Filtration and Sampling of Aerosols by Fibrous Filters - Aerosols: Measurement, Dosimetry, and Health Effects - page 455

17.6 ON TESTING SUBMICRON AND NANOFIBER
FILTERS WITH NANOPARTICLES
The researched mechanisms of aerosol particle deposition on ibers from a low enable inding the
size of the most penetrative particles for a ilter with given parameters depending on the low veloc-
ity. It leads to the possibility to choose correctly a method for testing high effective ilters.
There are several reasons explaining the discrepancy between experimental and calculation data.
Such discrepancy may relate to the coagulation of tested solid particles; to partial clogging of a il-
ter; to the inluence of charges on particles or ibers that are not taken into account in calculation; to
incomplete drying out of particles (e.g., to the presence of a “coat” on polystyrene latex particles);
to polydispersity of particles; to defects in the ilter structure (e.g., thin through-holes in a ilter), etc.
Besides, it is important while testing ilters to take into account the polydispersity of aerosols and
particles shapes. There are also many unsolved problems related, for example, to calculation errors,
taking into account charges. The mechanism of blowing out particles and iber fragments at ilter
vibrations is not clear. The mechanism of particles rebound from ibers is not investigated when
there is a local crack in the particle deposit on the ilter and velocity in the local crack increases
multiply. It is dificult to evaluate the inluence of hidden defects of highly effective ilters on the
penetration of particles. One can discover ilter defects with the help of nanoparticles with diameter
size of order of 0.01 μm: probability of penetration of such small particles is negligibly small at a
low low velocity and their presence after a ilter testiies the presence of a defect. It is necessary to
carefully check the measurement technique if the experimental data sharply differ from the calcu-
lated one. It is especially important while measuring the deposition of particles smaller than MPPS.
As it was mentioned earlier, modern highly effective iltering materials are manufactured from
very thin ibers; that is why the area of mostly penetrative particles moved to the range lower than
0.1 μm. At the same time according to commonly accepted standards one should measure the pen-
etration of particles around its maximum point. Not many photoelectric counters have sensitivity
suficient to detect particles in a range of 1 μm that is why for their registration now as half a century
ago one has to use method of enlargement of particles.
At present, a procedure of measuring size and concentration of nanoparticles based on applica-
tion of condensational counters is widely used. Test particles are obtained while separating polydis-
perse particles charged with one elementary charge in a differential mobility analyzer. Particles are
obtained while spraying of liquids—dioctyl phthalate or common salt solution.
It should be taken into account while measuring penetration of small particles through ilters
that not all particles get enlarged in condensational counters using oversaturated spirit vapors but a
part thereof depending on their size and concentration. It affects the test accuracy of highly eficient
ilters. This explains why penetration curves may differ from the calculated ones. Penetration of
particles in the area of sizes less than 0.07 μm varies very little with the variation of particle sizes,
especially in the case of К < 0.01.
Applicability of condensational counters to testing of high-eficiency ilters is easily proved by
measuring the penetration of particles through several sequentially placed similar units of iltering
materials. In accordance with the basic iltration law a correlation should be fulilled n/n 0 = ( n i / n 0 ) m ,
where m is the number of units of material and n i is the concentration of particles with radius r j after
one ilter layer.
Measurements that we conducted at a TSI mod 6030 installation (which incorporated a gen-
erator of monomobile particles of dioctyl phthalate and sodium chlorate as well as two condensa-
tional counters) with ive layers of iltering material, demonstrated that this correlation was strictly
observed only for particles with the size larger than 0.07 μm. For particles with a diameter of
0.02 μm at velocity 2 cm/s the value of n 0 / n 1 for the irst ilter was 115 and for the ifth n 4 / n 5 = 6 only.
Regretfully, most of the authors do not take that into account. As a result, not quite true generaliza-
tions and conclusions are published. The most reliable method of testing high-eficiency ilters is a
comparison method with standard reference ilters [92].