WO2012022844A1

WO2012022844A1 - Electrical impactor

Info

Publication number: WO2012022844A1
Application number: PCT/FI2011/050731
Authority: WO
Inventors: Kauko Janka
Original assignee: Individual
Current assignee: Individual
Priority date: 2010-08-20
Filing date: 2011-08-19
Publication date: 2012-02-23
Anticipated expiration: 2013-02-20
Also published as: WO2012022843A1; WO2012022842A1; EP2606344A1; FI20110067A0; FI20110066A0; FI20110065A0; FIU20100360U0

Abstract

Electrical impactor (1) where the particles entering the impactor (1) are classified by inertial separation on at least one stage (4). Impactor (1) comprises a functional component (22) which switches or modulates the operational mode of the electrical discharge unit (20) or/and the ion trap. Process for particle measurement where the particles are classified by inertial separation, Unwanted particle deposition on the upper stages used for inertial separation is minimized by using switching or modulating either the electrical source used for particle charging or the ion trap used to trap free ions and ultrafine particles.

Description

Electrical impactor

Field of invention

The present invention relates to an electrical apparatus for measuring particles according to the preamble of claim 1 and specifically to an electrical impactor comprising a component which switches or modulates at least a parameter affecting the measurement mode of the electrical impactor. The present invention further relates to a process for measuring particles according to the preamble of claim 12.

Background of the invention

There is a constant increase in the demand for real-time particle measurement. Such demand exists e.g. in the measurement of atmospheric aerosols, in particle filter development and within different high-temperature processes, such as combustion processes. Especially the real-time exhaust control of combustion engines, such as vehicles, requires reliable and non- expensive particle monitoring, there are also various applications where also information on the electrical charge of the particles is required, preferably as a function of the particle size. Such applications include e.g. research and development of electrostatic precipitators as well as different atomizers and nebulizers.

Particle measurements are frequently carried out by cascade impactors, where particulate matter is withdrawn (preferably isokinetically) from a source and segregated by size, Cascade impactors use the principle of inertial separation to size segregate particle samples from a particle laden gas stream. Conventional cascade impactors cannot be used in real-time particle size distribution (PSD) measurement and especially in measuring changes in PSD in real time. In the Electrical Low Pressure Impactor (ELPI™, Dekati Oy, Finland) The particles are first charged into a known charge level in the corona charger. After charging, the particles enter a cascade low pressure impactor with electrically insulated collection stages. The particles are collected in the different impactor stages according to their aerodynamic diameter, and the electric charge carried by particles into each impactor stage is measured in real time by sensitive multi-channel electrometers. This measured current signal is directly proportional to particle number concentration and size. The particle collection into each impactor stage is dependent on the aerodynamic size of the particles. Measured current signals are converted to (aerodynamic) size distribution using particle size dependent relations describing the properties of the charger and the impactor stages. The result is particle number concentration and size distribution in real-time. By using ELPI™ without operating corona charger, it can be used for particle charge distribution measurements. As the electrical impactor collects the particles, it is possible also to carry out post-sampling measurements, i.e. weight the samples collected on each impactor stage and/or analyze the composition of the particles collected on each stage.

A major problem in any particle measurement device where particles are charged, including the electrical impactor, is the deposition of the charged particles on the inner surfaces of the measurement device. This is caused by the repulsive Coulombic force of the charged particles on each other. This repulsive force is especially effective in the space where the particles are charged by an electric discharge, because the strength of the electrostatic field affecting the particle deposition is further increased by free ions and potentially also by the high-voltage electrode required for the electric discharge. Deposition of the charged particles on the inner surfaces of the electrical impactor may block the aerosol flow channels

(pathways) of the impactor or they may reduce the electrical insulation capacity of the electrical insulation of the collection stages. Blocking the pathways reduces the sample flow through the electrical impactor and thus changes the sensitivity and the size selectivity of the measurement unit. Thus there is a need to monitor the sample flow through the measurement unit and a need to clean the electrical impactor frequently. Frequent maintenance of the measurement unit is a problem especially when the electrical impactor is used in long-term measurements.

One factor causing errors in instruments based on particle charging and

determination of the final charging state is the fact that the original charging state of particles affects the results. In many cases, e.g. when measuring particles from combustion process or particles after electrostatic precipitators, the resulting measuring errors may be significant. The prior art technique uses a separate aerosol neutralizer prior to the charger unit to neutralize the particles before they are charged. However, the neutralizes are not usually used, because it is an extra component producing cost and man work, and it increases the particle losses yielding inaccuracy to the measurement result. There are no reliable and easily applied criteria to decide when use of a neutralizer is necessary. In addition to the problems mentioned above, electrical impactors face the problem of deposition of charged fine particles on the collection stages for larger particles. Due to the collection mechanism, i.e. inertial separation, cascade impactors, including electrical impactors need to be constructed so that the 1^st (upper) collection stage (in the direction of the sample flow) collects the largest particle fraction, the 2^nd stage collects the 2^nd largest particle fraction, etc. This is a major difference to e.g. collection devices based on diffusion, where the smallest particle fraction is collected first (in the direction of the sample flow). However, as the Coulombic force accumulates smaller particles on the first or on the few upper collection stages, these fine particles cause an error both on the real-time measurement and on the post- sampling analysis. Obviously, as some of the fine particles are collected on the upper collection stages, there exists an error also on the measurement results of the lower collection stages collecting fine particles.

The measurement error generated by the problem mentioned above may be significant. Fine particles have a high electrical mobility and thus are highly affected by Coulombic force. In a typical particle sampling the amount of large particles collected by the few upper collection stages is low and thus the electrical current erroneously generated by the fine particles accumulated on the upper stages may even be as large as or larger than the current generated by the large particles.

