PARTICLE GENERATOR

I. TECHNICAL FIELD

The invention relates to a method and an apparatus adapted for monitoring a spark discharge particle generator utilized for generating particles, particularly nanoparticles.

II. BACKGROUND ART

11.1. GENERATING NANOPARTICLES BY SPARK DISCHARGE

Applying high-voltage spark discharges between electrically conductive electrodes, nanoparticles dispersed in gas can be generated from the material of the electrode, i.e. this phenomenon is preferably applicable for the generation of nanoscale-sized particles.

By modifying gas flow conditions in the discharge space and adjusting discharge parameters, the size and number concentration of the generated particles can be influenced, while the chemical composition of the generated nanoparticles can be adjusted by changing the material of the electrodes. Presently, the characteristics of the nanoparticles generated in the spark discharges (primarily size, concentration, etc.) can be determined only by costly and complex instruments/methods that cannot/hardly can be utilized under industrial conditions, while most of them cannot even be operated real-time (online).

In the following, a brief review of known nanoparticle generation methods (based on electric discharge) is included.

The study of nanoparticles (i.e. objects having a dimension in the range below 100 nm; there are a number of definitions for nanoparticle but this one is the most frequently applied in the field) has recently gained a lot of attention due to the wide range of the potential applications allowed by their special characteristics (being different from bulk or macroscopic characteristics). These characteristics, for example lower melting point, higher diffusion coefficient, modified thermal properties, size-dependent catalytic properties (Kruis, F. E., Fissan, H., Peled, A.: Synthesis of nanoparticles in the gas phase for electronic, optical and magnetic application-a review. J. Aerosol. Sci., vol. 29, 511-535 (1998)), modified optical properties (Kelly, K. L, Coronado E., Zhao, L. L, and Schatz, G. C.: The Optical Properties of Metal Nanoparticles: The Influence of Size, Shape, and Dielectric Environment. J. Phys. Chem. B., vol. 107, 668-677 (2003)), etc. have provoked intensive research, among others in the field of the medical sciences, electronics, sensorics, non-linear optics and renewable energy (Borra, J-P.: Nucleation and aerosol processing in atmospheric pressure electrical discharges: powders production, coatings and filtration. J Phys D., vol. 39 R19-R54 (2006)).

The increased interest in the research for characteristics of nanoparticles and for their possible fields of application has resulted in considerable progress in the development of methods and processes for producing nanoparticles. The most widespread method of nanoparticle generation is the chemical, so-called sol-gel method, but gas-phase techniques involving physical processes also see more and more widespread use. One of the advantages of gas-phase methods is that they do not require liquid solvents and precursors, and that they are also waste- free, while they generate high-purity particles due to the higher thermal stability of certain gases (Lehtinen, K. E. J., Backman, U., Jokiniemi, J. K., Kulmala, M.: Three-body collisions as a particle formation mechanism in silver nanoparticle synthesis. J. Colloid Interface Sci., vol. 274, 526-530 (2004)). For a number of applications, it is also a preferred characteristic that in the course of the gas-phase processes typically crystalline particles are generated. Another important consideration is that, in contrast to wet chemical processes, gas-phase processes allow for continuous particle generation (Tabrizi, N. S., Ullmann, M., Vons, V. A., Lafont, U., Schmidt-Ott, A.: Generation of nanoparticles by spark discharge. J. Nanopart. Res. vol. 11 , 315-332 (2009)).

Thanks to its versatility, cost-effectiveness and scalability to industrial sizes, nanoparticle generation utilizing electrical discharges stands out even from among the gas-phase methods. This method allows for the generation of high-purity nanoparticles practically from any electrically conductive or semiconductor material, even on industrial scales, in a cost-effective and environmentally friendly manner. (Stein, M., Kiesler, D„ and Kruis, F. E.: Effect of Carrier Gas Composition on Transferred Arc Metal Nanoparticle Synthesis. J. Nanopart. Res., vol. 15, 1400 (2013)).

The electric discharges applied for generating nanoparticles are by way of example spark discharges (Schwyn, S., Garwin, E., Schmidt-Ott, A.: Aerosol generation by spark discharge. J. Aerosol. Sci., vol. 9, 639-642 (1988)). During the discharge event, the conditions for particle formation are created in a controlled (typically, atmospheric-pressure inert) gas environment, in the plasma situated between two electrodes, by removing material from the electrodes. Under the effect of a gas flow, atoms of the electrode material leave the plasma and, during a cooling-off process that involves coalescence and aggregation stages, form nanoparticles. By modifying the concentration of the removed electrode material, the size of the generated particles can be adjusted on a wide scale, from atomic clusters through spherical nanoparticles with a diameter of a few nanometers to particle agglomerates measuring tens or hundreds of nanometers (Pfeiffer, T. V., Feng, J., Schmidt-Ott, A.: New developments in spark production of nanoparticles. Adv. Powder Technol., vol. 25, 56-70 (2014)).

In spark discharge generators (SDG) a repetitive spark is created by a high- voltage (typically a few kV) power supply charging a capacitor (having a typical capacitance of a few nF) that is connected to the anode and cathode made from the desired nanoparticle base material (because polarity is periodically changing during the discharge event, no anode and cathode exist in general, only an“initial” anode and cathode). When voltage between the electrodes reaches the breakdown voltage of the gas in the generator, the capacitor discharges, generating a locally high current density at the surfaces of the electrodes, which causes material removal from the electrodes (Tabrizi, N. S., Ullmann, M., Vons, V. A., Lafont, U., Schmidt-Ott, A.: Generation of nanoparticles by spark discharge. J. Nanopart. Res, vol. 11 , 315-332 (2009)).

The discharge events are generated in the spark discharge generator without considerable external control, which is favourable from a technical aspect as it allows for simpler implementation and operation. Discharge plasmas carrying out the atomization of the electrode material, are, however, relatively sensitive to external conditions (for example electrode distance, temperature, fluctuations of the operating parameters of the power supply, gas environment, etc.), which may lead to fluctuations in the properties of the generated particles, to stability problems, and occasionally to the shutdown of the generator. Accordingly, in order to provide a particle yield at uniformly high quality and adequate quantity there is a need for the real-time monitoring of the spark discharge plasma constituting the central element of the generators. Besides that, it is also an important phenomenon that the discharge plasmas present in nanoparticle generators emit a significant amount of electromagnetic radiation, which makes data acquisition required for the above mentioned monitoring process very difficult.

II.2. OBSERVING OF A SPARK DISCHARGE GENERATOR - PRIOR ART

In the work of A. Kohut et al. Characterization of a copper spark discharge plasma in argon atmosphere used for nanoparticle generation” (Plasma Sources Sci. Technol. vol. 26, 045001-045014; 2017), a spark discharge applied for nanoparticle generation in laboratory is examined through spatially and temporally resolved optical emission spectroscopy (OES) complemented by fast imaging. The spark discharge is established applying high voltage between electrodes arranged in a hermetically sealed discharge chamber, and the discharge is examined by way of optical signals leaving the discharge chamber through a transparent window of the discharge chamber facing the spark gap between the electrodes (the location of the plasma discharge) applying a spectrograph and a high- sensitivity imaging CCD (charged-coupled device, a widespread optical sensor type) sensor. To achieve that, the optical signal leaving the apparatus through the transparent window is simultaneously routed (by means of a semipermeable mirror) to the spectrograph and the CCD sensor, the output signals of the two examining devices are evaluated together. In the study it has been found that the nanoparticle generation process can be considered to start during the afterglow phase of the spark discharge, so the spectroscopic examination of this phase could prove a useful tool for determining the initial conditions of the nucleation and growth models intended to describe the production of nanoparticles. A similar arrangement is described in another article (J. M. Palomares et al.: A time- resolved imaging and electrical study on a high current atmospheric pressure spark discharge, Journal of Appl. Phys. vol. 118, 233305 (2015)) of the researchers.

The study by A. Kohut et al.„From plasma to nanoparticles: optical and particle emission of a spark discharge generator” (Nanotechnology vol. 28, 475603, 2017) addresses the non-invasive optical examination of particle generation in a spark discharge. More particularly, the authors propose to gather information on the particles generated in the spark discharge utilizing time-resolved optical emission spectroscopy. They investigate particularly the effects of the charge current applied for generating the discharge, and of the size of the spark gap distance on copper (Cu) and gold (Au) electrode loss. The results of the study indicate, on one hand, that the concentration of the metal vapour forming inside the spark discharge generator can be increased only to a threshold level by increasing the energy of the spark, above which saturation occurs. On the other hand, based on the results of the study, an estimate can be given for the temporal initiation of nanoparticle formation during a spark event, and it can be established that the peak concentration of the generated nanoparticles and the modus of their particle size distribution can be increased linearly by increasing the repetition frequency of spark ignition (“spark repetition rate”; notation: f _rep ). Based on that, a clear, strong correlation can be detected between the optical emission from the plasma discharge and the properties of the nanoparticles generated in the spark discharge generator.

The approaches presented hereinabove are investigation methods applicable in relation to nanoparticle generation in a laboratory environment, as well as to follow it. In laboratory environments, the nanoparticles to be produced can usually be generated in very small quantities. Large scale nanoparticle production is possible in an industrial environment simultaneously applying a large number of nanoparticle generators. In the majority of cases, these generators are of identical type, but the simultaneous application of different-type nanoparticle generators can also be preferred in many cases.

Another arrangement adapted for examining aerosol particles utilizing a spark discharge detector is disclosed in US 9,140,653 B2. In view of the known approaches, therefore, there is a demand for a method and an apparatus that can be applied for continuous following and further optimizing and stabilizing of nanoparticle generation in spark discharge generators based on information originating/acquired from the spatial location of the spark discharge events. Monitoring methods and apparatuses applicable for such purposes are currently not known.

III. DESCRIPTION OF THE INVENTION

The primary object of the invention is to provide a method and apparatus which are free from disadvantages of prior art solutions to the greatest possible extent, and are capable of monitoring a particle generation process.

In light of the above, the object of the present invention is to provide a method and apparatus that enable the continuous following of nanoparticle generation in spark discharge generators, while the discharge can be further optimized or stabilized based on information gathered from the location of the spark discharge.

A further object of the invention is to provide a method and apparatus that can provide the optimization and stabilization of nanoparticle generation in certain cases even in industrial environments, and can provide for the simultaneous operation and monitoring of multiple spark discharge generators, even of different types. This also helps the cost-effective parallelization and quantitative upscaling of the nanoparticle generation process.

The objects of the invention can be achieved by the method according to claim 1 and the apparatus according to claim 12. Preferred embodiments of the invention are defined in the dependent claims.

The invention basically is a method and apparatus that are capable of detecting the intermittent or prolonged disturbances in the operation of one or even more spark discharge (nano)particle generators operated in parallel, and of issuing the appropriate warning or error signals to the operators of the system. Such a method and apparatus are not contained in the above referenced documents, and are also not utilized in the industry despite a serious demand for them, because presently the generators are monitored only by simple visual inspection or occasionally by measuring some electrical parameters, which is not acceptable for industrial applications.

IV. BRIEF DESCRIPTION OF THE DRAWINGS

Preferred embodiments of the invention are described below by way of example with reference to the following drawings, where

Fig. 1 schematically illustrates an assembly of an embodiment of the monitoring apparatus according to the invention and a spark discharge generator, with the apparatus being connected to the generator,

Fig. 2 is the schematic block diagram of a single-channel embodiment of the apparatus according to the invention that is capable of monitoring the operation of a single spark discharge generator,

Fig. 3 is the schematic block diagram of an n-channel embodiment of the apparatus according to the invention, i.e. a further exemplary embodiment capable of simultaneously monitoring n identical or different spark discharge generators,

Fig. 4 is the block diagram of an embodiment of the method according to the invention,

Fig. 5 is the block diagram of another embodiment of the method according to the invention,

Fig. 6 schematically illustrates an assembly of another embodiment of the monitoring apparatus according to the invention and a spark discharge generator, with the apparatus being connected to the generator,

Fig. 7 is the schematic block diagram of another single-channel embodiment of the apparatus according to the invention that is capable of monitoring the operation of a single spark discharge generator, and

Fig. 8 schematically illustrates an assembly of a yet further embodiment of the monitoring apparatus according to the invention and a spark discharge generator, with the apparatus being connected to the generator.

V. MODES FOR CARRYING OUT THE INVENTION V.1. INTRODUCTION TO THE DESCRIPTION OF THE METHOD AND

APPARATUS ACCORDING TO THE INVENTION The apparatus according to the invention is adapted for monitoring a spark discharge particle generator. The primary function of the apparatus is therefore monitoring. It performs this function particularly effectively thanks to the built-in learning capabilities directed at adapting the apparatus to a particular spark discharge particle generator (applying a reference acquisition/learning time period - see below for details).

This learning phase enables the monitoring apparatus and method to adapt its operation to a given generator (or even an n-member set of generators) without there being a need for the operator to manually enter an overly large amount of data or to calibrate the apparatus. In addition to improving ease of use, this also reduces the possibility of human error, and thus generators with even significantly different architectures and operation can be monitored simultaneously. It is to be noted here that it is even possible to add or remove generators to or from the monitored set of generators dynamically, as commanded by the operator during operation. This can be achieved by designing the monitoring software in such a way that it can receive such individual operator commands; i.e. the learning phase is interpreted by the software separately for each“channel” as being initiated when the given generator is started. These solutions collectively provide high flexibility, scalability, uninterrupted production, and cost effectiveness for the user.

