Embed Size (px)
Citation preview
Stephen N. RudnickJames J. McDevittMelvin W. FirstJohn D. Spengler
Air Transportation Center of Excellence for Airliner Cabin Environment ResearchHarvard School of Public HealthBoston, MA 02115
April 2009
Final Report
Inactivating Influenza Viruses on Surfaces Using Hydrogen Peroxide or Triethylene Glycol at Low Vapor Concentrations
DOT/FAA/AM-09/7Office of Aerospace MedicineWashington, DC 20591
OK-09-0434-JAH
Federal AviationAdministration
NOTICE
This document is disseminated under the sponsorship of the U.S. Department of Transportation in the interest
of information exchange. The United States Government assumes no liability for the contents thereof.
___________
This publication and all Office of Aerospace Medicine technical reports are available in full-text from the Civil Aerospace Medical Institute’s publications Web site:
www.faa.gov/library/reports/medical/oamtechreports
�
Technical Report Documentation Page 1. Report No. 2. Government Accession No. 3. Recipient's Catalog No.
DOT/FAA/AM-09/7 4. Title and Subtitle 5. Report Date
April 2009 Inactivating Influenza Viruses on Surfaces Using Hydrogen Peroxide or Triethylene Glycol at Low Vapor Concentrations 6. Performing Organization Code
7. Author(s) 8. Performing Organization Report No. Rudnick SN, McDevitt JJ, First MW, Spengler JD 9. Performing Organization Name and Address 10. Work Unit No. (TRAIS)
11. Contract or Grant No.
Air Transportation Center of Excellence for Airliner Cabin Environment Research Harvard School of Public Health 665 Huntington Avenue Boston, MA 02115
12. Sponsoring Agency name and Address 13. Type of Report and Period Covered
Office of Aerospace Medicine Federal Aviation Administration 800 Independence Ave., S.W. Washington, DC 20591
14. Sponsoring Agency Code 15. Supplemental Notes Work was accomplished under Public law 108-76 16. Abstract Any of the exposed surfaces in airplanes can become contaminated with infectious viruses, such as influenza, and facilitate transmission of disease. In this study, we disinfected surfaces contaminated with influenza viruses. Selection of the specific decontamination agents used in this study was based on three criteria: 1) no harm would be caused to the mechanical components or avionics of the airplane, 2) no potentially harmful residue would be left behind, and 3) the airplane could quickly be returned to service. We chose two decontamination agents that we believed fulfilled these criteria: 1) hydrogen peroxide (HP) at vapor concentrations in air below 100 ppm and 2) triethylene glycol (TEG) saturated air, which has a TEG vapor concentration of about 2 ppm at 25°C. For influenza viruses deposited on stainless-steel coupons and exposed for 15 min to 10 to 90 ppm of HP vapor, the number of log reductions of active viruses ranged from 3.6 to 4.7. The number of log reductions, however, was not linear with time; log reduction rate decreased significantly with increasing exposure time. For example, at a HP vapor concentration of 57 ppm, the number of log reductions was 3.2 after 2.5 min but just 4.0 after 10 min. Even after 60 min, the number of log reductions was only 5.6. At a HP vapor concentration of 10 ppm, the number of log reductions was 2.0 after 2.5 min. This corresponds to 99% inactivation of viruses, a significant reduction for such a low HP vapor concentration. For air saturated with TEG at 25-29°C, the number of log reductions of influenza viruses versus exposure time followed a linear relationship reasonably well. The decontamination rate was equal to 1.3 log reductions per hour. The potential for damage to the mechanical components or avionics of the airplane at a TEG vapor concentration of 2 ppm would be expected to be minimal. In addition, at a 2 ppm TEG vapor concentration, there is essentially no health risks to people.
17. Key Words 18. Distribution Statement
Surface Decontamination, Hydrogen Peroxide Vapor, Vaporized Hydrogen Peroxide, Triethylene Glycol Vapor, Influenza Virus
Document is available to the public through the Defense Technical Information Center, Ft. Belvoir, VA 22060; and the National Technical Information Service, Springfield, VA 22161
19. Security Classif. (of this report) 20. Security Classif. (of this page) 21. No. of Pages 22. Price Unclassified Unclassified 14
Form DOT F 1700.7 (8-72) Reproduction of completed page authorized
���
ACKNOWLEDGMENTS
The authors thank Dr. W�ll�am Gale for prov�d�ng gu�dance dur�ng th�s project and for h�s ass�stance �n
procur�ng sta�nless-steel surfaces used dur�ng exper�ments. We also thank the techn�c�ans who worked on
th�s project and contr�buted to �ts complet�on: G�han Amaras�r�wardena, Kenneth Ch�n, Al�ssa Fernald,
Chr�sta James, and Ch�ka Uch�yama.
