EHCWEEitK* 6 UBRMW -<«. 1/1 B R.AR.Y OF THE U N IVLRSITY Of ILLI NOIS 62$ IJLBBc ENGINEERS com. ROOM B. The person charging this material is re- sponsible for its return on or before the Latest Date stamped below. Theft, mutilation, and underlining of books are reasons for disciplinary action and may result in dismissal from the University. University of Illinois Library N0V8 |V' JAN 27 rani LOW L161— O-1096 Digitized by the Internet Archive in 2013 http://archive.org/details/studiesonviabili21poon CIVIL ENGINEERING STUDIES SANITARY ENGINEERING SERIES NO. 21 STUDIES ON THE VIABILITY OF AIRBORNE BACTERIA By CALVIN P. C. POON Supported by DIVISION OF OCCUPATIONAL HEALTH U. S. PUBLIC HEALTH SERVICE RESEARCH PROJECT OH-00127 DEPARTMENT OF CIVIL ENGINEERING UNIVERSITY OF ILLINOIS URBANA, ILLINOIS MAY, 1964 STUDIES ON THE VIABILITY OF AIRBORNE BACTERIA by CALVIN PO CHUEN POON Supported by Division of Occupational Health U. S. Public Health Service Research Project OH-00127 Department of Civil Engineering University of Illinois Urbana, Illinois May, 1964 !■ STUDIES ON THE VIABILITY OF AIRBORNE BACTERIA Calvin Po Chuen Poon Department of Civil Engineering University of Illinois, 1964 It is beyond doubt that the transmission of respiratory diseases is brought about by airborne microorganisms „ In order to understand how airborne microorganisms survive in the air, the factors affecting their viability must be delineated. Such information will assist in establishing more effective methods of air disinfection. The present work involved the study of airborne Escherichia coli in two phases, namely, the investigation of the factors which had immediate as well as those with long term effects on their viability, and the investigation of the conditions under which "inactivated" airborne E coli could be reactivated. To study the significance of various factors on the viability of airborne E coli , two storage chambers were constructed „ One of them provided a storage time from one-half second to four and one-half seconds while the other one provided a storage period of several hours. Using these storage chambers, it was possible to study the immediate effects as well as the long term effects. It was observed that relative humidity (R, H,) and temperature together had an instantaneous effect on the airborne cells by evaporating pro- tein bonded water from within cells. The death rate increased in direct pro- portion to the decrease of R, H„, but increased exponentially as the temperature increased. The evaporation of water droplets expressed by Fuch°s equation gave identical results. The rate of death of airborne E, coli was found to be directly proportional to the rate of water evaporation. In comparison to the case of E, coli aerosols sprayed from distilled water, the presence of sodium chloride in the aerosol tended to increase the rate of death at low temperature and high R H„ levels because of its immediate dehydration effect on the cells c At high temperature and low R„ H« levels, however, sodium chloride tended to protect the cells from rapid water evaporation and consequently gave a lower rate of death Because of its hygroscopic property, glycerol at low concen- tration reduced the rate of water evaporation and therefore the rate of death of airborne cells . Airborne E, coli were found to be able to synthesize metabolites at room temperature and R c H„ The rate of death of E„ coli during a long storage period, as determined by comparing the number of living bacteria before and after storage, was lower because of the synthes-is of new cells Active cells, harvested in accelerating growth phase, showed a lower death rate than that of non-active cells, harvested in stationary growth phase, because the former readily utilized metabolites for growth and mult iplicat ion , In the presence of several intermediate metabolites involved in the Krebs cycle reactions, airborne E„ coli cells inactivated by "desiccation" were found partially capable of resuming multiplication at room temperature and R Ho It was believed that the reactions in the Krebs cycle were blocked in the process of desiccation „ The enzymes involved in the production of aspartic acid and glutamic acid were inactivated,, By supplying precursory metabolites such as citric, a-ketoglutaric and oxaloacetic acids, the cyclic reactions were restored as well as the ability of the cells to multiply „ The degree of disinfection of air should therefore be raised to such an extent that metabolic recovery would fail Ill ACKNOWLEDGEMENTS Much of the data in this report is taken from the experimental works supported by the National Institute of Health, U. S, Public Health Service, Research Project USPH OH-00127. The author wishes to take this opportunity to express his gratitude for being authorized to use all this data in presentation,, The author also wishes to thank particularly the following persons; Dr. I. Hayakawa, Assistant Professor of the Department of Civil Engineering, for his supervision of the experimental works, his valuable com- ments, advice and suggestions as well as his kindness in allowing the author to use his equipment and space in the Air Pollution Laboratory, without which the author's work could not have been satisfactorily accomplished. Professor R. S. Engelbrecht, of the Department of Civil Engineering, not only for offering advice and assistance in the author's experimental works and taking his valuable time to read over the paragraphs while this report was being prepared, but also for his constant advice on the academic as well as personal matters during the author's course of study at the University of Illinois. Professor B. B. Ewing, of the Department of Civil Engineering, for his advice and assistance in the course of the author's work, particularly in purchasing and handling radioactive material. The author is indebted to Mr. N. M„ Norton who helped in construct- ing much equipment necessary for the experimental works ? as well as to all Sanitary Engineering Laboratory personnel who have helped the author in many ways which led to the completion of his work. This report is a partial fulfillment of the requirements of the degree of Doctor of Philosophy in Sanitary Engineering, Department of Civil Engineering^ University of Illinois, Urbana, Illinois, IV TABLE OF CONTENTS Page ACKNOWLEDGEMENTS iii LIST OF TABLES vi LIST OF FIGURES vii I. INTRODUCTION 1 A Airborne Infection 1 B. Factors Governing the Survival of Airborne Microorganisms 2 C Organisms Investigated 3 Do Purpose of the Study 3 II. HISTORICAL REVIEW 5 A Studies on Droplet Infection 5 Bo Effects of Humidity on Viability 6 Co Effects of Other Factors 8 Do Physical Loss and Viable Decay 9 III. THEORETICAL CONSIDERATIONS 12 A. The Effect of Relative Humidity on the Viability of Airborne Bacteria 12 B„ Protein Structure and Binding Water 13 C. Denature of Protein 15 D. Theories of Water Evaporation 16 E. Evaporation of Aerosols 22 F. The Effect of Salt on the Evaporation of Aerosols 24 G„ The Effect of Growth Phases and the Presence of Growth Supporting Medium on the Viability of Airborne Bacteria 26 Ho The Restoration of the Viability of Inactivated Airborne Bacteria 27 IV. EXPERIMENTAL EQUIPMENT AND PROCEDURES 30 A. Short Storage Chamber 30 Bo Sampling Unit 38 C. Aerosol Generating Unit 39 D„ Bacterial Culture 39 E„ Experimental Procedures in Short Storage Study 40 F. Calculation 43 G. Long Storage Chamber 44 H„ Experimental Procedures in Long Storage Study 46 V. PRELIMINARY EXPERIMENTS 48 A. Size of Aerosol 48 B„ Growth of Culture with P and without P 50 C. P32 Leached Out from Cells 50 Page VI. EXPERIMENTAL RESULTS 54 A. The Rate of Death and Its Calculation in Short Storage Study 54 B. Summary of k Values in Short Storage Study 57 C. Viability of E. coli Sprayed from 5 Percent Saline Suspension 70 D. Viability of E. coli Sprayed from 0.375 Percent Glycerol Solution " 70 E. The Effect of Air Flow Rate on the Viability of Airborne E. coli 73 F. The Death Rate After the First Half Second of Storage 75 G. The Rate of Death of E. coli Aerosols in the Long Storage Study . ^, ^, 75 H. The Effect of Growth Supporting Medium and Growth Phase on the Viability of Airborne E. coli 84 I. The Reactivation of Airborne E. coli with Metabolites 86 J. The Stability of Airborne E. coli Aerosolized with Glutamic and Aspartic Acids " 90 VII. DISCUSSION OF RESULTS 92 A. General Discussion 92 B. Future Works 112 /III. CONCLUSIONS ~ 113 A. Factors Affecting the Viability of Airborne Bacteria During Short Storage 113 B. Factors Affecting the Viability of Airborne Bacteria During Long Storage 113 C. Reactivation of Airborne Bacteria 114 IX. BIBLIOGRAPHY 115 APPENDIX A: THE CALCULATION OF THE TIME PERIOD REQUIRED TO EVAPORATE A WATER DROPLET FROM 13 y DIAMETER DOWN TO 0.9 u DIAMETER 118 APPENDIX B: RELATIVE HUMIDITY DETERMINED BY WET BULB AND DRY BULB TEMPERATURE 120 APPENDIX CI: RELATIVE COUNT CURVE AND CALCULATIONS 121 APPENDIX C2: RELATIVE COUNTING RATE CURVE 122 APPENDIX D: SURVIVABILITY OF E. coli 123 VI LIST OF TABLES Table Page I The Calculation of the Average Size of Bacterial Aerosols *+9 32 II Loss of P from Cells in Aqueous Suspension at U°C and 25°C as Shown by Increased Supernatant Activity 52 III k Values of E. coli Aerosols Sprayed from Distilled Water Suspension 58 IV k Values of E. coli Aerosols Sprayed from Distilled Water Suspension 6la V The Calculation of the Function of Water Evaporation 65 VI k Values of E. coli Sprayed from 5 Percent Saline Water 71 VII k Values of E. coli Sprayed from 0.375 Percent Glycerol Solution 71 VIII Summary of k Values of E. coli Aerosols During Long Storage Period 80 IX The Effects of Growth Supporting Medium and Growth Phases on the k Values of Airborne E. coli 85 X E. coli Reactivation with 0.2 Percent Buffered Oxaloacetic Acid 87 XI The Effect of Reactivation on the Rate of Death of Airborne E. coli 89 LIST OF FIGURES vii Figure 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Short Storage Chamber Devilbiss Nebulizer No. 40 (Atomizer) Short Storage Chamber Apparatus Inlet Section of the Short Storage Chamber Apparatus Outlet Section of the Short Storage Chamber Apparatus Special Connecting Tube for Lower Rate of Flows other than 431/min. 32 32 Flow Diagram of Sampling Unit Growth Curve of E. coli in the Presence of P Long Storage Chamber Apparatus Growth Curves of E. coli in Different Concentrations of P Survival of Airborne E. coli The Change of Rate of Death with Relative Humidity at Constant Temperature ( E. coli Sprayed from Distilled Water Suspension) The Change of Rate of Death with Temperature at Constant Relative Humidity ( E. coli Sprayed from Distilled Water Suspension) The Change of Rate of Death with Relative Humidity at Constant Temperature ( E. coli Sprayed from 0.85 Percent Saline Suspension) The Change of Rate of Death with Temperature at Constant Relative Humidity ( E. coli Sprayed from 0.85 Percent Saline Suspension) The Change of Rate of Water Evaporation with Temperature The Relationship of the Rate of Death of Airborne E. coli to the Rate of Water Evaporation ( E. coli Sprayed from Distilled Water) The Relationship of the Rate of Death of Airborne E. coli to the Rate of Water Evaporation ( E. coli Sprayed from 0.85 Percent Saline Suspension) Comparison of the Relationship of the Rate of Death of Airborne E. coli to Relative Humidity from Various Sprayed Suspensions at 30°C and Varying R. H. The Relationship of the Rate of Death of Airborne E. coli to the Rate of Flow Page 31 33 34 35 35 37 38 41 45 51 55 59 60 62 63 66 68 69 72 74 Vlll Page 21 The Change of Rate of Death with Relative Humidity at Constant Temperature 76 22 The Change of Rate of Death of Airborne E. coli with Temperature at Constant R. H. 77 23 The Relationship of the Rate of Death of Airborne E . coli to the Rate of Water Evaporation 78 24 The Change of Population of Airborne E. coli with Respect to the Time for Long Storage Period 79 25 The Change of Rate of Death with Relative Humidity at Constant Temperature in Long Storage Period ( E, coli Sprayed from Distilled Water Suspension) 81 26 The Change of Rate of Death with Temperature at Constant Relative Humidity in Long Storage Period ( E. coli Sprayed from Distilled Water Suspension) 82 27 The Rate of Death of Airborne E. coli to the Rate of Water Evaporation in Long Storage Period ( E. coli Sprayed from Distilled Water Suspension) 83 28 The Comparison of the Death Rates of Airborne E. coli with and without Reactivation by Employing o 2 Percent Buffered Oxaloacetic Acid Solution ( E. coli Sprayed from Distilled Water Suspension) 88 29 The Comparison of the Death Rates of Airborne E, coli from Different Origins (in Long Storage Period) 91 30 The Comparison of the Relationship of the Rate of Death of Airborne E, coli from Various Sources to the Rate of Water Evaporation 98 I. INTRODUCTION A. Airborne Infection The belief that disease may be airborne is centuries old, but only since the growth of modern microbiological science has the medical significance of bacterial aerosols become clear. In recent decades epidemiological investi- gations of respiratory diseases such as pulmonary tuberculosis, influenza and the common cold, have established beyond doubt the dissemination of disease by airborne microorganisms and methods of air hygiene such as disinfection by ultraviolet light and by chemical agents have been applied increasingly to their control. Various bacteria, viruses and molds represent the types of organisms most commonly involved in airborne infection. A typical bacterial cell is 1 to 2 u in diameter while virus particles are all smaller than 1 V with some as small as 0.01 u ; the size of mold spores ranges from 3 y to 50 u or more. One of the most important methods of spreading certain microbial disease of man is by the expulsion of germ- laden droplets of fluid matter from the human respiratory tract. In sneezing, coughing, and even in talking, great numbers of droplets, mostly between 1 and 100 \i in diameter, of saliva and other secretions, some containing microorganisms, are expelled into the air with considerable velocities (1). The larger droplets are deposited on nearby objects or fall to the ground before they can dry, but subsequently they evaporate and their residue may be lifted again into the air, as dust, by air currents and mechanical action. Smaller droplets evaporate before reaching the ground, thus leaving any microorganisms suspended in the air, called "droplet-nuclei." Such "droplet-nuclei" may remain airborne for long periods of time and be carried long distances. Another method of spreading microbial disease of man is the indirect liberation of microorganisms into the air from secretions that have dried on handkerchiefs or bedclothes. Many investigators have found that blankets 2 and bedding can become highly infective, and air sampling in hospital wards has shown that great numbers of bacteria are dispersed from blankets, etc., during bedmaking. The effect may be equally serious in domestic houses when a member of a family is infected with a respiratory disease, e.g., tuberculosis or influenza. It is therefore evident that the mechanical transmission of airborne infection depends entirely on the ability of the bacteria to survive in the air. In order to understand the problem of airborne infection i a study of the viability of airborne bacteria seems promising. The inconsistent results of the research work to date also add to the importance and urgent need for further study. B. Factors Governing the Survival of Airborne Microorganisms Factors governing the viability of airborne bacteria are numerous. Among them are temperature, relative humidity (R.H.), particle size, the presence of toxic material, salts and growth medium. Studies in this area have been conducted over the past two decades. The effect of relative humidity and temperature on airborne cells has formed the greater part of these studies, but conclusions drawn from these studies differ widely. Some researchers have stated that the death of the airborne cell is greatest at high R.H. (4, 5), while others have stated that the