Finnish patent publication FI 104127 B, Dekati Oy, 28.7.1999, describe a method for minimizing Coulombic losses in impactors and an impactor, in which the Coulombic losses have been minimized. The method according to the invention is based on restricting the force effect of the charges accumulated in impactor' s insulators on the charged particles. This restriction of the force effect can be realized, for example, by placing an electricity conductive surface between the insulator and the flow, or by forming an insulator (24) so that the force effect is minimized. The solution described in FI 104127 B leads to difficult mechanical constructions and requires high-purity electrical insulators, as even small impurities on the insulator surfaces cause leaking currents which adversely affect the measurement.

An electrostatic instrument for measuring particle concentrations and possibly sizes in aerosols, such as an Electrostatic Low Pressure Impactor or Differential Mobility Analyser suffers from errors which limit the useful response bandwidth of the device. United States Patent application Publication US 2004/0080321 Al, Kingsley St. John Reavell, et al., 29.4.2004, describes modifications made to the design of electrostatic particle measurement instruments to compensate or eliminate the transient currents produced by the rate of change of charge near the sensing electrodes, and hence reduce the transient errors in measured particle concentrations. The invention minimizes or eliminates these transient errors which are caused by changing particle concentrations in the aerosol. A system may be added to an otherwise conventional instrument to compensate for the transient effects based on a model of the charge production mechanism. Alternatively, a screening electrode placed over the sense electrodes in the instrument, and held at controlled electrical potential difference, is added to the instrument to eliminate the effect. A third embodiment adds compensating electrodes which provide a direct measurement of the transient effect which can be subtracted from the signal. The publication also provides a hint for volumetric flow measurement where the travel time between two electrodes, the penultimate electrode and the final electrode could potentially be used. However, this measurement method suffers from the problem discussed earlier, i.e. the need for at least two sensing elements. Also, the publication specifically states that the screening electrode is held at controlled potential difference and does not teach potential difference modulation.

US patent publication US 4,837,440 A, Burtscher et al., 6.6.1989, describes a method for the characterization of particles in aerosol, comprising: using an aerosol, which has been brought to at least one predetermined temperature sufficient for evaporation or decomposition or preventing condensation of molecules on particles of said aerosol; exposing said aerosol to an electromagnetic radiation for activating particles contained in the aerosol to cause said aerosol particles to emit electrons and attain an electric charge correspondingly; and measuring the electric charge of said aerosol particles. The patent publication also describes an apparatus for characterization of particles in an aerosol, said apparatus comprising: an essentially closed chamber comprising a gas inlet, a gas outlet, and means for establishing an aerosol flow through said chamber; means mounted in the interior of said essentially closed chamber in the region of said inlet for electrically neutralizing said aerosol; means for heating said aerosol, said heating means comprising measuring and control means for establishing a predetermined aerosol temperature; radiation means arranged within the interior of said essentially closed chamber and defining an irradiation zone for activating the heated aerosol flowing there through; said irradiation zone comprising a wall surface held at a defined electric potential and having a conductivity which is high enough to conduct the charge of ions produced by irradiation and movable to said wall surface to said defined electric potential; and at least one collector electrode mounted in the interior of said essentially closed chamber downstream of said irradiation zone for collecting electrically charged particles and electrically connected to a current or charge meter; said at least one collector electrode comprising a size selective particle filter mounted downstream of said irradiation zone and arranged in the flow of said aerosol, said size selective particle filter comprising an electric capacitor having two electrodes in spaced relationship and an electric field between said two electrodes, said aerosol flowing through the space between said two electrodes and said electric field having a magnitude which is high enough to deposit at least a part of the charged particles of said aerosol on the negative electrode of said capacitor, and control means connected to said electrodes of said capacitor in order to periodically change the field strength of said electric field between a lower value and a higher value. The patent publication US 4,837,440 describes a measurement method which is based on the use of filters connected in series. The probability that a particle is deposited in the filter depends on its diffusion coefficient which, in turn, depends on the size of the particle. Smaller particles have a higher diffusion velocity resulting in a higher probability that they are deposited in the filter. As the probability that a particle is deposited in the filter is also increased, if the filaments of the cluster or web of the filter are arranged tightly to each other, it is obvious for a person skilled in the art that in order to provide size selectivity to a measurement unit described in US 4,837, 440, the smaller particles are collected first (in the direction of the sample flow) and then, subsequently, larger and larger particles. The operation procedure is thus totally opposite to the one of cascade impactors and a person seeking for a solution for charged fine particle deposition in the upper collection stages of electrical impactors would not seek the solution from diffusion-based particle measurement units, such as described in US 4,837,440 or from any other particle measurement method or apparatus, where particle collection is not based on inertia. Such collection apparatuses which are not based on the use of inertial separation include e.g. US 2006284077 Al, TSI Inc., 21.12.2006, where particle collection is based on HEPA filters and US 7,406,855 B2, Dekati Oy, 5.8.2008, where particles are not collected at all. Real-time measurements always have a certain time constant and it is obvious that there is a tendency towards faster and faster measurements. The problem in speeding up the measurements with the electrical impactor is that different collection stages have different time responses. In order to compensate the differences and thus to achieve a high-speed PSD measurement, the difference in the time responses have to be compensated by a compensation algorithm. However, such algorithms cannot be generally developed because the parameters that define the time response change not only between individual measurement units, but also due to different operation conditions, such as temperature and pressure.

As described in Kaarle Hameri, "Instrumentation for aerosol physical properties", About Atmospheric Aerosols, Summer Workshop, 19.-20.6.2006, the idealized transfer function of an impactor, i.e. the deposition efficiency as function of particle size, can be described by means of a step-function at particle size d50 (particle size with 50% collection efficiency on the collection stage). However, real impactors exhibit a deposition

characteristic, Le. their transfer function deviates from the idealized transfer function. The impactor transfer functions are often plotted as function of the square root of the Stokes number for 50% collection efficiency, Stk₅o. A theoretical calculation of real impactor transfer functions, i.e. deposition efficiencies, is possible using numerical methods. The particle size distribution can be then reconstructed if the transfer functions of the different stages are known.

