The American Journal of‘ Medicine

VOL. VII SEPTEMBER, 1949 No. 3

Editorial

Disinfection of the Air with Triethylene Glycol Vapor / I

HE present premature large-scale coli and subtilis (vegetative fornl)Z a number commercialization of glycol vapors of common non-pathogens of the air and T for prevention of acute respiratory certain molds appear to bc susceptible to disease, together with the accompanying the action of the vapor.” No reports have propagation of much misinformation con- been made 01’ such stud& o11 air-l)ornc* cerning the use and effects of this form 01‘ tubercle bacilli. Naturally, occurring dust- aerial disinfection, make it seem particu- borne bacteria have been found to be much larly appropriate to review our knowledge more resistant to the killing action of the of the field. Furthermore, in the two and vapor than are those experimentally dis- one-half years intervening since a previous bersed into the air. summary of this subject,’ considerable new The recent development of certain quan- information has been acquired. titative technics for study of this subject has Among the many chemical compounds made possible much more precise experi- which have been tested as vapors or mists mentation, the results of which amplify for their lethal action on air-borne infecti- previous knowledge of the activity of tri- ous particles triethylene glycol still remains ethylene glycol vapor and provide new the agent of choice for use in environments interpretations. First, fundamental to a occupied by human beings. Germicidal more exact understanding of its germicidal concentrations of this vapor are odorless? effects was determination of the amounts of tasteless, non-irritating, non-toxic, invisible glycol vapor which could exist in the air at and have no deloterious effect on walls? varying humidities and temperatures. By fabrics, books or other objects in the treated means of a suitable method devised for this space. The presence of as little as 1 cc. of particular purpose4 curves were constructed vaporized triethylene glycol in several indicating the saturation concentrations of hundred million cc. of air is, under labora- triethylene glycol under conditions of rela- tory conditions, highly lethal for the com- tive humidity from 0 to 90 per cent and at mon respiratory bacteria, pathogenic and temperatures from 20”~. to 29”~. It was non-pathogenic, as well as for the viruses found that increasing humidity resulted in of influenza,’ psittacosis and meningopneu- tnonitis.2 Other bacteria including Bacillus pnrumonitis and psittacosis viruses with trirthylcnt glycol vapor. J. Exfier. Med., 85: 65-76, 1947. 1Roswrso~, 0. H. New methods for the control of 3B~~~, E. and MELLODY, M. Fungicidal action of air-borne infection with especial reference to the use of triethylene glycol. 3. Infect. Dis., 79: 45-56. 1946. triethylenr glyrol vapor. M’i.wvrsi?~ .M. ,7., 46: 31 l-31 7. d PUCK, T. T. and WISE, H. Stndies on vapor liquid 1947. equilibria. I. A new dynamic method for the determina- 2 ROSEBURY, T., MEIKLEJOHN, G., KINGSLAND, L. C. tion of vapor prrssure of liquids. J. P&r. Cb~wz., 50: 329. and ROLDT, M. H. Disinfection of clouds of meningo- 339, 1946.

