspection. By the time mold damage is visible in the occupied areas massive structural damage can take place. The second impediment to reducing public health risks associated with indoor mold growth is understanding the biological mechanisms that make occupying moldy buildings hazardous. Carbon Dioxide as a Screening Test for Hidden Mold Growth Despite over ten years of research an d limited field use, measuring microbial VOCs has not proven to be a pr actical tool for detecting hi dden mold growth. The high cost of laboratory sample analysis and the many well known inte rferences by building product emissions make interpretation of the data difficult at best. However, since the development of sensitive, hand-held, por table carbon dioxide sensors, real-time measurement of confined spaces can be performed. If fungal growth on common building materials could be determined to generate sufficient c oncentrations of CO 2 to elevate the concentration w ithin a wall cavity or plumbing chase above the background
56 concentration found in occupied spaces, it may serve as a useful screening test for microbial metabolic activity. A small probe can be inserted to draw air from the wall cavity into a monitor. Concentrations with in a wall cavity without any metabolically generated CO 2 should be the same or lower than indoor concentrations. By also measuring the pressure differential, the air fl ow direction can be determined and the air exchange rate may be estimated for the wall cavity. Monitoring of CO 2 levels in laboratory studies has been used as a test to indicate sterility of control samples (Borjesson et al., 1992). If it were found that fungi produced sufficient CO 2 to elevate interstitial concentrations, the absence of elevated CO 2 levels within wall cavities may be a useful indicator for the absence of mold gr owth, an especially useful test after water damage has occurred. Whether or not mold gr owth was truly prevented is a question that often remains after drying efforts have concluded. CO 2 Generation Rates During the Li fe Cycles of Two Mold Species In an attempt to derive carbon dioxide generation rates necessary to model CO 2 concentrations within wall cavities and static HVAC ducts two types of fungi often found in water damaged buildings were chosen. Stachybotrys chartarum was chosen for growth on gypsum wallboard and Aspergillus versicolor was chosen for growth on fiberglass ductboard. Both of these fungi ha ve been associated with occupant health complaints in water damaged buildings a nd both are capable of producing allergens, irritants, VOCs, and mycotoxins (Hodgson et al., 1998; Johanning et al., 1996). Because fungi are living, growing organisms, a singl e point measurement was not considered
57 predictive. Therefore, a series of experiments were designed to measure CO 2 production throughout their life cycles, ranging from 24 hours after inoculation to 32 weeks. Potential for Antimony Release from Sb 2 O 3 Flame Retardant as Trimethylstibine Speculation about antimony mobilization and release from environmental and indoor sources has existed since the earl y 1990s (EPA, 1992; EPA 1995; Grleyk, 1996; Jenkins et al., 1998a). Laboratory experiments published in 1997 and 1998 demonstrated the ability of certain fungi and bacteria to li berate antimony as trimet hylstibine. The first experimental evidence that antimony was converted to trimethylstibine by microorganisms was published in 1997 (Grley k et al., 1997). Ae robic biomethylation of soluble inorganic antimony by the fungus Scopulariopsis brevicaulis was first demonstrated in 1998 (Andrews et al., 1998; Je nkins et al., 1998a). Recognition of this mechanism prompted the question, could fungal growth on fire retard ant treated building materials release antimony as trimethylstibin e? In light of the widespread use of antimony-base fire retardants and the myriad of adverse health effects antimony exposure can cause, the impact of fungal growth on treat ed building materials should be examined. Previous studies have been careful ly controlled laboratory tests using homogeneous substrates amended with antimony compounds. These studies were intended to examine the abil ity of specific fungal species to biomethylate antimony. Their results indicated the biological mechan ism was possible and appeared to occur with many types of antimony additives, including antimony trioxide (Andrews et al., 1998; Jenkins et al., 1998a). Based on the positiv e results from earlier studies and the
58 speculation of the authors I chos e to focus this research on actual building materials that contained antimony trioxide. Using manufactured building materials did not allow for precise control over the amount and form of antimony added, and introduced many possible interferences due to limited knowledge of the proprietary material content. Conversely, use of building materials enables me to directly address the question of antimony mobilization from fire retardant treated building materials that have been demonstrated to support fungal growth in buildings. Measuring Antimony Release from Mold Growth on Sb 2 O 3 Flame RetardantTreated Building Materials Using a commercial laboratory, samples of various materials reported to contain antimony were analyzed for antimony content. For varnish and paint, the products were applied to oak panels and gypsum wallboard re spectively for analysis. Some pieces were wetted and inoculated with funga l spores to determine if they could grow on the material. Table 2 shows the results of ICP analysis for total antimony in test materials of various sizes. These data show the comparable amount of antimony present in each type of material tested. Differences between ne w materials and materials with mold growth were observed, but variations in application rates of an timony-containing coatings and fungal biomass density could have accounted for them. Additionally, confidence in the data was not high, as the laboratory reported their calibration consisted of a blank and a single point standard. Inconsistencies in measurement results, reported units and the lack of quality assurance information reduced my confidence in the results from the
59 commercial laboratory. Due to low confidence in these results I decided to perform all further analyses for antimony in the research laboratory with available methods described later in the methods section. Table 2: ICP analysis results from p reliminary study to identify materials containing Sb 2 O 3 fire retardant Material Description Total Sb ( g) in Sample g (Sb) cm -2 Date Fiberglass Liner Coating (no mold growth) 286 185 3/12/2004 Fiberglass Liner Coatin g (Heavy mold growth) 323 210 3/12/2004 Fiberglass Liner Coating (moderate mold growth) 281 182 3/12/2004 Fiberglass Liner Coating New 334 220 3/12/2004 Painted Gypsum Paper with mold growth 4.40 0.870 1/13/2004 Painted Gypsum Paper with mold growth 3.80 0.864 1/13/2004 Painted Gypsum Paper with mold growth 2.40 0.784 1/13/2004 Painted Gypsum Paper with NO mold growth 3.85 0.802 1/13/2004 UnPainted Gypsum Paper with mold growth <0.3 (BDL) <0.07 1/13/2004 UnPainted Gypsum Paper with mold growth 2.45 1.17 1/13/2004 Varnished Oak Panel with mold growth on 1-side 17.8 4.12 1/13/2004 Varnished Oak Panel with mold growth on 1-side 10.6 3.08 1/13/2004 Varnished Oak Panel with mold growth on 1-side 11.6 5.55 1/13/2004 Fiberglass Liner Coatin g with mold growth 402 227 1/13/2004 Fiberglass Liner Coat ing with mold growth 364 205 1/13/2004 Fiberglass Liner Coat ing with mold growth 338 191 1/13/2004 Fiberglass Liner Coating with NO mold growth 420 238 1/13/2004 Foil backing of ductboard with adhesive & mold growth 210 119 1/13/2004 Foil backing of ductboard with adhesive & mold growth 256 145 1/13/2004 Foil backing of ductboard with adhesive & mold growth 231 130 1/13/2004 Foil backing of ductboard with adhesive NO mold growth 327 185 1/13/2004 NEW Materials Varnished Oak Panel 4.55 2.28 12/29/2003 Varnished Oak Panel 5.35 2.68 12/29/2003 Varnished Oak Panel 5.25 2.62 12/29/2003 Painted gypsum paper w/ gypsum removed 0.85 0.85 12/29/2003 Painted gypsum paper w/ gypsum removed 0.85 0.85 12/29/2003 Painted gypsum paper w/ gypsum removed 2.40 0.60 12/29/2003 Fiberglass Liner Coating with NO mold growth 288 288 12/29/2003 Fiberglass Liner Coating with NO mold growth 238 238 12/29/2003 Fiberglass Liner Coating with NO mold growth 236 236 12/29/2003 Foil backing of ductboard with adhesive NO mold growth 218 218 12/29/2003
60 Initial results revealed the amount of antimony in th e varnish and paint was 1001,000 times less than the fiberglass ductboard. Mobilization of antimony from the ductboard material was assumed to be easier to detect if it occurred. Additionally, the consistency of the application rate for varnish and paint was expected to be greater than the coatings and adhesives applied to ductboard during its factory manufacturing process. The fungal species chosen to inoculat e the ductboard test materials was Aspergillus versicolor A series of experiments were designed to measure volatile antimony emissions and the residual amount of antimony remaining in the test material beginning at ~24 hours after inoculati on and continuing for 29 weeks.
61 MATERIALS AND METHODS CO 2 Generation from Fungal Growth on Gypsum Wallboard and Fiberglass Ductboard Carbon dioxide generation was studied from two types of fungi that are often found growing on water damaged building materials. Stachybotrys chartarum was inoculated onto gypsum wallboard and Aspergillus versicolor was inoculated onto fiberglass ductboard. The fiberglass ductboard was delaminated into two major components, the air-side coat ed surface, which normally faces the air stream of constructed ducts, and the foil faced skrim kraft paper (FSK) exterior wrapping with adhesive. Pieces of each material type were inoculated and the carbon dioxide concentrations measured inside an envir onmental chamber at regular periods during fungal growth to calculate the CO 2 production rate. After ca rbon dioxide measurements were taken, the fungal biomass density was es timated by collecting core samples from the materials and measuring -Nacetylhexosaminidase activit y. This endogenous enzyme correlates with mold biomass and a conversion fa ctor was reported for the strains of fungi chosen for this study (Reeslev et al., 2003).
62 By knowing the approximate surface area of inoculated building materials with mold growth, the biomass density of fungi and the carbon dioxide production rate at equilibrium in the chamber, the carbon dioxide production rate per surface area of mold and per biomass of mold were calculated using equation 1. Gypsum wallboard (GWB) test pieces were prepared from a single 4 ft x 8 ft (1.22 m x 2.44 m) sheet of materi al purchased from The Home Depot in Tampa, Florida. Individual pieces were cut from the 1/2 (1.27 cm) thick sheet, each measuring 10 cm x 10 cm. The edges were sealed with commer cial grade seam tape commonly used to finish interior walls (The Home Depot in Ta mpa, Florida). The edges were sealed to protect the interior gypsum core from dama ge and loss during handling. No paint or coating was applied to the finish or back side. Two holes were drilled through the upper edge of each piece to allow insertion of tw o 3 mm (1/8 inch) diameter brass rod. The brass rods were used to hang the test pieces in a vertical position. To help maintain a constant spacing between each test piece, a 1.27 cm ( inch) length of polyethylene tubing was placed on the brass rod between each piece of GWB test material. Five GWB test pieces were used for the experiment a nd altogether constitute d a test rig. See Figure 8.
Figure 8: Photograph of test rig for gypsum wallboard Fiberglass ductboard test pieces were also prepared from a single sheet of 4 ft x 10 ft (1.22 m x 3.05 m) sheet of material (CertainTeed ToughGard Duct Board) purchased from Jim Air Distributors Supply in Tampa, Florida. Individual pieces were cut from the 1 (3.81 cm) thick sheet, each measuring 10 cm x 10 cm. Each test piece was delaminated into two sections. The first section was the air-side surface coating with 1.27 cm ( inch) of fiberglass insulation. The second section consisted of the foil-faced skrim kraft (FSK) paper exterior wrapping with adhesive. Only 2 to 4 millimeters of fiberglass insulation was left on this material to ensure the adhesive was readily inoculated. Test pieces were placed onto brass rods in a similar manner to the gypsum wallboard pieces described earlier. 63
64 Both test organisms, Stachybotrys chartarum (IBT 9695) and Aspergillus versicolor (IBT 16000) were obtained from the cu lture collection of Biocentrum, Technical University of Denmark (CDC PHS Permit N o. 2003-09-094). The cultures were revived and grown on malt extract agar (M EA) in covered sterile Petri Dishes for 15 weeks (Fisher Scientific, Catalog #B11403). Stachybotrys chartarum cultures were thin until inoculated onto MEA supplemented with rice nutrient (Uncle Bens Instant Rice) to obtain heavy growth. Culture plates were sealed and maintained in an incubator at 25 o C. Spore suspensions of test organisms were harvested from agar plates by adding 80 ml of distilled water with 0.01 ml of Tween 80 and stirring w ith a sterile loop. The spore suspensions were then diluted to a final volume of 100 ml. Each spore suspension was then placed into a 250 ml stainless steel pr essurized spray mister and the hand pump was activated 20 times. The delivery rate of spore suspension was earlier determined to be at a rate of 1 ml/sec when filled with 100 ml of liquid and pressurized with 20 strokes. Gypsum wallboard test pieces were soaked in a bath of distilled de-ionized water for ten minutes and allowed to equilibrat e for 30 minutes. The test organism Stachybotrys chartarum (IBT 9695) was inoculated onto the entire front and back surfaces of 5 pieces of gypsum wallboard (GWB ) within a biosafety cabinet. Each piece was inoculated with 3 ml of spore susp ension on each side (30 ml total). The concentration of spores present in the suspen sion was not determined. The goal of this inoculation procedure was to en sure equal inoculation for all test pieces in an attempt to encourage homogeneous growth.
65 Five control test pieces of gypsum wallboa rd were prepared, but not inoculated with any spore suspension, or wetted with wate r. These control piec es were kept dry, but when tested for CO 2 production, a 30 ml beaker of distilled de-ionized water was set in the chamber in an attempt to account for possi ble carbon dioxide release from the water. The test organism Aspergillus versicolor (IBT 16000) was inoculated onto 5 pieces of fiberglass ductboard surface coati ng, each measuring 10 cm x 10 cm x 1.5 cm. Surface coating test pieces were placed in to a biosafety cabinet and each piece was coated with 3.45 mg/cm 2 of Malt Extract Agar (MEA) powder to serve as a nutrient source and then inoculated with 5 ml of s pore suspension onto the entire coated surface side (25 ml total), but not the back surface. Previous studies have reported fungal growth on fiberglass material can be slow or inhi bited in the absence of accumulated nutrients such as dust (Samimi & Ross, 2003). MEA powder was chosen because it should not have introduced other fungal colonies. Control test pieces of fibe rglass ductborad were prepared in the same manner as the GWB. Aspergillus versicolor (IBT 16000) was also inoculated onto the adhesive side of 5 pieces of fiberglass ductboard FSK exteri or wrapping with adhesive. No MEA powder was applied to these pieces as the adhesive was expected to provi de sufficient nutrient source.