It should be noted that Hameri describes a transfer function which is used to determine the PSD. This is not the same as a dynamic transfer function which can be used to describe the time response of the collection stage. One way for describe the time-wise behavior of a collection stage is to model it by a delay ¾ and a first-order low-pass filter (which, as obvious for a person skilled in the art is equivalent to the time-wise behavior of a completely mixed reactor). The Laplace-type transfer function of a low-pass filter is, (s) =

^(s) ⁼ _i+_Ts' ^wnere > ° ^ref^{er t0 me} input and output of the collection stage, s is frequency and τ is the time constant .

As stated above, the transfer function for the PSD determination as well as the dynamic transfer function changes due to different operational conditions and thus theoretical calculation of the transfer functions or even experimental determination of the transfer functions with fixed operational conditions will not provide accurate results for real-life measurements in continuously changing operational conditions. In addition to the particle concentration and PSD, information on the electrical charge of the particles under measurement is often required. Real-time particle charge measurement as a function of the particle size with the prior art technology requires two measurement units, such as two ELPF^Ms, one for measuring the particle concentration (as a function of the particle size) and another for measuring the particle charge, in which case the corona charger used to charge the particles entering the measurement unit is switched off. Such a measurement method is expensive and the use of two parallel instruments increases the measurement errors. The measurement error is further increased by the "naturally" charged particles (i.e. particles which carry charge before entering the measurement unit), as also the charging may differ between the sample flows of the two measurement units.

There is still another problem within the particle measurements carried out by the electrical impactors: the measurement of very small electrical currents, even down to fA- level. Especially the drifting of the current over time creates measurement errors and these are obviously most severe with long measurement periods and low particle concentrations.

Thus the electrical impactors of the prior art possess the following technical problems: deposition of fine particles to the few upper collection stages, limitations in the transfer function determination (both the transfer function for the PSD as well as the dynamic transfer function), determination of the charge of the particles entering the measurement unit, measurement errors caused by naturally charged particles and measurement current drifting.

Brief description of the invention

The inventor has surprisingly found a method which will solve the prior art electrical impactor problems described above. The problems are solved by implementing a process where at least either (1) the particle charging or (2) free ion and small particle collection is switched, preferably on and off, or modulated, over time.

The invented electrical cascade impactor, in the following text also referred to as "electrical impactor" or "impactor" ensures reliable, size-selective particle concentration measurement, makes sure particle charge measurement, provides a method for eliminating a measurement error caused by naturally charged particles, confirms reliable particle measurement results even with low particle concentrations and with long measurement periods and provides tools for precise correction algorithms due to the determination of time responses of different collection stages. All this can be achieved simultaneously with decreased soiling or particle deposition into the measurement unit and thus the maintenance period is prolonged. The invented method also provides a method to measure or control the sample flow flowing through the electrical impactor.

The mechanical and electrical construction of the impactor resembles the

construction of ELPI™: the impactor comprises an electrical discharging unit, typically a corona charger, which charges the particles flowing into the impactor, and at least one collection stage which is based on inertial separation of particles so that the largest particles are collected on the first (upper) collection stage. Prior to, in the direction of the sample flow, the collection stage or typically collection stages lays usually an ion trap which removes free ions or extremely small particles (typically having a diameter of few nanometers or less than 10 nanometers) from the sample flow. In some embodiments of the present invention the ion trap can remove particles up to about 100 nm particle diameter, i.e. also particles larger than the nucleation mode particles. Obviously such measurement mode is not used when measuring particles smaller than the particles which are trapped. The impactor may also comprise a neutralizer which neutralizes the electrical charge of the particles before they enter to the collection stage(s). The neutralizer may be a separate unit or the neutralization may be carried out by the electrical discharging unit, e.g. by such a way that the corona charger comprises two charging units with opposite electrical potential or by using a single corona charger in AC (alternating current) mode which produces ions with opposite charges.

The electrical discharging unit of the present invention has a unique feature that it can be switched or modulated between different charging modes. In an embodiment of the present invention the electrical discharging unit is a corona charger which is switched periodically between the ON-mode and OFF-mode, i.e. the corona voltage is periodically switched ON and OFF. Essentially continuous comparison of the measurement results between the ON and OFF modes provides the base for calculating the natural charge of the particles as well as for calculating the particle concentration. The term "periodical switching" means that the switching may have a fixed frequency or the frequency may vary and also the length of the ON and OFF modes may vary. The length of the ON mode may be less than 100 seconds, less than 10 seconds or even less than a second. The duty cycle may vary between 1 and 99%, preferably between 5 and 50% and more preferably between 5 and 20%. Especially with high-speed real-time measurements the response time of the impactor should be as short as possible. In another embodiment of the present invention the response time is decreased by using, in the calculation of the particle concentration or PSD, the time constants of each collection stage which can be obtained from the response which the switching, or more generally modulating, of the corona charger voltage creates to the current measured from the collection stage(s). Another alternative is to modulate or switch the trap voltage. The time constants can even be determined continuously and thus the changing operational conditions, such as temperature or pressure, will not adversely affect the measurement accuracy.

In the prior art ELPI™ measurements the particle concentration is determined from the total charge collected from a collection stage. In an embodiment of the present invention the concentration is determined from the difference between the collection stage current of the ON-mode and the OFF-mode. Using the switched mode measurement makes it possible to continuously determine the drifting of the measured current and eliminate it.

In yet another embodiment of the present invention, the electrical discharging unit is switched between two opposite voltages, ON⁺ and ON^". When the particle concentration or PSD is determined from the difference of the response between these two modes, the natural charge of the measured particles affects the measurement significantly less than if the result is determined from the difference between the ON and OFF modes.

In yet another embodiment of the present invention the electrical discharging unit is switched between three different modes: ON-, OFF-, and NE-modes, where the NE mode describes a mode where either a separate neutralizer is used or the electrical discharging unit is used to neutralize the particles entering the impactor. Using such an embodiment is especially beneficial when measuring the natural particle charge with low particle

concentrations. With this embodiment the current drifting and low-frequency noise is eliminated by determining the natural particle charge from the difference of the OFF-and NE- modes.

