METHOD AND APPARATUS FOR GENERATING TEST AEROSOL

FIELD OF THE INVENTION

The present invention relates to generating aerosol, which comprises singly ionized particles. The present invention also relates to an apparatus for generating aerosol.

BACKGROUND

The term aerosol refers to liquid or solid particles suspended in a gas. Aerosol measurements are needed e.g. in air pollution studies, for monitoring clean rooms in semiconductor industry, and for testing particle emissions from engines and power plants. Aerosol measuring instruments need to be tested or calibrated.

Referring to Fig. 1 , an aerosol measuring instrument may be connected to a test aerosol generator in order to calibrate and/or test the measuring instrument.

Aerosol particles may be electrically neutral, singly charged, doubly charged, or triply charged, etc. The charge of a singly charged particle is equal to the elementary charge ±e. The charge may be positive or negative. The charge of a doubly charged particle is equal to ±2 e, respectively. The number of particles of an aerosol flow may be determined by collecting an electric current carried by the particles, provided that the number of multiply charged particles is negligible.

US patent application 2005/0180543 discloses an apparatus for generating aerosol, said apparatus comprising an X-ray source, and a differential mobility analyzer arranged to separate particles of predetermined electrical mobility. The apparatus of US 2005/0180543 can be arranged to generate singly charged aerosol particles such that the relative fraction of doubly charged particles is equal to or less than 5%, by adjusting operating current and voltage of the X-ray source. In the apparatus of US 2005/0180543, there may be a need to adjust e.g. the operating current and the operating voltage of an X-ray particle charger according to size of the charged particles.

An article "Characteristics of the Berner impactor for sampling inorganic ions", by H. -C.Wang and W.John in Aerosol Science and Technology, Vol. 8 (1988), pp. 157-172, discloses a method for generating monodisperse submicron aerosol for test purposes (see Fig. 3 of said article and the related discussion). Aerosol particles were generated by coating solid ammonium fluorescein particles with oleic acid. The coated aerosol particles were subjected to charging by a radioactive source, and charged particles of a predetermined size were separated by an electrostatic classifier. Said predetermined size was set to be well above the peak of the aerosol size distribution guided to the electrostatic classifier in order to suppress the relative contribution of doubly charged particles. The separated particles were subsequently neutralized.

An article "Modification of the University of Washington Mark 5 in-stack impactor", by E.I.Kauppinen and R.E.Hillamo, in Journal of Aerosol Science, Vol. 20, No. 7 (1989), pp. 813-827, discloses calibrating impactor stages of a low-pressure impactor by using dioctyl phthalate (DOP) particles (see Fig. 2 of said article). Charged test particles are provided by using an electrostatic classifier. The particle size extracted by the classifier may be set to be larger than the mean diameter of an aerosol fed into the classifier in order to reduce the number of multiply charged particles (see end of page 819). Chapter 4.2. of said article discusses corrections, which are needed when a fraction of the test particles are multiply charged.

SUMMARY

An object of the present invention is to provide an apparatus for generating test aerosol comprising singly charged particles, wherein the relative fraction of neutral and multiply charged particles is low. A further object of the present invention is to provide a method for generating test aerosol. Yet, an object of the present invention is also to provide a method for testing aerosol measuring instruments.

According to a first aspect of the invention, there is provided a test aerosol generating apparatus according to claim 1.

According to a second aspect of the invention, there is provided a method for generating test aerosol according to claim 8.

According to a third aspect of the invention, there is provided a method for testing a measuring instrument according to claim 9.

Test aerosol comprising singly charged particles may be generated by charging small primary particles, and by increasing the size of the charged primary particles in a particle growing unit so as to form larger charged particles. The presence of multiply charged particles may be substantially avoided by limiting the upper size of charged primary particles prior to increasing the size of said particles. Singly charged particles may be subsequently separated from neutral particles and from particles of opposite polarity e.g. based on electrical mobility of the particles so as to provide the test aerosol. The relative fraction of multiply charged particles may be negligible, thanks to the small size of the particles introduced to the particle growing unit. The generated test aerosol may substantially consist of either positive or negative singly charged particles suspended in a gas. Consequently, the number of particles of the test aerosol may be accurately determined by collecting an electric current carried by the charged particles of said test aerosol.

The relative fraction of multiply charged particles may be smaller than in a case where the particles of a test aerosol are charged without increasing the size of the particles after charging. In other words, the generated test aerosol may provide more accurate calibration or test results. Thanks to limiting the upper size of the charged particles, it is not necessary to tune the particle size separated by the separating unit according to the size distribution of particles guided to the separating unit.

Thanks to limiting the upper size of the charged particles, it is not necessary to adjust the operating parameters of the charging unit according to the size distribution of particles guided to the charging unit. Furthermore, there is no need to accurately measure or adjust the ion concentration in the charging unit.

Multiply charged particles may be substantially avoided even if the ion concentration used for the charging would be high. Consequently, it is not necessary to neutralize primary particles prior to charging, because the charging unit may be arranged to act as a neutralizer for multiply charged primary particles. Consequently, the primary particles may also be generated by a particle generator, which generates a significant amount of multiply charged particles. This may also allow more freedom to select the material of the generated particles.

The method and apparatus according to the invention may be used to generate aerosol for testing and/or calibration of e.g. aerosol measuring instruments, aerosol sampling probes, or aerosol sampling lines.

The method and apparatus according to the invention may be used to provide a known number of aerosol particles. The method and apparatus according to the invention may be used to provide a known number density of aerosol particles, i.e. a known number of particles per unit volume. Hence, also a standard for the number of aerosol particles and/or a standard for number density of aerosol particles may be established.

The embodiments of the invention and their benefits will become more apparent to a person skilled in the art through the description and examples given herein below, and also through the appended claims. BRIEF DESCRIPTION OF THE DRAWINGS

In the following examples, the embodiments of the invention will be described in more detail with reference to the appended drawings, in which

Fig. 1 shows a measuring instrument to be tested, wherein the measuring instrument is connected to an output of the test aerosol generating apparatus,

Fig. 2 shows an aerosol flow consisting of charged particles suspended in a gas,

Fig. 3a shows a charge detector connected to the output of the test aerosol generating apparatus,

Fig. 3b shows a charge detector connected arranged to detect particles transmitted through the measuring instrument,

Fig. 3c shows distributing a test aerosol flow to a measuring instrument and to a charge detector,

Fig. 4 shows a charge detector,

Fig. 5a shows functional units of the test aerosol generating apparatus,

Fig. 5b shows method steps for generating test aerosol,

Fig. 6 shows a primary particle generator based on condensation,

Fig. 7 shows a primary particle generator based on nebulizing and drying, Fig. 8 shows a particle charger based on ionizing radiation,

Fig. 9 shows a particle charger based on corona discharge,

Fig. 10 shows a particle separating unit based on electrical mobility,

Fig. 11 shows a particle growing unit,

Fig. 12 shows behavior of neutral and singly charged particles during charging and growing,