All test and control pieces were placed onto brass rods in a similar manner as the GWB and a 1.27 cm ( inch) space left between each piece to allow adequate chamber air mixing. The assembled test rigs were self-supporting and readily fit into the test chambers and growth chamber described below. Test chambers were constructed of 5 liter stainless steel food-grade containers with clear acrylic lids (d = 19 cm, h = 18 cm) (OGCI 18/8 Stainless Steel Anaheim, CA) (See Figures 9 and 10). The lids were sealed using a silicone gasket, clamping ring closure and Teflon tape to reduce leakage around the gasket. Each test chamber was fitted with two stainless steel Swagelok fittings in the acrylic lid. The exhaust valve was fitted with a inch stainless steel tube that extended to the bottom of the chamber, enabling air to flow into the chamber near the top and exit near the bottom. This air flow pattern was intended to prevent stratification of the dense carbon dioxide gas. Figure 9: Photograph of test chamber, acrylic lid, inlet and exhaust fittings 66
67 During periods when the test materials were not being measured for emissions they were kept in growth chambers. Grow th chambers were 5 liter stainless steel containers with acrylic lids, but without the Swagelok valve fittings. The growth chambers were kept at room temperature 23-26 o C (74-79 o F) during the periods between laboratory experiments. Supply air to the chamber was filtered through a 37 mm polycar bonate filter (0.8 m pore size) as it entered and exited the chamber. Filtering the air into the chamber reduced the chance of inocul ation by other fungal spores and cross contamination. Filtering the air being exhausted out of the ch amber helped to stabilize the air flow and enabled easier control of the air flow rate using the in-line adjustment valve and rotameter. Air flow was supplied using th e laboratory vacuum which provided longterm, steady air flow. During chamber tests for carbon dioxide th e 5 liter chamber with test rig was placed in the work area of a biosafety cabinet. Loading and unloading of the chamber was performed under the hood to prevent e xposure to the fungal spores and cross contamination of other materials.
Figure 10: Diagram of test equipment 68 A typical chamber test of carbon dioxide generation proceeded as follows. Inoculated test materials were removed from a sealed growth chamber and placed into a test chamber. Carbon dioxide monitors were placed in-line with each chamber exhaust and data logging began. A single carbon dioxide monitor was used to measure and record intake air conditions (Temp., RH, and CO 2 ). Air flow rates were set to 250 ml/min using an in-line rotameter that was periodically checked against a primary calibration device (BIOS Dry-Cal Model-H). The chambers were allowed to reach equilibrium before recorded concentrations were used to calculate generation rates. The carbon dioxide monitors had internal data logging capability and concentrations were recorded every 10 minutes for the ambient air entering the chambers and for the exhaust air exiting the chambers. Once the chambers reached equilibrium, after 2 to 3 hours (6 to Legend 1 Ambient CO 2 probe 2 Air intake filter 3 Chamber (5L) 4 Test materials 5 Air exhaust filter 6 Chamber CO 2 probe 7 HgCl 2 -coated silica gel sorbent tube 8 Air flow rotameter 9 Flow adjustment valve 10 Laborator y Vacuu m 9 6 7 8 1 5 10 2 4 3 Chamber CO 2
69 9 air changes), recorded carbon di oxide concentrations were us ed to calculate the average generation rate described in equation 3. Carbon dioxide was measured using a mu lti-function, integrated monitor with a non-dispersive infrared sensor placed in-line w ith the chamber exhaust air flow (AQ5001 Pro, Quest Technologies, Oconomowoc, W I). The non-dispersive infrared (NDIR) sensor was reported to have a range of 0 to 5,000 parts per million (ppm), an accuracy of 3%, and resolution of 1 part per million (ppm). The CO 2 sensor calibration was verified prior to use with nitrogen (>99.80%) for zero adjustment and 1,000 ppm carbon dioxide for span adjustment (Air Liquide Americ a, Cambridge, MD). The monitor used a resistance temperature detection sensor for temperature measurement with a range of 0 o C to 60 o C (+32 o to + 140 o F) and an accuracy of 0.5 o C ( 0.9 o F) with a resolution of 0.06 o C (0.1 o F). The monitor used a capacitive sensor to measure relative humidity (RH). The accuracy is reported to be 3% at 25 o C and the resolution is 0.1% RH. These monitors had internal data l ogging capability and simultane ously recorded temperature, dew point temperature, relati ve humidity, and carbon dioxide. Recorded data from each monitor wa s immediately downloaded after each chamber test onto a laptop computer with Quest Suite Proprietary software supplied by the carbon dioxide monitor manufacturer (Que st Technologies Oconomowoc, WI). Data were then exported into Microsoft Excel for analysis, graphing and calculation.
70 For quality control, chamber tests we re conducted on empty chambers and on chambers with un-inoculated test materials pl us beakers of distilled, de-ionized water. These data are reported in Appendix 1. After carbon dioxide measurements were complete the test rig was removed from its test chamber and one core sample was cut fr om each test piece, tota ling 5 in all. Core samples were collected using a 1.4 cm diam eter aluminum metal tube core cutter, providing samples with an area of 1.54 cm 2 Samples were placed into beakers for immediate analysis of fungal enzyme activity, to determine the mold biomass density (p. 76). For each test material with mold gr owth carbon dioxide measurements were determined at different points in time after inoculation. For Stachybotrys chartarum inoculated onto gypsum wallboard CO 2 generation rates were determined at 24 hours after inoculation, and at weeks 1, 4, and 9. Gypsum wallboard test material previous ly wetted and placed in a chamber was later tested to evaluate CO 2 production from older colonies of growth at 27, 29, and 32 weeks post wetting. These test pieces were not inoculated in the same manner as described above. Instead, they were satura ted with tap water and placed in a sealed chamber. Growth occurred from spores that naturally occurred on these materials. Microscopic analysis of these pieces revealed the predominant colonization by Stachybotrys sp. but colonies of other fungal species were also observed to be present.
71 Comparable tests for CO 2 production from Aspergillus versicolor were performed on test rigs containing fiberglass ductboard surface coating and FSK exterior wrap with adhesive. Prior to each test the test rig wa s transferred from its growth chamber to the test chamber. During the transfer moisture content of the test pieces was measured using a Delmhorst BD-2000 pin-type moisture meter. When the materials were found to be drying out distilled, deionized water was added to support continued microbial activity. Chamber Performance Parameters Test chambers were constructed, evalua ted and operated in accordance with ASTM D 5116-97 Standard Guide for Small-Scale Environmental Chamber Determinations of Organic Emissions From In door Materials/Products (ASTM, 1999). The air change rate of the chamber was calculated using equation 4. The chamber was designed and operated to ensure good mixing an d the tracer gas decay test for quantifying mixing found the mixing level to be 71% using equation 5. Equation 1: Emission Rate per Surface Area of Mold Growth Emission Rate ER area = C s ( N/L A ) Where ER area = emission rate g (CO 2 ) cm -2 h -1 C s = steady state chamber concentration g/L C s = C out C in C out = CO 2 concentration of exhaust air from chamber (ppm) x 1.8 = g/L C in = CO 2 concentration of ambient ai r into chamber (ppm) x 1.8 = g/L N = air change rate, h -1 L A = Loading factor, cm 2 /L
Equation 2: Emission Rate per Mass of Mold Growth Emission Rate ER mass = C s (N/L M ) Where ER mass = emission rate g (CO 2 ) mg -1 h -1 C s = steady state chamber concentration g/L C s = C out C in C out = CO 2 concentration of exhaust air from chamber (ppm) x 1.8 = g/L C in = CO 2 concentration of ambient air into chamber (ppm) x 1.8 = g/L N = air change rate, h -1 L M = Loading factor (mass of mold growth), mg/L Equation 3: Average Emission Rate (mass or area) nERERnihourlyMA1 Where ER MA = Average emission rate mass: g (CO 2 ) mg -1 h -1 or area: g (CO 2 ) cm -2 h -1 ER hourly = Emission rate at each hour after chamber reaches equilibrium n = Number of discrete measurements taken hourly with chamber at equilibrium Equation 4: Chamber Air Change Rate N=Q/V (3 ACH) Where N = Air change rate (hr -1 ) Q = volume of air in to the chamber (L/hr) V = Volume of chamber (L) The chamber mixing level was calculated using equation 5. Equation 5: Chamber Mixing Level %10011111xtttCtttCtCniiiiniiiiiA 72
73 Mixing level = 71% Where = mixing level N = Chamber air change rate in units of inverse time (hr -1 ) t n = time constant of chamber = N -1 C A (t i ) = tracer gas concentration in chamber exhaust C(t i ) = concentration for perfectly mixed system, calculated by C(t) = C oe Nt n = number of discrete concentration measurements t i = time of the i th concentration measurement C o = tracer gas concentration at t=0 Chamber product loading was 200 cm 2 (mold growth) L -1 for gypsum wallboard, and 100 cm 2 (mold growth) L -1 for fiberglass ductboard surface coating and FSK exterior wrap with adhesive. Air exchange rate for the test chambers was 3 air changes per hour (ACH) with a flow rate Q of 250 ml/min. Mobilization of Antimony Trioxide Due to Fungal Growth on Fiberglass Ductboard The amount of antimony trioxide present on fiberglass ductboard (CertainTeed ToughGard), was reported on an MSDS dated June 21, 1999 to be 0.9%, but reported on a later MSDS from August 1, 2003 to contain up to 3.0% antimony trioxide by weight. Preliminary samples were collected and sent to First Environmental Laboratories, Inc. in Naperville, Illinois for ICP analysis (results re ported in Table 2). The total concentration of antimony found in fiberglass ductboard surf ace coating plus the FSK exterior wrap was 451 g/cm 2 and 417 g/cm 2 when measured using the ICP and Spectrophotometer methods respectively. This equated to only 0.23% to 0.25% antimony by weight when the reported product density of 48 kg/m 3 is used. This discrepa ncy indicated that either
74 the manufacturer was greatly over-reporting the amount of Antimony Trioxide present in the product or the acid recovery method used to analyze the product was inefficient. If the ductboard contained 0.9% to 3% as described, the mass of antimony trioxide would have been 1,641 to 5,472 g Sb/cm 2 A possible explanation of this discrepancy is that antimony concentrations of individual compone nts used to produce the final ductboard product, such as the adhesive, foil-paper b acking and surface coat ing contained greater percentages of antimony. Once combined with the fiberglass and resin which constitute the majority of the final products mass, the percentage of antimony trioxide falls well below the amounts reported. The effect of mold growth on antimony mobilization was assessed in two ways. The first method measured stibine, as total antimony, in the exhaust air of each chamber during a 4 hour period after the chambers reached equilibrium. NIOSH Method 6008 was used to collect and measure stibine as total antimony (NIOSH, 1994). The method calls for collection of air samp les onto mercuric -chloride (HgCl 2 )-coated silica gel sorbent tubes (SKC Cat No. 226-10-02). Samples were collected at a flow rate of 250 ml/min. The total volume collected was 60 L. Samples were immediately extr acted for analysis after the end of the collection period. The sorbent was placed into an acid washed 50-ml beaker along with 25 ml of concentrated HCl (Fisher Scientif ic, Trace Metals Grade Cat # A508-4). After an extraction time of 30 minutes 15 ml of the extract was tr ansferred by pipette to a 125 ml separatory funnel. Ceric sulfate was added (15 mg) and allowed to dissolve for 1 min.
75 Isopropyl ether (Fisher Scientific CAS 10820-3) was then added (15 ml) and the separatory funnel shaken for 30 sec. The two phases were allowed to separate for 1 min and the aqueous layer discarded. A working solution of Rhodamine B (20 ml) was added and shaken for 60 sec. The two phases were again allowed to separate for 1 min and the lower aqueous layer discarded. The remaining solution was drained into a 15 ml centrifuge tube, capped tightly and centr ifuged for 2 minutes at 2,000 rpm. The supernatant was transferred to an abso rption cell and capped for measurement. Samples were placed into a matched sili ca absorbance cell. Isoprolyl ether was used in the reference cell. Sample abso rbance was read at 552 nm using a Cary UV/Visible Spectrophotometer. Before each days analysis a calibration curve was determined using an isopropyl ether blank and 6 different c oncentrations of antimony in concentrated HCl, ranging from 0.05 g/ml to 1.50 g/ml. NIOSH Method 6008 was eval uated at a relative humidity of 85% and over a wide range of air concentrations, from 0.119 to 1.01 mg/m 3 using 20 L air samples. The reported limit of detection (LOD) for the method is 0.4 g SbH 3 per sample (NIOSH, 1994). This equates to a LOD for antimony emissions of 0.1 g Sb/hr from the 500 cm 2 test rig, or 0.0002 g Sb cm -2 hr -1 Calibration standards for the sample anal ysis were created using a Certified Reference Standard Solution of 1,000 ppm % Antimony in dilute HNO 3 (VWR Scientific-Ultra Scientific, Catalog Number ULICM-214). Calibration standards were
76 placed into polypropylene bottles A sufficient volume of calibration standard solutions were made to ensure that all tests could be performed using the same initial solutions. A positive control was not used in the chamber te st to verify the collection and recovery efficiency of the complete test system because trimethylstibine standards are not available for such uses (Krachler et al ., 2001). A supplier of trimethylstibine for semiconductor manufacturing was contacted, but the purity of their product was uncertain and handling of the liquid was a hazardous operation. The high concentration trimethylstibine was reported to be a pyrophoric gas that sp ontaneously combusts when mixed with air (MSDS Trimethylstibin e, 2003). Generation of stibine (SbH 3 ) using an electrolytic metallic hydride generator was considered, but due to the explosive hazard it posed, attempts to generate stibine were not carried forward (Saltzman, 1961). The second assessment for antimony mobilization was analysis of bulk samples from the test materials for antimony content. Core samples (n = 5) were collected after each carbon dioxide and stibine gene ration test. After analysis for -Nacetylhexosaminidase activity, described later, 1.4 cm diameter core samples were extracted using concentrated HCl and an alyzed using a modified NIOSH 6008 method (NIOSH, 1977). Extracts were diluted to a gr eater degree so the final solution was within the measurement range of the method, 0.05 to 1.5 g Sb/ml. Results using this method agreed with measurements of the same ma terials analyzed using the ICP/MS method 3050B/6010B. See Table 2 for ICP/MS preliminary data.