Using the method of the present invention the natural charge of the particles can be determined by using a single measurement apparatus. Obviously there is a practical advantage of avoiding the use and handling of two measurement units. Also, because the aerosol sampling is taken from a single point, the differences between two sampling points do not affect the measurement accuracy and neither do the calibration differences and instabilities of two different measurement equipment.

In all embodiments described above, the electrical discharge switching decreases the deposition of the charged particles on the measurement unit interior and especially beneficially decreases the deposition of fine particles on the upper collection stage(s) due to the Coulombic force. The best results are achieved when the duty cycle is small, preferably around 10% or less. It has been surprisingly found that in typical realizations of the detection electronics of electrical impactors the low-frequency (1/f) noise dominates the noise level of the measurement result. The consequence of this feature is that even low duty cycles and synchronous detecting technique may yield better noise and sensitivity performance to the instrument than the prior-art measurement mode without switching/modulating. So, this improved noise performance can be achieved in addition to the elimination of current drifting. Additionally, as the soiling of the measurement unit is considerably decreased, the

maintenance period of the instrument is substantially prolonged.

In one embodiment of the present invention, where particles with a high amount of natural charging are measured, it is advantageous to use switching between ON and NE modes to minimize the particle deposition, as the NE-mode decreases the Coulombic deposition of the naturally charged particles. If, in addition to the particle concentration and PSD, also the natural charge of the particles is to be measured, the electrical discharge unit can be switched between ON, OFF and NE-modes. With such an embodiment it is preferable that the length of the NE-mode is longer than the length of the ON- and OFF-modes and thus the Coulombic deposition can be minimized.

In another embodiment of the current invention, the measurement error caused by the fine particles accumulated on the upper collection stage(s) is determined by switching the operation mode of the ion trap. As described above prior to (in the direction of the sample flow) the collection stage(s) lays usually an ion trap which removes free ions or extremely small particles (typically having a diameter of few nanometers or less than 10 nanometers) from the sample flow. Increasing the ion trap voltage increases the trapping of fine particles and with high trap voltage the fine particles which would accumulate on the upper collection stage(s) are removed. The ion trap may be switched between high and low voltages, V_hjgh and V_low and measuring the response, i.e. the current from the upper collection stage(s), the measurement error on the upper collection stage(s) caused by the fine particles in the sample may be determined and simply decreased from the measurement current of the upper stage(s). This measurement result may also be used to estimate the error of the lower collection stage(s), which receive too little amount of fine particles, as compared to the actual sample flow.

In all embodiments described above, the duty cycle or the length of the ON⁺, ON^", OFF and NE-modes can be varied during the measurement and thus optimize the operation of the measurement apparatus. The duty cycle control may be based on an internal signal of the measurement apparatus e.g. the particle concentration or on the time wise derivative of the particle concentration. The duty cycle control may also be based on an external signal, e.g. when measuring particle exhaust from a combustion engine; the external signal may include a change in the momentum of the engine or in the revolution speed of the engine. An advantageous feature is that in high-concentration conditions, where the soiling rate is high, extremely low duty cycles can be used without practical degrading the noise properties of the signal. The reason for this is the fact that during even extremely short ON times (t₀N) the high signal levels can yield enough noise-free measurement data.

In the preferred embodiment of the present invention the response from the switched or modulated mode of the electrical discharge unit is determined by synchronic detection. Synchronic detection can be realized by using either analog electronics or digitally. The digital realization can obviously be carried out in a separate computing unit or it may be integrated to a common controller or computing unit, where other control functions of the electrical impactor are carried out as well.

In yet another embodiment of the present invention, the switching or modulation of the electrical discharge unit is used in determining the dynamic transfer function of each collection stage, keeping in mind that the typical main parameters, delay time, t_s, and characteristic time, τ, may vary for each collection stage. The main parameters are determined by modulating the electrical discharge unit or trap voltage, i.e. the switching of the electrical discharge unit between at least two of the modes ON⁺, ON^", OFF or NE. Within this embodiment the main parameters can be determined even continuously and thus the time response of each collection stage may be determined even continuously and the measurement result of each collection stage may be corrected even continuously and thus the changes in the measurement environment do not adversely affect the measurement result. If the changes in the measurement environment are not remarkable on a short time interval, and when the maximum time response of the measurement is required, the determination of the main parameters may be carried out with longer intervals, and thus use the determined parameters over longer intervals.

In one embodiment of the present invention, the correction of the measurement result of each collection stage is based on a model where the time wise behavior of each stage is modeled by delay and filter units which are connected together. Although the optimal equivalent circuit used in the modeling depends on the flow characteristics of the collection stage, the serial collection of delay and transfer units is usually a reasonably accurate model. The filter units are typically first-order low-pass filters, the behavior of which is determined by the characteristic time constant r. The time correction of the collection stages may be realized by compensating the combined delay times t_s by analog or digital means and by analog or computational (digital) filter units carrying out inverse time constants τ. The time correction may be carried out continuously, it may be carried out with longer intervals or it may be even carried out during the calibration of the electrical impactor and store the time correction parameters on the control unit of the measurement equipment. The interval time for determining the time correction parameters depends mainly on the use of the instrument, i.e. if it is used in steady or varying environmental conditions.

A surprising benefit of using a modulated signal instead of one sensor signal is that, in spite of simpler and cheaper practical solution, it can yield better performance. The reason of this feature is that one of the signals to be compared/correlated has practically no noise or other disturbances, whereas in prior-art solutions two noisy signals are compared/correlated.

Brief description of the drawings

In the following, the invention will be described in more detail with reference to the appended schematic drawings, where

Fig. 1 shows a summary of the invention with its benefits to the electrical impactor; Fig. 2 shows a schematic drawing of the invented apparatus;

Fig. 3 describes the time behavior of the collection stages;

Fig 4 describes the determination of the parameters used in the time correction of the collection stages;

Fig 5 describes the time wise correction of the measurement signal;

Fig 6 describes an embodiment where the determination of the parameters used in time correction and the time wise correction of the measurement signal are integrated together; and

Fig 7 describes another embodiment where the determination of the parameters used in time correction and the time wise correction of the measurement signal are integrated together.