Fig. 13 is a comparative example showing behavior of large and multiply charged particles,

Fig. 14a shows, by way of example, the relative fraction of singly charged particles versus particle diameter, and the relative fraction of doubly charged particles after charging,

Fig. 14b shows the ratio of the number of doubly charged particles to the number of singly charged particles versus particle diameter in the situation of Fig. 14a,

Fig. 15 shows behavior of neutral and singly charged particles in the test aerosol generating apparatus,

Fig. 16 shows a test aerosol generating apparatus, wherein the output of the growing unit is connected to the input of the separating unit,

Fig. 17 shows behavior of particles in the apparatus of Fig. 16,

Fig. 18a shows a particle size distribution, which has a wide descending slope,

Fig. 18b shows positioning of a pass band of a separating unit on the descending slope of a particle size distribution, Fig. 18c shows a particle size distribution, which has a narrow descending slope,

Fig. 19 shows a test aerosol generating apparatus comprising a neutralizer arranged to neutralize electrically charged particles prior to charging, and

Fig. 20 shows a test aerosol generating apparatus comprising a further size selecting unit to limit the width of the particle size distribution prior to charging.

The drawings are schematic.

DETAILED DESCRIPTION

Referring to Fig. 1 , a test aerosol generating apparatus 500 may be arranged to generate test aerosol A5, which comprises singly charged particles.

A measuring instrument 900 may be tested or calibrated by guiding a flow of the test aerosol A5 to the measuring instrument 900.

An exhaust port of the measuring instrument 900 may provide an exhaust flow EX9.

The measuring instrument 900 may capture or neutralize particles. Thus, the exhaust flow EX9 may comprise substantially 100% of the test aerosol A5, more than 0% and less than 100% of the particles of the test aerosol A5, or substantially 0% of the particles of the test aerosol A5.

The test aerosol A5 may also be guided to a chemical and/or physical process e.g. in order to study gas-particle interactions. Referring to Fig. 2, the test aerosol A5 may comprise singly charged particles VLP1 such that the relative fraction of neutral and multiply charged particles is low. The number density of multiply charged particles may be e.g. smaller than or equal to 5% of the number density of singly charged particles VLP1 , advantageously less than 2%, preferably less than 1 %. The number density of neutral particles may be e.g. smaller than or equal to 5% of the number density of singly charged particles, advantageously less than 2%, preferably less than 1 %.

If the generated test aerosol A5 comprises a major amount of positive particles, then the number density of negative particles may be e.g. smaller than or equal to 5% of the number density of the positive particles, advantageously less than 2%, preferably less than 1 %. If the test aerosol comprises a major amount of negative particles, then the number density of positive particles may be e.g. smaller than or equal to 5% of the number density of the positive particles, advantageously less than 2%, preferably less than 1 %.

The number of particles VLP1 traveling through a cross-sectional area AR5 may be determined by measuring the total charge of particles traveling through the cross-sectional area AR5. The number of particles traveling through the cross-sectional area AR5 may be determined accurately, thanks to the small relative fraction of neutral and multiply charged particles. The area AR5 may be the cross- sectional area of a duct 510. The particles may be carried by a gas G5.

The number of particles traveling through the cross-sectional area AR5 per unit time may be determined by collecting all electrical charges carried by the particles traveling through the area AR5 or impinging on said area AR5, and by measuring a current corresponding to the collected charges.

If the test aerosol A5 comprises singly charged particles, and if the number of neutral particles and the number of particles of opposite polarity are low, then the number of particles traveling through the cross-sectional area AR5 per unit time (dN/dt) may be given by:

dN ₌ i dt e

where I denotes the electrical current traveling through the area AR5, and e denotes the elementary charge 1.602-10 ^"19 C.

Referring to Fig. 3a, a current corresponding to the charge of the particles traveling through the cross-sectional area AR5 per unit time may be measured by a charge detector 600. Total charge of particles passed through the area AR5 may be determined by integrating the detected current I with respect to time.

The charge detector 600 may provide an exhaust flow EX1. The exhaust flow may or may not comprise particles. Particles of the exhaust flow EX1 may be substantially neutral, because the charges of the particles have been collected by the detector 600.

In some cases, a measuring instrument 900 or a sampling line may also be calibrated by determining the number or number density of particles which have passed through a measuring instrument 900. Referring to Fig. 3b, the number or number density of particles may be determined by guiding an output flow A9 of the measuring instrument to the charge detector 600. The set-up of Fig. 3b may be applied provided that the measuring instrument does not significantly neutralize, charge, or agglomerate particles.

In some cases, a measuring instrument 900 or a sampling line may be tested by comparing a result obtained by the set-up of Fig. 3a with a result obtained by the set-up of Fig. 3b. E.g. the determined particle numbers, particle concentrations, current values, or total charges may be compared.

Referring to Fig. 3c, the output aerosol A5 of the test aerosol generator 500 may be divided or distributed to a measuring instrument 900 and to a charge detector 600. The aerosol flow A5 may be divided into flows A5a and A5b by a branching device 700. The branching device 700 may be e.g. a Y-shaped tube, which is preferably dimensioned so that it does not capture a significant amount of particles.

The number density of particles of the test aerosol A5 may be determined based on a current signal provided by the charge detector 600 and based on the flow rate of gas through said charge detector 600. The gas flow rate may be measured by a flow meter 710.

The number of particles guided into the measuring instrument 900 may be determined if the gas flow rate through the measuring instrument is determined in addition to the particle number density. The gas flow rate through the measuring instrument may be determined e.g. by a flow meter 720 connected to the exhaust flow EX9 of the measuring instrument.

The branching device 700 may also be a multi-port valve, in particular a three-port valve. The flow of test aerosol A5 may be guided to a three-port valve, which is connected to the input of the measuring instrument 900 and to the charge detector 600. The aerosol flow A5 may be guided during a first time period to the measuring instrument 900 and during a second time period to the charge detector 600.

The multiport valve may also have four ports, wherein the test aerosol A5 may be guided to a first port, clean gas may be guided to a second port, a third port may be connected to the measuring instrument 900, and a fourth port may be connected to the charge detector 600. The aerosol flow A5 may be guided during a first time period to the measuring instrument 900 and during a second time period to the charge detector 600. The clean gas flow may be guided during the first time period to the charge detector 600 and during the second time period to the measuring instrument 900, respectively.

Referring to Fig. 4, a charge detector 600 may comprise an electrically conductive element 610 which is arranged to operate as a Faraday cup. Electric charges carried by the particles of the aerosol A5 may be collected by the conductive element 610 and the collected charges may be transferred to a current-measuring device 650 by a conductor 655. The letter "A" denotes measurement of electrical current.

