Content uploaded by William A. Anderson
Author content
All content in this area was uploaded by William A. Anderson on Apr 09, 2020
Content may be subject to copyright.
Draft: April 9, 2020
1
UV Disinfection of N95 Filtering Facepiece Respirators: A Brief Review
Prof. William A. Anderson, P.Eng.
Department of Chemical Engineering
University of Waterloo, Waterloo ON Canada N2L 4L5
wanderson@uwaterloo.ca
Updates: corrections and clarifications were added to the discussion on the Fisher and Shaffer paper.
Occupational maximum exposure limits for UV-C have been added to the section on Health and Safety
Considerations.
Introduction
In pandemic situations the supply of N95 filtering facepiece respirators (FFRs) becomes limited, leaving
healthcare professionals to either forego protection, use less protective types of masks, or consider the re-
use of the available N95 masks. Re-use of masks raises the risk of contamination of the wearer or
environment by pathogens already present on the mask surface or depths of the filter material, unless a
decontamination or disinfection process is employed. This review considers the available published peer-
reviewed literature on the use and effects of germicidal UV disinfection on N95 masks, with specific focus
on virucidal activity. This information can be used to inform infection prevention risk assessment.
UV Disinfection Background
Germicidal ultraviolet light (UV) has been employed for many decades for water, air and surface
disinfection. Germicidal UV, also known as UV-C, is typically generated with mercury-based lamps that
operate at 253.7 nm (often rounded off to 254 nm). Light at this wavelength is relatively strongly absorbed
by nucleic acids, resulting in damage to RNA and DNA molecules in pathogens, which prevents their
function or reproduction. However, 254 nm light is also absorbed by other organic molecules, especially
those containing aromatic rings, and this can reduce the efficacy of UV disinfection.
According to data compiled by Kowalski (2009), the UV dose required to inactivate various pathogens on
surfaces can be quite wide-ranging. This is likely a combination of inherent properties of the pathogen, as
well as experimental differences in measuring these processes. On average, Kowalski reports that
vegetative bacteria require 1.6 mJ/cm2 to inactivate 90%, while bacterial spores require 12.6 mJ/cm2.
Viruses
1
require 7.3 mJ/cm2, and fungi and yeasts may require 32.9 mJ/cm2 for 90% inactivation. The data
compiled by Kowalski for surfaces is relatively scarce however, so these values must be used with caution.
N95 FFR Disinfection
The literature specifically addressing N95 FFR disinfection by UV is relatively sparse, but there are some
studies that can provide good insights. There are several key issues that need to be considered for the
potential use of UV disinfection, namely: 1) recommended UV dose for effective disinfection; 2) the effects
of UV on FFR performance and/or fit and usability; and 3) UV exposure methodology and geometry of
systems. These issues will be reviewed below, to the extent that the available literature addresses them.
1
Kowalski’s virus data average appears to be mainly based on a combination of animal viruses and bacterial phages,
and is quite sparse for surface studies.
Draft: April 9, 2020
2
UV Dose Requirements
Several studies have tested the efficacy of UV-C exposure of N95 masks on various microbes. Since
experimental methods differ between these studies, rigorous comparisons will be difficult but general
guidance may be sought.
Lin et al. (2018) measured the disinfection performance of UV-C on N95 FFRs using Bacillus subtilis
spores, which would be relatively UV resistant. They measured a 0.2% survival after 2 minutes, with no
measurable survival at 5 minutes, at an intensity of 18.9 mW/cm2. This would correspond to doses of 2,268
mJ/cm2 and 5,670 mJ/cm2 respectively. The use of UV-A (365 nm, “tanning UV”) was also tested and
found to be largely ineffective.
Mills et al. (2018) reported on a more extensive set of tests on N95 masks using H1N1 viruses, and included
the effects of soiling agents (artificial saliva and/or skin oil) that could reduce the efficacy of UV exposure.
Fifteen different N95 models were tested from a variety of manufacturers, and both the facemask and straps
were monitored. All FFRs were disinfected to a level of at least 3 log (i.e. 99.9%), even in the presence of
soiling agents, when the UV dose was 1,000 mJ/cm2.
