Self-leveling mortar as a possible cause of symptoms associated with "sick building syndrome".
Modern building materials under certain conditions may be responsible for adding to the pollution load, such as the generation of formaldehyde gas. While these irritants are usually lower than levels known to induce problems in humans, there is concern that mixtures may act synergistically, and as a result produce unhealthy conditions. The term 'sick building syndrome' has been used to define a collection of symptoms including eye, skin and throat irritation, fatigue, headache, and noxious odors reported by occupants of new buildings. The present study investigates the effect of a casein, a protein additive compounded with cement to increase viscosity, as an etiologic agent of 'sick building syndrome'. Samples of casein-doped cement were examined for their ability to support bacteria. The samples were prepared and incubated under laboratory conditions and other samples of mortar were collected from buildings designated as 'sick' and from control buildings. The bacterial studies demonstrated various common bacteria in levels that were considered normal. Amines, ammonia, and sulfhydryl compounds were found in samples prepared with casein and in sample from some buildings that used this type of mortar, which were at a level known to produce the reported symptoms. However, the problem is extremely complex and many other possible contaminants, known and unknown, may be contributing to the problem. (Consumer Summary produced by Reliance Medical Information, Inc.)
Self-Leveling Mortar as a Possible Cause of Symptoms Associated with "Sick Building Syndrome"
DURING THE PAST FEW YEARS, indoor air pollutants have caused specific illnesses and many symptoms in inhabitants. [1] Materials present in modern buildings contribute to the indoor air pollution by emitting agents, e.g., formaldehyde and other volatile organic compounds. Concentrations of the individual pollutants, if identifiable, are usually below hygiene standards for working environments; lower concentrations, however, may act synergistically with other substances, and the lower limits of a dose-response relationship over time (i.e., [is greater than] 8 hr) are often poorly defined. The expression "sick building syndrome" refers to the symptoms, i.e., eye and throat irritation, fatigue, headache, skin irritation, and a bad odor [2] reported by the occupants of a dwelling or building. In newly constructed houses and buildings in Sweden where self-leveling mortar was used, inhabitants and office workers have complained of such symptoms and have, in some cases, reported so-called "allergic reactions." [3,4] These symptoms probably result from the degradation products emitted from the self-leveling mortar. The mortar usually contains proteins, e.g., casein, in concentrations up to about 4%. This component controls the viscosity during pouring.
Short-chain organic compounds with a high vapor pressure were suspected to be emitted from the mortar that contained protein. The proteins were assumed to be degraded either chemically at the initially high pH ([is greater than or equal to] 12) or by microorganisms growing in the mortar that used casein as a nutrient. Degradation of protein, i.e., casein, can give rise to several products that contain [-NH.sub.2] and/or -SH groups. Compounds that contain -SH groups are well known to cause offensive odors. Milk casein develops a bad odor after storage for 6 mo, which is due in part to volatile N-containing bases. [5] The odor threshold for trimethylamine is 0.002 ppm [6] and [is less than] 1 ppm for butylamine [7]; both chemicals are sufficiently volatile at room temperature to cause symptoms. Respiratory problems and eye irritation appear when exposure to amines in air at a level of a few ppm occurs. [8,9]
The present study investigates whether the cause of the inconveniences observed in houses where self-leveling mortar has been used is of microbiological or chemical origin and if volatile amines and other agents can be found in the mortar and in the affected houses.
Material and methods
Mortar specification. The main ingredients in self-leveling mortar are cement (25-50%), ballast (50-70%), ashes (5-20%), small amounts (0.1-4%) of casein or melamine resins (for viscosity during casting), and polymers and thickening substances (i.e., cellulose derivates, polyvinyl alcohol).
Microbiological investigations: Sampling and culture techniques. Dry self-leveling mortar powder that contained casein obtained from 7 manufacturers and 10 melamine-resin powders were analyzed for microorganisms.
Hardened mortar with and without casein and concrete samples were taken asceptically from 7 damaged and 8 undamaged control buildings at 2 to 4 different sites per room. A sample of the weight of 1-5 g material was ground in a sterile mortar and shaken in 10-ml thioglycollate broth in a vortex mixer for 15-30 s or in a Stomacher washer [10] for 5 min. Aliquots (0.1 and 0.5 ml) and, if necessary, serial dilutions were plated onto standard agar plates. In preliminary experiments, 0.1- and 0.5-ml aliquots were placed on Tryptone Soy Broth agar at pH 7.5 and 9.5 and were grown aerobically and anaerobically at +37 [degrees] C and, in most cases, also at +22 [degrees] C for up to 20 d. The number of fungi were determined on Sabouraud agar at +26 [degrees] C. Multiple samples of each mortar type were cultured.
The bacteria were identified by standard microbiological and biochemical techniques, [11-15] including gas liquid chromatography. A few strains were sent to Virginia Polytechnic Institute (Blacksburg, VA [USA] for further identification.
Survival and growth of Bacillus and Clostridia in self-leveling mortar. Twenty different strains of Clostridium and bacillus species isolated from cement and 8 strains (i.e., C. septicum NCTC 547, C. sporogenes ATCC 3584, C. putrificum ATCC 25784, C. butyricum CCUG 4217, C. novyi CCUG 1796, C. bifermentanssorelli group CCUG 9284, C. ghoni CCUG 9282, C. bifermentans CCUG 6329) were inoculated into wet sterilized mortar that contained casein and/or melamine resins. The pH was [is greater than or equal to] 12 or adjusted to approximately 8. Samples of 1 g for bacterial culture were taken during the first day and at intervals during 6-12 mos. The infected mortar and control plates were kept anaerobically at +22 [degrees] C. After 6 mo, the mortars were presented to 20 persons together with a questionnaire concerning the smell (no odor, odor, specific type of odor).