v
ABBREVIATIONS
As used in this report, the following abbreviations/acronyms have the meanings indicated
AbbreviAtion MeAning ACGIH - -- -- Amer�can Conference of Governmental Industr�al Hyg�en�stsBSA - -- -- -- -- Bov�ne serum album�nDPBS - -- -- -- Dulbecco’s phosphate buffered sal�neDPBS++ - -- -- DPBS w�th calc�um and magnes�umEPA -- -- -- -- Env�ronmental Protect�on AgencyFFU -- -- -- -- Fluorescent focus un�tsHP - -- -- -- -- Hydrogen perox�deMDCK - -- -- Mad�n-Darby can�ne k�dneyOSHA -- -- -- Occupat�onal Safety and Health Adm�n�strat�onPEL -- -- -- -- Perm�ss�ble exposure l�m�tRH - -- -- -- -- Relat�ve hum�d�tySARS - -- -- -- Severe acute resp�ratory syndromeTEG -- -- -- -- Tr�ethylene glycolTLV® - -- -- -- Threshold L�m�t ValueUV - -- -- -- -- Ultrav�oletUVGI -- -- -- Ultrav�olet germ�c�dal �rrad�at�on
MATHEMATICAL SYMBOLS
As used in this report, the following symbols have the meanings indicated
SyMbol MeAning f -- -- -- -- -- -- Fract�on of v�ruses rema�n�ng act�ve
Ulog - -- -- -- Mean of the logar�thms of UoUlog -- -- -- Mean of the logar�thms of U
o
n - -- -- -- -- -- Number of log reduct�onsn - -- -- -- -- -- Average number of log reduct�onsp
TEG -- -- -- -- Part�al pressure of TEG �n a�r
P - -- -- -- -- -- Amb�ent pressureo
TEGP -- -- -- -- Vapor pressure of pure l�qu�d TEG Us log -- -- -- -- Standard dev�at�on of log U
oUs log -- -- -- -- Standard dev�at�on of log Uo
ns -- -- -- -- -- Standard dev�at�on correspond�ng to n T -- -- -- -- -- TemperatureU -- -- -- -- -- Number of FFU per volume of r�nsate from an exposed couponU
o- - - - - - - - - - - - - - - - - - - -Number of FFU per volume of r�nsate from an unexposed coupon
TEGy -- -- -- -- TEG mole fract�on �n the gas phase
v��
CONTENTS
INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Hydrogen Perox�de Vapor Decontam�nat�on . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Tr�ethylene Glycol Vapor Decontam�nat�on . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
METHOD AND MATERIALS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
Influenza V�rus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
Preparat�on and Treatment of Test Surfaces. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
Exposure Chamber and Test Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
Calculat�ons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
RESULTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
Hydrogen Perox�de Vapor Decontam�nat�on . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
Tr�ethylene Glycol Vapor Decontam�nat�on . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
V�ab�l�ty Tests at Amb�ent Cond�t�ons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
DISCUSSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Hydrogen Perox�de Vapor Decontam�nat�on . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Tr�ethylene Glycol Vapor Decontam�nat�on . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1
inActivAting influenzA viruSeS on SurfAceS uSing Hydrogen Peroxide or trietHylene glycol At low vAPor concentrAtionS
INTRODUCTION
BackgroundPrevent�ng the spread of smallpox, �nfluenza, SARS,
and other v�rus-caused d�seases on commerc�al a�rplanes and other publ�c venues �s a s�gn�ficant challenge for the publ�c health commun�ty (Musher 2003; Olsen, Chang, Cheung, et al. 2003). Transfer of v�ruses from an �nfected �nd�v�dual to an un�nfected one can occur through var�ous modes of transm�ss�on: 1) by d�rect contact, 2) v�a fom�tes (�nan�mate objects capable of carry�ng �nfect�ous v�ruses), 3) from the d�rect spray of large droplets from an �nfected person, and 4) from droplet nucle�, wh�ch are very small dr�ed droplets that can stay suspended �n the a�r for long per�ods of t�me (Roy & M�lton 2004). In th�s study, we evaluated the efficacy of var�ous relat�vely gentle methods for decon-tam�nat�ng fom�tes.
Any of the exposed surfaces �n a�rplanes or other veh�cles used for publ�c transportat�on can become con-tam�nated w�th �nfect�ous v�ruses and be respons�ble for d�sease transm�ss�on. In th�s study, we d�s�nfected surfaces contam�nated w�th �nfluenza A v�ruses, whose subtypes may have the potent�al to cause a pandem�c propagated worldw�de by commerc�al travel. It may not be necessary to ster�l�ze an a�rplane cab�n; s�gn�ficant reduct�on �n the potent�al for d�sease transm�ss�on would be benefic�al.
Select�on of spec�fic decontam�nat�on methods used �n th�s study was based pr�mar�ly on three cr�ter�a: 1) the method would not be expected to cause damage to the mechan�cal components or av�on�cs of the a�rplane, 2) the method would leave no potent�ally harmful res�due, and 3) the method would requ�re a relat�vely br�ef per�od of t�me, so that an a�rplane or other means of publ�c transportat�on could be put back �nto serv�ce qu�ckly. We chose three methods that fulfill these cr�ter�a: 1) relat�vely low vapor concentrat�ons (<100 ppm) of hydrogen perox�de (HP), 2) very low vapor concentrat�ons of tr�ethylene glycol (TEG), and 3) thermal decontam�nat�on us�ng heated a�r. The efficacy of the decontam�nants HP and TEG �s the subject of th�s report. A separate report d�scusses thermal decontam�nat�on. As a basel�ne for compar�son and because of �ts �mportance, the length of t�me that �nfluenza v�ruses rema�n act�ve on surfaces under amb�ent cond�t�ons was also determ�ned.
We el�m�nated many other decontam�nat�on methods because they d�d not adhere to the des�red cr�ter�a. Spec�fi-cally, methods that rely on chlor�ne d�ox�de, formalde-hyde, ethylene ox�de, and methyl brom�de were deemed l�kely unacceptable due to the�r potent�al for damage to the a�rplane and the�r tox�c�ty to humans. In add�t�on, desp�te �ts effect�veness for �nact�vat�ng v�ruses and bac-ter�a, we d�d not choose ultrav�olet germ�c�dal �rrad�at�on (UVGI) because v�ruses can be protected from UV rays by lurk�ng �n the shadows and because the UV rays are not very penetrat�ng, allow�ng a coat�ng of dust or other mater�al to protect v�ruses from the UV rays. UVGI �s more appropr�ate for a�r decontam�nat�on (F�rst, Rudn�ck, Banahan, et al. 2007; Rudn�ck & F�rst 2007).
