els-jbs-prod-cdn.jbs.elsevierhealth.com · Journal Pre-Proof Mayo Clinic Proceedings Filtration...

Journ

al Pre-

Proof

Mayo Clinic Proceedings Filtration properties of cloth and cloth masks

© 2020 Mayo Foundation for Medical Education and Research. Mayo Clin Proc. 2020;95(x):xx-xx.

Forgotten Technology in the COVID-19 Pandemic. Filtration Properties of Cloth and Cloth Masks: A Narrative Review

Catherine M Clase1,2, MB, FRCPC; Edouard L Fu3, BSc; Aurneen Ashur1; Rupert CL Beale5, MB, PhD, Imogen A Clase1; Myrna B Dolovich4, P Eng; Meg J Jardine6, MBBS PhD; Meera Joseph1, MD, FRCPC; Grace Kansiime1, MBChB MMed; Johannes FE Mann7, MD, PhD; Roberto Pecoits-Filho8, MD, PhD; Wolfgang C Winkelmayer9, MD, ScD; Juan J Carrero10, Pharm, PhD

1Division of Nephrology, Department of Medicine, McMaster University, Hamilton, Canada 2Department of Health Research Methods, Evidence and Impact, McMaster University, Hamilton, Canada 3Department of Clinical Epidemiology, Leiden University Medical Center, Leiden, TheNetherlands

4Division of Respirology, Department of Medicine, McMaster University, McMaster University,Hamilton, Canada5Francis Crick Institute, London, UK 6The George Institute for Global Health, Sydney, Australia. Concord Repatriation GeneralHospital, Sydney, Australia

7Department of Nephrology, Hypertension & Rheumatology at Munich GeneralHospitals.University of Erlangen-Nürnberg, KfH Kidney Center, Munich-Schwabing, Germany8DOPPS Program Area, Arbor Research Collaborative for Health, Ann Arbor, Michigan, USA;School of Medicine, Pontifical Catholic University of Paraná, Curitiba, Brazil9Section of Nephrology, Department of Medicine, Baylor College of Medicine; Houston, Texas,United States10Department of Medical Epidemiology and Biostatistics, Karolinska Institutet, Stockholm,Sweden

Key words: COVID-19; SARS-CoV-2; cloth masks; face masks; filtration efficiency; leak

Journ

al Pre-

Proof



Abstract

We searched Medline and Embase, and used Google, including articles reporting the filtration

properties of flat cloth, or cloth masks. We reviewed the reference lists of relevant articles and

review articles, and identified articles the press. We found 25 articles. Study of protection for

the wearer often used a manikin wearing a mask, with airflow to simulate different breathing

rates. Studies of protection of the environment, also known as source control, used

convenience samples of healthy volunteers. The design and execution of the studies was

generally rigorously described. Many descriptions of cloth lacked the detail required for

reproducibility; no study gave all the expected details of material, thread count, weave, and

weight. Some of the homemade mask designs were reproducible.

Successful masks were muslin at 100 threads per inch (TPI) in 3-4 layers (4-layer muslin or a

muslin-flannel-muslin sandwich); tea towels (also known as dish towels), studied as one-layer,

and two-layer expected to be better; and good-quality cotton T shirts in 2 layers (with a

stitched edge to prevent stretching). In flat-cloth experiments, tea towel, cotton 600 TPI in two

layers, and cotton 600 TPI with flannel 90 TPI, performed well, but two-layer cotton 80 TPI did

not. Multiple layers should be used, at least two, and preferably three or four; however there is

a trade-off in that this increases the resistance to breathing.

This is not a systematic review; however, we included all the articles that we identified in an

unbiased way. We did not include grey literature or preprints.

Abbreviations

ASTM: American Society for Testing and Materials; AFNOR: Association française de normalisation; COVID-19: coronavirus disease 2019; NIOSH: National Institute for Occupational Safety and Health; PPE: personal protective equipment; RCT: randomized controlled trial; SARS-CoV-2: severe acute respiratory syndrome coronavirus 2; TPI: threads per inch, the sum of the warp plus weft thread count per inch; WHO: World Health Organisation

Journ

al Pre-

Proof



Introduction

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) resulting in coronavirus disease

2019 (COVID-19) has, at the time of writing, claimed at least 600,000 lives.1 The management of

this global crisis requires detailed appraisal of evidence to support clear, actionable, and

consistent public health messaging. The use of cloth masks for general public use is being

debated, and is in flux: in April 2020, the World Health Organisation (WHO) changed its position

from ‘not recommended under any circumstance’2 to ‘there is no current evidence to make a

recommendation for or against their use,’3 recognizing that, ‘decision makers may be moving

ahead with advising the use of non-medical masks’, as was indeed occurring.4-8 On June 5, 2020,

the WHO updated its guidance further ‘to advise that to prevent COVID-19 transmission

effectively in areas of community transmission, governments should encourage the general

public to wear masks in specific situations and settings as part of a comprehensive approach to

suppress SARS-CoV-2 transmission.’9

In early March, the WHO estimated that 89 million masks would be needed each month,

globally, for medical purposes alone,10 highlighting the importance of directing the supply of

medical masks amd respirator-type masks (e.g., N95s) to medical use. Non-medical masks will

be needed for the other purposes outlined. Cloth masks potentially offer a reusable,

sustainable and environmentally-friendly solution.

This review summarizes a century of evidence on the efficiency of cloth and cloth masks to

reduce transmission of droplets and aerosols (Box, supplementary table).11-36 We argue that

this body of work should inform decisions in the context of reducing the transmissibility of

COVID-19. Physical distancing, hand washing and disinfection of surfaces remain the

cornerstones of policy, and we stress that we are not discussing cloth masks as a means of

relaxing these interventions, or as a replacement for formal personal protective equipment

(PPE) for high-risk workers.

What are the standards in this literature?

When we breathe, eat, speak, sing, cough or sneeze, particles are released into the

environment. The size distribution of these particles varies with the activity, as does their

Journ

al Pre-

Proof



velocity and their trajectory. Though technically all these particles of liquid (respiratory

secretions) suspended in gas (air) are aerosols, we recognize a useful distinction between

coarse particles, sometimes called droplets, which are usually defined as > 5µm aerodynamic

diameter, and aerosols, which are particles of < 5 µm aerodynamic diameter. Virus particles are

nanoparticles, much less than 1 µm; exhaled secretions may contain virus particles.

Filtration efficiency is the proportion of particles blocked by a filter, usually expressed as a

percentage, and assessed using surrogate markers, not directly with transmissible pathogens

(figure 1, supplementary figure 1). Some surrogates are non-biological, such as ambient

particles, or aerosols of diesel combustion or saline; others are bioaerosols, usually bacterial.

Filtration standards specify detail for testing of mask materials (equipment, surface area tested,

air flow, particle type and size). Medical masks (also known as dental masks and surgical masks)

are certified according to the standards set by the American Society for Testing and Materials

(ASTM) standards.37 Canada uses these US standards for mask materials, which define 3 levels

(1-3) of mask according to particle filtration efficiency, greater than 95%, 98% and 98% for the

flat material, respectively. Increasing resistance to splashing with synthetic blood further

distinguishes level 2 and level 3 masks. Particle filtration efficiency of the flat mask material is

assessed using latex spheres at 0.1 µm; bacterial filtration efficiency using aerosolised

Staphylococcus aureus at a mean particle size of 3 µm.37 The material for respirator-type masks,

in North America called N95s, is certified according to standards set by the US National Institute

for Occupational Safety and Health (NIOSH).38 The relationship between particle size and

filtration efficiency is not linear, with small particles having consistently lower efficiency, but U-

shaped, with the lowest filtration efficiency usually around 0.3 µm, which is sometimes called

the most-penetrating-particle size.23, 38 Mask material for respirators is therefore tested at 0.3

µm, and particle filtration efficiency greater than 95% is required.38 The US Occupational Safety

and Health Administration and Canadian Standards Association standard Z94.4 further require

that N95 masks be fitted to the individual who will wear them.39, 40 Fit assesses both

penetration through the mask material and leak around the mask edge. A quantitative fit

testing device, the Portacount (TSI, Auburn, IL, US), measures saline particles in the 0.02 – 1 µm

range, inside and outside the mask. A ratio of 100 particles outside the mask to 1 particle

Journ

al Pre-

Proof



inside, known as a fit factor, is required: this is equivalent to filtration efficiency of 99%. A non-

quantitative alternative standard is to test with a hood and a strong-tasting aerosol such as

saccharin. 39 No fit testing is required for medical masks.

The diameter of SARS-CoV-2 virus has been reported as between 0.065 and 0.140 µm.41 In

contrast, the space between threads in many woven cloths are visible to the naked eye, and

even in high-thread count fabric, the gaps between fibres are of the order of magnitude of 5 to

15 µm.23 In lower thread-count fabric, gaps as large as 50 to 200 µm are expected and

observed.23, 42, 43 It is counterintuitive that cloth stops particles smaller than 5 µm; however,

particles of this size encounter cloth fibres and are filtered through the three physical principles

of impaction, sedimentation and diffusion.44

Transmission of virus is usually not as isolated virions, but in larger particles combined with

respiratory secretions. Though the literature describing the size distribution of particles

generated by activities such as breathing, coughing and sneezing is not completely consistent, it

appears that even for the less explosive activities, a proportion of particles are >1 µm, and for

coughing and sneezing, particles in the 10 µm and even 100 µm range have been observed45

though other reports suggest peak particle size around 1 – 5 µm46; the reasons for these large

differences between studies are not apparent. The particle size used for testing medical masks

is 0.1 µm.37 If individual particles contain more than one virion, and larger particles contain

more virions than smaller particles, filtration efficiency for virions reaching the environment

may exceed expectations based on testing using nano latex test particles.

Cloth is woven (crossing threads, known as warp and weft), knitted (interlocking loops of fibre)

or felted (compressed disorganized fibres). Woven cloth is further described by its weave. In

plain weave, fibres cross at 90 degrees. Twisted weave gives a diagonal stripe to the finish, and

is known as twill weave: a common example is denim. When the warp and the weft are

different numbers of threads in a given distance (conventionally, an inch), thread count may be

expressed by two numbers, e.g., 20x14. Thread count expressed as a single number, threads

per inch (TPI), is the sum of the warp plus weft thread count per inch. The finish may be plain or

raised to fuzziness, which is called a nap. Some fabrics, called terry, have projecting loops of

Journ

al Pre-

Proof



fibre to increase absorbance. The overall heaviness is described by its weight per surface area.

Very high thread counts (>300) are usually obtained by using very thin fibre and the resulting

material may be very fine (light-weight, such as some bed-linen).

Surgical gauze, plain woven cotton or linen (such as most dish-towels [US] or tea-towels [UK],

i.e., flat cloths used for drying dishes), muslin and buttercloth (a cloth used for straining in the

manufacture of butter), and some bed-linen, are plain-weave unnapped cloth. Flannel,

commonly used for nightwear and some bed linen, is a plain-weave napped cloth, often of

cotton. T-shirt material is usually knitted jersey; the proportion of cotton to man-made fibre,

and the weight, varies. Terry is used for most bath and hand-towels.

Commercial disposable masks are made from non-woven synthetic fibres in bonded layers.

These are unsystematically called medical masks, face masks, surgical masks, dental masks and

procedure masks. We use the term medical masks.

Can fabric block coarse and fine particles?

The increasing effectiveness of multiple layers of cloth to reduce transmission was

demonstrated in 1919, in a series of experiments using controlled sprays and real coughing to

create bioaerosols.34 Bacterial counts were used as the surrogate marker. Filtration efficiency

increased with thread count and layers, and was consistently greater, at any given total thread

count, the fewer the layers (e.g., one layer with mean thread count 42 provided greater

filtration than two layers with thread count 22, total thread count 44) (supplementary figures

2-4). At all distances, total thread counts above ~300 TPI were associated with >80% filtration

efficiency. Others confirmed these observations using similar designs16, 19, 22, 24, one study

observing that twill weave cotton was associated with 94% filtration efficiency, compared with

98 and 99% for material from two medical masks.16

Filtration efficiencies of 28-73% were reported for single layers of bath towel and cotton shirt

tested with 2 µm bacterial particles.18 For tea towel fabric, filtration efficiency for bacteria was

83% with one layer and 97% at two layers, compared with medical mask material at 96%.14 For

virus, one layer of tea towel had 72% efficiency, and one layer of T-shirt fabric 50%, compared

with 89% for mask material.14

Journ

al Pre-

Proof



Results are dependent on the type of cloth studied. For NaCl aerosol, three commercially-

available cloth masks, and single layers of scarfs, most sweatshirts, T-shirts and towel were

associated with filtration efficiency of 10-40%.29 The cloth from the mask studied in the single

randomized controlled trial (RCT) was tested using a TSI filter tester, according to Australian

and New Zealand standards for respirators.26 Filtration efficiency for the cloth was 3%,

compared to the medical mask, which tested at 56%. The trial is described in detail below.

Table 1 summarizes studies that use modern methodology (supplementary figure 1A) to test

the filtration efficiency of flat cloth, organized by fabric type, and includes information on

medical-mask-material and respirator-material comparators.14, 16, 20, 21, 23, 26, 29, 35, 47 Few of these

studies used standardized methodology and their results are not directly comparable.

Collectively, they show that even at low thread counts and layers, and for aerosols, some kinds

of cloth block substantial percentages of transmission. Filtration efficacy >50% has been

observed for single layers of high-thread-count cotton, for linen and cotton tea towels, for

some T shirt materials, and some towels; efficiency increases with layers, and efficiency for

virus is of the same order of magnitude as that for bacteria.

An important variable is airflow; in general, other experimental conditions being constant,

lower filtration efficiencies would be expected with higher airflows, though this is not

consistently observed, perhaps because of random error.23, 29 Most testing of flat materials

aims to simulate breathing through a mask, sometimes at high minute ventilation to simulate

exertion.14, 21, 23, 35, 37, 38 Lower flow rates, as observed in some studies,23 and flow rates as

velocities,29 require consideration in interpretation. Peak flow rates of 200 – 1300 L/min, and

peak velocities of 29 m/s have been observed for human coughs46. None of the experiments

that we identified on flat cloth aimed to simulate these conditions.

Does wearing a cloth mask prevent coarse and fine particles from reaching the environment

(outward protection, or source control; supplementary figure 1C and D)?

In a design in which volunteers, talking or coughing, sat at a table on which agar plates were

arranged, masks of 3-8 layers of buttercloth (TPI ~90), blocked 100% of bacteria at all

distances.15 Similar results were obtained by others.27, 48

Journ

al Pre-

Proof



A 3-layer 46x46 gauze mask (i.e., TPI 92 in each layer), compared with no mask, reduced

bacterial counts by 64% in the zone immediately in front of healthy volunteers, who were

talking.31 In a controlled box experiment, using volunteers talking, a mask made of a sandwich

of thin muslin and 4oz flannel (136 g/m2) reduced bacteria recovered on sedimentation plates

by >99%, compared with the recovery from unmasked volunteers.17 Total airborne

microorganisms were reduced by >99%, and bacteria recovered from aerosols (<4 µm) by 88-

99%, compared with those recovered from unmasked volunteers. Another controlled-box

experiment with talking volunteers compared 4 medical masks and one commercially produced

four-layer cotton muslin (92 TPI49) reusable mask.28 Filtration efficiency, assessed by bacterial

counts, was 96-99% for the commercial disposable masks and 99% for the commercial 4-layer

muslin mask. For aerosols (<3.3 µm) filtration efficiencies were 72-89%, and 89% respectively.

Using a pattern based on a pleated medical mask, but without assistance, volunteers made 2-

layer T-shirt masks with over-the-head elastic.14, 50 Wearing the mask they had made,

volunteers coughed twice into a box: at each particle size, homemade and medical masks were

similar, with 79% and 85% efficiency respectively at the smallest particle size measured, 0.65 –

1 µm, P=0.24 (table 2).

We identified one disconfirming report: studied using a Portacount (0.02-1 µm) on a manikin,

the efficiency of a one-layer tea-towel mask in reducing aerosols reaching the environment was

17% (table 3).32

These studies show that some multilayered cloth masks can show remarkable filtration

efficiency in the outward direction, reducing all particles by 64-99% and aerosols by 72-99%

emitted by the wearer: for some designs comparable or better than commercial medical masks.

Does wearing a cloth mask prevent inhalation of coarse and fine particles (inward protection,

or personal protection)?

Table 4 summarizes studies of cloth masks worn by volunteers or on a manikin (supplementary

figure 1B). Three authors made personalized cloth mask from heavy-duty T-shirt material,

including three sets of ties and 8 layers of material at the front.13 Filtration efficiency, measured

Journ

al Pre-

Proof



using a Portacount (0.02-1 µm), was 97%, 92% and 94% for the three individuals tested. (On

this test, a respirator performing at 99% or above would be considered a good fit.)

In a study using healthy volunteers and a Portacount (0.02-1 µm), a 2-layer home-made T-shirt

mask provided 50% inward filtration efficiency during a range of activities, compared with 80-

86% from a medical mask14, 50.

Three cloth masks and one medical mask purchased from street vendors in Kathmandu, Nepal,

and two N95 masks, were tested using a manikin.30 Test particles were polystyrene latex and

diesel combustion particles. Cloth mask 1, which had a conical shape and an exhalation valve,

performed as well as the two N95 masks: all three masks had ≈80% filtration efficiency for

polystyrene latex particles across the range of particle sizes, from 30 nm to 2.5 µm. Cloth masks

2 and 3, simple rectangles with ear-loops, had filtration efficiency of 15-65% for 30 nm particles

and 65-75% efficiency for 2.5 µm particles. For diesel particles between 30-500 nm in diameter,

filtration efficiency was 25-85%, 10-70%, and 10-25% for cloth masks 1, 2, and 3 respectively;

and 55-85% for the surgical mask.

Similar results, filtration efficiencies of 55 to 77%, tested with aerosols 0.2-1.0 µm, were

reported for a one-layer tea towel mask (table 3).32

In experiments using a manikin to identify leakage around the interface between mask and

face, leakage was reduced by taping, or by holding material in place with pantyhose.12

These studies show that one specific cloth mask performed as well as an N95 in excluding

aerosols from the wearer,30 that complex, multi-layer homemade masks can perform above

90%,13 and that simple one-layer masks can perform similarly to medical masks.32 The poorest-

performing masks showed some inward filtration efficacy for aerosols.

Does wearing a cloth mask prevent disease in animal experiments?

In rabbits, exposed to aerosolised tubercle bacilli, tightly-fitting 3- or 6-layer 40x44 (84 TPI)

gauze masks reduced the number of tubercles per rabbit from 28.5 in unmasked and 1.4 in

masked, representing filtration efficacy of 95%, p=0.003 (our calculations).25 This controlled

Journ

al Pre-

Proof



animal experiment shows significant reduction in aerosol transmission of tuberculosis, usually

considered an airborne organism, by multilayered cloth masks.

Have RCTs on the effectiveness of cloth masks in any setting been conducted?

