WO2006037704A1

WO2006037704A1 - Device and method for characterizing plasma

Info

Publication number: WO2006037704A1
Application number: PCT/EP2005/054317
Authority: WO
Inventors: Sébastien DINE; Jean-Paul Booth
Original assignee: Ecole Polytechnique; Centre National De La Recherche Scientifique (Cnrs)
Priority date: 2004-10-07
Filing date: 2005-09-01
Publication date: 2006-04-13
Also published as: FR2876536B1; FR2876536A1

Abstract

The invention concerns a method and a device for measuring electronic density in a plasma reactor (19) comprising: an emitting antenna (17) for emitting surface waves (10), a receiving antenna (18), means for electromagnetic continuity (19) between the emitting antenna and the receiving antenna for carrying the waves to the receiving antenna, means (24) for detecting the waves, processing means (16) connected to said detecting means for: determining a transmission threshold frequency corresponding to an emission frequency above which the waves are emitted with substantial power, deducing therefrom an electronic density in the plasma (14).

Description

DEVICE AND METHOD FOR CHARACTERIZING PLASMA

The present invention relates to a device for measuring the electron density in a plasma reactor, said plasma being distributed in a space contained in the reactor. And the invention also relates to an electronic density measurement method associated with such a device.

It is specified that in this text "plasma" means a sufficiently ionized gas for the charged species that it contains to have a collective behavior - that is to say that they adapt their movement according to the movement of the other species, electronic density and density of other charged species around them.

The electron density, n _e , is the number of electrons per unit volume often expressed in electrons / cm ³ .

And "plasma space" is in this text defined as the space occupied by the plasma inside the plasma reactor.

It is recalled that before any surface exposed to a plasma forms an area called "electrostatic sheath" (shown for example in Figure 2 by the reference 45). The typical thickness of this area is a few millimeters. The electrostatic sheath is an electron-poor zone because the electron density is much lower than that in the plasma. We can thus assimilate this zone to a vacuum.

This electrostatic sheath is automatically formed around any object immersed in a plasma. As will be seen, the invention applies particularly advantageously - but not limited - to low electron density plasmas and high gas pressure.

A "low density" plasma is defined here as a plasma in which the electron density is less than a value of 10 ¹⁰ electrons / cm ³ . And one likewise defines a plasma "high pressure" as a plasma in which the pressure is greater than a value of the order of 65 pascals.

There are already many types of electronic density measuring devices in a plasma reactor.

These devices are often able to perform so-called "global" measurements, ie they provide a unique value characterizing the overall electronic density.

Now it is sometimes desirable to carry out "local" measurements, which characterize the electronic density at a specific place in the space occupied by the plasma.

Some devices have been designed to perform such local density measurements.

The Langmuir probe is a first example of such a device. But if the probes of this type make it possible to carry out local measurements of density, they are associated with certain limitations.

Firstly, this type of probe is not suitable for chemically reactive plasmas: the Langmuir probes are indeed sensitive to deposits and other pollutions generated by such plasmas. In addition, this type of probe needs to be compensated in the case

- frequent - a plasma generated by radiofrequency waves (RF plasma).

In addition, the operation of this type of probe is relatively complex. The ion flow probe is another type of device for performing local measurements.

This type of probe - an illustration of which is given in the patent

FR 2 738 984 - requires the knowledge of certain parameters associated with the plasma (such as the Bohm speed and the negative ion density if the plasma contains them) to access, from these parameters, an electronic density. Now it would be advantageous to access direct measurements of electronic density, that is to say not requiring the knowledge of other parameters of the plasma in question.

A third type of device for performing local density measurements is the absorption plasma probe.

This type of probe, illustrated in particular by the document US Pat. No. 6,339,297, exploits the detection of absorption peaks of a stationary surface wave, from which is indirectly deduced a frequency representative of an electron density. . But this type of probe is also associated with limitations.

Its operation is indeed quite complex - it requires in particular an extrapolation processing from the observations made, and thus constitutes an indirect method - which does not give direct access to a density. In addition, the implementation of probes of this type requires significant digital processing, and a prior calibration.

This type of probe also has a limited sensitivity of measurement.