greatest death occurs at intermediate R.H. levels (6, 7, 8). Some have reported, on the other hand, i that the lowest R.H. levels are most lethal (9, 10). More recently, a modern technique of using radioactive cells in aerobiological studies has provided a means to distinguish between real and apparent death due to loss of viability of the airborne cells and particulate settling (10, 11). Besides, all experi- ments which were reported by these researchers were conducted with different rates of flow in storage chambers and consequently environmental conditions were different. The situation is further complicated when airborne cells of 3 different growth phases are employed for study. Another variable concerns the nature of the suspension medium used for bacterial aerosol atomization. It is obvious that any attempt to compare the data among the workers would be meaningless if all existing factors were not taken into consideration. In order to eliminate the complicated situation, experiments in this present study were so designed that many factors were investigated separately. C. Organisms Investigated The choice of suitable bacteria for the study of airborne bacteria is largely dictated by the problem to be investigated. For practical purposes, pathogenic organisms involved in airborne infection are most appropriate for use. However, due to the lack of facilities in the laboratory for handling pathogens, it was decided to use non-pathogenic bacteria in the present study so as to eliminate any possible health hazard to the laboratory personnel. Escherichia coli was chosen for the present study. This organism has been used widely by research workers studying its viability. Old findings therefore can support in the interpretation of the experiments and permit a comparison of results. E. coli has the ability to grow in a chemically defined medium. Its growth is reasonably rapid, and it is easy to control. Another virtue of E. coli which makes it more convenient for study is that its membrane is suf- ficiently permeable (12). The metabolically active centers of the cells are in intimate contact with the environment. This permits rapid growth of E. coli cultures and pronounced effects on viability by the environment. D. Purpose of the Study It is the purpose of this study to use E. coli to investigate the effect of different factors on the viability of airborne bacteria, and, if death of the bacteria occurs, to determine the mechanism involved. Radioactive P^2 was used to label the bacteria in order to differentiate the physical loss of bacteria in the storage chamber or confined environment from the actual death of the organisms. The present study was divided into two phases: 1. The first phase utilized a short storage chamber to provide storage times ranging from one-half second to four and a half seconds. This equipment was used primarily for investigating the immediate effects of temper- ature and R.H. as well as the general characteristics of the bacterial suspension on the death of airborne bacteria. 2. There are also factors which govern the death of airborne bacteria which do not exert an immediate effect. The second phase of the invest- igation was executed in order to study the long term effect of such factors as the presence or absence of growth supporting medium in the bacterial aerosols and the phases of growth of the bacteria when used in an aerosol. A long storage chamber with storage time from minutes to a few hours was provided for this purpose. It is not the purpose of this study, however, to investigate different kinds of bacteria and to compare the results, but rather to concentrate the study on one single bacterial species.' The results, therefore, cannot be applied to all the different kinds of bacteria which might be found in the atmosphere. However, the general significance of the effects on airborne bacteria under different environmental conditions will be determined. 5 II o HISTORICAL REVIEW Ao Studies on Droplet Infection Accepting the facts that the human being is the recipient of infection and that the control of disease of man is the principal point at issue, diseases and their transmission may be classified generally as follows; lo Diseases of lower animals transmitted directly to man (rabies, tularemia, glanders, etc), 2 Diseases of animals or man transmitted by insect vectors (typhus fever, bubonic plague, spotted fever, malaria, etc), 3 Diseases of animals or man transmitted indirectly by water, milk, food and inanimate objects, 4 Diseases of man transmitted directly a, by infective droplets „ <. o ..airborne infection (the respiratory diseases and others) and bo by direct contact (the respiratory and venereal diseases in particular) o Of these modes of transmission, airborne infection is one of the most important. As early as 1933, Wells (13), as a pioneer in this field of studies, studied bacterial behavior in air Wells and his associates shed new light on the mechanism of transmission of disease 9 especially respiratory disease, from man to man by experimental work involving droplet infection (14) They considered that under the usual conditions of humidity, particles or drop- lets smaller than 0„1 mm in diameter would evaporate completely before reaching the ground, leaving suspended nuclei consisting essentially of organic matter 8 salts and bacteria, to become, for all practical purposes, a part of the atmosphere o The survival of pathogenic microorganisms in such nuclei is, as suggested by Wells, largely a matter of their resistance to drying „ Wells' (13) postulation was entirely correct, for it has been shown that some of the respiratory pathogens, including viruses such as that of 6 influenza, may remain viable, and infective for many hours (15). The study of viability of airborne bacteria became important. Since then, many factors governing their viability have been investigated (16, 17, 18, 19, 20, 21, 22, 23). For the past three decades, there has been a growing appreciation of the effect of temperature and humidity on the survival of airborne organisms. Unfortunately, the results to date have been so different and in fact are so contradictory that present knowledge neither gives us a logical conclusion nor leads us to an understanding of the mechanism involved in the death of airborne organisms* B. Effect of Humidity on Viability Early in 1935, Wells (16) suggested that the death rate of E. coli increased as the humidity approached saturation. Curves plotted from the data furnished by Wells and Riley (17) showed that the 50 percent survival time of E, coli atomized from broth was decreased from approximately 24 to 10 minutes when the R.H, was increased from 41 to 89 percent. Williamson and Gates (18) dispersed several species of organisms, Serratia marcescens , Escherichia coli , Staphylococcus albus , Staphylococcus aureus , and Streptococcus salivarius , from distilled water into a room and demonstrated that the death rate increased as the R.H. was raised from 32 to 74 percent. Edward, Elford and Laidlaw (19) noticed that the recovery of influenza virus decreased at high humidities but attributed the loss to in- creased settling. The fact that the increased settling might account for the increased disappearance of aerosols was not yet recognized in those days among most workers in this field. Therefore, Loosli et al. (20) reported that the sur- vival time of influenza virus decreased as the R.H. increased. DeOme (4) reported Salmonella pullarum 's rate of death, atomized from an aqueous suspen- sion, increased steadily with an increase in R.H. from 15 to 80 percent. 7 However, they failed to notice that with increasing R.H,, the coagulation of aerosols also increased. With greater coagulation, sedimentation increased rapidly, and the settling loss, but not the true death, therefore was largely attributed to the disappearance of viable cells.. It was because of this fact that their results were misleading and incorrectly interpreted. The very same reason also explained why Dunklin and Pick (6) reported that at intermediate levels of R.H. the death rate was the highest for many organisms. They studied the viability of Pneumococcus , type 1, and found a high mortality rate in the vicinity of 50 percent R.H. , with lower and higher R.H. values having less effect. The same effect was found on hemolytic Strep- tococcus group C and Staphylococcus , although the morality rates were smaller than that of the Pneumococcus . They concluded that at intermediate levels of R.H., the microorganisms became much more sensitive to toxic agents than at either higher or lower R.H. levels. In their studies on the effect of settling, they stated that the disappearance of microorganisms from the air was a true lethal process, rather than a manifestation of an aerosol collision process (or settling loss). However, in a figure in which they presented their data, it may be seen that the greatest settling loss (about 40 percent) occurred at 50 percent R.H. level, It is evident that the settling loss accounted for a major portion of the disappearance of microorganisms, and their assumption of no manifestation of settling loss was not correct. Furthermore, the phenomenon of highest mortality rate at intermediate R.H. levels reported by Dunklin and Pick (6) only occurred when bacterial aero- sols were sprayed from solutions containing sodium chloride o 3ecause of this fact, their findings do not have general sirrnif icance without considering the nature of the suspending medium. There are workers who have reported highest mortality of airborne organisms at low levels of R.H. Wells (9) showed that the rate of death of airborne bacteria at low R.H. was much greater than at 50 percent R.H. The 8 rapid dehydration, as suggested by Wells, might contribute to the rapid death of airborne microorganisms. The rate of evaporation of small droplets of water in unsaturated air is so great that it would seem perfectly logical to correlate it with the rapid death of bacterial cells. A study, evidently stimulated by this idea, was performed by Webb (10) from which he proposed that a breakdown of a protein-water bond was the cause for death of bacteria when they were aerosolized from distilled water suspensions „ However, Webb (10) did not study the common factors which affected both the rate of water evapo- ration and the rate of death of airborne bacteria, Co Effect of Other Factors In past investigations, temperature has been considered along with Ro H. in all studies with airborne bacteria. As pointed out by Davis and Bateman (21), such results are generally complicated by extrinsic solutes, i.e., foreign material in the bacterial suspension from which bacterial aerosols are atomized. This means that different suspending media used in generating bac- terial aerosols have a remarkable effect on the death rate of airborne bacteria, A comprehensive study has been reported by Kethley, Cown and Fincher (22) in which they revealed the significance of evaluating the nature and composition of the suspending aerosols. In their study, distilled water, beef extract broth, glycerol and a combination of all these materials were employed a It is interesting to note from their data that the aerosols, sprayed from different solutions, were very different in nature. Therefore it is easily understood why they emphasized the importance of a study of the nature of the suspending aerosols in order to properly evaluate the factors affecting the viability of airborne bacteria. A term "immediate environment" was suggested by Kethley et al. (23) as referring to the bacterial aerosol surrounded by a layer of non-living material when they are subjected to various suspensions before aerosolization. This "immediate environment" is important in all bacterial 9 aerosol studies, whether they are made to determine the effect of temperature and R.H. or whether they are concerned with the possible effectiveness of aerial disinfectants. D. Physical Loss and Viable Decay Bacterial aerosols in a storage chamber are subjected to viable decay as well as to physical loss. A failure to differentiate these two factors could lead to a significant error. In fact it has been mentioned previously that if physical loss in a storage system is neglected, analysis of the data often results in incorrect interpretations, giving a completely erroneous picture of how humidities affect the viability of airborne bacteria. This error was not recognized until Edward, Elford and Laidlaw (19) showed that the greater disappearance of influenza virus at high humidities was due to increased settling. It was Henderson (24) who first reported a technique to differentiate losses due to trauma from those of mechanical origin, i.e., any deposition of aerosols during passage from the spray atomizer to the sampling point . Since Henderson's study, a variety of techniques have been developed to detect physical loss. Henderson confirmed that physical losses can be assessed separately by spraying a suitable chemical substance in solution and then measuring the amount collected per unit volume of flow. The results so obtained were expressed in percentage of recovery of the amount of substance sprayed. He referred to this ratio as the absolute spray factor (A.S.F.). Tn estimating the A.S.F., operating conditions should be kept constant, and a stable indicator substance should be used. The dye, sky blue F.F. (color index No. 518), was found satisfactory for this purpose. However, any material such as a dye which has a different physical characteristic from that of the bac- terial cells would not settle at the same rate as the bacterial aerosols and, consequently, the physical losses would not be the same. 10 Another method for detecting physical loss was to expose agar plates at the bottom of the storage chamber to determine the bacterial settling. The settling samples were taken with respect to time so that the rate of fallout under the experimental condition could be calculated, according to Kethley, Fincher and Cown (25). Dunklin and Pick (6), instead of measuring the number of organisms, dyed the bacteria in the sprayed solution and measured the color in the col- lected sample by using a spectrophotometer. The difference of the color intensity was then correlated with the number of organisms lost due to depo- sition. All the methods described so far depend entirely on the efficiency of sampling, like any other analytical technique. This limits their use in most works. Two other methods which eliminate the disadvantage of poor efficiency of sampling have been proposed. One was described by Dimmick (26) in which a recording light-scattering photometer and electronic particle counter were attached to the storage chamber. This technique employed an improved right- angle optical system and a single electron multiplier phototube. O'Konski and Doyle (27) also used a counter photometer in their study of aerosols. The bacterial aerosol behavior is not disturbed since samples are not removed from the aerosol chamber. This light scatter decay method yields data in terms of rates, and although some of the rates are approximations, a sensitive measure of aerosol presence and behavior is continuously available. The calculation of the physical decay curve requires only a relatively short time. However, it is not applicable to small aerosol chambers where the storape time is extremely short. Another method developed only recently, is to label bacterial cells with radioactive phosphorous, P^ (10, 11). Dead tracer cells labelled with P are mixed with living non-labelled cells under study in the same suspension, 11 The radioactive tracer cells are subjected only to physical loss in the chamber. Therefore the percentage of recovery of these radioactive cells permits calcu- lation of the physical loss in the system. The loss of non-radioactive living cells under investigation would suffer the same physical loss provided that both labelled cells and non-labelled cells were thoroughly mixed in the system and their physical characteristics were not different . Theoretically, this tracer method is more direct than the light scatter decay method. Because it yields absolute values, it can be employed to determine recovery percentages and collection efficiencies upon initial aerosolization. However, this method is relatively expensive in terms of time, both in the sense of man-hours and of the waiting period before physical behavior data become available for analy- sis. Details of the technique involved in this method can be found in the works of Webb (10) and Harper et al. (11). 