The conductive element 610 may comprise e.g. a filter, which is disposed in a duct 640. The dimensions of the wires or pores of the filter may be dimensioned such that at least 95% of the charges carried by the particles VLP1 may be collected, according to the flow velocity, gas density, and size and density of the particles. The exhaust flow EX1 comprises neutral particles VLPO. In order to further improve the accuracy, the filter may be typically arranged to collect at least 99% or even at least 99.99% of the charges carried by the particles VLP1.

Said filter may be electrically conductive. For example, the filter may be made of sintered grains of stainless steel, or the filter may comprise one or more stacked layers of conductive mesh. The filter may also comprise electrically insulating material if the electrically insulating material is enclosed within an electrically conductive structure. The electrically conductive structure may be coupled to the current- measuring device 650. The conductive element 610 may e.g. comprise glass or plastic fibers enclosed within a stainless steel mesh.

The detector 600 may comprise one or more insulators 630 which support the element 610, which insulate the element 610 from the duct 640 and/or the surroundings, and which may prevent leaking through a gap between the element 610 and the wall of the duct 640.

The other pole of the current measuring device 650 may be connected to an electrical ground GND of the apparatus 500. The conductive element 610 is advantageously enclosed within an electrostatic shield, which protects the the conductive element 610 from electric disturbances. In particular, the duct 640 may be electrically conductive and it may be arranged to act as an electrostatic shield for the conductive element 610. Also the duct 640 may be connected to the electrical ground GND of the apparatus 500.

Referring to Fig. 5a, the test aerosol generating apparatus 500 according to the present invention comprises a primary particle generator 100, a particle charger 200, and a particle growing unit 400.

The apparatus 500 may further comprise a particle separating unit 300. For certain applications, the output of the charging unit 200 may also be coupled directly to the growing unit 400.

The particle generator 100 provides an aerosol flow A1 , which comprises primary particles. The aerosol flow A1 may be guided to the charger 200, which provides an aerosol flow A2 comprising charged primary particles, by charging primary particles of the aerosol flow A1. The charging unit 200 provides charged primary particles by charging primary particles.

The charged aerosol flow A2 may now comprise both positive and negative particles so that a total charge carried by positive charged particles is substantially equal to a total charge carried by negative charged particles. In other words, the charged aerosol A2 may be quasi-neutral.

A separating unit 300 may be arranged to separate either positive or negative particles of the aerosol flow A2 so as to provide a separated aerosol flow A3, which substantially consists of positive charged primary particles suspended in a gas. Alternatively, the separated aerosol flow A3 may substantially consist of negative charged primary particles suspended in a gas. In other words, the particle separating unit 300 may be arranged to separate positive charged particles from neutral particles and from negative charged particles, or to separate negative particles from neutral particles and from positive charged particles.

The separating unit 300 may be arranged to separate particles based on electrical mobility.

The aerosol flow A3 may be guided to a particle growing unit 400, which is arranged to increase the size of the primary particles of the aerosol flow A3.

Multiply charged particles may be substantially avoided by limiting the upper size of charged primary particles guided to the particle growing unit 400. The upper size of the charged primary particles guided to the growing unit 400 may be selected such that most of the particles of the aerosol flow A3 are singly charged, and only a small number of particles are multiply charged.

Particle charging phenomena will be discussed in more detail in the context of Figs. 14a and 14b.

Ideally, all charged primary particles guided to the particle growing unit 400 have mobility diameters smaller than a predetermined maximum mobility diameter D _P,MAX -

In practice, the upper size or a lower size of an aerosol particle population may also be interpreted to be a statistical quantity. The upper size may be defined based on the number of particles. For example, at least 95% of the charged primary particles guided to the growing unit 400 may have mobility diameters smaller than a predetermined maximum mobility diameter D _P,MAX - The maximum diameter D _P,MAX may be e.g. 40 nm, advantageously 25 nm, and preferably 15 nm. Said percentage may also be e.g. 99%, or even 99.9% instead of 95%, in order to improve accuracy. The growing unit 400 may provide an output aerosol flow A5, which comprises larger singly charged particles and only a negligible amount of multiply charged and neutral particles.

The aerosol A3 provided by the separating unit 300 may contain a major fraction of singly charged positive charged particles. Alternatively, the aerosol A3 may contain a major fraction of singly charged negative particles. Consequently, the output aerosol A5 may contain a major fraction of singly charged positive particles, or a major fraction of singly charged negative particles.

The upper size of the charged primary particles guided to the growing unit 400 may be limited e.g. in the following ways:

- the upper size of primary particles guided to the charging unit 200 may be limited by using a primary particle generator 100 which is arranged to operate such that primary particles generated by the primary particle generator 100 have mobility diameters smaller than a predetermined maximum mobility diameter D _P,MAX>

- the upper size of primary particles guided to the charging unit 200 may be limited by separating large particles away from an aerosol flow AV provided by the primary particle generator 100 (see Fig. 20),

- the upper size of the primary particles may be limited prior to charging,

- large charged particles may be eliminated from the aerosol flow A2 provided by the charging unit 200, and/or

- large charged particles may be eliminated from the aerosol flow A3 provided by the separating unit 300.

The percentage of particles smaller (or greater) than a predetermined size may be verified e.g by using a Scanning Mobility Particle Sizer

(SMPS). The scanning mobility particle sizer comprises a differential mobility analyzer (DMA) arranged to provide a monodisperse distribution by classifying particles according to electrical mobility, and a condensation particle counter arranged to determine particle number and/or particle concentration associated with said monodisperse distribution. The differential mobility analyzer comprises electrodes connected to a variable voltage. The electrodes cause an electric field, which deflects particles through a laminar gas layer towards an opening such that only particles within a predetermined mobility range are guided through the opening to the condensation particle counter, The center of said mobility range is determined by the magnitude of the voltage. The size distribution of the aerosol may be determined by changing the voltage and by recording the particle number and/or concentration associated with each voltage.

Aerosol particles are distinguished from molecular clusters e.g. by their size. In this context, the geometrical size, i.e. the greatest dimension of aerosol "particles" may range e.g. from 2 n m to 1 00 μ m . The geometrical size of "molecular clusters" may be smaller than 2 nm, respectively. The geometrical size may be measured e.g. by an electron microscope.

In some cases, an aerosol flow A2 guided out of the charging unit 200 may contain gas ions or charged molecular clusters, in addition to the primary particles and the carrier gas.

However, the test aerosol A5 should not contain a significant number of gas ions or charged molecular clusters, because the presence of gas ions and/or charged molecular clusters may cause an error in the subsequent testing or calibration.

Ions and molecular clusters may be eliminated from the aerosol A3 or A5 e.g. by using a separating unit 300, which is arranged to separate particles P1 based on electrical mobility before the aerosol A3 is guided to the particle growing unit 400. Gas ions and charged molecular clusters may also be separated from the aerosol A3 or A5 e.g. by using an electric field or by using a diffusion collector.