Hydrogen Peroxide Vapor DecontaminationWe were unable to find any peer-rev�ewed publ�ca-
t�ons on surface decontam�nat�on of �nfluenza v�ruses us�ng HP at vapor concentrat�ons below 100 ppm. In the few publ�shed stud�es on surface decontam�nat�on of �nfluenza v�ruses at h�gher HP concentrat�ons, the dr�ed v�rus suspens�on was exposed to a relat�vely large dose of HP vapor; that �s, the HP vapor concentrat�on was relat�vely h�gh and exposure t�me relat�vely long. None of these stud�es gave results on v�rus �nact�vat�on versus dose. For example, the effect of HP vapor on surface-depos�ted �nfluenza v�ruses was evaluated by Heckert, Best, Jordan, et al. (1997) at a HP vapor concentrat�on of about 1200 ppm and exposure t�me of about 30 m�n. Although Heckert, Best, Jordan, et al. showed an overall reduct�on of �nfluenza v�ruses of about 6 logs, wh�ch was the�r l�m�t of detect�on, only about 3 logs were at-tr�butable to HP vapor; the rema�nder was due to 16 h of dry�ng at amb�ent cond�t�ons and heat exposure at 30-40°C. Other stud�es us�ng �nfluenza as the challenge v�ruses had s�m�lar l�m�tat�ons. In a recent rev�ew art�cle, De Bened�ct�s, Beato, and Capua (2007) conclude that “reports on the spec�fic efficacy aga�nst av�an �nfluenza v�ruses of hydrogen perox�de are contrad�ctory, and for th�s reason add�t�onal �nformat�on on �ts v�r�c�dal efficacy �s necessary.”
Triethylene Glycol Vapor DecontaminationAlthough we were unable to find any publ�cat�ons on
the use of TEG vapor to decontam�nate surfaces, �t has
2
been used to d�s�nfect a�r. Although other glycols can also be used to d�s�nfect a�r (Robertson, B�gg, M�ller, et al. 1941; Robertson, Loosl�, Puck, et al. 1941), TEG vapor �s the most su�table because of �ts extremely low vapor pressure, wh�ch results �n very low a�r concentrat�ons. In add�t�on, TEG �n aerosol form �s commonly used for theatr�cal “smoke” such as �n venues for Broadway produc-t�ons (Burr, Van G�lder, Trout, et al. 1994). As a result of th�s and other propert�es, TEG vapor �s bel�eved to do no harm to humans (EPA 2005a) or damage to env�ronmental surfaces (Lester, Kaye, Robertson, et al. 1950). There are a large number of journal publ�cat�ons, pr�mar�ly from the 1940s, on the use of TEG vapor for a�r d�s�nfect�on. TEG vapor has been shown to exert lethal act�on aga�nst a w�de var�ety of a�rborne �nfect�ous agents �nclud�ng bacter�a, v�ruses, and fung� (Lester, Kaye, Robertson, et al. 1950). In part�cular, TEG vapor was found to be an effect�ve decontam�nant agent for a�rborne �nfluenza v�ruses (Robertson, Puck, Lemon, et al. 1943).
METHOD AND MATERIALS
Influenza VirusVirus Stock. A frozen suspens�on of �nfluenza v�ruses
(A/PR/8/34 H1N1) was purchased from Advanced B�o-technolog�es (Columb�a, MD). It was thawed, d�v�ded �nto s�ngle-use packets, refrozen, and stored at −80°C unt�l needed.
Virus Assay. A fluorescent focus reduct�on assay (Hartshorn, Wh�te, Tecle, et al. 2007) was used to measure the t�ter of v�rus suspens�ons before and after decontam�nat�on. Confluent monolayers of Mad�n-Darby can�ne k�dney (MDCK) cells were prepared �n 96-well plates. Each well was �noculated w�th 50 μL of coupon r�nsate (or ser�al d�lut�ons of the r�nsate) and �ncubated at 37°C for 45 m�n �n a 5% CO
2 env�ronment. After
wash�ng the �nfected cells us�ng assay med�a composed of Dulbecco’s Mod�fied Eagle’s Med�um (Med�atech, Herndon, VA) w�th 0.1% bov�ne serum album�n (BSA) (SeraCare, M�lford, MA), the cells were �ncubated for 7 h at 37°C �n a CO
2 env�ronment. After �ncubat�on, the
cells were washed three t�mes w�th Dulbecco’s phosphate buffer sal�ne w�th calc�um and magnes�um (DPBS++) (Hyclone Laborator�es, Logan, UT) and fixed w�th an aqueous solut�on of 80% acetone for 10 m�n at 4°C. The �nfected MDCK cells were then labeled for 30 m�n at 4°C w�th 50 μL of nucleoprote�n ant�body solut�on, wh�ch was made by add�ng 50 μL of mouse monoclonal ant�bod�es (Centers for D�sease Control, Atlanta, GA, catalog #VS2366) to 5 mL of Dulbecco’s phosphate buffer sal�ne (DPBS) (Hyclone Laborator�es, Logan, UT) conta�n�ng 1% BSA, 1% heat �nact�vated human serum (Med�atech, Herndon, VA), and 0.02% sod�um
az�de. After wash�ng three t�mes w�th DPBS, the cells were �ncubated w�th tagg�ng solut�on, wh�ch was made by add�ng 50 μL of rhodam�ne-labeled goat ant�-mouse IgG (Jackson ImmunoResearch Laborator�es, West Grove, PA, catalog #115026062) to 5 mL of DPBS conta�n�ng 1% BSA, 1% heat �nact�vated human serum, and 0.02% sod�um az�de. The number of cells hav�ng a fluorescent foc�, wh�ch are referred to as fluorescent focus un�ts (FFU), were then counted us�ng an Olympus CKX-41 �nverted fluorescent m�croscope (Olympus, Center Valley, PA). Based on th�s assay, the t�ter of the s�ngle-use packets of �nfluenza v�rus suspens�on after be�ng thawed was about 109 FFU/mL.