12 III. THEORETICAL CONSIDERATIONS A « The Effect of Relative Humidity on the Viability of Airborne Bacteria It has been confirmed that the persistence of airborne bacteria depends on their ability to survive at different relative humidity and tem- perature levels. Also, it was found recently by research workers that the lethal process is much more rapid at low R.H. levels than at higher levels. When a bacterial culture is suspended in distilled water as the spraying solution, each aerosol particle thus formed contains an individual bacterial cell or cells surrounded by a layer of pure water. Since there is no foreign material in the droplet, no extrinsic effect is expected on the viability of the bacterial cell(s), except those associated with the tempera- ture and R.H. in the storage chamber. The question has been raised concerning the osmotic pressure effect on the bacterial aerosols. The osmotic pressure is changed when bacterial cells are transferred from the culture medium to the distilled water. In other words, they are exposed to a hypotonic solution after the transfer. In general, bacterial cells suspended in distilled water do not survive more than a few hours, although spores will survive for many weeks. Death of the organisms results from a variety of factors, one or more of which has been studied in most of the reported experiments. Water from a metal still, for example, often contains sufficient traces of metal ions to be toxic, and water sterilized in soda or "soft" glass contains alkali dissolved from the glass. Water freshly distilled from hard glass is neutral, but upon standing absorbs carbon dioxide from the air and becomes acid. pH can influence the survival of bacterial cells in water. Other factors , such as the number of bacteria suspended in a given amount of water, dissolved oxygen and accessibility to oxygen, etc., are likewise known to affect the survival of these organisms. The osmotic pressure of distilled water, which might be thought of as having considerable significance, is not an important factor because bacteria are 13 remarkably resistant to changes in osmotic pressure and are not disturbed or plasmolyzed in hypotonic or hypertonic solutions. For example, Aerobacter aerogenes can grow in media with salt concentrations ranging from less than 0.1 percent to about 12 percent. In fact, except for halophilic organisms and animal cells such as red blood cells, bacteria are relatively insensitive to hypotonic solutions as a general rule. By eliminating the effect of osmotic pressure from consideration, the only factors that could cause damage to airborne bacterial cells, sprayed from distilled water suspension, are temperature and R.H. since no extrinsic effect by foreign matter is present. If the temperature in the storage chamber is kept within a range in which the bacteria can survive, this property is eliminated in concerning the effect on the viability of cells. In this study, for example, temperatures ranging from 20° C to 50° C were investigated with 60° C being the critical temperature at which E. coli would be killed by a continuous exposure of 30 minutes. In considering R.H., it must be remembered that humidity is closely allied to hydration and dehydration processes, or an exchange of water mole- cules in and out the cell membrane. Thus, R.H. and temperature, which in turn affects the R.H., could have a combined effect on the death of bacteria by influencing the water content of cells. The rate of death is governed by the rate of change of temperature and R.H. in the environment to which the bac- terial cells are exposed. The greater the rate of evaporation of water from aerosol particles, the greater the rate of bacterial death is anticipated. B. Protein Structure and Binding Water Just how the evaporation of water influences the viability of airborne bacteria requires an understanding of the structure of protein mole- cules. The primary structure of a protein molecule is the peptide chain built up by L-amino acids. Although most workers believe that there is little or no 14 binding of water by the peptide groups, the works of Mellon et al. (28, 29, 30) have suggested that the peptide groups can participate in water binding. Perhaps the best evidence of the presence of water bonded to protein molecules is a study reported by Fraser and Macrae (31). According to their work, investigations of the infra-red spectrum of collagen in the 2 u region showed that hydrated collagen contains a proportion of water molecules which are preferentially oriented with respect to the polypeptide chains. They stated: "In the spectrum, absorption band at 5150 cm" is considered attributed to sorbed water. This absorption band may be assigned to a combination of the antisymmetrical stretching and symmetrical deformation modes of the water molecules so that the associated transition moment is parallel to the line joining the two hydrogen atoms. " The preferred orientation, as suggested by Fraser and Macrae (31), was as follows: 3 O (6) with d being the density of the liquid drop, (y = d) This equation is more appropriate for use in considering small droplets and aerosols of small sizes. Another approach to the derivation of an equation of water droplet evaporation was developed by Fuchs in 1934 and reviewed by Green and Lane (37) , Fuchs considered the diffusion process as starting not directly at the surface of the evaporating sphere but at a distance of A apart from the droplet surface. In other words, the evaporation starts from the surface of an enveloping sphere of radius r + A s where A is the thickness of the enveloping layer which is of the order of the mean free path of the diffusing molecules. Very few molecules will be present in the spherical shell of thickness A which represents the distance traveled by an evaporating molecule before it collides with a gas molecule, and can be calculated from the formulas, 1 A = A m 1 + m 2 m i (7) where X is the mean free path of the evaporating molecules, m, is the mass of an air molecule, and m. is the mass of a diffusing molecule, 2 The rate of evaporation into a vacuum is 4 tt r vaC molecules per 2 second, 4 tt r is the surface area of the droplet 8 with a vapor concentration at the surface near saturation^ C , a is the evaporation coefficient , i e, 8 the fraction of the molecules striking the surface which condense , and v is a t constant which is equal to (R c T/2T»m ) with R' J being the gas constant per molecule. If the vapor concentration at a distance r + A from the center of 19 drop is C;l molecules/cm 3 , then molecules arriving at the surface which is r + A away from the center of the droplet would be evaporated at the rate of k it r 2 vct(C - C^). At equilibrium, this should be equal to the rate at which molecules leave the surface by diffusion. Therefore 4 tt r 2 vot (C Q - C x ) = 4 * (r + A) V D C x (8) where D is the diffusion coefficient in Maxwell's equation. Rearranging the equation, it becomes: C-,[(r + A) D + r 2 vo ] = C n r 2 va and, c, = . fr' 1 *. ( 9 ) 1 (r + A) D + r 2 va From Equation ( 5 ) , dm •3t = MrDC o (10) which expresses the loss of mass per second at the droplet surface into the surrounding vacuum. Substituting Equation (9) into Equation (10), with C^ replacing C Q and r + A replacing r, the loss of mass per second at the surface of the enveloping shell of thickness A becomes : *» , A , c o r2va - ar = 4 ir ( r + A) D (r + A) „ + r2 = U tt D C Q r2va(r + A) va (r + A) D + r z va = <4 it D C d™L D + r 2 va/(r + A) " U W D r C ° D/(rvo) + r/(r + A) (11) 20 When r is large in the case of large droplets at atmospheric pressure, A is comparatively very small and r/(r + A) approaches unity, whereas D/rva will be negligible compared to unity. Then Equation (10) is derived. On the other hand, for very small drops, A is large compared to r, therefore the term r/(r + A) approaches zero. Dividing both sides of Equation (11) with a unit area, the expression of the rate, of evaporation per unit area becomes: 1 dm _ M- w D C p A dT " LD/(rva ) + r/(r + A)J 4 tt r 2 i&L D/va + r 2 /(r + A) (12) When r is very small, 1 dm _ " A dT ' va C ° and this is identical to the rate of evaporation into a vacuum as it has been defined previously. Therefore the assumption that the evaporation is propor- tional to the surface area of droplet is verified. Equation (12) can be transformed to give the rate of change of surface with time similar to the form expressed in Equation (6), dA 8 tt D Cp_ dt ' d * D/rva + r/(r + A) , 8 7T D M f v 1 R T d (P ° " c OS bC i—l 2s >» 5? /— s H s CD CD o <-H >• co ft O g s o in CU 5 -c >H C5 1-4 o B 4-> a> u. fe •i-t 4-» ♦-» 2 CP E t^ o •a CO » CU •— t <^ >«^ 4-> a u X X X o • o t^) V >-. »* c U CM CO >, S>> s o eo 3 u CU CU o o ca v.. CO x> N A •> 4-» • •c JS o i— 1 o X 3 • CO •1-4 •«-4 «•> o H »-) U CO CO 4-> < J-. u. CU CO >> » s A +-i *-* •»H E CU c CM • a V) o o o u A *-> «* • g ©> «M <■> o *4 M ta V4 u CO u B a V { CD CD »~4 o •*>4 CO •»* c J CM 4J ♦J a. c *-4 ^ c E o rt; o CO CO u -c • o o o •r4 ♦■» 4>> o tO U o E E 5 C/i e u (0 • Jatf J-i o V CO a O v> § 4) CD 0) I-. u < 1—4 • o» 4-> J5 J= >— » 4) O M (0 » -C o +J o X) JQ J8 3 0) <0 (0 U •M 8 4-* f^ 1—4 CQ 02 3= •H f 5 V) V) 3 ca 3 CO CO CM U ft fl S v CO to 0) •M 3 V •rH c >> «-> CO • 4J O O lb s £ z o a 0) 3t O as CO SB *2 > b y < co u a td tl. U tC M > 32 size and sealed with sealent to eliminate leakage completely. When conducting experiments, joint A was connected to an atomizer, Figure 2, and the outlet end of the main storage chamber was connected to a flowmeter, recirculation pump, bacterial filter, dehumidifier , in succession, back to the joint B, Figures 3, 4, and 5„ A Fisher laboratory flowmeter, with tapered graduated tube No. 2L-150 and a 0.25 inch stainless steel float, was used in all systems having air flows less than 38 l/min„ For air flow of 43 1/min., another Fisher flowmeter was used^. Following the flowmeter was a Gelman vacuum-pressure pump, Model 13152, of the single piston positive displacement type equipped with self lubricating hard graphite piston rings and skirts. After passing through the Gelman pump, the flow in the system was divided into two parts; one part was exhausted, and the remaining portion was recirculated back into the storage chamber after passing through a control valve, a bacterial filter and a dehumidifier. The exhaust line was connected to a gas washing bottle containing a strong sodium hydroxide solution which was used to retain the R.A. cells in the air flow. This was followed immediately by a glass wool filter. The gas flow passed out from the filter into a fume hood with a stack leading to the roof of the laboratory. The bacterial filter was a Matheson 410 brass line filter 4, with a membrane filter of 0.45 u pore size added. It effectively filtered out all the bacterial aerosol from the gas stream so that clean air was ensured before it entered into the storage chamber again. The dehumidifier was simply a 2. RTV 102 Silicone Rubber Sealent, General Electric. 3. Flowmeter, Pyrex brand glass with hollow stopper mounted with four orifices, measuring approximately 0.25, 0.5, 1 and 2 mm in diameter. The flowmeter was mounted on a stand, filled with mercury and was calibrated with a wet test meter of Precision Scientific Company, Chicago Model 63123. *+. The Matheson Co., Inc., Joliet, Illinois. 33 A. Air Inlet B. Reservoir for Bacterial Suspension C. Spraying Nozzle D. Rubber Stopper (Opening for Supplying Bacterial Suspension) E. Aerosol Outlet FIGURE 2. DEVILBISS NEBULIZER NO. 40 (ATOMIZER) 34 «9 W CO CD Oj c/i 3 CO O E 3 U o co <-n > X X t-i -o CD V 0) c w u ta CO >> >» •iM a. a. E O CD > o u •» •> +» r— ( 3 JS • < 1-4 o CO > CO CO X 3 • *■• M cd o H ►h ♦J f-4 u M o B O Cu >» D> s rr o ^ Ou c CM • u 4-> C ♦J •» ■(-> •» o 2 o •fH 4-> o |M z o o 3E s 0) 1— 1 •*■< c -Q >- u. ^■N a, o c E o E »-3 o CO N v< J-l 4J 3 o JS •i-4 cy o CO 0. co o r- ( 4-> CD 3 o Cs. > X T4 a C5 ►H ce Z a. •i-< ft dj ♦J CO u >. XS o CO CD •»-t CO > <»> r-H CD CO CO u +-» M-t U ■a •l-H 3 *J Oi 4-> T e t/5 1— 1 3 •rt - *■* CD v> O CD c co HH 1— t CO E o i— l a> )H £5 •l-t J- u •a •ft CO 3 +-> T3 »— i •»H 2 ■M CD 3 0) •rt > CD JS o 4) «-> o O CO s o <§ £ CD a. CO 00 J £ CD i-H I 4-30 5 s s gas I ts S S \ \ V .53* ' -3-0*'- -227- ./•«r. /v?7- n 1 3ZS3 i it: jrv^ =JJ p vs>>>J5Z5 J jo B ycc^ r\ s ggs^s ^ . 3X A, Outlet JEhd of Four-way Glass Joint, Pyrex B' . 7/8" I.D. Connecting Tube with Five Sampling Outlets of 0.39" 0.0. , Lucite C. Main Storage Chamber G. Glass Baffle FIGURE 6 SPECIAL CONNECTING TUBE FOU LOWER ilATE OF FLOWS OTHER THAN 43l/rain. 38 B. Sampling Unit There are two methods which may be used for sampling bacterial aerosols with a radioactive tracer. One method is to draw the sample onto an agar plate for determining the viable count while a separate sample is drawn onto membrane filter for radioactive count. One disadvantage of this method is that the aerosols may escape deposition on the agar surface which, as a result, gives a lower sampling efficiency. Another method is to draw the sample into an impinger containing collecting fluid. After sampling, the col- lecting fluid can be used for making a viable count as well as radioactive count. A higher sampling efficiency is expected by using the impinger method provided that the flow is low enough. Although sampling efficiency was not of much concern in the radioactive tracer technique used in the present study, impingers were used exclusively for sampling. n All glass, graduated midget impingers were used with 10 ml of 0.85 percent saline water as collecting fluid. A Gast vacuum pump** was used to draw an air sample from the short storage chamber at a rate of 2.6 1/min which was controlled by a needle valve on the Matheson flowmeter . This equip- ment was arranged in a series as shown in the following flow diagram; From Sampling Outlet -*— A. B. C. D. Impinger with 10 ml of collection fluid Matheson Flowmeter with Control Valve Bacterial Filter Gast Vacuum Pump FIGURE 7: FLOW DIAGRAM OF THE SAMPLING UNIT 7. MSA all glass midget impinger for dust sampling. 8. Gast 0211-V36 pump, Gast Manufacturing Corp., Benton Harbor, Michigan, 9. Matheson Model 603, Tube A-2-15-B, Stainless steel float. 39 C. Aerosol Generating Unit A Nebulizer No. 40 atomizer , Figure 2, was used for generating bacterial aerosols, along with a Devilbiss Compressor 501 equipped with a bacterial filter attached to the inlet side of the compressor so as to ensure a clean air supply. The compressor had a screw adjustment so that it could be controlled for a desired pressure. This provided a definite rate of flow for the atomization operation. The rate of flow was found to be 4.5 1/min., at a pressure of 2.5 psig. When this atomizer unit was connected to a Precision wet test meter, it was found that it took on the average 59.5 minutes to com- pletely spray 6 ml of bacterial solution in the atomizer. This rate of atomi- zation, 6/59,5 = 0.101 ml/min, was used throughout all experiments in the short storage study. D. Bacterial Culture A pure culture of E. coli was obtained from the Department of Microbiology, University of Illinois. A stock culture was maintained by trans- ferring the culture to a new tube of nutrient agar every month. A chemically defined medium was used for culturing E. coli prior to all experiments. This medium was one of the few defined media which have been employed in biosynthesis studies using E. coli (12): NH 4 C1 2.0 g Na 2 HP04 0.6 g KH 2 P0 4 0.3 g NaCl 5.0 g Mg (as MgCl2> 0.01 g S (as Na 2 S04) 0.026 g Glucose 1.0 g (in 100 ml dist. water) Distilled water 900 ml P 32 H (as H 3 P0 4 in HC1 solution) 10 mc 10. The Devilbiss Co., Summerset, Pennsylvania. 