Alternatively, the upper size may also be defined based on the total mass of particles. For example, at least 95% of the mass of the charged primary particles guided to the growing unit 400 may be contained in particles having mobility diameter smaller than a predetermined maximum mobility diameter D _P,MAX - The maximum diameter D _P,MAX may be e.g. 40 nm, advantageously 25 nm , and preferably 15 nm. Said percentage may be e.g. 99% instead of 95%, in order to improve accuracy.

Fig. 5b shows method steps corresponding to the apparatus of Fig. 5a. Primary particles may be generated in a primary particle generating step 101. The primary particles may be charged in a charging step 201. Positive or negative primary particles may be separated in a separating step 301. The size of the particles may be increased on a growing step 401. The generated test aerosol A5 may be outputted in an outputting step 501.

Referring to Fig. 6, the primary particle generator 100 may be based on evaporation and condensation. An amount of solid or liquid substance 130 may be heated by one or more heating elements 120 in order to produce vapor e.g. by sublimation. The vapor may be mixed with a carrier gas GO. The vapor and the carrier gas GO may be guided through a duct 110. Primary particles PO may be formed by condensation of the vapor in a cooling zone 1 15 in order to provide an aerosol flow A1 comprising the primary particles PO. The primary particles PO may comprise or consist of e.g. a metal, an oxide, or carbon. The substance 130 may be e.g. sodium chloride, and the gas GO may be e.g. air. Advantageously, the material of the primary particles PO has a low vapor pressure in the temperature range 250 - 350 K in order to provide stable particles. The material of the primary particles PO may be selected such that the vapor pressure of said material is smaller than or equal to 100 Pa (1 mbar) when the particles pass through the growing unit 400.

A particle generator based on evaporation and condensation has been described e.g. in an article "Generation of monodisperse Ag- and NaCI aerosols with particle diameters between 2 and 300 nm", by H.G.Scheibel, and J.Porstendόrfer, in Journal of Aerosol Science Vol. 14 pp. 113-126, 1983.

A particle generator based on spark atomization has been described e.g. in an article "Aerosol generation by spark source discharge, by S.Schwyn, E.Garwin", and A. Schmidt-Ott, in Journal of Aerosol Science Vol. 19 pp. 639-642, 1988.

Referring to Fig. 7, the primary particle generator 100 may also be based on nebulization and drying. A solution LO comprises a low vapor pressure substance and a high vapor pressure solvent. The solution LO may be nebulized by an ultrasonic or pneumatic nebulizer 140. Nebulized droplets SO may be guided along a duct 110 and heated by heating elements 120 so as to evaporate the solvent. The residual substance may form primary particles PO. In other words, the primary particles PO may be provided by drying the droplets SO. The droplets SO may be dryed in a drying zone 117 of the duct 110.

A particle generator based on nebulization and evaporation has been disclosed e.g. in an article "Operating characteristics of some compressed air nebulizers" by TT. Mercer, M.I.Tillery, and H.Y.Chow in American Industrial Hygiene Association Journal, Vol. 29 pp. 66-78, 1968.

The primary particles PO may comprise or consist of a metal, oxide, or carbon. Advantageously, the material of the primary particles PO has a low vapor pressure in the temperature range 250 - 350 K. The material of the primary particles PO may be selected such that the vapor pressure of said material is smaller than or equal to 100 Pa (1 mbar) when it passes through the growing unit 400. Primary particles PO may also be generated by chemical reactions, e.g. by oxidizing hydrocarbons or silicon tetrachloride (not shown).

Referring to Figs. 8 and 9, the primary particles PO may be charged e.g. by exposing them to ionized gas containing positive and negative gas ions. Charging by both positive and negative ions is known as bipolar charging. Ionized gas may be quasi-neutral, i.e. the total charge of all positive ions may be substantially equal to the total charge of all negative ions. Ionized gas having positive and negative ions may be generated e.g . by exposing a neutral gas to ionizing radiation generated e.g. by a radioactive source or an X-ray source. Ionized gas may also be generated by a corona discharge.

Referring to Fig. 8, the primary particles PO of the aerosol flow A1 may be charged by exposing the particles PO to ions generated by ionizing radiation hv. An aerosol flow A2 comprising positive P1 and negative PN1 particles may be provided.

A charger 200 may comprise a duct 210 for guiding the aerosol flows A1 , A2, and a radiation source 220 to provide ionizing radiation. The duct 210 may comprise an enlarged portion for increasing residence time τ of particles in an ion cloud. The enlarged portion may be a chamber.

When ions are generated by a radiation source, then the ion concentration nio _N may depend on the activity of the radiation source, on the flow rate of the gas exposed to the ionizing radiation, and on the dimensions of the charging device. The activity of an X-ray tube may be controlled by adjusting operating voltage and/or current. When the radiation source is a radiating isotope element, the effective activity of the radiation source may be controlled by using absorbers and/or radiation apertures. An isotope element may comprise e.g. a radiating isotope of krypton (Kr-85), americium (Am-241 ), or nickel (Ni-63) Referring to Fig. 9, primary particles PO may be charged by exposing primary particles to ions generated by a corona discharge.

A charger 200 may comprise electrodes 250, 252 for generating a corona discharge. The electrodes 250, 252 may be connected to a voltage supply 260, which provides a corona voltage V _c . The electrode 250 may be supported by an insulator 251. The primary particles PO may be guided along a duct 210, which contains ions generated by the electrodes 250, 252.

In case of a corona charger, the ion concentration nio _N may depend on the geometry of the electrodes, voltage, pressure, type of the gas, and on the spatial location between the electrodes 250, 252.

The charger 200 may further comprise one or more flow guiding structures (not shown) arranged to prevent access of primary particles PO to the vicinity of the first electrode 250, i.e. to keep the primary particles PO away from the zone where electrical breakdown occurs.

A corona charger may be arranged to operate as a unipolar charger or as a bipolar charger. A charger comprising a single needle electrode 250 or wire electrode 250 typically acts as a unipolar charger, wherein the majority of generated gas ions are either positive or negative. Substantially equal concentrations of positive and negative gas ions for bipolar charging may be provided e.g. by using several electrode pairs.

Referring to Fig. 10, a particle separating unit 300 may be arranged to separate particles based on electrical mobility of the primary particles. In other words, the particle separating unit 300 may be an electrical mobility classifier arranged to classify particles according to their electrical mobility.

Electrodes 330, 340 may be arranged to create an electric field E3.

Particles PO, P1 , PN1 of the aerosol flow A2 may be introduced to the electric field E3 along a first duct 310. PO denotes neutral particles, P1 denotes singly charged positive particles, and PN 1 denotes singly charged negative particles.

Particle-free gas G2 may be introduced to the electric field E3 by a second duct 320 such that charged particles P1 or PN 1 , which are deflected towards an opening 350 by the electric field E3 have to pass through a layer of the gas G2. Said layer of gas G2 may contain a negligible amount of neutral particles PO and a negligible amount of particles which are not driven by the electric field E3. The gas G2 may be substantially particle-free when exits from the second duct 320 to the electric field E3. The layer of the gas G2 may be substantially laminar at least in the vicinity of the opening 350.

