CoChat

The protective performance of reusable cloth face masks, disposable procedure masks, KN95 masks and N95 respirators: Filtration and total inward leakage

Open access
Duncan, Scott, Bodurtha, Paul, Naqvi, Syed · 2021 · PLOS ONE

Published version of a preprint

The DOI matches its Crossref record; 102 of the 102 references it cites were found in a registry or an index; 57 of the 132 sources its claims cite were checked for support.

Abstract

Face coverings are a key component of preventive health measure strategies to mitigate the spread of respiratory illnesses. In this study five groups of masks were investigated that are of particular relevance to the SARS-CoV-2 pandemic: re-usable, fabric two-layer and multi-layer masks, disposable procedure/surgical masks, KN95 and N95 filtering facepiece respirators. Experimental work focussed on the particle penetration through mask materials as a function of particle diameter, and the total inward leakage protection performance of the mask system. Geometric mean fabric protection factors varied from 1.78 to 144.5 for the fabric two-layer and KN95 materials, corresponding to overall filtration efficiencies of 43.8% and 99.3% using a flow rate of 17 L/min, equivalent to a breathing expiration rate for a person in a sedentary or standing position conversing with another individual. Geometric mean total inward leakage protection factors for the 2-layer, multi-layer and procedure masks were <2.3, while 6.2 was achieved for the KN95 masks. The highest values were measured for the N95 group at 165.7. Mask performance is dominated by face seal leakage. Despite the additional filtering layers added to cloth masks, and the higher filtration efficiency of the materials used in disposable procedure and KN95 masks, the total inward leakage protection factor was only marginally improved. N95 FFRs were the only mask group investigated that provided not only high filtration efficiency but high total inward leakage protection, and remain the best option to protect individuals from exposure to aerosol in high risk settings. The Mask Quality Factor and total inward leakage performance are very useful to determine the best options for masking. However, it is highly recommended that testing is undertaken on prospective products, or guidance is sought from impartial authorities, to confirm they meet any implied standards.

Overview

CoChat Summary

TL;DR

This study measured particle filtration through mask materials and total inward leakage (TIL) for reusable cloth, disposable procedure, KN95, and N95 masks. It found that mask performance is dominated by face-seal leakage: N95 FFRs provided high filtration and high TIL protection, while cloth and procedure masks provided only marginal TIL protection despite material differences.

Key findings

  • Geometric mean fabric protection factors (FPF) ranged from 1.78 for 2-layer fabric masks to 144.5 for KN95 materials, corresponding to overall filtration efficiencies of 43.8% and 99.3% at 17 L/min; N95 FFR materials had a geometric mean FPF of 69.8 (~1.4% overall penetration).
  • Geometric mean total inward leakage protection factors (TILPF) were 1.42 for 2-layer fabric masks, 1.78 for multi-layer fabric masks, 2.26 for disposable procedure masks, 6.20 for KN95 masks, and 165.7 for N95 FFRs, corresponding to TIL penetration of 70%, 56.5%, 44.2%, 16.1%, and 0.60%, respectively.
  • Adding filtering layers to cloth masks did not improve TILPF; multi-layer masks with polypropylene filtering material were indistinguishable in penetration from 2-layer masks, and PM 2.5 filter inserts gave FPF ~71 but TILPF only 1.8.
  • N95 FFRs were the only group with both high filtration efficiency and high TIL protection; their geometric mean quantitative fit factor was ~644, equivalent to a face-seal efficiency of 99.84%.

Methods

  • Five mask groups were tested: reusable fabric 2-layer and multi-layer masks, disposable procedure/surgical masks, KN95 masks, and N95 FFRs. Filtration testing used NaCl aerosol with SMPS/APS over ~0.027–5.0 μm at 17 L/min; TIL testing used a TSI PortaCount 8038 with seven CSA-recommended activities.
  • TIL tests recruited 11 trained subjects from DRDC Suffield, selected for a cross-section of sex, age, height, weight, and facial size; all were quantitatively fit tested to FFRs (fit factor >100).
  • Analyses included geometric mean FPF and TILPF, harmonic mean FPF, ANOVA with Bonholm/Tukey tests, and Kruskal-Wallis with Dunn’s test where normality failed.

Limitations

  • The paper states that its estimates of penetrating virion numbers should not be taken as absolute, because the assumption that virion number depends directly on particle volume has not been verified.
  • The paper notes that KN95 performance has been found to be inconsistent across models, and recommends testing prospective products or seeking impartial guidance to confirm implied standards.
  • Evident from the text: TIL testing used a small, trained volunteer pool (11 subjects) and only one N95 model (3M 9210) for TIL, so results may not generalize to untrained users or other N95 models; filtration testing was at 17 L/min rather than standard 28 or 85 L/min, and some tests had a maximum upper particle range of ~3.5 μm due to insufficient larger particles.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1371/journal.pone.0258191