40 32 32 Isotope P (P as H PO in weak HC1 solution) was obtained from Oak Ridge National Laboratory „ Each shipment contained about 10 mc at the time of arrival and was diluted with distilled water into a 25 ml stock solution,, To prepare the medium, 100 ml of the defined medium was placed in a 200 ml Erlen- meyer flask „ The flask was innoculated with E„ coli followed by 1 mc of 32 P . The culture was grown at 37° C» Two growth curves of a typical E coli culture, plotting time of growth against percentage of light transmittance of the culture measured at 560 mp X by a spectrophotometer , are shown in Figure 8 This figure indi- cates that the same growth curve can be reproduced provided that the initial population is within the limits in the range under invest igat ion „ It can be seen from Figure 8 that the logarithmic growth phase occurred from after 8 to 12 hours of growth, and that the culture entered the stationary growth phase after 20 hours of growth „ All experiments in the short storage chamber were conducted with cultures harvested after they had reached the stationary growth phase, or, in other words, cultural suspensions for aerosol generating were made from E„ coli cultures after 20 to 26 hours of growth „ E» Experimental Procedures in Short Storage Study In initiating each experiment, an E„ coli culture was harvested and washed three times by repeated centrifugation and resuspension in solution in the following steps „ The culture from a 200 ml Erlenmeyer flask was trans- ferred to 50 ml capacity centrifuge tubes „ After centrifuging at 8000 rpm in 12 a Servall centrifuge for 10 minutes, the supernatant was decanted „ The 11 o Coleman Model 14 Universal Spectrophotometer „ 12 o Weston Electrical Instrument Corp , Model 1332 41 CM 0) en CO h- CM in O o o to © CM o o o c • • • • • • • • • • • • 3 o u S s CO co s <* ^ g 8 8 lO CO 8 •—4 • O w o sO •o in o o LO or> >© to CO to A > B • • • • • • o • • • • W *< (0 •o If) to TT 18 10 o to to to 3 to a M o o o r* "3» ^ CO co CO CO o 0) • to " * s >H o I-* ^T o »-« co to o ^•1 CM ^r o •r* JC i-H 1— 1 1— t 1— t CM CM CM CM H COl a, o to «>4 4 \ *0 *s 1 s 4 (l ^ y \ o V * s, •N ^s r» to •-0 A i 4> 'i I 3 fr O c0 w u s o EC E efieiuaoiaj 'aoue:u|uisuB.ii m fi n cells were resuspended in one of the following solutions 9 depending on which bacterial suspension was going to be studied in the aerosol systems distilled water; 0„85 percent saline water; 5 percent saline water or 0.375 percent glycerol solution. The same procedure was repeated three additional times. After these washings, the supernatant usually gave radioactive counts below 32 10 cpm per ml. It was assumed that all the radioactive P in the final bacterial suspension was retained inside the cells. The cells were resus- pended in 10 ml of the appropriate solution. This final suspension was mixed thoroughly and 4 ml were placed in the atomizer so that the formation of clumps of bacteria was reduced in the spraying suspension. One-tenth ml of this suspension was removed and serially diluted in a saline solution. The viable count was determined by the drop plate method using EMB agar. One milliliter of the same suspension was placed in a 2-inch diameter aluminum planchet with concentric rings. The R. A, count on this suspension, expressed as cpm per milliliter, was obtained by placing this 13 planchet with the dried aliquot in a proportional counter and counting it for 10 minutes. The atomizer was then connected to the short storage chamber and the pressure pump was turned on to start generating the bacterial aerosol. The recirculation pump and the moisture control air supply were turned on at the same time. The flow in the system was regulated by operating the needle valve. The R. H. was regulated by adjusting the hosecocks on the dry and wet air lines, while the temperature was regulated by controlling the powerstat connected to the heating tapes. After a steady state condition had been reached, constant temperature and R, H. at the desired level, air samples from the storage chamber were taken with the sampling unit from the various outlets. All samples, collected in the impingers, were serially diluted, and the viable 13. NMC internal proportional counter, Model DS-1A 9 Indianapolis, 43 count as well as the R„ A„ count were determined in the same manner as in the spraying suspension 9 zero time sample „ After all samples had been taken, the temperature and R„ H control, the compressor, and the recirculation pump were turned off The atomizer was detached and replaced with another atomizer containing M- ml of formaldehyde, 37 percent HCH0„ This formaldehyde solution was sprayed with the recirculation pump operating for 5 minutes so as to disinfect the system „ Then everything was turned off The system was completely closed and let stand for about one hour» Recirculation was initiated again only with wet air supply so as to remove the excess formaldehyde in the system,, This was necessary so that the viability of the bacterial aerosols used in the following experiment would not be affected o The EMB agar plates were stored at 37°C for 24 hours and viable bacteria were counted with the aid of a Quebec colony counter,, Planchets containing 1 ml of sample were dried in an oven and were counted in an internal proportional counter „ F„ Calculation If the number of viable cells per 0.1 ml of spray suspension is equal to V , the R A„ count per minute per CI ml of spray suspension is V (Ro A ) , the ratio j^ 9. . - -. represents the number of viable cells per unit O v K A /_ radioactive count „ The total cells regardless of their viability in each of the samples could be calculated as follows; where T = total cells in Od ml of sample (collecting fluid) with storage time t, and 44 (R. A.) = Ro A. counts per minute per 0.1 ml of collecting fluid. Because a portion of the bacteria died off during storage time t, the total cells T were composed of living and dead cells. The number of viable cells per 0.1 ml of the collecting fluid can be determined by drop plate counts, as V . The difference between T and V was the actual death of bacteria. V t V — - x 100 was the percentage of survival after storage time t, or 100 - =-— x 100 4 t l t was the percentage of death of bacteria after storage time t. G„ Long Storage Chamber The chamber used in the long storage studies was made of a 5,56 inch 14 inside diameter Lucite tube, 4 feet long. A Serdex humidity indicator , a thermometer and a manometer were attached to the chamber as indicated in Figure 9, Bacterial aerosols were sprayed from an atomizer located on the upper part of the chamber. The atomizer was operated together with a Devilbiss 15 Compressor 501. The bacteria laden air was recirculated by a Dynapump pump through a 0,5 inch diameter Pyrex glass tube back into the main storage chamber at the bottom. The rate of recirculation was 3,7 1/min. The flow coming in from the bottom part of the chamber was distributed through a baffle with 23 holes of one-half inch diameter. Three sampling outlets, each of 3/8 inch diameter, were located along one side of the chamber at various elevations, 30, 10, and 3 inches from the bottom, A R. H. control line was attached to the recirculation line ahead of the pump, in order to supply dry or moist air to the storage chamber when correction of the R. H, in the storage chamber was necessary. 14. Serdex Relative Humidity Indicator Model 201, Bacharach Industrial Instrument Co., Pittsburgh, Pennsylvania. 15. Pressure-Vacuum Dynapump, Model 12-3097, Fisher Scientific Company, Chicago, Illinois. 45 A, B, C, D, E, F, G< H. I, h K. L, M, N. 0. Devilbiss Compressor 501 Devilbiss Nebulizer No. 40 (Atomizer) Main Storage Chamber Serdex Relative Humidity Indicator Recirculation Line Thermometer Exhaust Line Relative Humidity Control Gas Flow Recirculation Pump Powerstat (heat control) Bacterial Filter Vacuum Pressure Pump Hosecock Flowmeter Sampling Outlets Manometer FIGURE 9 LONG STORAGE CHAMBER APPARATUS 46 When a sample was being removed from the storage chamber, a clean air supply was introduced into the chamber at the same rate as the aerosol sampling so as to replenish the volume sampled. This was made possible by using a vacuum pressure pump (Gelman Model 13152) with a bacterial filter attached to the inlet side. The clean air thus obtained went through a Fisher flowmeter into the chamber with its flow rate regulated by a hosecock in accordance with the rate of sampling, 2 1/min. The temperature control unit and the sampling unit were the same instruments used in the short storage chamber „ Ho Experimental Procedures in Long Storage Study An Eo coli culture was harvested and washed three times by repeated centrifugation and resuspension in solution using the same procedures followed in the short storage study. Distilled water, 2.5 u mole glucose solution, cultural medium, and various metabolites of the tricarboxylic acid cycle were used for washing and final preparation of the suspensions for aerosols genera= tion. The storage chamber was purged with cleaned dry air to reduce the R« H„ down to approximately 20 percent. Temperature was adjusted to a constant value before the aerosols were introduced. When aerosols were sprayed into the chamber, the recirculation pump was turned on and the exhaust line was opened o The concentration of bacteria in the chamber increased as the R„ H increased. When the R. H. had reached the desired level, the aerosol supply was stopped and the exhaust line was closed. A sample was taken immediately as the zero time sample. Subsequent, hourly samples were taken for four hours total storage time. The settlement loss in the system with continuous recir= culation was found to be very great. Therefore, the recirculation pump was 47 turned on every five minutes and the air circulated for 30 seconds „ Through this procedure, the loss due to settlement was reduced so that sufficient bacteria were found in the last sample for both viable and radioactive counts „ Due to the decreasing concentration of bacteria in the chamber air, increasing amounts of sampling volume, 2/3, 2, 3, 5, and 10 liters of air were drawn from the chamber for the time, 1, 2, 3, and 4 hour samples respectively. The increasing sampled volume provided a measurable number of bacteria for both viable and R A counts for each sample „ Eo coli aerosols sprayed from various solutions were sampled at a flow rate of 2 l/min using impingers, each containing 5 ml of 85 percent sodium chloride solution « Following collection, these samples were plated immediately o In order to determine the effect of recovery of activation and growth, different buffered metabolite solutions were used instead of the isotonic saline solution. Samples taken in such metabolite solutions were plated and then incubated for 24 hours at 37°C before they were plated again on EMB agar The 24 hour incubation period was found necessary for the inactivated cells to become activated to the metabolites and to resume their growth ability o After all samples had been taken, the system was cleaned with a formaldehyde mist, followed by wet air 9 as was practiced in the short storage study o Calculation of the rate of death was made in much the same way as in the short storage study „ 48 V„ PRELIMINARY EXPERIMENTS A, Size of Aerosol Sizes of aerosols sprayed from an atomizer depend on a number of factors, such as the type of atomizer used, salt concentration of the generating solution (41, 42), and jet location relative to the surface of the generating solution in the atomizer (43), In the present study, a bacterial suspension 9 having approximately 2,0 x 10 cells per ml was prepared, and aerosols were generated with a Devilbiss Nebulizer No. 40 under 2,5 psig and 4,5 liters per minute of air flow. The average size of aerosol particles generated from this atomizer under the specified condition was determined indirectly by dividing the volume of suspension sprayed by the number of aerosol particles generated from this volume of sprayed suspension, A simple experiment was conducted by spraying the aerosols on a clean glass slide held approximately 3 inches away from the outlet of the atomizer for a few seconds , The number of bacteria in each aerosol particle was counted under a microscope and the average number of bacteria in each aerosol particle for all observations was determined. By dividing this average number with the concentration of the sprayed suspension, the volume of an average aerosol particle could be calculated. The results are given in Table I. As the results show, the average size of aerosols increased with the concentration of bacteria of the sprayed suspension. The increase is not of particular significance from the practical standpoint, because bacterial aerosols under natural conditions have a greater variation of sizes. The results also show that the atomizer does not yield monodispersed aerosols. lo E, Leitz Wetzlar microscope, with magnification x430, 49 to o CO S3 < a Si « Cm O N M M CO 3 a B o a o M H < o < C_> w EC H H o 4- 0) H r» X H o> o in + H o o en X CN CN V. in in o H CN en I o CO CN cn o 3- CM CO O CO CO o CN 3. CO G hO J2 H O > H ■P O to •H o < O •H W H to bO to » O rH c > CO o fc CO p CO < sx u CO CO O H +J ex +-» CO 4-> < O o to c < C 4h to m 3 m 4n 3 G O O « CO o c o m U 4n rfl o o co •4H c o fc a CO to to to to o < O •H rH CO »4h •+H -H f-i >-i rH rH rH £5 Q> O •P O o H rH H rH H c o to e to co H CI) CO CO CO CO o nj o rH 3 O e CO « rH O O o o O faO c rH rH U rtt N O CD > c o CO o CO °H •H 1 SS O CN CO J- IT) CO < 'H o o > < O CO | 50 Although the size of aerosol particles could influence the death of bacteria within the aerosols, nevertheless an average size of aerosols could be used in the study without any significant error. 32 32 B. Growth of Culture with P and without P It seemed desirable to investigate whether there was any effect on 32 the growth characteristic of E. coli in the presence of P at the concentra- tion used in the present study. Figure 10 summarized the data on the growth 32 of E. coli cultures at different P concentrations in the growth medium. Although it appears that all cultures were growing at different rates, part of the difference could be attributed to the errors in the tech- niques involved. Notwithstanding the difference, all cultures entered the different growth phases almost at the same time as shown in Figure 10. It was 32 concluded that a concentration of 1 mc of P per 100 ml of medium would not significantly change the growth characteristic of E. coli cultures insofar as this study was concerned. 32 C. P Leached out from Cells 32 On storage, P '" is slowly lost from the cells into the suspending fluid. The loss is greater for living cells than dead cells, probably because of the metabolism involved in living cells. When living cells are stored at 32 4°C, however, metabolism for all practical purposes ceases . The loss of P from cells stored under this condition should be negligible, 32 The loss of P " from cells could introduce an error into aerosol experiments if the conditions of generation of the aerosols were such that particles were formed containing no R. A. cells. These particles being radio- active would give a count as if R. A. cells were present. A comparison of 32 the loss of P '" from dead and living cells is shown in Table II, o 3 51 3 O o a o CA a > 3 o M o CM CO CM CM CM &, CO CO CO a. eu a. o C i-t <-• i— t o a m uj f- aSeq.ueoaad aoueq.ri.TuisuBaj, }i-l2n 52 >- H M > M H O < H % t- • < M W > g M w O < Q U 55 CO O < IH a CO 55 o W 55 Cm M CO 3 >" CO n J 55 < » H o O ac H CO Cm CO o < u C_) o o < m H CM 55 M W M Q C_) 5= Cm a < W Cm cq cj < o CO E- .3- < H 55 < * O 55 X o CO M CO 55 55 O U M Cm CO s n •J CO o CO CO o CO o o w o S < cy < 55 M 55 M CO J >> J E- W HH c_> > M £ H o O PS < Cm CM CO a, Cm o CO CO to o o in CM (0 0) o +J CO CM CO O O it +J m to o> m b0 Q rfl h o CO CM 0.33 0.37 0.37 0.31 CO 1 CM I • 1 O 1 0.30 0.23 H to CM CM cm cn CM H o o d- oo H H o o o o o o CM O o o o o H lO © • c -H a> c H d) o o -H e fl) ^> O e u •^ O •h o a> U •^ a> +-> O 4J © +-» a 3 H « CO a£ u CO in O w cv CO CO -o o t3 bO T3 © id V r-i .H C a> o <1) r0 •C ro •H J3 a 0) 0} x: e > CO X! •H Q m +j *« •H tfl +-> .