Lore et al. (2011) tested UV disinfection of N95 FFRs with H5N1 influenza virus and found > 4 log
reduction (i.e. >99.99%), using UV-C doses of approximately 1,800 mJ/cm2 when irradiating the convex
side of the FFR. Similarly, Heimbuch et al. (2011) noted at least 4 log (99.99%) reduction of H1N1 virus
during UV irradiation of the convex side of FFRs, with a dose of approximately 1,620 mJ/cm2. Heimbuch
also notes that microwave irradiation was similarly effective
2
.
Vo et al. (2009) measured a 3 log reduction (99.9%) of MS2 bacteriophage on N95 samples, with a UV-C
dose of 4,320 mJ/cm2.
Fisher and Shaffer (2010) tested circular coupons of N95 materials with MS2 coliphage and determined
that a minimum dose of 100 mJ/cm2 was required for a 3 log reduction (99.9%). Doses of less than 100
mJ/cm2 were still effective, but generally in the range of 1 to 2 log reduction (90 to 99%), with the lowest
dose tested being 30 mJ/cm2. Doses higher than 100 mJ/cm2 were able to achieve 3.4 to > 5.1 log reduction,
depending on the model of N95 material. It is important to qualify these numbers, in that Fisher and Shaffer
reported them in terms of “internal filtering medium” doses using a mathematical model. The translation
into the “applied doses” on the surface of the masks reported by all other above authors is not
straightforward and depends on the N95 construction.
Comments on UV Dose Requirements
From the published literature on UV disinfection using N95 masks or related materials, it is clear that the
minimum dose for reasonably effective disinfection (log 3 or 99.9%) is likely in the order of 1,000 mJ/cm2
or higher, assuming that the pathogen of interest has similar susceptibility to UV radiation as the ones
mentioned above.
This dose requirement is much higher than doses reported for surfaces, for example the 7 mJ/cm2 mentioned
by Kowalski, reflecting the fact that mask materials are not “surfaces” in the normal sense of a hard and
flat material. The N95 masks are in-fact porous depth filters, with an approximate thickness of 0.5 to 2.5
mm (for three samples mentioned by Suen et al., 2020). Therefore, for effective disinfection the UV light
rays must penetrate the full depth of the mask. Some UV energy is lost through absorption and reflection
processes as it penetrates the depth of the mask material, meaning that an effective dose must be much
higher than what is published for hard and flat surfaces.
2
Although not the focus of this review, microwaves may be another disinfection route of interest. Contact the author
of this review for further details if interested.
Draft: April 9, 2020
3
UV Effects on FFR Performance
It is well-known that UV radiation has degrading effects on organic materials, and therefore the potential
for UV to reduce the filtration performance of an N95 material must be considered. Unfortunately, there is
not extensive information about the effects, but the limited information is summarized in this section.
Viscusi et al. (2009) measured the effect of UV on a variety of N95 FFRs and surgical N95 respirators (as
well as some P100 models), specifically determining air flow and filtering effects. No significant effects
of UV exposure were found on aerosol penetration, airflow resistance or physical appearance. However,
the total UV dose applied was only approximately 340 mJ/cm2 (half the dose applied to each side).
Viscusi et al. (2011) assessed the fit factor for a variety of N95 FFRs after UV irradiation with a total dose
of approximately 3,240 mJ/cm2. No significant changes in fit, odour detection, comfort, or donning
difficulty were detected in this study with a small group (18) of experienced respirator users. The effect on
filtration performance was not assessed in this study, however.
Bergman et al. (2010) tested a variety of N95 FFRs with a higher UV dose of approximately 4,860 mJ/cm2
and found no significant change in filtration performance, including aerosol penetration.
Comments on UV Effects on FFR Performance
Although the available literature is sparse, the few available studies suggest that UV doses of up to almost
5,000 mJ/cm2 have no effect on the key parameter, i.e. filtration performance. FFR fit and wearablity do
not appear to be affected either, although strap breakage was mentioned as an occasional problem in at least
one of the above studies.
UV Treatment Methodology and Geometries
From a process engineering perspective, reliable delivery of a suitable dose of UV requires careful
consideration of the lamp power, exposure time, lamp geometry, orientation of the FFRs with respect to the
lamps, and distance from the lamps to the target FFRs. A detailed or recommended design has not been
published for this purpose, but the literature provides some insight into possibilities.
A relatively simple arrangement was used by Heimbuch et al. (2011) as illustrated in Figure 1, utilizing a
fixture that is similar to a typical fluorescent light, but with a UV-C bulb.