Fig. 1 shows an apparatus assembly of a generic spark discharge particle generator 50 and an embodiment of the monitoring apparatus according to the invention denoted by a reference number 100, with the apparatus 100 (preferably portable, and it can also be called an appliance) being connected to the particle generator 50. The particle generator 50 does not form a part of the apparatus 100 but can only be connected to it, the apparatus being adapted for monitoring and optionally, in a possible embodiment, for controlling the generator (see below).

In Fig. 1 , the particle generator 50 is indicated by a curly bracket, and the apparatus 100 is indicated by a dashed line. It is noted here that in Fig. 1 those components of all components shown in the figure are encircled by a dashed line that form a part of the apparatus 100. The apparatus 100 is also shown in Fig. 2; in this figure, further components of the apparatus 100 are also shown. In relation to Fig. 1 , a particle generator 50 that can be considered to have a generic configuration for which the apparatus according to the invention can be utilized, is described below. The particle generator 50 comprises a gas-tight discharge chamber KK, in which - usually opposite each other - a first electrode E1 (for example, at the start of the discharge, anode) and a second electrode E2 (for example, at the start of the discharge, cathode) is arranged (the reference signs are shown beside the portion of the corresponding electrode extending into the discharge chamber). The electrodes E1 , E2 are connected to electrode positioning devices (units) E1 and E2, respectively.

A gap forming a spark gap SZK is provided between the electrodes E1 , E2; the effective longitudinal extension of the spark gap SZK (i.e. its extension between the electrodes E1 , E2) can be adjusted/controlled by means of the electrode positioning devices EP1 , EP2 in line with the loss of material the electrodes E1 , E2 are subjected to during the spark discharge. Accordingly, the electrode positioning devices EP1 , EP2 are configured as being adapted for applying a feed to the electrodes E1 , E2 if required. The electrode positioning devices EP1 , EP2 preferably enable the gas-tight introduction of the electrodes E1 , E2 into the discharge chamber (through the wall thereof).

In one of the modes that will be described below (by analyzing primary indicators or even by utilizing image information) the monitoring apparatus is able to detect and signal the need for applying electrode feed in order to keep the spark gap distance at the desired value. This information is output by the monitoring apparatus preferably together with the required amount of feed. The feed can be performed by issuing a control signal from the monitoring apparatus having control option (for a detailed description, see below), however, feed can also be applied without including the control option, for example by external intervention.

The free ends of the electrodes E1 , E2 situated outside the discharge chamber KK are connected to the conductive plates of a (high-voltage) capacitor NK, and the conductive plates of the capacitor NK are also connected to a capacitor charging unit KT for charging the capacitor NK by the required amount (in Fig. 1 a grounding GND is also shown). Besides that, the discharge chamber KK is also connected to a gas control unit GK (gas handling unit or subsystem) via a gas line adapted for providing gas inflow (this is illustrated by a line interconnecting the discharge chamber KK and the gas control unit GK), that is, the gas control unit GK provides for filling up the discharge chamber KK with the desired gas through said gas line, and for discharging the nanoparticles generated (produced) during the spark discharge from the discharge chamber through said gas line (filing up the discharge chamber with gas and discharging the nanoparticles with the gas are not performed applying the same gas line, this, however, is not intended to be illustrated in the schematic drawing of Fig. 1 ). As illustrated in Fig. 1 , the gas handling unit GK belongs to the particle generator 50, i.e. it is not arranged in the region encompassed by the line surrounding the apparatus 100.

In Fig. 1 there is schematically illustrated the connection of a light collecting optics FO (for example a collecting lens or a collimator lens KL) to the particle generator 50. In Fig. 1 it is illustrated that the optics (lens) is placed on the optical window OA that is formed in the wall of the particle generator 50, and which is of course transparent in the appropriate wavelength range (and of course forms a part of the particle generator 50; Fig. 1 illustrates that the spark gap SZK can be seen through the optical window OA). From the light collecting optics FO, a first optical conduit 10 (waveguide) leads to an optical input OB of the apparatus 100 (it is shown in Fig. 2 that the optical conduit 10 is implemented applying fiber optics 45). It is also shown in Fig. 2, that the optical input OB also comprises a first optical input OB1 (and, in a similar embodiment, an optical input OB2, see Fig. 7), at which the first optical conduit 10 coming from the light collecting optics FO is terminated.

Therefore, in this embodiment a monitoring optical detection device comprising a spectrometer unit is applied. An optical detection device comprising one or more optical detector each selective to a given wavelength can also be applied. A spectrometer is essentially none other than a multitude of wavelength-selective optical detectors; however, the one or more detector (each selective to a given wavelength) applied as an alternative would only select one or more (for example, a maximum of five) given wavelength, i.e., a narrow wavelength window. Monitoring could therefore be implemented by not recording full spectra but monitoring light intensity only at certain well-selected fixed wavelengths utilizing individual detectors/sensors. In the embodiments comprising a spectrometer, the spectrometer can be replaced by a wavelength-selective optical detector, as well.

The particle generator (for example a particle generator 50) subjected to monitoring comprises a discharge space and an optical window facing it, and, in the course of the method, the spectrometer unit is connected to the particle generator through a first optical conduit (for example, an optical conduit 10 or fiber optic) by means of light collecting optics arranged at the optical window of the particle generator. In embodiments comprising a spectrometer unit, the monitored data (to be introduced later on) comprise spectral data recorded applying the spectrometer unit (see also in the tables below).

In Fig. 2 further subcomponents of the monitoring apparatus 100 are shown. A communication module KM shown at the bottom of the figure is adapted to provide an external communication link (to an operating personnel) over which the maintenance or control requirements of the generators can be directly or indirectly communicated. The connection options offered by the communication module KM preferably comprise visual and audio signals, sending raw or processed sensor data to external computers, and/or optionally sending direct control signals. The details of the optional maintenance requirements and the exemplary control options will be described in detail below.

In Fig. 2, a single-channel exemplary embodiment of the monitoring apparatus (an apparatus 100) according to the invention is shown schematically. A light signal is received into the apparatus 100 via the fiber optics 45 from the particle generator 50 (see Fig. 1 ). The input of the fiber optics 45 facing the particle generator 50 is adapted to monitor the spark gap SZK of the particle generator 50 through the light collecting optics FO (shown in Fig. 2) that is to be arranged at the side of the particle generator 50 (that is, the optical signal imaged by the light collecting optics FO is fed to the fiber optics 45). Inside the housing of the apparatus 100 (enclosed by a dotted line in Fig. 2) the fiber optics 45 bifurcates (splits) into a first and a second optical fiber 20a, 20b as it extends from the optical input OB1 towards the inside of the housing, which fibers constitute a bifurcated fiber optics. In other words, the selected light signal is transferred very effectively by the first optical input OB1 shown in Fig. 2 (and similarly by the optical multiplexer OM shown in Fig. 3) to a two-forked (bifurcated) fiber optics. The optical fiber 20b is connected to a fiber optics input In1 of a (fast) photodiode unit FD arranged in the apparatus 100, and the optical fiber 20a is connected to an input In1 of the spectrometer unit SP.

In this embodiment, therefore, in the course of the method (in the method), a photodiode unit and a trigger unit adapted for providing to the spectrometer unit a trigger signal corresponding to the start of a discharge event in the particle generator 50 based on the signal thereof is applied, the first optical conduit has a first section implemented as a common optical fiber (in the embodiment described above, an optical fiber 45), and a second section branching off to a first optical fiber leading to the spectrometer unit (see the optical fiber 20a above) and to a second optical fiber leading to the photodiode unit (see the optical fiber 20b above), and an optical delay unit is inserted into the first optical fiber 20a for enabling that the spectrometer unit SP to be brought in a measuring-capable state by the trigger signal by a time of receiving a signal transferred over the first optical fiber 20a.

As it is illustrated in Fig. 2, the apparatus 100 utilizes the output signals of the following sensors (some of the considered sensors are arranged at the outside surface of the wall of the discharge chamber KK of the particle generator 50, while other sensors are arranged in the housing of the monitoring apparatus 100 in accordance with the fact that the particle generator 50 does not form a part of the monitoring apparatus according to the invention; in the list below the term “internal” refers to that; in the embodiment shown in Fig.7, external units, i.e. units arranged outside the housing are also applied):

- photodiode unit FD (internal, fiber optics connection to the generator);

- spectrometer unit SP (internal, fiber optics connection to the generator).

The photodiode unit FD shown in Fig. 2 comprises a photodiode (preferably with short rise time and high sensitivity) and an impedance matching electric circuit (transimpedance amplifier). The - preferably fast and sensitive - photodiode and the included electronic circuitry enable the photodiode unit FD to perform very fast electric processing of the incoming light signals. The latter unit is known in this technical field and is required for performing readout of the fast photodiode. Its function is to match the high-impedance (very low-loadability) output signal to the much lower input impedance of the data acquisition units such that it also provides low-distortion amplification of the output signal of the photodiode over the required broad frequency range.

The applied photodiode preferably operates in the visible and ultraviolet light ranges. The photodiode unit PD does not typically comprise a colour filter or other wavelength-selective optical elements (though it may comprise). The photodiode detects the rate of the flashes related to the spark discharge, and its electronic preferably also allows for sensitivity adjustment. Because there are many different excited species present in the spark gap, emitting light at different intensities and wavelengths, the light emission is detected by the photodiode as being continuous but (but variable-intensity) over the lifetime of an oscillating spark discharge. Therefore, the temporal length of the light signal (wavelength-integrated emitted light intensity) detected by the photodiode corresponds to the time interval corresponding to the complete depletion of the charge stored in the capacitor of the generator (the lifetime of the oscillating spark discharge, t _spark ), while its temporal shape is characteristic of the given generator and the operating conditions thereof.

An electric output signal (Out1 ) of the photodiode unit FD is fed to two further units, i.e. to the input In1 of a(n electronic) trigger unit TR, and to the input In1 of a(n electronic) data acquisition unit (see below for the further inputs).

The signal shape (so-called profile) fed to the input In1 of the data acquisition unit AD is digitized at a high speed (at least tens of MHz) by a high-speed A/D converter, and is stored in the buffer storage thereof. The signal profile is passed on to the central computer SZ through the connector Out/ln1 of the data acquisition unit AD via a bidirectional digital connection (there is a bidirectional connection between the data acquisition unit AD and the central computer SZ; the central computer SZ communicates through its connector Out/ln1 with the data acquisition unit AD). Thus, in this embodiment, the central unit (to be described in detail later on) of the apparatus according to the invention is implemented as a central computer SZ. In general, the central unit can be implemented applying any such unit that is capable of performing the functions detailed below. These functions can be performed even by a single device, but relatively separated units can also be applied for performing them.

As illustrated in Fig. 2, the trigger unit TR of the apparatus 100 has two digital electronic outputs. One of the output signals (Out1 ) is a trigger signal (to the spectrometer unit SP), the rising edge thereof indicating to the spectrometer unit SP that it may start recording of the emission spectrum, because another oscillating spark discharge has started. The spark discharge events are effectively temporally shifted from each other. The parameter f _rep („spark repetition rate”) specifies the number of oscillating spark discharges occurring per second. While the value of f _rep is usually 10-1000 Hz, the oscillations have a frequency of 1-10 MHz. In essence, therefore, the data acquisition rate of the spectrometer unit SP is synchronized to the spark discharge rate by the trigger signal. The trigger signal has a fixed length (for example 0.5-1 ms, which corresponds to the range specified for frep) that is safely sufficient for detection by the spectrometer unit SP, but at the same time does not hinder monitoring of the spark discharges with high repetition frequency.

Another output signal of the trigger unit TR (Out2, towards the data acquisition unit AD, this signal is fed to an input In2 of the data acquisition unit AD) is also a pulse signal, with the length of the pulse carrying the information: it indicates the temporal length of the spark discharge event (the above defined t _spark value; this is typically a couple of times 10 ms), i.e. information for the data acquisition unit AD related to the time instant when the relevant signal portion appears. The trigger unit TR therefore generates such electric (trigger) signals at its outputs that indicate start, lifetime and frequency of the light signal of the generator unit, as well as the temporal trend of the intensity of the light signal.

The digital output signals of the trigger unit TR are also fed to digital inputs In2 and In3 of the data acquisition unit AD. A digital counter is located at the digital input In3 of the data acquisition unit AD (the lowermost of the three inputs) that is adapted to determine, by counting the digital pulses received from the output Out1 of the trigger unit TR (i.e. trigger events indicating spark discharge events), the repetition frequency (f _rep ) of the spark discharge events, and stores it in its buffer storage. This frequency specifies the number of discharges per second of the capacitor arranged in the discharge circuit of the generator. The electronics behind the digital input In2 (the middle input) of the data acquisition unit AD is adapted to determine the length of the pulse received from the output Out2 of the trigger unit TR ( t _spark ), and to store the determined value in its buffer storage. The acquired and stored data (f _rep and t _spark ) are transferred by the data acquisition unit AD applying a bidirectional data link through the connector Out/ln1 to the central computer SZ (through the connector Out/ln1 of the latter) upon a request received therefrom.