-1 * •H O J * •H f0 N_S a mh \~s » CO 53 Table II shows a greater loss from living E. coli cells, especially when they were stored at 25°C temperature. However, the loss was not great in that the cells could be left for days without risk of significant error, and it did not make much difference whether the cells were living or dead. 54 VI. EXPERIMENTAL RESULTS A. The Rate of Death and Its Calculation in Short Storage Study Experiments were conducted in a temperature range of 20°C to 50°C and a R. H. range of 20 to 80 percent. Facilities were not available to study temperatures below 20°C. After aerosols were generated from the atomizer, a large number of cells were found to have died within the first half second of storage. The following 4 seconds of storage showed a much lower rate of death. The abrupt change in the rate of death was also found by Webb (10), although his findings differed in rate from the results found in these experiments. This change in death rate is shown for one of the experiments in Figure 11. All experiments for short storage gave the same results, the only difference being the slope of the line or the rate. It was observed that a straight line relationship existed from zero to 1/2 second storage time, while another straight line fitted the data for the rest of the period. Figure 11 indicates that the death rate is proportional to the number of cells remaining alive, or J=kN (17) dt After integration, it becomes , 1 , o k " t ln N^ (18) This is the first order equation from which the rate of death, k, for any storage period can be calculated, provided that the number of cells at the beginning and at the end of the storage time, N and N , respectively, are known. 55 rf o bJ «-* £ •-H S 03 u: OS 5 < u l-H fe J ^ to c o f/l ' C V V) (0 U 0) CO 3 • r 4-> 14 ■o > ■ o 5 § 8 1 • • "6 "O >» re B5 3 u Q. w 0) i— I o r? ° C - a; e u: j ,. „*. .-- ^^ «■ *kJ- ,— ' ^^ i^ - — — * >0 n M CO e o o 0) E 0) Ds CO o CO V & §> £ <5 a s XBajajiis jo afie^uaojaj 56 The following is a typical calculation for the viability of airborne E. coli under specified experimental conditions. Experimental Condition: Culture: R. A. E. coli harvested in stationary growth phase Sprayed Suspension: Washed cells in distilled water suspension Temperature: 30°C R. H. 50% Rate of Air Flow: 43 1/min. Time R.A. Count Drop Plate Total Bacteria Percentage Survival Percent Sec. cpm/ml Count cells/ml Death 184,200 cells/ml 1.77 x 10 9 1/2 253 2.52 X 10 b 2 3.66 x IO 3 x x 10 5 253 2.52 X 10 b io 5 - 27.2% 72.8 = 9.26 9.26 X 1 242 2.34 X io 5 3.66 x 10 3 x x 10 5 242 2.34 X io 5 io 5 - 26.6% 73.4 = 8,80 8.80 x 2 228 2.10 X 10 5 3.66 x 10 3 x x IO 5 228 2.10 x io 5 io 5 _ 25.3% 74.7 = 8.30 8.30 X 3 256 2.26 x io 5 3.66 x IO 3 x x IO 5 256 2.26 X io 5 io 5 _ 23.0% 77.0 = 9.33 9.33 X 4 253 1.83 X io 5 3.66 x IO 3 x 253 1.83 X io 5 i n r>4 = 9.20 x IO 5 9.20 x 10 5 The k values can be calculated for various storage times, by using equation VI. 2, u w r , • • ,, 4 ■ . 1.77xl0 9 c/mlx0.10: N = number of living cells at zero time = ' .,<■> -, , » o & 43 1/min, 9 3 1.77x10 c/mlxO.lOl ml/min. = 4.15 x 10 c/liter of air 1. R. A. counts were corrected with self -absorption coefficient (Appendix C) V 9 2. j ■ = ■ * = 3.66 x 10 , cells per unit radioactive count. ^ K * A,; o 4.842 x IO 5 3. 6.0/59.5 = 0.101 ml/min. rate of atomization (see page 39) 57 c N- /2 = number of living cells at half-second = 27.2% x 4.15 x 10 = 1.13 x 10 6 c/1 c 1 4 15 x 10 k- /5 = rate of death for the half -second = Tprln ' ' -. Q b = 2 In 3.68 = 2.60 k^ = rate of death during the storage time from the first one-half second to the fourth second (effective storage time = 3.5 seconds), 1 t 27.2 x 4.15 x 10 6 . OQ _ . . _, = •=— r In - = 0.286 x In 1.37 d,b 19.9 x 4.15 x 10 b = 0.089 B. Summary of k Values in Short Storage Study The k values of E. coli aerosols sprayed from distilled water suspension are shown in Table III on the following page. Correction was not made in the calculation for the natural die off of E. coli when suspended in distilled water and stored at room temperature. Since the natural decay was found to be very slow for the first 45 minutes within which time all samples would have been taken, this die off was insignificant. The decay was less than 5 percent in 45 minutes (Appendix D). From Table III, a study of the effect of R. H. on the viability of E. coli aerosols could be made. When k . values were plotted against a function of R. H., (100 - R. H.), a family of straight lines was found. Each line represents the change of the rate of death, at a constant temperature, with respect to the change of R. H. in the aerosol system, Figure 13. The data suggest that the rate of death of E. coli aerosol increases as the R. H. in the environment decreases and a linear relationship exists. In other words, the lower the R. H., the higher would be the rate of death. Also from Table III, a study of the effect of temperature on the viability of E. coli aerosols was made in the same manner. If k.. .- values 58 o m o o •H CN CN o a> CO CO W H .H >-* rH o o o M H I o M co ss w Ol, en D CO w < M CO M Q £ o o V CO CN 3 Q U CO CO J o CO o OS o o w Cm o CO < > OS e Q) H o 4) W O CD CO CN os o o o o o o o in H CD d- 00 O H P» CN CD «H CO CD O 9 • » • • • • & A CO CO CN CN CN o o o o o o o cm co d- m co t> co U o o o in o m co cd cm cm 00 O P» t"» CD CD CO O rH O O O O O o o o o o o o 00 O CD CD in t*» d- co co & H m cm o CM CM CN CM H «H H in H cm cn co oo 00 CD zt CN H 00 H H • ••••• in in ^ * n cm o o o o o o o cn co .* in co t> co O O CO o • o CO O CN in co o CO CM CM CJ> CO CM O HO 00 00MD H rH O O O O O O O O O O O •H CM CM O CM O O A CM CO CO m CM o CO CO CN CM CN CN CM O O O o o o o cn co 3r m to r» co fc *S • O O o. 00 E o • 0> O CO H o o» CN CN o o o o o o o cm co Jt m co »> co O o co O • o CO 43 +j c 4-» 3 0) s c Oh O «H CO fc £ 3 oo CO h >> a> fc k a, CD co +-> C CO ITJ o k 5 •H Q) ■H -M X) ffl «H a» +J >H ■H 10 •~\ CO •rH C ■* +-» •H CO •H T3 •• X) CD » •»-> o c W rH •H a) Mh > CO fc fc .H CO -H H .C fO cu o CD >4h fc O • 3 < +j a) rH 4-» • a: 3 m O OS 59 < H g < b u; B s S u 3 5c 2 0> > 4) tL a CD <»n IS g £ a s c W u.- CO •o e h: H H o •*H w 3 •H *-4 M It < i-H •-* e g £ o o it ♦J ex U X fc • (0 •IS (0 3 g h B tul ■a to H 3 o S»» V $ H o o a M co 2 W 0- to CO Pi U < i^ CO IE O OS U. Q Cm CO CO o CO S3 o o w Cm o CO -J o 0) C4 CM e O a> CO CM OS ex E o o o o o o o CO CM O lO O 00 ZJ- loncn * no en it It CO CO CO CO CM o o o o o o o cm co :t in co r> oo. o o 00 a • O CO O oo CO I CO CM CM CM CM o o o o o o o cm co it m *o r> co u *J O O 00 cj . O CO o o» CM CM cm in C* in cm oo CO m d" CM CM o o o o o o o O O CM lO CM CO H co t H oo m m it A it co co o o o o o o o cm co & in to t> oo u * O O 00 o • O CO O CM CO CO O 00 rH It CO CM O A cm cm en oo r- r> ■H iH H O O O O o o o o o o o CO 00 it 00 A CM CO O f> It O Cn 00 00 A CO CO CO CM CM CM o o o o o o o cm co it in 10 f* co U X. o o 00 o • O CO o o CO CO c o <1) •H to CO ro a -C a> a, ex a) to 43 +j 3 ■(-> 3 to * c O • H a> fc e c bO •H t. rH >> a> ITJ u a to td c CO ■H o U C •H a) a) ■M +j o m •H fc +-> rH Q) to cu CO c J- IT) •H oo s -o «. o a) » ■m o c w rH •H > Cm to fc fc rH (T5 •H rH .c < 0) O 0) Uh fc O 3 < -P » § sz £ J- b. g Cx] H v> O M H •ri bJ P 3 «-H o ce fc o « u SB o Cd O (0 ca ♦J c o a> Cu •« If) 3 CO ■—< • o O w c o V0 •O M '( TTf r " T 3 auaoqajv jo qieaa jo a*ea J'fy 63 u 4) B U o XI u o eg o Q » u (0 u u JU u < c o u o a « > *» (B O 0> ♦■» CO 03 1000 900 800 700 600 500 4oo 300 ZOO 4 I 100 90 8o 70 60 50 40 50 20 10 20% 30% 40% 5o% 60% 70% 0^ o. r-© ,0 0^. ,—- J> 0- •■^ ■a"' 80% -o- o FIGURE 16 / THE CHAN (GE OF RATE OF WATER E VAPORATION WITH TEMPERATURE — 283 293 303 3/3 323 Temperature, T in absolute degree 67 rate of airborne E. coli can be seen in Figures 13 and 15 „ From Table V, it is also possible to determine the combined effect of R. H. and temperature on water evaporation as compared to the death rate of airborne E. coli . It is for this purpose that Figures 17 and 18 are given. These figures show the correlation of k-, /2 values and the evaporation of water under various combinations of R. H. and temperature. An arbitrary unit was chosen for the function of water evaporation, /, in plotting Figures 17 and 18. Straight lines of good fit were found for both plots. Since small water droplets evaporate instantaneously, it is believed that the function of water evaporation plotted in these two figures truly represents water evapora- tion from bacterial aerosols, and that instantaneous evaporation is the factor which governs the viability of the bacteria. Apparently the rate of death of airborne E. coli cells was in direct proportion to the rate of water evaporation from the bacterial aerosols as indicated by Figures 17 and 18. The presence of salt in the bacterial aerosols definitely affected the death rate by increasing the rate of death of bacterial cells under the same conditions of temperature and R. H. s the only exception being with environments of high temperature and very low R. H„ Under these conditions, the airborne bacteria died more rapidly if they were sprayed from distilled water suspension. A question was raised, consequently, as to the mechanism whereby the salt concentration affected the viability of E. coli in the air. Since it is assumed that the presence of salt increases water evaporation by exerting its dehydration properties on the protein-water bonded within the cells , an increase of salt concentration would certainly affect the rate of dehydration and consequently the death rate of bacterial cells. Based on this assumption, an experiment was performed by spraying E. coli aerosols suspended in 5.0 percent saline water in order to see how this increase in sodium chloride 5 ft y hi 4) 4) 4-> E O l-t >» 2 w o u w \ \ \ \ o o o \ °\ n G \° CD \ A °o\ o V o V £ ,\ \ oo vo UN KN CM O o 00 O o 00 O CM 68 c o ♦J o (8 > w u Oi *J re 3 o 4) *-> (0 0C ffoo '3 aiuoqjtjV jo mead jo a*BN ■ y[ M H 3 I OS s s u u OS M < a I ♦J s 0) B CO 10 00 g •o >» u a. M O o w \ 1 i i < 1 i \ o o \o o P 1 ° 1 ° 1° I 1 , 1 o o o o o O *1 o o o 00 69 •^ o CM CD a i O dr <\i « o o CX) O V0 O o CO s o u u CO *J (0 $• K\ N 1 euaoqjfV jo ineeci jo aieji ' % 70 concentration affected the death rates with respect to different environments of R. H. Table VI presents these data, C. Viability of E. coli Sprayed from 5 Percent Saline Suspension Table VI summarizes the results regarding the effect of high salt concentration on the viability of airborne E. coli . Correction for the natural die-off of culture was made in this case since rapid decay was found in 5 per- cent saline water at room temperature (Appendix D), D. Viability of E. coli Sprayed from 0.375 Percent Glycerol Solution In order to see whether a reduction in water evaporation would influence the rate of death of airborne bacteria, R. A. E. coli were suspended in a 0,375 percent glycerol solution for aerosol generation. Table VII indi- cates that the presence of glycerol, which reduces water evaporation because of its hygroscopic characteristic, decreases the rate of death to a great extent. To compare the effects of aerosols sprayed from different suspensions on the viability of E. coli , Figure 19 was prepared. This figure provides the evidence that bacterial viability is dependent upon the extent of water evap- oration from cells. With salt included in the bacterial aerosols, the organisms died at a greater rate due to the dehydration effect from the salt, and the rate increased as the concentration of salt in the sprayed suspension increased. In the presence of some hygroscopic material such as glycerol which reduced water evaporation, the death rate showed a considerable decrease as compared to bacterial aerosols sprayed from distilled water and salt suspensions. 71 TABLE VI k VALUES OF E D coli SPRAYED FROM 5 PERCENT SALINE WATER Ro H, % ^i/9» sec ^U» sec ~ 20 4.97 0.142 30 4.30 0„137 40 3,90 0ol35 50 3.54 0„095 60 3.40 o 095 70 3.10 0.073 80 3.12 0.079 E. coli culture harvested in stationary growth phase Rate of air flow; 43 1/min. Temperature: 30°C or 303. 8°K TABLE VII k VALUES OF E. coli SPRAYED FROM 0.375 PERCENT GLYCEROL SOLUTION 1/2' S6C 4' 20 2.29 0.075 30 2.18 0.066 40 2.23 0.065 50 2.17 0.074 60 2.03 0.059 70 1.94 0.070 80 1.97 0.070 E. coli culture harvested in stationary growth phase Rate of air flows 43 1/min. Temperature: 30°C or 303„8°K 72 FIGURE 29 COMPARISON OF THE RELATIONSHIP OF THE RATE OF DEATH OF AIRBORNE E. COLI TO RELATIVE HUMIDITY FROM VARIOUS SPRAYED SUSPENSIONS AT 30° C AND VARYING R.H. a A O 5.00% Saline Suspension 0,85% Saline Suspension Distilled \1J) Suspension 0.375%Glycerol 5 4 3 73 E. The Effect of Air Flow Rate on the Viability of Airbor ne E, coli The factor f ' (wind factor) may be included in the equation of water droplet evaporation as suggested by Frossling (38), ** a 8 I DM ( P . P ). f . dt R T d v o » ; (21) The wind factor f equals to (1 + K /Re) and it exists only when the Reynold's number is large enough because K, a value obtained through experiments, depends 1/2 on Re. The K value becomes extremely small when Re is below unity (39). To investigate whether the wind factor has any effect on the viability of air- borne E. coli , a series of experiments were conducted at 30°C and 50 percent R, H. while the rate of air flow in the system was varied from 10 . 8 to 43 liters per minute. The results are summarized in the following s -1 Air Flow Rate, 1/min. 1/2' SeC 43.0 2.93 21.6 2.84 17.3 2.70 14.4 2.68 12.7 2.80 10,8 2,77 Standard Deviation, a 0,084 The k. , values plotted against the air flow rate for this series of experiments are shown in Figure 20. Within the range of flow rates used in this study, the rate of death was not affected to any extent. This was to be expected since the Reynold's number with all the flows investigated was less than unity with an average aerosol size of approximately 13 u. The wind factor, as a consequence, was sufficiently small so that it may be neglected. 