Only particles P1 whose electrical mobility is in a predetermined range are deflected by the electric field E3 to the vicinity of the particle- extracting opening 350. PTH 1 denotes the path of singly charged positive particles P1 , whose electrical mobility is in the predetermined range. Positive particles P1 in the vicinity of the opening 350 may be carried to a third duct 355 by a gas stream to form a separated aerosol A3.

Neutral particles PO are carried along the gas flow G2 without being deflected towards the opening 350. Negative particles PN 1 are deflected away from the opening 350. The remaining particles PO and PN 1 may be carried along a fourth duct 356 to form an exhaust aerosol or exhaust flow EX3.

Alternatively, negative particles PN 1 may be extracted through the opening 350 by reversing the polarities of the electrodes 330, 340.

An upper limit of the concentration of the charged particles P1 separated by an electric mobility classifier 300 may be e.g. in the order of 10 ⁵ particles/cm ³ . A first electrode 330 and a second electrode 340 may be coupled to a voltage supply 360, which provides a field voltage V _M . In case of planar and parallel electrodes 330, 340, the electric field E3 is given by:

E3 = ^- (3)

L3

where L3 denotes the distance between the electrodes 330, 340.

The electrodes 33, 340 and the ducts 310, 320, 355, 356 may also be arranged in a concentric manner. In particular, the electrodes may be arranged coaxially.

Electrical mobility Z is defined as:

v _TE

Z = (4)

where v _TE denotes terminal electrostatic velocity and E denotes electric field.

The electrical mobility of a particle may be calculated by the equation:

Z = J (5)

3πηD _P

where j e denotes the charge of the particle, j denotes an integer (-1 , 0, 1 , 2, 3, ...), e denotes elementary charge, C _c denotes the Cunningham slip correction factor, η denotes viscosity of the carrying gas, and D _P denotes the mobility diameter of the particle.

A mechanical mobility b _P of a particle is defined as:

K =^- (6) where F _P denotes a driving force acting on said particle and v _P denotes a terminal velocity of said particle caused by said driving force F _P .

The mobility diameter D _P of a particle is a diameter of a sphere having the same mechanical mobility b _P as the particle under consideration.

According to eq. (5), the separating unit 300 of Fig. 10 may classify particles based on size and based on the charge j e.

Singly charged positive particles P1 may impinge on the electrode 340 if the electrical mobility Z of said particles is greater than an upper limit Z _MAX> i.e. when the diameter D _P of said particles is smaller than a lower limit D ₃ . Singly charged positive particles P1 may impinge on the wall of the duct 356 if the electrical mobility Z of said particles is smaller than a lower limit Z _{MIN >} i.e. when the diameter D _P of said particles is greater than an upper limit D ₄ . Thus, only singly charged particles P1 , whose diameter D _P is greater than or equal to D ₃ and smaller than or equal to D ₄ are separated into the duct 355. The limits D3 and D4 define a first pass band of the separating unit 300 (see Figs. 18a-18c).

The position and the width of the first pass band with respect to the size distribution of particles guided into the separating unit 300 may be adjusted e.g. by selecting the position and the dimensions of the opening 350, by changing the flow velocity of the gas G2 and/or by changing the magnitude of the electric field E3.

A particle separating device has been described e.g. in an article "Aerosol classification by electric mobility apparatus, theory and application", E.O.Knutson, K.T.Whitby, Journal of Aerosol Science, Vol. 6, pp. 443-451 , 1975.

The terms "electrostatic classifier" and "differential mobility analyzer" (DMA) are often used in this context. An "electrostatic classifier" comprises a charger and an electrical mobility classifier. In the literature, the term "differential mobility analyzer" is often used to mean the same thing as the electrical mobility classifier. However, it is emphasized that an electrical mobility classifier does not need to comprise a charger and/or a charge detector.

Referring to Fig. 11 , a particle growing unit 400 may be based on e.g. condensation or chemical deposition of further material on the charged primary particles P1.

In case of condensation, a heated gas stream G4 may be mixed with the aerosol stream A3. A portion of the duct 410 and/or the gas stream G4 may be heated by one or more heating elements 420. The mixture may guided along a duct 410. The gas flow G4 contains a vapor (or vapors) which may condense on the primary particles P1. The gas stream G4 may contain e.g. oil vapor, hydrocarbon vapor, or salt vapor. The temperature of the gas stream G4 and/or the temperature of the mixture may be selected such that the primary particles P1 are entrained into the vapor, which is in the saturated state or in a nearly saturated state. The temperature of the vapor and the temperature of the particles may be reduced by guiding the mixture of the aerosol A3 and the saturated vapor to a cooling zone 415. Upon cooling, the vapor may condense on the primary particles P1 thereby providing larger singly charged particles VLP1 (or VLPN1 ), without generating a significant amount of new particles by nucleation.

The primary particles P1 are enclosed within the larger particles VLP1. In other words, the primary particles P1 act as nuclei for the larger particles VLP1.

Nucleation of new particles may substantially be avoided or minimized by selecting the temperature of the cooling zone 415. Nucleation of new particles would provide new neutral particles, which in turn could reduce calibration accuracy.

The size of the produced larger particles VLP1 may be selected e.g. by adjusting the temperature of the vapor, by adjusting partial pressure of the vapor, by changing the chemical composition of the vapor and/or by adjusting the flow velocity through the duct 410. The temperature of the vapor may be adjusted e.g. by changing the temperature of the heating elements 420.

The size of the produced larger particles VLP1 may also be controlled by introducing dilution gas to the duct 410, in particular to the cooling zone 415 of the duct 410.

Two or more growing units 400 may be connected one after another in order to further increase the size of the particles VLP1.

A particle growing device 400 based on condensation has been described e.g. in a patent US 5,500,027.

The particle growing unit 400 may be arranged to increase the size of the primary particles P1 substantially without agglomerating particles

P1. Agglomeration may be minimized e.g. by reducing the time period during which the primary particles are small, and/or by limiting the particle number density. The apparatus 500 may be arranged to operate such that the particle density is e.g. smaller than or equal to 10 ⁵ particles/cm ³ .

The growing unit 400 may be arranged to operate such that when a larger particle VLP1 is provided by increasing the size of a primary particle P1 , at least one dimension of said larger particle VLP1 may be greater than or equal to two times the corresponding dimension of said primary particle P1. At least one dimension of the larger particle VLP1 may even be greater than ten times the corresponding dimension of the primary particle P1. The mass median diameter of the particles VLP1 may be e.g. in the range of 0.1 to 1 μm, wherein the mass median diameter of the primary particles may be e.g. in the range of 10 to 40 nm. The mass median diameter of the particles VLP1 may even be in the range of 1 to 10 μm.