For the sake of clarity, the figure only shows the details necessary for understanding the invention. The structures and details which are not necessary for understanding the invention and which are obvious for a person skilled in the art have been omitted from the figures in order to emphasize the characteristics of the invention.

Detailed description of preferred embodiments

The present invention is an electrical impactor 1 (numbers refer to Figure 2)where the particles entering the impactor 1 are classified by inertial separation on at least one stage 4, characterized in that impactor 1 comprises a functional component 22 which switches or modulates the operational mode of the electrical discharge unit 20 or/and the ion trap. The electrical discharge unit 20 is switched or modulated at least between two of the modes: high positive voltage (ON+), high negative voltage (ON-), essentially zero voltage (OFF) and neutralization (NE). The electrical discharge unit 20 or the ion trap is preferably switched or modulated on frequency higher than 0,01 Hz, more preferably higher than 0, 1 Hz and most preferably higher than 1 Hz. The duty cycle of the switched/modulated signal is preferably between 1-99%, more preferably between 5-50% and most preferably between 5-20%. In most embodiments of the present invention the frequency and the duty cycle can vary freely during the measurement and also the form of the modulated pulse may vary. Especially with synchronic detection such variations do not adversely affect the measurement.

The electrical impactor 1 further comprises functional component 24 connected to functional component 22 or to the electrical discharge unit 20 and to electrometer 6 for providing the natural charge of the particles on stage 4, as well as functional component 25 connected to functional component 22 or to the electrical discharge unit 20 and to

electrometer 6 for providing particle concentration of the particles on stage 4. Electrical impactor 1 further comprises functional component 26 connected to functional component 22 or to the electrical discharge unit 20 and to electrometer 6 for eliminating measurement current drifting in electrometer 6.

Electrical impactor 1 comprises functional component 27 for determining the essential parameters of transfer function Fi of stage i 4. Functional component 23 connected to functional component 27 for monitoring the volumetric flow flowing through impactor 1.

The present invention also includes a process for particle measurement where the particles are classified by inertial separation, characterized in that unwanted particle deposition on the upper stages used for inertial separation is minimized by using essentially continuous switching or modulating either the electrical source used for particle charging or the ion trap used to trap free ions and ultrafine particles. The electrical source used for particle charging is switched or modulated at least between two of the modes: high positive voltage (ON+), high negative voltage (ON-), essentially zero voltage (OFF) and neutralization (NE). The electrical source used for particle charging is switched or modulated on frequency higher than 0,01 Hz, preferably higher than 0,1 Hz and most preferably higher than 1 Hz. The electrical source used for particle charging is switched or modulated with duty cycle between 1-99%, preferably between 5-50% and most preferably between 5-20%.

The difference in the measurement results between at least two of the modes: high positive voltage (ON+), high negative voltage (ON-), essentially zero voltage (OFF) and neutralization (NE) is used to determine particle concentration, or particle size distribution, or natural charge of particles and to eliminate the measurement current drifting as well as to determine the essential parameters of the transfer function Fi of stage i. In a preferred embodiment of the current invention, the transfer function Fi of stage i comprises a first-order low-pass filter transfer function and the essential parameters are delay time t i of stage i and time constant rof the first-order low-pass filter. The process further includes correcting the measurement signal Soi by the inverse of the transfer function Fi.

The switching or modulation frequency as well as the duty cycle can be adjusted during the measurement. The duty cycle does not have to stay constant during particle monitoring, but a beneficial advantage of the present invention is that the duty cycle can be dynamic, i.e. its value may be optimized during the measurement. Not only the duty cycle may be varied, but also the length of the t₀N may be varied as well. The variation and optimization may be based e.g. on the sensor measurement result (particle concentration, time-wise derivative of particle concentration) or on external factors (like combustion engine revolution speed or torque).

Figure 2 shows an embodiment of the present invention, which comprises an electrical impactor 1 where the particles entering the impactor 1 are classified by inertial separation on at least one stage 4. Impactor 1 comprises a functional component 22 which switches or modulates the operational mode of the electrical discharge unit 20 or/and the ion trap.

Impactor 1 also comprises means for switching or modulating the electrical discharge unit 20 at least between two of the modes: high positive voltage (ON+), high negative voltage (ON-), essentially zero voltage (OFF) and neutralization (NE). The switching or modulating frequency is preferably higher than 0,01 Hz, more preferably higher than 0,1 Hz and most preferably higher than 1 Hz and the duty cycle is between 1-99%, preferably between 5-50% and most preferably between 5-20%.

Electrical impactor 1 also comprises functional component 24 connected to functional component 22 or to the electrical discharge unit 20 and to electrometer 6 for providing the natural charge of the particles on stage 4 and functional component 25 connected to functional component 22 or to the electrical discharge unit 20 and to

electrometer 6 for providing particle concentration of the particles on stage 4. Impactor 1 also comprises functional component 26 connected to functional component 22 or to the electrical discharge unit 20 and to electrometer 6 for eliminating measurement current drifting in electrometer 6.

Another feature of the present invention is that electrical impactor 1 may comprise functional component 27 for determining the essential parameters of transfer function Fi of stage 4i. It is possible to monitor the volumetric flow passing through the impactor lby determining the essential parameters of transfer function Fi and thus impactor 1 may comprise functional component 23 connected to functional component 27 for monitoring the volumetric flow flowing through impactor 1.

The present invention also includes a process for particle measurement where the particles are classified by inertial separation, so that unwanted particle deposition on the upper stages used for inertial separation is minimized by using switching or modulating either the electrical source used for particle charging or the ion trap used to trap free ions and ultrafme particles. The electrical source used for particle charging is switched or modulated at least between two of the modes: high positive voltage (ON+), high negative voltage (ON-), essentially zero voltage (OFF) and neutralization (NE). The electrical source used for particle charging is switched or modulated on frequency higher than 0,01 Hz, preferably higher than 0,1 Hz and most preferably higher than 1 Hz and the electrical source used for particle charging is switched or modulated with duty cycle between 1-99%, preferably between 5-50% and most preferably between 5-20%.