Chemical investigations: Amine and ammonia determinations. Indoor air samples, air samples from beneath floor covering material on self-leveling mortar, samples from self-leveling mortar itself, and casein samples were investigated.
Amines in air were collected in an impinger flask containing an absorbing solution of 10 ml diluted acid (usually 0.1 M [H.sub.2.SO.sub.4]) with a flow rate of 1.0 l/min for a period of 8 hr. This resulted in a sampling volume of approximately 500 l and detection limits below 0.0002 ppm for amines and 0.02 ppm for ammonia. The accuracy of this sampling procedure has been shown previously. [16,17] Casein was analyzed by direct head-space technique only.
Self-leveling mortar was either extracted with dilute [H.sup.2.SO.sup.4] or investigated by head-space technique. The sample was divided into two portions, one of which was alkalinized and transferred with a stream of nitrogen to an acidic collecting solution according to the procedure described by Chriswell and Fritz. [18] Low volatile or very water-soluble amines were analyzed after a permethylation procedure, where the amines--after derivatization--were extracted into a di-isopropyl ether solution as described earlier. [19] In some cases, self-leveling mortar was extracted with toluene.
Head-space samples were obtained by extracting an air volume of 0.1-1 ml from a 250-ml container with alkalinized mortar or casein. An alkalinized casein sample was also investigated using the gas purging method described above.
All quantitative analyses were performed by direct injection of air samples or of alkalized aqueous samples on a Varian 3700 gas chromatograph equipped with a nitrogen-sensitive detector (Varian TSD). Reference compounds were used for peak identification. Further identification was made using a Finnegan gas chromatograph-mass spectrometer. The reliability of the air sampling step, the work-up step to a permethylated amine, and the final gas chromatography (GC) analysis step has been shown previously. [17,19,20] When extracting mortar with dilute sulfuric acid, samples spiked with 1 ppm of both ethylamine and trimethylamine gave a recovery that exceeded 95%. At the low concentrations found in indoor air samples and in mortar, the precision of determinations is about 20%.
Gas chromatography-mass spectrometry (GC-MS) was used for the study of head-space samples from alkalinized sterilized casein or casein in water that had remained stationary for 3 mo. Reference spectra for different short-chain aliphatic amines, e.g., methylamine, dimethylamine, trimethylamine, and propylamine were obtained after alkalinizing acidic solutions of the reference substances contained in test tubes with screw caps and septa. Because the resolution in the GS-MS system was incomplete for the most volatile amines (i.e., methyl-, dimethyl-, and trimethylamine), mass fragmentography was used for reliable identification of these substances.
Identification of sulfur-compounds in alkalified casein. Dimethyl sulfide [(CH.sub.3).sub.2S] and dimethyl disulfide [(CH.sub.3).sub.2S.sub.2] were identified in head-space samples of the same alkalified casein sample as discussed above by use of CG-MS.
Results
Microbiological investigation: bacterial contamination of dry mortar powder and of samples from buildings. All the dry mortar powder (81 samples from 18 different types of self-leveling mortar) contained bacteria. The median level was [10.sup.2] culture-forming units (cfu)/g material (Table 1). High numbers ([10.sup.4] cfu/g) of bacteria were found in only a few samples. However, bacteria were found in 45 of 82 samples from 7 "affected" buildings ([10.sup.2] cfu/g, median value). Three samples that contained [10.sup.4] cfu/g consisted mainly of aerobic bacteria, e.g., Bacillus and Micrococcus. In 9 of 11 control samples from 8 buildings where foul smell or other damage did not occur, bacteria were found (median value, [10.sup.2] cfu/g). In 16 of 20 concrete samples from 2 damaged buildings and in 4 of 6 control samples, bacteria were found (median value, [10.sup.2] cfu/g). The number of bacteria isolated was higher on J-agar compared to TSB agar at pH 7.5 and 9.5
Species of microorganisms. Most bacteria were gram-positive spore-forming bacteria Bacillus (Table 1), e.g., B.licheniformis, B.circulans, B.sphaericus. In addition, microorganisms originating from the normal skin flora such as Micrococcus and Propionibacterium were found in most samples. In dry mortar powder, Clostridium sp. (mainly C.sporogenes, C.butyricum, C.beijerinckii) were isolated in 7 of 8 casein mortar types, but Clostridium was only found in samples from one "affected" building. No Clostridium species were found in melamine resin mortar powder. Fungi, in low numbers, were also isolated (Streptomyces, Aspergillus, Penicillium, Cladosporium, and Pullularia). No pathogens were found.
Laboratory experiments. The bacterial survival of Bacillus and Clostridium was slightly better in casein mortar adjusted to pH 7-8 than at pH [is greater than or equal to] 12. In melamine resin mortar, the reduction factor was [is greater than] [10.sup.5] cfu/g during the first 24 hr. When melamine resin mortar was supplemented with 4% casein, the reduction factor was [10.sup.1] cfu/g during the first 24 hr and [isgreater than] [10.sup.5] cfu/g after 1 week. The melamine resin mortar releases formaldehyde during hardening ([|] 800 mug/g mortar analyzed). [21] Melamine resin mortar does not, of itself, contain casein and does not form amines. No formaldehyde was found in the casein mortar ([is less than] 200 mug/g). In melamine resin mortar supplemented with casein, the values were initially low ([is less than] 200 mug/g) but increased after the first day.