77 Measurement of Fungal Biomass on Gyps um Wallboard and Fiberglass Ductboard Fungal biomass was estimated by measuring -Nacetylhexosaminidase enzyme activity and applying a conversi on factor previously report ed for the specific fungal cultures that were used (Reeslev et al., 2003). Core samples cut from test materials (1.4 cm diameter) were analyzed using a fl uorometric method (MycoMeter, Copenhagen, Denmark) that provided rapid results within an hour. Five samples were taken from the test material after each cham ber test for carbon dioxide and stibine. Each sample was individually incubated for 30 minutes at 23 o C, or adjusted to compensate for temperature variations. After extraction, 100 l of the enzyme substrate solution was transferred to a developer solution and read using a Picofl uor fluorometer (Turner Designs, Sunnyvale, CA) according to the manufacturer s instructions (MycoMeter 2002). Prior to analysis the fluorometer was calibrated using a supplied standard a nd all reagent blanks were measured. All sample fluoroescent valu es were corrected using reagent blank measurements. Fungal biomass density was estimated usi ng the conversion fact ors reported for Stachybotrys chartarum (IBT 9695) and Aspergillus versicolor (IBT 16000) (Reeslev et al., 2003). The fungal growth area for each sample was 1.54 cm 2 The limit of quantification (LOQ) for fungal biomass fo r the method, using the conversion factors reported by Reeslev et al., were 0.8 g mold/cm 2 and 0.5 g mold/cm 2 for Stachybotrys chartarum (IBT 9695) and Aspergillus versicolor (IBT 16000) respectively (2003). The sample mass was not measured because of th e high moisture content that could not be dried without affecting the funga l colonies metabolic activity.
Equation 6: Calculating fungal biomass density CFAFBMD Where BMD = Biomass density of mold growth (mg/cm 2 ) F = Fluorescent units formed by sample in 30 min at standard temperature A = Area of sample (cm 2 ) CF = Conversion Factor of Fluorescent units formed per milligram mold biomass dry weight per 30 min at standard temperature CF SC = 8,275 F/mg for Stachybotrys chartarum (IBT 9695) CF AV = 12,370 F/mg for Aspergillus versicolor (IBT 16000) Statistical Analysis of Measurements Correlation coefficients were used to evaluate relationships between fungal biomass density and carbon dioxide emission rates, and between fungal biomass density and antimony concentrations in the test materials. To determine if measurements of antimony in test materials changed over time the Student-t test was used (Microsoft Excel Statistical Package). Antimony measurements taken at each period after mold growth had been established were compared with initial antimony concentration measurements taken at the time of inoculation, before mold growth had established. Two-tailed t-tests were performed to determine if the sample means for each data set were statistically different at the alpha level of 0.05, with an assumption of unequal variance. 78
79 RESULTS CO 2 Production Rates on Gypsum Wallboard Carbon dioxide production from building materials varied with time, mold biomass density and the material surface area with mold growth. With exception of the chamber test performed ~24 hours after inoculation, the measurements of CO 2 were made after the mold growth was assumed to have reached the stationary phase. Measures of biomass density suggest the colony reached st ationary phase by week four, but did not reach its maximum biomass density until week 29. Gypsum drywall inoculated with Stachybotrys chartarum produced measurable amounts of CO 2 in all experiments, from 24 hours to 32 weeks post inoculation. Table 3 shows the carbon dioxide production rates measured during chamber experiments of gypsum wallboard inoculated with Stachybotrys chartarum over the course of 32 weeks. The amount of CO 2 produced per cm 2 of mold growth was calculated using equation 1, and the amount of CO 2 produced per mg of mold was calculated using equation 2. The average rate of CO 2 produced per hour was calculated using equation 3 for both units. Table 3 also shows the estimated biomass density of mold at the time of each test. The biomass was calculated using equation 6.
TABLE 3: Carbon dioxide production rates for S. chartarum on gypsum wallboard Week g (CO 2 ) cm -2 hr -1 g (CO 2 ) mg -1 (Mold) hr -1 Biomass density (g/cm 2 ) (S. chartarum)* 0 (24 hours) 0.67 135 5 1 0.61 19.2 32 4 1.30 44.0 30 9 1.20 60.9 19 27 0.48 4.05 118 29 0.87 1.91 458 29** 1.10 1.97 570 31 0.30 1.58 190 32 0.32 4.04 79 *Biomass Density (mg mold/cm 2 ) = (MV/1.54 cm 2 )/8,275 Conversion Factor (Reeslev, 2003) ** Replicate test on a separate piece of GWB with higher biomass density of mold growth. Figure 11: CO 2 generation and biomass density on gypsum wallboard Gypsum Wallboard0.670. 3 0.300.870.481.181.330.61050100150200250300350400450500051015202530Weeks of Mold Growth (S. chartarum)Mold Density (ug/cm2 ) Mold density (ug/cm2) Carbon Dioxide (ug/hr .cm2) Carbon dioxide production rates for S. chartarum grown on gypsum wallboard ranged from 0.30 to 1.30 g (CO 2 ) cm -2 hr -1 and 1.58 to 135 g (CO 2 ) mg -1 (mold biomass) hr -1 throughout its growth cycle. Compared to other studies where this strain of Stachybotrys chartarum was grown on gypsum wallboard, the maximum biomass density 80
81 achieved was low and did not reach its peak un til much later (Reeslev, 2003). In light of these results, the values obtained may be reflective of slow mold growth. CO 2 Production Rates on Fiberglass Ductboard Fiberglass ductboard (surface coating) inoculated with Aspergillus versicolor produced measurable amounts of CO 2 in all experiments, from 24 hours to 29 weeks post inoculation. Table 4 shows the carbon di oxide production rates measured during two separate chamber experiments of fibe rglass ductboard inoculated with Aspergillus versicolor over the course of 29 weeks and 9 weeks. The amount of CO 2 produced per cm 2 of mold growth was calculated using equation 1, and the amount of CO 2 produced per mg of mold was calculated using equation 2. The average rate of CO 2 produced per hour was calculated using equation 3 for both units. Table 4 also shows the estimated biomass density of mold at the time of each test. The biomass was calculated using equation 6. TABLE 4: Carbon dioxide production rates for A. versicolor on ductboard surface coating Test Number 1 Test Number 2 Week g (CO 2 ) cm -2 hr -1 g (CO 2 ) mg -1 (mold) hr -1 Biomass density ( g/cm 2 ) ( A. Versicolor )* g (CO 2 ) cm -2 hr -1 g (CO 2 ) mg -1 (mold) hr -1 Biomass density ( g/cm 2 ) ( A. Versicolor )* 0 (24 hrs) 1.66 176 9 3.08 598 5 1 0.82 10.5 78 2.65 115 23 2 2.02 8.83 228 4 1.15 3.56 324 1.18 3.18 371 7 1.91 12.86 148 9 1.60 15.65 103 1.17 2.48 470 29 1.13 2.96 381 *Biomass Density (mg mold/cm 2 ) = (MV/1.54 cm 2 )/12,370 Conversion Factor (Reeslev, 2003) ** Values represent a second e xperiment on the same material. -: No Data Collected
Figure 12: CO 2 generation and biomass density on ductboard surface coating test #1 Ductboard Surface Coating (#1)29 Weeks1.660.821.131.601.911.152.02050100150200250300350400450500051015202530Weeks of Mold Growth (A. versicolor)ug/cm2 of Mold00.511.522.533.544.55ug/cm2 x hr of CO2 Mold density (ug/cm2) Carbon Dioxide (ug/hr .cm2) Figure 13: CO 2 generation and biomass density on ductboard surface coating test #2 Ductboard Surface Coating (#2)9 Weeks1.173.082.651.18050100150200250300350400450500012345678910Weeks of Mold Growth (A. versicolor)ug/cm2 of Mol d 0.000.501.001.502.002.503.003.504.004.505.00ug/cm2 x hr of CO2 Mold density (ug/cm2) Carbon Dioxide (ug/hr .cm2) 82
83 Carbon dioxide production rates for A. versicolor grown on fiberglass ductboard surface coating ranged from 0.82 to 2.02 g (CO 2 ) cm -2 hr -1 and 2.96 to 176 g (CO 2 ) mg -1 (mold biomass) hr -1 in Test #1, and 1.17 to 3.08 g (CO 2 ) cm -2 hr -1 and 2.48 to 598 g (CO 2 ) mg -1 (mold biomass) hr -1 in Test #2. Measures of biomass density suggest the colony reached stationary pha se by week four in Test #1, but may not have reached stationary phase in Test #2 by week nine. Fiberglass ductboard FSK exterior wrap inoculated with Aspergillus versicolor produced measurable amounts of CO 2 in all experiments, from 24 hours to 29 weeks post inoculation. Measures of biomass density s uggest the colony did not reach stationary phase. Table 5 shows the carbon dioxide production rates measured during chamber experiments of fiberglass ductboard FSK exterior wrap inoculated with A. versicolor over the course of 29 weeks. The amount of CO 2 produced per cm 2 of mold growth was calculated using equation 1, and the amount of CO 2 produced per mg of mold was calculated using equation 2. The average rate of CO 2 produced per hour was calculated using equation 3 for both units. Table 5 also sh ows the estimated biomass density of mold at the time of each test. The bioma ss was calculated using equation 6.
TABLE 5: Carbon dioxide production rates for A. versicolor on ductboard FSK exterior wrapping Week g (CO 2 ) cm -2 hr -1 g (CO 2 ) mg -1 (mold) hr -1 Biomass density (g/cm 2 ) (A. Versicolor)* 1 1.75 392 4 3 1.93 218 9 5 0.50 94.6 5 9 0.02 4.92 4 14 0.68 18.6 36 29 1.76 3.95 445 *Biomass Density (mg mold/cm 2 ) = (MV/1.54 cm 2 )/12,370 Conversion Factor (Reeslev, 2003) Figure 14: CO 2 generation and biomass density on FSK exterior wrapping Ductboard Foil Backing29 Weeks1.751.930.500.020.681.76050100150200250300350400450500051015202530Weeks of Mold Growth (A. versicolor)ug/cm2 of Mold0.000.501.001.502.002.503.003.504.004.505.00ug/cm2 x hr of CO 2 Mold density (ug/cm2) Carbon Dioxide (ug/hr .cm2) Carbon dioxide production rates for Aspergillus versicolor grown on fiberglass ductboard FSK exterior wrapping ranged from 0.02 to 1.93 g (CO 2 ) cm -2 h -1 and 3.95 to 392 g (CO 2 ) mg -1 (mold biomass) h -1 throughout its growth cycle. 84
Modeled Concentration of CO 2 in Wall Cavities To model CO 2 concentrations within an enclosed cavity such as a wall cavity or HVAC duct, the model described in equation 7 was used. This equation was solved for the following assumptions, and presented as equation 8, air exchange rate, I = (Q/V), generation rate (G), and the outdoor concentrations (Co) remain constant. Equation 7: Model for CO 2 concentration inside a wall cavity and HVAC Duct GCCVQdtdCtot)( Equation 8: Used to calculate equilibrium concentration of wall cavity and HVAC duct at equilibrium ItoteQGCC1 where C t = Concentration of CO 2 at time t, C o = outdoor CO 2 concentration, G = generation rate of CO 2 in the space, Q = volumetric airflow rate into and out of the space, V = volume of cavity, I = air exchange rate (Q/V) The concentration of carbon dioxide at equilibrium in a wall cavity could only become elevated above ambient concentrations by 25 ppm if over of the surface area on one side of the cavity, or of the total internal surface area, was supporting mold 85
86 growth (S. chartarum). In a single wall stud bay, with a volume of 88 L, it would take ~0.45 m 2 of mold growth to elevate CO 2 concentrations by ~25 ppm. However, this rise in CO 2 could only be expected during the first 9 weeks of mold gr owth. In later stages of growth, measured between 27 and 32 weeks, the generation of CO 2 could not be expected to cause an elevation of more than 25 ppm. When the situation of a larger wall cavity was modeled the same ratio was nece ssary to cause an elevation of CO 2 by 25 ppm or more. An elevation of 25 ppm in CO 2 was chosen as a significant rise based on the reported accuracy of the hand-held CO 2 monitor used. At 375 ppm the accuracy of the monitor is reported to be 3% or 12 ppm. In order for the measurements taken from inside the cavity and in the ambient air to be distinctly different they must differ by at least 25 ppm from ambient concentrations. At a higher ambient c oncentration of 1,000 ppm, the difference necessary to demonstrate elevated CO 2 is 60 ppm. Table 6 shows the modeled CO 2 concentrations within a wall cavity given the described parameters and assumptions. Table 6: Modeled concentrations of CO 2 inside a wall cavity colonized with (S. chartarum ) Age of mold growth Single Stud Bay 16 in x 8 ft (1) Single Stud Bay 16 in x 8 ft (2) Single Wall 10 ft x 8 ft (3) Single Wall 10 ft x 8 ft (4) 24 hours 413 ppm 394 ppm 414 ppm 395 ppm 4 weeks 451 ppm 413 ppm 453 ppm 414 ppm 31 weeks 392 ppm 384 ppm 392 ppm 384 ppm
87 (1) 9,000 cm 2 of mold growth (~ entire wall surface) (2) 4,500 cm 2 of mold growth (~1/2 wall surface) (3) 74,322 cm 2 of mold growth (~entire wall surface) (4) 37,161 cm 2 of mold growth (~1/2 wall surface) 16 Stud Bay Volume = 88 L 10 Wall Volume = 708 L Air Change per hour from outdoors = 1 Outdoor concentration of CO 2 = 375 ppm Modeled Concentration of CO 2 in HVAC Systems The amount of mold growth (A. versicolor), in a static HVAC duct or air handling unit (AHU) cabinet, necessary to cause an elevation of CO 2 by more than 25 ppm was ~1/2 of the total internal su rface area. Approximately 8 m 2 of mold growth would be necessary to elevate CO 2 by 25 ppm in a single duct measuring 40.6 cm x 43.2 cm x 0.975 m (16 in x 17 in x 32 ft) with an internal surface area of 16.35 m 2 Approximately 1.1 m 2 of mold growth on fiberglass lining with in an AHU cabinet would be necessary to elevate CO 2 by 25 ppm, or ~ of the total internal surface area. Table 7: Modeled concentrations of CO 2 inside a static supply duct ( A. versicolor ) Age of mold growth Single Supply Duct 16 in x 17 in x 32 ft (1) Single Supply Duct 16 in x17 in x 32 ft (2) AHU Cabinet 2 ft x 2 ft x 3 ft (3) AHU Cabinet 2 ft x 2 ft x 3 ft (4) 24 hours 463 ppm 419 ppm 430 ppm 403 ppm 4 weeks 436 ppm 405 ppm 413 ppm 394 ppm 29 weeks 435 ppm 405 ppm 413 ppm 394 ppm