Cites

102 of 102 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Figures

Fig 1 (a) Disposable procedure mask: Henan Liwei Biological Pharmaceutical Co. Ltd; (b) Multi-layer mask: four layered system comprised of two layers of quilt batting between an outer and inner layer of cotton; (c) Non-woven polypropylene ‘craft’ material (manufactured by smart-fab®); (d) “N95-like” electret filter membrane material (purchased from Amazon.ca) used as the middle filter layer, as part of a multi-layer mask; (e) 2-Layer mask: quilt batting/cotton mask; (f) KN95 mask: MedSup Canada Protective Face Mask, and; (g) N95 FFR: 3M™ model 9210.
Fig 1 (a) Disposable procedure mask: Henan Liwei Biological Pharmaceutical Co. Ltd; (b) Multi-layer mask: four layered system comprised of two layers of quilt batting between an outer and inner layer of cotton; (c) Non-woven polypropylene ‘craft’ material (manufactured by smart-fab®); (d) “N95-like” electret filter membrane material (purchased from Amazon.ca) used as the middle filter layer, as part of a multi-layer mask; (e) 2-Layer mask: quilt batting/cotton mask; (f) KN95 mask: MedSup Canada Protective Face Mask, and; (g) N95 FFR: 3M™ model 9210.CC BY · PubMed Central
Fig 2 Schematic of aerosol swatch penetration set-up. A polydisperse NaCl aerosol is generated in the mixing chamber and pulled through the charge neutralizer and swatch test rig using a regenerative blower. Air flow is regulated using the flow meter. Aerosol concentration, upstream and downstream, is measured with the SMPS/APS as shown.
Fig 2 Schematic of aerosol swatch penetration set-up. A polydisperse NaCl aerosol is generated in the mixing chamber and pulled through the charge neutralizer and swatch test rig using a regenerative blower. Air flow is regulated using the flow meter. Aerosol concentration, upstream and downstream, is measured with the SMPS/APS as shown.CC BY · PubMed Central
Fig 3 Total inward leakage test being performed on the 3M™ model 9210 N95 FFR. The Sampling probe for the TSI Portacount 8038 is shown to be connected to the respirator by use of a sampling probe supplied by TSI. The operator is shown to be wearing a loose-fitted PAPR and nitrile gloves as a health and safety precaution from working in close proximity to the test volunteers during the COVID-19 pandemic.
Fig 3 Total inward leakage test being performed on the 3M™ model 9210 N95 FFR. The Sampling probe for the TSI Portacount 8038 is shown to be connected to the respirator by use of a sampling probe supplied by TSI. The operator is shown to be wearing a loose-fitted PAPR and nitrile gloves as a health and safety precaution from working in close proximity to the test volunteers during the COVID-19 pandemic.CC BY · PubMed Central
Fig 4 Mean penetration profile as a function of aerodynamic particle diameter for the material of five fabric masks comprised of 2 layers. Note, the penetration range is 0 to 100%. Materials included quilt batting, cotton, nylon, polyester and silk.
Fig 4 Mean penetration profile as a function of aerodynamic particle diameter for the material of five fabric masks comprised of 2 layers. Note, the penetration range is 0 to 100%. Materials included quilt batting, cotton, nylon, polyester and silk.CC BY · PubMed Central
Fig 5 Mean penetration profile as a function of aerodynamic particle diameter for the material of seven fabric masks with an internal layer to augment filtration. Note, the penetration range is 0 to 100%. Multi-layer material systems included: i) cotton/electret furnace filter/cotton, ii) cotton/procedure mask/cotton, iii) cotton/electret filter membrane/cotton, iv) cotton/quilt batting/cotton, v) cotton/polypropylene shopping bag/cotton, vi) cotton/polypropylene craft 1/cotton, and vii) cotton/polypropylene craft 2/cotton.
Fig 5 Mean penetration profile as a function of aerodynamic particle diameter for the material of seven fabric masks with an internal layer to augment filtration. Note, the penetration range is 0 to 100%. Multi-layer material systems included: i) cotton/electret furnace filter/cotton, ii) cotton/procedure mask/cotton, iii) cotton/electret filter membrane/cotton, iv) cotton/quilt batting/cotton, v) cotton/polypropylene shopping bag/cotton, vi) cotton/polypropylene craft 1/cotton, and vii) cotton/polypropylene craft 2/cotton.CC BY · PubMed Central
Fig 6 Mean penetration profile as a function of aerodynamic particle diameter for the filter material of six disposable procedure masks. Note, the penetration range is 0 to 100%. The bump in the penetration data between the particle range 0.5 to 0.67 μm is related to averaging the SMPS and APS measurements.
Fig 6 Mean penetration profile as a function of aerodynamic particle diameter for the filter material of six disposable procedure masks. Note, the penetration range is 0 to 100%. The bump in the penetration data between the particle range 0.5 to 0.67 μm is related to averaging the SMPS and APS measurements.CC BY · PubMed Central
Fig 7 Mean penetration profile as a function of aerodynamic particle diameter for the filter material of two KN95 masks (five replicates each). Note, the penetration range is 0 to 20%.
Fig 7 Mean penetration profile as a function of aerodynamic particle diameter for the filter material of two KN95 masks (five replicates each). Note, the penetration range is 0 to 20%.CC BY · PubMed Central
Fig 8 Mean penetration profile as a function of aerodynamic particle diameter for the filter material of five N95 FFRs (three replicates each). Note, the penetration range is 0 to 20%.
Fig 8 Mean penetration profile as a function of aerodynamic particle diameter for the filter material of five N95 FFRs (three replicates each). Note, the penetration range is 0 to 20%.CC BY · PubMed Central
Fig 9 Protection performance results for the five mask groups; fabric 2-layer masks (2-L), multi-layer masks (M-L), disposable procedure masks (PMs), KN95 masks and N95 FFRs. TILPF—total inward leakage protection factor. FPF—fabric protection factor (overall protection of fabric material against entire particle size distribution). FF—fit factor (total inward leakage measured through face seal only (by QNFT). Sample sizes: n = 8 to 26 for TILPF; n = 6 to 20 for FPF; and; n = 30 for FF. The geometric mean is shown by the square marker.
Fig 9 Protection performance results for the five mask groups; fabric 2-layer masks (2-L), multi-layer masks (M-L), disposable procedure masks (PMs), KN95 masks and N95 FFRs. TILPF—total inward leakage protection factor. FPF—fabric protection factor (overall protection of fabric material against entire particle size distribution). FF—fit factor (total inward leakage measured through face seal only (by QNFT). Sample sizes: n = 8 to 26 for TILPF; n = 6 to 20 for FPF; and; n = 30 for FF. The geometric mean is shown by the square marker.CC BY · PubMed Central
Fig 10 QF values for the fabric 2-layer, multi-layer, disposable procedure/surgical and KN95 mask groups. As the face velocities used in filtration testing for the N95 FFR group were dissimilar to the other three mask groups, their QF values were not plotted for comparison in this figure. DPM = disposable procedure mask.
Fig 10 QF values for the fabric 2-layer, multi-layer, disposable procedure/surgical and KN95 mask groups. As the face velocities used in filtration testing for the N95 FFR group were dissimilar to the other three mask groups, their QF values were not plotted for comparison in this figure. DPM = disposable procedure mask.CC BY · PubMed Central