The larger singly charged particles VLP1 form an aerosol A4 together with a carrier gas. The composition of the carrier gas depends mainly on the composition of the gas flows G4 and G2 (Fig. 10). The aerosol A4 may be used as an output aerosol A5 of the test aerosol generating apparatus 500 (see Fig. 5a).

The size of the primary particles P1 may also be increased by chemical deposition. The vapor or vapors of the gas stream G4 may react with the surface of the primary particles P1 so that further material may be deposited on the surfaces of the primary particles P1.

For example, the material of the primary particles P1 may be e.g. silicon dioxide or titanium dioxide. The gas stream G4 may contain e.g. tetraethylortosilicate (TEOS) vapor or titanium tetraisopropoxide (TTIP) vapor.

Fig. 12 illustrates the method steps of Fig. 5b. Bipolar charging of primary particles PO provides an aerosol A2, which comprises singly charged positive primary particles P1 and singly charged negative primary particles PN1. A fraction of the primary particles PO may remain neutral.

Singly charged positive primary particles P1 may be separated from the neutral particles PO and from the negative particles PN 1 based on electrical mobility.

The size of the singly charged positive primary particles P1 may be subsequently increased e.g. by condensation or chemical deposition so as to provide larger singly charged particles VLP1.

Fig. 13 shows a comparative example, where the charged aerosol A2 comprises a significant fraction of large particles PLO, PL1 , PL2, PL3, and multiply charged particles PN2, P2, PL2, PN3, P3, PL3.

If singly charged positive particles P1 are separated based on electrical mobility, then the separated aerosol A3 may further comprise larger doubly charged positive particles PL2 (See eq. (5) above). The size of the doubly charged primary particles PL2 may be further increased in the growing step. Thus, the output aerosol A5 may comprise larger doubly charged positive particles VLP2, in addition to the larger singly charged particles VLP1. Multiply charged particles may reduce accuracy in the subsequent testing or calibration.

The situation of Fig. 13 may be avoided if the primary particles have a narrow size distribution and/or if it is ensured that only a negligible fraction of the primary particles are multiply charged.

The solid curve of Fig. 14a shows, by way of example, the relative fraction n-|/n _τ of singly charged particles as a function of particle mobility diameter D _P in a situation, where particles have been charged by exposing them to bipolar ions generated by X-ray radiation. The dashed curve of Fig. 14a, shows, by way of example, the relative fraction n ₂ /n _τ of doubly charged particles as a function of particle mobility diameter D _P in said charging situation. n ₊₁ denotes the number of singly charged particles in an unit volume, n ₊₂ denotes the number of doubly charged particles in said unit volume, and n _τ denotes the number of all aerosol particles in said unit volume.

Kinetics of particle charging has been discussed e.g. in US 2005/0180543, and in an article "Ion-aerosol attachment coefficients and steady-state charge distribution on aerosols in a bipolar environment" by W. H. Hoppel, G.Frick in Aerosol Science and Technology, Vol. 5, pp. 1-21 , 1986.

Fig. 14b shows the ratio of the number n ₊₂ of doubly charged particles to the number n ₊₁ of singly charged particles, for the case of Fig. 14a.

It may be noticed that only a small fraction of the particles are multiply charged when the mobility diameter of the charged particles is kept below a predetermined limit D _P,MAX -

If the predetermined limit D _P,MAX is selected to be e.g. 60 nm, this means in case of Fig. 14b that approximately up to 5% of the particles are doubly charged. Thus, the doubly charged particles may carry approximately 10% of the total charge. This in turn means that the corresponding error in the particle number is in the order of 10%.

If a better accuracy is desired, then a lower predetermined limit D _P,MAX should be selected. If the predetermined limit D _P,MAX is selected to be e.g. 30 nm, this means in case of Fig. 14b that less than 1 % of the particles are doubly charged.

In order to ensure a sufficient accuracy, the predetermined upper limit Dp. _MA x may be e.g. 40 nm, advantageously 25 nm, and preferably 15 nm.

The relative fractions of neutral, singly charged, and multiply charged particles may depend on the concentration nio _N of ions, and on a residence time τ during which the particles are exposed to said ion concentration. In particular, the population of the different charged states of the particles may depend on the value of the product nι _O ι\rτ. If the particles are initially neutral, the relative fraction of multiply charged particles typically increases with an increasing value of until a steady state is reached.

To the first approximation, the relative fraction of charged particles may be estimated by the Boltzmann equation:

where n _j denotes the number of j-charged particles in a unit volume. Each j-charged particle has a charge j e. n ₀ denotes the number of neutral particles in said unit volume. ε ₀ denotes vacuum permittivity (8.85-10 ^"12 J ^~1 C ² rτϊ ¹ ), k denotes Boltzmann constant (1.38-10 ^"23 JK ^"1 ), and T denotes temperature expressed in kelvins (K). In practice, equation (7) tends to underestimate the relative fraction of charged particles smaller than 50 nm. However, eq. (7) may still be used to qualitatively describe how the relative fraction of multiply charged particles depends on the particle size and temperature. It may be noticed from eq. (7) that the number of doubly charged particles (j=2) may be reduced by decreasing the diameter D _P of the particles and/or by decreasing the temperature T.

However, limiting the upper size of the particles may be technically easier than charging particles in cryogenic temperatures.

Particles, whose diameter is smaller than or equal to 20 nm, are typically neutral or singly charged in the Boltzmann equilibrium when the temperature T is in the range of 250 to 350 K. Some particles greater than 20 nm may be multiply charged. The relative fraction of multiply charged larger particles may be controlled, to some extent, by selecting the value of riio _N 'τ according to the size of the larger particles. However, when the relative fraction of the larger particles is small, there is no need to accurately measure or adjust the value of the product riioN'τ.

Thus, the predetermined upper diameter D _P,MAX of the particles during the charging may be selected according to:

- a predetermined maximum relative fraction of multiply charged particles, and

- the temperature T of the particles during the charging process.

The predetermined maximum relative fraction of multiply charged particles may be selected according to the accuracy desired for the intended use of the test aerosol.

The temperature T may be e.g. in the range of 200 K to 400 K, preferably in the range of 250 K to 350 K, which is the temperature range typically encountered when studying environmental reactions. In particular, the temperature T during the charging may be in the vicinity of the normal room temperature, i.e. in the range of 290 K - 310 K. The gas and the particles may be substantially in the same temperature during the charging.

Referring back to Fig. 14a, the relative fraction n-|/n _τ of singly charged particles may be too low for very small particles. The particles are charged by gas ions impinging on neutral primary particles PO. If the primary particles PO are very small, the collision cross section for the charging process may also be very small. In other words, the charging rate may be too low. The mobility diameter D _P of the primary particles PO may be kept above a predetermined limit in order to ensure efficient charging. The primary particle generator 100 may be arranged such that e.g. at least 90% of the primary particles PO generated by the generator 100 may have mobility diameters greater than equal to a predetermined limit D _P>MI N- The limit D _P , _MI N may be e.g. 4 nm, advantageously 7 nm.

