Based on preliminary studies (submitted for peer review but not yet published; pending suggestions for improvement from our peers):
We have developed a rapid testing protocol for loose-fitting type masks that provides information on particle removal efficacy and an interpretation of the results that independently assesses the quality of the mask fit and of the mask material. This takes less than 2 hours to test a new mask design (n=3 masks tested, n=3 tests per mask).
The protocol was validated on N95 masks; for the well-fitted mask we record >99% particle removal efficiency, and for the poorly-fitted mask the mean removal efficiency is just about 90%.
We have tested three different brands of commercial surgical masks. Worn as designed these had mean removal efficiency ranging from 50% to 75%, however all of them had close to 90% particle removal efficiency when a nylon layer was added. This indicates that the weakness of these masks is primarily a poor fit rather than a poor filter material.
Fabric masks have highly variable results, from less than 30% to nearly 90% particle removal efficiencies when worn as designed. The table below provides detailed data for each mask tested, including average (mean) particle removal efficiency and whether or not improvement was seen by adding a nylon layer. Some were as effective as commercial medical masks.
The addition of a nylon stocking overlayer improved the removal efficiency for nearly all of the surgical-style loose-fitting masks but few of the cone-shaped masks. This may indicate that cone-shaped masks have fewer air-leakage pathways, although our tests were run using only a single mask-wearer, so this result could vary from face-to-face.
The masks that achieved the highest levels of filtration when using the nylon stocking layer (the indicative test for material quality) each included a filter layer (organic cotton batting, Pellon, or loosely-woven cotton muslin) in addition to two layers of cotton fabric.