71+ $ J b. % B a g o H M fe o CM fe • £ 5 £ & w O 3 OS t-1 u < PS £ fe H E I c p c p o c o la o 41 i— 1 t*4 b. U •»* < «M o » «•> o AS OS KN CM 75 F. The Death Rate of Airborne E. coli After the First Half Second of Storage As previously shown, the death rate after the first half second changed abruptly and may be expressed as an exponential function with time, or first order reaction. The k values reported in Tables III and IV were the death rates during storage from 1/2 to 4 seconds. When the k^ values are plotted with respect to R, H, at a constant temperature, or with respect to temperatures at a constant R, H , as shown in Figures 21 and 22, respectively, straight line relationships exist. Figure 23 is plotted in much the same way as Figure 17, with k^ values against the water evaporation function, /, and smooth curves were obtained. In both curves, the same trend was found which established a definite relationship between the k u values and the rate of water evaporation. More important was the fact that k u values increased with the function of water evaporation, though not linearly. This indicated clearly that both R. H. and temperature combined had the same effect on the viability of air- borne E. coli during the storage period from 1/2 to 4 seconds and during the storage period of 1/2 second. I G. The Rate of Death of E. coli Aerosols in the Long Storage Study These experiments were performed for the purpose of investigating the difference in the death rate of airborne bacteria from distilled water origin when stored for a longer period of time. The result of a typical experiment at 50°C and 50 percent R. H. is shown in Figure 24 . When the change in the number of living bacteria with respect to time is plotted, it 1 N o ! appears that the rate of death, defined as k = — In rs°-, does not hold constant ! * N t for the whole 4 hour period. The death rate during the first two hours of ! storage was calculated in this experiment to be 0.043 min. whereas the death j rate of the subsequent two hours decreased to 0.007 min. . The same phenomenon 76 os 8 JI»3 "3 aowqjjv }<> «*»» B > 7 3-0° ' V / / ' ' r / ( / / i /f > / / / / / f / / // / / / 1 / // // t // 12 3 4 Time - hour FIGURE 2*. THE CHANGE OF POPULATION OF AIRBORNE E. COLI WITH RESPECT TO THE TIME FOR LONG STORAGE PERIOD 80 was observed in all experiments with E. coli aerosols sprayed from distilled water suspension. The death rate during the first two hours of storage may be expressed as k« and that of the subsequent two hour period as k . Table VIII summarizes the results found in this phase of study. Temperature 30°C 40°C 50°C TABLE VIII SUMMARY OF k VALUES OF E. coli AEROSOLS DURING LONG STORAGE PERIOD R. H. % k«, min -1 40 50 60 0.041 0,033 0.021 40 50 60 0.044 0.037 0.029 40 50 60 0.047 0.043 0.038 ^2-4* m * n ° 0.007 0,006 0,005 0.008 0.007 0.005 0,010 0.007 0,006 E„ coli culture harvested in stationary growth phase Sprayed from distilled water suspension The change of k values, k and k 2-4 ! with respect to R. H, and temperature is plotted in Figures 25 and 26 respectively. A linear relation- ship was found to exist for most of the data. Similar to the analysis of the short storage study, k values are plotted against the function of water evaporation, /, in Figure 27. The k value increases as the value of the water evaporation function increases, but the slope gradually levels off. A linear relationship does not appear to exist. It appears that there is a certain limit for k above which it is not affected by temperature and R„ H. When this limit is reached, k does not increase even though the temperature EC H H 8 e o Q < •«* 6 E s a a ' a B a fed 6 (A "O S CM < s ^* M 0) CO as H C9 M FIGURE OF TFE IS H 3 * IS a. CO CO i-H o •a 8 g & fc c #2 • t* 3B e 5 o ExJf ••-> CO H U o o o o o o O o o O O O us ^- m u\ sf ^ \ ' N* (M 1 JC — c V -60% 0'030 k 2 ' 0020 0-010 , 4o% 0-009 0-008 -^7 s 50% 0007 / / f \ - • "~~" t 0-006/ f<2-4 , \— *■ """^ t s / ' 60% 0005 ( •) -s } 0-004 FIGURE 26 THE CHANGE OF RATE OF DEATH WITH TEMPERATURE AT CONSTANT RELATIVE HUMIDITY IN LONG STORAGE PERIOD $• C°li sprayed from distilled water suspension) 0-003 0-002 0-001 293 303 31? 32 5 353 Temperature, T in absolute degree 83 TO THE RATE OF WATER PERIOD rater suspension ) i CM J* < ORE 27 E E. COLI G STORAGE istilled i FIGl] 'ATH OF AIRBORN 'ORATION IN LON sprayed from d 1< " RATE OF DE EVAF ( E. coli °\ \ \ 1 O CM "=* /— s a. 1 H O Q. CM • CM Experiments were therefore conducted using E. coli aerosols sprayed from growth supporting medium suspension. The same synthetic medium for cul- turing E. coli was used to wash the bacterial cells and to prepare the final — suspension for spraying. Experiments were also conducted with E. coli sprayed from a glucose solution. Although glucose by itself cannot support the growth of bacteria, it was intended to see whether an exogenous energy source could help to maintain the population. All the experiments were conducted at 30°C and 50 percent R. H„ Results of these experiments are presented in Table IX. The growth phase of the bacterial culture was also taken into consideration in these experiments. Whether the metabolites of the growth | supporting medium will be readily taken up by the bacterial cells depends i ■ipon the synthesis activities of the cells. These activities are considerably jiifferent during the different growth phases. From Table 9, it can be seen tthat in the absence of a growth supporting medium or any energy source in the 85 aerosols, there was no difference in their death rates even though E, coli of different growth phases were used. However, with growth supporting medium present, a very slight change in the death rate was noticed. The change occurred because the bacterial cells were able to carry on the necessary meta- bolic processes in an environment of 30°C and 50 percent R, H„ Undoubtedly those cells which survived the desiccation were able to utilize the food material and multiplied accordingly , As a consequence 8 more living bacterial cells were recovered at the end of the storage period , Since bacterial cells in the accelerating growth phase are more active in their synthesis processes, they could readily take up the metabolites and grow while those at the station- ary stage appeared to require a lag period before the active synthesis occurred „ The results showed that the rate of death was slower in the former case than in the latter case, TABLE IX THE EFFECTS OF GROWTH SUPPORTING MEDIUM AND GROWTH PHASES ON THE k VALUES OF AIRBORNE E, coli Spraying Suspension E, coli harvested in stationary growth phase k 2 k 2-4 Eo coli harvested in accelerating growth phase k 2 k 2-4 Culture Medium 0,029 0,028 0,028 0,0053 0,0055 0,0054 0,027 0,027 0,0041 0,0048 0,2% Glucose H O CO CO CO CO o o o o o o 0,0053 0,0055 0,0059 — =— Disto Water 0,033* 0,0060* 0,034 0,0064 *From Table VIII NOTE; Temperature 30°C, Relative Humidity 50% Glucose by itself can maintain the viability of bacterial cells in [water suspension (40), In an airborne condition^ glucose showed very little 1 effect in maintaining the bacterial population. Although the rate of death was 86 on the average smaller than in the case of a control ( E. coli sprayed from distilled water), the difference was not significant , Therefore no definite conclusion can be drawn that glucose can maintain bacteria cells in air. I, The Reactivation of Airborne E, coli with Metabolites The reactions in the Krebs cycle serve to supply intermediates for synthesis of protoplasm. The protein products of the cycle depend therefore upon these intermediates as their precursors. It is possible that by the desiccation effect, some of the reactions in the cycle are blocked and, con- sequently, the formation of amino acids is prevented « This leads to the inactivation of the bacterial cells. The following experiments were designed to investigate the possibility of reactivating some of the intermediate steps in the cycle. The following intermediates were used in the study; 0.2% oxaloacetic 0,2% sodium citrate 0.2% ot-ketoglutaric acid 0„2% succinic acid 0,2% malic acid All these acid solutions were prepared in 1;1 dilution with J M buffer solu- tion (Na^HPO + KH PO ). Samples from the long storage chamber were taken with midget impingers containing one of these metabolite solutions. It was found that in order to reactivate the cells, the collected samples should be incubated at 37°C for 24 hours. The population did not increase during this incubation period because the light transmittance of the sample solution, measured by a spectrophotometer, remained constant in this 24~hour period. The number of living cells was determined by pouring EMB agar plates before and after the incubation period. The results of a typical experiment are presented in Table X, Again here, the first order equation was used to 87 express the rate of death of the first two hours of storage period as well as the subsequent two hours (the rate of death here is defined as the true death which cannot be reactivated with intermediate metabolites), Figure 28 , TABLE X E. coli REACTIVATION WITH 0,2 PERCENT BUFFERED OXALOACETIC ACID Concentration of Percentage Rate of Death Bacterial Cells of . -1 Time in Sampling Surviving min k 2 k 2-4 hours Solution No./mlo Bacteria 1.53 x io 5 100,00 1 2.07 x 10 3 11.70 Samples not 2 1.98 x io 2 1.18 0,037 0,006 Incubated 3 1.12 x io 2 0,83 4 6,40 x io 1 0,58 1.13 x io 7 100.00 Samples 1 3.49 x io 5 27.80 Incubated 2 6.30 x io" 5.34 0.024 0.005 at 37°C for 3 3.54 x 10* 3.85 24 Hours 4 2.07 x 10* 2.85 E, coli harvested in stat ionary growth phase, sprayed from distilled water suspension, at 30°C and 50% R„ H„ with samples collected in 0,2% buffered oxaloacetic acid. The results of the experimental rate of death of airborne E. coli with reactivation are summarized in Table XI. The results indicated that all investigated metabolites, except succinic acid and malic acid, reactivated a j significant number of bacterial cells which otherwise would have been con- sidered dead because of their failure to grow in ordinary growth supporting ! medium. It is possible, from the results obtained, to compare the percentage i I of recovery of living cells before and after incubation. Using Table X as an 88 OB e o i H « .c e O I; *j CO ° ln-& © N. e t 5! 4 CD OS o JB) O 1 FIGDRE 28 THE COMPARISON OF THE DEATH RATES WITH AM) WITHOUT REACTIVATION BY E BUFFERED OXALOACETIC AC (E. coli sprayed from distilled \ OF AIRBORNE E. COLI MPLOYING 0.2 PERCENT ID SOLUTION water suspension) Without Reactivation^ *^~ ^^ / With Reactivation __, s / «*f // // // , // // 2 3 Time - hour 89 example, the percentage of recovery for the first hour, before incubation (without reactivation), was 11,7, and that after incubation (with reactivation) was 27,8. The ratio of recovery was (27. 8/11. 7\ or 2.36, The ratio of recovery for the second hour of storage was then (5.34/1.17), or 4.52, for the third hour, (3,85/0.83), or 4.64 and for the fourth hour, (2.86/0,58), or 4.93. It can be seen that the ratio of recovery increases with an increase in percentage of death o In other words, with a greater number of apparently dead cells, the percentage of recovery was higher . This is somewhat expected since a greater probability of reactivation exists when more inactivated cells are subjected to reactivation. The same phenomenon was observed for all experiments using a-ketoglutaric acid and sodium citrate as reactivation agents „ TABLE XI THE EFFECT OF REACTIVATION ON THE RATE OF DEATH OF AIRBORNE E„ coli . -1 mm. k 2-* Incubation Solution Rate of k 2 Death, Oxaloacetic acid 0.024 0.0050 Sodium citrate 0.022 0.0050 a-ketoglutaric acid 0.018 0.0040 Succinic acid 0.032 0.0054 Malic acid 0.034 0.0051 0.85% NaCl (control) 0.033 0.0060 E. coli harvested in stationary growth phase, sprayed from distilled water at 30°C and 50% R. H. The results indicated that the desiccation of airborne E. coli had blocked the operation of the Krebs cycle in some cells to some degree, possibly by inactivating some enzymes involved in the synthesis of proteins. By supply- ing certain intermediate metabolites, the blocked reactions of the cycle were restored in some cells and the normal flow pattern of the cyclic operation was completely recovered. 90 j The Stability of Airborne E„ coli Ae rosolized with Glutamic and Aspartic 1 Acids Since it was established that a portion of the death of airborne E 3 coli was the result of the interference of the synthesis of glutamic acid and of aspartic acid, the presence of either of these two acids in the aerosols would certainly help to stabilize the bacterial cells „ In order to verify this fact, E. coli aerosols were sprayed from glutamic acid and aspartic acid solutions, both were prepared as 0»2 percent solutions with 0,1 H phosphate buffer, NaJUPO^ plus KH.PO.o The death rates observed in these experiments definitely showed that airborne E, coli population could maintain life much longer than when sprayed from a distilled water suspension „ Figure 29 compares the death rates affected by the presence of glutamic acid or aspartic acido Death rates of E, coli aerosols sprayed from a culture medium suspension as well as from distilled water were also plotted for comparison „ Evidently glutamic acid was the most effective in stabilizing airborne cells whereas aspartic acid and the culture medium had about the same effect „ 91 0> e O c (D -1.3a CD U 0> c > «M o ♦J CO oc 4) XI o In a I o E. doli Sprayed from Distilled Water E. coli Sprayed from Culture Medium A E. coli Sprayed from 0.2 % Buffered Aspartic Acid a E. coli Sprayed from 0.2 % Buffered Glutamic Acid ^^ t c y/^^< °y // w / //// / I /A / 7 0' / / ft / /{ > / 0" y& / 2 3 4 Time - hour FIGURE 29. THE COMPARISON OF THE DEATH RATES OF AIRBORNE E. COLI FROM DIFFERENT ORIGINS (IN LONG STORAGE PERIOD) 92 VII. DISCUSSION OF RESULTS General Discussion It was the purpose of this study to determine if one or more factors influence the death of airborne bacteria using a single species, E. coli c A number of investigators have observed that viability of airborne bacteria varied from genus to genus „ In spite of this fact, it is felt that there should be some common factors affecting the viability of airborne cells „ Webb's (10) data showed that S albus , B„ subtilis , S. marcescens and E. coli varied considerably in their death rates „ All of these species indicated an extremely high rate of death during the first second of storage, while the death rate during the next 9 seconds was relatively slow The same phenomenon was observed in the present study „ This readily suggests one common mechanism which might be responsible for the rapid decay of all these bacteria, regard- less of structural differences between the organisms „ The results also indicated that the highest mortality occurred at low R. H, values, with a decrease in rate of death with increasing R H„ Paral= lei to this phenomenon is the evaporation of a water droplet which takes place readily when it is exposed to an unsaturated air environment Appendix A shows that a water droplet of 13 V size at room temperature and R H evaporates to 0.9 y diameter within four-tenths of a second. Consider an aerosol of about 1 13 V size, composed of a distilled water layer surrounding a bacterial cell mass of 0.9 V size, it would take therefore less than four-tenths of a second to evaporate the water layer completely and leave the bacteria cell suspended and exposed in the air. The evaporation of a water droplet at rest can be expressed by the following equation dA 8*M , dt = R-T-d (f °P T - Pe } rva = Constant. =■ (f „p T - p Q ) 93 The half-second storage time, the time period after the bacterial cell has been entirely exposed to the air without the water layer appears to be critical to the viability of the airborne E poli o The rate of death for this one-half second storage time has been shown in this study to be directly proportional to the decrease of R„ H when temperature is constant, Figures 12 and 14 The same relationship is found in the evaporation of a water droplet since at con- stant temperature, the equation becomes -rr = Constant „ (f „p T - p Q ) Furthermore, the rate of death of airborne cells during the first half-second storage period at constant R» H„ was shown to be exponentially proportional to ; increasing temperature, Figures 13 and 15 „ It is important to note that the rate of evaporation of water droplet with varying temperature holds the same relationship, Figure 16 „ Webb (10) observed a decrease in death rate with increased tempera- tures during the first second of aerosolization This was 9 however, not observed in the present study „ Webb explained his findings as being the result i ; of a sudden chilling effect „ It is obvious that the evaporation of water drops would result in a sudden lowering of the temperature at the surface of the drops o Therefore it is conceivable that some portion of the initial death of bacterial aerosol may be due to a large temperature drop When the process of ; evaporation is carried out at a higher temperature the chilling effect is less Q Considering Equation 14 for temperature change, A - L M D (f ° PT " Pq) ■ T 6 " FTT (D/rva + l) + T the decrease in temperature, as calculated from this equation, would not lower 94 the droplet temperature down to 4°C at all the temperature and R„ H Q levels investigated in the present study, Table V. Hence, the sudden chilling effect on the bacterial aerosols cannot entirely account for their rapid initial death, The presence of bacterial cells in an aerosol could possibly reduce the evapo- ration coefficient a, so that the temperature decrease would be even smaller „ Therefore, the calculation of the temperature decrease with a equal to 0.04 for the case of pure water gave higher values of temperature decrease (Table V) The calculation of 6 requires trial and error procedures which in themselves introduce some inaccuracy which is not significant for the present study. Also, the radius was taken as 13 u in all calculations, though it was only true as an average condition „ The following is a sample calculation of 8 for a water droplet of 13 u size in a 30°C and 50 percent R H„ environment „ 579.5 x 18 x 0.255 (0.5 x 33.321 - p ) + 303.8 0,0000288 x 62360 x 303.8 [0.255/(13/2) x 10 x 0.04 x 14950 + 1] (16.66 - p ) = 4,87 - — . .