Verification

Claims the paper makes

23 Valid · 3 Not valid · 37 Not checked · 139 Own results and interpretations

117 more claims · 7 with their cited sources checked

Create a free account to see all 202 claims and to verify the 4 references still open.

See all claims — free account
  • It may state two separate claims as one. 54
  • Probably not a substantial claim. 34

Create a free account to see these claims.

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

5 of 5

Citations

Cites

Show all 102 references

Research works printed but not linked: 2

  1. L Bourouiba, 2016, Images in Clinical Medicine. A Sneeze, N Engl J Med, 375, e15
  2. J Gawn, 2008, Evaluating the protection afforded by surgical masks against influenza bioaerosols

Cited by

See all 75

Metadata

Authors
Duncan, S., Bodurtha, P., & Naqvi, S.
Affiliations
Defence Research and Development Canada (CA)
Corresponding author
Duncan, Scott
Venue
PLOS ONE
Published
2021-10-06
Volume
16
Issue
10
Pages
e0258191
ISSN
1932-6203 (ISSN-L)
OpenAlex
W3202327157
Cited by
75
References
91
Citations by year
2026: 8 · 2025: 7 · 2024: 21 · 2023: 16 · 2022: 20 · 2021: 3
FWCI
9.73
Type
journal article
Language
en
Open access
Locations
  • PLoS ONE (publishedVersion, cc-by)
  • medRxiv (acceptedVersion, cc-by-nc-nd)
  • Europe PMC (PubMed Central) (publishedVersion, other-oa)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
Topics
  • Health Sciences › Medicine › Pulmonary and Respiratory Medicine › Infection Control and Ventilation
  • Physical Sciences › Engineering › Electrical and Electronic Engineering › Aerosol Filtration and Electrostatic Precipitation
  • Social Sciences › Sociology and Political Science › COVID-19 Prevention and Impact
Keywords
face masks, cloth masks, KN95 masks, N95 respirators, surgical masks, filtration efficiency, particle penetration, SARS-CoV-2, aerosol protection, face coverings, respiratory protection
MeSH terms
COVID-19, SARS-CoV-2, N95 Respirators, Disposable Equipment, Equipment Design, Filtration, Humans, Masks, Respiratory Protective Devices, Textiles, Equipment Reuse

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

The claim verdicts this report counts are experimental: their accuracy is not yet measured.

A language model drew the claims from the paper on 2026-10-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-10 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Cited sources are checked with a language model’s help (gpt-6-sol and qwen3.8-omni-flash).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: The protective performance of reusable cloth face masks, disposable procedure masks, KN95 masks and N95 respirators: Filtration and total inward leakage. https://app.cochat.ai/papers/10.1371/journal.pone.0258191
The paper
Duncan, S., Bodurtha, P., & Naqvi, S. (2021). The protective performance of reusable cloth face masks, disposable procedure masks, KN95 masks and N95 respirators: Filtration and total inward leakage. PLOS ONE. https://doi.org/10.1371/journal.pone.0258191

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.