293 294 Editorial a progressive although not straightline killed in two to three minutes; 80 per cent or diminution in quantity of glycol vapor re- more of them were destroyed in the first quired to saturate the air. Raising the tem- minute. At high relative humidities (60 to perature was found to increase the capacity 80 per cent) the rate of action was much of the air to hold glycol.5 With these data reduced but still appreciable. Likewise, at available chemical analyses of glycol-con- very low humidities (5 to 10 per cent) killing taining air could then be interpreted in was somewhat retarded. The more nearly relation to per cent saturation-a figure the concentration of the glycol vapor in the which has been found to be much more air approached the saturation value the significant in respect to bactericidal effect more rapid the kill. However, the increase than was the absolute quantity of glycol in in effectiveness at levels about 70 per cent the atmosphere. was slight.7 Another important advance was the Observations of an analogous nature on further development of the glycostat, or influenza virus in which white mice were glycometer, an instrument for measuring exposed to atmospheres containing freshly and controlling the concentration of tri- atomized virus and glycol vapors showed ethylene glycol vapor in the air.6 This in- that the presence of the glycol in the air at strument is extremely sensitive, being capa- saturations of 70 to 90 per cent afforded ble of detecting as little as 1 microgram of the mice complete protection against lethal glycol/L. of air under ordinary conditions concentrations of the virus. Protection was of humidity and temperature and responds optimum at relative humidities of 15 to 40 rapidly to changes in concentration of or 50 per cent, as was the case with bac- vapor. By means of attachment to a record- tericidal activity of the glycol. However, at ing device the glycometer provides a con- high humidities, 70 to 80 per cent, the tinuous record of the concentration of glycol glycol had relatively little effect as most of in the atmosphere in terms of per cent satu- the test mice died even when the air was ration. It can be used also to control the saturated or supersaturated with triethylene output of the glycol vaporizer. Unfortu- glycol. Under optimum conditions of humid- nately this apparatus is not yet commercially ity and glycol saturation the virucidal available. action of the vapor was found to be very Employing the glycostat in specially con- rapid.s structed experimental rooms in which at- Studies with dried bacteria in the form mospheres could be maintained at any of droplet nuclei or as a fine dust made from desired temperature and relative humidity, desiccated saliva suspensions have shown a large series of observations has been made them to be about as susceptible to the lethal on the bactericidal and virucidal action of action of triethylene glycol vapor as are varying concentrations of glycol vapor freshly atomized suspensions. The impor- under a wide range of environmental con- tance of this finding lies in the fact that most ditions. Optimum conditions for the rapid pathogenic bacteria present in an infected action of the glycol vapor at ordinary room atmosphere are probably in the dried temperatures were found to be relative state.8 humidities of 15 to 40 per cent and vapor These more recent observations bring saturations of 40 to 100 per cent. In such out the fact that, in adequate concentra- atmospheres freshly atomized bacteria were tions, triethylene glycol vapor is effective over a wider range of relative humidities 5WISE, H. and PUCK, T. T. The saturation concentra- tions of triethylene glycol vapor at various relative ~LESTER, W., ROBERTSON,0. H., PUCK, T. T., WISE, humidities and temperatures. Science, 105: 556-557, H. and SMITH, M. The rate of bactericidal action of 1947. triethylene glycol vapor on microorganisms dispersed in- s PUCK, T. T. An automatic dewpoint meter for the to the air in small droplets. Am. 3. Hyg.,to be published. determination of condensible vapors. Rev. Scient. Instru- *ROBERTSON, 0. H., LESTER, W. and DUNKLIN, E. ments, 19: 16-23, 1948. Unpublished experiments.

AMERtCAN JOURNAL OF MEDICINE Editorial 295 than was formerly thought to bc the case.* of air which goes on to some degree even It has not been found possible, however, to when windows and doors are closed. This sterilize the air of inhabited rooms. Numer- loss amounts to 70 to 90 per cent of the ous tests under a variety of conditions have glycol vaporized. 8 Hence in order to secure shown that the air-borne bacterial popula- and maintain adequate concentrations of tion can be reduced by not more than 70 to glycol in the air it is necessary to vaporize 75 per cent and the rate of killing is much usually four to five times the amount of slower than in the case of atomized bacterial triethylene glycol calculated to produce the suspensions. Investigation of dust-borne desired concentration in a given space. bacteria (non-pathogens with rare excep- When air exchange in the room is increased tions) indicates that the reason for their by opening of doors and windows, the relative insensitivity to the lethal effect glycol requirement is even higher. of triethylene glycol vapor lies in the physi- Air currents are necessary for uniform cal state of the bacterial particle which dispersion of glycol vapor throughout the obviously differs from that of the glycol- treated space. In moderate-sized rooms of sensitive micro-organisms desiccated under several thousand cu. feet the natural convec- laboratory conditions. tion currents are usually adequate for this The utilization of glycol vapor for pur- purpose. In larger spaces an electric fan or poses of aerial disinfection involves a num- two depending on the size and shape of the ber of considerations which may be out- space to be glycolized accomplishes the lined briefly as follows: since triethylene desired result. Air conditioning systems glycol has a very low vapor pressure (boiling offer the most satisfactory means of glycol point is 550”~.) heat is essential for vaporiza- vapor distribution. tion, However, the temperature which can Maintenance of an adequate concentra- be employed for this purpose is limited by tion of triethylene glycol vapor in the the fact that this glycol begins to decompose atmosphere offers the principal problem in at temperatures far below its boiling point. the field of practical application. Until These properties of triethylene glycol in- glycostats become available the only means troduce very definite requirements for the of knowing whether sufficient glycol vapor design of vaporizers. In order to disperse is present, aside from chemical determina- sufficient glycol into the air extended tion, is to produce a slight fog. The presence evaporating surfaces must be employed. of a mist, best determined by the Tyndall Heating of the pool or reservoir of glycol effect (from a focused flashlight) does not should be avoided. Certain glycol vapor- necessarily indicate saturation of the air. izers commercially available embody these A beam may be detectable at any concen- principles, others do not. tration above 50 per cent saturation depend- Continuous dispersion of the glycol vapor ing on the dustiness of the air. However. the into the treated space is essential because of occurrence of a Tyndall beam under ordi- the constant loss of vapor from the air, due nary conditions indicates the presence of a to invisible condensation on all surfaces germicidal concentration. The optimum (including dust in the air) and the exchange concentrations in terms of per cent satura- tion probably lie between 60 and 80 per * Certain earlier studies,g which indicated that dried cent. bacteria were relatively unsusceptible to glycol vapor action at humidities below 30 per cent, were carried out Prior to its use in environments inhabited before any means was available for measuring concentra- by human beings, prolonged tests (twelve tions of triethylcne glycol in the air. Subsequent de- to eighteen months) for chronic toxicity velopments have shown that the amounts of glycol used in those experiments which were calculated to saturatt were carried out on monkeys and rats ex- the air actually produced concentrations so low as to posed continuously to atmospheres satu- be almost negligible. rated with triethylene glycol vapor. In 9 ROBERTSON,0. H. Sterilization of air with glycol vapors. Harvey Lecture Series, 38: 227-234, 1942-43. none of this large group of animals nor in