More specifically, the implementation of this type of probe is more particularly delicate at low density (<10 ¹⁰ electrons / cm ³ ) and high pressure (> 65 pascals,> 0.5 torr), because it requires a recourse to a complex analysis of observed signal spectra. Furthermore, the measuring sensitivity of this probe decreases as the pressure increases - it is therefore a function of the pressure.

The sensitivity of this type of probe is thus typically 10 ⁹ electrons / cm ³ below 13 pascals (0.1 torr) and it is of the order of 10 ¹⁰ electrons / cm ³ between 13 pascals and 133 pascals (1 torr).

By comparison, in the case of the invention the tests carried out by the applicants showed a sensitivity of 10 ⁹ electrons / cm ³ up to 133 pascals (1 torr). Thus, in the case of plasmas between 0.1 and 1 Torr, the implementation of this type of known probe is limited to plasmas whose density is greater than about 10 ¹⁰ electrons / cm ³ - which does not allow to characterize low density plasmas.

It thus appears that there remains a need for a device for measuring the density of a plasma, able to perform local density measurements and freed from the limitations set out above.

An object of the invention is to meet this need.

In particular, an object of the invention is to make it possible to carry out density measurements simply, quickly and directly.

Another object of the invention is to make it possible to carry out such measurements without resorting to a large calibration or complex digital processing.

Another object of the invention is to enable the characterization of reactive plasmas which generate deposits and pollutions.

Another object of the invention is to be able to apply density measurement to low density plasmas, high pressure plasmas - and plasmas which are both low density and high pressure.

In order to achieve these aims, the invention proposes, according to a first aspect, an electronic density measuring device in a plasma reactor, said plasma being distributed in a plasma space contained in the reactor, characterized in that it comprises :

Means for generating a periodic electrical voltage associated with a so-called emission frequency,

Control means for controllably varying said transmission frequency; at least one transmitting antenna connected to said emission voltage generating means for emitting surface waves on the surface of the plasma space;

At least one receiving antenna for receiving said surface waves emitted by the transmitting antenna (s), electromagnetic continuity means between the transmitting antenna (s) and the antenna or antennas; s) receiver (s), to constitute a continuous transmission sheath corresponding to the interface between an electrostatic sheath and the plasma between each transmitting antenna and at least one receiving antenna, the boundary of said electrostatic sheath with the plasma for conveying the surface waves emitted by each transmitting antenna to at least one receiving antenna,

Means for detecting the surface waves received by the receiving antenna (s),

Processing means connected to said control means and to said detection means, for:

determining a transmission threshold frequency, which corresponds to an emission frequency above which the surface waves are transmitted to the receiving antenna with a substantial power, and deriving from said transmission threshold frequency an electronic density in the plasma.

Preferred, but not limiting, aspects of the device according to the invention are the following:

The topology of said electromagnetic continuity means is controlled to prevent the plasma from entering between a transmitting antenna and an associated receiving antenna and to guarantee the continuity of the electrostatic sheath between the two antennas,

For each transmitting antenna and an associated receiving antenna, said transmission sheath is located in continuity with the electrostatic sheaths associated respectively with said transmitting antenna and with said receiving antenna,

Said means of electromagnetic continuity are obtained by the dimensioning and the arrangement of each transmitting antenna with respect to each receiving antenna intended to receive the surface wave emitted by said transmitting antenna, Said electromagnetic continuity means correspond to a dielectric continuity element,

Said processing means comprise a spectral or frequency-selective analysis device capable of analyzing the electromagnetic energy received by each reception antenna as a function of the transmission frequency,

Said electromagnetic continuity means comprise an electrostatic continuity structure between each transmitting antenna and each receiving antenna intended to receive the surface wave emitted by said transmitting antenna,

Said electrostatic continuity structure is made of a dielectric or insulating material,

The device comprises an emitting antenna arranged opposite an associated receiving antenna and coaxially with said receiving antenna, so that the space separating the ends of the two antennas is substantially centered on the common axis of the two antennas,

Said electrostatic continuity structure is a cylinder arranged coaxially between a transmitting antenna and an associated receiving antenna,

The device comprises a transmitting antenna disposed non-coaxially with an associated receiving antenna, the ends of the transmitting antenna and the receiving antenna being located substantially opposite one another; said electrostatic continuity structure is a a piece which blocks the space between the respective ends of the transmitting antenna and the receiving antenna so as to prevent the plasma from occupying said space,

The transmitting antenna and the receiving antenna are carried on the same support,

• the antennas are isolated from the plasma, • the antennas are dipolar.