Preparation and Treatment of Test SurfacesOne-�nch by three-�nch sta�nless-steel coupons were
used as test surfaces. For tests us�ng h�gher concentra-t�ons of HP vapor or TEG vapor, 50 μL of �nfluenza v�rus suspens�on was seeded onto a predeterm�ned number of coupons. For lower decontam�nant concentrat�ons, the �nfluenza v�rus suspens�on was d�luted pr�or to be�ng seeded onto coupons. All of the coupons were placed �nto a b�olog�cal safety cab�net unt�l the depos�ted l�qu�d had evaporated. The requ�red dry�ng t�me was between 20 and 30 m�n, depend�ng on amb�ent cond�t�ons. Some of these seeded coupons, wh�ch are referred to as “control” coupons, were left �n the b�olog�cal safety cab�net, where they cont�nued to be exposed to filtered room a�r at amb�ent cond�t�ons. The rema�n�ng seeded coupons (along w�th a clean coupon used as a negat�ve control) were placed �n an exposure chamber conta�n�ng HP vapor or TEG vapor.
The coupons were removed from the exposure chamber after predeterm�ned exposure t�mes. The control cou-pons that were left �n the b�olog�cal safety cab�net had no exposure to HP or TEG and, thus, were cons�dered to be unexposed coupons. Immed�ately after the last coupon was removed from the exposure chamber, each of the seeded coupons was r�nsed w�th DPBS++ us�ng the follow�ng procedure: The clearly marked port�on of the coupon where the v�ruses had been �n�t�ally depos-�ted was r�nsed 25 t�mes w�th a s�ngle 500-μL port�on of DPBS++ us�ng a p�pette. No v�s�ble res�due rema�ned. A fluorescent focus reduct�on assay was then done on the r�nsate and/or d�luted r�nsate from each coupon. Only about 50% of the �nfluenza v�ruses that were seeded onto the control sl�des were recovered.
Exposure Chamber and Test MethodologyHydrogen Peroxide Vapor Tests. Tests �n wh�ch
�nfluenza v�ruses depos�ted on sta�nless-steel coupons were exposed to HP vapor were done �n a 130-L cub�-cal plex�glass chamber located w�th�n a laboratory fume
3
hood. A shallow pool of an aqueous solut�on of 35% HP (VWR, West Chester, PA), d�luted w�th water to a predeterm�ned HP concentrat�on that was calculated to prov�de the des�red HP vapor concentrat�on, covered much of the floor area of the chamber. Pred�ct�on of the HP concentrat�on �n an aqueous solut�on requ�red for a spec�fic HP vapor concentrat�on was based on publ�shed correlat�ons (Schumb, Satterfield & Wentworth 1955).
The a�r �ns�de the chamber was kept well m�xed through the use of two small fans. To ma�nta�n the des�red rela-t�ve hum�d�ty (RH), 17 L/m�n of dry a�r was added to the chamber. Temperature and RH were mon�tored and recorded every 30 s us�ng a HOBO (Onset Computer Corp., Buzzards Bay, MA). RH and temperature were also measured per�od�cally w�th a hygrometer (Omega Eng�-neer�ng, Stamford, CT) and mercury thermometer.
HP vapor concentrat�on was mon�tored cont�nuously and data logged us�ng a newly purchased, cal�brated ATI C16 PortaSenII w�th a hydrogen perox�de sensor (Analyt�-cal Technology, Collegev�lle, MD), wh�ch has a measure-ment range of 0 to 100 ppm HP vapor. The cal�brat�on was done on 9/6/07, d�rectly before our exper�mental tests began; the manufacturer stated that the cal�brat�on had ±10% accuracy. D�rectly after our exper�mental tests were completed (1/30/08), the �nstrument was sent back to the manufacturer to be re-cal�brated. The �nstrument was read�ng 17% h�gher than �t should have been—aga�n, w�th�n ±10% accuracy. Nevertheless, w�th�n the accuracy of the cal�brat�on method, the �nstrument’s cal�brat�on rema�ned stable dur�ng our exper�mental tests.
After a constant HP vapor concentrat�on had been reached �n the exposure chamber, test coupons were �nserted �nto the chamber through a vert�cally open�ng sl�d�ng door. When �nsert�ng test coupons, the door was l�fted only very sl�ghtly so that the HP vapor concentra-t�on would rema�n essent�ally constant.
Triethylene Glycol Vapor Tests. Tests �n wh�ch �nfluenza v�ruses depos�ted on sta�nless-steel coupons were exposed to TEG vapor were also done �n the same well-m�xed 130-L cub�cal plex�glass chamber used for the HP vapor tests. Greater care, however, was taken to seal the chamber, and dry a�r was not added to the chamber. A shallow pool of 99% pure l�qu�d TEG (VWR, West Chester, PA) covered much of the floor area of the cham-ber. A beaker of water was placed w�th�n the chamber �n order to help ma�nta�n a reasonably constant RH. If the beaker of water was not present, the RH �n the chamber would decrease over t�me because TEG �s very hydroscop�c. Temperature and RH were mon�tored and recorded every 30 s us�ng a HOBO (Onset Computer Corp., Buzzards Bay, MA). RH and temperature were also measured per�od�cally w�th a hygrometer (Omega Eng�neer�ng) and mercury thermometer. After equ�l�br�um cond�t�ons
had been ach�eved, test coupons were �nserted �nto the chamber by m�n�mally open�ng a vert�cally sl�d�ng door. The coupons were �nserted as qu�ckly as poss�ble so as to m�n�m�ze d�srupt�on of equ�l�br�um cond�t�ons.