In the invented process the difference in the measurement results between at least two of the modes: high positive voltage (ON+), high negative voltage (ON-), essentially zero voltage (OFF) and neutralization (NE) is used to determine particle concentration, or particle size distribution, or natural charge of particles. The difference in the measurement results between at least two of the modes: high positive voltage (ON+), high negative voltage (ON-), essentially zero voltage (OFF) and neutralization (NE) may also be used to eliminate the measurement current drifting and to determine the essential parameters of the transfer function F, of stage i. In one embodiment of the present invention, the transfer function F, of stage i comprises a first-order low-pass filter transfer function and the essential parameters are delay time of stage i and time constant rof the first-order low-pass filter. The invented process further includes correcting the measurement signal Soi by the inverse of the transfer function E,.

Figure 1 shows a summary of the invention, describing the beneficial effects that the invention has on the operation of the electrical impactor. The main inventive step of the present invention is that the modulation (time wise change) of the electrical discharge element input/output (e.g. modulation of the voltage of a corona discharge unit resulting in the modulation of the ion current emitted from the corona needle) and/or modulation of the ion trap voltage decreases the particle deposition on the interior of the electrical impactor and especially minimizes fine particle deposition on the upper collection stages caused by

Coulombic force. Modulation may be a step-wise switching, e.g. switching between ON and OFF modes, or it may be an essentially smooth change between at least two separate modes.

In such embodiments of the present invention where the ion trap voltage is modulated, it is possible to use the measurement results especially from the upper collection stages to determine the parasitic effect which the deposition of fine particles on the upper collection stages produces. Determination of the parasitic effect allows correction of the error caused by upper-stage fine particle deposition. The correction can be made both to the upper stages (additional deposition of fine particles) or lower stages (reduced deposition of fine particles).The correction can be made both to the electrical signal measured from the stages and/or to the weighted mass on the stages.

As Figure 1 shows, the modulation provides additional benefits to the electrical impactor in addition to the decrease of the particle deposition. These additional beneficial effects include definition of the natural charge of the particles in the sample flow, elimination of measurement current drifting, reduction of the low-frequency noise, and elimination of the measurement error caused by the deposition/accumulation of the fine particles on the upper collection stages (which not only affects the measurement signal of the upper stages but also of the lower stages collecting fine particles) and improving the representativeness of the sample. As the present invention also allows the determination of the time response which can be typically presented by two parameters: the delay time tj and the time constant Tof each collection stage, it may be used to speed up the time response of the electrical impactor, improve the PSD determination and monitor the sample flow through the measurement apparatus. Figure 2 shows the schematic drawing of the invented apparatus. Electrical impactor 1 comprises an electrical discharge unit 20 for charging particles which enter impactor 1 through the inlet nozzle 12. The electrical discharge unit 20 is preferably a corona charger or a dielectric barrier discharge unit and it may also comprise a neutralizer or it may be used in a neutralizing mode, e.g. by using an alternating current (AC) power supply with corona charger. Impactor 1 comprises at least on chamber 2, having side surfaces and forming collection stage 4, which is electrically connected to an electrometer 6. Collection stage 4 is electrically isolated from the impactor body 10 and potential other stages 4 by electrical insulator 8. Aerosol sample flow 14 is passed through the inlet nozzle 12 into chamber 2. The bottom of chamber 2 comprises a nozzle 16 which is essentially perpendicular to the sample flow 14. Nozzle 16 is in galvanic contact to stage 4. Nozzle 16 comprises holes through which sample flow 14 flows. After nozzle 16 the sample flow impacts on a collection plate 18 and particles larger than a designed particle size attach on the collection plate 18. The collection plate 18 may be in galvanic contact to stage 4, but impactor 1 may also be constructed so that only the collector plate 18 is connected to the electrometer 6, or impactor 1 may also be constructed so that only the surface of collector plate 18, e.g. a metal foil, is connected to the electrometer 6. The particle collection in electrical impactor 1 is based on inertial separation, where particle collection is based on sudden change in the direction of the aerosol flow.

Impactor 1 further comprises a exit nozzle 30 to guide flow 14 to the next chamber 2 or to exit the flow from impactor 1. Apparatus 1 of the embodiment shown in Figure 2 further comprises component 22 to modulate the operation mode of the electrical discharge unit 20. In one embodiment of the present invention, component 22 controls the corona voltage of a corona charger 20 between ON and OFF modes with 0,01 Hz - 1 Hz frequency. The duty cycle is preferably about 10%, so that 90% of time the corona charger is in OFF-mode and thus the particles entering apparatus 1 are not charged by the corona charger 20.

The embodiment of Figure 2 further comprises component 24, which component 24 should be considered as a functional component, i.e. component 24 may be a separate component carrying out the required function or it may be an integral part of some other component such as the main control unit of the electrical impactor 1. Component 24 may carry out the required function by analog or by digital means. Component 24 is connected to the electrical discharge unit 20 or to the controlling component 22 or to both and it is also connected to the electrometer 6. The function of component 24 is to provide information on the natural charge of the particles entering the impactor 1, based on the current measured by the electrometer 6. Component 24 compares the current values provided by the electrometer 6 at the different operation modes of the electrical discharge unit 20, which is controlled by the control unit 20. Typically the electrical discharge unit 20 is a corona charger and its voltage is switched or modulated between OFF and NE modes. Switching between OFF and NE modes provides an additional benefit as it allows the elimination of the current drift of the electrometer 6. Elimination of the current drift is carried out by another functional component 26, which may be a separate component carrying out the required function or it may be an integral part of some other component such as the main control unit of the electrical impactor 1. Component 26 may carry out the required function by analog or by digital means.

Practically switching between stages ON and OFF modes gives satisfactory results also for the natural charge of the particles, if the particle concentration is so high that the electrometer current drift elimination is not required. ON/OFF switching also provides satisfactory results for electrometer current drifting elimination, for particle concentration measurement and PSD measurement if the natural charge of particles is not too high.