We identified a single RCT that compared continuous wear of a cloth mask with continuous and

with as-needed wear of medical masks.26 The cloth masks used were tested on an industry-

standard TSI device, according to the standards used for N95-type mask material, and were

found to be unusually inefficient at 3%. Though data are not exactly comparable between

studies, the observed filtration efficiency of 3% for this particular cloth mask material is the

lowest we identified in any study (table 1).13, 14, 26, 50 The medical mask comparator, assessed at

56% filtration efficiency (as flat material), performed substantially better.37 Unsurprisingly,

given these properties, continuous cloth mask use, compared with continuous medical mask

use, was associated with increased incidence of influenza-like illness, relative risk (RR) 13.3

(95% CI 1.74-101). Participants in this study were healthcare workers on high-risk medical

wards. The comparator groups were continuous medical mask use and medical mask use where

indicated by the patient’s isolation status. The use of a cloth mask continuously meant that

health care workers caring for patients requiring respiratory isolation wore a cloth mask in this

context instead of a medical mask. This study has been widely discussed in the press, and has

not always been accurately represented. One report summarizes it as “actually increased the

rate of infections among health care workers compared to those who wore surgical masks,”

which could be interpreted as cloth masks actually causing harm.51 A 2015 article on this study

carries the title “Cloth masks: Dangerous to your health?” and refers to “harm caused by cloth

masks.”52 The study leaves us unable to draw conclusions about the efficacy or harms of

wearing a cloth mask, compared with no mask, because there is no ‘no mask’ comparison

group. What we can infer from this study, however, is that in a healthcare setting, a device with

56% filtration efficiency prevents clinical illness compared with one with 3% filtration efficiency.

There is absence of evidence, then, rather than evidence of absence, or evidence of harm, on

whether cloth masks prevent transmission of clinical illness.

Does wearing a medical mask in a community context protect oneself or others?

Journ

al Pre-

Proof



Greenhalgh and coworkers, on 9 April 2020, identified five peer-reviewed systematic reviews

on public mask wearing to prevent transmission of a wide range of respiratory pathogens, and

summarized them as absence of evidence; citing the precautionary principle, the authors

advocated for public mask wearing.53 Using the framework of evidence-based medicine and the

concept of risk-based decision making under uncertainty (i.e., the absence of clear clinical

evidence of benefit), we supported this position.54 Subsequently, in a meta-analysis of

observational studies of risk of infection from the coronaviruses SARS-CoV-1, MERS and SARS-

CoV-2, use of masks (respirators, medical masks, or 12-16 layer cloth masks) compared with no

mask, was protective in both health-care settings (RR 0.30, 95% CI 0.22-0.41; I2 50%) and non-

health-care settings (RR 0.56, 95% CI 0.40-0.79; I2 48%).55

Does wearing a cloth mask in a community context protect oneself or others?

The meta-analysis identified three observational studies of mask use in the community.55 The

primary studies, reports of of SARS-C0V-1 transmission in Hong Kong, Beijing, and Vietnam, did

not identify the mask type used.56-58 One of these reports includes only nine participants

wearing masks.56 In another of these reports, compared with not visiting the index case, the

odds ratio (OR) for infection associated with a visit with a mask was 1.8 (95% confidence

interval [CI] 0.8-4.0) for one person wearing a mask, 1.9 (0.9-4.0) for both persons wearing a

mask, and 4.2 (2.4-7.3) for neither wearing a mask.57 The third study reports OR for infection of

0.5 (0.2-0.9) for sometimes wearing a mask when going out, and 0.3 (0.2-0.5) for always

wearing a mask when going out, compared with the referent of never wearing a mask when

going out.56, 58

The meta-analysis and detailed review of the primary studies advance our understanding from

‘absence of evidence’ to the point where we have somewhat-consistent observational evidence

of a protective effect from mask wearing in the community, with a large effect size. It is

plausible that masks protect people and there is coherence between the data on community

mask wearing and mask wearing in health care.59 However, the evidence is somewhat indirect:

SARS-CoV-1 transmission may differ from SARS-CoV-2. RCTs have not been conducted.

Journ

al Pre-

Proof



Symptomatic people should follow public health guidance and self-isolate. The point of

community mask wearing is to prevent presymptomatic and asymptomatic transmission.

Though asymptomatic transmission undoubtedly occurs,60-63 the proportion of transmission

that occurs from asymptomatic individuals is the subject of controversy.64, 65 However, evidence

from transmission pairs suggests that in individuals who will eventually develop symptoms,

peak infectivity may occur before the onset of symptoms, and that the highest levels of viral

shdding may occur in a period 2-3 days before the appearance of symptoms and 1 day after.66

Data on viral load in the days after symptom onset are congruent with this,67 and

presymptomatic transmission has been documented.63, 68 Modeling studies show that facemask

use depresses the effective basic reproductive rate over a range of plausible values for mask

use and cloth mask effectiveness, and that in conjunction with periods of lockdown, even 50%

adherence to a 50% effective cloth mask dramatically alters the total numbers affected.69

What materials and designs should be used? An evidence-informed cloth mask.

A pleated mask design based on the common pleated design for ASTM level 1 masks results in a

mask with subjectively good fit that is relatively simple to make. Paper fasteners, florists’ or

electricians’ wire, or pipecleaner can be inserted across the top to improve fit at the nose.

Though data are not available that conform with any modern standard method, from the

studies available, cotton, muslin (a type of unfinished cotton), and flannel are the best

supported and are our suggestions for an evidence-informed cloth mask. Successful masks have

used muslin at TPI of ~100 in 3-4 layers (4-layer muslin28 or a muslin-flannel-muslin sandwich17),

tea towels (also known as dish towels), studied as one-layer,14, 32, 50, 70 and two-layer expected

to be better,12, 14, 15, 18-24, 27, 34 and good-quality cotton T shirts in 2 layers14, 47, 50; in flat-cloth

experiments cotton 600 TPI in two layers,23 or cotton 600 TPI with flannel 90 TPI,23 performed

well. (Two-layer cotton 80 TPI did not perform well.23) Multiple layers should be used, at least

two, and preferably three or four. With fabric that stretches, such as T shirt fabric, it may be

important to use a design with edge stitching to prevent transmission of tension to the cloth,

which will increase the size of gaps in the material and affect filtration. There is a trade-off with

increased layers: they provide increased filtration efficiency, but also increase the resistance to

breathing, which increases the work of breathing, and may lead to discomfort and even to

Journ

al Pre-

Proof



reduced adherence. Increased resistance with increased layers also leads to increased edge

leak, decreasing the efficiency of the mask. People making masks for sale should specify the

materials (composition, weave, weight, thread count) for each layer and the number of layers

(e.g., cotton 100%, plain weave, 150 g/m2, 300 TPI; 3 layers). People making their own masks or

choosing a mask should consider these same factors, and also their planned activities while

wearing a mask. It might be sensible, for example, to choose a higher number of layers for

quietly sitting at a desk in a shared workspace for the duration of a working day, than for

grocery shopping in a ventilated environment with physical distancing. WHO guidance, 5 June

2020,9 based on expert opinion, recommends a three-layer mask, the outer layer and middle

layers hydrophobic (e.g., polypropylene, polyester and their blends) and the inner layer

hydrophilic (e.g., cotton or cotton blends).

Our recommendations for materials, above, are the same if using a bandana or scarf-type

design, though we would anticipate that this would be less efficient. Optimally, this will include

a prefolded shape, and a clear differentiation of outside and inside, such as this multi-layered

suggestion.71 Evidence on household filters is limited. The one included study of tissue paper

and paper towel did not report high efficiencies35: we think a third or fourth layer of cloth is

preferable to a disposable filter.

The information above, intended for the general public and for mask manufacturers, on

materials, designs, and correct use, can be found at clothmasks.ca. We will update this site as

new information is published.

What are the research priorities? An evidence-based cloth mask.

Reproducibly-described cloth and cloth masks should be tested in aerosol laboratories. The

effects of activity, time, and moisture18, 24 on effectiveness should be studied. The trade-off

between increased protection on the one hand, and decreased tolerability and increased leak

on the other, with higher thread counts and additional layers, should be explicitly explored.72

Women, children, and people who wear glasses require special consideration. Optimal methods

of laundering (home and industrial) and the effect of laundering on mask properties should be

studied.

Journ

al Pre-

Proof



Human trials should focus on best-performing cloth masks, should include healthcare workers,

other essential workers, and clients of essential services who can tolerate mask wearing, and

should study both inward and outward protection. Considerations are multi-faceted

educational interventions, measures of unintended consequences (e.g., incorrect mask use;

complacency about physical distancing and hand hygiene; mitigation of effects on people who

do not hear well); and address the impact of adherence on outcomes.

If reproducible designs of cloth mask that meet ASTM standards can be identified, this will have

direct and immediate impact in low- and middle-income countries. Widespread adoption in any

setting, including high-income countries, of reusable evidence-based cloth masks that meet the

standards of PPE would reduce the environmental impact of PPE, and mitigate supply problems

in this and future pandemics.

Standards for cloth masks have been developed in by the French standardization association,

Association française de normalisation (AFNOR): testing flat cloth using 3 µm particles, 70-<90%

filtration efficiency is designated category 2 (for use by the general public in a group of mask-

wearers) and 90-<95% category 1 (for use by non-health-care professionals, e.g., police, in

contact with the public).73-75 At the time of writing, a database of >1200 tested mask-material

combinations (many of them including non-woven synthetic materials such spunbond and

meltbond, that are used in formal PPE) has been compiled.75, 76

Businesses and academics in textiles, design and fashion are critical in embracing evolving

information on evidence-informed and evidence-based masks, and in using their specific

knowledge and skills to create a variety of masks that are not only functional, but comfortable

and stylish, to maximize the acceptability of mask wearing, particularly for young people.

Conclusions

Cloth masks can offer substantial filtration, in some cases equivalent to some medical masks.

This knowledge on hand can be used to create evidence-informed cloth masks to mitigate

transmissibility of viruses such as COVID-19. Aerosol laboratory testing of these masks may lead

to the design of evidence-based cloth masks, reproducibly-described so as to be manufactured

in diverse settings. No direct data with clinically-important outcomes are available.

Journ

al Pre-

Proof



Advocating for the public to wear cloth masks shifts the cost of a public health intervention

from society to the individual. In low-resource areas and for people living in poverty this may be

unacceptable and could be mitigated by public health interventions, with local manufacture

and distribution of evidence-informed and evidence-based cloth masks.

Acknowledgments

This work was created, in part, on the traditional territory shared between the Haudenosaunee confederacy and the Anishinabe nations, which was acknowledged in the Dish with One Spoon wampum belt.

We thank Melanie Chiarot, librarian, for her assistance in retrieving articles during a holiday period, and Robinson Healthcare Limited in the UK, who volunteered details of materials and methods for, and photographs of, the commercial 4-layer muslin mask.28 We thank the authors who responded promptly for additional details of their studies.14, 23, 32, 50 We thank the editors and anonymous reviewers whose suggestions greatly improved this work. We thank Dr Zain Chagla, infectious diseases consultant, McMaster University, for his peer review of clothmasks.ca. We dedicate this paper to Esta H McNett (“It was largely due to her unrelenting efforts that this study was undertaken”),25 and Charlotte Johnson,34 nurses who made important contributions to the field of mask design, but who were not acknowledged as authors in the manuscripts which included descriptions of their work. We thank the editors and the anonymous reviewers who encouraged us and improved our work immensely. We recognize the animals that suffered and died.25

Author contributions

The concept for the manuscript originated with C.M.C., M.J.J., J.F.E.M., R.P-F, W.C.W. and J.J.C.; C.M.C. and E.L.F. performed the original literature review; and C.M.C. wrote the first draft. E.L.F., A.A., I.A.C., G.K., and M.J. reviewed citations and performed data extraction; J.J.C. and E.L.F. collated discrepancies, and with C.M.C., resolved them by consensus. E.L.F. performed statistical analysis on reported data, and created graphs and tables. M.B.D. provided input on aerosol science and R.C.L.B. on virology and relevance to COVID-19. All authors contributed references and important revisions of content and meaning. C.M.C. is the guarantor for the work.

Disclosures

C.M.C. declares having received consultancy fees from Amgen, Astellas, Baxter, Boehringer-Ingelheim, Janssen, Johnson & Johnson, LEO Pharma, Pfizer, and Ministry of Health Ontario; is

Journ

al Pre-

Proof



expected to receive fees from Ministry of Health Ontario for future consultancy work; and speaker honoraria from Sanofi. M.J.J. is supported by a Medical Research Future Fund Next Generation Clinical Researchers Program Career Development Fellowship; is responsible for research projects that have received unrestricted funding from Gambro, Baxter, CSL, Amgen, Eli Lilly, and MSD; has served on advisory boards sponsored by Akebia, Astra Zeneca, Baxter, Boehringer Ingelheim, and Vifor; serves on Steering Committee for trials sponsored by Janssen and CSL; spoken at scientific meetings sponsored by Janssen, Amgen, Roche and Vifor; with any consultancy, honoraria or travel support paid to her institution. J.F.E.M. declares receipt of speaker honoraria from AstraZeneca, Bayer, Boehringer Ingelheim, Fresenius, Medice, Novartis, Novo Nordisk, Roche; research support from European Union, Canadian Institutes Health Research, Celgene, Novo Nordisk, Roche, Sandoz; and consultation fees from AstraZeneca, Bayer, Boehringer Ingelheim, Celgene, Novo Nordisk. R.P-F. declares having received consultancy fees from Akebia, AstraZeneca, Fresenius Medical Care, and Novo Nordisk; speaker honoraria from AstraZeneca and Novo Nordisk; and research support from Fresenius Medical Care. W.C.W. declares having received consultancy fees from Akebia, Amgen, AstraZeneca, Bayer, Daichii-Sankyo, Janssen, Merck, Relypsa, Vifor Fresenius Medical Care Renal Pharma and speaker honoraria from Fibrogen. J.J.C. declared having received consultancy fees from AstraZeneca and Baxter; speaker honoraria from AstraZeneca and Viforpharma; and research support from the Swedish Research Council (#2019-01059), Swedish Heart and Lung Foundation, AstraZeneca, Viforpharma and Astellas. E.L.F. and M.J. have no conflicts of interest.

References

1. Center for Systems Science and Engineering (CSSE) at Johns Hopkins University (JHU). COVID-19 dashboard, https://coronavirus.jhu.edu/map.html (2020, accessed last accessed 2020-06-19). 2. World Health Organisation. Advice on the use of masks the community, during home care and in health care settings in the context of the novel coronavirus (2019-nCoV) outbreak, https://www.who.int/docs/default-source/documents/advice-on-the-use-of-masks-2019-ncov.pdf (2020, accessed 2020-04-08). 3. World Health Organisation. Advice on the use of masks in the context of COVID-19: interim guidance, 6 April 2020, https://apps.who.int/iris/handle/10665/331693 (2020, accessed 2020-07-18). 4. Centers for Disease Control. Recommendation Regarding the Use of Cloth Face Coverings, Especially in Areas of Significant Community-Based Transmission, https://www.cdc.gov/coronavirus/2019-ncov/prevent-getting-sick/cloth-face-cover.html (2020, accessed 2020-04-09). 5. Council of Chief Medical Officers of Health Communication: Use of Non-Medical Masks (or Facial Coverings) by the Public, https://www.canada.ca/en/public-health/news/2020/04/ccmoh-communication-use-of-non-medical-masks-or-facial-coverings-by-the-public.html (2020, accessed 2020-

Journ

al Pre-

Proof



04-09). 6. Feng S, Shen C, Xia N, et al. Rational use of face masks in the COVID-19 pandemic. The Lancet Respiratory medicine 2020 2020/03/24. DOI: 10.1016/S2213-2600(20)30134-X. 7. Mahase E. Covid-19: What is the evidence for cloth masks? BMJ 2020; 369: m1422. 2020/04/09. DOI: 10.1136/bmj.m1422. 8. Japan Govt to Give 2 Masks to Every Household over Virus, https://www.nippon.com/en/news/yjj2020040101147/japan-govt-to-give-2-masks-to-every-household-over-virus.html (2020, accessed 2020-05-31). 9. World Health Organisation. Advice on the use of masks in the context of COVID-19: interim guidance 5 June 2020, https://www.who.int/publications/i/item/advice-on-the-use-of-masks-in-the-community-during-home-care-and-in-healthcare-settings-in-the-context-of-the-novel-coronavirus-(2019-ncov)-outbreak (2020, accessed 2020-06-07). 10. World Health Organisation. Shortage of personal protective equipment endangering health workers worldwide, https://www.who.int/news-room/detail/03-03-2020-shortage-of-personal-protective-equipment-endangering-health-workers-worldwide (2020). 11. Capps JA. Measures for the prevention and control of respiratory infections in military camps. JAMA 1918; 71: 448-451. 12. Cooper DW, Hinds WC, Price JM, et al. Common materials for emergency respiratory protection: leakage tests with a manikin. American Industrial Hygiene Association journal 1983; 44: 720-726. 1983/10/01. DOI: 10.1080/15298668391405634. 13. Dato VM, Hostler D and Hahn ME. Simple respiratory mask. Emerging infectious diseases 2006; 12: 1033-1034. DOI: 10.3201/eid1206.051468. 14. Davies A, Thompson KA, Giri K, et al. Testing the efficacy of homemade masks: would they protect in an influenza pandemic? Disaster medicine and public health preparedness 2013; 7: 413-418. DOI: 10.1017/dmp.2013.43. 15. Doust BC and Lyon AB. Face masks in infections of the respiratory tract. JAMA 1918; 71: 12216-11218. 16. Furuhashi M. A study on the microbial filtration efficiency of surgical face masks--with special reference to the non-woven fabric mask. The Bulletin of Tokyo Medical and Dental University 1978; 25: 7-15. 17. Greene VW and Vesley D. Method for evaluating effectiveness of surgical masks. Journal of bacteriology 1962; 83: 663-667. 18. Guyton HG, Decker HM and Anton GT. Emergency respiratory protection against radiological and biological aerosols. AMA archives of industrial health 1959; 20: 91-95.