Figure 1: Schematic of UV irradiation system, as shown by Heimbuch et al. (2011), utilizing a 120 cm,
80 W UV-C lamp irradiating the exterior convex surface of the FFRs for a period of 15 minutes.
Draft: April 9, 2020
4
Lin et al. (2018) used a 6 W handheld UV lamp, placed 10 cm above the FFRs, which were exposed for up
to 20 minutes per side (convex and concave). The relatively low lamp power necessitated longer exposure
times to achieve the desired doses.
Fisher and Shaffer (2010) used a 40 W lamp (Philips TUV 36T5) inside a biological safety cabinet to expose
circular coupons of N95 filter materials.
Woo et al. (2012) used a 4 W UV-C lamp adjusted to a height of 10 cm to irradiate petri dishes used to
quantify viral aerosol decontamination.
Bergman et al. (2010) used a 40 W UV-C lamp lifted about 25 cm above the working surface to expose the
exterior surfaces of FFRs.
Vo et al. (2009) also used a biological safety cabinet, equipped with a 40 W lamp (TUV 36TS 4P SE). Lore
et al. (2011) used a laminar flow cabinet with a dual-bulb 15 W lamp
3
mounted 25 cm above the working
surface. Viscusi et al. (2009) also used a laminar flow cabinet with a 40 W lamp system (dimensions
uncertain).
Mills et al. (2018) employed a custom-built UV irradiation device (shown in Figure 2), employing an
aluminum housing and eight 32 inch (about 81 cm) UV-C bulbs having an irradiance of 0.39 W/cm2 at 1
metre. This device incorporated extra features for temperature control.
Figure 2: UV irradiation device fabricated and employed by Mills et al. (2018).
3
Some of the lamp details are uncertain.
Draft: April 9, 2020
5
The foregoing systems all employed some variation of a linear/planar exposure system. This has some
advantages in simplicity of layout and operation, but the UV dose must be carefully measured and planned,
as FFRs placed in the centre of a lamp will have a greater dose than those place towards the ends of the
lamp. The quantification of UV dose requires the use of a UV-C radiometer to calibrate the time required
to achieve the desired dose.
As an alternative to the planar systems, some healthcare institutions already have UV disinfection systems
used for room decontamination, such as the example shown in Figure 3.
Figure 3: mobile “tower” UV disinfection system (from https://prescientx.com/)
For a tower system like that shown in Figure 3, it could be feasible to set up a holder such that FFRs are
arranged around the circumference of the radiation field at some fixed distance, as illustrated in Figure 4.
This arrangement benefits from the radial geometry that helps to irradiate the sides of the FFRs more
thoroughly than a planar arrangement might accomplish. The placement of the FFRs must be carefully
considered as those placed towards the top or bottom of the lamps will receive a lower dose. Again, the
use of a UV-C radiometer would be required to monitor and calibrate the dose delivery time.
Draft: April 9, 2020
6
Figure 4: schematic (top view) of a UV irradiation set-up using a “tower” disinfection device, and FFRs
attached on a holding device around the tower.
Comments on UV-C Irradiation Equipment and Layout
There a several options that may permit rapid deployment of a UV disinfection system for FFR treatment
under urgent conditions.
• The planar arrangement illustrated in Figure 1 can be readily set-up anywhere using stock items
such as fluorescent light fixtures, if UV-C mercury discharge lamps can be obtained to fit. These
will typically be 24, 36, or 48 inch lamps with a bi-pin G13 base, with T5 or T8 diameter. Other
UV-C lamps are available from a variety of suppliers.
• The cylindrical arrangement shown in Figure 4 can be readily set-up in locations where the UV
tower disinfection units already exist (or if they can be purchased). Fabrication of a “shell” where
FFRs can be mounted is required, but any sheet metal, plastic, or rigid screen material could suffice.
In all cases, a UV-C radiometer will be required (borrowed or purchased) to calibrate the exposure time to
ensure an adequate dose will be achieved. The time required will depend on the lamp power, layout and
distance from the lamps to the FFRs.
Health and Safety Considerations
There are several caveats and H&S considerations to be aware of, as listed below.
1. Disinfection can never be considered 100% effective. There are risks that active pathogens may
remain on the FFR or buried within the filter material. Pathogenic spores, such as C. difficile, may
be more resistant to UV than viruses and vegetative bacteria.
2. In disinfection, pathogen risk is not only dependent on the degree of UV inactivation, but also the
initial load of pathogens. For example, 99% disinfection might be quite adequate when few
pathogens are present, but when a very high load are initially present 99.9% disinfection may still
leave enough viable to supply an infectious dose.