-- U + 303.8 1. bob By trial and error, it was found that at 21„7°C, p was equal to 19.468, 6 = 2.94 (16.86 - 19.468) + 303.8 = 295.5° absolute temperature 9 or 21.7°C Thus, the decrease in temperature was 303.8° - 295.5° a 8.3°C. The calculation ;of droplet temperature for other environmental conditions followed the same procedure. The results were tabulated in Table V. The lowest droplet tempera- ture in the present study, as calculated B was 281.2° absolute degrees or 7„4°C 9 which was above 4°C. Therefore the lethal effect of a sudden chilling to 4°C was not applicable in interpreting the data collected in this study (48). 195 Since temperature and R. H, affect the viability of airborne bacteria simultaneously, it is necessary to study their combined effect. It was shown in Figure 17 that the rate of death of airborne E. coli was directly propor- tional to the rate of water evaporation. The rate of evaporation of protein- bonded water within cells is much different from the rate of water evaporation from a water droplet. The difference can be attributed to the fact that the diffusing water molecules were crossing boundaries of different thicknesses in completing the evaporation process. In the evaporation of water droplets, a water molecule diffuses from the spherical droplet into an imaginary boundary whereby the water molecule will have traveled a certain distance, Equation 7, before it collides with a gas molecule which completes the evaporation. Diffu- sion of a water molecule out of a bacterial cell is further complicated by the presence of the cellular membrane and the protein-water bond. The protein- water bond requires more energy in order for it to rupture before the water molecule is free to travel. The cell membrane adds to the effective thickness of the imaginary boundary through which the water molecule must pass before completing the course of evaporation. It is, therefore, evident that the rate of evaporation of water from within cells is much slower than that from water droplets alone. From Figure 17, it appears that as the function of water evaporation becomes smaller, the spread of the individual points from the straight line becomes larger. This probably was due to the fact that at low temperature and j high relative humidity, the evaporation of water was so slow that there was a significant lag period before the protein-bonded water could start diffusing | out of the cell. As a result, the relationship between k . and the function I of water evaporation would not hold. In the case of aerosols from saline water j suspensions, this possible lag period did not exist because of the immediate 96 onset of dehydration of the cellular water. Therefore the straight line plot in Figure 18 was a perfect fit throughout all f- values obtained in the study . Glycerol in high concentration is "toxic" to bacteria, but at a concentration of 0.375 percent, it causes no harmful effect on bacteria. It is a trihydroxy alcohol highly hygroscopic in nature. When bacterial aerosols are sprayed using a glycerol suspension, water is not as readily evaporated as in aerosols sprayed from distilled water suspensions. Two different characterise tics were encountered with airborne bacteria sprayed from glycerol suspensions a. Owing to the hygroscopic nature of glycerol, the aerosols appeared to hold more moisture on the cells. Thus, a longer period of time was required to evaporate the water. In fact, it was found in the present ' study that E. coli suspended in 0,375 percent glycerol, as examined by a Carl Zeiss phase microscope, showed an average size of somewhere between 1 y and 2 u, while the same culture suspended in distilled water showed an average size of less than 1 p. This indicated in glycerol suspension E. coli had a higher water content. b. The glycerol as a trihydroxy alcohol would reduce the rate of water evaporation. It is a known fact that water, with added impurities , evaporates at a slower rate. As shown in Appendix A, the equation for calcu- lating the time required for water evaporation indicates that anything which reduced the a value would result in a longer time for evaporation of the same droplet. It may be assumed, therefore, that glycerol reduced the a value A reduction in the rate of water evaporation by fatty alcohols was reported by Eisner, Quince and Slack (44). The results showed a significant decrease in the death rate, as anticipated. This suggests the theory that the rate of water evaporation governs the rate of death of bacterial aerosols. 97 The study of viability of airborne bacteria sprayed from a saline solution indicated a similar result compared to that obtained from aerosols of distilled water origin. The difference between the two was the slopes of the curves shown in Figure 30, It was shown in the case of E. coli aerosols from saline solution that the rate of change in k . values was less than in the case of aerosols sprayed from distilled water. Another difference was that, at low temperature and high R. H,, the k . values were higher in the case of bacterial aerosols from saline suspension, but the reverse was true at high temperatures and low R, H. as shown in Figure 30. The mechanism of death of E. coli in both cases was identical. Regardless of the difference in slopes, the rate of water evaporation governed the rate of death. There were other factors besides temperature and R. H. which affected the evaporation of cellular water in the presence of sodium chloride. The presence of sodium chloride altered the rate of cellular water evaporation in the following ways : a. After the water had been evaporated from the NaCl colution surrounding the E. coli aerosols, NaCl would form either a layer of crystals I or, when the evaporation was not complete, a very high concentration layer of NaCl solution. In either case, the evaporation process was retarded because the thickness of the barrier through which a diffusing water molecule had to travel was increased. Sodium chloride also reduced water evaporation by lowering the vapor pressure in the solution (52), b. "Dehydration" of protein molecules in the presence of high salt content accelerates water evaporation within cells. The higher the concentra- tion, the more charged groups were formed and more polarized water particles were extracted from the cell proteins. c. It has been mentioned that an osmotic pressure change can be tolerated by most bacteria to a certain extent. In spite of this fact, the ft ft 0= ft O H C/5 o o o OS 5 o E ri o o a. < • > u: fc 8 < OJ 3 ce 1— 1 < ft ft B H < a M 0) 4-> i 4-> i 0) c ** T5 f* a CO i-h (A i— i •»H 4-> in co oo •it • -a o § § >- u «M «M TJ •a a CD >» >» CO CO M 1* a. cl trt w •*« •* pH «-H O o o u c • W cut o « 00 vD Irs xj- KN «sl — lioo "3 auaoqjjv jo meag jo 9%b^ ' ^ o 98 CL dr e o 'f\ +•> CO o CL CO > hi u rotein carbon of E„ coli cells (12), it is apparent that the Krebs cycle is :he major mechanism for producing the necessary compounds for protein synthesis, 'he operation of the Krebs cycle is also observed in other organisms 9 various kinds of yeast, mold and algae „ In the present study the flow pattern of the .rebs cycle was used to demonstrate the ability of some metabolites to react iv- te airborne bacterial cells . This demonstrated that inactivation of certain rotein molecules due to the lack of these metabolites or to desiccation in irborne bacteria was responsible for the death of cells, and that the respons- ble metabolites could restore the activity of bacterial cells „ Although the rebs cycle is not the mechanism synthesizing the necessary amino acids for all he proteins in the cell, it is felt that the same phenomenon of inactivation lad reactivation of protein molecules affecting the activity of cells occurs i protein molecules other than those derived from the Krebs cyclic operation lis means that other amino acid families such as serine, histidine and i somatic amino acids or their precursors could be inactivated and reactivated 9 107 and the activity of the bacterial cells is affected accordingly. A Krebs cycle is shown in the following to facilitate discussion regarding the mechanism of the reactivation process! CO, 2 PYRUVIC^—* ACETIC CO ^ — > OXALOACETIC MALIC CITRIC ASPARTIC ACONITIC FUMARIC SUCCINIC CO. GLUTAMIC ISOCITRIC OXALOSUCCINIC a-KETOGLUTARIC CO, The metabolites investigated were: Oxaloacetic acid Citric acid (Sodium citrate) a-ketoglutaric acid Succinic acid Malic acid 108 Since oxaloacetic acid and a-ketoglutaric acid are immediate precur- sors of aspartic acid and glutamic acid respectively, they are most important insofar as their positions in the operation of the Krebs cycle are concerned. Results showed that both oxaloacetic acid and a-ketoglutaric acid, particularly oxaloacetic acid, restored the activity of some inactivated bacterial cells if a proper incubation period of approximately 24 hours at a temperature of 37°C was provided. This observation suggested two possibilities of what may have happened to the airborne bacteria in the study: 1. The synthesis of aspartic acid and of glutamic acid in the cells was blocked so as to cause inactivation of the cells. Supplemented with their immediate precursors which would induce enzyme formation for the synthesis, these two amino acid compounds could be synthesized and the activities of cells were partially reestablished. 2. The formation of oxaloacetic acid and of a-ketoglutaric acid was blocked. When they are supplied as exogenous metabolites, the operation of the Krebs cycle was restored during the incubation time. The concept of reactivation of "killed" bacteria here suggests that some basic functional activities of the cellular processes were left essen- tially intact during the lethal process of desiccation. The cells may regain activity if they are brought to function specifically to reestablish cellular livision. Two modes of action of the metabolites investigated in this study :an be suggested in justifying the reactivation of "killed" bacteria. The -irst restores the inactivated enzyme systems necessary for metabolism, ■nox et al. (50), working with E. coli , demonstrated that chlorine in bacteri- cidal concentrations reduced the activity of various sulfahydryl enzymes and I'ther enzymes which were sensitive to oxidation. Oliver (51) found that the ctivity of the decarboxylating enzymes of E. coli ceased above 80°C but re- ;Umed their activity on return to lower temperatures. He also observed that 109 E, coli in suspensions "killed" by heat and chemicals were still sterile after 7 days' incubation in culture medium but possessed almost normal decarboxylase activity,. This indicated that loss of viability did not represent the loss of all enzyme activity. All those experiments reported by previous workers, as well as the experiments in the present study, support a speculation that if a route of entry into the respiratory cycles could be found with a proper metabo- lite, a partial restoration of cyclic process could take place „ Such a phenom- enon could be explained as a stepwise operation in a chain process where a metabolite induces enzyme formation, and subsequent enzymatic conversion of metabolite would provide a new metabolite for other enzyme format ion „ In such a manner, a complete cycle could become operative The second or alternate node of action of the metabolites in reactivation could be interpreted as providing a supply of more readily available energy „ The supplemented tricar- boxylic acids which are immediately utilizable compounds would serve as energy sources o It was also shown that citrate was able to restore partially the Inactivated airborne bacteria whereas succinic acid and malic acid did not I ■how any definite reactivation ability „ Table XI showed that a-ketoglutaric icid was most effective in reactivation, sodium citrate being less effective nd oxaloacetic acid the least effective among the three „ The result can be nterpreted with respect to the positions of these compounds in the Krebs cycle peration,, Since oxaloacetic acid and a-ketoglutaric acid are immediate pre- ursors of aspartic acid and glutamic acid respectively 8 which are the primary j ompounds for protein synthesis in E, coli 9 their ability to reactivate "killed" ells is readily visible „ Regardless of whether protein molecules or enzyme /stems are inactivated, oxaloacetic acid and a-ketoglutaric acid are in a .Loser position with the amino acids and protein products B and the presence of I ;iese two compounds certainly gives a higher efficiency of reactivation,, 110 letween aspartic acid and glutamic acid, the former is a much better amino acid donor (12) „ It seems that internally synthesized aspartic acid acts as an amino acid donor in many transamination reactions , even in the absence of exogenous aspartic acido On the other hand, glutamic acid is a relatively poor amino acid donor, and needs more exogenous supply to carry on the metabo- lism. The supplement of a-ketoglutaric acid replenishes the glutamic acid and very likely induces enzyme formation in order to clear the pathways for metabo- lism o It was also shown that a large fraction of exogenous glutamic acid can be converted into aspartic acid through the Krebs cycle operation (12) „ Con- sequently, supplemented a-ketoglutaric alone could restore partially the cyclic flow and therefore reestablish activities of the "killed" bacteria,. Though the supply of exogenous oxaloacetic acid alone may induce the formation of aspartic acid, it has to be incorporated with an exogenous acetate supply in order to carry out further steps in the Krebs cycle , Without the acetate supply, the synthesis of glutamic acid is significantly retarded „ Therefore the efficiency of the reactivation process carried out by oxaloacetic acid alone cannot be expected to be as great as that by a-ketoglutaric acido When exogenous citrate jis supplied, the Krebs cycle is carried on with the formation of glutamic acid which in turn produces aspartic acido As a result, it gives a better effi- ciency of reactivation than oxaloacetic acido Succinic acid and malic acid were also supplied in the present study as exogenous sources for the purpose of demonstrating reactivation Their positions in the Krebs cycle, however, are such that they do not lead to the synthesis of aspartic acid and glutamic acid as readily and directly as oxaloacetic acido This makes them extremely ineffec- tive to restore the cyclic operation if glutamic acid needs to be supplied From the discussion above, it is felt that the inactivation of air- pome cells does not occur specifically in some particular protein molecule or Enzyme systems. Very likely most of the proteins and enzymes involved in the Ill bacterial synthesis are partially damaged because of broken protein-water bonds and lack of readily utilizable metabolites Therefore , each investigated metabolite had more or less some effect in react ivat ion „ The degree of activa- tion appears to depend upon how readily the metabolite can be utilized to synthesize aspartic acid as well as glutamic acid„ The glutamic acid synthesis is particularly important because internal synthesis of glutamic acid is not as good an amino acid donor as aspartic acid„ The exogenous aspartic acid supply may be deficient, but the same acid compound synthesized internally can carry on the cycle operation though the rate of the flow in the operation is much lesso Since aspartic acid and glutamic acid were of primary concern in the reactivation of bacterial cells, the last experiment was designed to see how their presence in the bacterial aerosols could affect the viability of air- k° rne E° coli . It was observed that both aspartic acid and glutamic acid stabilized the airborne E coli Glutamic acid was the most effective while aspartic acid was about as effective as the growth supporting