The spectrometer unit SP of the apparatus 100 receives an optical signal at its (optical) input In1 from the input of the fiber optics 45 through the optical fiber 20a (optical fiber branch). The spectral data is acquired by the spectrometer unit SP are made for the central computer SZ (Out/ln2) with which it has a permanent data connection through the connector Out/ln1 , and the data acquisition process is started by the spectrometer unit SP - as it was described above - upon receiving a trigger signal at its input In2 (the spectrum of each spark discharge is recorded separately and passed on to the central computer SZ). The spectrometer unit SP has a permanent bidirectional digital electric connection to the central computer SZ, the central computer SZ passes on the recording parameters to the spectrometer unit SP and receives the recorded spectrum from the spectrometer unit SP via this link.

The optical fiber 20a may optionally comprise an optical delay unit (e.g. an optical fiber having a length corresponding to the delay time; for example, an optical fiber section having a length of 30 m - preferably wound up for saving space - causes a delay of 90-100 ns) that is adapted to achieve that the light signal is subjected to a delay sufficient for the spectrometer unit SP to enter a measuring-capable state after detecting the arrival of the starting signal from the trigger unit TR.

The wavelength range of the spectrometer unit SP is preferably the UV-visible (and/or near infrared) range, particularly at least the range of 200-600 nm, expediently with an optical resolution of 0.1 nm or better. The spectrometer unit SP preferably comprises a semiconductor serial detector and a readout electronics that allow it to record the emission spectrum of the spark gap in a wide range, very quickly, with high resolution, and without utilizing any moving parts. The detector can for example be implemented as a PDA (photodiode array), CCD (charge- coupled device), or CMOS (complementary metal-oxide-semiconductor) electronic quantum photodetector. The sensitivity and the electronics of the detector preferably allow for applying an integration interval in the ms range, and a wake-up time of 1-100 ns.

The initial programming of the spectrometer unit SP is performed by the control software of the central computer SZ before starting the data acquisition process, in part taking into account the data acquired during the learning time period, since the spectrometers typically have a built-in microcontroller (microcomputer) to which the data acquisition parameters have to be sent in advance, after which the spectrometer usually records the spectral data without further intervention at the specified rate, and forwards them on to the central computer SZ.

The apparatus can be called an “opto-electronic status monitoring apparatus (appliance)” that is adapted for monitoring spark discharge-based nanoparticle generators. The subassemblies of the apparatus are therefore preferably arranged in an electrically shielded housing for providing protection against environmental electromagnetic disturbances, and the apparatus has only optical connection to the nanoparticle generators (hereinafter simply referred to as“generators") to be monitored.

The apparatus (device) is capable of continuous operation in case the power connector (e.g. arranged on the front panel) is connected to the mains supply (in an example, all the connectors and other components led out to the side of the apparatus are led out at the front panel of the apparatus). To provide for disturbance-free and independent operation, the apparatus preferably comprises a built-in battery power supply that enables independent operation for a time period of expediently more than 24 hours; the battery of the power supply preferably automatically starts charging when the power connector is connected to the mains supply. According to the above, the invention therefore relates to a method for monitoring a spark discharge particle generator adapted for generating particles, preferably nanoparticles, and to an apparatus for carrying out the method. The method and apparatus are applied for following (monitoring) the events going on in the particle generator, and optionally, in an embodiment of the invention, for intervening in the particle generation process by feeding control signals to the generator.

The spark discharge particle generator itself does not form a part of either the method or the apparatus according to the invention, the method and apparatus according to the invention being only applied for the particle generator; in the case of the apparatus according to the invention preferably such that the apparatus is connected to the generator by means of suitable connections in order that the above described functions can be performed applying the monitoring apparatus according to the invention. Monitoring is carried out by the method according to the invention, expediently applying the apparatus according to the invention.

An embodiment of the monitoring method according to the invention is illustrated in the flow diagram of Fig. 4. In the course of the method according to the embodiment illustrated in Fig. 4:

In step S320 (the steps of the method may also be called operational steps) during the generation of the particles in the particle generator (for example in the particle generator 50 shown in Fig. 1 ), the value of monitored data by means of a monitoring optical detection device adapted for investigating generation of particles (typically at a specific sampling rate; the process start event and the optionally included stop event - see below - also indicate that the values are recorded continually) is recorded, and one or more primary indicator is derived from the monitored data. For that, it is necessary to initiate the spark discharge- based particle generation process in the generator by carrying out the steps for starting the generator (see also below; certain aspects of starting the generator have already been addressed), even immediately prior to start the monitoring process.

According to the above, monitoring also comprises applying a sensor based on an optical principle, tracking (following) the temporal progress of certain parameters, i.e. recording said data utilizing the sensor, followed by generating one or more primary indicator (selected, prioritized indicator, i.e. indicators applicable for monitoring) from the data. Primary indicators have a special role during the monitoring process (examples for primary indicators are included below). For example, a monitoring apparatus may provide feedback to the user if a primary indicator has an unwanted value, i.e. by watching the values of primary indicators during the monitoring process the user can make sure that the process is running normally (also these are the parameters that are modified in the case of a necessary intervention and the ones applied for controlling the process).

As it was mentioned above, not only the primary indicators play a role in the course of the monitoring method, but - in a manner described hereinabove - the so-called “secondary indicators” (non-selected, non-prioritized indicators, i.e. indicators that have a less direct role in monitoring than primary indicators, they may also be termed e.g.“basic indicators”). In principle, these can also become primary indicators, but typically they are other parameters that are not directly being monitored, i.e. their values are only logged such that changes of their values can be found later. According to the above, a corresponding reference value is assigned to each primary indicator, but the secondary indicators do not have associated reference values because their values are not tracked.

In step S330, in a reference acquisition time period (warm-up or learning period) following (after) a start of the recording of the monitored data, one or more reference value is determined for the one or more primary indicator, respectively. The reference acquisition time period is therefore adapted for identifying the reference values of the primary indicators, in other words, for training the monitoring apparatus for the typically occurring values of the given parameters in the case of the currently tracked particle generator. This is important because during the monitoring method (in course of, i.e. while the monitoring method is performed) these reference values are utilized for performing certain monitoring tasks (logging, establishing a warning event) and, in a further embodiment, for the decision on the necessity and manner of control intervention (see below, for example in relation to Fig. 8). !n a production time period following the reference acquisition time period, in a (continuously executed, i.e. continuous) step S340 it is checked, in step S350 according to (as per) a checking time interval (each time a checking interval has elapsed; the checking time interval can be a longer, predetermined checking time interval, or can even be chosen to be very short, essentially implying continuous error monitoring) whether an error state (error event) can be identified (can be determined; i.e. has occurred) in the particle generator,

- in case an error state can be identified (it has occurred), an error signal is issued in step S355, and the recording of the value of monitored data (i.e. the monitoring of the particle generation process) is terminated (stopped),

- in case an error state cannot be identified, actual value of the one or more primary indicator (a respective actual value of each of the one or more primary indicator), or actual value and one or more earlier value thereof, or a weighted value calculated from actual value and one or more earlier value thereof is compared with one or more corresponding reference value in step S360 (i.e. according to the“or” options it is possible to choose that only the actual value is used for the comparison, or earlier values of the indicator are also included in the comparison, or the actual and earlier values are applied for calculating a weighted value, for example an average value, which is then used for the comparison), and monitoring of generation of particles is continued, identifying a warning event in case a difference between the value of one or more primary indicator and the corresponding reference value (the indicator values are of course compared with the reference values corresponding thereto) exceeds a first threshold value (this is checked in step S365; accordingly, we return to the particle monitoring step S340 along the“yes” or the“no” branch issuing from this checking step; the warning event preferably specifies the generator subassembly - if there are more subassemblies - or setting parameter that is likely to be responsible for the deviation; accordingly, in the“yes” branch of step S365 there appears a step S370 corresponding to issuing the warning event; usually even the over-the-threshold difference of a single primary indicator value from the corresponding reference value will lead to a warning event). According to the above, after starting the generation of the particles, recording of the values of monitored data is also started (and then the primary indicator values can be obtained for these values), and then, after the recording of monitored data has been started, the reference values are determined over a reference acquisition time period. Monitoring is continued even after the reference acquisition time period, periodically checking the monitored data for an error state. Recording of monitored data is stopped if an error state is detected (or, of course, if the monitoring process is stopped).

An error signal is generated in case the generator needs to be shut down and needs to be inspected/serviced by the operator (an operator personnel must intervene) for reasons of safety or damage prevention. If necessary, error state indications can also result in the automatic shutdown of the generator. Signals corresponding to the warning events indicate minor deviations from normal operational parameters that do not necessitate immediate intervention. Upon system startup, the operator of course has to specify to the monitoring system the tolerance limits of the parameters (normal operation, warning and error limits) and the applied priority order of the parameters (for easy handling and for adaptive operation, from the aspect of programming this implies that the parameters can be selected from lists and the tolerance limit values can be specified in %).

Fig. 5 shows the flow diagram of another embodiment of the method according to the invention that can be carried out applying the monitoring apparatus according to the invention. According to Fig. 5, the method comprises at least the following steps (in the case of some steps the operational steps of the given embodiment are described from the aspect of the operator; in the figure, the preferred (in many cases, required) execution order of the steps is specified by the arrows shown in the flow diagram):

First, in step S410, the monitoring apparatus is turned on by the operator. For that, first it is necessary to connect the optical sensors of the monitoring apparatus to the spark discharge generator to be investigated (by establishing a fiber optics connection). In step 415, the conditions and tolerances (threshold values) are entered to the monitoring apparatus (preferably to the software thereof) by the operator, i.e. the initial parameters are specified (for example, the material of the electrodes, the type of the gas applied in the discharge chamber, the tolerance levels (threshold values), etc.). This is preferably performed before starting the monitoring process, i.e., as indicated by the next step, before starting the generator, that is when the monitoring apparatus is already operating (making it possible to enter the data), but particle generation has not yet been started.

In step S420, the generator is started by the operator, preferably under operational conditions. Then, after turning on the generator, the operator starts the monitoring apparatus in“training” operation mode (the operating time period corresponding to this mode is the so-called “reference acquisition time period” or learning time period, see below).

After the monitoring apparatus has been started in learning operation mode, in step S425 measurement data (monitored data) are acquired by the monitoring apparatus over a fixed time period (learning time period, reference acquisition time period), from which the reference values of the indicators are calculated. On the one hand, this step of the present embodiment specifies that indicators are derived from measurement data (monitored data), and on the other hand it also specifies that during this initial time period reference values are specified (cf. steps S320 and S330 in Fig. 4). In other words, in this step, by processing the data acquired by the sensors, the apparatus determines and records the reference range of the generator’s operating parameters (i.e. it records the reference values of the primary indicators - as it can be seen in the figure, parameters recorded as reference values and the corresponding threshold values). The recording the reference values (selecting or specifying the values) is preferably performed only after the operation of the generator has stabilized.

In relation to the above it has to be noted that the switching on of the monitoring apparatus (unit) does not necessarily to be mentioned because the subsequent steps evidently can only be performed if the monitoring apparatus is switched on. Because the method does not explicitly include the application of the monitoring apparatus itself, it is sufficient to include the threshold value into the method according to Fig. 4. Such a case may in principle occur wherein the threshold value is available a priori, but it can also be specified/entered at the start of the process.

In step S435, the monitoring apparatus automatically switches to production mode, that is, the "live” (real) operation is started (i.e. after the reference acquisition time period has elapsed, the production time period is started), and in step S440 the measurement data (monitored data) are acquired, with the actual values of the indicators being established, that is, the parameters are continuously followed (tracked) during the operation of the generator.

At certain time intervals (preferably at time intervals - check time intervals - that are identical for the steps S445, S455 and S465, according to the hierarchy shown also in Fig. 5, but the question points occurring during these steps can also be checked at different rate) it is checked, in step S445, if the entire expected operating time period has elapsed, i.e. if the production and monitoring time period has ended. As soon as the operating time period has elapsed, in step S450 a normal shutdown of monitoring is performed (the desired length of the monitoring process is typically set by the length of the operating time period). If necessary, the parameter values stored as a function of time are exported, and a report is compiled.

Expediently, soon after the monitoring process has ended the particle generator is also shut down, because the monitoring operation time period typically also pertains to the length of the particle generation time period. It is expedient if monitoring is stopped sooner from the monitoring and the particle generation, otherwise the shutdown event of the particle generator (“plasma off’) would be entered into the monitoring error log.

If the entire expected operating time period has not elapsed, in step S455 it is investigated whether an error state can be identified (whether there is an error state). This step essentially corresponds to the step S350 of Fig. 4. In case an error state can be identified, in step S460 an error signal is issued, and monitoring is stopped. ln case an error state cannot be identified, in step S465 it is investigated whether there is an out-of-tolerance difference (a difference greater than the threshold value) between the indicator and the reference values. This step essentially corresponds to the step S360 of Fig. 4. During this step, therefore, a comparative statistical evaluation of the difference between the actual and the reference values of the operating parameters of the generator (i.e. between the actual value of the primary indicator and the reference value) is performed, and the parameter values and sensor data are continuously stored.