The development of a bad odor was related to the presence of casein in the mortar and a high pH ([is greater than or equal to] 12). Melamine-resin mortar and casein-containing mortar, which was adjusted to a pH of 7-8 during pouring, did not develop any bad odor during the 6 mo that followed. Melamine-resin mortar supplemented with 4% casein, however, developed a bad odor, which was similar to that associated with casein mortar. There was no difference between sterile reference mortar and mortar infected with bacteria.
Chemical investigation: Amine and ammonia determination. The predominant peak in direct head-space sampling from alkalinized casein is isopropylamine, whereas the gas-purging technique reveals high concentrations of the very water soluble ethyl- and trimethylamines in the casein material. The occurrence of all these substances was confirmed by mass fragmentography. No measurable amounts of amines were obtained with casein in non-alkalinized water.
Head-space samples were also obtained from alkalinized samples of self-leveling mortar that contained casein from five affected buildings. The total concentration of short-chained amines found was in the 0.5 - 3 ppm range. No amines were found in mortar from reference buildings.
The pattern of amines in air samples from "affected" buildings is very much the same as in the head-space samples of alkalinized casein. However, the relative concentrations of the most volatile amines (methyl-, dimethyl-, ethyl-, and trimethylamine) are higher compared to long-chained amines. Concentrations of amines in different houses where casein-containing self-leveling mortar are listed in Table 2. In some cases where the peak resolution was inadequate, quantification was made assuming the same detector response for the substances considered. This is normally a good approximation using nitrogen-selective detection, where the response is mainly dependent on the nitrogen content in the molecule and, therefore, differs only slightly for similar substances. The total concentration values of amines were obtained by summing the values of amines were obtained by summing the values emanating from acid extraction, toluene extraction, and the permethylation procedure. Using the permethylation procedure, 1, 4-diaminobutane and 2-phenylethylamine could be identified using GC with nitrogen-selective detection and comparison with standards. Concentrations of these substances were [|] 0.03 ppm and 0.1-0.4 ppm, respectively.
Results from measurements of ammonia using GC techniques after optimizing the detector for ammonia sensitivity are shown in Table 3. These samples were also analyzed using Drager tubes, and an agreement within 30% was usually found. Drager tubes are less sensitive and selective compared to GC. The amine and ammonia concentrations given in Tables 2 and 3 were obtained in buildings with normal ventilation. No amines and ammonia were found in reference buildings where self-leveling casein mortar was not used. The concentration of amines in air is usually in the ppb range, whereas the concentration of same in self-leveling mortar is [|] two orders of magnitude larger. Concentrations of ammonia are much higher (Table 3). Ammonia concentrations that exceed 1 000 ppm (w/w) have been found in extracted self-leveling mortar analyzed by the gas purging technique.
Identification of sulfur compounds in alkalinized casein. Dimethyl sulfide [(Ch.sub.3).sub.2.S] and dimethyl disulfide [(CH.sub.3).sub.2.S.sub.2] was found using a head-space sample of alkalinized casein. Comparison with library compounds showed excellent agreement.
Discussion
There are several plausible explanations for the multifaceted sick building syndrome experienced by residents in newly constructed houses. In 1981-1983, air sampling in affected houses in which self-leveling mortar was used showed the presence of 2-ethyl-1-hexanol, [9] but these concentrations were not at toxic levels for humans. [22]
In this investigation, several types of amines were detected in the affected buildings. The total concentration of short-chained amines found in room air was in the 0.003-0.013 ppm range. No amines were detected in reference buildings. In laboratory experiments with hydroxide-degraded casein, several amines similar to the trace amounts found in buildings whose inhabitants reported symptoms were identified together with sulfhydryl compounds. The volatile amines, heretofore found in room air, were in the ppb range, which is below the levels accepted for working environments.
Respiratory problems have occurred after exposure to sub-ppm levels of amines in air [23]; as well, eye irritation (so-called "blue haze") has been reported [24] after 8 hr exposure to [is less than] 5 ppm triethylamine. The values found in this investigation are much lower but may, after prolonged exposure, be within the limits of dose-response values. However, very low odor threshold valuesin air were reported for volatile amines, e.g., trimethylamine, 0.0002 ppm. [6] The concentration of amines found in this study may contribute considerably to the reported sanitary inconveniences caused by a foul smell.
The concentration of ammonia in air in affected buildings was, in some cases, 2 ppm or more, and it is probable that the eye and throat irritation reported are due, in part, to ammonia exposure. Good indoor air quality in Czechoslovakia and the USSR has been assigned limit values for ammonia of 0.42 ppm and 0.28 ppm, respectively, and the 24-hr exposure value in Poland is only 0.04 ppm. [6]
One major question is whether the problems discussed above are caused by microbiological activity or by chemical degradation. It is known that amines and fatty acids may be formed by microbiological growth during protein degradation. The results show, however, that the median number of bacteria isolated from the affected buildings was only [10.sup.2] cfu/g material, which is low for an environment that is not treated aseptically, but the median value did not differ from the number isolated from reference buildings.
The bacteria isolated from the floor materials have mainly been aerobic and anaerobic gram-positive rods (i.e., Bacillus and clostridium) and bacteria found in the normal skin flora, e.g., Propionibacterium and Micrococcus. [25]
No species have been shown to dominate in the damaged material. Clostridium was isolated from most dry mortar powder samples, but in samples from houses Bacillus dominated in damaged and reference buildings. Clostridium and Bacillus are normally found in the environmental flora. Clostridium is an obligate anaerobic organism, [26] but is not very susceptible to oxygen, [27] and it can survive in air for long periods. The risk of not isolating Clostridium from the samples with the methods used here is negligible. None of the few pathogens in these families of bacteria have been isolated from mortar. There is no suspicion that microorganisms in floor mortar cause infection in persons staying in the environment. In laboratory experiments, the foul smell was associated with casein in mortar at a high pH. In melamine resin mortar that contained no casein, no odor developed within the experimental time.