Referring to Figs. 15 and 5a, a part of larger singly charged primary particles PL1 may be eliminated from the aerosol A3 based on electrical mobility when the separating unit 300 is positioned between the charging unit 200 and the growing unit 400. Thus, the size distribution of the aerosol A3 provided by the separating unit 300 may be substantially narrower than the size distribution of the singly charged aerosol A2 provided by the charging unit 200.

If the primary particles guided to the growing unit 400 have a narrow size distribution, this may, in turn, facilitate providing output aerosol A5 which has a narrow size distribution. Furthermore, when the primary particles P1 introduced to the growing unit 400 have a narrow size distribution, the probability of clustering (agglomeration) in the growing unit 400 may be minimized.

Referring to Figs. 16 and 17, the separating unit 300 may also be positioned after the growing unit 400. Thus, neutral primary particles PO and charged primary particles P1 , PN1 may be guided to the growing unit 300 before separating the positive (or negative) primary particles. Positive or negative particles may be separated from the output A4 of the growing unit 400 by the separating unit 300 in order to provide the test aerosol A5.

In certain operating conditions, it is possible that a small number of new neutral particles may be generated in the growing unit 400 by homogeneous nucleation. An advantage associated with positioning the separating unit 300 after the growing unit 400 is that the new neutral particles may be eliminated from the test aerosol A5.

Positioning of the pass bands of the separating unit 300 with respect to the particle size distribution will now be discussed with reference to Figs. 18a-18c.

Referring to Fig. 18a, a particle size distribution dn _T /d _P of an aerosol guided to the separating unit 300 may reach its maximum value (max) at a mobility diameter D _PEA κ- Di denotes a mobility diameter value where the distribution reaches e.g. 1 % of the maximum value. W _DES c indicates the width of a descending slope DESS. W _DES c is equal to the difference between D ₁ and D _PEA κ-

An electrical mobility classifier may be arranged to separate singly charged particles, whose mobility diameter D _P is between values D ₃ and D ₄ . The values D ₃ and D ₄ define the first passband PB1 of the electrical mobility classifier. The first pass band PB1 may be adjusted to substantially coincide with the peak of the particle size distribution, i.e. D _PEA κ may be between the values D ₃ and D ₄ .

Doubly charged particles P2 whose mobility diameter is equal to a value D ₂ may have the same electrical mobility Z as singly charged particles P1 whose mobility diameter is equal to D ₃ . The electrical mobility classifier cannot separate singly charged particles P1 from doubly charged particles P2 if they have the same electrical mobility Z. Consequently, the electrical mobility classifier also has a second pass band PB2 for doubly charged particles P2, whose mobility diameter D _P is greater than or equal to a value D ₂ and smaller than or equal to a value D ₇ . The portion of the distribution within the first pass band PB1 is indicated by a diagonal hatch pattern. The portion of the distribution within the second pass band PB2 is indicated by a cross hatch pattern. It is emphasized that the first pass band PB1 applies only to singly charged particles, and the second pass band PB2 applies only to doubly charged particles.

In case of Figs. 18a and 18b, the size distribution has a wide descending slope DESS, which means that the value of D ₂ is smaller than or equal to the value of D ₁ when the first passband PB1 is set to coincide with the peak D _PEA κ-

In other words, the size distribution has a wide descending slope if the second pass band PB2 of an electrical mobility classifier for doubly charged particles substantially overlaps the descending slope DESS of the particle size distribution in a situation where the first pass band PB1 of the electrical mobility classifier for singly charged particles P1 coincides with the peak D _PEA κ of the size distribution.

The electrical mobility Z is a function of the charge j e and the mobility diameter D _P (see eq. (5)). Singly charged smaller particles cannot be separated from doubly charged larger particles based on electrical mobility if the particles have the same electrical mobility Z. This condition may be expressed mathematically as follows:

Z(J = 2,D _P = D _{P 1} ) = Z(J = I ₉ Dp = Dp ₁ ) (8a)

Eq. (8a) may also be expressed in the following simplified form:

2C _C (Z ⁾ _Λ2 ) C _c (D _Pλ )

(8b)

D P, 2 D Pλ

where the slip correction factor C _C (D _P ) is a function of the mobility diameter D _P . D _{P 1} denotes the mobility diameter of a singly charged smaller particle, and D _{P 2} denotes the mobility diameter of a doubly charged larger particle.

The value of D ₂ may be calculated by using eq. (8a) or (8b) when the value of D ₃ is known, or vice versa. Also the value of D ₇ may be calculated from D ₄ , respectively.

If desired, the position D _PEA κ of the peak of the size distribution may be determined e.g. by measurements and/or by theoretical analysis. For example, D _PEA κ may be estimated by using the apparatus 500 itself. For example, the position of the first pass band PB1 may be varied and the measurement results provided by the charge detector 600 may be monitored as a function of the position of the first pass band PB1 , wherein the charge detector 600 may be arranged to monitor the charges carried by the particles of the test aerosol A5.

Referring to Fig. 18b, the size distribution may reach e.g. 50% of its maximum value at points D ₅ and D ₆ . The limits D ₅ and D ₆ may define a peak region TOP ₅₀ of the size distribution. For example, the pass band PB1 may be set to overlap the peak region TOP ₅₀ , i.e. at least one of the values D ₃ and D ₄ may be located between the values D ₅ and D ₆ .

Thanks to limiting the upper size D _P,MAX of the charged primary particles, there is a considerable freedom to set the location of the first pass band PB1 with respect to the peak. In fact, it may be sufficient if the first pass band PB1 is set to any position where it overlaps the particle size distribution.

Thanks to limiting the upper size of the charged primary particles guided to the growing unit 400, the relative fraction of multiply charged particles may be kept so low that it is not necessary to adjust the size range of the aerosol A3 provided by the separating unit 300 according to the size distribution of particles guided to the separating unit 300, or at least there is no need to fine-tune the size range of the aerosol A3 according to the size distribution guided to the separating unit 300. It is not necessary to measure the position D _PEA κ of the peak of the size distribution, respectively.

In an embodiment, the field voltage V _M of the separating unit 300, the dimensions of the opening 350 (Fig. 10), and the position of the opening 350 may even be pre-adjusted at a factory to predetermined values, wherein the same predetermined values can be used when changing the material of the primary particles PO generated by the test particle generator 100, when changing the material of the test particles VLP1 , and/or when changing the size of the test particles VLP1. In an embodiment, the separating unit 300 may even be sealed such that a user can not alter the voltage V _M or dimensions of the opening 350 without breaking the seals.

On the other hand, the position of the first pass band PB1 of the separating unit 300 may be adjusted according to the size distribution such that the pass band PB2 for doubly charged particles is outside the particle size distribution. In that case the upper size of particles guided to the charging unit 200 may be slightly larger than the upper size D _P,MA x of the primary particles guided to the growing unit 400. This arrangement may allow a greater freedom to select the material of the primary particles PO, P1 and/or to select the operating principle of the primary particle generator 100.