Journ

al Pre-

Proof



19. Haller DA and Colwell RC. The protective qualities of the gauze face mask - experimental studies. JAMA 1918; 71: 1213-1215. 20. Jang JY and Kim SW. Evaluation of Filtration Performance Efficiency of Commercial Cloth Masks. J Environ Health Sci 2015; 41: 203-215. 21. Jung H, Kim J, Lee SA, et al. Comparison of Filtration Efficiency and Pressure Drop in Anti-Yellow Sand Masks, Quarantine Masks, Medical Masks, General Masks, and Handkerchiefs. Aerosol and Air Quality Research 2014; 14: 991-1002. 22. Kellogg WH and MacMillan G. An experimental study of the efficacy of gauze face masks. American Journal of Public Health 1920; 10: 34-42. 23. Konda A, Prakash A, Moss GA, et al. Aerosol Filtration Efficiency of Common Fabrics Used in Respiratory Cloth Masks. ACS nano 2020; 14: 6339-6347. DOI: https://dx.doi.org/10.1021/acsnano.0c03252. 24. Leete HM. Some experiments on masks. The Lancet 1919; 193: 392–393. 25. Lurie MB and Abramson S. The Efficiency of Gauze Masks in the Protection of Rabbits against the Inhalation of Droplet Nuclei of Tubercle Bacilli. American Review of Tuberculosis 1949; 59: 1-9. 26. MacIntyre CR, Seale H, Dung TC, et al. A cluster randomised trial of cloth masks compared with medical masks in healthcare workers. BMJ open 2015; 5: e006577. DOI: 10.1136/bmjopen-2014-006577. 27. Paine CG. The aetiology of puerperal infection. BMJ 1935; 1: 243-246. 28. Quesnel LB. The efficiency of surgical masks of varying design and composition. The British journal of surgery 1975; 62: 936-940. DOI: 10.1002/bjs.1800621203. 29. Rengasamy S, Eimer B and Shaffer RE. Simple respiratory protection--evaluation of the filtration performance of cloth masks and common fabric materials against 20-1000 nm size particles. The Annals of occupational hygiene 2010; 54: 789-798. DOI: 10.1093/annhyg/meq044. 30. Shakya KM, Noyes A, Kallin R, et al. Evaluating the efficacy of cloth facemasks in reducing particulate matter exposure. Journal of exposure science & environmental epidemiology 2017; 27: 352-357. DOI: 10.1038/jes.2016.42. 31. Shooter RA, Smith MA and Hunter CJ. A study of surgical masks. The British journal of surgery 1959; 47: 246-249. DOI: 10.1002/bjs.18004720312. 32. van der Sande M, Teunis P and Sabel R. Professional and home-made face masks reduce exposure to respiratory infections among the general population. PloS one 2008; 3: e2618. DOI: 10.1371/journal.pone.0002618.

Journ

al Pre-

Proof



33. Weaver GH. The value of the face mask and other measures in prevention of diphtheria, meningitis, pneumonia, etc. JAMA 1918; 70: 76-78. 34. Weaver GH. Droplet Infection and Its Prevention by the Face Mask. The Journal of Infectious Diseases 1919; 24: 218-230. 35. Zhao M, Liao L, Xiao W, et al. Household Materials Selection for Homemade Cloth Face Coverings and Their Filtration Efficiency Enhancement with Triboelectric Charging. Nano letters 2020; 20: 5544-5552. DOI: 10.1021/acs.nanolett.0c02211. 36. Konda A, Prakash A, Moss GA, et al. Correction to Aerosol Filtration Efficiency of Common Fabrics Used in Respiratory Cloth Masks. ACS Nano 2020. 37. ASTM International. ASTM Reading Room. COVID-19 Related Standards: Masks, https://www.astm.org/READINGLIBRARY/VIEW/PHMSA.html (2020, accessed 2020-04-27; behind paywall). 38. The National Institute for Occupational Safety and Health (NIOSH). NIOSH Guide to the Selection and Use of Particulate Respirators Certified under 42 CFR 84: DHHS (NIOSH) Publication Number 96-101, https://www.cdc.gov/niosh/docs/96-101/default.html (1996, accessed 2020-06-13). 39. Occupational Safety and Health Administration. Assigned Protection Factors for the Revised Respiratory Protection Standard, https://www.osha.gov/Publications/3352-APF-respirators.pdf (2009, accessed Last accessed 2020-06-13). 40. Canadian Standards Association. CAN/CSA Z94.4-2018: Selection, Use, And Care Of Respirators, https://webstore.ansi.org/Standards/CSA/CSAZ942018?gclid=Cj0KCQjwuJz3BRDTARIsAMg-HxX4F2MYi5rxnJQXstCHVUXJz-2ugsaSNEoW3A_douYB5OvGwd23Xm0aApG8EALw_wcB (2018, accessed Last accessed 2020-06-15 (behind paywall)). 41. Varga Z, Flammer AJ, Steiger P, et al. Electron microscopy of SARS-CoV-2: a challenging task - Authors' reply. Lancet 2020; 395: e100. DOI: 10.1016/S0140-6736(20)31185-5. 42. SEMLab Inc. Fiber Analysis, https://www.semlab.com/services/materials-analysis/fiberanalysis/ (2020, accessed Last accessed 2020-06-14). 43. Verma D. Effectiveness of DIY masks: How do they work?, https://www.nanoscience.com/applications/materials-science/effectiveness-of-diy-masks-how-do-they-work/ (2020, accessed Last accessed 2020-06-14). 44. Laube BL and Dolovich MB. Aerosols and Aerosol Drug Delivery Systems. Middleton's Allergy. Eighth Edition ed., 2014, pp.1066-1082. 45. Han ZY, Weng WG and Huang QY. Characterizations of particle size distribution of the droplets exhaled by sneeze. Journal of the Royal Society, Interface 2013; 10: 20130560. DOI: 10.1098/rsif.2013.0560.

Journ

al Pre-

Proof



46. VanSciver M, Miller S and Hertzberg J. Particle Image Velocimetry of Human Cough. Aerosol Science and Technology 2011; 45: 415-422. 47. Davies A, Thompson KA, Giri K, et al. Frequently asked questions - homemade facemasks study, https://www.researchgate.net/publication/340253593_Frequently_asked_questions_-_homemade_facemasks_studypdf (2020, Last accessed 2020-06-10; requires membership). 48. McKhann CF, Steeger A and Long AP. Hospital infection - A survey of problem. Arch Pediatr Adolesc Med 1938; 55: 579-599. 49. Thompson ME and Gillespie WA. Nasal carriage of Staphylococcus aureus by nurses. The Journal of pathology and bacteriology 1958; 75: 351-355. DOI: 10.1002/path.1700750213. 50. Davies A, Thompson KA, Giri K, et al. Facemask instructions COVID-19.doc, https://www.researchgate.net/publication/340023785_facemask_instructions_COVID-19doc (2020, accessed 2020-05-02; requires membership). 51. Dyer E. Some health experts questioning advice against wider use of masks to slow spread of COVID-19, https://www.cbc.ca/news/politics/covid-19-pandemic-coronavirus-masks-1.5515526 (2020, accessed 2020-07-18). 52. Wheelahan D. Cloth masks: Dangerous to your health?, https://www.sciencedaily.com/releases/2015/04/150422121724.htm (2015, accessed Last accessed 2020-06-10). 53. Greenhalgh T, Schmid MB, Czypionka T, et al. Face masks for the public during the covid-19 crisis. BMJ 2020; 369: m1435. DOI: 10.1136/bmj.m1435. 54. Clase CM, Beale RCL, Dolovich MB, et al. Cloth masks may prevent transmission of COVID-19: an evidence-based, risk-based approach, https://www.acpjournals.org/doi/10.7326/M20-2567 (2020). 55. Chu DK, Akl EA, Duda S, et al. Physical distancing, face masks, and eye protection to prevent person-to-person transmission of SARS-CoV-2 and COVID-19: a systematic review and meta-analysis. Lancet 2020. DOI: 10.1016/S0140-6736(20)31142-9. 56. Tuan PA, Horby P, Dinh PN, et al. SARS transmission in Vietnam outside of the health-care setting. Epidemiology and infection 2007; 135: 392-401. DOI: 10.1017/S0950268806006996. 57. Lau JT, Lau M, Kim JH, et al. Probable secondary infections in households of SARS patients in Hong Kong. Emerging infectious diseases 2004; 10: 235-243. DOI: 10.3201/eid1002.030626. 58. Wu J, Xu F, Zhou W, et al. Risk factors for SARS among persons without known contact with SARS patients, Beijing, China. Emerging infectious diseases 2004; 10: 210-216. DOI: 10.3201/eid1002.030730.

Journ

al Pre-

Proof



59. Hill AB. The environment and disease: association or causation? 1965. Journal of the Royal Society of Medicine 2015; 108: 32-37. DOI: 10.1177/0141076814562718. 60. Rothe C, Schunk M, Sothmann P, et al. Transmission of 2019-nCoV Infection from an Asymptomatic Contact in Germany. N Engl J Med 2020; 382: 970-971. 2020/02/01. DOI: 10.1056/NEJMc2001468. 61. Pan X, Chen D, Xia Y, et al. Asymptomatic cases in a family cluster with SARS-CoV-2 infection. The Lancet Infectious diseases 2020; 20: 410-411. DOI: 10.1016/S1473-3099(20)30114-6. 62. Bai Y, Yao L, Wei T, et al. Presumed Asymptomatic Carrier Transmission of COVID-19. JAMA 2020. DOI: 10.1001/jama.2020.2565. 63. Kimball A, Hatfield KM, Arons M, et al. Asymptomatic and Presymptomatic SARS-CoV-2 Infections in Residents of a Long-Term Care Skilled Nursing Facility - King County, Washington, March 2020. MMWR Morbidity and mortality weekly report 2020; 69: 377-381. DOI: 10.15585/mmwr.mm6913e1. 64. Miller A. WHO backtracks on claim that asymptomatic spread of COVID-19 is 'very rare', https://www.cbc.ca/news/health/who-covid-19-asymptomatic-spread-1.5604353 (2020). 65. Li R, Pei S, Chen B, et al. Substantial undocumented infection facilitates the rapid dissemination of novel coronavirus (SARS-CoV-2). Science 2020; 368: 489-493. DOI: 10.1126/science.abb3221. 66. He X, Lau EHY, Wu P, et al. Temporal dynamics in viral shedding and transmissibility of COVID-19. Nat Med 2020; 26: 672-675. DOI: 10.1038/s41591-020-0869-5. 67. Zou L, Ruan F, Huang M, et al. SARS-CoV-2 Viral Load in Upper Respiratory Specimens of Infected Patients. N Engl J Med 2020; 382: 1177-1179. DOI: 10.1056/NEJMc2001737. 68. Wei WE, Li Z, Chiew CJ, et al. Presymptomatic Transmission of SARS-CoV-2 - Singapore, January 23-March 16, 2020. MMWR Morbidity and mortality weekly report 2020; 69: 411-415. DOI: 10.15585/mmwr.mm6914e1. 69. Stutt ROJH, Retkute R, Bradley M, et al. A modelling framework to assess the likely effectiveness of facemasks in combination with ‘lock-down’ in managing the COVID-19 pandemic. Proc R Soc A 2020: 20200376. 70. Davies A, Thompson KA, Giri K, et al. Frequently asked questions - homemade facemasks study, https://www.researchgate.net/publication/340253593_Frequently_asked_questions_-_homemade_facemasks_studypdf (2020, 2020-06-10; requires membership). 71. Mister Bandana. Make a bandana face mask in under 1 minute with just two materials, https://www.youtube.com/watch?v=7TasM4uRj74 (2020, accessed 2020-06-15).

Journ

al Pre-

Proof



72. Podgórski A, Balazy A and Gradoń L. Application of nanofibers to improve the filtration efficiency of the most penetrating aerosol particles in fibrous filters. Chem Eng Sci 2006; 61: 6804-6815. 73. EuraMaterials, Institut Français du Textile et de L'Habillement and Techtera. Actualisation de la base selon la note interministerielle du 29 Mars 2020 [Updating of the database in keeping with to the interministerial note of March 29, 2020] Updated 26/05/2020, https://euramaterials.eu/wp-content/uploads/2020/05/base-de-donnees-matieres-resultats-dga-maj-26052020.pdf (2020, accessed 2020-07-18). 74. Euramaterials, Institut Français du Textile et de L'Habillement and Techtera. Note d'information à destination des industriels. Base de données du groupe matière: masques à Usage Non Sanitaire (UNS) [Information note for manufacturers. Material group database: Non-Medical Masks], https://euramaterials.eu/wp-content/uploads/2020/04/2020-04-28-Note-dinformation-BdD-mati%C3%A8res.pdf (2020, accessed 2020-07-18). 75. Euramaterials. Masques à usage non sanitaire : les résultats des tests de caractérisation matière de la DGA [Non-medical masks: results of tests for the Direction générale de l'armement], https://euramaterials.eu/masques-anti-projections-les-resultats-des-tests-de-caracterisation-matiere-de-la-dga/ (2020, accessed 2020-07-18). 76. Afnor groupe. Protective masks: download our reference document for free, https://www.afnor.org/en/news/protective-masks-download-our-reference-document-for-free/ (2020, accessed 2020-07-24).

Journ

al Pre-

Proof

Box

We searched Medline and Embase, and used Google, using search terms ‘cloth’, ‘fabric’, ‘gauze’, ‘cotton’, ‘mask’ and ‘filtration’ and synonyms for articles on the filtration properties of flat cloth or cloth masks or face coverings. We reviewed the reference lists of relevant articles and review articles to identify further articles. Our selection of articles for this review was unbiased; i.e., it did not depend on the direction of the results. We did not conduct a systematic review or search grey literature. We identified 25 articles that described filtration properties of cloth or cloth masks, some of which included medical masks and N95 respirators as comparators. Most studies used surrogates for filtration, sometimes graded by particle size. Some studies used bioaerosols, usually bacterial. A minority of papers used virus; one study used SARS-CoV2. Studies of the filtration properties of flat cloth used a variety of methods, few of which were the equivalent to the standard methods used by the American Society for Testing and Materials. Study of protection for the wearer often used a manikin wearing a mask, with airflow to simulate different breathing rates. Studies of protection of the environment, also known as source control, used convenience samples of healthy volunteers, often the investigators themselves; the volunteers are rarely characterized in any way. The design and execution of the studies was generally rigorously described; the number of replicates and variance of estimates less well described, and it was unusual to find statistical comparisons between different types of cloth or types of mask. Many descriptions of cloth were lacking in the detail required for reproducibility; no study gave all the expected details of material, thread count, weave, and weight. Some of the homemade mask designs were reproducible.

Journ

al Pre-

Proof

Table 1. Summary of filtration efficiency for flat cloth, including details of methodology: aerosol used, aerosol size and flow rate.

Composition Weight, Weave, Thread Count

1 layer 2 layer Aerosol Measured Aerosol Size

Flow rate (L/min)

Standard

Cotton ‘Quilters’ cotton’

80 TPI 4% 32% NaCl solution 75-100 nm ~ 3.5‡ No Konda 202023, 36

80 TPI 3% NaCl solution 75-100 nm ~ 9‡ No Konda 2020 80 TPI 6% 50% NaCl solution 2-3 µm ~ 3.5‡ No Konda 2020 80 TPI 34% NaCl solution 2-3 µm ~ 9‡ No Konda 2020

Cotton 600 TPI (#1 in hybrids, below) 600 TPI 76% 85% NaCl solution 75-100 nm ~ 3.5‡ No Konda 2020 600 TPI 98% 99.5% NaCl solution 2-3 µm ~ 3.5‡ No Konda 2020

Gauze, cotton NA 1% 1% 4 layer 4%

NaCl solution 75 nm§ 85 NIOSH Jung 201421

T shirt: 100% cotton Knit 69% 71% B atrophaeus NA 30 No Davies 201314

Knit 51% Bacteriophage MS2 NA 30 No Davies 2013 T shirt, Hanes: 100% cotton Knit 9% NaCl solution 1 µm 5.5 cm/s No Rengasamy

201029

12% NaCl solution 1 µm 17 cm/s No Rengasamy 2010 T shirt, cotton Knit

157 g/m2 22% NaCl solution 75 nm§ 32 No Zhao 202035

Sweater, cotton Knit 360 g/m2

26% NaCl solution 75 nm§ 32 No Zhao 2020

Towel, Pem America: 100% cotton NA 23% NaCl solution 1 µm 5.5 cm/s No Rengasamy 201029

49% NaCl solution 1 µm 17 cm/s No Rengasamy 2010 Towel, Pinzon: 100% cotton NA 30% NaCl solution 1 µm 5.5 cm/s No Rengasamy 2010

58% NaCl solution 1 µm 17 cm/s No Rengasamy 2010 Towel, Aquis: 100% cotton NA 33% NaCl solution 1 µm 5.5 cm/s No Rengasamy 2010

0% NaCl solution 1 µm 17 cm/s No Rengasamy 2010 Scarf: cotton NA 62% 71% B atrophaeus NA 30 No Davies 201314

NA 49% Bacteriophage MS2 NA 30 No Davies 2013 Scarf, Pocket Square: 100% cotton NA 0% NaCl solution 1 µm 5.5 cm/s No Rengasamy

201029

0% NaCl solution 1 µm 17 cm/s No Rengasamy 2010

Journ

al Pre-

Proof

Scarf, Seed Supply: 100% cotton NA 1% NaCl solution 1 µm 5.5 cm/s No Rengasamy 2010 7% NaCl solution 1 µm 17 cm/s No Rengasamy 2010

Pillowcase

NA 61% 62% B atrophaeus NA 30 No Davies 201314

NA 57% Bacteriophage MS2 NA 30 No Davies 2013 Pillowcase: 100% cotton 116 g/m2 5% NaCl solution 75 nm§ 32 No Zhao 202035

Handkerchief, cotton NA 1% 2% 4 layer 4%


Linen Tea towel NA 83% 97% B atrophaeus NA 30 No Davies 201314

NA 72% Bacteriophage MS2 NA 30 No Davies 2013 Linen NA 60% B atrophaeus NA 30 No Davies 2013

NA 62% Bacteriophage MS2 NA 30 No Davies 2013 Silk Silk NA 58% B atrophaeus NA 30 No Davies 2013

NA 54% Bacteriophage MS2 NA 30 No Davies 2013 Silk 100% (#2 in hybrids, below)

39 g/m2 145 TPI†

54% 65% 4 layers: 84%

NaCl solution 75-100 nm ~ 3.5‡ No Konda 202023

39 g/m2 145 TPI†

55% 66% 4 layers: 89%

NaCl solution 2-3 µm ~ 3.5‡ No Konda 2020

Napkin: Silk Woven 84 g/m2


Manmade Chiffon: 90% polyester, 10% spandex (#3 in hybrids, below)

195 TPI† 58% 86% NaCl solution 75-100 nm ~ 3.5‡ No Konda 202023

195 TPI† 24% NaCl solution 75-100 nm ~ 9‡ No Konda 2020 195 TPI† 73% 90% NaCl solution 2-3 µm ~ 3.5‡ No Konda 2020 195 TPI† 53% NaCl solution 2-3 µm ~ 9‡ No Konda 2020

Interfacing material: polypropylene Spunbond 30 g/m2


Scarf, Walmart Fleece: 100% polyester NA 25% NaCl solution 1 µm 5.5 cm/s No Rengasamy 201029

14% NaCl solution 1 µm 17 cm/s No Rengasamy 2010 Toddler wrap: polyester Knit


Exercise pants: nylon Woven 23% NaCl solution 75 nm§ 32 No Zhao 2020

Journ

al Pre-

Proof

164 g/m2 Composites Cotton mix NA 75% B atrophaeus NA 30 No Davies 201314

NA 70% Bacteriophage MS2 NA 30 No Davies 2013 Flannel: 65% cotton, 35% polyester (#4 in hybrids, below)

90 TPI† 55% NaCl solution 75-100 nm ~ 3.5‡ No Konda 202023

90 TPI† 13% NaCl solution 75-100 nm ~ 9‡ No Konda 2020 90 TPI† 44% NaCl solution 2-3 µm ~ 3.5‡ No Konda 2020 90 TPI† 46% NaCl solution 2-3 µm ~ 9‡ No Konda 2020

Norma Kamali sweatshirt: 85% cotton, 15% polyester

NA 8% NaCl solution 1 µm 5.5 cm/s No Rengasamy 201029

NA 26% NaCl solution 1 µm 17 cm/s No Rengasamy 2010 Hanes sweatshirt: 70% cotton, 30% polyester