In summary, no evidence of bacterial growth or bacterial degradation of protein compounds in self-leveling mortar has been found. The results presented here show the presence of chemical substance, e.g., amines, ammonia and sulfhydryl compounds in affected houses but not in control houses.
The microbiologic investigation did not demonstrate any differences between affected and control houses. Thus, it is more likely that the foul smell and other reported symptoms result from a chemical degradation of casein at an initially high pH and high humidity in the mortar. The small amounts amines, ammonia, and sulfhydryl compounds formed may, in several cases, be responsible for the foul smell and may effect the symptoms in the inhabitants. However, the situation is very complex, and all the compounds that contribute to this complex air contamination pattern remain to be identified. Further investigations are needed to fully understand this multifaceted problem.
References
[1] Gammage RB, Kaye SV, eds. Indoor air and human health. Chelsea, Michigan: Lewis, 1985.
[2] Stolwijk JAJ. The "sick building" syndrome indoor air. Vol. 1. In: Berglund B, Lindvall T, Sundell J, eds. Recent advances in the health sciences and technology, 1983.
[3] Hermansson J. Medicinsk rapport angaende Dalen, Stockholms Kommun, 1982.
[4] Steineck G. Slutrapport angaende miljoutredning, Nya Forvaltningsbyggnaden, Gallivare, 1982.
[5] golovnya RV. Investigation of organic bases in the specific odour of casein and coprecipitate during storage. Die Nahrung 1982;26:603-13.
[6] Fuselli S, Cerquiglini S, Chiacchierini E. 1973. Air pollution by methylamines. Sampling and gas chromatographic determination. La Chimical E L'Industria 1973;60:711-14.
[7] Sutton WL. Aliphatic and alicyclic amines: In: Industrial hygiene and toxicology. New York: Interscience, 1963;2037-67.
[8] Belin L, Wass U, Audunsson G, Mathiasson L. Amines: possible causative agents in the development of bronchial hyperactivity in workers manufacturing polyurethanes from isocyanates. Br J Ind Med 1983;40:251-57.
[9] Ericsson H, Hellstrom B, eds. BFR Report 193:1984. Skador i golv pa underlag av flytspacklad betong under tiden 1977-1983; 1984.
[10] Sharpe AN, Jackson AK. Stomaching: a new concept in bacteriological sample preparation. Appl Microbiol 1972;124:175-78.
[11] Holdeman LV, Cato EP, Moore WEC. Anaerobe laboratory manual. Blacksburg, VA: Virginia Polytechnic Institute and State University, 1977.
[12] Barjac H, Bonnefoi A. Essai de classification biochemique de 64 "Bacillus" de groupes II et III representant II especes differentes. Ann Inst Pasteur 1972;122:463-73.
[13] Barjac H, Cosmao-Dumanoir V. Interet de certain criteres biochemique supplementaires pour la classification des souches de Bacillus. Ann Microbiol 1975;126A:83-95.
[14] Cowan ST. Cowan and Steel's manual for the identification of medical bacteria. Cambridge, MA: Cambridge University Press, 1979 (2nd ed).
[15] Wilson GS, Miles AA, Parker MT, eds. Topley and Wilson's principles of bacteriology, virology and immunity. London: Butler and Tanner Ltd, 1983.
[16] Audunsson G, Dalene M, Jonsson JA, Lovkvist P, Mathiasson L. Analysis of amines in the industrial environment. Int J Environ Anal Chem 1985;20:85-100.
[17] Audunsson H, Jonsson JA, Mathiasson L. The efficiency of air pollutant sampling by midget impingers using amines as model substances. Am Ind Hyg Assoc J (in press).
[18] Chriswell CD, Fritz, JS. Selective concentration of amines from aqueous solutions by a gas purging technique. J Chromat 1977;36:371-77.
[19] Dalene M, Lundh T, Mathiasson L. N-per-methylation of polyamines at trace levels for gaschromatographic analysis. J Chromat 1985;322:169-76.
[20] Audunsson G, Mathiasson L. Direct analysis of free amines in salt solutions at sub-ppm levels by gas-liquid chromatography. J Chromat, 1984;315:299-312.
[21] Hanlos V. Determination of formaldehyde residuals in autoclave-sterilized materials. Arch Pharm Chem Sci 1975;5:163-69.
[22] Opdyke DLJ. Food and cosmetics toxicology. Monograph on fragrance raw materials. Special Issue V 1979;17:775-77.
[23] Brubaker RE, Muranko HJ, Smith DB, Beck GJ. Evaluation and control of a respiratory exposure to 3-(dimethylamino)propylamine. J Occup Med 1979;21:688-90.
[24] Akesson B, Floren I, Skerfving S. Visual disturbances after experimental human exposure to trimethyl amine. Br J Ind Med 1985;42:848-50.
[25] Noble WC, Somerville DA. Microbiology of the huma skin. London: WB Saunders Co., 1974.
[26] Gottschalk G. Bacterial metabolism. New York: Springer Verlag, 1979.
[27] Balows A, Hauser WJ Jr. Diagnostic procedures for bacterial, mycotic and parasitic infections. Washington, DC: American Public Health Association, 1981.