The device comprises:

> two dipole antennas inserted in the plasma, each disposed at a first end of a respective coaxial cable whose second end opens out of the chamber of a plasma reactor;

a dielectric tube in which the coaxial cables are inserted;

> a cylindrical piece, made of a dielectric material, disposed between and in contact with two dipole antennas in which the latter are inserted over their entire length; said cylindrical part making it possible to form between the two antennas an electrostatic sheath whose plasma separation interface is continuous between the two antennas,

an axisymmetric coaxial assembly formed by the dielectric cylinder and the rectilinear dipole antennas whose axes coincide with the axis of symmetry of this assembly,

The device also comprises a plug made of a dielectric material plugging the end of the dielectric tube in its portion exposed to the plasma, a first linear dipole antenna being disposed on the axis of the tube protruding from the tube at its end located at inside the enclosure, passing through the cap,

a second rectilinear dipole antenna being disposed parallel to the axis of the tube near its inner wall, protruding from the tube at its end located inside the enclosure, passing through the plug,

the device further comprising:

S a piece made of a dielectric material, in which is inserted the antenna, S a piece made of a dielectric material, in which is inserted the antenna. the device comprises:

> two rectilinear dipole antennas, each disposed at a first end of a respective conductor whose second end opens out of the enclosure of a plasma reactor,

external means to the enclosure for supplying a first antenna with a high-frequency sinusoidal voltage delivered by a source, whose frequency is adjustable, and connected to said second end of the conductor whose first end is associated with said first antenna; ,

external means to the enclosure for measuring the power sensed by the second antenna and connected to said second end of the conductor whose first end is associated with said second antenna; means arranged in series on the conductor associated with said first antenna; to subtract from the sinusoidal voltage component the frequency of the plasma generating source and a possible DC voltage component on an inner electrode of said conductor associated with said first antenna,> means arranged in series on the conductor associated with said second antenna for subtracting from the sinusoidal voltage component the frequency of the plasma generating source and a possible DC voltage component q on an internal electrode of said conductor associated with said second antenna,> a display and processing device for developing the power measured by the device according to the fr quence of the source, and for deriving the threshold frequency above which the sensed power increases significantly to determine the value of the electron density of the plasma located around the antennas. Preferred, but not limiting, aspects of this process are:

According to a second aspect, the invention also proposes a method for measuring electronic density in a plasma reactor, said plasma being distributed in a plasma space contained in the reactor, characterized in that it comprises the steps of:

establishing between at least one transmitting antenna located in said plasma and at least one receiving antenna separate from said transmitting antenna also located in the plasma a continuous electrostatic sheath,

generating at the interface between said electrostatic sheath and the plasma a surface wave propagating from said transmitting antenna to said receiving antenna, by applying to said transmitting antenna a periodic electrical voltage whose frequency is controlled - said frequency resignation,

acquire the electromagnetic signal received by said receiving antenna and quantify its power,

> vary the transmission frequency while continuing to acquire said received electromagnetic signal by the receiving antenna, to detect a transmission frequency corresponding to a transmission frequency above which the power of said signal received by the receiving antenna increases significantly,

> derive from the value of said transmission frequency a density of electrons in the plasma near the transmitting antenna and the receiving antenna.