NA 19% NaCl solution 1 µm 5.5 cm/s No Rengasamy 2010 NA 15% NaCl solution 1 µm 17 cm/s No Rengasamy 2010

Faded Glory sweatshirt: 60% cotton, 40% polyester


Dickies T shirt: 99% cotton, 1% polyester NA 8% NaCl solution 1 µm 5.5 cm/s No Rengasamy 2010 NA 20% NaCl solution 1 µm 17 cm/s No Rengasamy 2010

Faded Glory T shirt: 60% cotton, 40% polyester


Paper Paper towel: cellulose Bonded


Tissue paper: cellulose Bonded 33 g/m2


Copy paper: cellulose* Bonded 73 g/m2

99.9% NaCl solution 75 nm§ 32 No Zhao 2020

Hybrids Cotton/Silk (1 layer of #1 above**, 1 layer of #2 above)

96% NaCl solution 75-100 nm ~ 3.5‡ No Konda 202023

97% NaCl solution 2-3 µm ~ 3.5‡ No Konda 2020 Cotton/Chiffon (1 layer of #1 above**, 1 layer of #3 above)

97% NaCl solution 75-100 nm ~ 3.5‡ No Konda 2020 99.5% NaCl solution 2-3 µm ~ 3.5‡ No Konda 2020

Cotton/Flannel (1 layer of #1 above**, 1 layer of #4 above)

95% NaCl solution 75-100 nm ~ 3.5‡ No Konda 2020 96% NaCl solution 2-3 µm ~ 3.5‡ No Konda 2020

Cloth mask material Commercial mask fabric A, bleached cotton

96 TPI 69% Staph aureus NA 8 US military standard,

Furuhashi 197816

Journ

al Pre-

Proof

1978 Commercial mask fabric D, bleached cotton

86 TPI 43% Staph aureus NA 8 US military standard, 1978

Furuhashi 1978

Commercial mask fabric B, calico 160 TPI 73% Staph aureus NA 8 US military standard, 1978

Furuhashi 1978

Commercial mask fabric C, twill weave NA 94% Staph aureus NA 8 US military standard, 1978

Furuhashi 1978

Cloth mask A: 50% nylon, 40% polypropylene, 10% polyurethane

1.22 mm thick

29% 59%, 4 layer 75%

NaCl solution 0.3-0.5 µm NA No Jang 201520

1.22 mm thick

60% 70%, 4 layer 94%

NaCl solution 2-5 µm NA No Jang 2015

Cloth mask B: 84% nylon, 12% polyester, 4% spandex

0.62 mm thick

28% 32% 4 layer 67%


0.62 mm thick

63% 71% 4 layer 77%


Cloth mask C: 100% polyester 0.29 mm thick

18% 50%, 4 layer 55%


0.29 mm thick

45% 78%, 4 layer 81%


Cloth mask D: 100% polyester microfibre

0.30 mm thick

9% 45% 4 layer 62%


0.30 mm thick

45% 59% 4 layer 99%


Journ

al Pre-

Proof

Cloth mask E: 100% polyester microfibre 2.77 mm thick

27% NaCl solution 0.3-0.5 µm NA No Jang 2015

2.77 mm thick

80% NaCl solution 2-5 µm NA No Jang 2015

Cotton mask: surgical type, 4 distinct masks

NA 23% SD 27


Cloth mask NA 3% NaCl solution 75 nm§ NA AS/ NZS1716

MacIntyre 201526

Respro Bandit NA 22% NaCl solution 1 µm 5.5 cm/s No Rengasamy 201029

NA 34% NaCl solution 1 µm 17 cm/s No Rengasamy 2010 Breath Health NA 13% NaCl solution 1 µm 5.5 cm/s No Rengasamy 2010

NA 44% NaCl solution 1 µm 17 cm/s No Rengasamy 2010 Breath Health Fleece NA 22% NaCl solution 1 µm 5.5 cm/s No Rengasamy 2010

NA 13% NaCl solution 1 µm 17 cm/s No Rengasamy 2010 Medical mask material Mölnlycke Health Care Barrier 4239 96% B atrophaeus NA 30 No Davies 201314

90% Bacteriophage MS2 NA 30 No Davies 2013 Hopes Fine glass

fiber with non-woven fabric

98% Staph aureus NA 8 “US military standard” 1978

Furuhashi 197816

Medispo Fine glass fiber with non-woven fabric

99% Staph aureus NA 8 “US military standard” 1978

Furuhashi 1978

Medical mask material NA 56% NaCl solution 75 nm§ AS/NZS1716

MacIntyre 201526

R class respirator material 1.81 mm thick

91% NaCl solution 0.3-0.5 µm NA No Jang 201520

100% NaCl solution 2-5 µm NA No Jang 2015 Medical mask: surgical type, 4 distinct masks

NA 41% SD 38


Medical mask: dental type, 5 distinct masks

NA 71% SD 12


Journ

al Pre-

Proof

For experimental details and additional studies, see supplementary table

We extracted information on weight, weave, thread count and thickness for each material when available.

NA not available; TPI threads per inch (number of threads in warp plus number of threads in weft); SD standard deviation

§TSI filter tester generates NaCl aerosol with count mean diameter 75 nm and geometric standard deviation 1.75

* this is writing paper, and obviously not breathable

** The 600 TPI cotton was used in the hybrid experiments (personal communication, Supratik Guha)

†calculated from pitch, the distance between the centre of one thread and the next

‡Konda et al23 measured cloth using a system that produced initial flow rates of 35 L/min and 90 L/min respectively; however, when cloth was inserted, increasing the resistance, the flow rate fell, probably by an order of magnitude (personal communication, Supratik Guha), and erratum.36 We have reflected this by reporting a flow rate that is the initial divided by 10, and by indicating that it is approximate (~); we thought this preferable to giving no indication. The experiments in this paper include some readings done with a gap past the filter, to simulate edge leak. We have extracted these data in the supplementary material but here present the results of flat cloth with no gap.

When multiple data points were available, we extracted data closest to 100 nm (used for testing particle filtration efficiency for medical masks, according to ASTM standards) and 3000 nm (3 µm) (used for testing bioaerosol filtration efficiency for medical masks, according to ASTM standards). To be conservative, we selected the closest point below the target particle size.

Many original papers provided a measure of error variance. We have not extracted these data to this table for readability. They are often wide, in the 10 – 30% range. We report the SD for Jung et al 2014 because it reflects the differences in properties of a number of distinct masks of different materials (4 surgical, 3 dental and 5 cotton; tested in triplicate for each design) that are not reported separately, not the error variance of a single mask.

We did not extract data for N95 mask material and medical mask material from Konda et al,23 because the methodology used for testing fabric by them were under conditions different than are those used for specifying fitted protective equipment such as the N95 respirators, which are tested under higher differential pressures and flows (personal communication, Supratik Guha).

Journ

al Pre-

Proof

Table 2. Filtration efficiency for homemade 2-layer T shirt, and disposable commercial medical masks according to particle diameter from 21 volunteers coughing, recalculated from Davies and colleagues14; P values are our calculations, difference between two proportions, using R (R foundation, Vienna, Austria).

Particle Diameter (μm) 2-layer T Shirt Mask Medical Mask P >7 67% 44% 0.14 4.7-7 61% 61% 1.00 3.3-4.7 20% 20% 1.00 2.1-3.3 85% 89% 0.70 1.1-2.1 84% 94% 0.31 0.65-1.1 71% 86% 0.24 Total 79% 85% 0.62

Bacterial filtration efficiency was calculated as (Bacterial counts without mask - Bacterial counts with mask)/Bacterial counts without mask.

Journ

al Pre-

Proof

Table 3. Protection factor and filtration efficiency for homemade and medical masks according to particle diameter from 28 volunteers (inward, immediate), 22 volunteers (inward, after 3 hours), and data from a manikin wearing a mask (outward), recalculated from van der Sande and colleagues.32

Filtration efficiency Inward Outward Immediate After 3 hours Immediate 1-layer Tea-Towel Mask 55 – 69% 63 - 77% 17% Medical Mask 76 – 81% 74 – 83% 58%

Filtration efficiency is calculated as 1- (1/Protection factor). Both adults and children were studied in short term, with somewhat lower performance in children; we have extracted the adult data for consistency with the rest of the literature. For each experimental condition, we have extracted the highest and lowest median efficiencies from the data provided. Outward data were read from graphs. Because medians were reported, statistical testing was not possible.

Journ

al Pre-

Proof

Table 4. Summary of filtration efficiency for cloth masks, inward protection (protecting the wearer), assessed at < 1μm particle size. From the paper by Shakya and colleagues, we excluded cloth mask 1 because it had an exhalation valve that may have improved its performance; we included the latex particle data because they are comparable with other experiments, but not the data obtained with diesel combustion particles. (These data are shown in the supplementary table.) When a medical mask was included as a comparator, we have also shown the data for the medical mask.

First Author,

Year

Cloth Mask Detailed

Description Testing Device and Particle Size Details

Cloth Mask Filtration Efficiency

Medical Mask Filtration Efficiency

Dato 200613

T-shirt mask made by the authors to fit their own faces; 8-layer high quality preshrunk cotton T-shirt fabric (Hanes Heavweight T-shirt) with 3 sets of ties

The authors as volunteers

Portacount 0.02-1 µm

Author 1 99% Author 2 92% Author 3 94%

Cloth Mask Medical Mask Initial After the

other exercises

Initial After the other

exercises Davies 201314

T-shirt mask by unskilled volunteers, to a pattern, without assistance; 2-layer T-shirt fabric, pleated design

Volunteers Portacount 0.02-1 µm

Normal breathing

50% 50% 83% 80%

Heavy breathing

50% 86%

Shaking head

50% 80%

Nodding 50% 80% Bending over

0% 67%

Talking 50% 83% Overall 50% 80%

Cloth Mask Medical Mask Flow rate Flow rate 8L/min 19L/min 8L/min 19L/min Shakya 201730

Purchased from street vendor, Kathmandu Nepal in 2014; simple cloth rectangles (layers unknown) with ear loop,

Manikin Particle counter 30 nm

Cloth mask 2

89% 15% 91% 62%

Cloth mask 3

54% 26%

100 nm Cloth mask 2

57% 32% 94% 70%

Cloth mask 3

57% 27%

500 nm Cloth mask 2

47% 57% 92% 65%

Cloth mask 45% 31%

Journ

al Pre-

Proof

cloth not specified

3 1 µm Cloth mask

2 69% 54% 99% 96%

Cloth mask 3

85% 49%

Cloth Mask Medical Mask Short

term After 3 hours

Short term

After 3 hours

Van der Sande 200832

Cloth mask, homemade, made of TD Cerise Multi teacloths (tea-towel), Blokker; one layer mask

Volunteers Portacount 0.02-1 µm

Sitting quietly

60% 69% 76% 77%

Nodding 55% 63% 79% 78% Shaking head

55% 66% 80% 76%

Reading 69% 77% 81% 83% Walking 58% 66% 76% 74%

Journ

al Pre-

Proof

Figure 1. Schematic showing different types of filtration experiments.

A: an experiment on a flat cloth sample or mask material sample (the filter). The surface area of the sample tested, the particle size, particle composition and the flow rate should be defined. The pressure drop across the filter under these or other, specified, conditions can be measured. There is no edge leak. All the particles that contribute to the concentration on the protected side of the filter have penetrated the material. The TSI 8130 filter tester (TSI, Auburn, IL, US) is an example of such a system. Using this type of experiment, the American Society for Testing and Materials (ASTM) defines the standards for testing material for medical masks, and the National Institute for Occupational Safety and Health (NIOSH) the standards for testing material for respirators (N95-type masks).

B: an inward protection experiment on a mask using a human volunteer or a manikin. For human volunteers, the concentration inside the mask is measured using a thin-walled tube, called a probe, that fits across the mask material. For a manikin, a pump will simulate breathing and the concentration inside the mask can be measured at any point in the circuit. The concentration outside the mask is measured from the surrounding air. The TSI Portacount (TSI, Auburn, IL, US) is an example of such a system, and can be used with human volunteers or with a manikin. When the particles to be measured are inert and harmless, such as the saline aerosol typically used with the Portacount, the experiment can be conducted in an ordinary room without special conditions. The concentration of particles on the protected side (the inside) is a combination of penetration of the mask (through the material) and edge leak (around the material); it measures both the material and the fit. In experiments using human volunteers, a variety of activities can be undertaken to further challenge the mask (e.g., deep breathing, head movement, bending). In experiments using manikins, the flow rate can be adjusted to simulate different levels of minute respiration, corresponding to different activity levels. For medical masks, no relevant standards have been defined. For respirators (N95-type masks), the Occupational Safety and Health Administration requires that N95 masks be fitted to the individual who will wear them. This can be done quantitatively using a device such as the Portacount, or non-quantitatively, using a strong-tasting substance such as saccharin.

C: an outward protection experiment on a mask using a manikin. The aerosol is generated and passed through the manikin into the mask. The concentration of the aerosol on the source side can be measured at any point in the circuit. The concentration of the aerosol on the protected side is measured from the environment. The apparatus is protected in a chamber filled with filtered air, to ensure that all particles outside the mask have come from the manikin. The concentration of particles on the protected side (the outside) is a combination of penetration of the mask (through the material) and edge leak (around the material). There are no standards that relate to this design.

D: an outward protection experiment using human volunteers. The aerosol is generated through human activity – breathing, talking, or coughing. Usually these are bioaerosol experiments, measuring normal human mouth flora, or pathogens from volunteers who are unwell. (In some experiments, to standardize the experiment and to increase the concentration of bacteria in the aerosol, volunteer-investigators contaminated their mouths with non-pathogenic bacteria and studied transmission specifically of that species.) There are no standards that relate to this design.

Journ

al Pre-

Proof

chigh particle concentration on the source side of the filter; clow particle concentration on the protected side of the filter; Delta P pressure drop across the filter

Journ

al Pre-

Proof

Supplementary table. Experimental design and detailed results of included studies

First author, year

Masks tested

Description of cotton mask including material, weight, weave (twill or twisted), thread count, and number of layers

Details of experiments performed

Details of what was sampled

Outward, inward both, neither

Filtration efficiency

Capps 19181

Cloth Mask

Rectangular mask measuring 5 x 7 inches made of 3 or 4 layers of gauze, weight NA, weave NA, thread count NA

No experiment was set up; rather it was an observation of patients and physicians wearing the masks at the infirmary, ambulance, office and wards.

The study measured the effectiveness of the entire method of masking and cubical quarantining for prevention of the spread of respiratory infectious diseases like measles and scarlet fever

Both No mask efficiency was reported but the system as a whole (cubical quarantining plus masking) was 95% and 100% effective for preventing scarlet fever and measles, respectively

Cooper 19832

Cloth Mask (cotton/polyester shirt material, cotton handkerchief material, toweling) Surgical mask (Johnson & Johnson

In total, 3 cloth masks were tested: 1. Shirt made with oxford cloth 65% fortel polyester and 35% cotton, weight NA, weave NA, thread count 46/inch by 46/inch, 4 layers 2. Handkerchief white broadcloth 100% cotton, weight NA, weave NA, thread count 66/inch by 58/inch, 4 layers 3. Toweling terryweave 88% cotton and 12% dacron polyester, weight NA, weave NA, thread count NA, 1 or 2 layers

Different material masks were fastened on a mannequin head and aerosol inward leakage and penetration were measured using fluorescent aerosols and a filter located inside the mannequin head’s mouth. The filter was 47mm in diameter and was cleaned after every experiment. Aerosols (fluorescent dioctyl phthalate aerosol 1.8 μm in diameter) were generated using a Thermo-Systems incorporated (TSI)

The dioctylphthalate fluorescence on the filter was sampled, the fluorescence was measured and concentration determined by comparison with a standard curve and linear regression. Four tests were performed and mean of leakage plus penetration was calculated.

Inward 3M nylon hosiery: 99.42% 3M fully taped: 98.5% 3M strapped: 81% J&J fully taped: 95.8% J&J tied: 64% Shirt-oxford cloth fully taped: 69% Shirt-oxford cloth corners taped: 26% Handkerchief fully taped: 76% Handkerchief corners taped: 32% Handkerchief nylon

Journ

al Pre-

Proof

Co., Model HRI 8137) Disposable face mask (3M Corp., Model #8710)

Model 3050 vibrating orifice generator. Volume of air inhaled per minute was 37 L, respiratory rate of 23 cycles per minute. Mannequin was U.S. army design used for testing military respirators, facial features are based upon average male. Masks were fastened on the head by taping using different methods: 1) completely seal all edges with plastic tape over nose, around cheeks and under chin 2) loosely hold material with four pieces of tape on corners of mask 3) using nylon hosiery to hold mask in place by placing nylon hosiery over the head entirely.

Filtration efficiency was calculated using formula FE = 1-TIL.

hosiery: 72% Toweling washcloth fully taped (1 layer): 61% Toweling washcloth (1 layer) corners taped: 40% Toweling washcloth (2 layers) corners taped: 70%

Dato 20063 Cloth mask

Hanes Heavyweight 100% preshrunk cotton T-shirt (made in Honduras) was boiled for 10 minutes and air-dried to maximize shrinkage and sterilize material in manner available in developing countries. Scissor, marker and ruler were used to cut out 1 outer layer (37x72 cm; used to fasten mask to head with 3 straps) and 8 inner layers (≤18 cm2, layered as follows: 2 cross grain, 2 straight

Three authors of this paper made their own cloth masks to fit their faces. A quantitative fit test was performed using the Portacount Plus Respirator Fit Tester with N95 Companion, which measured the concentration of aerosol outside and inside the prototype mask. Ambient

Aerosol concentration (ambient dust and other aerosols present in air) outside and inside the prototype mask were measured. A fit factor was calculated,

Inward Cloth mask filtration efficiency, 98.5%, 92.3%, 94.1% N95 filtration efficiency, 99%

Journ

al Pre-

Proof

grain, 2 cross grain, 2 straight grain), weight NA, weave NA, thread count NA

dust and other aerosols present in the air were measured. Workplace activities were simulated (series of exercises, each 1 minute in duration).