Irrespective of the result of step S465, i.e., independent of the result of the comparison of the actual value of the indicator and the reference value, execution returns to monitoring the particle generation process. Accordingly, the“yes” and “no” branches of step S465 both lead back to step S440, i.e. the acquisition of measurement data and the establishing of the actual values of the indicators. At the same time, however, as it is shown in Fig. 5, if an out-of-tolerance deviation is experienced (i.e. the deviation is greater than a threshold value), a warning signal S470 is issued (this essentially corresponds to establishing a warning event in step S370), i.e. the deviation is recorded and introduced among the logged data of the monitoring process (if the deviation is smaller, then no logged event will be generated).

As illustrated in Fig. 5, in case an out-of-tolerance deviation is detected between the indicator and the reference values, in addition to the issuing of the warning signal (before, after, or simultaneously with it), optionally, i.e. in a variant of the process of Fig. 5, in step S475, proportional control signals (amounting to a certain percentage of the signal) are derived and sent towards the generator. This step is optional, i.e. in the default case no other measures are taken in addition to recording the out-of-tolerance deviation (issuing and logging a warning signal), in accordance with the fact that in the default case, monitoring of particle generation is performed that is aimed at recording/logging out-of-tolerance deviations. However, as it is described in detail below, if a difference greater than the tolerance limit, i.e. than the threshold value, in an embodiment an intervention is also performed applying control signals in addition to recording/logging the event. Based on the temporal trends and the relative magnitude of the acquired sensor data, the detailed error state can also be established. The error states that can be signalled by the software include, but are not limited to, the following:

^■ The discharge has stopped - This is the most important error state from the aspect of operating the generators, its detection is based on multiple independent measurement data: 1 ) the spatially and spectrally integrated emission signal of the discharge has dropped below the minimal detectable value (essentially, the signal is lost); 2) the emission spectrum of the discharge does not contain the spectral lines that were recorded during the learning phase or that are characteristic of the material system specified by the user; 3) the cameras detect that there is no discharge channel.

■ The discharge is unstable - The discharge has not stopped (i.e. none of the criteria formulated in the above point is fulfilled) but the spatially integrated and camera-recorded indicators leave their ranges defined during the learning phase that correspond to normal operation. There occur such instabilities which can be restored by proper intervention and control; it is described in detail below which one of the possibilities for intervention (for example, actuation of the electrode positioning device, intervening into the charging of the capacitor, control of the chemical composition) can be resorted to.

The material quality of the gas is not as intended - The emission spectrum of the discharge does not conform to the reference spectrum (reference values) recorded during the learning phase and/or to the values specified by the user that are characteristic of the experimental setup.

■ The material quality of the electrode material is not as intended - The emission spectrum of the discharge does not conform to the reference spectrum recorded during the learning phase and/or to the values specified by the user that are characteristic of the experimental setup.

■ The electrode material is running out - A camera (see below) detects an increase of the distance between the electrodes (preferably, an intervention can be carried out also in this case, by actuating the electrode positioning device).

The position of the discharge plasma channel is not as intended - The camera detects a change in the position of the discharge channel and/or the emission spectrum contains specific emission lines (specified during configuring the software) characteristic of components of the discharge chamber other than the electrodes (in the case of spark discharges, usually the stainless steel wall of the discharge chamber).

^■ Gas pressure is not constant - It is possible that there is a gas leak inside the discharge chamber.

A question concerning the details of the control algorithm of the apparatus is how much of the information previously entered by the user is used (for example, the material of the electrode, the gas material, etc.). The control software can of course be written such that it“believes” everything it is presented with during the training time period, in this case, however, the possibility of detecting malfunctions caused by user error (for example, the user would like to generate Cu nanoparticles applying Cu electrodes, but the technician accidentally installs Ni electrodes; if during training the control software records the spectrum it detects, then it cannot give a warning on the incorrect electrode material).

V.2. EXTENSION OF THE METHOD AND APPARATUS ACCORDING TO THE INVENTION TO THE SIMULTANEOUS MONITORING OF MULTIPLE GENERATORS

Another possible embodiment of the monitoring apparatus according to the invention is adapted for the simultaneous monitoring of n (n is a positive integer) particle generators having an identical configuration. This embodiment is illustrated by an apparatus 200 shown schematically in Fig. 3. The apparatus 200 is therefore adapted for simultaneously monitoring a plurality of particle generators 50, i.e. it is an n-channel apparatus with multiple inputs, its sensors being read out cyclically (sequentially, in a time-sharing manner). This is an economical solution for the mass production of nanoparticles, where a large number of parallel- connected generators have to be monitored simultaneously.

An embodiment of the method is therefore adapted for monitoring a plurality of particle generators, recording, by means of the monitoring optical detection device, the value of monitored data in each particle generator. In this embodiment, the monitored data of the optical detectors each corresponding to a respective particle generator are preferably processed in a time-multiplexed manner (see below).

This operating regime can be realized by connecting the fiber optics inputs of the light collecting optics FO to the input of an optical multiplexer OM unit as shown in Fig. 3, which unit is controlled (i.e. selection among input channels) by a central computer SZ having a similar functionality as the computer described above in relation to the apparatus 100 through the Out/In connector (that is shown in Fig. 3 at the left side of the optical multiplexer OM unit and provides a direct, bidirectional connection to the central computer SZ) of the optical multiplexer OM unit, selecting the light signal of the next one of the generator units (connected to one of the inputs of the optical multiplexer OM unit) according to the sequence of the time- multiplexed operation. The central computer SZ controls the measurement cycles, starting and stopping the operation of other components, receiving, storing and evaluating the measurement data and displaying the results.

In the case of such an application, the apparatus according to the invention spends a few seconds reading each of the generators before moving on to the next one, so with a reasonable number of generators a cycle time of a few minutes or even lower can be provided. This allows for a more than satisfactory data refresh rate and error signaling rate, particularly if we take it into account that the operating malfunctions of the generators usually have to be fixed by humans (although automation is also a possibility here as well).

V.3. ADDING OPTIONAL SENSORS TO THE METHOD AND APPARATUS ACCORDING TO THE INVENTION

The assembly of a further embodiment of the apparatus according to the invention and the particle generator 50 (and thus certain aspects of the related monitoring method) are illustrated in Fig. 6, with the functional block diagram of the internal configuration of the apparatus 300 shown in Fig. 6 being shown in Fig. 7.

The arrangement is similar to the assembly shown in Fig. 1 , with the monitoring apparatus 300 having additional functionality compared to the embodiment of Fig. 1. In Fig. 6 it is illustrated that measurement data (monitored data) are supplied to the monitoring apparatus 300 - i.e. according to Fig. 6, to the electrical input EB interface thereof (as shown in the figure, this can preferably be grounded/shielded, or can be electrically separated utilizing an optocoupler) - by a current probe AM through a first electrical connection 22, by a voltage probe FM through a second electrical connection 24, by a camera optics KO through an optical conduit 12, and by a gas control unit GK through a third electrical connection 26. It is also indicated in Fig. 6 (in dashed lines) that in certain variants of the apparatus even one of them, some of them or all of these optional units adapted for supplying measurement data (a current probe AM, a voltage probe FM, a camera optics KO, and a gas control unit GK) can be included.

Data supplied by these units can be valuable in the course of the monitoring method for following the particle generation method for example according to the following: The camera optics KO can be utilized for tracking the changes in plasma morphology (shape, size, symmetry conditions, etc.), and even for tracking the changes of the electrode gap. The current meter (current probe) can be utilized for tracking the direction of discharge oscillations, for measuring the frequency of the oscillations, for deriving indirect plasma temperature information, etc. The voltage probe can be utilized for example for detecting the start and the end of the discharge event (i.e. for the trigger signal), etc.

In the embodiment shown in Fig. 6, in addition to the light collecting optics FO that is also shown in Fig. 1 a camera optics KO is also arranged on the optical window OA of the particle generator 50, which camera optics KO is adapted for supplying monitored data to an optical input OB of the apparatus 300 through a second optical conduit 12. This is shown in detail in Fig. 7, where it can be observed that the optical conduit 12 is implemented applying a fiber optics bundle 47.

The internal configuration of the monitoring apparatus 300 is illustrated in Fig. 7. In Fig. 7 an optional input adapted for connecting the camera optics KO is shown, and the other optional data supply units are shown in Fig. 6: a current probe AM, a voltage probe FM, and a gas control unit GK (unlike the other units, the gas control unit GK is shown in a continuous-line frame because in this embodiment it is not an optionally included unit, what is optional is whether it supplies data to the interface of the electrical inputs EB). These are all connected to the electrical inputs EB interface which in turn is in connection with (supplies data to) the central computer SZ.

A digital camera DK comprising an optical input OB2 is adapted for the continuous monitoring, with the help of the camera optics KO shown in Fig. 7, of the spark discharge occurring in the discharge chamber KK, more particularly in the spark gap SZK (see below for more details). The digital camera DK has a data communication connection to the communication module KM via the central computer SZ.

In addition to the apparatus 100, the apparatus 300 shown in Fig. 7 also utilizes the output signal of the camera objective KO (has an external optical connection, implemented by a fiber optics bundle, to the generator - in the following such an option is also described wherein the digital camera is applied for direct investigations through the optical window; the measures to be taken for that are described below). For that, in a manner illustrated in Fig. 7 an imaging-capable fiber optics bundle 47 has to be applied at the digital camera DK, and in such a case the digital camera DK is arranged in the housing (casing) of the monitoring apparatus 300 (like the connection of the fiber optics and the SP or FP, wherein the fiber optics functions solely as a light conductor). The fiber optics bundle 47 is required for transmitting image information, i.e. the image information is transmitted by the fiber optics bundle 47 between the camera objective KO placed on the particle generator 50 and the digital camera DK connected to the second optical input OB2 inside the housing.

According to the configuration of Fig. 1 , basically a single fiber optics - connected to the optical window (optical port) of the generator through a lens - extends between each generator and the monitoring apparatus (appliance), i.e. the fiber optics input ports thereof (the assignment of the ports to the generators has to be specified by the operator in the software, because only the operator has information on where the particular fiber optics are connected). The basic monitoring operation (see Fig. 1 ) therefore requires that a single fiber optics (only one optical port) has to be included for each generator. According to the above, in a manner illustrated in Figs. 6 and 7, optionally a second optical channel can be provided for each generator for monitoring it utilizing a digital camera; in which case a fiber optics is connected to the (second) optical window of the generator through an objective, but in this case the fiber optics is a fiber optics bundle (having as many individual fiber cores as the number of pixels to be recorded in the image). For the operation of this, a digital camera (either one camera for each generator, or a single camera utilized in a time-multiplexed manner) has to be disposed in the housing of the monitoring apparatus

The illustrated embodiment is therefore an embodiment wherein a monitoring optical detection device comprising a camera device is applied (the type of camera, i.e. an analogue or digital camera is generally not specified, but from a number of aspects it is more preferable to utilize a digital camera as a camera device). In the above embodiment, furthermore, the particle generator (for example a particle generator 50) subjected to monitoring comprises a discharge space and an optical window facing it, and, in the course of the method, the spectrometer unit is connected to the particle generator through a second optical conduit (for example, an optical conduit 12, e.g. fiber optids) bundle by means of a camera optics arranged at the optical window of the particle generator.

Due to the above described solution applying fiber optics bundles, as with the embodiment of Fig. 1 , the embodiment described above is also configured such that in the course of the method the monitored data, by means of the monitoring optical detection device, is acquired exclusively through optical conduit, particularly a fiber optics or a fiber optics bundle. In such a case, therefore, the monitoring system collects exclusively (only) optical signals (including spectroscopic signals) from the spark discharge generator(s), the optical signals being transmitted between the generator and the monitoring device by fiber optics (optical waveguides). This allows for an operation that is highly protected from external electrical interference (this is especially important in an industrial environment). This arrangement also allows for arranging the generator and the monitoring device at a significant distance from each other, and also enables easy “upscaling” of the system, i.e. monitoring n pieces of generators utilizing a single monitoring device.

According to the above, the monitoring system preferably comprises the fiber optics lines extending from the housing of the monitoring apparatus (the components from the optical/optomechanical connection to the generator: collimator lens and camera optics also belong to the monitoring apparatus, although they typically extend outside the housing thereof), data acquisition/storage and evaluation subsystems, decision-making algorithms, communication module, and power supply.

Through the connector Out/ln3, the digital camera DK and the second optical input OB2, the computer SZ - which is also included in the apparatus 100 - is also connected to the camera objective KO which also watches the spark gap. In accordance with Fig. 6 that illustrates the optional interconnections, this further monitoring of the particle generator 50 is optional, which is indicated by the dashed-line interconnection shown in Fig. 7. In Figs. 6 and 7 there are shown further dashed-line interconnections, all of which are optional. In accordance with the illustration such embodiments can be conceived wherein all dashed-line interconnections are included, while embodiments wherein only one or more of them are included are also possible.

During operation of the monitoring apparatus 100, the camera objective KO is mounted on the spark discharge generator (see Fig. 1 ), and is connected to the second optical input OB1 through a fiber optics bundle adapted for imaging purposes. The fiber optics bundle is adapted to galvanically separate the camera objective KO from the apparatus 100 in order that the electromagnetic disturbances occurring near the spark discharge generator cannot propagate into the monitoring apparatus (the fiber optics connections to the photodiode unit FD and the spectrometer unit SP are included for the same reason).