The establishment of said electrostatic sheath is obtained by placing between each transmitting antenna and each associated receiving antenna an electromagnetic continuity structure,

• the density is obtained by the relation n _e (cm ^{~ 3} ) = 2.5x \ 0 ¹⁰ f £ (GHz), with:> n _e = density /

> f, _r = -F = = transmission threshold frequency

> ^r J f P = ^ 2%

P ^ε 0 ^m e

> e = elementary electronic charge,

> ε ₀ = vacuum permittivity,

> m = mass of the electron,

Said antennas are dipole antennas,

Said antennas are disposed at the end of a tube, made of a dielectric material, and inserted into the plasma through a wall of the enclosure of a plasma reactor,

Said antennas are arranged in a plane, separated by a planar substrate made of a dielectric material, and inserted into the plasma through a wall of the enclosure of a plasma reactor, or flush with the wall of the reactor so as to to be in contact with the plasma by disturbing it as little as possible,

The method comprises determining, from the measurement of the power transmitted between two antennas, the minimum frequency, _fr , for which a significant electromagnetic power is transported by a surface wave guided by the interface between the plasma and the electrostatic sheath,

• said significant electromagnetic power corresponds to a microwatt.

The method comprises: feeding a transmitting antenna with a high frequency sinusoidal voltage delivered by a source whose frequency of the delivered signal can be modified, measurement of the electromagnetic power picked up by a receiving antenna arranged near the transmitting antenna, as a function of the frequency of the source.

The method comprises the continuous and increasing variation of the frequency of said sinusoidal voltage, while measuring the electromagnetic power picked up by another antenna in order to determine the frequency above which the power picked up by the antenna and due to the propagation of a surface wave, grows significantly. Other aspects, objects and advantages of the invention will appear better on reading the following description, made with reference to the appended drawings in which:

FIG. 1 is a schematic representation of a plasma reactor that can be used in the context of the invention, FIG. 2 is a schematic representation of a measuring device according to the invention, according to a first embodiment. realization,

FIG. 3 is a more detailed view explaining, in connection with this first embodiment, the configuration of the elements introduced into the plasma reactor, FIG. 4 is a detailed view of the electromagnetic continuity means implemented in the first embodiment of FIG. embodiment of the invention, in which an axis AA 'and a cutting plane BB' are defined,

FIG. 5 is a sectional view of these electromagnetic continuity means, along the plane BB 'defined in FIG. 4; FIG. 6 is a schematic representation of a measuring device according to the invention, according to a second mode realization,

FIG. 7 details, in connection with this second embodiment, the configuration of the transmitting antenna, the receiving antenna, and the associated electromagnetic continuity means, an axis AA 'and a cutting plane BB' being defined on this figure FIG. 8 is a sectional view of the electromagnetic continuity means and antennas implemented in this second embodiment, along the plane BB 'defined in FIG. 7,

FIG. 9 is an overall diagram more particularly detailing the means of voltage generation, control, detection and processing implemented in the invention,

FIG. 10 is a graph illustrating surface wave dispersion relationships,

FIG. 11 is a theoretical graph illustrating the power that must theoretically be transmitted between a receiving antenna and a transmitting antenna, as a function of the frequency of the voltage supplying these antennas,

FIG. 12 is a representation of the electrodes making it possible to generate a plasma in a plasma reactor used in the context of the invention,

FIG. 13 comprises FIGS. 13a and 13b, which respectively represent a general view of a practical embodiment made by the applicants of the transmission probe according to the invention - this embodiment corresponding to a first embodiment. And FIG. 13b is a detailed view of this probe, extracted from the view of FIG.

13a,

• Figure 14 includes two graphs:

a graph a) whose curve shows the shape of the transmitted power observed between a transmitting antenna and a receiving antenna of the device, as a function of the frequency of the supply voltage of the device. This curve shows in particular two frequency thresholds,

> a graph b) which declines the same type of curve by highlighting the influence of the power, P _RF , of the radiofrequency generator injected to maintain the plasma (and therefore the electron density) on the position of the two frequency thresholds just mentioned,

• Figure 15 includes two graphs:

a graph which gives the values of the two threshold frequencies highlighted in FIGS. 14 a and 14b, for different powers,

> a graph b) which establishes that the ratio of the two threshold frequencies is always close to a constant factor,

• Figure 16 includes two graphs:> a graph a) which illustrates the influence of the pressure on the power transmitted,

a graph b) which illustrates the influence of the electron density on the power transmitted,

FIG. 17 illustrates transmitted power results obtained in a hydrogen plasma at a relatively high pressure of 65 pascals.

Presentation of a plasma reactor

Prior to the description of embodiments of the invention which follows, some characteristics of a plasma reactor that may be implemented in the context of the invention will be set forth below.