Davies 20134-6

Cloth mask and medical mask cut in circular shape and used as a filter

Different materials were used to make a “homemade” mask. Materials included 100% cotton shirt, scarf, tea towel, pillowcase, vacuum cleaner bag, cotton mix, linen, and silk. Weight NA, weave NA, thread count NA, 1 or 2 layers Medical mask (Mölnlycke Health Care Barrier face mask 4239, EN14683 class I)

This paper consisted of three experiments: 1. Measuring filtration efficiency as a measure of inward protection, done by cutting masks made of different household materials in circular pieces and then placing in airtight cases as a filter. A Henderson apparatus allows closed-circuit generation of microbial aerosols from a Collison nebulizer at a controlled relative humidity and was used to deliver aerosol across each material at 30L/min. Aerosol particle size and distribution NA. 2. Measuring fit factor of homemade mask made of 100% cotton t-shirt fabric, by comparing concentration of microscopic particles outside and inside the respirator using the TSI

For inward experiment, there was an empty filter (used as a reference point) and then the chosen filter (used as the experimental group) to determine concentration of the different microbial aerosols in and out to determine filtration efficiency. B. atrophaeus and Bacteriophage MS2 were used, and can be compared in size to influenza virus. For outward protection, many variables were measured including fit,

Both Filtration Efficiency (first experiment) given in percentage, number in parentheses is for 2 layers. First numbers using B atrophaeus, second numbers using Bacteriophage MS2 100% cotton T-shirt: 69.42% (70.66%), 50.85% Scarf: 62.30%, 48.87% Tea towel: 83.24% (96.71%), 72.46% Pillowcase: 61.28% (62.38%), 57.13% Antimicrobial pillow case: 65.62%, 68.90% Medical mask: 96.35%, 89.52% Vacuum cleaner bag: 94.35%, 85.95%

Journ

al Pre-

Proof

PortaCount Plus Respirator Fit Tester and N95 Companion module model 8095. During the fit test volunteers performed following consecutive exercises, each lasting 96 seconds: normal breathing, deep breathing, head moving side to side, head moving up and down, talking aloud, bending at waist as if touching toes and normal breathing. 3. A mobile sampling chamber, or cough box, was used for the purpose of sampling aerosols and droplets from healthy volunteers outward protection. Four settling plates with Tryptose soya agar were used as the culture medium placed inside this cough box, and the number of colony forming units were counted. Volunteers coughed twice into the box, wearing homemade mask, surgical mask and no mask

median and interquartile range, and colony forming units from “droplets”

Cotton mix: 74.60%, 70.24% Linen: 60.00%, 61.67% Silk: 58.00%, 54.32% Filtration Efficiency (second experiment), given as protection factors and converted in to filtration efficiency Normal breathing Homemade mask, 50% Medical mask, 83% Heavy breathing Homemade mask, 50% Medical mask, 86% Head moving side to side Homemade mask, 50% Medical mask, 80% Head moving up and down Homemade mask, 50% Medical mask, 80% Bending over Homemade mask, 0% Medical mask 67%

Journ

al Pre-

Proof

Talking Homemade mask, 50% Medical mask, 83% Normal breathing again Homemade mask, 50% Medical mask 80% All data Homemade mask, 50% Medical mask, 80% Filtration Efficiency (Third experiment), given as number of colonies and converted in to filtration efficiency, 3 different sampling methods, air, settle plates and total Air Homemade mask, 83.3% Medical mask, 83.3% Settle plates Homemade mask, 0% Medical mask, 100% Total sampling methods Homemade mask, 50% Medical mask, 100% Filtration efficiency

Journ

al Pre-

Proof

(third experiment), given as number of colonies and converted in to filtration efficiency, different particle diameters >7 μm Homemade mask, 66% Medical mask, 44% 4.7-7 μm Homemade mask, 61.1% Medical mask, 61.1% 3.3-4.7 μm Homemade mask, 20% Medical mask, 20% 2.1-3.3 μm Homemade mask, 85.1% Medical mask, 89.4% 1.1-2.1 μm Homemade mask, 84% Medical mask, 94% 0.65-1.1 μm Homemade mask, 71.4% Medical mask, 85.7% All particle sizes Homemade mask,

Journ

al Pre-

Proof

78.5% Medical mask, 85%

Doust 19187

Cloth masks

Coarse gauze, medium gauze, buttercloth, hemmed on the edges with 4 plaits on each lateral edge, equipped with tapes on 4 corners to tie behind the head. 6x8 inches, weight NA, weave NA, thread count NA, number of layers varying from 2 to 10 layers.

This paper performed 4 experiments with no masks, coarse gauze, medium gauze and buttercloth comparing the colony count on agar plates in different breathing conditions by volunteers sitting at a table with exposed plates arranged at distances of 1, 2, 3, 4, 5, 6, 7, 8, 9 and 10 feet. Volunteers were instructed to talk in ordinary conversational tone for five minutes, talk in a loud tone for 5 minutes, or cough as much as possible for 5 minutes with the no mask and the different mask conditions

Volunteers were contaminated with B. prodigiosus

Outward Coarse gauze Speaking in a loud tone for 5 minutes 1ft, 2 layers, 68.2% 1ft, 3 layers, 92.4% 1ft, 4 layers, 90.7% 1ft, 5 layers, 99.2% 1ft, 6 layers, 97.9% 1ft, 7 layers, 98.7% 1ft, 8 layers, 99.2% 1ft, 9 layers, 100% 1ft, 10 layers, 100% 2ft, 2 layers, 50% 2ft, 3 layers, 50% 2ft, 4 layers, 50% 2ft, 5 layers, 100% 2ft, 6 layers, 100% 2ft, 7 layers, 100% 2ft, 8 layers, 100% 2ft, 9 layers, 100% 2ft, 10 layers, 100% 3ft, 4ft, for all layers 100% expect 4ft, 3 layers, which is 0% 5ft and 6ft, the control is 0, unable to calculate Coarse gauze Coughing for 5 minutes 1ft, 2 layers, 0% 1ft, 3 layers, 54.5% 1ft, 4 layers, 0% 1ft, 5 layers, 49% 1ft, 6 layers, 76.3%

Journ

al Pre-

Proof

1ft, 7 layers, 75.2% 1ft, 8 layers, 97.1% 1ft, 9 layers, 97.8% 1ft, 10 layers, 96.7% 2ft, 2 layers, 0% 2ft, 3 layers, 54% 2ft, 4 layers, 0% 2ft, 5 layers, 12.3% 2ft, 6 layers, 89% 2ft, 7 layers, 65.5% 2ft, 8 layers, 99.5% 2ft, 9 layers, 97.8% 2ft, 10 layers, 98.9% 3ft, 2 layers, 0% 3ft, 3 layers, 64.1% 3ft, 4 layers, 0% 3ft, 5 layers, 0% 3ft, 6 layers, 88.4% 3ft, 7 layers, 82.6% 3ft, 8 layers, 100% 3ft, 9 layers, 97.7% 3ft, 10 layers, 100% 4ft, 2 layers, 0% 4ft, 3 layers, 93.4% 4ft, 4 layers, 54.1% 4ft, 5 layers, 44.9% 4ft, 6 layers, 94.8% 4ft, 7 layers, 100% 4ft, 8 layers, 100% 4ft, 9 layers, 100% 4ft, 10 layers, 100% 5ft, 2 layers, 0% 5ft, 3 layers, 89.2% 5ft, 4 layers, 78.4% 5ft, 5 layers, 59.5% 5ft, 6 layers, 100%

Journ

al Pre-

Proof

5ft, 7 layers, 100% 5ft, 8 layers, 100% 5ft, 9 layers, 100% 5ft, 10 layers, 100% 6ft, 2 layers, 0% 6ft, 3 layers, 73.3% 6ft, 4 layers, 60% 6ft, 5 layers, 86.7% 6ft, 6 layers, 86.7% 6ft, 7 layers, 100% 6ft, 8 layers, 100% 6ft, 9 layers, 100% 6ft, 10 layers, 100% Medium gauze Speaking in a loud tone for 5 minutes 1ft, 2 layers, 6.8% 1ft, 3 layers, 99.6% 1ft, 4 layers, 100% 1ft, 5 layers, 100% 1ft, 6 layers, 99.6% 1ft, 7 layers, 99.2% 1ft, 8 layers, 100% 1ft, 9 layers, 100% 1ft, 10 layers, 100% 2ft, all layers, 100% Except 2ft, 2 layers, 0% 3ft, all layers, 100% Except 3ft, 3 layers, 42.9% 4ft, all layers, 100% Except 4ft, 4 layers, 0% 5ft and 6ft, the control is 0, unable to calculate

Journ

al Pre-

Proof

Medium gauze Coughing for 5 minutes 1ft, 2 layers, 94.5% 1ft, 3 layers, 85.8% 1ft, 4 layers, 83.6% 1ft, 5 layers, 99.3% 1ft, 6 layers, 100% 1ft, 7 layers, 97.1% 1ft, 8 layers, 98.2% 1ft, 9 layers, 98.9% 1ft, 10 layers, 99.6% 2ft, 2 layers, 96.7% 2ft, 3 layers, 86.3% 2ft, 4 layers, 88.5% 2ft, 5 layers, 99.5% 2ft, 6 layers, 99.5% 2ft, 7 layers, 94.5% 2ft, 8 layers, 100% 2ft, 9 layers, 96.7% 2ft, 10 layers, 100% 3ft, 2 layers, 98.7% 3ft, 3 layers, 86.1% 3ft, 4 layers, 90.7% 3ft, 5 layers, 100% 3ft, 6 layers, 100% 3ft, 7 layers, 91.9% 3ft, 8 layers, 98.8% 3ft, 9 layers, 98.8% 3ft, 10 layers, 100% 4ft, 2 layers, 98.7% 4ft, 3 layers, 97.4% 4ft, 4 layers, 98.7% 4ft, 5 layers, 100% 4ft, 6 layers, 100% 4ft, 7 layers, 88.2% 4ft, 8 layers, 98.7%

Journ

al Pre-

Proof

4ft, 9 layers, 98.7% 4ft, 10 layers, 98.7% 5ft, 2 layers, 97.3% 5ft, 3 layers, 97.3% 5ft, 4 layers, 100% 5ft, 5 layers, 100% 5ft, 6 layers, 100% 5ft, 7 layers, 81.2% 5ft, 8 layers, 97.3% 5ft, 9 layers, 89.2% 5ft, 10 layers, 100% 6ft, 2 layers, 100% 6ft, 3 layers, 100% 6ft, 4 layers, 100% 6ft, 5 layers, 100% 6ft, 6 layers, 100% 6ft, 7 layers, 20% 6ft, 8 layers, 100% 6ft, 9 layers, 86.7% 6ft, 10 layers, 100% Buttercloth Speaking in loud tone for 5 minutes 1ft, 2ft, 3ft, 4ft, all layers 100% 5ft, 6ft, the control is 0, unable to calculate Buttercloth Coughing for 5 minutes 1ft, 2ft, 3ft, 4ft, 5ft, 6ft, all layers 100% Except 2ft, 2 layers, 98.9%

Furahashi Cloth Cloth from 4 different cloth masks A test apparatus (US Total number of Non- FE is given as % (SD)

Journ

al Pre-

Proof

19788 masks, surgical masks

and mask material from 2 commercial made masks were tested A. Bleach cotton fabric, weight NA, weave NA, thread count 46/inch by 50/inch B. Calico, weight NA, weave NA, thread count 80/inch by 80/inch C. Twill weave cotton, weight NA, weave NA, thread count NA D. Bleached cotton fabric, weight NA, weave NA, thread count 40/inch by 46/inch E. Fine glass fiber with non-woven fabric (commercial mask 1; Hopes) F. Fine glass fiber with non-woven fabric (commercial mask 2; Medispo)

military standard) was used to determine bacterial filtration efficiency. Bacterial agar plates were used within the apparatus where they compared the number of colony counts on control plates versus the plates with the filter interposed. Flow rate 8L/min. Aerosol size not specified.

bacterial colony counts of Staphylococcus aureus (used with all masks) and Serratia marcescens (only used with commercial mask made by Hopes).

directional A. Bleached cotton fabric: 68.8% (SD 3.65) B. Calico: 73.2% (SD 3.55) C. Twill weave cotton: 93.6% (SD 1.16) D. Bleached cotton fabric: 43.1% (SD 8.93) E. Fine glass fiber with non-woven fabric (Hopes): 98.1% (SD 1.02) for Staph. aureus and 96.4% (SD 0.65) for Serr. marcescens F. Fine glass fiber with non-woven fabric (Medispo): 99.4% (SD 0.45)

Greene 19629

Cloth mask

2 layers of thin muslin, inner lining of 4-oz outing flannel, weight NA, weave NA, thread count NA

A Sampling chamber was used made of a plywood box (5 ft by 16 inch by 16 inch) mounted vertically on an angle iron frame. This allowed the participant to insert their head only into this isolated chamber. Participants were instructed to say “sing and chew” at 10 second intervals. Thereafter, the air was sampled on blood agars using an Anderson sampler

The number of airborne microorganisms was sampled on the sedimentation plates or by the sample chamber for masked and unmasked individuals

Outward Filtration efficiency as %, taken from sedimentation plates, talking Subject 1, 99.9% Subject 2, 99.6% Subject 3, 99.9% Subject 4, 99.3% Airborne microorganisms, taken from sampling chamber particles less than 4 μm Subject 1, 95.7%

Journ

al Pre-

Proof

at different ranges Subject 2, 87.6% Subject 3, 99.0% Subject 4, 98.6% Average of all subjects, 96.7% All particles Subject 1, 99.5% Subject 2, 99.5% Subject 3, 99.8% Subject 4, 99.7% Average of all subjects, 99.6% Airborne particles, sampling chamber, talking >8 μm, 99.8% 4-8 μm, 99.8% <4 μm, 96.7% Total particles, 99.6% Airborne particles, unconfined space, talking >8 μm, 97.3% 4-8 μm, 96.5% <4 μm, 95.4%

Guyton 195910

Cloth masks

8 different items were tested. 1. Men’s cotton handkerchief, weight NA, weave NA, thread count 80/inch by 10/inch. 2. Toilet paper Waldorf Scottissue 3. Towel, bath Cotton terry weave Federal Spec, Bi, DDD-T-551 B Type 2, Class D, weight NA, weave

Four subjects were used for each material. A mouth collector was placed in the mouth of participants and the mask material was placed on top of the collector, subjects held the mask in place. Subjects

Bacillus subtilis var. niger was sampled as the “exposure aerosol”

Inward Mean filtration efficiency, number of layers (95% confidence interval). - Men’s cotton handkerchief 16 layers: 94.2 (92.6-95.5) - Toilet paper 3 layers:

Journ

al Pre-

Proof

NA, thread count NA 4. Bed sheet muslin, Pepperell Red Label (fine muslin), weight NA, weave NA, thread count 131 per square inch 5. Shirt cotton Arrow Dart, weight NA, weave NA, thread count NA 6. Women’s handkerchief, cotton lawn fabric, weight NA, weave NA, thread count 76/inch by 72/inch 7. Dress material, cotton, Rondo Percale, weight NA, weave NA, thread count NA 8. Slip, rayon, Barbizon Jaunty Fit, acetate and rayon, weight NA, weave NA, thread count NA

were put into an exposure chamber which released the “contaminants” in the air and anything that was not filtered by the mask was collected by the mouth collector. This allowed for the measurement of filtration efficiency of the different masks. Aerosols had particle size of 1-5 microns.

91.4 (89.8-92.8) - Men’s cotton handkerchief 8 layers: 88.9 (85.5-91.6) - Men’s cotton handkerchief, crumpled: 88.1 (85.1-90.5) - Towel bath, 2 layers: 85.1 (83.3-86.8) - Towel bath, 1 layer: 73.9 (70.7-76.8) - Bed Sheet Muslin, 1 layer: 72.0 (68.8-74.9) - Towel bath, 1 layer wet: 70.2 (68.0-72.3) - Shirt cotton, 1 layer wet: 65.9 (57.9-72.3) - Shirt cotton 2 layers: 65.5 (60.8-69.6) - Women’s cotton handkerchief, 4 layers wet: 63.0 (57.3-67.9) - Men’s cotton handkerchief, 1 layer wet: 62.6 (57.0-67.5) - Cotton Dress Material, 1 layer wet: 56.3 (49.6-62.0) - Women’s cotton handkerchief, 4 layers: 55.5 (52.2-58.7) - Rayon Slip, 1 layer: 50.0 (46.2-53.6) - Cotton Dress Material, 1 layer: 47.6

Journ

al Pre-

Proof

(41.4-53.2) - Shirt cotton, 1 layer: 34.6 (29.0-39.9) - Men’s cotton handkerchief, 1 layer: 27.5 (22.0-32.5)

Haller 191811

Cloth masks

Four different cloth masks were tested, gauze used was Bauer and Black’s or equivalent of their specimens called 1. B and B, weight NA, weave NA, thread count 32/inch by 26/inch 2. L and L, weight NA, weave NA, thread count 28/inch by 24/inch 3. Lakeside, weight NA, weave NA, thread count 24/inch by 20/inch 4. Dearborn, weight NA, weave NA, thread count 20/inch by 14/inch Masks tested varied from 1-8 layers

Two different experiment were performed to demonstrate inward and outward protection. The first experiment had one infected subject wear masks with different layers and then cough at a constant pace and pressure toward a petri dish placed horizontally 12-14 inches away. The second experiment required the same infected subject to caught at a petri dish covered with different layers of mask to demonstrate inward protection.

Pneumococci (Type IV)

Both First experiment, mask over face of person coughing 1 layer B and B, 59.3% L and L, 60.0% Lakeside, 56.3% Dearborn, 59.5% 2 layers B and B, 86.2% L and L, 85.3% Lakeside, 70.0% Dearborn, 61.3% 3 layers B and B, 91.5% L and L, 83.7% Lakeside, 85.0% Dearborn, 76.5% 4 layers B and B, 99.2% L and L, 90.0% Lakeside, 91.5% Dearborn, 84.5% 5 layers B and B, 100% L and L, 98.2%

Journ

al Pre-

Proof

Lakeside, 93.2% Dearborn, 81.0% 6 layers B and B, NA L and L, 100% Lakeside, 96.3% Dearborn, 88.0% 7 layers B and B, NA L and L, NA Lakeside, 100% Dearborn, 96.7% 8 layers B and B, NA L and L, NA Lakeside, NA Dearborn, 100% Second experiment, mask over Petri dish Lakeside, 5 layers, 100% No other data given

Jang 201512

Cloth from cloth masks, medical mask

Cloth mask A, shape: plate type, 50% nylon, 40% polypropylene, 10% polyurethane, thickness 1.22 mm, weave NA, thread count NA, 1, 2, and 4 layers Cloth mask B, shape: plate type, 84% nylon, 12% polyester, 4% spandex, thickness 0.62mm,

Polydisperse NaCl aerosols were generated by an atomizer (Atomizer 9302, TSI, USA) and introduced into an aerosol chamber and then passed through the fabric that was being tested. The concentration of particles was measured

Polydisperse NaCl aerosols of the size range 0.3~10 µm

Non-directional

Cloth mask A 0.3-0.5 µm 1 layer: 29% 2 layers: 59% 4 layers: 75% 2-5 µm 1 layer: 60% 2 layers: 70%

Journ

al Pre-

Proof

weave NA, thread count NA, 1, 2, and 4 layers Cloth mask C, shape: plate type, 100% polyester (cool comfort fabrics), thickness 0.29 mm, weave NA, thread count NA, 1, 2, and 4 layers Cloth mask D, shape: plate type, 100% polyester (microfiber), thickness 0.30 mm, weave NA, thread count NA, 1, 2 and 4 layers Cloth mask E, shape: cup type, 100% polyester (microfiber), 2.77 mm, weave NA, thread count NA, 1 layer R, Class 1 disposable respirator, shape: cup type, non-woven fabrics, thickness 1.81 mm, weave, thread count and layers all not relevant (N95 type mask)

by an optical particle counter (OPC) in five channels of the size range 0.3~10 µm. The mask fabric was either tested in 1, 2, or 4 layers. Flow rates of 30 LPM, 95 LPM and 85± LPM were mentioned but due to a language barrier it is not clear which one was used for which test.