An embodiment wherein a digital camera DK is directly connected to the optical window OA of the discharge chamber KK can also be conceived.

The digital camera DK (digital camera unit) preferably has an at least 1 -megapixel resolution and a sensitivity that allows for taking images with sufficient contrast. Expediently (i.e. taking into account the aspects of sensitivity, resolution, cost and functionality), the camera is a monochrome unit, with an A/D converter with at least 12-bit resolution, and with a global trigger (electronic shutter). It preferably has such a speed and triggering functionality that allow for taking full-resolution images with a wake-up time and integration time on the order of a few ms (microseconds).

The optics of the digital camera DK is adapted for monitoring the entire spark gap (spatial region between the electrodes) with a contrast such that the spark gap fills the entire image (to achieve that it has to have an appropriate magnification, focal length, and depth of field). The application of a telecentric optics is preferable but not essential. Telecentric optics or lenses are widely used in industry, for example for monitoring products progressing on a production line. This type of lens has the advantageous characteristic that it has a constant magnification throughout the entire depth of field range, which minimizes distortions.

The format and size of the sensor of the digital camera DK is not of essential importance, but it has to match the above separation solution. The wavelength- sensitivity curve and optical transmission of the camera preferably allow for performing monitoring in the UV-visible range (at least in the wavelength range of 300-800 nm). Light with a wavelength outside this wavelength range can be prevented from entering the chamber expediently by applying colour filters. For functional reasons, this range is slightly different from the range typically applied in the case of the spectrometer unit SP (200-600 nm). The camera is primarily dimensioned to take visible - observable with the eyes - images, while the spectrometer has to function well in the UV and near-Vis ranges that have primary relevance as far as atomic spectroscopy is concerned. The other aspect taken into account is that cameras are much less sensitive in the UV range.

As illustrated in Fig. 7, the digital camera DK can be triggered (synchronized) from the central computer SZ, or optionally from outputs Out1 or Out2 of the trigger unit TR (this latter option is not shown in Fig. 7). The only difference between these two options is the programming of the digital camera DK, since through the output Out1 of the trigger unit TR a “start signal” is issued - basically towards the spectrometer unit SP -, and a pulse having a length corresponding to the required integration time of the digital camera DK being issued - basically towards the data acquisition unit AD - through the output Out2 of the trigger unit TR. For triggering spectrometers, a rising or falling (trailing) signal edge is usually required, indicating the beginning of integration (of course, in addition to the wake-up time). However, in the case of cameras it is often the length of the incoming pulse that defines the length of the period during which image information is received. The hardware and software of the two devices (SP and DK) ideally have identical functions, so it can in principle be provided that both signals are identical, but

1.) for the sake of generality it was important to treat this case separately, and

2.) it is sufficient to process the image much less frequently, as described below.

The digital camera DK is adapted for recording images of the spark discharge with an integration time that is longer than t _spark but is shorter than 1lf _rep „ i.e. it is adapted for taking a separate image, integrated over time and the wavelength range, of each spark discharge event. However, due to the nature of the information supplied by the digital camera DK it is not necessary to record the image of each spark discharge, i.e. it is sufficient to take an image intermittently, for example one image every second, or every minute. The digital camera DK can be programmed during the learning phase of the monitoring process (see below), taking into account the differences arising from the galvanic separation and from the different lengths of the fiber optics cables, the values of t _spark and 1/ f _rep , etc.

Taking into account also the optional route shown in Fig. 7, the light signals originating in the spark discharges can be preferably transmitted to the apparatus along two different routes. On the one hand, the light emitted by the spark discharge events is transmitted, preferably through fiber optics, to a fiber optics connector (first optical input OB1 ) arranged on the apparatus (preferably on the front panel thereof), and, on the other hand, the signal of the camera objective KO (that is mounted on the generator units and receives power supply from the apparatus) transmitting image information related to the spark discharge events is fed to a second optical input OB2 (for example also disposed on the front panel of the apparatus). For the fiber optics inputs (leading to the photodiode unit FD and the spectrometer unit SP) application of an objective is not required, i.e. the free end of the fiber optics in a suitable grip can be sufficient (as illustrated in the figure), while an actual imaging lens (camera optics KO) - adapted for projecting the image of the spark gap SZK to the fiber optics - is required to be installed at the end of the fiber optics bundle that faces the discharge chamber KK. The optical inputs OB1 and OB2 (collectively referred to as optical input OB) are connected to the particle generator 50, preferably in a manner shown in Fig. 6, through a light collecting optics FO and a camera optics KO. Monitoring optical detection devices monitoring particle generation (and recording the values of the monitored data) are connected to the particle generator 50 through the light collecting optics FO and also through the camera optics KO. The light signals of the first optical inputs OB1 are absolutely necessary for the operation of the apparatus.

In contrast to the arrangement of Fig. 7, in the embodiment of Fig. 3 no camera optics and digital cameras (arranged in the monitoring apparatus) corresponding to them are applied for the particle generator. The digital camera can also be included in the embodiment of Fig. 3. A multi-channel optical input for this could perform similar functions as the optical multiplexer OM by selecting the camera optics corresponding the light collecting optics assigned to the selected particle generator.

V.4. DESCRIPTION OF THE INDICATORS; ADDITION OF AN OPTIONAL CONTROL OPTION TO THE METHOD AND APPARATUS ACCORDING TO THE INVENTION

As we have pointed out above in relation to step S475 of Fig. 5, in an embodiment of the method, an intervention can optionally be made (by issuing control signals to the generator) if the difference between the value of an indicator and the reference value exceeds a threshold value (tolerance limit). At the investigation whether the threshold value has been exceeded involves checking the difference between the value under investigation (which can be a derived value, calculated for example by averaging multiple values) and the reference value, i.e. whether the difference is greater than a threshold; in other words, it is checked if the difference stays within a tolerance range.

In such embodiments, therefore, control tasks are also included in the monitoring process (in addition to logging the warning events, a feedback is also provided based on the monitored data and the indicators derived therefrom). Accordingly, in an embodiment, in case of identifying a warning event, a control signal is determined for and issued to the particle generator based on a difference between the one or more primary indicator and the reference value corresponding thereto exceeding a first threshold value (wherein the first threshold value is adapted for comparing the one or more primary indicator and the one or more reference value; in general, the threshold values specify the difference between the monitored values at which difference is regarded as significant, and also the difference at which a warning event is established when monitoring the first threshold value, and at which an intervention is made in the case of quantities applied in the control). The interconnections that also allow for control intervention are shown in Fig. 8.

The monitoring system (framework) therefore also includes the possibility of implementing control (feedback), because the decision-making algorithms utilized during monitoring also supply information based on which control signals can be derived. The feedback signals allowing for controlling are electrical, corresponding to the electronic sub-units that can be controlled utilizing them (optical separation can be applied for these signals). By its nature, the control process is dependent on the configuration of the generator, i.e. on the path and magnitude of the control signals that have to be/can be applied and also on the interfaces required; therefore, it may become necessary to adjust the monitoring process to the given generator by adjusting certain parameters.

In Fig. 8, an assembly - similar to the arrangements shown in Figs. 1 and 6 - of the particle generator 50 and a yet further embodiment of the apparatus according to the invention (a monitoring apparatus 400) is shown. In Fig. 6, there are shown additional data acquisition/data supply devices (components) other than the data acquisition device (component) based on the light collecting optics FO. In contrast to that, in Fig. 8 there can be seen such additional (compared to Fig. 1 ) electrical connections with the particle generator 50 through which certain controls (control tools; modules, sub-units) of the particle generator 50 can be adjusted (this is illustrated by the arrows pointing in the direction of the given modules). It is of course possible to combine the embodiments of Fig. 6 and Fig. 8, i.e. in an embodiment, the optional interconnections indicated by dashed lines can be incorporated from both former embodiments. ln the embodiment illustrated in Fig. 8, dedicated feedback connections lead to respective controllable modules of the particle generator 50: a first feedback connection 14a and a second feedback connection 14b lead to the electrode positioning devices EP1 and EP2, respectively, and a third feedback connection 16 and a fourth feedback connection 18 leading to the capacitor charging unit KT and the gas control unit GK, respectively (accordingly, in a conceivable embodiment the gas control unit GK may supply monitored data - indicating, for example, if there is a gas flow while the unit can also be subjected to control action, i.e. a potential problem with the gas feed can be addressed by control intervention, giving a reaction). In such a case, the (preferably optically separated) interface of the electrical outputs EK are connected to the control subsystems (the gas control unit GK, the electrode positioning devices EP1 , EP2, the capacitor charging unit KT) of the particle generator 50 for passing on the signals. It is noted that the interface of the electrical outputs EK preferably form a part of the central unit, while certain (sub)units of the latter can be implemented in the interface of the electrical outputs EK. In some embodiments, all of the feedback connections shown in dashed lines can be included, but optionally only one or more of these feedback connections can be included. In line with the above, the control options are described below, addressing in particular the control implementations in different embodiments.

In the following, the process of deriving, by means of the monitoring apparatus according to the invention (e.g. the apparatus 100, 200, 300, or 400), the operating parameters (primary indicators) from the sensor information collected in the course of the monitoring method according to the invention (e.g. the method according to Fig. 4 or Fig. 5) is described. The monitoring-control algorithm relies on input data (measurement/sensor data, user-specified constants) for deriving indicators. The indicators, and especially the 2-3 primary (prioritized, composite) indicators are the status indicators that collectively describe the status of the system.

Based on the above, an exemplary process of implementing the controllingmonitoring functionality performed by the apparatus and method according to the invention is the following: 1. First, the so-called“constants”, i.e. the basic operating parameters (see Table 2) are needed to be specified by the user. These constants are parameters that are kept at a constant value by external systems during the operation of the generator, as well as basic spectroscopic data characteristic of the components of the chemical system, etc. The“weight factors” (Si,, see Table 2 and the detailed description below) are also entered at this time. They are applied for determining the significance of certain primary indicator data (see below) during the control process.

2. The generator is started under such operating conditions and with such parameters (for example, electrode material composition, characteristics of the gas atmosphere, electrical parameters, etc.) that are appropriate for producing nanoparticles with the desired qualities (e.g. yield, size distribution, composition, etc.).

3. Utilizing its optical sensors, an embodiment of the apparatus and method according to the invention performs measurement and recording of the parameters and data vector listed below in Table 1 (the so-called "input data”, among them are the monitored data) during the reference acquisition time period by monitoring the operating generator (i.e. it also monitors whether the parameter values settle after an initial time period has elapsed).

4. Based on the reference values, the present embodiment of the method and apparatus according to the invention determines the optimal data acquisition parameters required during the subsequent continuous operation (for example tm.up, tm,dnp, t _update , etc., see Table 3), after which the continuous operation monitoring is started.

5. The input data are utilized by an embodiment of the method and apparatus according to the invention for calculating - taking into account theoretical relationships - derived parameters: the so-called secondary indicator data (see Table 4), and also the primary (selected) indicators. The values of these latter indicators, established during the “learning” time period are utilized as reference values. . After the data collecting time period t _update has elapsed (which expediently corresponds to the checking time interval) the actual values are calculated again. Based on the comparison of these data with the corresponding reference values, and on the actual values of the sensor input data, it is identified by the present embodiment of the method and apparatus according to the invention whether error phenomena (error state) have occurred in the generator (whether these can be identified). If yes, then it reacts by issuing error signals. If there has been no error phenomenon, but the actual values of the indicators still differ from the reference data, then such an output control signal is generated (taking into account the relevant weight values) for the driver units of the generator that “urges” the generator back towards the reference state.

7. Then, returning to point 6 above, the present embodiment of the method and apparatus according to the invention continues operation until it is shut down. Table 1 below comprises the description of sensor input (measurement) data (among them exemplary monitored data) that originate from the generator and are valid for the monitoring and control system.

Notation Description

PD(t) The wavelength-integrated, time-dependent intensity signal profile of the photodiode (corresponding to the photodiode unit FD or arranged in the phototrigger unit FT)

TRIG A digital pulse with a steep rising edge (trigger; a TTL [transistor- transistor logic] or CMOS level) derived from the PD(t) signal. The event (emission of the TRIG pulse) occurs at the time instant when the condition PD(t) ³ Ith,up is fulfilled for the first time (i.e. the signal of the photodiode exceeds a threshold value); its length is ttng (see

Tables 2. and 3).

SP(l, t) The spectrum acquired by the spectrometer unit SP. The spectral lines l _a,i,j , l _k,i,j , l _g,i,j preferably all fall into the measured wavelength range (UV-Vis[ible]; for the definition of the wavelengths see Table 2; these wavelengths are characteristic of the anode, the cathode, and the carrier gas). The sensitivity and spectral resolution have to be high enough to allow the measurement of these spectral lines in a disturbance-free manner and with a favourable signal-to-noise ratio. Data acquisition by the spectrometer is initiated by the TRIG event, so the data to be acquired by the spectrometer are not available if there is no spark discharge event due to an error d distance between the electrodes (electrode gap, spark gap)

P pressure of the gas environment in the chamber. Its value is considered to be constant.

u _br breakdown voltage value, at the given pressure, of the carrier gas located in the spark gap

U _ch charging voltage set on the capacitor charging power supply. Under operating conditions it is assumed that U _ch > u _br

l _ch charging current set on the capacitor charging power supply

Table 1

The monitoring and control algorithm requires the values of certain constants that have to be specified by the user at the start of the monitoring process. These parameters are listed below in Table 2:

Notation Description Typical

(initial) value l _a,i,j The j-th atomic or ionic spectral line of the i- th elemental

component of the anode material of the generator that

can be measured without disturbance, and with a

favourable signal-to-noise ratio, by the spectrometer. For

a reliable measurement, at least 2-3 such measurement

wavelengths are required for each component.