The invention applies in fact to the measurement of the electron density in an ionized gas or plasma - in particular a plasma in a chamber constituting a plasma reactor.

It is specified that the electron density is the number of electrons per unit of volume, volume density expressed most often in electrons per cubic centimeter. Plasma reactors can be used to coat a sample of a thin layer of material, to etch a sample (eg ion bombardment, by a flow of atoms and reactive radicals, or both together), or more generally to modify the structure or the chemical composition of a surface.

A plasma reactor may also be used as a light source or as a waste gas treatment device for pollution control applications or as a thermonuclear fusion reactor.

FIG. 1 represents, schematically and in section, an example of a plasma reactor to which the present invention applies. This is, for example, a so-called radiofrequency excitation (RF) reactor by capacitive or inductive coupling.

Such a reactor comprises a vacuum chamber 1. Near a first wall 2 of this chamber is placed on a substrate holder 3, a sample to be treated 4. The sample 4 has in this figure the shape of a disk a surface directed towards the interior of the enclosure 1 constitutes the surface to be treated. This form is not limiting.

The enclosure 1 is filled with a gas at low pressure, for example of the order of a few pascals to a few hundred pascals. The gas comes from a source 12 to be injected into the enclosure of the reactor via a gas supply pipe 10, the gas flow rate being regulated by a mass flow controller 11.

When a gas mixture is used, multiple sources, flow controllers and supply pipes are used in parallel. The gas is discharged from the enclosure 1 by a discharge pipe 13 connected to a pumping system 14 consisting of one or more vacuum pumps in series. The pumping volumetric flow rate is adjusted by means of a valve 15.

The pressure in the chamber is controlled with the valve 15 or the flow controller 11. Several means can be used to generate the plasma 16. For example, in a configuration called "capacitive coupling reactive ion etching", a radio frequency voltage is applied to the substrate holder. It is also possible, as shown in FIG. 1, to generate the plasma 16 by means of a source 6 independent of the substrate holder 3.

This source 6 can be for example:

• a polarized electrode with a DC voltage (DC voltage source),

An electrode powered by a radio frequency generator (capacitive source),

• a coil powered by a radiofrequency generator (inductive source)

• a microwave generator.

This source may possibly be associated with means allowing the application of a static magnetic field. In the case of the use of a source 6 independent of the substrate holder, the latter may be polarized by a radiofrequency source 5 to establish self-polarization and thus increase the impact energy of the ions on the surface to be treated .

When the source 6 is a radiofrequency source, the latter may optionally be biased at a higher frequency than that applied to the substrate holder 5 in order to control the electronic density preferentially.

Plasmas concerned by the present invention include chemically reactive plasmas (and in which chemical reactivity in addition to ion bombardment may be used).

The only reactivity of the gas or gas mixture injected into the chamber is sometimes the only phenomenon exploited.

But in general this reactivity is improved or even generated by collisions of electrons on neutral atoms or molecules thus producing radicals, unstable chemical species absent in the gas without the presence of electrons, as well as reactive ions. These radicals that reactive ions are responsible for depositing or etching. In the case of deposition, this is known as plasma-enhanced chemical vapor deposition.

This reactivity initiated by the electrons avoids the need for a significant heating of the gas or substrate holder, which would damage the sample to be treated. The rate of production of radicals produced by electronic collisions is therefore a function of the electron density.

Similarly, the flow of charged particles (electrons and ions) that arrive at the surface to be treated is proportional to the electron density. Chemical reactivity and ion bombardment generally act synergistically in these plasmas.

In a process of plasma deposition or etching, it is important to know the characteristics of the plasma to be able to control the implementation of the process and its reproducibility, in particular to control the speed of deposition or etching as a function of the thickness. deposition or depth of the desired etching. The electronic density is one of those characteristics whose measurement and control is decisive in the implementation of these deposition or etching processes. The invention applies to a measurement of this density in a plasma reactor.

First embodiment of the probe

Referring now to Figure 2, there is shown an electronic density measuring device (which we will name the portion immersed in the plasma "plasma probe") according to a first embodiment of the invention. As will be seen, the operation of the probe according to the invention is based on the determination of the "transmission frequency" of a surface electromagnetic wave propagating in an electrostatic sheath (typically cylindrical) established between the plasma and the probe. The probe is inserted through the wall 7 of a vacuum chamber 1 of a plasma reactor so as to be brought into contact with the plasma 16.