Modern building materials under certain conditions may be responsible for adding to the pollution load, such as the generation of formaldehyde gas. While these irritants are usually lower than levels known to induce problems in humans, there is concern that mixtures may act synergistically, and as a result produce unhealthy conditions. The term 'sick building syndrome' has been used to define a collection of symptoms including eye, skin and throat irritation, fatigue, headache, and noxious odors reported by occupants of new buildings. The present study investigates the effect of a casein, a protein additive compounded with cement to increase viscosity, as an etiologic agent of 'sick building syndrome'. Samples of casein-doped cement were examined for their ability to support bacteria. The samples were prepared and incubated under laboratory conditions and other samples of mortar were collected from buildings designated as 'sick' and from control buildings. The bacterial studies demonstrated various common bacteria in levels that were considered normal. Amines, ammonia, and sulfhydryl compounds were found in samples prepared with casein and in sample from some buildings that used this type of mortar, which were at a level known to produce the reported symptoms. However, the problem is extremely complex and many other possible contaminants, known and unknown, may be contributing to the problem. (Consumer Summary produced by Reliance Medical Information, Inc.)
Self-Leveling Mortar as a Possible Cause of Symptoms Associated with "Sick Building Syndrome"
DURING THE PAST FEW YEARS, indoor air pollutants have caused specific illnesses and many symptoms in inhabitants. [1] Materials present in modern buildings contribute to the indoor air pollution by emitting agents, e.g., formaldehyde and other volatile organic compounds. Concentrations of the individual pollutants, if identifiable, are usually below hygiene standards for working environments; lower concentrations, however, may act synergistically with other substances, and the lower limits of a dose-response relationship over time (i.e., [is greater than] 8 hr) are often poorly defined. The expression "sick building syndrome" refers to the symptoms, i.e., eye and throat irritation, fatigue, headache, skin irritation, and a bad odor [2] reported by the occupants of a dwelling or building. In newly constructed houses and buildings in Sweden where self-leveling mortar was used, inhabitants and office workers have complained of such symptoms and have, in some cases, reported so-called "allergic reactions." [3,4] These symptoms probably result from the degradation products emitted from the self-leveling mortar. The mortar usually contains proteins, e.g., casein, in concentrations up to about 4%. This component controls the viscosity during pouring.
Short-chain organic compounds with a high vapor pressure were suspected to be emitted from the mortar that contained protein. The proteins were assumed to be degraded either chemically at the initially high pH ([is greater than or equal to] 12) or by microorganisms growing in the mortar that used casein as a nutrient. Degradation of protein, i.e., casein, can give rise to several products that contain [-NH.sub.2] and/or -SH groups. Compounds that contain -SH groups are well known to cause offensive odors. Milk casein develops a bad odor after storage for 6 mo, which is due in part to volatile N-containing bases. [5] The odor threshold for trimethylamine is 0.002 ppm [6] and [is less than] 1 ppm for butylamine [7]; both chemicals are sufficiently volatile at room temperature to cause symptoms. Respiratory problems and eye irritation appear when exposure to amines in air at a level of a few ppm occurs. [8,9]
The present study investigates whether the cause of the inconveniences observed in houses where self-leveling mortar has been used is of microbiological or chemical origin and if volatile amines and other agents can be found in the mortar and in the affected houses.
Material and methods
Mortar specification. The main ingredients in self-leveling mortar are cement (25-50%), ballast (50-70%), ashes (5-20%), small amounts (0.1-4%) of casein or melamine resins (for viscosity during casting), and polymers and thickening substances (i.e., cellulose derivates, polyvinyl alcohol).
Microbiological investigations: Sampling and culture techniques. Dry self-leveling mortar powder that contained casein obtained from 7 manufacturers and 10 melamine-resin powders were analyzed for microorganisms.
Hardened mortar with and without casein and concrete samples were taken asceptically from 7 damaged and 8 undamaged control buildings at 2 to 4 different sites per room. A sample of the weight of 1-5 g material was ground in a sterile mortar and shaken in 10-ml thioglycollate broth in a vortex mixer for 15-30 s or in a Stomacher washer [10] for 5 min. Aliquots (0.1 and 0.5 ml) and, if necessary, serial dilutions were plated onto standard agar plates. In preliminary experiments, 0.1- and 0.5-ml aliquots were placed on Tryptone Soy Broth agar at pH 7.5 and 9.5 and were grown aerobically and anaerobically at +37 [degrees] C and, in most cases, also at +22 [degrees] C for up to 20 d. The number of fungi were determined on Sabouraud agar at +26 [degrees] C. Multiple samples of each mortar type were cultured.
The bacteria were identified by standard microbiological and biochemical techniques, [11-15] including gas liquid chromatography. A few strains were sent to Virginia Polytechnic Institute (Blacksburg, VA [USA] for further identification.
Survival and growth of Bacillus and Clostridia in self-leveling mortar. Twenty different strains of Clostridium and bacillus species isolated from cement and 8 strains (i.e., C. septicum NCTC 547, C. sporogenes ATCC 3584, C. putrificum ATCC 25784, C. butyricum CCUG 4217, C. novyi CCUG 1796, C. bifermentanssorelli group CCUG 9284, C. ghoni CCUG 9282, C. bifermentans CCUG 6329) were inoculated into wet sterilized mortar that contained casein and/or melamine resins. The pH was [is greater than or equal to] 12 or adjusted to approximately 8. Samples of 1 g for bacterial culture were taken during the first day and at intervals during 6-12 mos. The infected mortar and control plates were kept anaerobically at +22 [degrees] C. After 6 mo, the mortars were presented to 20 persons together with a questionnaire concerning the smell (no odor, odor, specific type of odor).