However, if the pass band PB1 does not coincide with the peak, then the fraction of particles which can be utilized is smaller than in a case where the pass band PB1 is positioned at the peak D _P . Also the temporal stability of the particle concentration of the test aerosol A5 may be better when the pass band PB1 substantially coincides with the peak of the size distribution, when compared with a situation where the pass band PB1 is tuned to the side of the size distribution.

Fig. 18c shows a size distribution which has a narrow descending slope. The limits D ₁ and D ₂ are in a different order than in Fig. 18a, and the pass band PB2 for doubly charged particles P2 is outside the particle size distribution even when the pass band PB1 coincides with the peak of the size distribution.

Because the size distribution of Fig. 18c has a narrow descending slope, singly charged particles can be completely separated from doubly charged particles based on electrical mobility even when the first pass band PB1 substantially coincides with the peak of the size distribution. However, with certain materials it may be difficult or even impossible to provide a size distribution, which has a sufficiently narrow descending slope.

Limiting the upper size D _P,MAX of the charged primary particles guided to the growing unit 400 may allow using a size distribution, which has a wide descending slope.

Limiting the upper size D _{P, MAX} of the charged primary particles guided to the growing unit 400 may allow allow considerable freedom to select the material of the primary particles PO, P1 and/or a considerable freedom to select operating parameters of the primary particle generator 100.

Thanks to limiting the upper size D _P,MA χ θf the charged primary particles guided to the growing unit 400, it is not necessary to adjust the position of the pass band PB1 of the separating unit 300 according to the size distribution of particles guided to the separating unit 300. This applies when the separating unit 300 is positioned before the growing unit 400 (Fig. 5a) as well as when the separating unit 300 is positioned after the growing unit 400 (Fig. 16).

Thanks to limiting the upper size D _P,MAX of the charged primary particles guided to the growing unit 400, the separating unit 300 does not need to have a capability of classifying particles according to size. The first pass band PB1 of the separating unit 300 may even be so wide that it overlaps the particle size distribution in most foreseeable applications, without a need to tune the field voltage V _M or the position of the opening 350. The first pass band PB1 of the separating unit 300 for singly charged particles P1 may be so wide that the pass band PB1 for singly charged particles may overlap the pass band PB2 for doubly charged particles (i.e. the limit D ₄ may be greater than the limit D ₂ ). The width of the first pass band PB1 may be greater than the width W _DES c of the descending slope DESS.

Thanks to limiting the upper size D _P,MA χ θf the charged primary particles guided to the growing unit 400, the size of the particles of the test aerosol A5 does not need to match with the size distribution guided to the separating unit 300. The size and/or material of the particles of the test aerosol A5 may be changed without a need to re-adjust the separating unit 300.

Thanks to limiting the upper size D _P,MA χ θf the charged primary particles guided to the growing unit 400, the size distribution of the test aerosol A5 may have a wide descending slope DESS.

Referring to Fig. 19, a neutralizer 250 may be arranged to substantially eliminate electrical charges of the primary particles before introducing them into the charging unit 200.

The neutralizer 250 may be based e.g. on ions generated by radiation or corona discharge.

However, if the ion concentration nio _N in the charging unit 200 is arranged to be high enough and/or if the residence time τ in the charging unit 200 is arranged to be high enough, it is not necessary to use the neutralizer 250 even if a significant amount of the particles generated by the particle generator 100 would be multiply charged.

A high ion concentration nio _N and/or a long residence time τ may, in fact, convert multiply charged small particles into neutral or singly charged particles. The ion concentration nio _N prevailing in the charging unit 200 may be so high that most of multiply charged primary particles P2 are converted into neutral PO or singly charged particles P1. The ion concentration nio _N may be e.g. in the range of 10 ⁸ - 10 ¹¹ ions/cm ³ . Referring to Fig. 20, a size selecting unit 150 may be optionally arranged to define an upper size and/or a lower size of the primary particles introduced into the particle charging unit 200.

The size selecting unit 150 may be e.g. a cyclone, an impactor, in particular a virtual impactor, or an electrical mobility classifier.

Referring back to Figs 1 and 3b, the test aerosol A5 provided by the test aerosol generating apparatus 500 may be used for testing and/or calibrating aerosol measuring instruments 900 or for studying gas- particle interaction. Gas-particle interactions may be related e.g. to formation of smog and acid rain. The measuring instrument 900 may be e.g. an impactor, or an optical particle counting device. The measuring instrument may be e.g. an optical particle counter (OPC), a laser particle counter (LPC), a condensation particle counter (CPC), a scanning mobility sizer (SMPS), a differential mobility sizer (DMPS), or an aerodynamic particle sizer (APS).

The measuring instrument 900 may be suitable for studying particles e.g. in the exhaust gases of engines, in industrial processes, or in power plant flue gases. The generated test aerosol A5 may be used for testing instruments and devices for clean rooms, semiconductor manufacturing processes, nanotechnology, or aerosol technology. The test aerosol A5 may be used for testing gas cleaning devices, e.g. mechanical filters or electrostatic precipitators.

The efficiency of an aerosol sampling probe or aerosol sampling line may be tested. For example, the test aerosol A5 may be used for testing whether a significant number of particles are lost in a sampling line.

Also in less demanding cases, the concentration of multiply charged particles may be reduced by charging the particles before increasing the size. A generated test aerosol may be used for studying e.g. gas- particle interactions even if it would contain e.g. up to 10% multiply charged particles.

The use of the method and apparatus according the present invention allows producing singly charged test aerosol having a large mass median diameter. The mass median diameter may be e.g. in the range of 0.1 μm to 1 μm, in the range of 1 μm to 2 μm, or in the range of 2 to 10 μm. There is a considerable freedom to change the mass median diameter and/or the material of the generated test aerosol.

The separating unit may be e.g. an electrical mobility classifier arranged to select a predetermined size, but it does not need to be. The first pass band PB1 of the separating unit may be very wide. It is sufficient if the separating unit is capable of separating positive particles from neutral and negative particles, or if the separating unit is capable of separating negative particles from neutral and positive particles.

In certain cases, the apparatus 500 does not even need to contain means for separating charged particles from neutral particles and from particles of opposite polarity. For example, the generated test aerosol

A5 may be used for testing an external instrument, wherein said instrument is itself capable of distinguishing charged particles from neutral particles and from particles of opposite polarity. Said instrument may be e.g. an electrical mobility classifier. If the electrical mobility classifier contains an internal means for charging, said means for charging may be switched off during said testing.

For the person skilled in the art, it will be clear that modifications and variations of the devices and the methods according to the present invention are perceivable. The particular embodiments described above with reference to the accompanying drawings are illustrative only and not meant to limit the scope of the invention, which is defined by the appended claims.