The density measurement is done by implementing an electromagnetic wave transmission between a transmitting antenna 17 and a receiving antenna 18 spaced a given distance (which may be about 1 cm). As will be seen in more detail, it is important that the space between the transmitting antenna and the receiving antenna is not too great.

The probe here comprises only a transmitting antenna and a receiving antenna.

It is however specified that it is also possible to provide more than one transmitting antenna and more than one receiving antenna, said antennas then being arranged in different locations of the reactor - with always a controlled spacing and means of electromagnetic continuity between each antenna transmitter and the receiver antenna (s) associated with it. The transmitting antenna 17 is connected by a coaxial cable 21 to a source circuit 22 which delivers a periodic voltage (typically a sinusoidal voltage) whose frequency can be modified.

This is modified so as to continually scan a given frequency range. In a preferred implementation example, this frequency range includes frequencies between 50 MHz and 6 GHz.

The receiving antenna 18 is connected by another coaxial cable 23 to a device 24 for measuring the power of the electromagnetic radiation picked up by this receiving antenna. This makes it possible to determine the electromagnetic power transmitted between the two antennas as a function of the frequency of the source 22.

A cylinder 19 of dielectric material about 1 millimeter in diameter and 1 centimeter in length is interposed between the two antennas.

The roll material (19) is a chemically inert dielectric such as alumina, glass, fused silica or polytetrafluoroethylene (PTFE).

The two antennas and the cylinder 19 are located in the plasma 16. This figure shows the electrostatic sheath (here referenced

45) which corresponds to the spontaneously forming zone in front of any surface exposed to a plasma, and whose typical thickness is a few millimeters.

Thus, FIG. 2 shows an electrostatic sheath 45 around the elements of the probe and in particular around the cylinder 19 and the antennas 17 and 18.

This electrostatic sheath zone is separated from the plasma 16 by an interface 20.

The cylinder 19 is disposed between the two antennas to ensure electromagnetic continuity between the transmitting antenna and the receiving antenna.

It is specified that more precisely "electromagnetic continuity" means "continuity of the electromagnetic properties of the dielectric structures (in this case electrostatic and dielectric sheath) between the two antennas".

This electromagnetic continuity is obtained because the cylinder prevents the plasma 16 from occupying the space between the two antennas and this cylinder ensures the continuity of the electrostatic sheath 45 and the plasma / electrostatic sheath interface 20 between the two antennas . This continuity of the electrostatic sheath and its interface between the transmitting antenna and the receiving antenna will allow a wave of surface to propagate along the interface 20, from the transmitting antenna 17 to the receiving antenna 18.

The cylinder 19 thus corresponds to an example of electromagnetic continuity means between a transmitting antenna and a receiving antenna.

It is specified that this electromagnetic continuity can also be established without having an additional element between the two antennas, but by providing that the distance separating these two antennas is sufficiently small so that the plasma can not penetrate between the antennas and the electrostatic sheaths generated. by the two antennas meet. In such a case, the "electromagnetic continuity means" simply correspond to the particular dimensioning and arrangement making it possible to obtain the desired continuity without additional element. Returning to the embodiment described with reference to Figure 2, the assembly formed by the cylinder 19, the electrostatic sheath 45 and the plasma 16 constitutes a waveguide (or a transmission line) for the surface waves.

The properties of this guide depend on the properties of the plasma and in particular its electron density.

The two antennas 17, 18 are as will be seen used to characterize in transmission the properties of this guide, and deduce the local electronic density between the two antennas.

The probe implemented in the context of the invention is thus a transmission plasma probe (this being the case for all embodiments of the probe). It differs from the absorption plasma probes that have been mentioned in the introduction to this text.