Chemical investigations: Amine and ammonia determinations. Indoor air samples, air samples from beneath floor covering material on self-leveling mortar, samples from self-leveling mortar itself, and casein samples were investigated.
Amines in air were collected in an impinger flask containing an absorbing solution of 10 ml diluted acid (usually 0.1 M [H.sub.2.SO.sub.4]) with a flow rate of 1.0 l/min for a period of 8 hr. This resulted in a sampling volume of approximately 500 l and detection limits below 0.0002 ppm for amines and 0.02 ppm for ammonia. The accuracy of this sampling procedure has been shown previously. [16,17] Casein was analyzed by direct head-space technique only.
Self-leveling mortar was either extracted with dilute [H.sup.2.SO.sup.4] or investigated by head-space technique. The sample was divided into two portions, one of which was alkalinized and transferred with a stream of nitrogen to an acidic collecting solution according to the procedure described by Chriswell and Fritz. [18] Low volatile or very water-soluble amines were analyzed after a permethylation procedure, where the amines--after derivatization--were extracted into a di-isopropyl ether solution as described earlier. [19] In some cases, self-leveling mortar was extracted with toluene.
Head-space samples were obtained by extracting an air volume of 0.1-1 ml from a 250-ml container with alkalinized mortar or casein. An alkalinized casein sample was also investigated using the gas purging method described above.
All quantitative analyses were performed by direct injection of air samples or of alkalized aqueous samples on a Varian 3700 gas chromatograph equipped with a nitrogen-sensitive detector (Varian TSD). Reference compounds were used for peak identification. Further identification was made using a Finnegan gas chromatograph-mass spectrometer. The reliability of the air sampling step, the work-up step to a permethylated amine, and the final gas chromatography (GC) analysis step has been shown previously. [17,19,20] When extracting mortar with dilute sulfuric acid, samples spiked with 1 ppm of both ethylamine and trimethylamine gave a recovery that exceeded 95%. At the low concentrations found in indoor air samples and in mortar, the precision of determinations is about 20%.
Gas chromatography-mass spectrometry (GC-MS) was used for the study of head-space samples from alkalinized sterilized casein or casein in water that had remained stationary for 3 mo. Reference spectra for different short-chain aliphatic amines, e.g., methylamine, dimethylamine, trimethylamine, and propylamine were obtained after alkalinizing acidic solutions of the reference substances contained in test tubes with screw caps and septa. Because the resolution in the GS-MS system was incomplete for the most volatile amines (i.e., methyl-, dimethyl-, and trimethylamine), mass fragmentography was used for reliable identification of these substances.
Identification of sulfur-compounds in alkalified casein. Dimethyl sulfide [(CH.sub.3).sub.2S] and dimethyl disulfide [(CH.sub.3).sub.2S.sub.2] were identified in head-space samples of the same alkalified casein sample as discussed above by use of CG-MS.
Results
Microbiological investigation: bacterial contamination of dry mortar powder and of samples from buildings. All the dry mortar powder (81 samples from 18 different types of self-leveling mortar) contained bacteria. The median level was [10.sup.2] culture-forming units (cfu)/g material (Table 1). High numbers ([10.sup.4] cfu/g) of bacteria were found in only a few samples. However, bacteria were found in 45 of 82 samples from 7 "affected" buildings ([10.sup.2] cfu/g, median value). Three samples that contained [10.sup.4] cfu/g consisted mainly of aerobic bacteria, e.g., Bacillus and Micrococcus. In 9 of 11 control samples from 8 buildings where foul smell or other damage did not occur, bacteria were found (median value, [10.sup.2] cfu/g). In 16 of 20 concrete samples from 2 damaged buildings and in 4 of 6 control samples, bacteria were found (median value, [10.sup.2] cfu/g). The number of bacteria isolated was higher on J-agar compared to TSB agar at pH 7.5 and 9.5
Species of microorganisms. Most bacteria were gram-positive spore-forming bacteria Bacillus (Table 1), e.g., B.licheniformis, B.circulans, B.sphaericus. In addition, microorganisms originating from the normal skin flora such as Micrococcus and Propionibacterium were found in most samples. In dry mortar powder, Clostridium sp. (mainly C.sporogenes, C.butyricum, C.beijerinckii) were isolated in 7 of 8 casein mortar types, but Clostridium was only found in samples from one "affected" building. No Clostridium species were found in melamine resin mortar powder. Fungi, in low numbers, were also isolated (Streptomyces, Aspergillus, Penicillium, Cladosporium, and Pullularia). No pathogens were found.
Laboratory experiments. The bacterial survival of Bacillus and Clostridium was slightly better in casein mortar adjusted to pH 7-8 than at pH [is greater than or equal to] 12. In melamine resin mortar, the reduction factor was [is greater than] [10.sup.5] cfu/g during the first 24 hr. When melamine resin mortar was supplemented with 4% casein, the reduction factor was [10.sup.1] cfu/g during the first 24 hr and [isgreater than] [10.sup.5] cfu/g after 1 week. The melamine resin mortar releases formaldehyde during hardening ([|] 800 mug/g mortar analyzed). [21] Melamine resin mortar does not, of itself, contain casein and does not form amines. No formaldehyde was found in the casein mortar ([is less than] 200 mug/g). In melamine resin mortar supplemented with casein, the values were initially low ([is less than] 200 mug/g) but increased after the first day.