The embodiment of Figure 2 further comprises component 25, which component 25 should be considered as a functional component, i.e. component 25 may be a separate component carrying out the required function or it may be an integral part of some other component such as the main control unit of the electrical impactor 1. Component 25 may carry out the required function by analog or by digital means. Component 25 is connected to the electrical discharge unit 20 or to the controlling component 22 or to both and it is also connected to the electrometer 6. The function of component 25 is to provide information on the concentration of the particles entering the impactor 1, based on the current measured by the electrometer 6. Component 25 compares the current values provided by the electrometer 6 at the different operation modes of the electrical discharge unit 20, which is controlled by the control unit 20. Typically the electrical discharge unit 20 is a corona charger and its voltage is switched or modulated between ON and OFF modes or between ON and NE modes or between ON⁺ and ON^" modes. Switching between different modes provides an additional benefit as it allows the elimination of the current drift of the electrometer 6. Elimination of the current drift and low-frequency noise suppression is carried out by another functional component 26. The embodiment of Figure 2 further comprises component 27, which component 27 should be considered as a functional component, i.e. component 27 may be a separate component carrying out the required function or it may be an integral part of some other component such as the main control unit of the electrical impactor 1. Component 27 may carry out the required function by analog or by digital means. Component 27 is connected to the electrical discharge unit 20 or to the controlling component 22 or to both and it is also connected to the electrometer 6. The function of component 27 is to determine the main parameters, especially delay time ^and time constant rof the transfer function of stage 4. These parameters can be used to optimize the time-wise behavior of stage 4.

The embodiment of Figure 2 further comprises component 23, which component 23 should be considered as a functional component, i.e. component 23 may be a separate component carrying out the required function or it may be an integral part of some other component such as the main control unit of the electrical impactor 1. Component 23 may carry out the required function by analog or by digital means. Component 23 is connected to the component 27. The function of component 23 is to use the delay and time constants obtained by component 27 to monitor the flow passing through the electrical impactor 1.

In a preferred embodiment of the present invention the function of component 27 is realized by describing the time- wise behavior of stage 4 so that it comprises a delay component t_s and a time constant which describes the behavior of a first-order low-pass filter. The time-wise behavior of the firs-order low-pass filter is similar to the one of a fuU-mixed reactor. The first-order low-pass filter can be described in Laplace notation as

_F(s) = K ^- (i)

J Si(s) l +TS ' where So(s) is the output signal, Si(s) is the input signal, K is the filter passband gain (which in this case can be set to unity), t is the time constant and s is the Laplace transform variable.

The derivative of function F(s ) in time domain is dSo(t) _ Si(t)-So(t) _(r).

dt ~ τ ^ In the embodiment shown in Figure 2 component 27 receives signals at a certain frequency and with a certain duty cycle, the frequency being preferably higher than 0,01 Hz, more preferably higher than 0,1 Hz and even more preferably higher than 1 Hz and the duty cycle being preferably between 5-20%. The differential equation (2) can be approximated by difference equation (3):

Δ ο _ Si-So

At ~ τ ^W

In the electrical impactor 1 the stages 4 are connected to a cascade and thus the time response of the whole impactor can be described as a serial connection of first-order low-pass filters and delay times, as shown in Figure 3.

Delay time ts and time constant ¾ of each collection stage A can be determined by setting them to such values that the model describing the time-wise behavior of each stage has a maximum correlation to the actual measured signal. In one embodiment of the present invention, the required impulse for the response is generated by modulating the electrical discharge unit 20 between ON and OFF modes. In another embodiment of the present invention, the required impulse for the response is generated by modulating the electrical discharge unit 20 between ON and NE modes. In yet another embodiment of the present invention, the required impulse for the response is generated by modulating the electrical discharge unit 20 between ON⁺ and ON^" modes. In yet another embodiment of the present invention, the required impulse for the response is generated by modulating the trap voltage of the ion trap.

Figure 4 shows a block diagram of an embodiment where the time- wise parameters are determined. Component 27 providing the time-wise model receives an excitation Cy, from component 22 controlling the corona charger 20. Component 27 comprises a function Con,- which is either an analog or digital unit connected to correlator X. Correlator X compares the measurement signal So, from stage 4j to the calculated signal E, and the time parameters t_<u and Ti are set to the values where the correlator X provides a maximum signal C„ i.e. the correlation between the model and the measured signal is set to maximum.

When the parameters and ¾ have been determined, the time response of the measured signal can be compensated or correlated by modifying the signal with the inverse function. With the transfer function following the Laplace notation this means multiplying the signal with the inverse of the transfer function F(s).

Compensating the time delay is merely applying a time shift. Compensating the first- order low-pass filter is achieved by the difference equation Si = So + T ^ (4)

An embodiment with the compensation algorithm based on time delay and first-order low-pass filter is shown as a block diagram in Figure 5. The corrected output signals Set from each stage i are calculated by modifying the stage output signals So, with the inverse transfer functions 7/F,-. It is obvious for a person skilled in the art that modeling the time-wise behavior with a first-order low-pass filter is only given here as an example and any suitable model which describes the behavior of stage 4 and which can be presented in analog or digital form can be used for modeling. The best model depends on the construction of the electrical impactor.

In one embodiment of the present invention the current from each stage is not measured only from the collection plate 18, but from essentially all surfaces of collection stage 4, i.e. stage 4 forms a Faraday cage into which sample 14 enters from nozzle 12 and exits from nozzle 30, as shown in Figure 2. The current from the charged particles which stay in collection stage 4 is measured with electrometer 6. The charged particles leaving stage 4 obviously carry current with them to the next stages. The time- wise correction to the currents /, measured from each stage i can be made as:

and thus the corrected current la of the first stage is: lc = ( + + + ·" ) - ( + + - ) (6)

1

Figures 6 and 7 show two embodiments for determining the correction parameters and realizing the signal correction. Only two first stages are shown in the figures, but it is obvious for a person skilled in the art that similar protocol may be followed also for the subsequent staes 3,4, and so on. The embodiment of Figure 6 maximizes the correlation between the signal Soi measured from stage i and the modeled transfer function into which signal Ch provides the impulse. The embodiment of Figure 7 maximizes the time-wise correlation between the corrected signal Sc, and the signal Ch representing the control or output signal of the electrical discharge unit 20. Both embodiments lead in principle to the same result.