E _a,i,j The excitation energy of the j-th atomic or ionic spectral

line (with a wavelength of l _a,i,j ) of the i-th elemental

component of the anode material of the generator

A _a,i,j The transition probability of the j-th atomic or ionic

spectral line (with a wavelength of l _a,i,j of the i-th elemental component of the anode material of the generator

9 _a,i,j The degeneracy of the j-th atomic or ionic spectral line

(with a wavelength of l _a,i,j of the i-th elemental component of the anode material of the generator l _a,i,j The j-th atomic or ionic spectral line of the i-th elemental component of the cathode material of the generator that can be measured in an undisturbed manner, with a favourable signal-to-noise ratio, by the spectrometer of the assembly. For reliable operation, at least 2-3 such measurement wavelengths are required for each component.

E _k,i,j The excitation energy of the j-th atomic or ionic spectral line (with a wavelength of l _k,i,j of the i-th elemental component of the cathode material of the generator

A _k,i,j The transition probability of the j-th atomic or ionic spectral line (with a wavelength of l _k,i,j of the i-th elemental component of the cathode material of the generator

g _k,i,j The degeneracy of the j-th atomic or ionic spectral line

(with a wavelength of l _k,i,j of the i-th elemental component of the cathode material of the generator l _g,i,j The j-th atomic or ionic spectral line of the i-th elemental component of the carrier gas medium of the generator that can be measured in an undisturbed manner, with a favourable signal-to-noise ratio, by the spectrometer of the assembly. For reliable operation, at least 2-3 such measurement wavelengths are required for each component.

l _e,i,j The j-th atomic or ionic spectral line of the i-th elemental component of the atmospheric environment (air) of the generator that can be measured in an undisturbed manner, with a favourable signal-to-noise ratio, by the spectrometer of the assembly; it is necessary to measure

it in order to detect the potential leaking of the chamber.

For reliable operation, at least 2-3 such measurement

wavelengths are required for each component.

t _w Reference acquisition (warm-up) time period 5-20

minutes t _trig Length of the TRIG pulse; its value is expediently chosen < 1 ms

such that the trigger signal can be applied as a start

signal for all electronic sub-units of the apparatus. Its

value cannot be too long due to its interdependence with

other parameters, so for example the condition

has to be fulfilled at all times.

t _update Cycle time of the status check (data evaluation) process 1

is t _w = N . t _update - where N is an integer and is greater than minute

10.

S _i weight factors specifying the weight of the i-th primary 1

indicator during the control process (see below for the

introduction of the primary indicators)

Table 2

The value specified for the reference acquisition (warm-up) time period in Table 2 is 5-20 minutes. This time period is needed for the sub-assemblies of the generator to warm up, but at the same time it is important to consider that the reference values have to be based on a sufficient number of spark discharge events. According to our experience, the reference acquisition time period is preferably constituted by a warm-up period having a length of 5-10 minutes; however, the reference acquisition time period can also be chosen to be as long as 10 to 20 minutes. Besides these, other intermediate variables also emerge during operation that are preferably also determined automatically, utilizing hardware (e.g. electronic) or software components. These variables have to be introduced at this point in order to keep the subsequent sections of the description concise and clear. Notation Description Typical

(initial) value

I _th,up Intensity threshold value corresponding to the rising edge

of the signal profile PD(t) of the photodiode

t _th,dn Intensity threshold value corresponding to the falling edge

of the signal profile PD(t) of the photodiode

t _sp Temporal length of an oscillating spark discharge that is 10 ms

estimated in the course of the measurements by the time

difference between the rising edge threshold value

transition I _th,up and the falling edge threshold transition

tth.dn of the signal profile PD(t) of the photodiode.

t _int Integration time of the spectrometer. The relation t _int £ t _sp 1-10 ms has to be fulfilled at all times.

Table 3

With the help of the above described input data, constants and intermediate variables, the following secondary indicator parameters, listed in Table 4 below, are derived by the algorithm (software; the secondary indicators are optical, typically spectroscopic parameters):

Corresponds to the value of SP( l _k,i,j , t).

l _g,i,j Net intensity of the j-th atomic or ionic spectral line of the i-th elemental component of the carrier gas medium of the generator.

Corresponds to the value of SP( l _g,i,j , t).

l _e,i,j Net intensity of the j-th atomic or ionic spectral line of the i-th elemental component of the atmospheric environment (air) of the generator. Corresponds to the value of SP( l _e,i,j , t).

w Oscillation frequency of a single spark discharge event, which is determined approximately from the fluctuation of the signal profile

PD(t) of the photodiode based on the smallest characteristic frequency value appearing in the Fourier transform of the signal profile PD(t) acquired during the period t _sp .

Table 4

The most important 2-3 primary (prioritized) indicators on which the control algorithm can be based are the described below; in the descriptions, the evaluation process of the indicators are also specified through concrete examples. 1. Spark repetition frequency (f _rep )

It is derived from the photodiode signal PD(t); it is primarily related to the yield. There are two possible ways for determining the value of this indicator:

a) it is equated with the smallest characteristic frequency value appearing in the Fourier transform of the signal profile PD(t) acquired during the time period t _update , or

b) the number of TRIG events that have occurred per second (this latter option does not work if there are no spark discharge events due to an error).

The control signal corresponding to this primary indicator is the value of the capacitor charge current (l _ch ).

Value of f _rep is influenced primarily by the capacitor charge current (l _ch ). f _rep could possibly increase in the case of a decrease in U _br (breakdown voltage, see Table 1 ), which is, however, not realistic, because it could primarily be caused by the reduction of the electrode gap, but under operating conditions that is not a realistic scenario (by decreasing of the material of the electrodes, the electrode gap increases). It can then be presumed that l _ch has decreased, so it is expedient to increase it in order to restore the value of f _rep . A decrease of f _rep mostly occurs due to the increase of l _Ch and due to electrode loss (the increase of the electrode gap); the latter would also be indicated by an increase in T _pl (plasma temperature, see below in point 2) and A _sp . Therefore, if these parameters (T _pl and A _sp ) have not changed, then U _b r also has not changed, so it is expedient to reduce l _ch . 2. Plasma temperature (T _pl )

It is derived from the emission spectrum, and is primarily related to the size distribution of the particles. This temperature value can be determined by the so- called Boltzmann method (the Saha-Boltzmann method is an extension of the Boltzmann method, in relation to that see the corresponding reference), utilizing the intensity values (/ _a,i,j and/or I _k,i,j ) and other basic parameters (E, g, l) of the spectral lines characteristic of the electrode material.

The control signal corresponding to this primary indicator is the electrode gap size

(d).

An increase in T _pl (together with an increased A _sp and reduced f _rep ) indicates an increase in U _br , which, with p being constant, can occur mostly in case the electrode gap has increased, i.e. there is significant electrode erosion. It is then necessary to decrease (restore) d, which practically involves applying electrode feed (re-supply).

Under normal operating conditions a permanent decrease of T _pl is not expected because the electrodes cannot“grow”. Therefore it can only be caused by an error phenomenon, i.e. the occurrence of an error state (for example, changes of the parameters of the discharge circuit), which is also indicated by a decrease of A _sp and a significant change in w (oscillation frequency, see Table 4). Plasma temperature is indirectly related to the size distribution of the generated particles. The relationship is based on that the higher the breakdown voltage (which, according to the Paschen formula, is determined primarily by the distance between the electrodes, and the temperature and pressure of the gas in the electrode gap [Friedrich Paschen, Annalen der Physik. 273 (1889) 69-75.]), the higher the“plasma energy” (“energy per spark”), because it is usually considered to be the same as the energy stored in the capacitor. With increasing energy, plasma temperature increases [A. Kohut, L. Ludvigsson, B.O. Meuller, K. Deppert, M.E. Messing, G. Galbacs, Zs. Geretovszky, Nanotechnology vol. 28, 475603 (2017)] while the mass that is ablated (eroded) from the electrode material (per spark) also increasing (F. Llewellyn-Jones, Br. J. Appl. Phys. 1 , 60-65 (1950)). This, in turn, causes the formation of larger-sized nanoparticles from the higher- concentration material. In relation to plasma temperature see also A. Kohut, L. Ludvigsson, B.O. Meuller, K. Deppert, M.E. Messing, G. Galbacs, Zs. Geretovszky, Nanotechnology vol. 28 475603 (2017), and J. Feng, G. Biskos, A. Schmidt-Ott, Scientific Reports, vol. 5 15788 (2015).

Particle size distribution is of course an important parameter during the generation process, it is, however, not directly measured by the apparatus according to the invention; there are commercially available devices adapted for that purpose, for example scanning mobility particle sizers (SMPS). The output information of such devices can in principle be fed into the present apparatus, which may allow for the direct utilization of their data for control purposes. However, it is intended that the apparatus according to the invention applies sciential expressions, which allows for manufacturer-independent, autonomous operation.

3. The chemical composition of the plasma (X _P )

Derived from the emission spectrum, this parameter directly determines the composition of the particles. This indicator is applied for binary (dual-component) nanoparticles, in relation to generators with different anode and cathode compositions. Such nanoparticle generation methods are more widespread than the methods involving the generation of so-called “single metal” nanoparticles applying an anode and a cathode that have identical material composition. Accordingly, this indicator can be applied in typical cases. Its value corresponds to a concentration (atomic) ratio, calculated in an anode/cathode manner. The value of this parameter can be estimated applying an equilibrium plasma model that is also required for obtaining T _pl .

The control signal corresponding to this primary indicator is: electrode polarity reversal.

If the value of X _p slightly differs from the reference value, then the erosion rate of one of the electrodes has to be increased, while the erosion of the other electrode has to be slowed down. This can be achieved by reversing the polarity of the electrodes.

The state of the system can also be represented in a coordinate system where the values of the primary indicator are illustrated on the axes of the coordinate system. The system’s state can then be described as a point. For the sake of a simpler description, let us place the origin of the coordinate system at the point that corresponds to the reference values of the primary indicators. In this parameter space, the“distance” of the current state of the generator from the reference state (the state during the reference acquisition time period) corresponds to the (Euclidean) distance measured from the origin. In a given data acquisition cycle, the control actions performed by the method and apparatus according to the invention attempt to bring closer to the reference value the value of the primary indicator that has the largest relative change compared to the reference value. The corresponding control action can be changing, e.g. by 10%, the control signal assigned to the particular parameter. In an embodiment, therefore, more (a plurality, more than one) primary indicators are applied, and a control signal is generated for the primary indicator for which the difference between the primary indicator and the corresponding reference value is the largest (of all of the differences between primary indicators and the corresponding reference values). The control signal determined on the basis of theoretical relationships modifies an appropriate operating parameter (or parameters) of the generator that urge(s) the state of the generator characterized by the primary indicators towards the reference state. The above mentioned 10% modification is an example that is in line with the general principles of process control technology. There is a proportional relationship between the feedback (control) signals and the primary indicators (if one increases, the other increases too, etc.), however, the exact value of the proportional factor depends on the properties of the system established including the particle generator, which value is not initially known to a universal control apparatus (but of course can be measured for a given generator system).

Therefore, the value of a control (feedback) signal to be issued in response to the change of a primary indicator by a particular amount cannot be established in a general manner, although it is known that its value has to be reduced. It is therefore expedient to decrease or increase (i.e. change) the feedback signal value by for example 10% with respect to the earlier value in order to prevent an overshoot. Because the tupdate value of the cycle time of the status check is usually short (for example, 1 minute), on the one hand it can be expected that there is only a small change in the primary indicator values during each cycle, and on the other hand, it can be expected that the phenomenon can be addressed by 10% small control intervention. Since in the next cycle it is checked if the primary indicators are within their respective tolerance ranges (i.e. if their deviation from the corresponding reference values is under the threshold value), a potentially required further intervention can be performed in the next cycle, so the generator can be brought into an appropriate state within a few cycles, preferably without overshoot.

The weight factors included in the bottommost row of Table 2 enable the user to set that which of the particular primary indicators with what priority (weight) are to be handled during the control process by the method and apparatus according to the invention. This is because the primary indicator values are compared by the method and apparatus at the given time instant (in the n-th data update cycle) with their respective values measured in the previous (n-1-th) cycle, and preferably an attempt is made to control back (by means of modifying the operating parameters, i.e. by the control (output) signals) that value which has the greatest deviation compared to the corresponding reference value. Since the different primary indicators have different ranges (the value of is between 1 and 1000 Hz, T _pl is typically in the range between 5000 and 15000 K, the chemical (fractional) composition is in the range of 0-1), preferably the normalized relative change of the indicators are monitored in the control loop (the primary indicator is denoted by El in the formula):

By specifying weight factors Si entered upon startup of the system (i.e. the apparatus and the method corresponding thereto), the user may select those primary indicators for which the deviations are to be treated with priority. If, for example, particle concentration is important for the user, then it is expedient to increase the weight of the f _rep indicator, while in case the important parameter is particle size or composition, then the weight of the T _pl or the Xi indicator, respectively, are expediently increased. If, for example, the indicator f _rep has a weight of 2 and the other indicators have a weight of 1 , then in case such a relative deviation of f _rep occurs that is only half as great as the relative deviation of other indicators, the apparatus (system) will still attempt to control back this parameter until the DEIi value remains the largest for this indicator.