88 (1) 163,509 cm 2 of mold growth (~entire duct surface) (2) 81,755 cm 2 of mold growth (~1/2 duct surface) (3) 22,311 cm 2 of mold growth (~en tire AHU cabinet surface) (4) 11,156 cm 2 of mold growth (~ 1/2 AHU cabinet surface) 32 ft long duct Volume = 1,716 L AHU Cabinet Volume = 372 L Air Change per hour from outdoors = 1 Outdoor concentration of CO 2 = 375 ppm Discussion Chamber tests for carbon dioxide generation from growth of Stachybotrys chartarum on gypsum wallboard reflected the different stages of growth commonly recognized. Early log-phase growth occurs once conditions allowed germination of spores, and was characterized by an exponential increase in biomass density. This phase of growth appeared to coincide with a disproportionately large CO 2 generation rate with respect to the biomass of mold present. B ecause the experimental design limited the area of mold growth to the surface area of test materials, only the biomass density could increase. Variations in car bon dioxide generation rates a ppear to be a function of metabolic activity and biomass density. As an indicator of mold growth, carb on dioxide clearly increased as fungal biomass increased during the first 24 hours when compared to un-inoculated controls for all test materials. Chamber tests of cl ean, dry test material s without inoculation
89 demonstrated no measurable generation of car bon dioxide. The primary focus of this experiment was carbon dioxide generation during the stationa ry phase of mold growth, not during the log-phase. Only a single cham ber test was performed during the log-phase within 24 hours after inoc ulation. Early stationary phase, weeks 1 through 9, demonstrated less carbon dioxide generation pe r area of mold growth or per mass of mold growth than during the log phase. Late stationary phase, weeks 27 through 32, demonstrated declining biomass density and CO 2 generation. This is not surprising as nutrient availability was probably the contro lling factor that leads to reduced fungal metabolic activity. However, it was observed that throughout the various growth phases of both species, on both types of test materi als, carbon dioxide generation rates were greater than 0.61 g (CO 2 ) cm -2 hr -1 of mold growth during early stationary growth at 1 week and 0.30 g (CO 2 ) cm -2 hr -1 of mold growth during late stage growth at 32 weeks. Dispersion of carbon dioxide throughout the wall cavity could impact the measured CO 2 concentration if not evenly disperse d. A study on the dispersion of carbon dioxide inside closed spaces demonstrated that regardless of the source location within an enclosed space, such as a wall cavity, the ga s is evenly dispersed throughout the entire cavity after 90 minutes (Rahamani, 2004). The time scale being evaluated for CO 2 generation by fungi is on the order of days or weeks, making the issue of stratification and uneven mixing irrelevant. The modeling results predict that det ectable elevations in carbon dioxide concentrations could be detected only from an area of mold growth th at is greater than
90 the internal surface area inside of a wall cavity. With larger areas of mold growth, or in cavities with lower ventilation rates, the c oncentration of carbon dioxide could become more elevated within a wall cavity or supply duct. However, the usefulness of CO 2 monitoring as an indicator of hidden mold growth is impractical and could not be used to demonstrate the absence of hidden mold growth. In addition to the concentr ation of carbon dioxide, a m onitoring relative humidity could be used as an indicator of high mois ture conditions. Elevated relative humidity levels within a wall cavity may serve as an indicator of either a water source, infiltration of moisture-laden outside air, or microbial respiration. Measuremen t of both elevated carbon dioxide levels and high relative humid ity could provide greater evidence that mold growth, or other metabolically active microorganisms, may be present within the building cavity. Measured Emission Rates of Stibine as Total Antimony All air samples collected for stibine analysis were below the detection limit of the method (0.4 g SbH 3 per sample or 0.0002 g Sb cm -2 hr -1 ). Relative humidity measurements and observed condensation within the chambers indicated extremely high humidity, surpassing 95%. When the sorben t media was transferre d from the collection tubes for extraction agglomeration of the sili ca beads was noted. It was hypothesized that high moisture levels in the chamber exhaust possibly interfered with collection of any stibine that may have been generated. The NIOSH method 6008 had only been evaluated at relative humidity as high as 85%.
91 Measured Antimony Levels in Test Materials with Fungal Growth Bulk samples collected from fiberg lass containing antimony trioxide (Sb 2 O 3 ) were inoculated with Aspergillus versicolor. The mean (n = 5) concentration of antimony in the surface coating of ductboard materials before fungal growth was 255 g (Sb) cm -2 in materials used for Test #1 and 221 g (Sb) cm -2 in Test #2. Nine weeks after inoculation Test #1 materials ha d a mean antimony level of 191 g (Sb) cm -2 and Test #2 materials had reduced to 166 g (Sb) cm -2 These results show a reduction in antimony concentration of 25% for each test after 9 w eeks of mold growth. The concentration of antimony after 9 weeks of fungal degradation was significantly reduced in both tests by 64 and 55 g (Sb) cm -2 or 25% (P-value <0.05). Tabl e 8 shows the results of antimony measurements and biomass density of mold. Table 8: Antimony in fiberglass duct board surface coating test material Test Number 1 Test Number 2 Week *Mold Density g (mold) cm -2 Antimony g (Sb) cm -2 Std Dev *Mold Density g (mold) cm -2 Antimony g (Sb) cm -2 Std Dev 0 9 255 40.43 5 221 41.32 1 78 192 12.61 23 2 228 208 26.7 4 324 209 19.8 371 192 39.34 7 148 202 16.27 9 103 191 22.28 470 166 20.42 29 381 197 7.68 -: No Data Collected The limit of quantification (LOQ) for fungal biom ass for the method, using the conversion factors reported by Reeslev et al., were 0.8 g mold/cm 2 and 0.5 g mold/cm 2 for Stachybotrys chartarum (IBT 9695) and Aspergillus versicolor (IBT 16000) respectively.
Figure 15: Sb concentration and biomass density on fiberglass ductboard test #1 Ductboard Surface Coating (#1)9 Weeks191*255202*209208192*050100150200250300350400450500012345678910Weeks of Mold Growth (A. versicolor)ug/cm2 of Mold & Antimony Mold density (ug/cm2) Sb/cm2 Asterisks indicates mean antimony sample measurement was significantly different from initial measurement taken of test material =0.05. Figure 16: Sb concentration and biomass density on fiberglass ductboard test #2 Ductboard Surface Coating (#2)9 Weeks166*192221050100150200250300350400450500012345678910Weeks of Mold Growth (A. versicolor)ug/cm2 of Mold & Antimony Mold density (ug/cm2) Sb/cm2 Asterisks indicates mean antimony sample measurement was significantly different from initial measurement taken of test material =0.05. 92
The mean concentration of antimony in new FSK Exterior Wrapping of ductboard materials at one week after inoculation, before fungal growth had become established, was179 g (Sb) cm -2 Nine weeks after inoculation these materials had a mean antimony level of 148 g (Sb) cm -2 and after 29 weeks the mean antimony level had reduced to 101 g (Sb) cm -2 These results show a reduction in antimony concentration of 17% after 9 weeks of mold growth and 43% after 29 weeks. The concentration of antimony after 29 weeks of fungal growth was significantly reduced by 78 g (Sb) cm -2 or 43% (P<0.01). Table 9: Antimony in fiberglass ductboard FSK exterior wrapping test material Week Mold Density g (mold) cm -2 Antimony g (Sb) cm -2 Std Dev 1 4 179 35.19 3 9 167 18.97 5 5 188 22.16 9 4 148 7.35 14 36 150 22.64 29 445 101 25.71 Figure 17: Sb concentration and biomass density on FSK exterior wrapping Ductboard Foil Backing29 Weeks101*150148188167179050100150200250300350400450500051015202530Weeks of Mold Growth (A. versicolor)ug/cm2 of Mold & Antimony Mold density (ug/cm2) Sb/cm2 Asterisks indicates mean antimony sample measurement was significantly different from initial measurement taken of test material =0.05. 93
94 Estimated Antimony Emission Rates Test #1 on fiberglass ductboard surface coating suggests the mobilization took place over the first week of exponential mold growth, but Test #2 suggests the mobilization rate was relatively constant over th e entire nine week period. Each scenario was examined. Loss of antimony from test materials was assumed to be from the antimony trioxide fire retardant reportedly applied during manufacturing. Because of limitations in the analytical method used, the molecular form of antimony present in the test material could not be directly determ ined. The available method only allowed for determination of total antimony in the solid substrate. Manufacturer documents indicated that antimony was applied to the product in the form of antimony trioxide as a fire retardant. In Test #1 the substrate lost 25% of its initial mass of antimony per cm 2 in the first week. This equates to 63 g (Sb) cm -2 over 1 week, or 168 hours, resulting in a theoretical generation rate of 0.38 g (Sb) cm -2 hr -1 (3.8 mg (Sb) m -2 hr -1 ). When mold growth slowed and biomass density began to subside, presumably due to nutrient limitations, antimony concentrations in the materi al stabilized. This accelerated release of antimony appears to occur during the rapid increase in mold biomass as indicated by the simultaneous increase in biomass density measurements. The material in Test #2 lost 25% of its initial mass of antimony per cm 2 at a relatively constant rate over the cour se of 9 weeks. This equates to 55 g (Sb) cm -2 over a 9 week, or 1,512 hours, resulting in a theoretical generation rate of 0.036 g (Sb) cm -2
95 hr -1 (0.36 mg (Sb) m -2 hr -1 ). However, in this experime nt mold growth did not subside during the testing. This patte rn of biomass density suggest s the fungal growth on this material had not reached its maximum and thus had not reached the stationary phase by 9 weeks of growth. This slower release of an timony appears to reflect the slower growth rate of this colony, as evidenced by the biomass density measurements. The FSK exterior wrapping test material #3 lost 43% of its initial mass of antimony per cm 2 over the course of 29 weeks. The trend of measurements appeared to be similar to Test #2 where the antimony wa s mobilized at a slow rate throughout the growth of the mold. This equates to 78 g of (Sb) cm -2 over the course of 29 weeks, or 4,872 hours, resulting in a theore tical generation rate of 0.016 g (Sb) cm -2 hr -1 (0.16 mg (Sb) m -2 hr -1 ). The slow growth rate observed over the first 9 weeks of this experiment possibly reflects a slow downward trend in antimony concentration, but once exponential growth began at 29 weeks, a significant reduction in the antimony level in the test material was observed. Initial antimony measurements were compared to antimony measurements taken at each subsequent period after mold growth had been established. Two-tailed t-tests of the sample means for each data set were perfor med to determine if they were statistically different at the alpha level of 0.05, with an assumption of une qual variance. Statistically different sample means are annotated w ith an asterisk on figures 15, 16 and 17.
Using the reduction of antimony concentrations from test materials, antimony aerosol generation rates were estimated. Generation rates for total antimony were calculated and used to model indoor air concentrations over the course of 1 week, 9 weeks and 29 weeks. A simple, single room mass balance model was used to estimate indoor air concentrations of antimony as the oxidative products of trimethylstibine or trimethylstibine oxides. The generation rate of antimony estimated as an emission factor was 0.357 g Sb cm -2 hr -1 for the condition of 25% reduction in 1 week; 0.0398 g Sb cm -2 hr -1 for the condition of 25% reduction in 9 weeks; and 0.022 g Sb cm -2 hr -1 for the condition of 43% reduction in 29 weeks. All of these theoretical emission factors are above the estimated limit of detection (0.0002 g (Sb) cm -2 hr -1 ) for the chamber test method used in the study. Modeled Concentration of Trimethylstibine Oxides as Antimony Equation 10: Used to calculate equilibrium concentration of antimony compounds in a 3,600 ft 2 (334 m 2 ) home ItoteQGCC1 where C t = Concentration of Sb at time t, C o = outdoor Sb concentration = 0, G = generation rate of Sb in the space, 96
97 Q = volumetric airflow rate into and out of the space, V = volume of home, I = air exchange rate (Q/V) Model assumptions: Volume = 815 m 3 (28,800 ft 3 ) of a 3,600 ft 2 home Air Exchange Rate: 1 ACH Entire Duct System contained 61 m 2 of mold growth on fire retardant treated duct. Containing 261.65 g of Sb 2 O 3 Single 32 ft (9.75 m) supply duct contained 8.18 m 2 of mold growth on fire retardant treated duct. Containing 20.44 g of Sb 2 O 3 All antimony was assumed to be released as the gaseous trimethylst ibine as a result of fungal decomposition and rapidly oxidized to a particulate-phase trimethylstibine oxide.