4 layers: 94% Cloth mask B 0.3-0.5 µm 1 layer: 28% 2 layers: 32% 4 layers, 67% 2-5 µm 1 layer: 63% 2 layers: 71% 4 layers: 77% Cloth mask C 0.3-0.5 µm 1 layer: 18% 2 layers: 50% 4 layers: 55% 2-5 µm 1 layer: 45% 2 layers: 78% 4 layers: 81% Cloth mask D 0.3-0.5 µm 1 layer: 9% 2 layer: 45% 4 layers: 62% 2-5 µm 1 layer: 45% 2 layers: 59% 4 layers: 99% Cloth mask E

Journ

al Pre-

Proof

0.3-0.5 µm: 27% 2-5 µm: 80% Class 1 disposable respirator, R 0.3-0.5 µm: 91% 2-5 µm: 100%

Jung 201413

Cloth masks, medical masks

5 types of cotton mask, all flat, weights NA, weaves NA, thread counts NA, layers NA 3 types of handkerchief 1 cotton, 1 gauze and 1 towel Shape NA, weights NA, weaves NA, thread counts NA, 1-4 layers 7 types of medical mask 4 surgical masks: 1 cotton and flat, 1 nonwoven and flat and 2 nonwoven and cup shaped. All weights, weaves, thread counts and layers NA 3 dental masks: all 3 nonwoven and flat. All weights, weaves, thread counts and layers NA

NaCl particles were tested by two TSI 8130 Automatic Filter Testers (AFTs). The AFT was designed in compliance with the KFDA protocol and the NIOSH regulation 42 CFR part 84 protocols. Before testing the fabric the tested aerosols were examined to meet size criteria of the NIOSH and KFDA with a scanning mobility particle sizer (SMPS, TSI-3910; TSI Inc., Shoreview, MN, USA). The fabric samples were attached to plates with hot-metal adhesive. Using the TSI 8130 automated filter tester the plate was placed into the lower chuck of tester with a space ring (20 cm in diameter and 10 cm in height) fitted with a gasket placed on top. Then a second plate was placed on top of the spacer ring, the pressure, when the

1% NaCl concentration and 2% NaCl solution

Medical masks were tested in both directions. Non-directional for cloth masks.

NIOSH protocol Medical masks Surgical inward: 40.9% SD 36.7 Surgical outward: 42.3% SD 33.7 Dental inward: 70.9% SD 12 Dental outward: 68.8% SD 14.3 General masks Non-woven: 54.75 SD 9.414 Cotton: 22.6% SD 26.8 Handkerchief Cotton 1 layer: 1.1% SD 0.666 2 layers: 2% SD 0.702 3 layers: 3.1% SD 0.379 4 layers: 3.8% SD 0.346 Gauze 1 layer: 0.7% SD 0.300 2 layers: 1.4% SD 0.493 3 layers: 2% SD 0.400 4 layers: 3.6% SD 0.351

Journ

al Pre-

Proof

AFT was closed, of the top chuck on the upper plate compressed the plates and spacer ring together, forming an airtight seal. The TSI uses two aerosol photometers to measure particle penetration, with one placed before and one placed after the filter. (NIOSH, 1996; TSI, 2006) The penetration was recorded at 1-min intervals. Six samples of each model were tested: three for the KFDA method and three for the NIOSH method. For the KFDA method all penetration tests were done at the flow rate of 95 L/min and a NaCl concentration of 1%. For the NIOSH method the tests were done at the flow rate of 85 L/min and a 2% NaCl solution was used.

Kellogg 192014

Experiment No. I.

Cloth Gauze, weight NA, weave NA, thread count 40 by 17, 6 layers

An unknown number of replicates coughing on petri dishes located 4ft in front of them.

Bacillus prodigiosus sprayed into the mouths of volunteers

Outward 82.20%

Kellogg 192014 Experiment No. II.

Cloth Gauze, weight NA, weave NA, thread count 20 by 17, 6 layers

An atomizer was placed 1, 2, and 3 ft away from petri dishes in jars.

Bacillus prodigiosus, saline

Inward At 1ft 73.9%, At 2ft 35.5%

Kellogg Cloth Gauze, weight NA, Weave NA, An atomizer was placed 3, Bacillus Inward 3 ft, 3 layers 12%

Journ

al Pre-

Proof

1920 Experiment No. III.

thread count 20 by 17, 6, 5, 4, and 3 layers

4, 5, 6, 7, and 8 ft away from petri dishes

prodigiosus, paraffin oil

3 ft, 4 layers 53.4% 3ft, 5 layers 89.1% 3ft, 6 layers 87% 4ft, 3 layers 26.4% 4ft, 4 layers 70.2% 4ft, 5 layers 87.5% 4ft, 6 layers 90% 5ft, 3 layers 37.1% 5ft, 4 layers 75% 5ft, 5 layers 90.8% 5ft, 6 layers 88% 6ft, 3 layers 23% 6ft, 4 layers 71.6% 6ft, 5 layers 95.1% 6ft, 6 layers 87.6% 7ft, 3 layers 26.7% 7ft, 4 layers 74.9% 7ft, 5 layers 91.6% 7ft, 6 layers 87.4% 8ft, 3 layers 55.8% 8ft, 4 layers 81.5% 8ft, 5 layers 94.4% 8ft, 6 layers 87.7%

Kellogg 192014 Experiment No. V.

Cloth Gauze, weight NA, weave NA, thread count reported as 24 by 18 but also as 24 by 28, possible error. 2, 3, 4, 5, 6, 7, 8, and 9 layers

An atomizer was placed 5ft away from petri dishes. Petri dishes were in jars with a whole in the lid and suction applied to the bottom of the jar to create air flow. The jars were covered in no layers or 2, 3, 4, 5, 6, 7, 8, or 9 layers of gauze. Atomizer was turned on and off and then was left to settle for 5min.

Bacillus prodigiosus

Inward 2 layers 25.9%, 3 layers 48.1%, 4 layers 78%, 5 layers 72.7%, 6 layers 85%, 7 layers 81.6%, 8 layers 97.4% 9 layers 98.3%

Journ

al Pre-

Proof

Kellogg 192014 Experiment No. VI.

Cloth Gauze, weight NA, weave NA, thread count reported as 24 by 18 but also as 24 by 28, possible error. 2, 3, 4, 5, 6, 7, and 8 layers

An atomizer was placed 5ft away from petri dishes. Petri dishes were in jars with a whole in the lid and suction applied to the bottom of the jar to create air flow. The jars were covered in no layers or 2, 3, 4, 5, 6, 7, 8 or 9 layers of gauze. Atomizer was turned on and off and then was left to settle for 3min. Identical to experiemnt No. V. expect for exposure time.


Inward 2 layers 12.5% 3 layers 0% 4 layers 15.9% 5 layers 17.4% 6 layers 28.1% 7 layers 55% 8 layers 59.2%

Kellogg 192014 Experiment No. VII.

Cloth Fine and extra fine gauze also called butter cloth by the author, weight NA, weave NA, thread count 42 by 44 threads, 2, 3, 3, 4, 5, 6, 7, 8, and 9 layers

An atomizer was placed 4 and 5.5 ft away from petri dishes in jars. Suction was applied to the jars to create air flow. The jars were either covered with no gauze or 2, 3, 4, 5, 6, 7, 8 or 9 layers. The DeVilbiss No. 15. Atomizer was used. The atomizer was turned on and off and then left to settle for 5min.


Inward 4 ft, 2 layers 10.1% 4 ft, 3 layers 0% 4 ft, 4 layers 31.4% 4 ft, 5 layers 68.9% 4 ft, 6 layers 96.7% 4 ft, 7 layers 98.9% 4 ft, 8 layers 98.6% 4 ft, 9 layers 97.5% 5.5 ft, 2 layers 0% 5.5 ft, 3 layers 0% 5.5 ft, 4 layers 11.6% 5.5 ft, 5 layers 37.3% 5.5 ft, 6 layers 94.8% 5.5 ft, 7 layers 98.3% 5.5 ft, 8 layers 99% 5.5 ft, 9 layers 97.1%

Kellogg 192014 Experiment No. VIII.

Cloth Fine and extra fine gauze also called butter cloth by the author, weight NA, weave NA, thread count 42 by 44 threads, 2, 3, 3, 4, 5, 6, 7, 8, and 9 layers

An atomizer was placed 4 and 5.5 ft away from petri dishes in jars. Suction was applied to the jars to create air flow. The jars

B. prodigiosus Inward 4 ft, 2 layers 76.3% 4 ft, 3 layers 86.3% 4 ft, 4 layers 90.5 4 ft, 5 layers 88.2% 4 ft, 6 layers 100%

Journ

al Pre-

Proof

were either covered with no gauze or 2, 3, 4, 5, 6, 7, 8 or 9 layers. The DeVilbiss No. 15. Atomizer was used. The atomizer was turned on and off and then left to settle for 3min. Identical to experiment No. VII. except for exposure time.

4 ft, 7 layers 100% 4 ft, 8 layers 100% 4 ft, 9 layers 99.5% 5.5 ft, 2 layers 84.3% 5.5 ft, 3 layers 93.7% 5.5 ft, 4 layers 85.8% 5.5 ft, 5 layers 89.8% 5.5 ft, 6 layers 100% 5.5 ft, 7 layers 99.2% 5.5 ft, 8 layers 100% 5.5 ft, 9 layers 100%

Kellogg 192014 Experiment No. IX.

Cloth Gauze, weight NA, weave NA, thread count 60 by 72, 1, 2, 3, 4, 5, 6, 7, 8, and 9 layers.

An atomizer was placed 4 and 5.5 ft away from petri dishes in jars. Suction was applied to the jars to create air flow. The jars were either covered with no gauze or 1, 2, 3, 4, 5, 6, 7, 8 or 9 layers. The DeVilbiss No. 15. Atomizer was used. The atomizer was turned on and off and then left to settle for 5 min.

B. prodigiosus Inward 4 ft, 1 layer 0% 4 ft, 2 layers 0% 4 ft, 3 layers 84.7% 4 ft, 4 layers 97% 4 ft, 5 layers 97.9% 4 ft, 6 layers 96% 4 ft, 7 layers 97.6% 4 ft, 8 layers 97.1% 4 ft, 9 layers 98% 5.5 ft, 1 layer 0% 5.5 ft, 2 layers 0% 5.5 ft, 3 layers 26.2% 5.5 ft, 4 layers 93% 5.5 ft, 5 layers 91.9% 5.5 ft, 6 layers 95.9% 5.5 ft, 7 layers 96.5% 5.5 ft, 8 layers 95.3% 5.5 ft, 9 layers 97.7%

Kellogg 192014 Experiment No. X.

Cloth Gauze, weight NA, weave NA, thread count 60 by 73, 5, 6, 7, 8, 9, and 10 layers

An atomizer was placed 5 ft away from a large jar. The jar had the petri dishes inside and two holes in it, the first covered by a wax nose with nostrils and the other one was open with a

B. prodigiosus Inward Nose without nostrils, 5 layers 92.9% Nose without nostrils, 6 layers 89.4% Nose without nostrils, 7 layers 95.4% Nose without nostrils, 8

Journ

al Pre-

Proof

wax nose without nostrils just above the hole. The jar was standing vertically. The atomizer was turned on and off and then left to settle for 3 min. The noses were covered with no gauze or 5, 6, 7, 8, 9 or 10 layers of gauze. The DeVillbiss No. 15. atomizer was used.

layers 97.6% Nose without nostrils, 9 layers 98.4% Nose without nostrils, 10 layers 99.6% Nose with nostrils, 5 layers 90% Nose with nostrils, 6 layers 86.7% Nose with nostrils, 7 layers 91.3% Nose with nostrils, 8 layers 90% Nose with nostrils, 9 layers 94% Nose with nostrils, 10 layers 94.7%

Kellogg 192014 Experiment No. XI.

Cloth Gauze, weight NA, weave NA, thread count 60 by 72, in 5, 6, 7, 8, and 9 layers and 24 by 28 in 6, 7, 8, 9, and 10 layers.

Atomizer placed 5 ft away from petri dishes that are placed in 2 vertical jars with holes on the top. There is suction on both jars to create air flow. The hole on the top of one jar is covered by a wax nose with nostrils and a gauze mask. The mask is piece of cloth over the nose. The second jar's hole is covered with a nose with nostrils but no mask. The atomizer was turned on and off and there was 3 min of settling. This experiment was performed twice once with

B. prodigiosus Inward Set 1 (gauze with thread count 60 by 72) 5 layers 57.7% 6 layers 77.3% 7 layers 77% 8 layers 98.2% 9 layers 100% Set 2 (gauze with thread count 24 by 28) 6 layers 38.8% 7 layers 77.3% 8 layers 56.5% 9 layers 94.7% 10 layers 94%

Journ

al Pre-

Proof

gauze with a thread count of 60 by 72 and another time with gauze with a thread count of 24 by 28. Both times with varying layers of cloth.

Konda 202015, 16

Cloth, surgical mask material, N95 mask material

15 different types of fabric were tested. Cotton quilt, filling: ~0.5cm, 90% cotton, 5% polyester, 5% other fibers, purchased from NA, weight NA, weave woven, thread count 120 TPI, 2 layers Quilters cotton, 100% cotton, purchased from NA, weight NA, weave woven, thread count 80 TPI, number of layers varies Cotton, 100% cotton, purchased from Wamsutta, weight NA, weave woven, thread count 600 TPI, number of layers varies Flannel, 65% cotton, 35% polyester, purchased from Walmart Fabric Center, weight NA, weave woven, thread count 90 TPI, 1 layer Chiffon, 90% polyester, 10% spandex, purchased from Jo-Ann Stores (1636949), weight NA, weave woven, thread count 195 TPI, number of layers varies Natural silk, 100% silk, purchased from NA, weight 9 momme or 39 g/m2 (personal communication Supratik Guha), weave woven,

A polydisperse, nontoxic NaCl aerosol was generated by a particle generator (TSI Particle Generator, model #8026) and introduced into a mixing chamber. Particle sizes were in the range of 10 nm to 10 μm. Here it was mixed with the help of a portable fan and passed through the material (area: ~59 cm2) that was being tested, which was held in place using a clamp for a better seal. The aerosol was sampled before and after passing through the material by two different particle analyzers, a TSI Nanoscan SMPS nanoparticle sizer (Nanoscan, model #3910) and a TSI optical particle sizer (OPS, model #3330) for measurements in the range of 10 to 300 nm and 300 nm to 6 μm, respectively. Cloth was measured using a system

Polydisperse, nontoxic NaCl aerosol

Non-directional

Note: standard deviations are available in the original manuscript. Not extracted here because of the large number of data points. Flow rate: 35 L/min (‘decreased by an order of magnitude,’ once cloth inserted, personal communication, Supratik Guha) ~3.5 L/min 75-100 nm: - N95 (no gap): 90% - N95 (with gap): 32.5% - Surgical mask (no gap): 79% - Surgical mask (with gap): 49% - Cotton quilt: 98% - Quilter’s cotton (80 TPI), 1 layer: 4% - Quilter’s cotton (80 TPI), 2 layers: 32% - Flannel: 55% - Cotton (600 TPI), 1 layer: 75.5% - Cotton (600 TPI), 2

Journ

al Pre-

Proof

thread count 145 TPI, number of layers varies Synthetic silk, 100% polyester, purchased from Jo-Ann Stores (1446277), weight NA, weave woven, thread count 102 TPI, number of layers varies Satin, 97% polyester, 3% spandex, purchased from Jo-Ann Stores (4488359), weight NA, weave NA, thread count 203 TPI, 1 layer Spandex, 52% nylon, 39% polyester, 9% spandex, purchased from Jo-Ann Stores (17026402), weight NA, weave woven, thread count 180 TPI, 1 layer Polyester, 100% woven polyester, purchased from Walmart Fabric Center, weight NA, weave woven, thread count 135, 1 layer Cotton/silk, cotton identical to 600 TPI cotton described above, silk not otherwise specified, order not specified, 1 layer of cotton, 2 layers of silk Cotton/chiffon, cotton identical to 600 TPI cotton described above, chiffon identical to chiffon described above, order not specified, 1 layer of cotton, 2 layers of chiffon Cotton/flannel, cotton identical to 600 TPI cotton described above, flannel identical to flannel described above, order not specified, 1 layer of cotton, 1 layer

that produced initial flow rates of 35 L/min and 90 L/min respectively during unrestricted flow; however, when cloth was inserted, increasing the resistance, the flow rate fell, by an amount that could be an order of magnitude or more than the original flow rate (personal communication, Supratik Guha). Some tests were carried out with two circular holes with a diameter of 0.635 cm in the material, to simulate the effect of gaps on the filtration efficiency. Each sample was tested 7 times.

layers: 85% - Chiffon, 1 layer: 57.5% - Chiffon, 2 layers: 86% - Natural silk, 1 layer: 54% - Natural silk, 2 layers: 65% - Natural silk, 4 layers: 84% - Silk, 1 layer: 53.5% - Silk, 2 layers: 64% - Silk, 4 layers: 83.5% - Hybrid 1 cotton/chiffon: 97% Cotton/chiffon, 2 layers: 98% - Hybrid 2 cotton/silk (no gap): 96% - Hybrid 2 cotton/silk (with gap): 34% - Hybrid 2 cotton/silk, 2 layers (no gap): 96% - Hybrid 2 cotton/silk, 2 layers (with gap): 33% - Hybrid 3 cotton/flannel: 95% 2-3 μm: - N95 (no gap): 100% - N95 (with gap): 7% - Surgical mask (no gap): 100% - Surgical mask (with gap): 45% - Cotton quilt: 95%

Journ

al Pre-

Proof

of flannel Surgical mask, not otherwise specified, weight, weave, thread count and layers all not relevant N95, not otherwise specified, weight, weave, thread count and layers all not relevant Natural silk and synthetic silk (polyester) are both described as materials. We have extracted data exactly as reported; where we have written ‘silk’ it was not otherwise specified in the original report.

- Quilter’s cotton (80 TPI), 1 layer: 6% - Quilter’s cotton (80 TPI), 2 layers: 50% - Flannel: 44% - Cotton (600 TPI), 1 layer: 98% - Cotton (600 TPI), 2 layers: 99.5% - Chiffon, 1 layer: 73% - Chiffon, 2 layers: 90% - Natural silk, 1 layer: 55% - Natural silk, 2 layers: 66% - Natural silk, 4 layers: 88.5% - Silk, 1 layer: 55% - Silk, 2 layers: 65% - Silk, 4 layers: 87% - Hybrid 1 cotton/chiffon: 98% - Hybrid 1 cotton/chiffon, 2 layers: 99.5% - Hybrid 2 cotton/silk (no gap): 97% - Hybrid 2 cotton/silk (with gap): 35% - Hybrid 2 cotton/silk, 2 layers (no gap): 98% - Hybrid 2 cotton/silk, 2 layers (with gap): 49% - Hybrid 3 cotton/flannel: 96%

Journ

al Pre-

Proof

Flow rate: 90 L/min (‘decreased by an order of magnitude,’ once cloth inserted, personal communication, Supratik Guha) ~9 L/min 75-100 nm: -N95 (no gap): 94% -N95 (with gap): 58% -Surgical mask (no gap): 59.5% -Surgical mask (with gap): 7.5% -Quilt cotton (80 TPI): 3% -Cotton quilt: 64.5% -Flannel: 13% -Chiffon: 24% -Synthetic silk: 10% -Satin: 13% 2-3 μm: -N95 (no gap): 100% -N95 (with gap): 66% -Surgical mask (no gap): 80% -Surgical mask (with gap): 9.5% -Quilt cotton (80 TPI): 33.5% -Cotton quilt: 80% -Flannel: 45.5% -Chiffon: 53% -Synthetic silk: 23.5% -Satin: 42%

Journ

al Pre-

Proof

Leete 191917

Cloth Gauze, weight NA, weave NA, thread count NA but described as very open weave, 2, 4, 8 and 12 layers Muslin, weight NA, weave NA, thread count 24 per cm, 2, 4, 6, 8, and 10 layers Damp muslin (soaked in water and then wrung out well), weight NA, weave NA, thread count NA, layers NA

Atomizer placed 9 inches away from a vertical Petri dish. Petri dish covered with nothing, gauze or muslin in varying layers. Cloth was fastened over top of petri dish and at a distance of 1.5 cm from the dish. For one of the experiments they set the atomizer to produce a coarser spray, still placed 9 inches away.