The wave to be transmitted is generated by the alternating current coming from the source 22, then injected at a first end of the waveguide mentioned above by the transmitting antenna 17, to be picked up at the output of the waveguide by the receiving antenna 18. And the waveguide authorizes the propagation of surface waves only at frequencies higher than a particular frequency called transmission threshold frequency (which may also be designated by the simple term "transmission frequency" in the following of this text), noted f _fr . This transmission frequency can be detected because it corresponds to a singular point in the frequency spectrum of the transmitted power. The transmission frequency corresponds to the frequency above which the electromagnetic power received by the receiving antenna 18 increases significantly. And the value of this transmission frequency is related to the electronic density between the two antennas. We can indeed show that the density n _e is deduced from the transmission frequency by the relation: n _e (cm- ³ ) = 2.5ΛO ^lo f _tr ² (GHz).

FIG. 3 shows in more detail the plasma probe of FIG. 2. This probe was inserted into a plasma reactor, for example through an opening in a wall 7 of the reactor, in order to be in contact with the plasma (on Figure 3 inside the reactor is on the left - unlike the representation of Figure 2).

The probe comprises two semi-rigid cables 21 and 23 inserted in the same electrically insulating and rigid tube 25. These cables are coaxial cables.

The two cables 21 and 23 may be identical.

The material of the tube may be alumina or fused silica, and the outer diameter of this tube may be about 1 centimeter. Each of the two cables has a protruding portion out of the tube

In the reactor.

These two projecting parts are bent so that the two cables each face one end facing each other, the two ends being coaxial. These two ends, stripped, constitute two respective dipole antennas 17 and 18. The other end of each cable is located outside the plasma reactor.

It is specified that each antenna 17, 18 is designated by the term "dipole" antenna because in the case of each of these antennas it is the dipole constituted by the antenna itself and the shielding sheath

(Conductor) coaxial cable of the antenna that creates the electric field that will excite the surface wave.

The two antennas 17 and 18 are inserted in the dielectric cylinder 19, the radius of which is equal to the outside diameter of the cables 21 and 23 (that is to say their diameter with the protection 26 which can surround them before their bare end, if such protection exists).

The curved configuration of the two cables forms a closed loop, the bases of the projections of the two cables being close to each other, while their stripped ends are facing each other, also close, and joined by the cylinder 19.

Note that it is not necessary (although it is often more convenient) to get the two coaxial cables out through the same port in the reactor. The recurved configuration of the two cables allows the plasma to occupy the entire space around the cylinder 19. It also ensures that no conductive surface is located near the cylinder 19 which allows not to disturb the propagation of surface waves along the waveguide. The ends of the coaxial cables 21 and 23 which open out of the enclosure are connected to a measuring circuit which will be described later in this text.

In a preferred configuration of the device, the antennas 17 and 18 are identical and they can be used interchangeably as transmitting antenna or receiving antenna. In the part where the cables 21 and 23 are in the plasma space these two cables can be covered by a dielectric layer 26.

This layer constitutes a protection in the case where it is desired to avoid pollution of the plasma space by the metal of the coaxial cables 21 and 23.

The material constituting this protective cladding must be able to withstand the temperatures of the gas in the enclosure without melting and without polluting the plasma. It must also not react chemically with the gases injected into the enclosure. This protection can be made of alumina, glass, fused silica or with a flexible PTFE tube.

In the part where the coaxial cables are housed in the dielectric tube 25, these cables are not exposed to the plasma.

A plasma-tight and electrically conductive plug 27 blocks the end of the tube 25 which opens into the reactor. This plug is pierced with two holes to pass the cables 21 and 23. This plug prevents the plasma from entering the tube 25.

A gland formed of the elements referenced 7, 29 and 30 is used to seal between the wall of the reactor chamber and the tube 25. The part 29 is a gland cover screwed onto a projection of the wall 7 so as to crush a flexible seal 30 against the rigid tube 25.

By slightly unscrewing the cover 29, the probe can be moved into the reactor while preserving the tightness of the vacuum passage of the probe. The gland allows the probe to be moved without "breaking the vacuum" in the reactor vacuum chamber.

A second plug 28 placed outside the reactor blocks the second end of the tube 25. This second plug makes it possible to seal the vacuum passage of the coaxial cables in the tube, the inside of the tube being kept under vacuum. FIG. 4 is a detailed view of a longitudinal section of the waveguide and antennas of the transmission plasma probe, according to the first embodiment of the invention.