Table 10: Modeled concentration of antimony as trimethylstibine oxides Figure 18: Sb concentration in a home with all ducts supporting mold growth Source Description Generation Rate (mg/hr) Mass of Sb Mobilized Resulting Concentration (mg/m 3 ) Depicted in Figures 18 & 19 as Entire Duct System 261.65 g 25% in 1 week 373.81 0.4587 FIG 16: A 25% in 9 weeks 41.53 0.0510 FIG 16: B 43% in 29 weeks 23.1 0.0283 FIG 16: C Single 32 ft duct 20.44 g 25% in 1 week 29.20 0.0358 FIG 17: D 25% in 9 weeks 3.25 0.0040 FIG 17: E 43% in 29 weeks 1.8 0.0022 FIG 17: F Estimate concentrations of Sb at equilibriumin a home with all ducts effected00.050.10.150.18.104.22.1680.40.450.50123456789101112Duration (Hours)Sb Air Concentration (mg/m 3 A B C 98
Figure 19: Sb concentration in a home with one duct supporting mold growth Estimate concentrations of Sb at equilibriumin a home with 1 effected duct00.0050.010.0150.020.0250.030.0350.040123456789101112Duration (Hours)Sb Air Concentration (mg/m 3 D E F Discussion Using Scopulariopsis brevicaulis, grown aerobically as a submerged culture in 500 ml flasks, volatile antimony compounds were measured. Using the reported data I estimated the generation rates from fungal growth in media amended with various sources of antimony. When antimony trioxide (Sb 2 O 3 ) was added as the source of antimony, 0.004 g Sb/hr was estimated to be mobilized over the 8 day experiment. 99
100 When potassium antimony tartrate (PAT) was added as the antimony source, 0.04 g Sb/hr was estimated to be mobilized over the 8 day experiment. Because the experiment was performed on cultures in a liquid medium comparison with surface areas of growth cannot be made. When the initial amount of antimony added to the cultures is compared, the experiments by Jenkins et al. suggest that only 0.002% of th e initial amount of antimony was mobilized in 8 days. However, th e researchers concluded that increases in antimony concentration effected mobilized am ounts in a disproportional manner. The amount of antimony mobilized only rose 2-fold when there was a 20-fold increase in the amount added to the cultures (Jenkins et al, 1998a). The antimony emission rates calculated from measured reduct ions in antimony concentrations on ductboard materials were on the order of 10 to 100 times higher than those calculated from mobilizat ion amounts reported by Jenkins et al. However, this was one of the few studies that reported meas uring volatile antimony compounds from adding antimony trioxide to fungal cultures. Estim ates of indoor antimony concentrations are maximum levels calculated using worst case conditions. Without data identifying antimony in a gaseous form (stibine or trimethylstibine) the pathway of antimony mob ilization from the antimony trioxide-treated building materials cannot be concluded. There are tw o possible mobilization pathways that may have occurred. The first is that fungal gr owth mobilized the antimony trioxide by uptake and biomethylation via the Challenger Mech anism. The trimet hylstibine may have quickly oxidized and deposited onto the su rfaces of the chamber or captured on the
101 exhaust filter before reaching the sorb ent tube. The environmental fate of trimethylstibine in the environment is reported in the literature to be a multi-step process that results in formation of a mixture of antimony oxides and insoluble polymers. The residence time of trimethylstib ine in the air is not expect ed to be long, requiring a constant source to result in a measurable amount. The second possibility is that antimony tr ioxide was mobilized via mechanical means enhanced by fungal decomposition of the test materials. The antimony trioxide would have been in particulate form, a nd may have been shed along with fungal fragments, spores or degraded test material. The two possible scenarios described above could result in occupant exposure to trimethylstibine oxides, antim ony trioxide, or a combination of both compounds. The inhalation reference conc entration (RfC) of anti mony trioxide is 0.0002 mg/m 3 (NAS, 2000). An inhaltion RfC has not been determ ined for trimethylstibine oxide or total antimony. Using the inhalation RfC for antimo ny trioxide a hazard index was estimated to range from 11, at the lowest emission rate to 2,283 for the highest emission rate. Since the estimated concentrations do not represent lifetime exposures, calculation of a cancer risk was not performed. When water damage occurs to building materials treated with antimony trioxide fire retardant, fungal growth can occur. This has been demonstrated on both components of fiberglass ductboard that contain an timony, specifically the surface coating and the
102 adhesive-FSK exterior wrap. Studies ha ve also found fungal growth to occur on insulation materials, with and without added biocides (Ezeonu et al., 1994; Price et al., 1994; Samimi et al., 2003; Van Loo et al., 2004). Laboratory studies have demonstrated that at least some fungal sp ecies can biomethylate antimony, m obilizing it initially as the highly toxic volatile compound trimethylstibine, but rapidly oxidiz ing it to non-volatile precipitates of antimony. Th e purpose of this study wa s to determine if antimony trioxide, present on commonly used building materials, could mobilize antimony aerosols when fungal growth occurred. The volatile form of antimony was not successfully collected and identified, but material tests demonstrated significantly lower concentrations of antimony (P <0.05) rema ining after fungal growth had occurred. Concentrations of antimony in test mate rials after fungal growth had occurred demonstrated a significant reduction from the original amount in three separate experiments. The major difference between each experiment was the rate at which the antimony concentrations reduced. The m echanisms causing variations in the mobilization rate of antimony from the test ma terial could not be determined using this experimental design.
103 CONCLUDING REMARKS Public Health Significance of Research The etiologic agent of bu ilding occupant complaints, adverse health effects and disease outbreaks are often not determined specifically determined for any one case (OReilly, 1998). Rather, rese archers and public health pr ofessionals often identify situations, conditions and contam inants that should not be pr esent or are in excessively high concentrations. Guidance is then give n to remove the contaminant sources and remedy the conditions as part of a trial and e rror approach. If the complaints cease or illnesses resolve, the recommendations are view ed as correct. Quite often the specific cause of an adverse health effect is not de termined because no biomarkers are known to exist that indicate cessation of exposure. This has been the situation with most cases of fungal growth in indoor environments. Fungal growth in indoor environments has been associated with adverse health effects and when found, its removal is recommended (EPA, 2001; IOM, 2004; Macher, 1999). Resear ch continues to identify fungal agents possibly responsible for occupant illnesses, but many symptoms and illnesses that are commonly attributed to fungal exposure have not been linked to measured fungal aerosols (IOM, 2004).
104 Some physicians and public health prof essionals in the United States posed a possible association between sudden infant death syndrome (SIDS) and exposure to moldy indoor environments. This proposed asso ciation resulted from the investigation of a cluster of infant deaths in the Clevel and area due to acute idiopathic pulmonary hemorrhage (AIPH) in 1993 (CDC, 1994; CDC, 1997; Etzel et al., 1998). The data collected during the initial investigation was reviewed by a CDC scientific task force and a very different conclusion was published in a March 2000 revision report. The post-hoc reviewers concluded that this study was not of sufficient quality to support an association between S. chartarum and AIPH. In addition, the review ers noted that evidence from other sources supporting a causal role of S. chartarum in AIPH is limited. The task force concluded that S. chartarum was not clearly associated with AIPH CDCCDC, 2000). Since the initial CDC repor t in 1994 two additional case reports of AIPH in infants have examined the link between pulmonary hemorrhage and exposure to S. chartarum. One case study published in 1999 identified likely exposure to S. chartarum and other fungi due to water damage in th e home (Flappan et al., 1999). The second case report published in 2002 focused on isolation of S. chartarum from the lung of a child diagnosed with AIPH and identification of a new serine proteinase described as satachyrase A (Kordula et al., 2002). Similarly in 1994 claims were made that fi re retardants in crib mattresses were, in part, responsible for SIDS (Blair et al ., 1995; Cook, 1994; Thompson & Faull, 1995). Researchers claimed that antimony and arsenic we re released as volatile toxic forms into
105 the lungs of sleeping infants, resulting in death. Independent study conducted in England concluded that transformation of the trace c oncentrations of arsenic by bacteria into arsine gas was not possible under the aerob ic conditions present in the mattress environment. However, possible inhala tion exposure to antimony could not be dismissed. The authors of the study ev entually concluded the emission of trimethylstibine would be too low to cause de ath and that SIDS had been occurring long before infants began sleeping on fire-retarda nt treated mattresses. The speculative identification of toxic gas fo rmation in mattresses was eventually dismissed after study by the government appointed panel (DeWolff 1995; Jenkins et al., 1998b; Limerick, 1998). The available evidence clearly indicate s that neither toxige nic fungi nor antimony fire retardants caused SIDS. However, th e CDC investigators did not consider the potential for microbial rel ease of antimony when evaluati ng the association between fungal growth on fire retardant-treated ma terials and AIPH. Antimony trioxide is strongly irritating to tissues and membranes. Of greater public health concern is the potential for chronic respirat ory irritation resulting from exposure to antimony oxides including antimony trioxide released by fungal degradat ion of fire retardants. In light of the documented mechanism and potential for antimony transformation by fungi from an immobile, non-volatile surface application to a volatile form that rapidly oxidizes and precipitates, the impacts on exposed occupant should be examined. The results of this limited study, while not conc lusive, do support the speculation by previous
106 authors that fungal growth on antimony-containing building materials can mobilize the applied fire retardant. Fu rther research on occupant exposures in environments where fungal growth on fire retardant-treated material s exists is justified. Populations with the greatest susceptibility such as infants, asthmatics and others with compromised respiratory systems could be adversely a ffected by exposure to antimony compounds in addition to allergens and irritants released from microbial sources. The potential for a synergistic effect should also be considered in both sus ceptible and healthy populations. Practical Applications of This Research Carbon Dioxide does not appear to be a sensitive indicator of hidden fungal growth in wall cavities and static HVAC Systems. Available sampling and analysis methods are impractical, expensive, or insens itive a new approach should be examined. Fluctuations in CO 2 generation rates that can be infl uenced by colony age, moisture deprivation and limited nutrien t availability, make carbon di oxide a poor indicator of hidden fungal growth outside of laboratory conditions. Carb on dioxide could not be used to demonstrate the absence of fungal gr owth in a hidden cavity, as transient environmental conditions can inhibit CO 2 production of even well established colonies. Study Limitations Carbon dioxide generation rates were onl y examined during the initial log phase, and stationary phase of growth ex tending to 32 weeks. Trends in CO 2 generation rates for later stages of growth indicated that detect able carbon dioxide generation eventually ceased. Based upon the model used to predict co ncentrations within enclosed cavities at
107 least to of the internal surface area must have fungal gr owth in order to detect a significantly elevated CO 2 concentration. While metabolism may be an important factor in the mobilization of antimony fire retardan ts, fungi also produce allergens and other irritants that may have an adverse imp act on building occupants and remediation personnel. Until a full understanding of all the hazards that indoor mold growth presents, care must be taken to prevent human exposur e and to effectively remediate conditions that allow fungal growth in indoor environments. Measurement methods for trimethylstib ine were based on detection of total volatile antimony, and were not capable of speciating the valence forms or molecular forms present. The rapid oxidation rate of trimethylstibine, on the order of 10 3 M -1 s -1 most likely caused the precip itation of any trimethylstibin e that was generated to particulate-phase oxides (Parris & Brinckman, 1976). The presence of an in-line filter may have captured antimony oxides before they reached the sorbent tube, but no analysis of the filter was ever performed. Additionall y, high humidity of the test chambers may have interfered with collection or recovery of any trimethylstibine that may have been generated. Because a positive control test wa s not performed, due to the unavailability of a safe gas-phase standard and the explosive nature of stibine generators, full confidence in the ability to detect trimethylstibine or trimethylstibine oxides was not achieved. Estimates of biomass present in samples were based upon conversion factors calculated from experiments performed in Denm ark by Reeslev et al (2003). While the same strains of fungal species used to calculate th e conversion factors were used in this
108 experiment, differences between growth substr ates and laboratory conditions could have biased the results in an undetermined ma nner. As the repeated experiments with fiberglass ductboard surface coating demonstrat ed, growth rates can be different due to unaccounted for variables. De spite recognized limitations in the calculated biomass, based on the fluorometric assay of -N-acetylhexosaminidase activity, these results were considered more reflective of biomass than traditional mycological methods that rely upon serial dilution and colony formation on nutr ient agar. The important component of fungal growth considered here is hyphae formation, which typically accounts for 95-97% of fungal colony biomass. Culture based methods rely upon quantifying viable spores which account for only 3-5% of fungal biomass. The use of mold colony counts (CFU) is, to a large extent, a measure of spor ulation, not fungal biomass (Schnurer, 1993). Recommendations for Future Research Examination of antimony trioxide fire re tardant application to indoor products and building materials should be performed. A broad-based assessment of fire retardant treated materials currently in buildings should be performed to evalua te the impact of age and microbial degradation. Better indicators of exposure to antimony should be researched. Measurement of antimony in ur ine and feces only evaluates the portion that is absorbed into the body. Damage to the respiratory system due to surface irritation in the lungs, may not reflect the damage caused by unabsorbed antimony compounds. An assessment of antimony exposures in homes could provide a better understanding of population risks. Begi nning in 2005, CDC's National Health and
109 Nutrition Examination Survey (NHANES) will collect dust samples from approximately 7,000 participants' homes each year. Analysis of dust for antimony could help establish what levels Americans are exposed to. When researchers perform studies that attempt to assess mold exposure, they should also documen t the type of materi al that is supporting mold growth within a building. Air samples for antimony in moldy indoor environments could be compared to air concentrations and personal exposure samples for antimony in environments without mold contamination. Field portable X-Ray Fluorescence devices, using X-ray tube technology or an Americium-source, can directly measure Sb on surfaces. If a survey determines that mold growth is occurring on antimony-contai ning materials, the difference between surfaces with mold growth and surfaces not supporting mold growth could indicate if mobilization of antimony into the indoor environment may be taking place. Future research on the public health imp acts of biological cont amination of indoor environments should include assessment of aerosols released from the microbial degradation of building materi als such as synthetic floori ngs, wall coverings, coatings, paints, adhesives and insulation. The potential for toxic and irritating aerosol emissions from the microbial break down or biomethylat ion of additives should be examined in depth. Understanding these seldom evaluated aerosols may help to explain the inconsistency of study findings contributing to the paradox of mold exposures and human health effects.