SEPTEMBER, 1949 2g6 Editorial others receiving oral doses many hundreds Environmental conditions of the several of times the amount they could absorb tests differed markedly and it is quite likely from inhalation were any deleterious effects that opportunities for the transmission of observed either during life or in histologic infection by routes other than the air likc- sections of the organs following sacrifice at wise varied. This latter unknown, namely, termination of the experiments. lo Subse- the percentage of cases of the various quent to these tests thousands of individuals diseases of the respiratory tract that are air- have resided in glycol-containing atmos- borne makes the evaluation of any control pheres, many of them continuously for measure most difficult. In environments months, without apparent disturbance. The such as hospital wards adequately con- presence of triethylene glycol vapor in the trolled tests of the effect of aerial disinfec- air offers no fire or explosive hazard. tion on the incidence of diseases acquired Evidence of the effect of triethylene glycol by way of the respiratory tract must include vapor derived from practical application rigid enforcement of ward technics intended is as yet scanty. Tests of the lethal action to reduce the possibility of channels of in- of glycol vapor on pathogens demonstrable fection other than through the air. in the air have thus far been confined to From the foregoing summary it is appar- observations on hemolytic streptococci. In ent that the usefulness of triethylene glycol hospital wards housing patients with strep- vapor as a means of protection against tococcal respiratory tract infections the respiratory infection remains to be de- dispersion of triethylene glycol vapor re- termined. However, the fact that this vapor sulted in reduction of air-borne streptococci has been shown to be a potent aerial disin- of approximately 70 per cent. However, fectant, at least for freshly dispersed patho- when in addition to glycol, dust control gens, indicates the worthwhileness of more measures (oiling of bedding and floors) were clinical trials in a variety of environments. instituted, the reduction was increased to The kind of control employed in such tests 95 per cent. These findings bring out the will of necessity vary with the particular importance of dust suppression as an ad- environment and will have to be planned iunct to the use of triethylene glycol vapor. l1 for each type of population under study. The few studies that have been reported It is only through an accumulation of results on the clinical effects of employing glycol from controlled experiments of this kind vapor for the prevention of air-borne infec- that an evaluation of the procedure will be tion are much more difficult to evaluate. possible. The results of these experiments which were In conclusion, reference should be made carried on in hospitals and military barracks to the nationwide sale of glycol vaporizers. vary from marked reduction of infections in Two recent editorials in medical periodi- the glycolized areai to inconclusive effects.13 cals14*15 have dealt with this subject pointing I0 ROBERTSON, 0. H., LOSSLI, C. G., PUCK, T. T., out the unfortunate aspects of unrestrained LEMON, H. M., WISE, H. and LESTER, W. Tests for the exploitation of glycol vapors and the fact chronic toxicity of propylene glycol and triethylene glycol on monkeys and rats by vapor inhalation and that the majority of the vaporizing devices oral administration. 3. Phnrmncol. & Exper. Thunp., 71 : on the market are completely unsatisfac- 52-76, 1947. tory. A few mechanically sound and effici- ‘I PUCK, T. T., HAMBURGER, M., ROBERTSON, 0. H. and HURST, V. The effect of triethylene glycol vapor on ent vaporizers are being produced and it air-borne beta hemolytic streptococci in hospital wards. seems likely that some means of official II. The combined action of glycol vapor and dust control certification of such apparatus will soon be measures. 3. I$ect. Dis., 76: 216-225, 1945. available. I2 ROBERTSON, 0. H. Disinfection of air by germicidal vapors and mists. Am. j’. Pub. Health, 36: 390-391, 1946. 0. H. ROBERTSON, M.D. 1a LOOSLI, C. G., SMITH, M. H. D., GAUI.D, R., ROB- ” Commercial exploitation of glycol vaporizers. Am. j’. ERTSON, 0. H. and PUCK, T. T. Control of cross infections Pub. Health, 39: 222-224, 1949. in infants’ wards by the use of triethylene glycol vapor. I6 The sale of glycol vaporizers. Cirrcitrruti 3. Med., ‘4m. 3. Pub. Health, 37: 1385-1398, 1947. April, 1949.

AMERICAN JOURNAL OF MEDICINE