The development of a bad odor was related to the presence of casein in the mortar and a high pH ([is greater than or equal to] 12). Melamine-resin mortar and casein-containing mortar, which was adjusted to a pH of 7-8 during pouring, did not develop any bad odor during the 6 mo that followed. Melamine-resin mortar supplemented with 4% casein, however, developed a bad odor, which was similar to that associated with casein mortar. There was no difference between sterile reference mortar and mortar infected with bacteria.
Chemical investigation: Amine and ammonia determination. The predominant peak in direct head-space sampling from alkalinized casein is isopropylamine, whereas the gas-purging technique reveals high concentrations of the very water soluble ethyl- and trimethylamines in the casein material. The occurrence of all these substances was confirmed by mass fragmentography. No measurable amounts of amines were obtained with casein in non-alkalinized water.
Head-space samples were also obtained from alkalinized samples of self-leveling mortar that contained casein from five affected buildings. The total concentration of short-chained amines found was in the 0.5 - 3 ppm range. No amines were found in mortar from reference buildings.
The pattern of amines in air samples from "affected" buildings is very much the same as in the head-space samples of alkalinized casein. However, the relative concentrations of the most volatile amines (methyl-, dimethyl-, ethyl-, and trimethylamine) are higher compared to long-chained amines. Concentrations of amines in different houses where casein-containing self-leveling mortar are listed in Table 2. In some cases where the peak resolution was inadequate, quantification was made assuming the same detector response for the substances considered. This is normally a good approximation using nitrogen-selective detection, where the response is mainly dependent on the nitrogen content in the molecule and, therefore, differs only slightly for similar substances. The total concentration values of amines were obtained by summing the values of amines were obtained by summing the values emanating from acid extraction, toluene extraction, and the permethylation procedure. Using the permethylation procedure, 1, 4-diaminobutane and 2-phenylethylamine could be identified using GC with nitrogen-selective detection and comparison with standards. Concentrations of these substances were [|] 0.03 ppm and 0.1-0.4 ppm, respectively.
Results from measurements of ammonia using GC techniques after optimizing the detector for ammonia sensitivity are shown in Table 3. These samples were also analyzed using Drager tubes, and an agreement within 30% was usually found. Drager tubes are less sensitive and selective compared to GC. The amine and ammonia concentrations given in Tables 2 and 3 were obtained in buildings with normal ventilation. No amines and ammonia were found in reference buildings where self-leveling casein mortar was not used. The concentration of amines in air is usually in the ppb range, whereas the concentration of same in self-leveling mortar is [|] two orders of magnitude larger. Concentrations of ammonia are much higher (Table 3). Ammonia concentrations that exceed 1 000 ppm (w/w) have been found in extracted self-leveling mortar analyzed by the gas purging technique.
Identification of sulfur compounds in alkalinized casein. Dimethyl sulfide [(Ch.sub.3).sub.2.S] and dimethyl disulfide [(CH.sub.3).sub.2.S.sub.2] was found using a head-space sample of alkalinized casein. Comparison with library compounds showed excellent agreement.
Discussion
There are several plausible explanations for the multifaceted sick building syndrome experienced by residents in newly constructed houses. In 1981-1983, air sampling in affected houses in which self-leveling mortar was used showed the presence of 2-ethyl-1-hexanol, [9] but these concentrations were not at toxic levels for humans. [22]
In this investigation, several types of amines were detected in the affected buildings. The total concentration of short-chained amines found in room air was in the 0.003-0.013 ppm range. No amines were detected in reference buildings. In laboratory experiments with hydroxide-degraded casein, several amines similar to the trace amounts found in buildings whose inhabitants reported symptoms were identified together with sulfhydryl compounds. The volatile amines, heretofore found in room air, were in the ppb range, which is below the levels accepted for working environments.
Respiratory problems have occurred after exposure to sub-ppm levels of amines in air [23]; as well, eye irritation (so-called "blue haze") has been reported [24] after 8 hr exposure to [is less than] 5 ppm triethylamine. The values found in this investigation are much lower but may, after prolonged exposure, be within the limits of dose-response values. However, very low odor threshold valuesin air were reported for volatile amines, e.g., trimethylamine, 0.0002 ppm. [6] The concentration of amines found in this study may contribute considerably to the reported sanitary inconveniences caused by a foul smell.
The concentration of ammonia in air in affected buildings was, in some cases, 2 ppm or more, and it is probable that the eye and throat irritation reported are due, in part, to ammonia exposure. Good indoor air quality in Czechoslovakia and the USSR has been assigned limit values for ammonia of 0.42 ppm and 0.28 ppm, respectively, and the 24-hr exposure value in Poland is only 0.04 ppm. [6]
One major question is whether the problems discussed above are caused by microbiological activity or by chemical degradation. It is known that amines and fatty acids may be formed by microbiological growth during protein degradation. The results show, however, that the median number of bacteria isolated from the affected buildings was only [10.sup.2] cfu/g material, which is low for an environment that is not treated aseptically, but the median value did not differ from the number isolated from reference buildings.
The bacteria isolated from the floor materials have mainly been aerobic and anaerobic gram-positive rods (i.e., Bacillus and clostridium) and bacteria found in the normal skin flora, e.g., Propionibacterium and Micrococcus. [25]
No species have been shown to dominate in the damaged material. Clostridium was isolated from most dry mortar powder samples, but in samples from houses Bacillus dominated in damaged and reference buildings. Clostridium and Bacillus are normally found in the environmental flora. Clostridium is an obligate anaerobic organism, [26] but is not very susceptible to oxygen, [27] and it can survive in air for long periods. The risk of not isolating Clostridium from the samples with the methods used here is negligible. None of the few pathogens in these families of bacteria have been isolated from mortar. There is no suspicion that microorganisms in floor mortar cause infection in persons staying in the environment. In laboratory experiments, the foul smell was associated with casein in mortar at a high pH. In melamine resin mortar that contained no casein, no odor developed within the experimental time.