Staphylococcus pyogenes aureus

Inward Controls: confluent colonies, too many to count. Filtration efficiency can therefore not be calculated. Number of colonies reported below Gauze, dry 2 layers: 17,500 4 layers: 4,200 8 layers: 2,000 12 layers: 700 Muslin, dry 2 layers: 4,300 4 layers: 1,400 6 layers: 100 8 layers: 40 10 layers: 0 Muslin, dry, 4 layers 12": 88 colonies 18": 14 colonies 24": 7 colonies Muslin, damp, 4 layers 9": 2000 colonies 12": 268 colonies 18": 127 colonies Muslin, dry, coarse spray 4 layers: 356 colonies 6 layers: 230 colonies 8 layers: 50 colonies

Journ

al Pre-

Proof

Lurie 194918

Cloth masks

Gauze, weight NA, weave NA, thread count 40/inch by 44/inch, 3 or 6 layers. Masks were sewn to fit the contour of a rabbit's head, neck and ears. The mask slipped over the rabbit's head like a hood. There were no seams in front of the rabbit's nose or mouth.

Rabbits were placed in an iris diaphragm collar which fitted closely around their necks. Their heads protruded into an exposure chamber in which a nebulizer generated droplet nuclei of tubercle bacilli. In total ten experiments were performed with 6 rabbits in each experiment.

Rabbits were sacrificed and the number of macroscopic tubercles in the lungs were counted

Inward 88% (authors calculation) 95% (our calculation from data provided)

MacIntyre 201519

Cloth masks, medical masks

Medical masks of non-woven material, weight NA, weave NA, thread count NA, 3 layers Cloth masks of cotton, weight NA, weave NA, thread count NA, 2 layers

A TSI 8110 Filter tester was used to test the filtration performance of both of the masks. To test the filtration performance, the filter is challenged by a known concentration of sodium chloride particles of a specified size range and at a defined flow rate. The particle concentration is measured before and after adding the filter material and the relative filtration efficiency is calculated.

Known concentration of sodium chloride particles, particle size not specified. TSI filter tester generates NaCl aerosol with count mean diameter 75 nm and geometric standard deviation 1.75.

Inward Cloth masks 3% Medical masks 56%

Paine 193520

Cloth masks

Silk, surgical gauze, fine dental gauze, all oblong shaped, all 6.5 inches by 4.5 inches. 1, 2, 3, 4, 5, 6, 7, and 8 layers of each material were used. There are tapes at each corner to tie the masks to the face.

An atomizer was attached to a tube which led to 3 holes in a “cast face” representing the mouth and 2 nostrils of a human face. Horizontal agar plates were placed at varying distances below the

M. lysodeiklicus Outward Distance in inches from mouthpiece, number of layers and filtration efficiency are reported, respectively Surgical Gauze 1", 2 layers 0% 1", 4 layers 0%

Journ

al Pre-

Proof

plaster face, which sprayed “droplets” at two different momentums. The different types of mask were tied to the plaster face in the same way as they are worn in practice and tied securely under the chin.

5", 2 layers 42.3% 5", 4 layers 65.4% 9", 2 layers 0% 9", 4 layers 12.5% 14", 2 layers 67.2% 14", 4 layers 91.4% 18", 2 layers 88.8% 18", 4 layers 96.9% 22", 2 layers 64% 22", 4 layers 100% (presumed) 26", 2 layers 22.2% 26", 4 layers 100% (presumed) 30", 2 layers 42.9% 30", 4 layers 100% (presumed) 8 layers 100% at all distances (presumed) Silk, 2 layers, 100% at all distances (presumed) Fine dental gauze, 100% at all distances (presumed)

Quesnel 197521

Cloth mask, medical masks

In total, five masks were tested. 1. Aseptex mask No. 1800 (3M Company, Medical Products Division), a rigid cup shaped mask of bonded polyester and rayon fibers held in place by an elastic band. 2. Cestra mask (Robinsons of

Volunteers wearing one of the masks put their heads inside a vertical chamber. Sliding panels around their heads formed a snug fit around their necks. Subjects then began to say the word chew at 1-second intervals for 5 seconds

Normal human mouth flora, number of colonies were counted with and without masks

Outward >3.3 μm Aseptex 98.9% Cestra 99.3% Surgine 99.7% Filtermask 99.3% Filtron 99.8% 0-3.3 μm Aseptex 80%

Journ

al Pre-

Proof

Chesterfield), four-ply cotton muslin, weave NA, weight NA, thread count NA, 4 layers 3. Surgine mask (Johnson & Johnson Ltd), outer layers made of bonded rayon, inner layers made of glass fibre, 3 pleats, weave not relevant, weight NA, thread count not relevant, 3 layers 4. Filtermask E-Z breathe (Deseret Pharmaceutical Co. Inc.), outer layers made of cellulose, inner layers made of glass fibre, simple folded design, 3 layers 5. Filtron mask (3M Company, Medical Products Division), outer layers made of cellulose, inner layers made of polypropylene fibre, single box-pleat design, 3 layers All masks except for the Aseptex were held in place by pairs of fabric ties. All masks except for the Cestra had metal contour strips across the nose and checks

followed by a 5-second rest, alternating for 4 minutes and saying the word a total of 120 times. After the subject finished saying the words they remained mute for 5 more minutes. Then they removed their heads from the chamber and took off their mask. This whole procedure was then repeated without masks. Samples were collected with blood-agar plates and the Andersen sampler, which was linked to the chamber by rubber tubing.

Cestra 89% Surgine 89.6% Filtermask 72.2% Filtron 88.3% All sizes Aseptex 96.5% Cestra 98.8% Surgine 98.8% Filtermask 95.8% Filtron 98.8%

Rengasamy 201022

Cloth mask, N95 mask

In total, 5 materials (cloth mask, sweatshirt, T-shirt, towel, scarf) were tested, each with 3 models. 1. Cloth mask fabric (presumed multi layered as manufactured)

All 15 fabric materials were tested using a TSI 8130 Automated Filter Tester. The material was cut into 100 cm2 samples and measured at two

Monodisperse NaCl particles. 500 to 1000 nm. We decided a priori to extract data for 1000 nm.

Non-directional

1000 nm particles, results are given for 5.5 and 16.5 cm/s, respectively 1. Cloth mask fabric

Journ

al Pre-

Proof

- Respro bandit mask, no details given - Breathe Health cloth mask, no details given - Breathe Health fleece mask, no details given 2. Sweatshirt fabric (presumed 1 layer) - Norma Kamali Tunic, 85% cotton, 15% polyester - Hanes, 70% cotton, 30% polyester - Faded Glory, 60% cotton, 40% polyester 3. T-shirt fabric (presumed 1 layer) - Dickies, 99% cotton, 1% polyester - Hanes, 100% cotton - Faded Glory, 60% cotton, 40% polyester 4. Towel fabric (presumed 1 layer) Pem America, 100% cotton, Pinzon, 100% cotton Aquis, 100% cotton 5. Scarf fabric (presumed 1 layer) - - Today's Gentleman Pocket Square, 100% cotton, - Walmart, fleece, 100% polyester, - Seed Supply, 100% cotton,

different face velocities, 5.5 and 16.5 cm/s, corresponding to 33 and 99 L/min. The fabric was tested against polydisperse NaCl particles.

- Respro Bandit mask, 22%, 34% - Breath Health Cloth mask, 13%, 44% - Breath Health Fleece mask, 22%, 13% 2. Sweatshirt fabric - Norma Kamali, 8%, 26% - Hanes, 19%, 15% - Faded Glory, 6%, 12% 3. T-shirt fabric - Dickies, 8%, 20% - Hanes, 9%, 12% - Faded Glory, 0%, 15% 4. Towel fabric - Pem America, 23%, 49% - Pinzon, 30%, 58% - Aquis, 33%, 0% 5. Scarf fabric - Today's Gentleman, 0%, 0% - Walmart, 25%, 14% - Seed Supply, 1%, 7% N95, 100%, 100%

Shakya 201723

Cloth masks, medical

Cloth mask 1, purchased from street vendors in Kathmandu, Nepal, has a plastic and latex

Experiment 1 A constant output atomizer (model 3076)

Generated polystyrene latex microsphere

Inward FE given for 8 L/min and 19 L/min, respectively

Journ

al Pre-

Proof

masks, N95 masks

exhalation valve, weave NA, weight NA, thread count NA, layers NA Cloth mask 2, purchased from street vendors in Kathmandu, Nepal, weave NA, weight NA, thread count NA, layers NA Cloth mask 3, purchased from street vendors in Kathmandu, Nepal, weave NA, weight NA, thread count NA, layers NA Surgical mask, purchased from street vendors in Kathmandu, Nepal, has pleats, weave NA, weight NA, thread count NA, layers NA N95 mask 1, 3M model (8200) N95 mask 2, Moldex model (2701), has a plastic and latex exhalation valve

generated Polystyrene latex (PSL) microspheres in different sizes. PSL drops were then added to deionized water (~300 mL) and pure nitrogen was used as the motive gas. The aerosol was passed through a silica-based water vapor denuder to dry the particles, and then into a controlled exposure chamber. In the chamber was a polystyrene mannequin head, fitted with one of the masks, a layer of parafilm was used around the edge of all the masks to minimize leaks. Tubes connected to the mannequins mouth also connected to 2 particle sizing classifiers, an aerodynamic particle sizer (APS; Model: TSI 3321) and a SMPS (SMPS; Model 3080 Electrostatic Classifier and TSI 3775 Condensation Particle Counter). This experiment was performed at 2 different flow rates, 19 L/min, and 8 L/min. For each mask type, 8 consecutive runs were made, the first run was

particles with sizes of 30nm, 100nm, 500nm, 1μm, 2μm

30 nm N95 mask 1, 86%, 81% N95 mask 2, 64%, 77% Cloth mask 1, 87%, 78.5% Cloth mask 2, 88.5%, 15% Cloth mask 3, 54%, 26% Surgical mask, 91%, 62% 100 nm N95 mask 1, 95%, 87.5% N95 mask 2, 86.5%, 84% Cloth mask 1, 94%, 86% Cloth mask 2, 56.5%, 32% Cloth mask 3, 56.5%, 27% Surgical mask, 94%, 69.5% 500 nm N95 mask 1, 93%, 85% N95 mask 2, 85%, 79% Cloth mask 1, 90%, 82% Cloth mask 2, 47%, 56.5% Cloth mask 3, 45%, 31% Surgical mask, 92%, 64.5% 1 μm N95 mask 1, 96%, 92% N95 mask 2, 96%, 68% Cloth mask 1, 94%,

Journ

al Pre-

Proof

discarded, and the remaining seven runs from each experiment were used for the analysis.

88.5% Cloth mask 2, 69%, 54% Cloth mask 3, 85%, 49% Surgical mask, 98.5%, 96% 2 μm N95 mask 1, 97%, 94% N95 mask 2, 95%, 76% Cloth mask 1, 90%, 80% Cloth mask 2, 75%, 74% Cloth mask 3, 82%, 65% Surgical mask, 99%, 97%

Shakya 201723

Same as above

Same as above Experiment 2 Primary diesel particles were generated in the laboratory to simulate urban conditions. Whole exhaust from a single-cylinder diesel generator (Yanmar L100) was injected into a 13m2 laboratory smog chamber made of fluorinated ethylene propylene. Then it was diluted with zero air to bring the concentration level down and passed into a small sealed chamber constructed of stainless steel and aluminum, which contained the mannequin head and mask. The experiment lasted several hours. Commercially

Laboratory generated diesel particles, ranging from 14.6–710.5 nm. Results are only given for particles of 30, 100 and 500 nm size range.

Same as above

30 nm N95 mask 1, 54% N95 mask 2, 51% Cloth mask 1, 87.5% Cloth mask 2, 81.5% Cloth mask 3, 10% Surgical mask, 90% 100 nm N95 mask 1, 71% N95 mask 2, 45% Cloth mask 1, 55% Cloth mask 2, 8% Cloth mask 3, 8.5% Surgical mask, 58% 500 nm N95 mask 1, 82% N95 mask 2, 29% Cloth mask 1, 30% Cloth mask 2, 62% Cloth mask 3, 26.5%

Journ

al Pre-

Proof

available, ultralow sulfur diesel was used for the combustion. A flow rate of 19 L/min was used.

Surgical mask, 92.5% In addition, overall efficiency for all particle sizes were given for cloth masks: Cloth mask 1 34% Cloth mask 2 40% Cloth mask 3 14%

Shooter 195924

Cloth masks

Three types of masks were tested 1. Filtration mask, bucket-shaped, fits fairly snugly over nose and chin, made of gauze, weight NA, weave NA thread count per layer is 46, 4 layers 2. Tail mask, deflexion mask, made of closely woven cambric, 7 1/2 in. by 8 1/2 in, attached to a tail of the same size that hangs down over neck and chest, fit was loose over the cheeks, weight NA, weave NA, thread count NA, 2 layers 3. Paper mask, deflexion mask, single use only, 6 1/2 in. by 5 3/4 in, outer and inner layer surrounding a pad of cellulose wadding, covers nose, mouth and chin, fit was loose over the cheeks, weight NA, weave NA, thread count NA, 3 layers

129 healthy volunteers sat for 15 minutes with their head from the neck up enclosed in a large box. The gap around the neck was sealed by allowing a rubber diaphragm to spring back into place. A plastic canopy, held up by poles and fastened to the edge of the table by battens was put over the table. 20 petri dishes were placed, horizontally, on the floor of the box. Filtered air entered from a pipe and was sucked out at a rate of 1 cu. ft. Volunteers were then asked either to remain silent or to talk and to make an attempt at quiet but continuous conversation. The interior of the canopy and the top of the table were disinfected 30 min before each test.

Human mouth flora

Outward Bacteria from plates Silent Area 1 (immediately in front of volunteer) Filtration mask, 0% Tail mask, 31.3% Paper mask, 37.5% Area 2 (further away but still in front of volunteer) Filtration mask, 0% Tail mask, 14.3% Paper mask, 0% Area 3 (directly behind volunteer) Filtration mask, 0% Tail mask, 6.3% Paper mask, 40% Area 4 (on both sides of volunteer) Control data not given Talking Area 1 Filtration mask, 64% Tail mask, 64% Paper mask, 66%

Journ

al Pre-

Proof

Area 2 Filtration mask, 0% Tail mask, 15.4% Paper mask, 0% Area 3 Filtration mask, 0% Tail mask, 0% Paper mask, 0% Area 4 Filtration mask, 50% Tail mask, 25% Paper mask, 40% Bacteria isolated from air Silent Filtration mask, 0% Tail mask, 0% Paper mask, 0% Talking Filtration mask, 4.2% Tail mask, 0% Paper mask, 16.7%

Van der Sande 200825

Cloth mask, surgical mask, FFP2 mask

Cloth mask, homemade, made of TD Cerise Multi teacloths, Blokker, weave NA, weight NA, thread count NA, layers NA Filtering Facepiece against Particles (FFP)-2 mask 18727V (3M), European equivalent to N95 Surgical mask, 1818 Tie-On, 3M

Healthy volunteers, 3 different experiments to assess 1) short term protection for different types of masks worn during 10-15 minutes by the same volunteer following a standardized protocol. 2) long-term protection of a specific mask worn continuously by a volunteer for 3 hours

For all experiments, candles were used in the room to increase the ambient particle count. Particles had a size of 0.02–1μm

Inward (experiment 1 and 2) Outward (experiment 3)

Experiment 1 - Tea cloth: no activity 60.0%, nodding 54.5%, shaking 54.5%, reading 68.8%, walking 58.3% - Medical mask: no activity 75.6%, nodding 78.7%, shaking 80.4%, reading 81.1%, walking 76.2% - FFP2: no activity 99.1%, nodding 98.7%, shaking 98.9%, reading

Journ

al Pre-

Proof

during regular activities 3) effectiveness of different types of masks in preventing outgoing transmission by a simulated infectious subject Experiment 1: 28 adult and 11 children (5-11 years) volunteers each wearing one of the types of mask were asked to perform 5 tasks in a fixed sequence, 1.5 minute of duration each: sit still and not do anything, nod head ("yes"), shake head ("no"), read aloud from a standard text and stationary walk. The concentration of particles on both sides of the mask were measured with a receptor fixed on both sides of the mask. The receptor was connected to a portable counter of all free floating particles via an electrostatic particle classifier and counter, the Portacount. Experiment 2: Adult volunteers were divided into 3 groups. Each group wore a single type of mask

98.5%, walking 99.0% Experiment 2 - Tea cloth: no activity 68.8%, nodding 63.0%, shaking 65.5%, reading 76.7%, walking 65.5% - Medical mask: no activity 77.3%, nodding 77.8%, shaking 75.6%, reading 83.1%, walking 74.4% - FFP2: no activity 98.1%, nodding 97.9%, shaking 97.6%, reading 98.9%, walking 97.7% Experiment 3 (outward protection), two measurements for each mask type - 30 L/min Tea cloth, 17%, 17% Surgical mask, 47.4%, 65.5% FFP2, 50%, 64.3% - 50 L/min Tea cloth, 17%, 17% Surgical mask, 64.3%, 47.4% FFP2, 50%, 68.3% - 80 L/min Tea cloth, 20%, 20% Surgical mask, 52.4%, 44.4% FFP2, 68.3%, 52.4%

Journ

al Pre-

Proof

for 3 hours. Participants were asked to carry on with their normal activities and after 3 hours they repeated the first experiment, performing the 5 tasks, no activity, nodding, shaking, reading aloud and walking. They did each of these for 1.5 minutes. Concentration of particles was measured similar to experiment 1. Experiment 3: The 3 different type of masks were fitted to an artificial test head, which was connected to PC-driven respirator (Bacou LAMA AMP, Modelref 1520307). Breathing frequency was varied to mimic different respiratory rates, this resulted in a breathing flow of 30, 50 and 80 liters per minute. Concentrations of particles were measured by a TSI Portacount Respirator Fit tester, model 8020.