110 REFERENCES ACCA. (1995). Residential Duct Systems Manual D 2 nd Printing. Washington, D.C.Air Conditioning Contractors of America. ACGIH. (1986). Documentation of the Threshold Lim it Values and Biological Exposure Indices. 5th ed. Cincinnati, OH: American Conference of Governmental Industrial Hygienists. Ahearn, D.G., Price, D.L., Simmons, R.B., & Crow, S.A. (1992). Colonization studies of various HVAC insulation materials. In Proceedings of IAQ '92 Environment for people. Atlanta, GA. American Society of Heating, Refrigeration and AirConditioning Engineers Inc. AIHA. (2001). Report of Microbial Growth Task Force. Fairfax, VA. AIHA Press. Amrane, A, Plibon, F, & Pringent, Y. (1999). Proton transfer in rela tion to growth of Geotrichum candidum and Penicillium ca memberti in synthetic liquid media. Enzyme and Microbial Technology, 24 (8-9), 561-568. Andrewes, P., Cullen, W.R., Feldmann, J., Koch, I., Polishchuck, E., & Reimer, K.J. (1998). The production of methylated organoantimony compounds by Scopulariopsis brevicaulis. Applied Organometalic Chemistry, 12, 827-842. Andrewes, P., Kitchin, K.T., & Wallace, K. (2004) Plasmid DNA damage caused by stibine and trimethylstibine. Toxicology and Applied Pharmacology, 194, 41-48. ASTM. (1999). D5116-97. Standard Guide for SmallScale Environmental Chamber Determinations of Organic Emission s From Indoor Materials/Products. West Conshohocken, PA. American Society for Testing & Materials. ASHRAE. (1990). Ventilation for Acceptable Indoor Air Quality 62-1989. Atlanta, GA. American Society of Heating, Refriger ation and Air-Conditioning Engineers. ATSDR. (1992). Toxicological Profile for Antimony. Agency for Toxic Substances and Disease Registry, Public Health Service. Atlanta, GA. US Department of Health and Human Services.
111 Baldwin, D.R. & Marshall, W.J. (1999). Heavy metal poisoning and its laboratory investigation. Annals of Clinical Biochemistry, 36, 267-300. Baughman, A.V. & Arens, E.A. (1996). Indoor Humidity and Human Health Part I: Literature Review of Hea lth Effects of Humidity-I nfluenced Indoor Pollutants. ASHRAE Transactions. 102, Part 1. Beliles, R.P. (1994). In Clayton & Clayton, (editors) Pattys Industrial Hygiene and Toxicology. New York. Wiley-Interscience Publication. Bentley, R. & Chasteen, T.G. (2002). Microbial Methylation of Metalloids: Arsenic, Antimony, and Bismuth. Microbial and Mo lecular Biology Reviews, 66 (2), 250271. Bhat, C.G., D.E. Maier, & K.E. Ileleji. (2003). Exploratory Use of a Portable CO 2 Sensor for Early Detection of Spoila ge in a Large Corn Storage Tank. Paper No. 036149. St. Joseph, MI: ASAE. Blackwell, M. & Robbins, A. (1979). Arsine (arsenic hydride) poisoning in the workplace (8/3/79) American Industrial Hygien e Association Journal, 40. Blair, P., Fleming, P., Bensley, D., Smit h, I., Bacon, C., & Taylor, E. (1995). Plastic mattresses and sudden infant death syndrome. Lancet, 345, 720. Borjesson, T., Stollman, U. & Schnurer, J. (1992). Volatile Metabolites Produced by Six Fungal Species Compared with Other I ndicators of Fungal Growth on Cereal Grains. Applied and Environmental Microbiology, 58 (8), 2599-2605. Burge, H.A.(1995). Bioaerosols. Boca Raton, FL. Lewis Publishers, CRC Press. Burge, H.A. (1997). The Fungi: How They Grow and Their Effects on Human Health. Heating/Piping/Air Conditioning. 69-74. Burge, H.A., Pierson, D.L., Groves, T.O ., Strawn, K.F., & Mishra, S.K. (2000). Dynamics of airborne fungal populations in a large office building. Current Microbiology 40 (1), 10-16. Canadian Public Health Association. (1987). Significance of Fungi in Indoor Air: Report of a Working Group. Canadian Journal of Public Health, 78, March/April. CDC. (1994). Acute pulmonary hemorrhage/hemosider osis among infants--Cleveland, January 1993-November 1994. MMWR, 43, 881-883.
112 CDC. (1997). Update: pulmonary hemorrhage/hemosid erosis among infants--Cleveland, Ohio, 1993-1996. MMWR, 46, 33-35. CDC (1998). Surveillance for Asthma United States, 1960-1995. MMWR. 47, (SS-1). CDC. (2000). Update: Pulmonary Hemorrhage/Hemosiderosis Among Infants --Cleveland, Ohio, 1993-1996. MMWR Weekly 49(09),180-184. Chang, J.C.S., Foarde, K.K., & VanOsdell, D.W. (1995). Growth evaluation of fungi (Penicillium and Aspergillus spp.) on ceiling tiles. Atmosphere and Environment. 29: 2331-2337. Cohen, B.S. & Hering, S.V. (1995). Air Sampling Instrument s for evaluation of atmospheric contaminants. 8 th Edition. Cincinnati, OH. American Conference of Governmental Industrial Hygienists. Crandall, M.S. & Si eber, W.K. (1996a). The National Institute for Occupational Safety and Health Indoor Environmental Evaluation Experience. Part One: Building Environmental Evaluations. Applied Occupational and Environmental Hygiene, 11(6). Crandall, M.S. & Si eber, W.K. (1996b) The National Institute for Occupational Safety and Health Indoor Environmental Evalua tion Experience. Part Two: Symptom Prevalence. Applied Occupational and Environmental Hygiene 11(6). Cummings, J.B., Withers, C/R/, Moyer, N., Fairey, P., & McKendry, B. (1996). Uncontrolled Air Flow in Non-Residen tial Buildings. Final Report April 15, 1996. Cocoa, Florida. Florida Solar Energy Center.. Dales, R.E., Miller, D., & McMullen, E. (1997). Indoor Air Quality and Health: Validity and Determinants of Reported Home Dampness and Moulds. International Journal of Epidemiology, 26 (1). Deacon, J.W. (1984). Introduction to Modern Mycology. Boston, MA. Blackwell Scientific Publications. DeWolff, F.A.(1995). Antimony and Health: Incriminating stibine in the sudden infant death syndrome is difficult on current evidence. British Medical Journal, 310, 1216-1217. Dillon, H.K., Heinsohn, P.A., Miller, J.D., Eds. (1996). Field Guide for the Determination of Biological Contaminants in Environmental Samples. Fairfax, VA. American Industrial Hygiene Association.
113 Doll, S.C. (2002). Determination of Limiting Conditions for Fungal Growth in the Built Environment. (Doctoral Dissertation, Harvard School of Public Health, 2002). EPA. (1983). Antimony Metal; Antimony Trioxide; and Antimony Sulfide Response to the Interagency Testing Committee. Federal Register, 48 (4), 717-725. EPA. (1992). Indoor Air Assessment: I ndoor Biological Pollutants. EPA 600/8-91/202 U.S. Washington, D.C. Envir onmental Protection Agency. EPA. (1995). HVAC Systems as Emission Sources A ffecting Indoor Air Quality: A Critical Review. EPA-600/R-95-014. Research Triangle Park, NC. Office of Environmental Engineering and Technology Demonstration, Environmental Protection Agency. EPA. (2001). Mold Remediation in Schools and Commercial Buildings. EPA 402-K-01001. Washington, D.C. Envir onmental Protection Agency. Etzel, R.A., Balk, S.j., Bearer, C.F., Miller M.D., Shannon, M.W., Shea, K.M., Falk, H., Goldman, L.R., Miller, R.W., Rogan, W., Coven, B., & Fedeyko, H.J. (1998). Toxic Effects of Indoor Molds. American Academy of Pediatrics. April. European Collaborate Action. (1993). Biological particles in indoor environments. Report No. 12. EUR 14988 EN. Luxembourg. Office for Offi cial Publications of the EC. Ezeonu, I.M., Noble, J.A., Simons, R.B., Price, D.L., Crow, S.A., & Ahearn, D.G. (1994). Effect of relative humidity on fungal colonization of fiberglass insulation. Applied and Environmental Microbiology 60 (6): 2149-2151. Flappan, S.M., Partnoy, J., Jones, P., & Barnes, C. (1999). Infant Pulmonary Hemorrhage in a Suburban Home with Water Damage and Mold (Stachybotrys atra). Environmental Health Perspectives, 107 (11). Foster, J.W. (1949). Chemical activ ities of fungi. New York. Academic Press Publishers. Fung, F. & Hughson,W.G., (2003). Health Effects of I ndoor Fungal Bioaerosol Exposure. Applied Occupational and Envi ronmental Hygiene, 18: 535-544. Gao, P., Korley, F., Martin, J., & Chen, B.T. (2002). Determination of Unique Microbial Volatile Organic Compounds Produced by Five Aspergillus Species Commonly Found in Problem Buildings. AIHA Journal 64, 135-140. Gorny, R.L., Reponen, T., Willeke, K., Schmec hel, D., Robine, E., Boissier, M., & Grinshpun, S. (2002). Fungal Fragments as Indoor Air Biocontaminants. Applied and Environmental Microbiology, 68 (7), 3522-3531.
114 Gravesen, S., Nielsen, P.A., Iversen, R., & Nielsen, K.F. (1999). Microfungal contamination of damp buildings Exam ples of risk constructions and risk materials. Environmental Health Perspect ives, 107, Supplemental 3, 505-508. Great Lakes Chemical Corporation. ( 2003). Retrieved December 18, 2004, from http://www.e1.greatlakes.com Grleyk, H., van Fleet-Stadler, V., & Chasteen, T.G. (1997). Confirmation of the biomethylation of antimony compounds. Applied Organometalic Chemistry, 11, 471-483. Grleyk, H. (1996). Investigation of the Biologic al Reduction and Methylation of Antimony Compounds. (Masters Thesis, Sam Hous ton State University, 1996). Hodgson, M.J., Morey, P., Leung, W.Y., Morrow, L., Miller, D., Jarvis, B.B., Robbins, H., Halsey, J.F., & Storey, E. (1998). Building-Associated Pulmonary Disease from Exposure to Stachybotrys chartarum and Aspergillus versicolor. Journal of Occupational and Environm ental Medicine, 40 (3). Horner, W.E., Wortham, A.G., & Morey, P.R. (2004). Airand Dustborne Mycoflora in Houses Free of Water Damage and Fungal Growth. Applied and Environmental Microbiology, 70 (11), 6394-6400. IOM. (2000). Clearing the Air: Asthma and Indoor Air Exposure. Washington, D.C. Institute of Medicine Committee on the Assessment of Asthma and Indoor Air. National Academies Press. IOM. (2004). Damp Indoor Spaces and Health. Washington, D.C. Institute of Medicine Committee on Damp Indoor Spaces and H ealth. National Academies Press. Jenkins, R.O., Craig, P.J., Goessler, W., Mille r, D., Ostah, N., & Irgolic, K.J. (1998a). Biomethylation of inorganic antimony compounds by an aerobic fungus: Scopulariopsis brevicaulis. Environmental Science and Technology, 32:882-885. Jenkins, R.O., Craig, P.J., Goessl er, W., & Irgolic, K.J. (1998b). Biovolatilization of antimony and sudden infant death syndrome (SIDS). Human Exposure Toxicology, 17(4):231-8. Jentschke, G., Godbold, D.L., & Brandes, B. (2001). Nitrogen limitation in mycorrhizal Norway spruce (Picea abies) seedlings induced mycelial fo raging for ammonium: implications for Ca and Mg uptake. Plant and Soil, 234 (1), 109-117.
115 Johanning, E., Biagini, R., Hull, D., Morey, P., Jarvis, B., & Landsbergis, P. (1996). Health and immunology study following exposur e to toxigenic fungi (Stachybotrys chartarum) in a water-damaged office environment. International Archives of Occupational and Enviro nmental Health 68:207-218. Kendrick, B. (1985) The fifth kingdom. Waterloo, ON, Canada: Mycologue Publications. Kordula, T., Banbula, A., Macomson, J., & Travis, J. (2002). Isolation and Properties of Stachyrase A, a Chymotrypsin-Like Serine Proteinase from Stachybotrys chartarum. Infection and Immunity, 419-421. Korpi, A., Pasanen, A.L., Pasanen, P., & Kalliokoski, P. (1997) Microbial Growth and Metabolism in House Dust. International Biodeterioration & Biodegradation, 40 (1), 19-27. Krachler, M., Emons, H., & Zheng, J. (2001). Speciation of antimony for the 21 st century: promises and pitfalls. Trends in Analytical Chemistry, 20 (2). Krause, J. D., Hammad, Y.Y. & Ball, L.B.. editor: Ashley, K. (2003). Analytical Instrument Performance Criteria: App lication of a Fluorometric Method for the Detection of Mold in Indoor Environments. Applied Occupational and Environmental Hygiene, 18, 499. Lacey, J. (Editor). (1985). Trichothecenes and Other Mycotoxins. In Proceedings of the International Mycotoxin Symposium he ld in Sydney, Australia, August 1984. New York. John Wiley & Sons. Lewis FA. (1995). Regulating indoor microbes: the OSHA proposed rule on IAQ. International Conference on Fungi and Bacteria in Indoor Air Environments: October 6, 1994; Saratoga Springs, NY. Eastern New York Occupational Health Program, 5-9. Lewis R, (1997). Metals. In: J LaDou, Occupational and Environmental Medicine. Stamford, CT. Appleton and Lange, 405-39. Limerick, L. (1998). Final Repor t of Expert Group of Cot De ath. Retrieved February 12, 2005, from http://www.dh.gov.uk/PublicationsAndStatistics Little, B. & Staehle, R. (2001). Fungal Influenced Corrosion in Post-Tension Structures. The Electrochemical Society Interface, 10 (4). Lovley, D.R. (Editor). (2000). Metals and Fungi in Environmental Microbe-Metal Interactions. Washington, D.C. ASM Press.