The concentrat�on of TEG vapor was not measured. Because we allowed a large pool of nearly pure l�qu�d TEG located on the floor of the well-m�xed exposure chamber to reach equ�l�br�um w�th the gas phase, the a�r was essent�ally saturated w�th TEG, and the part�al pres-sure ( TEGp ) of TEG was approx�mately equal to �ts vapor pressure ( o
TEGP ). TEG vapor pressure can be calculated from the Anto�ne equat�on (NIST 2005): (1)
where vapor pressure �s �n bars and temperature (T) �s �n degrees Kelv�n. Based on Equat�on 1, the vapor pres-sure of TEG at 25°C �s 0.00131 mm Hg1, wh�ch �s �n nearly perfect agreement w�th the value of 0.00132 mm Hg at 25°C g�ven by the EPA (2005a). The TEG mole fract�on ( TEGy ) �n the gas phase can be calculated from Dalton’s law:
(2)
where P �s amb�ent pressure. In actual�ty, because l�qu�d TEG �s so hydroscop�c, the pool of TEG on the chamber floor would tend to become d�luted w�th water over t�me so that the mole fract�on of TEG vapor would be some-what less than 1.7 ppm. However, because the amount of l�qu�d TEG �n the chamber was relat�vely large, d�lut�on would not be expected to have a very s�gn�ficant effect on the TEG vapor concentrat�on.
Tests of Natural Die Off Rate. Our normal meth-odology for evaluat�ng the loss of v�rus act�v�ty over t�me �nvolved seed�ng 50 μL of �nfluenza v�rus suspens�on onto each of the coupons that were to be used dur�ng an exper�mental test. All v�rus assays for an exper�mental test were then done at the same t�me and, when poss�ble, �n the same 96-well plate. Because grow�ng and ma�nta�n�ng cells �s somewhat of an art, perform�ng all assays for an exper�mental test at the same t�me �s �mportant �n order to get cons�stent results.
For tests to measure the natural d�e-off rate of �nflu-enza v�ruses, th�s methodology could not be used because the durat�on of the test was too long, so an alternat�ve procedure was employed. In preparat�on for an exper�-mental test to measure the natural d�e-off rate of �nfluenza v�ruses, s�ngle-use packets of �nfluenza v�rus suspens�on
1Although Equat�on 1 was spec�f�ed for a temperature range that d�d not �nclude 25°C, �t pred�cted the same vapor pressure at 25°C as was g�ven by the EPA (2005a).
ppm7.1107600013.0 6 =×=
≅=P
PP
py
oTEGTEG
TEG
299.13715757.6log10 −
−=T
P oTEG
4
were thawed, separated �nto 200-μL port�ons, and then re-frozen at –80°C. At the start of a natural d�e-off test, one of these 200-μL port�ons was thawed, and then each of three sta�nless-steel coupons was seeded w�th 50 μL of v�rus suspens�on pr�or to be�ng exposed to amb�ent cond�t�ons �n a small chamber w�thout a decontam�na-t�on agent present. After a predeterm�ned amount of t�me, another 200-μL port�on was thawed, and three add�t�onal coupons were seeded and then exposed to amb�ent cond�t�ons. Th�s procedure was repeated mul-t�ple t�mes. At the end of the test, the v�rus res�due on each coupon was extracted us�ng our standard method, and each extract�on was assayed at the same t�me and �n the same plate.
Calculations The number of fluorescent focus un�ts (FFU) per
volume of coupon r�nsate �s a measure of the quant�ty of cultureable v�ruses present on the coupon. The rat�o of the number of FFU per volume �n the r�nsate from an exposed coupon (U) to the number from an unexposed coupon (U
o) �s defined as the fract�on of v�ruses rema�n-
�ng act�ve (-f-):
oUUf = (3)
For cultureable v�ruses, the number of log reduct�ons (n) �s equal to the d�fference between the logar�thm of the �n�t�al FFU per volume and the logar�thm of the final FFU per volume: fUUn o logloglog −=−= (4)
where the logar�thms are to the base 10. Thus, n = 4 corresponds to 4 log reduct�ons, wh�ch �s equ�valent to 0.01% of the v�ruses rema�n�ng act�ve and 99.99% of the v�ruses �nact�vated; that �s, start�ng w�th 10,000 FFU �n the r�nsate from an unexposed coupon, only one would rema�n �n the r�nsate from the exposed coupon.
Because three coupons were exposed and three were not exposed dur�ng a spec�fic t�me per�od, the unexposed and exposed coupons could not be separated �nto pa�rs. Therefore, the mean number of log reduct�ons ( n ) was calculated from Equat�on 5: UUn o loglog −=
where oUlog and Ulog are the means of the logar�thms of U
o and U, respect�vely. The standard dev�at�on ( ns )
correspond�ng to n can be calculated from the standard dev�at�on of log U
o (
oUs log ) and the standard dev�at�on of U ( Us log ): 2
log2log
2UoUn sss +=
In figures �n wh�ch the number of log reduct�ons ( n ) �s plotted versus t�me, the error bars correspond to
±1 standard dev�at�on (sn), as g�ven by Equat�on 6.
Based on 109 FFU/mL for the �nfluenza v�rus suspen-s�on �n a s�ngle-use packet and 50% recovery of v�ruses from control sl�des, the theoret�cal l�m�t of detect�on �n terms of the number of log reduct�ons that could be detected by the methods descr�bed above was calculated to be 7.4. Th�s calculat�on �s based on the assumpt�on that a s�ngle FFU detected from any of the three coupons exposed at a spec�fic test cond�t�on corresponds to the l�m�t of detect�on.
RESULTS
Hydrogen Peroxide Vapor DecontaminationThe number of log reduct�ons based on Equat�on 5
versus exposure t�me for tests �n wh�ch �nfluenza v�ruses depos�ted on sta�nless-steel coupons were exposed at approx�mately 25°C and 58-65% RH to relat�vely low concentrat�ons of HP vapor �s shown �n F�gure 1. In th�s figure, error bars correspond to ± one standard dev�at�on, based on Equat�on 6. Even at a HP vapor concentrat�on as low as 10 ppm, about a two-log reduct�on was observed after 2.5 m�n of exposure. The reduct�on, however, d�d not �ncrease as much as would be expected w�th �ncreases �n e�ther exposure t�me or HP vapor concentrat�on. If a HP vapor concentrat�on of 10 ppm and an exposure t�me of 2.5 m�n are taken as the base, �ncreas�ng exposure t�me by a factor of s�x or concentrat�on by a factor of n�ne added only an extra 1.6 and 1.3 logs of reduct�on, respect�vely. For 15 m�n of exposure t�me, the h�ghest measured decontam�nat�on rate was 4.7 log reduct�ons at a HP vapor concentrat�on of 90 ppm. An add�t�onal test, not shown �n F�gure 1, �n wh�ch �nfluenza v�ruses were exposed at a HP vapor concentrat�on of 57 ppm for 60 m�n, resulted �n a decontam�nat�on rate of 5.6 log reduct�ons.