In an embodiment of the method, therefore, more primary indicators are derived from the monitored data, and if the difference between the values of more primary indicators and the corresponding reference values are larger than a first threshold value, a control signal is generated and issued, weighted by weight factors, for the primary indicator resulting in the largest difference, in case an error state cannot be identified in the particle generator.

In the following, examples of the error signaling and control mechanism of the method and apparatus according to the invention are described, i.e. specific embodiments of the invention are presented. The below included description of the three characteristic primary indicator deviations illustrates the error detection/signaling and control processes applied by method and apparatus according to the invention for the three primary indicators (f _rep , T _pl and X _p ). A) Primary indicator deviation I: Of the continuously monitored optical signals, a deviation (difference) from the reference value recorded during the reference acquisition/learning time period of the spark repetition frequency (f _rep ) - a primary indicator derived from the photodiode signal ( PD(t )) - is detected (i.e. one of the primary indicators monitored in the present embodiment is f _rep and a deviation thereof compared to the threshold value exceeding the threshold value is detected).

In this embodiment, the steps of the analysis process can be the following:

1. In this embodiment of the method and apparatus according to the invention it is investigated whether the parameters characterising the discharge circuit of the generator (R, L, C, i.e. resistance, inductance, capacitance) have changed compared to the reference values. This is expediently performed on the basis of the oscillation frequency ( w) calculated from the PD(t) signal. The relationship between the frequency w and the investigated electrical parameters is well known from the literature on RLC circuits (for spark discharge particle generators, see for example: N. S. Tabrizi, M. Ullmann, V. A. Vons, U. Lafont, and A. Schmidt-Ott, “Generation of nanoparticles by spark discharge,” J. Nanoparticle Res., vol. 1 1 , no. 2, pp. 315-332, (2009)). A change of w with compared to the reference value (i.e. w leaving its tolerance range) indicates that the parameters of the discharge circuit (RLC circuit) have changed. In this case the discharge circuit is presumably malfunctioning, which calls for the shutdown of the generator and operator intervention.

2. If w is unchanged (i.e. its difference from the reference value is under the threshold value or in other words its value stays inside the tolerance range) then it means that the electrical parameters of the discharge circuit have not changed, so f _rep has changed for a different reason.

3. Accordingly, in the next step, in this embodiment of the method and apparatus it is checked whether the breakdown voltage characteristic of the discharge has changed. To achieve that, a plurality of optically derivable parameters may be investigated simultaneously. One example of such parameters is the plasma temperature (T _pl ) that can be calculated from the emission spectrum of the spark discharge. This parameter is sensitive to changes in breakdown voltage. In order to calculate ( T _pl ), the method and apparatus makes use of the software-specified constants characteristic of the spectral lines of the carrier gas, and the line intensities obtained from the emission spectrum of the spark discharge. The calculation is based on the so-called Boltzmann, as well as Saha-Boltzmann method, the application of which for spark discharge generators is described for example in A. Kohut, G. Galbacs, Z. Marton, and Z. Geretovszky,“Characterization of a copper spark discharge plasma in argon atmosphere used for nanoparticle generation,” Plasma Sources Sci. Technol., vol. 26, no. 4, p. 045001 , Mar. (2017), and in other studies, for example in C. A. Bye and A. Scheeline,“Saha-Boltzmann statistics for determination of electron temperature and density in spark discharges using an echelle/CCD system,” Appl. Spectrosc., vol. 47, no. 12, pp. 2022-2030, (1993), C. Aragon and J. A. Aguilera, “Characterization of laser induced plasmas by optical emission spectroscopy: A review of experiments and methods,” Spectrochim. Acta - Part B At. Spectrosc., vol. 63, no. 9, pp. 893-916, (2008). If the T _pl value calculated from the emission spectrum has changed compared to the reference value (has left the tolerance range), then it is an indication of a change in the breakdown voltage. As it is taken for granted that a constant pressure is maintained in the monitored particle generators, a change in the breakdown voltage indicates a change of the spark gap distance. In such a case, utilizing the direction and magnitude of the change of T _pl the device gives an estimate for the amount and direction of the necessary correction of the spark gap distance in order to return to normal operation, and sends the corresponding control signal to the actuators moving the electrodes i.e. in this case the electrode positioning device is operated. The electrode positioning device is implemented, by way of example, applying actuators moving the electrodes. If the T _pl value has not changed compared to the reference (i.e. it stays within the tolerance range), then the breakdown voltage is also constant, so the change of f _rep is caused by a change of the capacitor charge current. In such a case, based on the direction and degree of the optically detected change of f _rep it is determined in which direction and by what amount the charge current has to be changed in order to restore the operating value of f _rep , and the corresponding control signal is issued to the power supply applied for charging the capacitor, i.e. the capacitor charge current is modified by issuing a control signal for modification.

In the above described embodiment, therefore, the method and apparatus is applied in relation to such a particle generator (such a particle generator is subjected to monitoring and control applying the method) that comprises electrodes providing a spark discharge, an electrode positioning device, and, in its discharge circuit, a capacitor connected parallel to the electrodes, and a corresponding capacitor charging unit.

In the embodiment described above, a generation spark repetition frequency of the particle generator is applied as one of the one or more primary indicator. As it is discussed below, in the case of this specially selected primary indicator the role of the first threshold value applied for identifying the warning event is performed by a second threshold value (identical threshold values may also correspond to the different individually specified primary indicators, but the values may also be different; however, in the discussion below they are referred to by different names). Accordingly, in the below described case (i.e. if the second threshold value is exceeded) a warning event is also issued, which is followed by performing the steps described below.

In the present embodiment, furthermore, a reference spark repetition frequency value is determined, in the reference acquisition time period, for the generation spark repetition frequency, and in case a difference between the generation spark repetition frequency and the reference spark repetition frequency value exceeds a second threshold value (wherein the second threshold value is adapted for comparing the generation spark repetition frequency and the value of the reference spark repetition frequency) it is investigated whether a deviation of the oscillation frequency of the discharge circuit from its reference value exceeds a third threshold value (wherein the third threshold value is adapted for comparing the oscillation frequency of the discharge circuit with the corresponding reference value), - in case it exceeds the third threshold value, an error signal is issued and an error state is identified in the particle generator,

- in case it does not exceed the third threshold value, it is investigated if a deviation of the breakdown voltage of the discharge circuit from the corresponding reference value exceeds a fourth threshold value (wherein the fourth threshold value is adapted for comparing the breakdown voltage of the discharge circuit with the corresponding reference value),

- in case it exceeds the fourth threshold value, an electrode positioning control signal is determined and issued to the electrode positioning device of the particle generator (this signal is responsible for controlling the electrode positioning device, i.e. for applying an appropriate feed to the electrodes with respect to their earlier position),

- in case it does not exceed the fourth threshold value, a capacitor charging control signal is determined and issued to (or collectively: is generated for) the capacitor charging unit of the particle generator (this signal is responsible for the appropriate control of the capacitor charging unit; i.e. it is essentially the capacitor charge current that is modified).

B) Primary indicator deviation II: In this embodiment of the method and apparatus according to the invention, there is a change of the plasma temperature ( T _pl ) calculated from the continuously monitored plasma emission spectrum. In a further embodiment, to be described below, therefore, the plasma temperature is followed as one of the primary indicators (in case of a deviation, the change of plasma temperature, i.e. the temperature value leaving the tolerance range, is detected in a first step).

In this embodiment, the steps of the analysis process are the following:

1 . As with the above described embodiment, it is investigated whether the R, L, C parameters characterising the discharge circuit of the generator have changed compared to the reference values. This is performed on the basis of the oscillation frequency (w) calculated from the PD(t) signal. The relationship between the frequency w and the above mentioned electrical parameters is well known from the literature on RLC circuits (for spark discharge particle generators, see for example: N. S. Tabrizi, M. Ullmann, V. A. Vons, U. Lafont, and A. Schmidt-Ott, “Generation of nanoparticles by spark discharge,” J. Nanoparticle Res., vol. 11 , no. 2, pp. 315-332, (2009). A change of w compared to the reference value indicates that the parameters of the discharge circuit (RLC circuit) have changed (for example, in Fig. 1 only a capacitor is shown, but of course in such circuits a non-zero (low) resistance and an inductance are also present). In such a case the discharge circuit is presumably malfunctioning, which calls for the shutdown of the generator and operator intervention. . If w is has not changed, the change of T _pl may be attributed to the changes in discharge voltage. The discharge voltage determines the energy of the spark, through which it is directly related to the size and amount of the generated particles (N. S. Tabrizi, M. Ullmann, V. A. Vons, U. Lafont, and A. Schmidt-Ott, “Generation of nanoparticles by spark discharge,” J. Nanoparticle Res., vol. 11 , no. 2, pp. 315-332, (2009), J. Feng, L. Huang, A. Maisser, G. Biskos, and A. Schmidt-Ott,“A general approach to the evolution of singlet nanoparticles from a rapidly-quenched point source,” J. Phys. Chem. C, no. December, (2015), J. Feng, G. Biskos, and A. Schmidt-Ott, “Toward industrial scale synthesis of ultrapure singlet nanoparticles with controllable sizes in a continuous gas- phase process,” Sci. Rep., vol. 5, no. October, p. 15788, (2015), A. Kohut et ai, “From plasma to nanoparticles: optical and particle emission of a spark discharge generator," Nanotechnology, vol. 28, no. 47, p. 475603, Nov. (2017), H. Horvath and M. Gangl, “A low-voltage spark generator for production of carbon particles,” J. Aerosol Sci., vol. 34, no. 11 , pp. 1581-1588, (2003), M. Seipenbusch, A. P. Weber, A. Schiel, and G. Kasper,“Influence of the gas atmosphere on restructuring and sintering kinetics of nickel and platinum aerosol nanoparticle agglomerates,” J. Aerosol Sci., vol. 34, no. 12, pp. 1699- 1709, (2003)).

3. As it is taken for granted that a constant pressure is maintained in the monitored particle generators, a change in the breakdown voltage indicates a change of the spark gap distance. In such a case, utilizing the direction and magnitude of the change of T _pl an estimate is given for the amount and direction of the necessary correction of the spark gap distance in order to return to normal operation, and the corresponding control signal is sent to the actuators adapted to move the electrodes (i.e. the electrode positioning device is operated).

Thus, in an embodiment, the respective embodiment of the method and apparatus according to the invention is applied in relation to such a particle generator (such a particle generator is subjected to monitoring and control applying the method) that comprises electrodes for generating a spark discharge, an electrode positioning device, and provided with (has) a discharge circuit.

In this embodiment, in the course of the method a generation plasma temperature of the particle generator is applied as one of the one or more primary indicator. In this embodiment, the role of the first threshold value utilized for identifying a warning event is performed by a fifth threshold value; the steps described below are carried out after the warning event has been identified.

In the present embodiment, furthermore a reference plasma temperature is determined, in the reference acquisition time period, for the generation plasma temperature, and, in case a difference between the generation plasma temperature and the reference plasma temperature exceeds a fifth threshold value (wherein the fifth threshold value is adapted for comparing the generation plasma temperature and the reference plasma temperature), it is investigated whether the deviation of the oscillation frequency of the discharge circuit from its reference value exceeds a third threshold value,

- in case it exceeds the third threshold value, an error signal is issued and an error state is identified in the particle generator,

- in case it does not exceed the third threshold value, an electrode positioning control signal is determined and is issued to the electrode positioning device of the particle generator (the method is simpler here because at the start of this embodiment it is identified that the plasma temperature has changed, so - as described above in relation to the primary indicator deviation I - the need for an electrode positioning intervention can be identified in fewer steps). C) Primary indicator deviation III: In a yet further embodiment of the method and apparatus, there is a change of the chemical composition (X _p ) calculated from the continuously monitored spark emission spectrum (a deviation is identified if a change in the chemical composition leaving the tolerance range is detected).

In this embodiment, the steps of the analysis process are the following:

1. The chemical composition (X _p ) is dependent upon the concentration ratio of the atoms of the electrode material present in the spark gap (this parameter is interpreted in the case of compound (composite) nanoparticles, i.e. electrodes having a different material). Since it is a relative parameter, it is not sensitive to the changes of f _rep .

2. In order to obtain X _p , the plasma temperature has also to be known (see for example in A. Kohut, G. Galbacs, Z. Marton, and Z. Geretovszky, “Characterization of a copper spark discharge plasma in argon atmosphere used for nanoparticle generation,” Plasma Sources Sci. Technol., vol. 26, no. 4, p. 045001 , Mar. (2017), B. Charfi,“The effect of temperature on the spectral emission of plasma induced in Water,” J. Spectrosc., vol. 1 , no. 1 , p. 6 pages, (2013)), so, in case of a change in X _p (since the intensity of the individual spectral lines in the atomic emission spectrum is proportional to the concentration of the chemical elements present, in a manner known per se the chemical composition is calculated accordingly), in this embodiment it is therefore investigated whether T _pl has changed. If in T _pl has been changed, the present embodiment jumps to the steps described above in relation to the primary indicator deviation II, and performs a correction of T _pl .