116 Macher, J. (Editor). (1999). Bioaerosols: Assessment and Control. Cincinnati, OH. American Conference of Govern mental Industrial Hygienists. MedTox. (2004). Antimony. Medical Toxicology Unit of the Guys and St. Thomas Hospital. Retrieved February 26, 2004, from http://www.medtox.org/lab/assay.asp?ID=82 Michalke, K., Wickenheiser, E.B., Mehri ng, M., Hirner, A.V., & Hensel, R. (2000). Production of Volatile Derivatives of Meta l(loids) by Microflora Involved in Anaerobic Digestion of Sewage Sludge. Applied and Environmental Microbiology 66 (7), 27912796. Miller, M., Paloja, A., Rangger, A., Reeslev, M., & Kjoller, A. (1998). The Use of Fluorogenic Substrates To Measure Fungal Presence and Activity in Soil. Applied and Environmental Microbiology, 64 (2), 613. Montana, E., Etzel, R.A., Allan, T., Horgan, T.E., & Dearborn, D.G. (1997). Environmental Risk Factors Associated with Pediatric Idi opathic Pulmonary Hemorrhage and Hemosiderosis in a Cleveland Community. Pediatrics, 99 (1). Morey, P.R. & Williams, C.M. (1990). Porous insulation in buildings: A potential source of microorganisms. In Proceedings of Indoor Air 5 th International Conference on Indoor Air Quality and Climate. Toronto, Ontario, Canada, 529533. Morey, P.R. & Williams, C.M. (1991). Is porous insulation inside an HVAC system compatible with a healthy building? In: Proceedings of IAQ Healthy Buildings. Washington, D.C., 128-135. MSDS Trimethylstibine. (2003). Material Safety Data Sheet for Trimethylantimony, elec. Gr. (99.999%-Sb), Trimethylstibine. Newburyport, MA. Puratrem Strem Chemicals. MSDS. (1999). Material Safety Data Sheet for Crown Trade Interior Wall and Ceiling Coating. London, England. Crown Trade. MycoMeter. (2002). Handbook of the MycoMeter Test: A Guide for the Detection and Quantification of Molds in Buildings. Copenhagen, Denmark. MycoMeter ApS. NAS. (2000). Toxicological Risks of Select ed Flame-Retardant Chemicals. National Academy of Sciences and U.S. Consumer Product Safety Commission. Washington, D.C. National Academy Press,
117 Nielsen, K.F. & Madsen, J. (2000). Determination of ergosterol on mouldy building materials using isotope diluti on and gas chromatographytandem mass spectrometry. Journal of Chromatography, 227. NIOSH (1977). Analytical Method, Stibine, Backup Data Report. S243 (Set 2). Rockville, Maryland. National Institute for Occupational Safety and Health. NIOSH. (1979). NIOSH Current Intelligence Bulletin #32: Arsine (Arsenic hydride) poisoning in the workplace (8/3/79). Rockville, Maryland. National Institute for Occupational Safety and Health. NIOSH. (1994). NIOSH Manual of Analytical Methods Fourth Edition. Washington, D.C. National Institute for O ccupational Safety and Health. NIOSH. (1996). National Occupational Research Agenda. Washington, D.C. National Institute for Occupational Safety and Health. NIOSH (2002). Respiratory Protection & Mold Cleanup. Retrieved February 6, 2005 from, http://www.aiha.org/committees/html/iac/ respiratoryprotection_moldcleanu p.doc NIST. (1994). Manual for Ventilation Assessmen t in Mechanically Ventilated Commercial Buildings. Gaithersburg, MD. National Institute of Standards and Technology. NYC DOH. (2000). Guidelines on Assessment and Remediation of Fungi in Indoor Environments. New York City Department of He alth Bureau of Environmental & Occupational Disease Epidemiology. OReilly, J.T., Hagan, P., Gotts, R., & Hedge, A. (1998). Keeping Buildings Healthy. New York. John Wiley & Sons, Inc. OSHA. (2003). A Brief Guide to Mold in the Workplace. Safety and Health Information Bulletins SHIB 03-10-10. Occupational Safety and Health Administration Retrieved February 6, 2005 from, http://www.osha.gov/dts/shib/shib101003.html Ozkaynak, H., Xue, J., Spengler, J., Wallace, L., Pellizzari, E., & Jenkins, P. (1996). Personal exposure to airborne particles and metals: results from Particle TEAM Study in Riverside, CA. Journal of Exposure Analysis and Environmental Epidemiology 6(1), 57-78. Parish, G.G., Glass, R., & Kimbrough, R. (1979) Acute Arsine Poisoning in Two Workers Cleaning a Clogged Drain. Archives of Environmental Health. July/August, 224-227.
118 Parris, G.E. & Brinckman, F.E. (1976). Reactions which relate to environmental mobility of arsenic and antimony. II. Oxidation of trimethylarsine and trimethylstibine. Environmental Science Technology, 10, 1128-1134 Pasanen, A.L., Korpi, A., Kasanen, J.P., Pasanen, P. (1998). Critical Aspects on the significance of microbial volatile metabolites as indoor air pollutants. Environment International, 24(7),703-712. Persily, A.K., Nabinger, S. J., & Dols, W.S. (1994). A Study of Ventilation and Carbon Dioxide in an Office Building. ASHRAE Transactions, 100 (2). Price, D.L., Simmons, R.B., Ezeonu, I. M., Crow, S.A., & Ahearn, D.G. (1994). Colonization of fiberglass insulation used in heating, ventilation, and air conditioning systems. Journal of Industrial Microbiology, 13: 154-158. Reeslev, M., Miller, M. & Nielsen, K. F. (2003). Quantifying Mold Biomass on Gypsum Board: Comparison of Ergosterol and Beta-N-Acetylhexosaminidase as Mold Biomass Parameters. Applied and Environmental Microbiology, 3996. Rhamani, M & Cooper, D. (2004). Dispersion of a Dense Gas in a Closed Space. Paper #230. In Proceeding of the 97 th Annual Meeting of the Air & Waste Management Association. June 22-25, 2004. Air & Waste Management Association. Robbins, C.A., Swenson, L.J., Nealley, M.L., Kelman, B.J., & Gots, R.E. (2000). Health Effects of Mycotoxins in Indoor Air: A Critical Review. Applied Occupational and Environmen tal Hygiene, 15 (10), 773-784. Roos, W. & Luckner, W. (1984). Relationships Between Proton Extrusion and Fluxes of Ammonium Ions and Organic Acids in Penicillium cyclopium. Journal of General Microbiology, 130, 1007-1014. Saltzman, B.E. (1961). Preparation and Analysis of Calib rated Low Concentrations of Sixteen Toxic Gases. Analytical Chemistry, 33 (8), 1100-1112. Samimi, B.S. & Ross, K. (2003). Extent of Fungal Growth on Fiberglass Duct Liners with and without Biocides Under C hallenging Environmental Conditions. Applied Occupational and Environmental Hygiene, 18(3), 193-199. Schnurer, J. (1993). Comparison of Methods for Estima ting the Biomass of Three FoodBorne Fungi with Different Growth Patterns. Applied and Environmental Microbiology, 59 (2), 552-555. Sharma, R.P. & Salunkhe, D.K. (editors). (1991). Mycotoxins and Phytoalexins. Boca Raton, FL. CRC Press.
119 Shelton, B.G., Kirkland, K.H., Fla nders, W.D., & Morris, G.K. (2002). Profiles of Airborne Fungi in Buildings and Outdoor Environments in the United States. Applied and Environmental Microbiology, 68 (4). Thayer, J.S. (1984). Organometallic compounds and living organisms. Orlando, FL. Academic Press, Inc. Thayer, J.S. (1995). Environmental chemistry of the heavy elements: hydrido and organo compounds. Weinheim, Germany. VCH Publishers, Inc. Thompson, M. & Faull, J.L. (1995). Do microbiota cause emission of stibine from cot mattresses? Lancet, Dec 9;346(8989):1557-8. Trimethylstibine. (1991). Kirk-Othmer Encyclopedia of Chemical Technology. 4 th ed. Vol 1. New York, NY. John Wiley and Sons, 3:385. USGS. (2004). Mineral Commodity Profiles: Antimony. Open-File Report 03-019. U.S. Washington, D.C. Geological Survey. Retrieved December 18, 2004, from http://pubs.usgs.gov/ of/2003/of03-019/of03-019.pdf US Gypsum. (2003a). U.S. Gypsum Company Introduces New Gypsum Panels Designed Specifically To Resist Moisture And Mold (Press Release April 22, 2003) Chicago, IL. Retrieved February 13, 2005, from http://22.214.171.124/navigate.do?resource= /USG_Marketing_Content/usg.com/w eb_files/products/prod_details/SH EETROCK_Brand_HUMITEK_Gypsum_Panel s.htm#top US Gypsum. (2003b) MSDS for SHEETROCK Brand HUMITEK Gypsum Panels. Retrieved February 13, 2005, from http://126.96.36.199/USG_Marketing_Conten t/usg.com/web_files/Documents/M SDS/ShtrkHmtk_Gypsum_Panels-MSDS_54-040-001.pdf US Gypsum. (2005). History and Innovations. Retrieved February 13, 2005, from http://www.usg.com/about/history.jsp van Alphen, M. (1998). Paint Film Components. National Environmental Health Forum Monographs General Series No. 2. Van Loo, J.M., Robbins, C.A., Swenson, L., & Kelman, B.J. (2004). Growth of Mold on Fiberglass Insulation Building Materi als A Review of the Literature. Journal of Occupational and Envi ronmental Hygiene, 1:349-354. Vice, S. J. (2000). Indicators of Mold Growth in Indoor Environments. (Masters Thesis, Virginia Polytechnic Institute and State University, 2000).