Triethylene Glycol Vapor DecontaminationThe number of log reduct�ons based on Equat�on
5 as a funct�on of exposure t�me for tests �n wh�ch �nfluenza v�ruses depos�ted on sta�nless-steel coupons were exposed to a�r saturated w�th TEG at 25-29°C and 45-55% RH �s shown �n F�gure 2. Based on Equa-t�ons 1 and 2, the concentrat�on of TEG vapor �n these tests was equal to between 1.7 and 2.5 ppm. Error bars �n F�gure 2 correspond to ± one standard dev�at�on, based on Equat�on 6. The number of log reduct�ons (n) versus exposure t�me (t) follows a l�near relat�on-sh�p reasonably well. The relat�onsh�p �s g�ven by the follow�ng equat�on: n-= 1.31t (7)where exposure t�me �s �n hours. Thus, the decontam�na-t�on rate attr�butable to TEG vapor was 1.3 log reduct�ons
(5)
(6)
5
0
1
2
3
4
5
0 0.05 0.1 0.15 0.2 0.25
Exposure Time, hours
Num
ber o
f Log
Red
uctio
ns
10 ppm20 ppm39 ppm57 ppm90 ppm
57 ppm
39 ppm
10 ppm
90 ppm
20 ppm
2.5 min 5 min 15 min10 min
Figure 1. Surface decontamination of influenza viruses with hydrogen peroxide vapor
n = 1.31 tr 2 = 0.78
-1
0
1
2
3
0 0.5 1 1.5 2Exposure Time (t), hours
Num
ber o
f Log
Red
uctio
ns (n
)
Figure 2. Surface decontamination of influenza viruses with triethylene glycol saturated air
per hour. Equat�on 7 �s equ�valent to Equat�on 8, the equat�on for exponent�al decay of the fract�on of v�ruses rema�n�ng act�ve ( f ) :
f-= exp(–3.02t) (8)
Viability Tests at Ambient ConditionsFor purposes of compar�son w�th chem�cal decontam�-
nat�on tests, the natural d�e-off rate at amb�ent cond�t�ons of �nfluenza v�ruses depos�ted on sta�nless-steel coupons was measured. The number of log reduct�ons versus t�me for two separate tests, each last�ng a few days, �s plotted �n F�gure 3.
Error bars �n th�s figure correspond to ± one standard dev�at�on based on Equat�on 6. No decontam�nat�on agent was used dur�ng these tests. Based on the data po�nts from both tests, the number of log reduct�ons (n) versus exposure t�me (t) follows a l�near relat�onsh�p g�ven by Equat�on 9: tn 0829.0= (9)where exposure t�me �s �n hours. Thus, the natural decay rate of �nfluenza v�ruses was 0.083 log reduct�ons per hour, wh�ch �s equ�valent to a half-l�fe of 3.6 h.
6
DISCUSSION
Hydrogen Peroxide Vapor DecontaminationTest results on the decontam�nat�on of �nfluenza v�ruses
us�ng HP vapor at concentrat�ons less than 100 ppm (F�g. 1) are somewhat surpr�s�ng �n that the number of log reduct�ons �n act�ve v�ruses versus exposure t�me �s very non-l�near; that �s, the fract�on of v�ruses rema�n�ng act�ve versus exposure t�me does not follow an exponent�al decay curve. As the exposure t�me �ncreases, the log reduct�on rate decreases s�gn�ficantly; thus, as shown �n F�gure 1, the number of log reduct�ons for the �n�t�al 2.5 m�n of exposure �s greater than the number of log reduct�ons from 2.5 to 15 m�n of exposure. Th�s trend �s true for all HP vapor concentrat�ons evaluated. For example, at a HP vapor concentrat�on of 10 ppm, the lowest concentra-t�on tested, the number of log reduct�ons was 2.0, 3.1, 3.4, and 3.6 after 2.5, 5, 10, and 15 m�n of exposure, respect�vely. S�m�larly, at a HP vapor concentrat�on of 90 ppm, the h�ghest concentrat�on tested, the number of log reduct�ons was 3.2, 4.5, and 4.7 after 2.5, 10, and 15 m�n of exposure, respect�vely.
Another unexpected outcome of these tests was that, �n the �n�t�al 2.5 m�n of exposure to 10 ppm HP vapor, the number of log reduct�ons was equal to two, wh�ch �s a 99% v�rus reduct�on. If the number of log reduct�ons at 10 ppm HP vapor versus exposure t�me was l�near, 15 m�n of exposure would result �n ster�l�zat�on (12 log reduct�ons). Instead, due to the nonl�near�ty of the curves, after 15 m�n of exposure to 10 ppm HP vapor, only 3.6 log reduct�ons were measured. Nevertheless, th�s �s a s�gn�ficant reduct�on for such a low HP vapor concentrat�on. The Threshold L�m�t Value (TLV®) and OSHA-perm�ss�ble exposure l�m�t (PEL) for occupat�onal HP vapor exposure �s an e�ght-hour t�me-we�ghted average
of 1 ppm (ACGIH 2008; OSHA 2008). Th�s suggests that 10 ppm of HP vapor �s a relat�vely safe concentrat�on over a short t�me per�od, although the TLV �ncludes the caveat that HP vapor �s a “confirmed an�mal carc�nogen w�th unknown relevance to humans.”