Weaver 191826

Cloth mask

Gauze mask, weight NA, weave NA thread count NA, layers 2. Shaped to fit closely over the face from chin up to well over the nose, held in place by two tapes

Before-after study. Over the course of 2 years and 7 months the number of nurses carrying diphtheria bacilli were counted. After

Diphtheria bacilli by throat culture, cases of scarlet fever

Inward Filtration efficiency was calculated from percentage of nurses who carried Diphtheria, and the percentage

Journ

al Pre-

Proof

tied behind the head

introduction of masks for nurses that covered nose and mouth, the number of carriers were counted again. The number of nurses with scarlet fever was also counted before (3 years 3 months) and after (1 year and 6 months) introduction of face masks.

who acquired clinical scarlet fever in the no mask and with mask periods. Diphtheria no mask: 10/43 (23.25%) Diphtheria with mask: 6/73 (8.2%) Filtration efficiency 64.7% Scarlet fever no mask: 9/112 (8.0%) Diphtheria with mask: 0/73 (0%) Filtration efficiency 100%

Weaver 191927

See supplementary material figures 2-4

Zhao 202028

Cloth, medical mask material

Polypropylene 1, particulate FFR, meltblown, nonwoven, weight: 25 g/m2, thread count and layers not relevant Polypropylene 2, medical face mask, meltblown, nonwoven, weight: 26 g/m2, thread count and layers not relevant Polypropylene 3, medical face mask, meltblown, nonwoven, weight: 20 g/m2, thread count and layers not relevant All meltblown fabric came from

Tests were conducted with an Automated Filter Tester 8130A (TSI, Inc.) using a 0.26 µm, mass mean diameter (0.075 ± 0.02 µm count median diameter) of sodium chloride (NaCl). The test size of the filter tester was 100 cm2, with a circular gasket outer diameter of approximately 13 cm. All samples were cut to a size greater than a 13 cm × 13 cm square. A flow rate of 32 L/min was chosen because it is similar

NaCl particles Non-directional

Polypropylene 1: 95.9% ± 2.0 Polypropylene 2: 33.1% ± 1.0 Polypropylene 3: 18.8% ± 0.5 Polypropylene 4: 6.2% ± 2.2 Cotton 1: 5.0% ± 0.6 Cotton 2: 21.6% ± 1.8

Journ

al Pre-

Proof

Guangdong Meltblown Technology Co., Ltd. Polypropylene 4, interfacing material, spunbond (Hongxiang New Geo-Material Co., Ltd.), nonwoven, weight: 30 g/m2, thread count NA, layers NA Cotton 1, pillow cover, woven, weight: 116 g/m2, thread count NA, layers NA Cotton 2, t-shirt, knit, weight: 157 g/m2, thread count NA, layers NA Cotton 3, sweater, knit, weight: 360 g/m2, thread count NA, layers NA Polyester, toddler wrap, knit, weight: 200 g/m2, thread count NA, layers NA Silk, napkin, woven, weight: 84 g/m2, thread count NA, layers NA Nylon, exercise pants, woven, weight: 164 g/m2, thread count NA, layers NA Cellulose 1, paper towel, bonded, weight: 42.9 g/m2, thread count NA, layers NA Cellulose 2, tissue paper, bonded,

to that in typical human breathing. This flow rate was used to test all samples.

Cotton 3: 25.9% ± 1.4 Polyester: 17.5% ± 5.1 Silk: 4.8% ± 1.5 Nylon: 23.3% ± 1.2 Cellulose 1: 10.4% ± 0.28 Cellulose 2: 20.2% ± 0.32 Cellulose 3: 99.9% ± 0.02

Journ

al Pre-

Proof

weight: 32.8 g/m2, thread count NA, layers NA Cellulose 3, copy paper, boded, weight: 72.8 g/m2, thread count NA, layers NA

SD standard deviation

References

1. Capps JA. Measures for the prevention and control of respiratory infections in military camps. JAMA 1918; 71: 448-451. 2. Cooper DW, Hinds WC, Price JM, et al. Common materials for emergency respiratory protection: leakage tests with a manikin. American Industrial Hygiene Association journal 1983; 44: 720-726. 1983/10/01. DOI: 10.1080/15298668391405634. 3. Dato VM, Hostler D and Hahn ME. Simple respiratory mask. Emerging infectious diseases 2006; 12: 1033-1034. DOI: 10.3201/eid1206.051468. 4. Davies A, Thompson KA, Giri K, et al. Testing the efficacy of homemade masks: would they protect in an influenza pandemic? Disaster medicine and public health preparedness 2013; 7: 413-418. DOI: 10.1017/dmp.2013.43. 5. Davies A, Thompson KA, Giri K, et al. Frequently asked questions - homemade facemasks study, https://www.researchgate.net/publication/340253593_Frequently_asked_questions_-_homemade_facemasks_studypdf (2020, 2020-06-10; requires membership). 6. Davies A, Thompson KA, Giri K, et al. Facemask instructions COVID-19.doc, https://www.researchgate.net/publication/340023785_facemask_instructions_COVID-19doc (2020, accessed 2020-05-02; requires membership). 7. Doust BC and Lyon AB. Face masks in infections of the respiratory tract. JAMA 1918; 71: 12216-11218. 8. Furuhashi M. A study on the microbial filtration efficiency of surgical face masks--with special reference to the non-woven fabric mask. The Bulletin of Tokyo Medical and Dental University 1978; 25: 7-15. 9. Greene VW and Vesley D. Method for evaluating effectiveness of surgical masks. Journal of bacteriology 1962; 83: 663-667. 10. Guyton HG, Decker HM and Anton GT. Emergency respiratory protection against radiological and biological aerosols. AMA archives of industrial health 1959; 20: 91-95. 11. Haller DA and Colwell RC. The protective qualities of the gauze face mask - experimental studies. JAMA 1918; 71: 1213-1215. 12. Jang JY and Kim SW. Evaluation of Filtration Performance Efficiency of Commercial Cloth Masks. J Environ Health Sci 2015; 41: 203-215. 13. Jung H, Kim J, Lee SA, et al. Comparison of Filtration Efficiency and Pressure Drop in Anti-Yellow Sand Masks, Quarantine Masks, Medical Masks, General Masks, and Handkerchiefs. Aerosol and Air Quality Research 2014; 14: 991-1002. 14. Kellogg WH and MacMillan G. An experimental study of the efficacy of gauze face masks. American Journal of Public Health 1920; 10: 34-42. 15. Konda A, Prakash A, Moss GA, et al. Aerosol Filtration Efficiency of Common Fabrics Used in Respiratory Cloth Masks. ACS nano 2020; 14: 6339-6347. DOI: https://dx.doi.org/10.1021/acsnano.0c03252. 16. Konda A, Prakash A, Moss GA, et al. Correction to Aerosol Filtration Efficiency of Common Fabrics Used in Respiratory Cloth Masks. ACS Nano 2020. 17. Leete HM. Some experiments on masks. The Lancet 1919; 193: 392–393. 18. Lurie MB and Abramson S. The Efficiency of Gauze Masks in the Protection of Rabbits against the Inhalation of Droplet Nuclei of Tubercle Bacilli. American Review of Tuberculosis 1949; 59: 1-9.

https://www.researchgate.net/publication/340253593_Frequently_asked_questions_-_homemade_facemasks_studypdf

https://www.researchgate.net/publication/340023785_facemask_instructions_COVID-19doc

https://dx.doi.org/10.1021/acsnano.0c03252

Journ

al Pre-

Proof

19. MacIntyre CR, Seale H, Dung TC, et al. A cluster randomised trial of cloth masks compared with medical masks in healthcare workers. BMJ open 2015; 5: e006577. DOI: https://dx.doi.org/10.1136/bmjopen-2014-006577. 20. Paine CG. The aetiology of puerperal infection. BMJ 1935; 1: 243-246. 21. Quesnel LB. The efficiency of surgical masks of varying design and composition. The British journal of surgery 1975; 62: 936-940. DOI: 10.1002/bjs.1800621203. 22. Rengasamy S, Eimer B and Shaffer RE. Simple respiratory protection--evaluation of the filtration performance of cloth masks and common fabric materials against 20-1000 nm size particles. The Annals of occupational hygiene 2010; 54: 789-798. DOI: 10.1093/annhyg/meq044. 23. Shakya KM, Noyes A, Kallin R, et al. Evaluating the efficacy of cloth facemasks in reducing particulate matter exposure. Journal of exposure science & environmental epidemiology 2017; 27: 352-357. DOI: 10.1038/jes.2016.42. 24. Shooter RA, Smith MA and Hunter CJ. A study of surgical masks. The British journal of surgery 1959; 47: 246-249. DOI: 10.1002/bjs.18004720312. 25. van der Sande M, Teunis P and Sabel R. Professional and home-made face masks reduce exposure to respiratory infections among the general population. PloS one 2008; 3: e2618. DOI: 10.1371/journal.pone.0002618. 26. Weaver GH. The value of the face mask and other measures in prevention of diphtheria, meningitis, pneumonia, etc. JAMA 1918; 70: 76-78. 27. Weaver GH. Droplet Infection and Its Prevention by the Face Mask. The Journal of Infectious Diseases 1919; 24: 218-230. 28. Zhao M, Liao L, Xiao W, et al. Household Materials Selection for Homemade Cloth Face Coverings and Their Filtration Efficiency Enhancement with Triboelectric Charging. Nano letters 2020; 20: 5544-5552. DOI: 10.1021/acs.nanolett.0c02211.

https://dx.doi.org/10.1136/bmjopen-2014-006577

Journ

al Pre-

Proof

1

Supplementary material

Forgotten technology. Filtration properties of cloth and cloth masks: a narrative review

Journ

al Pre-

Proof

2

Supplementary Figure 1. Definitions and relationship between filtration efficiency, protection factor and total inward leak. For consistency, we calculated filtration efficiency from data provided in the original work, rather than presenting the data in the units chosen by the authors. Protection factor and fit factor are synonyms. Modified and reproduced with the permission of the American College of Physicians.1 Q factor as described by WHO updated guidance 2020 June 05,2 and Podgórski and colleagues.3 When calculated as defined below, WHO guidance states that expert consensus recommends a minimum Q factor of 3 (as well as other criteria) for mask material. Other authors4 use log10 rather than ln to define Q; results need to be interpreted differently.

There are several types of experiments described in the literature. Some experiments test flat cloth samples, and others test masks worn by volunteers or on manikins (figure 1). For the latter two cases, some experiments test how well the mask protects the wearer from external particles, and some test how well the mask stops a wearer from transmitting particles to the environment.

In all cases, the experiments compare particle concentrations on both sides of the mask or cloth sample. We use the term filter below to refer to either a mask or a cloth sample depending on the type of experiment.

Let chigh be the concentration of particles on the source side of the filter. The source side is the upstream side for tests of cloth; the outside for tests of masks as a means of protecting the wearer, and the inside for tests where the wearer is the source of the infection.

Let clow be the concentration of particles on the protected side of the filter. This is the opposite of the source side.

When measuring the filtration properties of filters, there are two contributions to particles on the protected side: particles that have passed through the mask (penetration), and particles that have leaked around the edges of the mask (edge leak). Experiments on cloth samples use equipment designed to exclude edge leak. With masks on manikins and volunteers, edge leak can be a significant factor.

The definitions below can be applied in all types of experiments. Of course, the measured efficiency of a piece of cloth would not be the same as the efficiency of a mask made of that cloth because in the latter case, edge leak would reduce the measured efficiency.

The filtration efficiency (FE) of the filter is the ratio of particles removed by the filter; this is a number in the range 0 < FE < 1. This is calculated by the formula:

𝐹𝐹𝐹𝐹 = 𝑐𝑐ℎ𝑖𝑖𝑖𝑖ℎ − 𝑐𝑐𝑙𝑙𝑙𝑙𝑙𝑙

𝑐𝑐ℎ𝑖𝑖𝑖𝑖ℎ

The protection factor (PF) of the filter is the ratio of particle concentration high to low; this is necessarily at least 1, and the higher the number, the better protection afforded by the filter. As a formula:

Journ

al Pre-

Proof

3

𝑃𝑃𝐹𝐹 =𝑐𝑐ℎ𝑖𝑖𝑖𝑖ℎ𝑐𝑐𝑙𝑙𝑙𝑙𝑙𝑙

These are related:

𝐹𝐹𝐹𝐹 = 𝑐𝑐ℎ𝑖𝑖𝑖𝑖ℎ − 𝑐𝑐𝑙𝑙𝑙𝑙𝑙𝑙

𝑐𝑐ℎ𝑖𝑖𝑖𝑖ℎ= 1 −

𝑐𝑐𝑙𝑙𝑙𝑙𝑙𝑙𝑐𝑐ℎ𝑖𝑖𝑖𝑖ℎ

= 1 − 1𝑃𝑃𝐹𝐹

We also define the total leakage (TL) to be proportion of particles admitted by the filter. This is also in the range 0 < TL < 1. In the literature, the name total inward leakage (TIL) is often used for this quantity.

𝑇𝑇𝑇𝑇 =𝑐𝑐𝑙𝑙𝑙𝑙𝑙𝑙𝑐𝑐ℎ𝑖𝑖𝑖𝑖ℎ

Since a particle is either admitted by the filter or removed by the filter, it is apparent that

𝑇𝑇𝑇𝑇 + 𝐹𝐹𝐹𝐹 = 1

so

𝐹𝐹𝐹𝐹 = 1 − 𝑇𝑇𝑇𝑇

Furthermore,

𝑃𝑃𝐹𝐹 = 1𝑇𝑇𝑇𝑇

Let ∆P be the pressure drop across the material (under standard conditions of area and air flow), in kPa. It is a measure of breathability: the higher the pressure drop, the more the work of breathing increases to maintain flow, and the lower the breathability.

The quality factor, Q, is used to describe the trade-off between efficiency and breathability for a given material, and is defined as

𝑄𝑄 =−ln(1 − 𝐹𝐹𝐹𝐹)

∆𝑃𝑃

Journ

al Pre-

Proof

4

Supplementary figure 2. Data are from Table 3 in Weaver 1919.5 In this experiment, the transmission of bacteria through material was studied using woven gauze with varying thread count and number of layers, for example, 3 layers of 24x20 gauze. We calculated thread count as the mean of the warp and the weft, multiplied by the number of layers. The relationship between transmission and thread count is shown for each of the distances studied. Transmission decreases as thread count increases.

Journ

al Pre-

Proof

5

Supplementary figure 3. Data are from Table 3 in Weaver 1919.5 In this experiment, the transmission of bacteria through material was studied using woven gauze with varying thread count and number of layers, for example, 3 layers of 24x20 gauze. We show the relationship between number of layers and transmission, for each type of gauze, at each distance studied. Transmission decreases as number of layers increases.

Journ

al Pre-

Proof

6

Supplementary figure 4. Data are from Table 3 in Weaver 1919.5 In this experiment, the transmission of bacteria through material was studied using woven gauze with varying thread count and number of layers, for example, 3 layers of 24x20 gauze. We calculated thread count as the mean of the warp and the weft, multiplied by the number of layers. The relationship between transmission and thread count is shown for each of the distances studied. Transmission decreases as thread count increases.

We selected, a priori, all pairs of data in which similar total thread counts were achieved with varying numbers of layers. 6 pairs were identified and graphed. In each case, achieving the same thread count with a smaller number of layers resulted in less transmission. This extended to the case of 1 vs 2 layers: 1 layer with mean thread count 42 (1 layer of 44x40) resulted in less transmission that 2 layers of mean total thread count 44 (2 layers of 24x20).

To convert this to the modern unit, threads per inch (TPI), the warp and the weft are added together. For example, 24x20 gauze would now be described as 44 TPI. Total thread counts of 168 (in 3 layers) and 176 (in 4 layers), bottom right, would convert to 336 TPI and 352 TPI respectively.

Distance to plates

Perc

enta

ge tr

ansm

issi

on

6 inche

s1 fo

ot2 fe

et3 fe

et4 fe

et5 fe

et6 fe

et7 fe

et8 fe

et0

20

40

60

80

100Total 34 threads, in 2 layersTotal 30 threads, in 1 layer

Distance to plates

Perc

enta

ge tr

ansm

issi

on

6 inche

s1 fo

ot2 fe

et3 fe

et4 fe

et5 fe

et6 fe

et7 fe

et8 fe

et0

20

40

60

80

100Total 44 threads, in 2 layersTotal 42 threads, in 1 layer

Distance to plates

Perc

enta

ge tr

ansm

issi

on

6 inche

s1 fo

ot2 fe

et3 fe

et4 fe

et5 fe

et6 fe

et7 fe

et8 fe

et0

20

40

60

80

100Total 68 threads, in 3 layersTotal 60 threads, in 2 layers

Distance to plates

Perc

enta

ge tr

ansm

issi

on

6 inche

s1 fo

ot2 fe

et3 fe

et4 fe

et5 fe

et6 fe

et7 fe

et8 fe

et0

20

40

60

80


Distance to plates

Perc

enta

ge tr

ansm

issi

on

6 inche

s1 fo

ot2 fe

et3 fe

et4 fe

et5 fe

et6 fe

et7 fe

et8 fe

et0

20

40

60

80


Distance to plates

Perc

enta

ge tr

ansm

issi

on

6 inche

s1 fo

ot2 fe

et3 fe

et4 fe

et5 fe

et6 fe

et7 fe

et8 fe

et0

20

40

60

80


Journ

al Pre-

Proof

7

References

1. Clase CM, Beale RCL, Dolovich MB, et al. Cloth masks may prevent transmission of COVID-19: an evidence-based, risk-based approach, https://www.acpjournals.org/doi/10.7326/M20-2567 (2020, accessed 2020-07-18). 2. World Health Organisation. Advice on the use of masks in the context of COVID-19: interim guidance 5 June 2020, https://www.who.int/publications/i/item/advice-on-the-use-of-masks-in-the-community-during-home-care-and-in-healthcare-settings-in-the-context-of-the-novel-coronavirus-(2019-ncov)-outbreak (2020, accessed 2020-06-07). 3. Podgórski A, Balazy A and Gradoń L. Application of nanofibers to improve the filtration efficiency of the most penetrating aerosol particles in fibrous filters. Chem Eng Sci 2006; 61: 6804-6815. 4. Zhao M, Liao L, Xiao W, et al. Household Materials Selection for Homemade Cloth Face Coverings and Their Filtration Efficiency Enhancement with Triboelectric Charging. Nano letters 2020; 20: 5544-5552. DOI: 10.1021/acs.nanolett.0c02211. 5. Weaver GH. Droplet Infection and Its Prevention by the Face Mask. The Journal of Infectious Diseases 1919; 24: 218-230.

https://www.acpjournals.org/doi/10.7326/M20-2567

https://www.who.int/publications/i/item/advice-on-the-use-of-masks-in-the-community-during-home-care-and-in-healthcare-settings-in-the-context-of-the-novel-coronavirus-(2019-ncov)-outbreak



els-jbs-prod-cdn.jbs.elsevierhealth.com · Journal Pre-Proof Mayo Clinic Proceedings Filtration...

Documents

Transcript of els-jbs-prod-cdn.jbs.elsevierhealth.com · Journal Pre-Proof Mayo Clinic Proceedings Filtration...