It is checked whether the parameters of the discharge circuit have changed. If yes, that indicates the failure of the discharge circuit. If the parameters of the discharge circuit have not changed, the electrode positioning device is brought into operation, i.e., a control signal is issued thereto, to achieve control of T _pl ,.

3. If T _pl has not changed, a change in X _p indicates that a change in the relative erosion of the electrode has occurred. (J. Feng, N. Ramlawi, G. Biskos, and A. Schmidt-Ott, “Internally mixed nanoparticles from oscillatory spark ablation between electrodes of different materials”, Aerosol Sci. Technol., vol. 52, no. 5, pp. 505-514, May (2018), T. V. Pfeiffer, J. Feng, and A. Schmidt-Ott,“New developments in spark production of nanoparticles,” Adv. Powder Technol., vol. 25, no. 1 , pp. 56-70, (2014), A. Kohut, L. P. Villy, T. Ajtai, Z. Geretovszky, and G. Galbacs,“The effect of circuit resistance on the particle output of a spark discharge nanoparticle generator,” J. Aerosol Sci., vol. 1 18, pp. 59-63, Apr. (2018)). This can be the result of a permanent malfunction of the RLC circuit (this can be checked by examining the changes of the oscillation frequency (w) calculated from the PD(t) signal; for more details see the previous examples), which requires the shutdown of the generator and operator intervention.

4. If a permanent malfunction of the discharge circuit of the generator has not occurred, the control of X _p can be implemented for example by means of changing the initial polarity of the electrodes or the total resistance ( R ) of the discharge circuit. To achieve that, a control signal can be supplied by the present embodiment of the method and apparatus according to the invention (taking into account the magnitude and direction of the change of X _p ), in case the monitored spark discharge generator allows the modification of the above described parameters.

The control action that can be achieved through polarity change works along the same principle than the above described mechanism that involves for example a 10% modification, because through polarity change a 5-10% modification of chemical composition can be achieved, so the process can be controlled in a similar manner.

In the above described embodiment, therefore, the method and apparatus according to the invention are applied in relation to such a particle generator (such a particle generator is subjected to monitoring applying the method) that comprises electrodes for generating a spark discharge, an electrode positioning device, and provided with a discharge circuit.

In this embodiment of the method, a generation chemical composition detectable in the particle generator is applied as one of the primary indicators. In this embodiment, the role of the first threshold value utilized for detecting a warning event is performed by a sixth threshold value; the steps described below are carried out after the warning event has been identified.

In this embodiment, furthermore, a reference chemical composition is determined, in a reference acquisition time period, for a generation chemical composition, in case the difference between the generation chemical composition and the reference chemical composition exceeds a sixth threshold value (wherein the sixth threshold value is adapted for comparing the generation chemical composition and the reference chemical composition) it is investigated whether the deviation of the generation plasma temperature of the particle generator from the corresponding reference value exceeds a fifth threshold value,

- if it exceeds the fifth threshold value, it is investigated whether the deviation of the oscillation frequency of the discharge circuit from its reference value exceeds a third threshold value,

- in case it exceeds the third threshold value, an error signal is issued and an error state is detected in the particle generator,

- in case it does not exceed the third threshold value, an electrode positioning control signal is determined and is issued to the electrode positioning device of the particle generator,

- if it does not exceed the fifth threshold value, it is investigated whether the deviation of the oscillation frequency of the discharge circuit from its reference value exceeds a third threshold value,

- in case it exceeds the third threshold value, an error signal is issued and an error state is detected in the particle generator,

- in case it does not exceed the third threshold value, a chemical composition control signal is determined and issued for controlling the chemical composition detectable in the particle generator.

In the embodiment described above, the chemical composition control signal is preferably aimed at modifying the polarity of the electrodes or the total resistance of the discharge circuit.

V.5. CERTAIN ASPECTS OF THE APPARATUS ACCORDING TO THE INVENTION ln accordance with the above, some embodiments of the invention relate to an apparatus for monitoring a spark discharge particle generator adapted for generating particles (particularly nanoparticles).

In an embodiment, the apparatus according to the invention comprises an optical detection device adapted for monitoring generation of particles in the particle generator and for recording the value of monitored data, and a central unit by means of which

- one or more primary indicator is derived from the monitored data, and, in a reference acquisition time period following a start of the recording of the monitored data, one or more reference value for the one or more primary indicator is determined,

- generation of particles is monitored in a production time period following the reference acquisition time period, and it is checked according to a checking time interval whether an error state can be identified in the particle generator,

- in case an error state can be identified, an error signal is issued, and the recording of the value of monitored data is terminated,

- in case an error state cannot be identified, actual value of the one or more primary indicator, or actual value and one or more earlier value thereof, or a weighted value calculated from actual value and one or more earlier value thereof is compared with one or more reference value, and monitoring of generation of particles is continued, identifying a warning event in case a difference between the value of one or more primary indicator and the corresponding reference value exceeds a first threshold value.

The apparatus according to the invention is preferably suitable for carrying out the steps of the method according to the invention, so those described above also applies to the apparatus unless it is explicitly contradicted. Additionally, in the following some embodiments of the apparatus are specified explicitly.

As it is illustrated also by the examples described above, in an embodiment of the apparatus the monitoring optical detection device comprises a spectrometer unit. In this embodiment, the particle generator subjected to monitoring preferably has a discharge space and an optical window facing it, and the spectrometer unit is connected to the particle generator through a first optical conduit by means of light collecting optics arranged at the optical window of the particle generator.

In a further embodiment, the apparatus preferably comprises a photodiode unit and a trigger unit adapted for providing to the spectrometer unit a trigger signal corresponding to the start of a discharge event in the particle generator based on the signal thereof, the first optical conduit has a first section implemented as a common optical fiber and a second section branching off to a first optical fiber leading to the spectrometer unit and to a second optical fiber leading to the photodiode unit, and an optical delay unit is inserted into the first optical fiber for enabling that the spectrometer unit to be brought in a measurement-ready state by the trigger signal until a time of receiving a signal transferred over the first optical fiber.

In the embodiments of the apparatus comprising a spectrometer unit the monitored data comprise spectral data recorded by the spectrometer unit.

In line with the above description of the method, in an embodiment of the apparatus the monitoring optical detection device preferably comprises a camera device (for example, the above mentioned digital camera). In this embodiment, the particle generator subjected to monitoring preferably has a discharge space and an optical window facing it, and the camera device is connected to the particle generator via a bundle of second optical conduits by means of a camera optics arranged at the optical window of the particle generator.

Thanks to that, in an embodiment the apparatus exclusively has such a connector adapted to be connected to the particle generator that transfers the monitored data to the monitoring optical detection device through optical conduit, particularly a fiber optics or fiber optics bundle.

In a further embodiment, the apparatus comprises an audiovisual display adapted for displaying the warning event (for example a display provided with a speaker unit), and/or a logging unit adapted for logging the warning event.

According to the above, an embodiment of the apparatus is therefore adapted for monitoring a plurality of particle generators, recording, by means of a monitoring optical detection device, the values of monitored data in each particle generator. Furthermore, this embodiment preferably comprises an optical multiplexer unit OM adapted for time-multiplexed processing of the monitored data of the optical detectors corresponding to the different particle generators, i.e. the monitored data of the optical detection devices corresponding to each of the particle generators are processed in a time-multiplexed manner.

In line with the above, control is implemented as a possible functionality also in some embodiments of the apparatus. Accordingly, in an embodiment the apparatus comprises a control signal generation unit (control signal issuing unit) adapted for generating a control signal for the particle generator, which control signal generation unit, in case of identifying a warning event, generates a control signal for the particle generator based on a difference between the value of one or more primary indicator and the one or more reference value exceeding a first threshold value.

In a further embodiment, the central unit is adapted for deriving more primary indicators, and the control signal generation unit generates a control signal for the primary indicator for which the difference between the primary indicator and the corresponding reference value is the largest.

An embodiment of the apparatus according to the invention is adapted for carrying out the embodiment of the method that was discussed hereinabove in relation to the primary indicator deviation I. This embodiment of the apparatus is adapted for subjecting to monitoring such a particle generator that comprises electrodes providing a spark discharge, an electrode positioning device, and, in its discharge circuit, a capacitor connected parallel to the electrodes, and a corresponding capacitor charging unit. In this embodiment, the central unit is adapted for establishing, in the reference acquisition time period, a reference spark repetition frequency value for the spark repetition frequency, and for comparing the generation spark repetition frequency and the reference spark repetition frequency. in this and other embodiments adapted for controlling primary indicator parameter deviations the relationship of the deviations and the threshold values are examined by the central unit that also issues an error signal if necessary, and the control signals to be issued are generated by the control signal generation unit.

A further embodiment of the apparatus according to the invention is adapted for carrying out the embodiment of the method that was discussed hereinabove in relation to the primary indicator deviation II. This further embodiment of the apparatus is adapted for subjecting to monitoring such a particle generator that comprises electrodes for generating a spark discharge, an electrode positioning device, and provided with (has) a discharge circuit. In this embodiment, the central unit is adapted for establishing, in the reference acquisition time period, a reference plasma temperature for the generation plasma temperature, and for comparing the generation plasma temperature and the reference plasma temperature.

A yet further embodiment of the apparatus according to the invention is adapted for carrying out the embodiment of the method that was discussed hereinabove in relation to the primary indicator deviation III. This yet further embodiment of the apparatus is adapted for subjecting to monitoring such a particle generator that comprises electrodes for generating a spark discharge and an electrode positioning device, and is also provided with a discharge circuit. In this embodiment, the central unit is adapted for determining, in a reference acquisition time period, a reference chemical composition for the generation chemical composition, and for comparing the generation chemical composition and the reference chemical composition.

Of the threshold values mentioned above, those threshold values that have different names are typically related to different values (and correspond to different indicators). However, those threshold values that have identical names (although they may have been introduced independent of each other) can preferably assume identical values.

In the above described embodiment of the apparatus, the chemical composition control signal is preferably aimed at modifying the polarity of the electrodes or the total resistance of the discharge circuit. In an embodiment of the apparatus applying control, more primary indicators are derived from the monitored data, and if the difference between the values of more primary indicator and the corresponding reference values are larger than a first threshold value, a control signal, weighted by weight factors, is generated and issued for the indicator resulting in the largest difference in case an error state cannot be identified in the particle generator.

As set forth above, the apparatus according to the invention is also capable of signaling the status (normal operation, errors/malfunctions, etc.) of the spark discharge particle generators, and thereby it can be utilized for reducing production downtime (i.e. production efficiency increases), while a need for raw material resupply (due to electrode material loss) can also be predicted in advance.

The invention is, of course, not limited to the preferred embodiments described in details above, but further variants, modifications and developments are possible within the scope of protection determined by the claims.

Legends

KK discharge chamber

SZK spark gap

EP1 (first) electrode positioning device

EP2 (second) electrode positioning device

E1 (first) electrode

E2 (second) electrode

OB optical input

KT capacitor charging unit

NK (high-voltage) capacitor

GND grounding

GK gas control unit

EK electrical output

FO light collecting optic

KO camera optic

OA optical window KM communication module

FD photodiode unit

AD data acquisition unit

TR trigger unit

SP spectrometer unit

OB1 (first) optical input

OB2 (second) optical input

SZ central computer

DK digital camera

OM optical multiplexer

10 (first) optical conduit

12 (second) optical conduit

14a (first) feedback connection

14b (second) feedback connection

16 (third) feedback connection

18 (fourth) feedback connection

20a (first) optical fiber

20b (second) optical fiber

22 (first) electrical connection

24 (second) electrical connection

26 (third) electrical connection

45 (first) fiber optics

47 fiber optics bundle

50 (spark discharge) particle generator

65 (second) fiber optics

100 (first monitoring) apparatus

200 (second monitoring) apparatus

300 (third monitoring) apparatus

400 (fourth monitoring) apparatus

S320 monitored data are recorded and a primary indicator is derived S330 a reference value is identified in a reference acquisition time period S340 particle generation is monitored

S350 error state identified? S355 error signal

S360 according to a checking time interval, the primary indicator and the reference value are compared

S365 is the difference greater than a threshold value?

S370 warning event

S410 the monitoring unit is turned on by the operator

S415 conditions and tolerances are entered by the operator to software of the monitoring unit

S420 generator is started by the operator

S425 monitoring unit is started by the operator in learning mode

S430 measurement data are acquired by the monitoring unit in a fixed time period (learning time period); from which reference values of the indicators are calculated S435 monitoring unit automatically switches to production mode

S440 measurement data are acquired, actual indicator values are determined S445 has the expected operating time period elapsed?

S450 normal shutdown

S455 error state identified?

S460 error signal, shutdown

S465 is there an out-of-tolerance difference between the indicator and the reference values?

S470 warning signal

S475 derivation of optional proportional control signals and output towards the generator