Triethylene Glycol Vapor DecontaminationD�v�d�ng Equat�on 7 by Equat�on 9 �nd�cates that
TEG vapor �ncreases the natural d�e-off rate of �nfluenza v�ruses by a factor of 16. The decontam�nat�on rate for a�r saturated w�th TEG vapor at 25-29°C, wh�ch was measured to be 1.3 log reduct�ons per hour, �s cons�der-ably less than for HP vapor (F�g. 1), even at a concentra-t�on of 10 ppm. For example, for a 15-m�nute exposure per�od, the decontam�nat�on rate for TEG vapor was 0.33 log reduct�ons, compared to 3.6 log reduct�ons for HP vapor. Nevertheless, TEG vapor has some �mportant advantages.
For surface decontam�nat�on us�ng TEG vapor, am-b�ent a�r or warmed a�r could eas�ly be saturated w�th TEG pr�or to be�ng �ntroduced �nto an a�rplane cab�n. Alternat�vely, m�crometer-s�ze TEG droplets, wh�ch evaporate rap�dly, could be �njected �nto the supply a�r duct or d�rectly �nto the cab�n a�r. The standard method for �ntroduc�ng TEG droplets �nto a�r for the purpose of a�r decontam�nat�on �s through the use of a pressur�zed l�qu�d (EPA 2005a), although a nebul�zer could also be used. If a pandem�c were to occur, both surface and a�r decontam�nat�on could take place s�multaneously, even wh�le passengers were onboard.
Although an object�on could be ra�sed due to the potent�al health r�sk of us�ng TEG vapor for a�r decon-tam�nat�on, th�s �s l�kely an unwarranted concern because TEG �s an odorless chem�cal of no known tox�c�ty, and exposure of people to TEG �s already w�despread. TEG vapor �s used as a bacter�ostat to k�ll odor-caus�ng bacter�a for the purpose of a�r san�tat�on and deodor�zat�on. It was first reg�stered for use �n hosp�tals as an a�r d�s�nfectant �n 1947. Present appl�cat�on scenar�os �nclude spray�ng TEG �ns�de offices, schools, hotels, lobb�es, theaters, recept�on rooms, sleep�ng rooms, bathrooms, and hosp�tal rooms (EPA 2005b). In add�t�on, products conta�n�ng TEG packaged �n aerosol cans and des�gned to be sprayed �nto the a�r �ns�de homes to control odors are sold �n stores everywhere (e.g., Oust® and Febreze®).
Accord�ng to the U.S. Env�ronmental Protect�on Agency (EPA 2005a), “the Agency has no r�sk concerns for tr�ethylene glycol w�th respect to human exposure. Based on a rev�ew of the ava�lable tox�cology data, the Agency has concluded that tr�ethylene glycol �s of very low tox�c�ty by the oral, dermal, and �nhalat�on routes
n = 0.0829 t
0
1
2
3
4
5
6
0 10 20 30 40 50 60 70 80Exposure Time (t), hours
Num
ber o
f Log
Red
uctio
ns (n
)
7/14/2008
4/17/2007
best-fit line(r = 0.82)2
Figure 3. Natural die-off of influenza viruses on stainless-steel coupons at ambient conditions
7
of exposure. The tox�cology database �s adequate to character�ze the hazard of tr�ethylene glycol, and no data gaps have been �dent�fied. There are no �nd�cat�ons of spec�al sens�t�v�ty of �nfants or ch�ldren result�ng from exposure to tr�ethylene glycol.” In add�t�on, TEG has no known deleter�ous effects on fabr�cs or other surfaces (Lester, Kaye, Robertson & Dunkl�n 1950). Unl�ke HP vapor, TEG vapor �s not an ox�d�z�ng agent. TEG �nact�-vates v�ruses and bacter�a because �t �s very hydroscop�c; �t condenses on bacter�a- and v�rus-conta�n�ng part�cles unt�l the TEG concentrat�on becomes suffic�ently h�gh to be germ�c�dal (Puck, T.T. 1947a; Puck, T.T. 1947b).
To demonstrate the effect�veness of s�multaneous surface and a�r decontam�nat�on, m�crob�olog�cal stud�es need to be conducted �n a room-s�ze chamber us�ng both v�ruses and bacter�a as challenges. Methods also need to be developed to mon�tor TEG vapor concentrat�on so that TEG �ntroduct�on can be prec�sely controlled.
It �s reasonable to expect that the efficacy of TEG vapor w�ll �ncrease as �ts concentrat�on �s �ncreased. However, at 25°C, the concentrat�on of TEG �n a�r cannot exceed 1.7 ppm because a�r �s saturated at that concentrat�on. The only way of �ncreas�ng the concentrat�on �s to �ncrease temperature. As shown �n Table 1, wh�ch was calculated from Equat�ons 1 and 2, modest �ncreases �n temperature result �n s�gn�ficant �ncreases �n TEG vapor concentra-t�on. Thus, further work �nvest�gat�ng TEG vapor as a decontam�nat�ng agent �s warranted. Spec�fically, the effect�veness of TEG vapor for surface decontam�na-t�on at h�gher concentrat�ons—that �s, at temperatures greater than room temperature—should be determ�ned. In add�t�on, the �nfluence of RH on decontam�nat�on effect�veness should also be evaluated.
CONCLUSIONS
Our exper�ments show that HP vapor concentrat�ons as low as 10 ppm and TEG vapor concentrat�ons of 2 ppm can prov�de effect�ve decontam�nat�on of a commerc�al a�rplane cab�n. At these very low concentrat�ons, the potent�al for damage to the mechan�cal components or av�on�cs of the a�rplane would be expected to be m�n�mal. Although �t has somewhat lower efficacy than 10 ppm HP vapor, a�r saturated w�th TEG vapor at 25°C �s probably the better cho�ce for decontam�nat�on of a�rplane cab�ns; TEG �s safer w�th regards to both personnel and a�rplanes and �s eas�er to apply. If TEG vapor �s determ�ned to be a v�able cand�date for decontam�nat�on of a�rplane cab�ns, opt�m�z�ng temperature and RH would l�kely lead to greater efficacy.