Appendix 1: Carbon Dioxide Quality Control Data Table 11: Empty chamber #4 on 5/29/2004 Q= 0.25 L/min V= 5 L C= ppm x 1.8 g/L No test pieces in chamber Hr g/hr g hr -1 cm -2 Ambient (ppm) Chamber (ppm) Area (cm 2 ) Diff CO 2 2 -54 -0.11 402 400 500 -2 3 -54 -0.11 404 402 500 -2 4 -54 -0.11 404 402 500 -2 5 27 0.05 402 403 500 1 n=4 Mean -0.07 g hr -1 cm -2 -1.25 Figure 20: Monitoring data from an empty chamber on 5/29/2004 Empt y Chamber #4 5/29/200401002003004005006000.000.501.001.502.002.503.003.504.004.505.00Duration (hours)Carbon Dioxide (ppm)0102030405060708090100Relative Humidity CO2 ppm Ambient CO2 ppm Test Chamber %RH Ambient %RH Test Chamber 121
122 Appendix 1 (Continued) Table 12: Gypsum wallboard control CO 2 test on 5/22/2004 Hr g/hr g hr -1 cm -2 Ambient (ppm) Chamber (ppm) Area (cm 2 ) Diff CO 2 3 27 0.07 433 434 400 1 4 54 0.14 434 436 400 2 5 -27 -0.07 435 434 400 -1 6 -27 -0.07 442 441 400 -1 7 -27 -0.07 438 437 400 -1 8 -81 -0.20 445 442 400 -3 9 27 0.07 453 454 400 1 10 -54 -0.14 465 463 400 -2 11 54 0.13 456 458 400 2 12 0 0.00 439 439 400 0 13 0 0.00 428 428 400 0 14 27 0.07 421 422 400 1 15 0 0.00 415 415 400 0 n=13 Mean -0.01 g hr -1 cm -2 -0.08 Table 13: Ductboard surface coating control CO 2 test on 3/29/2004 Hr g/hr g hr -1 cm -2 Ambient (ppm) Chamber (ppm) Area (cm 2 ) Diff CO 2 9 -54 -0.11 420 418 500 -2 10 0 0.00 422 422 500 0 11 54 0.11 421 423 500 2 12 0 0.00 426 426 500 0 13 -81 -0.16 434 431 500 -3 14 0 0.00 439 439 500 0 15 108 0.22 437 441 500 4 16 54 0.11 446 448 500 2 17 243 0.49 432 441 500 9 18 216 0.43 418 426 500 8 19 81 0.16 407 410 500 3 20 162 0.32 409 415 500 6 21 54 0.11 405 407 500 2 22 27 0.05 402 403 500 1 23 0 0.00 400 400 500 0 24 54 0.11 399 401 500 2 n=16 Mean 0.11 g hr -1 cm -2 2.13
Appendix 1 (Continued) Table 14: Ductboard surface coating control CO 2 test on 4/49/2004 Hr g/hr g hr -1 cm -2 Ambient (ppm) Chamber (ppm) Area (cm 2 ) Diff CO 2 13 -81 -0.16 441 438 492.3 -3 14 378 0.77 399 413 492.3 14 15 81 0.16 396 399 492.3 3 16 27 0.05 395 396 492.3 1 17 -27 -0.05 395 394 492.3 -1 n=5 Mean 0.14 g hr -1 cm -2 2.51 Figure 21: Monitoring data from control chamber on 4/49/2004 Control Chamber Ductboard Surface Coating01002003004005006000123456789101112131415161718Duration (Hours)Carbon Dioxide (ppm)0102030405060708090100% Relative Humidity CO2 ppm Ambient CO2 ppm Control Chamber %RH Control Chamber %RH Ambient 123
124 Appendix 1 (Continued) Table 15: Ductboard surface coating control CO 2 test on 4/11/2004 Hr g/hr g hr -1 cm -2 Ambient (ppm) Chamber (ppm) Area (cm 2 ) Diff CO 2 9 -162 -0.33 416 410 492.3 -6 10 0 0.00 419 419 492.3 0 11 -135 -0.27 410 405 492.3 -5 12 -108 -0.22 404 400 492.3 -4 13 -81 -0.16 402 399 492.3 -3 14 -135 -0.27 407 402 492.3 -5 n=6 Mean -0.21 g hr -1 cm -2 -3.83 Table 16: Ductboard surface coating control CO 2 test on 4/25/2004 Hr g/hr g hr -1 cm -2 Ambient (ppm) Chamber (ppm) Area (cm 2 ) Diff CO 2 9 -54 -0.11 439 437 492.3 -2 10 -108 -0.22 442 438 492.3 -4 11 135 0.27 427 432 492.3 5 12 -54 -0.11 426 424 492.3 -2 13 108 0.22 411 415 492.3 4 14 -54 -0.11 407 405 492.3 -2 15 -54 -0.11 403 401 492.3 -2 16 -81 -0.16 400 397 492.3 -3 n=8 Mean 0.00 g hr -1 cm -2 0.00
Appendix 1 (Continued) Table 17: Ductboard FSK exterior wrap control CO 2 test on 5/2/2004 Hr g/hr g hr -1 cm -2 Ambient (ppm) Chamber (ppm) Area (cm 2 ) Diff CO 2 9 0 0.00 419 419 492.3 0 10 81 0.16 412 415 492.3 3 11 27 0.05 416 417 492.3 1 12 27 0.05 413 414 492.3 1 13 27 0.05 413 414 492.3 1 14 -54 -0.11 417 415 492.3 -2 15 27 0.05 414 415 492.3 1 16 135 0.27 405 410 492.3 5 17 -135 -0.27 424 419 492.3 -5 18 27 0.05 405 406 492.3 1 19 135 0.27 400 405 492.3 5 20 81 0.16 394 397 492.3 3 n=12 Mean 0.06 g hr -1 cm -2 1.17 Figure 22: Monitoring data from a control chamber on 5/2/2004 Control Chamber FSK w/ Adhesive 5/2/200401002003004005006000123456789101112131415161718192021Duration (Hours)Carbon Dioxide (ppm)0102030405060708090100% Relative Humidity CO2 ppm Ambient CO2 ppm Control Chamber %RH Ambient %RH Control Chamber 125
126 Appendix 2: Antimony Quality Control Data To evaluate the potential for antimony mobilization from fiberglass ductboard surface coating a control set of test materials were initially measured on 3/28/2004. The results are provided in Table 11 below. Table 18: Antimony analysis results for control test pieces on 3/28/2004 Sample ID Measured Absorbance g Sb/15 mL g Sb/mL (Analyte) g Sb/25 mL Extract g Sb/ cm2 Mean Std Dev 6 0.9381 12.2385 0.8159 305.9615 199 220.69 41.32 7 0.9465 12.3462 0.8231 308.6538 200 8 0.8286 10.8346 0.7223 270.8654 176 9 1.2136 15.7705 1.0514 394.2628 256 10 1.2926 16.7833 1.1189 419.5833 272 These same materials were later inoculat ed on 5/30 and measured for Sb content, and fungal enzyme activity, but not CO 2 production. The results are provided in Table 12 below. Table 19: Antimony analysis results for control test pieces on 5/30/2004 Sample ID Measured Absorbance g Sb/15 mL g Sb/mL (Analyte) g Sb/25 mL Extract g Sb/ cm2 Mean Std Dev 134 0.9182 10.4923 0.6995 262.3081 170 221.40 45.29 135 1.5038 17.4061 1.1604 435.1535 283 136 1.0551 12.1086 0.8072 302.7155 197 137 1.099 12.6269 0.8418 315.6730 205 138 1.3472 15.5573 1.0372 388.9315 253
127 Appendix 2 (Continued) The measurements were taken 10 weeks apart (March 28 until May 30). The mean sample concentration measured on March 28 was 220.69 g (Sb)cm -2 and 221.40 g (Sb)cm -2 on May 30. When these values are rounde d to three signifi cant digits they are identical. The sample results do not demonstrate a change in mean antimony concentration. All antimony analyses we re performed using the same calibration standards prepared on 3/28/2004. Calibration curves were determ ined before analysis of each sample set. The calibration curves are provided below in Table 13 and Figures 23 through 31.
Appendix 2 (Continued) Table 20: Antimony calibration dataCalibration Data 3-28-04 Calibration Data 4-4-04Calibration Data 4-11-04 ug Sb/15 ml Sample Measured Absorbance Predicted Absorbance ug Sb/15 ml Sample Measured Absorbance Predicted Absorbance ug Sb/15 ml Sample Measured Absorbance Predicted Absorbance 00.0049-0.016500.0030-0.002600.00620.0113 0.750.04910.04200.750.05230.05650.750.09350.0736 1.50.08880.10051.50.10510.11561.50.14480.1360 3.750.27150.27603.750.29320.29293.750.31360.3229 7.50.54510.56857.50.58440.58847.50.60530.6346 151.16071.1535151.21001.1794151.27081.2578 22.51.74171.738522.51.75101.770422.51.88101.8811 y = 0.078x 0.0165y = 0.0788x 0.0026y = 0.0831x + 0.0113 R^2 = 0.9995R^2 = 0.9994R^2 = 0.9995 x = (Abs + 0.0165)/0.078x = (Abs + 0.0026)/0.0788x = (Abs 0.0113)/0.0831 Calibration Data 4-24-04Calibration Data 5-2-04Calibration Data 5-16-04 ug Sb/15 ml Sample Measured Absorbance Predicted Absorbance ug Sb/15 ml Sample Measured Absorbance Predicted Absorbance ug Sb/15 ml Sample Measured Absorbance Predicted Absorbance 00.00580.015700.00570.029200.00680.0218 0.750.08810.07470.750.10320.09150.750.09920.0862 1.50.14860.13381.50.16740.15371.50.16640.1507 3.750.30520.31083.750.33140.34053.750.34940.3439 7.50.56660.60607.50.62860.65177.50.62800.6661 151.23981.1962151.34051.2742151.33651.3103 22.51.77151.786522.51.86011.896722.51.94771.9546 y = 0.0787x + 0.0157y = 0.083x + 0.0292y = 0.0859x + 0.0218 R^2 = 0.9984R^2 = 0.9976R^2 = 0.9991 x = (Abs 0.0157)/0.0787x = (Abs 0.0292)/0.083x = (Abs 0.0218)/0.0859 Calibration Data 5-30-04Calibration Data 6-27-04Calibration Data 7-31-04 ug Sb/15 ml Sample Measured Absorbance Predicted Absorbance ug Sb/15 ml Sample Measured Absorbance Predicted Absorbance ug Sb/15 ml Sample Measured Absorbance Predicted Absorbance 00.00770.029500.00480.020800.00510.0122 0.750.10440.09300.750.09900.08700.750.09140.0775 1.50.16840.15661.50.16250.15311.50.15930.1427 3.750.33500.34713.750.33370.35163.750.32370.3385 7.50.66980.66487.50.68680.68237.50.63080.6647 151.32071.3000151.36711.3438151.35461.3172 22.51.92171.935322.51.99102.005322.51.95581.9697 y = 0.0847x + 0.0295y = 0.0882x + 0.0208y = 0.087x + 0.0122 R^2 = 0.9995R^2 = 0.9995R^2 = 0.9989 x = (Abs 0.0295)/0.0847x = (Abs 0.0208)/0.0882x = (Abs 0.0122)/0.087 Calibration Data 3-28-04 y = 0.078x 0.0165R^2 = 0.9995 Calibration Data 4-4-04y = 0.0788x 0.0026R^2 = 0.9994 Calibration Data 4-11-04y = 0.0831x + 0.0113R^2 = 0.9995 Calibration Data 4-24-04y = 0.0787x + 0.0157R^2 = 0.9984 Calibration Data 5-2-04y = 0.083x + 0.0292R^2 = 0.9976 Calibration Data 5-16-04y = 0.0859x + 0.0218R^2 = 0.9991 Calibration Data 5-30-04y = 0.0847x + 0.0295R^2 = 0.9995 Calibration Data 6-27-04y = 0.0882x + 0.0208R^2 = 0.9995 Calibration Data 7-31-04y = 0.087x + 0.0122R^2 = 0.9989
Appendix 2 (Continued) Figure 23: Antimony calibration curve 3-28-040.0491 0.2715 0.5451 1.1607 1.7417 0.0888 0.00491 y = 0.078x 0.0165 R2 = 0.9995 x=(Abs+0.0165)/0.078 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 0510152025 ug Sb/15 mLAbsorbance @ 552 nm
Appendix 2 (Continued) Figure 24: Antimony calibration curve 4-4-040.003 0.0523 0.1051 1.751 0.2932 0.5844 1.21 y = 0.0788x 0.0026 R2 = 0.9994 x=(Abs+0.0026)/0.0788 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 0510152025 ug Sb/15 mLAbsorbance @ 552 nm
Appendix 2 (Continued) Figure 25: Antimony calibration curve 4-11-040.1448 1.2708 1.881 0.0062 0.0935 0.3136 0.6053 y = 0.0831x + 0.0113 R2 = 0.9995 x=(Abs-0.0113)/0.0831 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 0510152025 ug Sb/15 mLAbsorbance @ 552 nm
Appendix 2 (Continued) Figure 26: Antimony calibration curve 4-24-040.5666 0.0058 0.0881 0.1486 0.3052 1.7715 1.2398 y = 0.0787x + 0.0157 R2 = 0.9984 x=(Abs-0.0157)/0.0787 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 0510152025 ug Sb/15 mLAbsorbance @ 552 nm
Appendix 2 (Continued) Figure 27: Antimony calibration curve 5-2-040.0057 0.1032 0.1674 0.3314 0.6286 1.8601 1.3405 y = 0.083x + 0.0292 R2 = 0.9976 x=(Abs-0.0292)/0.083 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 0510152025 ug Sb/15 mLAbsorbance @ 552 nm
Appendix 2 (Continued) Figure 28: Antimony calibration curve 5-16-040.3494 0.628 1.9477 0.0068 0.1664 0.0992 1.3365 y = 0.0859x + 0.0218 R2 = 0.9991 x=(Abs-0.0218)/0.0859 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 0510152025 ug Sb/15 mLAbsorbance @ 552 nm
Appendix 2 (Continued) Figure 29: Antimony calibration curve 5-30-040.335 1.9217 0.0077 0.1684 0.1044 0.6698 1.3207 y = 0.0847x + 0.0295 R2 = 0.9995 x=(Abs-0.0295)/0.0847 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 0510152025 ug/15 mLAbsorbance @ 552 nm
Appendix 2 (Continued) Figure 30: Antimony calibration curve 6-27-040.1625 0.3337 0.0048 0.099 0.6868 1.991 1.3671 y = 0.0882x + 0.0208 R2 = 0.9995 x=(Abs-0.0208)/0.0882 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 0510152025 ug Sb/15 mLAbsorbance @ 552 nm
Appendix 2 (Continued) Figure 31: Antimony calibration curve 7-31-040.0051 0.0914 0.1593 0.3237 0.6308 1.9558 1.3546 y = 0.087x + 0.0122 R2 = 0.9989 x=(Abs-0.0122)/0.0870 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 0510152025 ug Sb/15 mLAbsorbance @ 552 nm
About the Author John David Krause received a Bachelor of Science Degree in Biological Sciences from Florida State University in 1989. While at Florida State he received an ROTC scholarship and the Distinguished Military Grad uate Award. He served in the U.S. Army from 1989 until 1993 in Germany. He completed active duty in 1993 and was honorably discharged with the rank of Captain. Mr. Krause served as the Florida Departme nt of Healths Indu strial Hygienist and Indoor Air Quality Programs Coordinator from 1993 to 1996. In 1997 he founded Indoor Air Solutions, an indoor air quality consulting firm. Wh ile continuing to work, Mr. Krause completed a Masters of Science in Public Health, in Toxicology at the University of South Florida, College of Public Health in 1999. While in the USF Ph.D. program he co-authored a paper in Applied Occupational and Environmental Hygiene, and presented at internati onal conferences in Helsinki, Finland and Monterey, California.
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001681040
007 cr mnu|||uuuuu
008 051220s2005 flu sbm s000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0001086
Krause, John D.
Generation of carbon dioxide and mobilization of antimony trioxide by fungal decomposition of building materials
h [electronic resource] /
by John D. Krause.
[Tampa, Fla.] :
b University of South Florida,
Thesis (Ph.D.)--University of South Florida, 2005.
Includes bibliographical references.
Text (Electronic thesis) in PDF format.
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
Title from PDF of title page.
Document formatted into pages; contains 149 pages.
ABSTRACT: Fungal contamination of buildings poses numerous challenges to researchers, building owners and occupants. Public health agencies promote prevention and remediation of mold and water damage, but lack sensitive methods to detect hidden mold growth and a complete understanding of the biological mechanisms that make occupying moldy buildings a hazard. The wide spread use of the fire retardant antimony trioxide (Sb2O3) on building materials and furnishings makes it inevitable that mold growth on treated materials will occur in some buildings with water damage. Several authors have speculation that microbial growth on materials treated with antimony trioxide could mobilize antimony through a volatile intermediate, trimethylstibine. The purpose of this study was to determine if fungal growth on a commonly used building material that contains antimony trioxide, fiberglass ductboard, results in the mobilization and release of antimony compounds.Additionally, CO2 generation rates were determined during fungal growth on fiberglass ductboard and gypsum wallboard. Results demonstrated a significant reduction of antimony concentration in fiberglass ductboard after fungal growth had occurred. Antimony emission rates and resulting concentrations of antimony oxide aerosols were estimated using an indoor mass balance mathematical model. Concentrations of CO2 were also modeled within a wall cavity and static HVAC ducts to determine if fungal growth could elevate CO2 levels above ambient concentrations. Although volatile phase antimony was not detected in chamber experiments, probably due to rapid oxidation and high humidity, mobilization of antimony trioxide from fiberglass ductboard components was demonstrated in several experiments.
Adviser: Yehia Y. Hammad.
x Public Health
t USF Electronic Theses and Dissertations.