medium,, The stabilization phenomenon may be attributed either to the presence of readily available energy for growth in the storage period, or to the effect of react iv= ation<> Since in this experiment no incubation was required for the samples before pouring agar plates, both acid compounds were probably more important in supplying an available energy source than in supplying metabolites to induce enzyme formation for reactivation,, Some practical aspects of the phenomenon of stabilization and reactivation of airborne bacteria certainly deserve attention,, If the sterility pf an air environment is tested by culture plate or culture tube techniques, the sterility is only arbitrary because the present study reveals that suitable metabolites could recover inactivated bacterial cells when incubated for a certain period,, Air hygiene such as the disinfection of air by ultraviolet 112 light or by chemical agents would not ensure the safety of the "cleaned" air„ A large amount of "killed" pathogenic bacteria might enter the body of a host and, through proper incubation in the presence of some suitable metabolites a be reactivated and resume their multiplication process „ The sterilizing of an air environment perhaps should be raised to such a standard that metabolic recovery and metabolic stabilization would fail in order to secure complete sterility.. Bo Future Works From the author's experience in the present study, it is felt that the particle size associated with aerosols containing bacteria affect the viability of cells » More bacterial cells in one aerosol for example would ■ protect each other from rapid cellular water evaporation,) It is suggested that works should be done in the future concerning this aspect „ Also, it seems important that studies of the bacterial stabilization in the presence of amino acids should be carried out more intensively in the future « A study covering a variety of amino acids should be done in order to determine a general signif- icance o Another aspect worth studying is the reactivation of airborne bacteria after sterilization by different disinfectants, such as sodium hypochlorite, resorcinol, propylene glycol, or by physical treatment such as exposure to ultraviolet light „ Standard concentrations of these disinfectants should be investigated in different environments so that no reactivation with metabolites can be demonstrated „ Finally,, different kinds of bacteria should be studied s including the most frequently found airborne pathogenic bacteria, so as to fulfill the practical purpose of controlling the viability of airborne bacteria. 113 VIII. CONCLUSIONS A. Factors Affecting the Viability of Airborne Bacteria During Short Storage 1. The mechanism of instantaneous killing of airborne bacteria is related to the evaporation of cellular water This finding and the results of other research workers suggest that this mechanism can be applied to all kinds of airborne bacteria, 2. Relative humidity in the environment determines the equilibrium water content in airborne bacteria „ It also determines the rate of transfer of water vapor between cell surface and the environment,, The rate of death of some airborne bacteria is directly proportional to the dryness of the environ- mental air. A linear relationship exists between the rate of death and the dryness of the air. 3. Temperature governs the rate of change of water vapor as well as the rate of exchange of heat between the cell surface and the environment . Therefore it affects the viability of airborne bacteria. The rate of death increases exponentially with an increase in temperature. 4. The rate of death of airborne bacteria is directly proportional to the rate of cellular water evaporation in an instantaneous lethal process. 5. Foreign material associated with bacterial aerosols either increases or decreases the rate of death of airborne bacteria by changing the rate of water evaporation from within the cells. 3. Fact ors Affecting the Viability of Airborne Bacteria During Long Storage 1, The rate of death also increases with an increased rate of water evaporation, but a linear relationship does not hold. Other factors affect :he viability simultaneously. . 114 2 The viability of airborne bacteria is affected by the presence >f a growth supporting medium,, At 30°C and 50 percent R H , airborne bacteria ire capable of synthesizing food material and therefore more living cells can i)e recovered. Consequently, the rate of death is less 3. When a growth supporting medium is present , the growth phase of >acteria also influences the viability of airborne bacteria. Active cells , in :he accelerating growth phase, appear to have a slower death rate because more lew cells are synthesized to offset the number of dying cells 4 In the presence of amino acids which can be readily utilized as .intermediate metabolites by the airborne cells, the viability is greatly .ncreasedo '• Reactivation of Airborne Bacteria Inactivated airborne bacteria can be recovered with proper incubation n the presence of intermediate metabolites The degree of disinfection of an ir environment should be raised to such an extent that metabolic recovery ould fail„ 115 ,. BIBLIOGRAPHY Duguid, J, P,, "The Size and the Duration of Air-Carriage of Respiratory Droplets and Droplet-nuclei," J, of Hygiene , 44, 471 (1946), 2, Lidwell, 0, M. , and Lowbury, E„ J„, "The Survival of Bacteria in Dust," J, of Hygiene , 48 , 6 (1950),, 3o Loosli, Clayton G,, Referree, Am, J, of Public Health and the Nation's Health , 37, 4, 353 (1947) . 4. DeOme, K, B., "Effect of Temperature and Humidity and Glycol Vapour on Viability of Airborne Bacteria," Am, J. of Hygiene , 40 , 239 (1944), 5, Loosli, C, Go, Lemon, H, M, , Robertson, 0, H,, and Appel, E„, "Experimental Airborne Influenza Infection," Proc, Soc. Exptl. Biol, Med, , 53 , 205 (1943), 6o Dunklin, E. W,, and Pick, T. T,, "The Lethal Effect of Relative Humidity on Airborne Bacteria," J, of Exptl. Med, , 87 , 87 (1948), 7, Ferry, R. M„, and Maple, L, G,, "Studies on the Loss of Viability of Stored Bacterial Aerosols," J. of Infectious Diseases , 95, 142 (1954), 8, Shechmeister, L. L, , and Goldberg, L, J„, "Studies on the Epidemiology of Respiratory Infections," J, of Infectious Diseases , 87, 116 (1950), 9, Wells, W, , and Zapposont, P,, "The Effect of Humidity on B-streptococci Atomized in Air," Science , 96 , 277 (1942), 0, Webb, S. J„, "Factors Affecting the Viability of Air-borne Bacteria," Canadian J, Microbio. , _5, 649 (1959), 1, Harper, G, J„, Hood, A, M,, and Morton, J, D„, "Airborne Microorganisms; A Technique for Studying their Survival," J, of Hygiene , 56, 364 (1958), 2, Roberts, R. B , Cowie, D„ B,, Abelson, P, H,, Bolton, E, T, 9 and Britten, R, J,, Studies of Biosynthesis in Escherichia coli s Carnegie Institution of Washington Publication, 607, Washington, D, C, 58 (1957), 3, Wells, W„ F, , "Apparatus for the Study of the Bacterial Behavior of Air," Am, J, Pub, Health , 23, 58 (1933), '+, Wells, W, F, , and Stone, W, R, , "On Airborne Infection, Study II," Droplets and Droplet Nuclei," Am, J, of Hygiene , 20 , 611 (1934), 3, Burrows, William, Bacteriology , Fifth ed,, W, B„ Saunders Co,, Philadelphia, 231 (1949), >» Wells, Wo Fo, "Airborne Infection and Sanitary Air Control," J, Ind, Hyg„ , 17, 253 (1935), !', Wells, W, F , and Riley, E, C, "An Investigation of the Bacterial Contami- nation of the Air of Textile Mills with Special Reference to the Influence of Artificial Humidification," J, Ind, Hyg, and Toxicol,, 19, 513 (1937), 116 18, Williamson, A, E., and Got ass 8 H G. s "Aerosol Sterilization of Airborne Bacteria/' Ind. Med. , 11, 40 (1942). 19, Edward, D. B„, Elford, W. J., and Laidlaw, P P., "Airborne Virus Infec- tions," J. of Hygiene , 43, 1 (1943), 20, Loosli, C. G„, Lemon, H, Mo, Robertson, 0. H„, and Appel, E,, "Experimental Airborne Infection, I, Influence of Humidity on Survival of Virus in Air," Proc. Soc. Exp. Biol, and Med. , 53 , 205 (1943). 21, Davis, Mary S,, and Baterman, J. B„, "Relative Humidity and the Killing of Bacteria; 1, Observations on Escherichia coli and Micrococcus lysodeikti- cus," J. of Bact. , 80 , 577 (I960). 22, Kethley, T, W„, Coun, W, B., and Fincher, E„ L,, "The Nature and Composition of Experimental Bacterial Aerosols," Applied Microbiology , 5, 1 (1957). 23, Kethley, T, W., Coun, W, B,, and Fincher, E L,, "A System for the Evalua- tion of Aerial Disinfectants," Applied Microbiology , 237 (1956). 24, Henderson, D, W. , "An Apparatus for the Study of Airborne Infection," J. of Hygiene , 50 , 53 (1952), 25, Kethley, T. W., Fincher, E. L., and Coun, W. B., "The Effect of Sampling Method upon the Apparent Response of Airborne Bacteria to Temperature and Relative Humidity," J. of Infectious Diseases , 100 , 97 (1957), 26, Dimmick, R, L, , "Measurement of the Physical Decay of Aerosols by a Light Scatter Method Compared to a Radioactive Tracer Method," J, of H ygiene, 58, 373 (1960), 27, 0'Konski, C, T,, and Doyle, G, T,, "Light Scattering Studies in Aerosols with a New Counter Photometer," Analyt. Chem. , 27, 694 (May, 1955). 28, Mellon, E, F,, Karn, A, H,, and Hoover, S„ R, , "Water Absorption of Pro- teins; 1, Effect of Free Amino Groups in Casein," J. Am Chemo S oc. , 69, 827 (1947), 29, Mellon, E, F,, Karn, A, H„, and Hoover, S„ R, , "Water Absorption of Pro- teins; Effect of Physical Structure," J. Am, Chem, Soc , 71, 2761 (1949). 30, Mellon, E, F. , Karn, A. H., and Hoover, S. R., "Water Absorption of Pro- teinss Effect of Guanidino Groups of Casein," J . Am , Chem . Soc . , 73, 1870 (1951). 31, Fraser, R. D„ B,, and Macrae, T, P., "Possible Role of Water in Collagen Structure," Nature , 183 , 179 (1959), 32, Crick, F. H. C, and Kendrew, J. C„, Adv. in Protein Chem. , 12 , 133, Academic Press, Inc., New York (1957). 33, Linderstrtfm-L.ang , K. V. , and Schellman, J. A,, "Protein Structure and Enzyme Activity," The Enzyme , 1, 2nd ed., edited by Boyer, P. D„, Lardy, H,, and Myrback, K,, 470, Academic Press, Inc., New York (1959), 34, Scott, W. J„, "The Effect of Residual Water on the Survival of Dried Bacteria During Storage," J. General Microbiology , !£, 624 (1958), 117 35o Ferry, R„ M., Brown 9 W„ F , and Damon, E. B„, "Studies of the Loss of Viability of Stored Bacterial Aerosols; II 9 Death Rate of Several non- pathogenic Organisms in Relation to Biological and Structural Character- istics," J, of Hygiene , 56 , 125 (1958). Fuchs, No A , Evaporation and Droplet Growth in Gaseous Media , Chapter 3, 14, Pergamon Press, New York (1959). Green, H. L., and Lane, W. R. , Particulate Clouds; Dust, Smokes and Mists , 84, D. Van Nostrand Company, Inc., New York (1957). 58. Frossling, N„, "Uber die Verdunstung Fallender Tropfen," Beitr„ Geophys. , 52, 170 (1938), }9. Kinzer, G. Do, and Gunn, R. , "The Evaporation, Temperature and Thermal Relaxation Time of Freely-falling Water Drops," J, Met., 8, 71 (1951). K). McGraw, S. B„, and Mallette, M F„, "Energy of Maintenance in E coli ," J, of Bacteriology , 83 , 844 (1962), \1, Dautrebande, L„ , Study on Aerosols , 4, University of Rochester, Rochester (1958). t2. Stalpott, "Aerosols Medicamenteux IX. De. 1'action Durelique de diverces substances administrees sous forme d' aerosols," Arch. Intern „ Pharmaco- dynamic , 72, 282 (1946). }3. Lanterbach, K E., Hayes, A D„, and Coelho, M„ A„, "An Improved Aerosol Generator," Arch, Indo Health , 13 , 156 (1956). ■4, Eisner, H. S., Quince, B W , and Slack, C„ s "The Stabilization of Water Mists by Insoluble Monolayers," Discussion Faraday Soc , 30 , 86 (1960) „ ■5. Alty, T,, "The Maximum Rate of Evaporation of Water," The Philosophical Magazine , Ser 7, 15, 82 (1933). j6„ Johnson, J. C, "Measurement of the Surface Temperature of Evaporating Water Drops," J. of Applied Physics , 21 (1), 22 (1950). 7„ Hilsenrath, J„, Toulaukian, Y. S„, Trans. Amer. Soc. Mech. Engr. , 76, 967 (1954). 8. Meynell, G. G., "The Effect of Sudden Chilling on E. coli ," J. Gen. Microbiol. , 19, 380 (1958). 9. Heinmets, F„, Taylor, W. W., and Lehman, J. J., "The Use of Metabolites in the Restoration of the Viability of Heat and Chemically Inactivated E. coli ," J. Bacteriology , 67 , 5 (1954). 0. Knox, W. E., Stump, P. K. , Green, D. E., and Auerbach, V. H., "The Inhibi- tion of Sulfhydryl Enzymes as the Basis of the Bactericidal Action of Chlorine," J. Bacteriology , 55, 451 (1948). |1. Oliver, W. H., "The Stability of Some Bacterial Enzymes Toward Heat and Chemical Bactericides," J. General Microbiology , _7, 329 (1952). | 2 ° Handbook of Chemistry and Physics , 41st ed., 2432, Chemical Rubber Pub- lishing Company, Cleveland (1959-1960 ) . 118 APPENDIX A THE CALCULATION OF THE TIME PERIOD REQUIRED TO EVAPORATE A WATER DROPLET FROM 13 y DIAMETER DOWN TO 0.9 y* DIAMETER By using equation (15) dA 8TTM ,. > dt = Rfd" rVa (f 'P T "Pe ) where A = surface area of water droplet ct = evaporation coefficient, 0.04 for pure distilled water d = density of water vapor, =1 f = relative humidity expressed as a fraction M = molecular weight of water, 18 p = water vapor pressure at ambient absolute temperature p = water vapor pressure at equilibrium temperature o r = radius of water droplet R = universal gas constant, 62,360 cc mm/mol„ deg t = time in seconds T = ambient absolute temperature 1/2 v = (RT/2ttM) , a constant at constant temperature. 2 Since A = 4nr , substitute into the previous equation, d(4Trr 2 ) 8ttM lc . _ = _ rV a (f ,p T - p Q ) _ dr 8ttM ,,. . 8ur dT = RTd" rVa (foP T " V k 0.9 y is the equivalent size of the bacterial cell mass in an average size aerosol of 13 y dia. (see Table I). 119 Rearrange and integrate, dt = RTd Mvct (f.p T - p e ) , dr or t = - RTd r " r t Mva (f.p - p ) where droplet temperature 8 and the corresponding vapor pressure p. can be calculated by using equation (14), a „ lmd (f 'P T - V , " KRT (D/rva + 1) At any instant, for example at 30°C and 50 percent R. H, 6 = 295,5° absolute temperature, and p = 19,458 mm Hg. (see page 77) ,-4 t = 62,360 x (303.8) x 1 (13/2 - 0.9/2) x 10 18 x 14,950 x 0.04 " [1/2 x (33.321) - 19.468] = 0,38 second 120 APPENDIX B RELATIVE HUMIDITY DETERMINED BY WET BULB AND DRY BULB TEMPERATURE (Figures in the table are the wet bulb temperature readings that go with the corresponding dry bulb temperatures to give the proper relative humidities) Dry Bulb R. H . Temp. 20% 30% 40% 50% 60% 70% 80% 20°C 9.3* 11 12.5 13.8 15.2 16.3 17.8 68°F 48.8 51.8 54.3 56.9 59.3 61.6 64 30°C 16.2 18.0 20.0 22.1 23.9 25.5 27.1 86°F 61.2 64.3 68.1 71.8 74.9 78.0 80.9 4Q°C 22.1 25.1 27.9 30,3 32.5 34.8 37.0 104°F 71.7 77.2 82.1 86.6 90.7 94.4 98.0 50°C 29.0 34.0 36.0 39.0 42.0 45.0 47.0 12 2° F 85.0 94.0 97,0 103 108 113 116 60°C 36.0 42.0 46.0 49.0 52.0 54.0 57.0 140°F 97.0 108 115 120 126 130 134 121 OS e 3 o u *i — 1 'H » C s a 0) IA CO u fc B & H E •* O Q O O O O O O Q Q Q E HlSQ'-< l NC>H«C>Oin \, ocoiH>oc>HTjoc; c\» -^ in r- u u c D> : a) a o> 0. c ° i +-> c o o cu "T^OQi-H— to Wrr io ir cm vOvONOh-h-l^-h-COOOJ • ••••«♦••♦• CMOJCNJCMCVJCMCMCMCOCOCC no >0 • •••••••••« CMCMCMCNJCMCVlOJCMCvJOJCN OHCMlOOlfiOO • ••• •••OlOO OOOhhCMUJhhiM cp r> s> 122 OJ B a. o a. o < j> H 3 W OS Or- 6 NO 6 <0 9^ey SunurtQQ aAn B I 9 H so £ s c CO 1—1 a. c o CO »0 O OS •l-t 0) 2 o 123- 30 45 Time in Minutes 124 VITA The author, Calvin P. C, Poon, was born on November 8, 1935, in Canton, China. After finishing high school in Hong Kong, the author began his formal engineering education at the National Taiwan University, Taipei, Taiwan, in 1954, and obtained the degree of Bachelor of Science in Civil Engineering in 1958. In 1959 the author enrolled for advanced study at the Graduate School, University of Missouri, Columbia, Missouri, in the field of Sanitary Engineering. After receiving his Master of Science degree in June of 1960, he joined the firm of J. S. Watkins Consulting Engineers. In September of 1960 the author began further studies at the University of Illinois at Urbana towards the degree of Doctor of Philosophy in Sanitary Engineering. Recently, the author has obtained a position with the firm of Gannet, Fleming, Corddry and Carpenter, Inc., consulting engineers in Harrisburg, Pennsylvania.