REFERENCES
ACGIH (2008). TLVs-and-BEIs, Amer�can Conference of Governmental Industr�al Hyg�en�sts, C�nc�n-nat�, OH.
Burr, G.A., Van G�lder, T.J., Trout D.B., W�lcox, T.G., and Dr�scoll, R.M. (1994). Health hazard evalua-t�on, HETA 90-0355-2449. Nat�onal Inst�tute of Occupat�onal Health and Safety.
De Bened�ct�s, P., Beato, M.S., and Capua, I. (2007). Inact�vat�on of av�an �nfluenza v�ruses by chem�cal agents and phys�cal cond�t�ons: a rev�ew. Zoonoses-Public-Health, 54(2), 51-68.
EPA (2005a). Reregistration- eligibility- decision- for- tri -ethylene- glycol, EPA739-R-05-002, Un�ted States Env�ronmental Protect�on Agency.
EPA (2005b). Revised-antimicrobials-division’s-review-of-the-disciplinary-sciences-for-issuance-of-the-reregistration-eligibility- decision- (RED) document, EPA-HQ-OPP-2005-0002, Un�ted States Env�ronmental Protect�on Agency.
F�rst, M.W., Rudn�ck, S.N., Banahan, K.F., V�ncent, R.L., and Br�ckner, P.W. (2007). Fundamental factors affect�ng upper room ultrav�olet germ�c�dal �rrad�at�on—Part I. Exper�mental, J-Occup-Environ-Hyg,-4(5), 321-31.
Hartshorn, K.L., Wh�te, M.R., Tecle, T., Tornoe, I., Sorensen, G.L., Crouch, E.C., and Holmskov, U. (2007). Reduced �nfluenza v�ral neutral�z�ng act�v�ty of natural human tr�mers of surfactant prote�n D. Respir-Res, 8(1), 9-19.
Temperature, °C Concentration, ppm20 1.125 1.730 2.835 4.440 6.9
Table 1. Triethylene Glycol Concentration in Air
8
Heckert, R.A., Best, M., Jordan, L.T., Dulac, G.C., Ed-d�ngton, D.L., and Sterr�tt, W.G. (1997). Eff�cacy of vapor�zed hydrogen perox�de aga�nst exot�c an�mal v�ruses. Appl-Environ-Microbiol, 63(10), 3916-8.
Lester, W., Kaye, S, Robertson, O.H., and Dunkl�n, E.W. (1950). Factors of �mportance �n the use of tr�ethylene glycol vapor for aer�al d�s�nfect�on. Am-J-Pub-Health, 40(7), 813-20.
Musher, D.M. (2003). How contag�ous are common resp�ratory tract �nfect�ons? N- Engl- J- Med, 348, 1256-66.
NIST (2005). Chem�stry WebBook, Nat�onal Inst�tute of Standards and Technology. <webbook.n�st.gov/cg�/cbook.cg�?ID=C112276&Un�ts=SI&Mask=4#Thermo-Phase>.
Olsen, S.J., Chang, H.L., Cheung, T.Y.Y., Tang, A.F.Y., F�sk, T.L., Oo�, S.P.L., Kuo, H.W., J�ang, D.D.S., Chen, K.T., Lando, J., Hsu, K.H.H., Chen, T.J., and Dowell, S.F. (2003). Transm�ss�on of the severe acute resp�ratory syndrome on a�rcraft. N-Engl- J-Med,-349, 2416-22.
OSHA (2008). L�m�ts for a�r contam�nants, Standard Number 29 CFR 1910.1000, Table Z-1, Electron�c Code of Federal Regulat�ons. <http://ecfr.gpoaccess.gov/cg�/t/text/text-�dx?c= ecfr&rgn=d�v5&v�ew=text&node=29:6.1.1.1.1&�dno=29#29:6.1.1.1.1.1.1.1>.
Puck, T.T. (1947a). The mechan�sm of aer�al d�s�n-fect�on by glycols and other chem�cal agents I. Demonstrat�on that the germ�c�dal act�on occurs through the agency of the vapor phase. J-Exp-Med, 85, 729-39.
Puck, T.T. (1947b). The mechan�sm of aer�al d�s-�nfect�on by glycols and other chem�cal agents II. An analys�s of the factors govern�ng the eff�c�ency of chem�cal d�s�nfect�on of the a�r. J-Exp-Med,-85, 741-57.
Robertson, O.H., Loosl�, C.G., Puck, T.T., B�gg, E., and M�ller, B.F. (1941). The protect�on of m�ce aga�nst �nfect�on w�th a�r-borne �nfluenza v�rus by means of propylene glycol vapor. Science, 94 (2452), 612-3.
Robertson, O.H., B�gg, E., M�ller, B.F., Baker, Z., and Puck, T.T. (1941). Ster�l�zat�on of a�r by certa�n glycols employed as aerosols. Science,- 93(2409), 213-4.
Robertson, O.H., Puck, T.T., Lemon, H.F., and Loosl�, C.G. (1943). The lethal effect of tr�ethylene glycol on a�r-borne bacter�a and �nfluenza v�rus. Science,-97(2510), 142.-4.
Roy, C.J. and M�lton, D.K. (2004). A�rborne transm�ss�on of commun�cable �nfect�on—the elus�ve pathway. N-Engl-J-Med, 350, 1710-2.
Rudn�ck, S.N. and F�rst, M.W. (2007). Fundamental factors affect�ng upper-room ultrav�olet germ�c�dal �rrad�at�on—Part II. pred�ct�ng effect�veness. J-Oc -cup-Environ-Hyg, 4(5), 352-62.
Schumb, W.C., Satterf�eld, C.N., and Wentworth, R.L. (1955). Hydrogen- peroxide, Amer�can Chem�cal Soc�ety Monograph, ser�es no. 128, Re�nhold Publ�sh�ng Corp., New York.