3. UV-C radiation is also damaging to human cells and exposure must be avoided. Ideally, remote
operation in a secure room with a door interlock switch is desirable to eliminate the potential for
accidental entrance of personnel during operation. Other protocols and procedures might be put
into place if necessary. Occupational exposure limits for UV-C (254 nm) exposure are in the order
UV Tower
Draft: April 9, 2020
7
of 6 mJ/cm2 (Kowalski, 2009), which can easily be exceeded in a few seconds or less with a typical
UV lamp (depending on the power level and distance). Kowalski reports that glass or plastic of at
least 1/8 inch thickness (3.2 mm) provides adequate protection against UV-C.
Disclaimer
The information in this document is provided for consideration, but there are no guarantees that any ideas
suggested here will be effective in any or all situations. The author accepts no responsibility for actions
taken by others based on the information presented here. This is a technical review of available literature,
and not a work of engineering.
Further Information
The author can be contacted by email (wanderson@uwaterloo.ca) for further explanations or discussion of
this material.
Draft: April 9, 2020
8
References
Bergman, Michael S., Dennis J. Viscusi, Brian K. Heimbuch, Joseph D. Wander, Anthony R. Sambol, and
Ronald E. Shaffer. "Evaluation of multiple (3-cycle) decontamination processing for filtering facepiece
respirators." Journal of Engineered Fibers and Fabrics 5, no. 4 (2010): 155892501000500405.
Fisher, Edward M., and Ronald E. Shaffer. "A method to determine the available UV‐C dose for the
decontamination of filtering facepiece respirators." Journal of applied microbiology 110, no. 1 (2011): 287-
295.
Heimbuch, Brian K., William H. Wallace, Kimberly Kinney, April E. Lumley, Chang-Yu Wu, Myung-
Heui Woo, and Joseph D. Wander. "A pandemic influenza preparedness study: use of energetic methods to
decontaminate filtering facepiece respirators contaminated with H1N1 aerosols and droplets." American
journal of infection control 39, no. 1 (2011): e1-e9.
Kowalski, W. Ultraviolet Germicidal Irradiation Handbook. Springer-Verlag, Berlin, 2009.
Lin, T‐H., F‐C. Tang, P‐C. Hung, Z‐C. Hua, and C‐Y. Lai. "Relative survival of Bacillus subtilis spores
loaded on filtering facepiece respirators after five decontamination methods." Indoor air 28, no. 5 (2018):
754-762.
Lore, Michael B., Brian K. Heimbuch, Teanne L. Brown, Joseph D. Wander, and Steven H. Hinrichs.
"Effectiveness of three decontamination treatments against influenza virus applied to filtering facepiece
respirators." Annals of occupational hygiene 56, no. 1 (2012): 92-101.
Mills, Devin, Delbert A. Harnish, Caryn Lawrence, Megan Sandoval-Powers, and Brian K. Heimbuch.
"Ultraviolet germicidal irradiation of influenza-contaminated N95 filtering facepiece
respirators." American journal of infection control 46, no. 7 (2018): e49-e55.
Suen, L. K. P., Y. P. Guo, S. S. K. Ho, C. H. Au-Yeung, and S. C. Lam. "Comparing mask fit and usability
of traditional and nanofibre N95 filtering facepiece respirators before and after nursing
procedures." Journal of Hospital Infection (2019).
Viscusi, Dennis J., Michael S. Bergman, Debra A. Novak, Kimberly A. Faulkner, Andrew Palmiero, Jeffrey
Powell, and Ronald E. Shaffer. "Impact of three biological decontamination methods on filtering facepiece
respirator fit, odor, comfort, and donning ease." Journal of occupational and environmental hygiene 8, no.
7 (2011): 426-436.
Vo, Evanly, Samy Rengasamy, and Ronald Shaffer. "Development of a test system to evaluate
decontamination procedures for viral droplets on respirators." Applied and Environmental
Microbiology (2009).
Woo, Myung-Heui, Adam Grippin, Diandra Anwar, Tamara Smith, Chang-Yu Wu, and Joseph D. Wander.
"Effects of relative humidity and spraying medium on UV decontamination of filters loaded with viral
aerosols." Appl. Environ. Microbiol. 78, no. 16 (2012): 5781-5787.