In summary, no evidence of bacterial growth or bacterial degradation of protein compounds in self-leveling mortar has been found. The results presented here show the presence of chemical substance, e.g., amines, ammonia and sulfhydryl compounds in affected houses but not in control houses.
The microbiologic investigation did not demonstrate any differences between affected and control houses. Thus, it is more likely that the foul smell and other reported symptoms result from a chemical degradation of casein at an initially high pH and high humidity in the mortar. The small amounts amines, ammonia, and sulfhydryl compounds formed may, in several cases, be responsible for the foul smell and may effect the symptoms in the inhabitants. However, the situation is very complex, and all the compounds that contribute to this complex air contamination pattern remain to be identified. Further investigations are needed to fully understand this multifaceted problem.
References
[1] Gammage RB, Kaye SV, eds. Indoor air and human health. Chelsea, Michigan: Lewis, 1985.
[2] Stolwijk JAJ. The "sick building" syndrome indoor air. Vol. 1. In: Berglund B, Lindvall T, Sundell J, eds. Recent advances in the health sciences and technology, 1983.
[3] Hermansson J. Medicinsk rapport angaende Dalen, Stockholms Kommun, 1982.
[4] Steineck G. Slutrapport angaende miljoutredning, Nya Forvaltningsbyggnaden, Gallivare, 1982.
[5] golovnya RV. Investigation of organic bases in the specific odour of casein and coprecipitate during storage. Die Nahrung 1982;26:603-13.
[6] Fuselli S, Cerquiglini S, Chiacchierini E. 1973. Air pollution by methylamines. Sampling and gas chromatographic determination. La Chimical E L'Industria 1973;60:711-14.
[7] Sutton WL. Aliphatic and alicyclic amines: In: Industrial hygiene and toxicology. New York: Interscience, 1963;2037-67.
[8] Belin L, Wass U, Audunsson G, Mathiasson L. Amines: possible causative agents in the development of bronchial hyperactivity in workers manufacturing polyurethanes from isocyanates. Br J Ind Med 1983;40:251-57.
[9] Ericsson H, Hellstrom B, eds. BFR Report 193:1984. Skador i golv pa underlag av flytspacklad betong under tiden 1977-1983; 1984.
[10] Sharpe AN, Jackson AK. Stomaching: a new concept in bacteriological sample preparation. Appl Microbiol 1972;124:175-78.
[11] Holdeman LV, Cato EP, Moore WEC. Anaerobe laboratory manual. Blacksburg, VA: Virginia Polytechnic Institute and State University, 1977.
[12] Barjac H, Bonnefoi A. Essai de classification biochemique de 64 "Bacillus" de groupes II et III representant II especes differentes. Ann Inst Pasteur 1972;122:463-73.
[13] Barjac H, Cosmao-Dumanoir V. Interet de certain criteres biochemique supplementaires pour la classification des souches de Bacillus. Ann Microbiol 1975;126A:83-95.
[14] Cowan ST. Cowan and Steel's manual for the identification of medical bacteria. Cambridge, MA: Cambridge University Press, 1979 (2nd ed).
[15] Wilson GS, Miles AA, Parker MT, eds. Topley and Wilson's principles of bacteriology, virology and immunity. London: Butler and Tanner Ltd, 1983.
[16] Audunsson G, Dalene M, Jonsson JA, Lovkvist P, Mathiasson L. Analysis of amines in the industrial environment. Int J Environ Anal Chem 1985;20:85-100.
[17] Audunsson H, Jonsson JA, Mathiasson L. The efficiency of air pollutant sampling by midget impingers using amines as model substances. Am Ind Hyg Assoc J (in press).
[18] Chriswell CD, Fritz, JS. Selective concentration of amines from aqueous solutions by a gas purging technique. J Chromat 1977;36:371-77.
[19] Dalene M, Lundh T, Mathiasson L. N-per-methylation of polyamines at trace levels for gaschromatographic analysis. J Chromat 1985;322:169-76.
[20] Audunsson G, Mathiasson L. Direct analysis of free amines in salt solutions at sub-ppm levels by gas-liquid chromatography. J Chromat, 1984;315:299-312.
[21] Hanlos V. Determination of formaldehyde residuals in autoclave-sterilized materials. Arch Pharm Chem Sci 1975;5:163-69.
[22] Opdyke DLJ. Food and cosmetics toxicology. Monograph on fragrance raw materials. Special Issue V 1979;17:775-77.
[23] Brubaker RE, Muranko HJ, Smith DB, Beck GJ. Evaluation and control of a respiratory exposure to 3-(dimethylamino)propylamine. J Occup Med 1979;21:688-90.
[24] Akesson B, Floren I, Skerfving S. Visual disturbances after experimental human exposure to trimethyl amine. Br J Ind Med 1985;42:848-50.
[25] Noble WC, Somerville DA. Microbiology of the huma skin. London: WB Saunders Co., 1974.
[26] Gottschalk G. Bacterial metabolism. New York: Springer Verlag, 1979.
[27] Balows A, Hauser WJ Jr. Diagnostic procedures for bacterial, mycotic and parasitic infections. Washington, DC: American Public Health Association, 1981.