In all embodiments described above the volumetric flow passing the electrical impactor 1 is inversely proportional to the sum of the time delay ^and time constant τ. Thus the determination of the time constants tj and Tprovides also a tool to monitor or measure the volumetric flow.

It is obvious for a person skilled in the art that the determination of the correction parameters may be utilized by various ways, and including either analog or digital electronics or both. As the components should be considered as functional components, the function provided by each component may also be produced in a main control unit or equivalent.

Claims

1. Electrical impactor (1) where the particles entering the impactor (1) are classified by inertial separation on at least one stage (4), characterized in that impactor (1) comprises a functional component (22) which switches or modulates the operational mode of the electrical discharge unit (20) or/and the ion trap.

2. Electrical impactor ( 1 ) of claim 1, characterized in that impactor (1) comprises means for switching or modulating the electrical discharge unit (20) at least between two of the modes: high positive voltage (ON⁺), high negative voltage (ON^"), essentially zero voltage (OFF) and neutralization (NE).

3. Electrical impactor (1) as in any of the previous claims,

characterized in that impactor ( 1 ) comprises means for switching or modulating the electrical discharge unit (20) or the ion trap on frequency higher than 0,01 Hz, preferably higher than 0,1 Hz and most preferably higher than 1 Hz.

4. Electrical impactor (1) as in any of the previous claims,

characterized in that impactor (1) comprises means for switching or modulating the electrical discharge unit (20) or the ion trap with duty cycle between 1-99%, preferably between 5-50% and most preferably between 5-20%.

5. Electrical impactor (1) as in any of the previous claims,

characterized in that impactor ( 1 ) comprises functional component (24) connected to functional component (22) or to the electrical discharge unit (20) and to electrometer (6) for providing the natural charge of the particles on stage (4).

6. Electrical impactor (1) as in any of the previous claims,

characterized in that impactor (1) comprises functional component (25) connected to functional component (22) or to the electrical discharge unit (20) and to electrometer (6) for providing particle concentration of the particles on stage (4).

7. Electrical impactor (1) as in any of the previous claims,

characterized in that impactor ( 1 ) comprises functional component (26) connected to functional component (22) or to the electrical discharge unit (20) and to electrometer (6) for eliminating measurement current drifting in electrometer (6).

8. Electrical impactor (1) as in any of the previous claims,

characterized in that impactor ( 1 ) comprises functional component (27) for determining the essential parameters of transfer function F,- of stage i (4).

9. Electrical impactor ( 1 ) as in claim 8, characterized in that impactor (1) comprises functional component (23) connected to functional component (27) for monitoring the volumetric flow flowing through impactor (1).

10. Electrical impactor (1) as in any of the previous claims, comprising means for detennining the parasitic effect which the deposition of fine particles on the upper collection stages produces and means for using the determination of the error to correct the measurement result of at least either the upper stages or the lower stages.

11. Process for particle measurement where the particles are classified by inertial separation, characterized in that unwanted particle deposition/accumulation on the upper stages used for inertial separation is minimized by using essentially continuous switching or modulating either the electrical source used for particle charging or the ion trap used to trap free ions and ultrafine particles.

12. Process of claim 11, characterized in that the electrical source used for particle charging is switched or modulated at least between two of the modes: high positive voltage (ON⁺), high negative voltage (ON^"), essentially zero voltage (OFF) and neutralization (NE).

13. Process of claim 11 or 12, characterized in that the electrical source used for particle charging is switched or modulated on frequency higher than 0,01 Hz, preferably higher than 0, 1 Hz and most preferably higher than 1 Hz.

14. Process as in any of claims 11-13, characterized inthatthe electrical source used for particle charging is switched or modulated with duty cycle between 1-99%, preferably between 5-50% and most preferably between 5-20%.

15. Process as in any of claims 11-14, characterized inthatthe difference in the measurement results between at least two of the modes: high positive voltage (ON⁺), high negative voltage (ON^"), essentially zero voltage (OFF) and neutralization (NE) is used to determine particle concentration, or particle size distribution, or natural charge of particles.

16. Process as in any of claims 11-15, characterized inthatthe difference in the measurement results between at least two of the modes: high positive voltage (ON⁺), high negative voltage (ON^"), essentially zero voltage (OFF) and neutralization (NE) is used to eliminate the measurement current drifting and/or to decrease low-frequency noise level.

17. Process as in any of claims 11-16, characterized inthatthe measured response for switching between at least two of the modes: high positive voltage (ON⁺), high negative voltage (ON^"), essentially zero voltage (OFF) and neutralization (NE) is used to determine the essential parameters of the transfer function F,- of stage i.

18. Process as in any of claims 11-17, characterized inthatthe measured response for modulation of the ion trap voltage used to trap free ions and ultrafine particles is used to determine the essential parameters of the transfer function F,- of stage i.

19. Process as in claim 17 or 18, characterized in that the transfer function Ft of stage i comprises a first-order low-pass filter transfer function and the essential parameters are delay time ½ of stage i and time constant rof the first-order low-pass filter.

20. Process as in claim 17 -19, characterized in that the process includes correcting the measurement signal So, by the inverse of the transfer function F,-.

21. Process as in any of the claims 11-20, characterized inthatthe switching or modulation frequency is adjusted during the measurement.

22. Process as in any of the claims 11-21, characterized in that the duty cycle is adjusted during the measurement.

23. Process as in any of the claims 11-22, characterized in that modulation of the ion trap voltage is used to determine the parasitic effect which the deposition of fine particles on the upper collection stages produces and the determination of the error is used to correct the measurement result, including both the electrical signal from the stages and/or the weighted mass on the stages, of at least either the upper stages or the lower stages.