The waveguide has a symmetry of revolution along the axis AA 'and is also symmetrical with respect to the plane containing the axis BB' and orthogonal to the axis AA '.

The plasma is separated from the cylinder 19 by an electrostatic sheath whose thickness s is a few millimeters.

A surface wave generated by one of the antennas will thus be concentrated mainly near the interface 20 which separates the plasma 16 from the electrostatic sheath. Such a wave will propagate parallel to the axis AA '.

Thanks to the electromagnetic continuity (provided here by the cylinder 19, but which is recalled that it can be obtained by a very close placement of the two antennas), the surface waves will propagate along the interface 20, the transmitting antenna 17 to the receiving antenna 18.

The antennas 17 and 18 are inserted along their entire length in the cylinder 19 in order firstly to protect these antennas from the plasma and, secondly, to ensure the mechanical blocking of the cylindrical part 19. This part 19 can also be glued to the insulation 32 which surrounds the cables.

The material of this insulator 32 as well as that constituting the cylinder 19 must be chosen in such a way that they can withstand the temperatures of the gas without deforming, as well as the etching of the plasma. Antennas should preferably not be in contact with the plasma for two reasons:

• avoid polluting the plasma with the metal constituting the antenna,

• and prevent the formation of a continuous self-biasing voltage of the antenna that can disturb the measurements made by the device. The length d of antennas 17 and 18 is approximately 5 millimeters. These antennas may be an extension of the inner conductor of the semi-rigid coaxial cables 21, 23.

For example, the outer diameter of the rigid coaxial cables may be about 2 millimeters, and the distance L between the ends of the antennas approximately 1 centimeter. The dielectric protection

26 whose thickness e is then less than 0.5 millimeter is disposed on the outer conductor 31 of the coaxial cables.

The cladding of the insulator 26 may possibly be extended on the cylinder 19 in order to improve its mechanical locking as well as the protection of the antennas.

FIG. 5 is a schematic transverse section along a plane orthogonal to the axis AA 'and containing the axis BB' of the waveguide shown in the previous figure. This figure shows that the plasma is separated from the cylinder 19 of diameter D by an electrostatic sheath of thickness s separated from the plasma by the interface 20.

Second embodiment of the probe

The transmission plasma probe described above with reference to the first embodiment uses a surface waveguide between the two antennas whose geometry is coaxial. The guide used has a symmetry of revolution about the axis AA ', as represented in particular in FIG. 4.

Figure 6 shows a second embodiment in which the probe is made more compact.

In this embodiment, the probe implements a surface wave guide whose geometry is radial and plane. The plane of this guide may also be coplanar with part of the reactor wall, in order to minimize the disturbance caused by the presence of the probe. This variant also uses two identical semi-rigid coaxial cables 21 and 23, inserted in an insulating and rigid tube 25 (elements identical or equivalent to those of the first embodiment are referenced in the same way). The material of the tube 25 is again alumina, glass or fused silica, and the outer diameter of the tube is about 2 centimeters.

As for the first embodiment, the sealing of the vacuum passage of the probe through the wall 7 of the reactor is provided by a gland formed of the elements 7, 29 and 30. The sealing of the vacuum passage of the cables coaxial 21 and 23 is made using a plug 28 which seals the end of the tube 25 located outside the reactor.

Each of the two cables is terminated by a stripped part forming dipole antenna (respectively antennas 17 and 18). These antennas extend in parallel and are arranged to protrude into the interior of the reactor, out of the end of the tube 25 which is located inside the reactor.

A dielectric plug 48, pierced with two holes for passing the cables 21 and 23, sealingly closes this end of the tube 25 to prevent the plasma from entering the tube.

Antennas 17 and 18 may be inserted throughout their length in insulating cylinders 46 and 47 if it is desired to protect these antennas from the plasma.

The materials of the protections 46 and 47 as well as that constituting the plug 48 must here again be chosen in such a way that they can withstand the chemical etching of the plasma and the temperatures of the gas without being deformed.

The material used must be a chemically inert dielectric such as teflon, alumina, fused silica or glass. Figure 7 is a schematic longitudinal section of the radial waveguide and antennas of the second embodiment of the plasma probe.

The transmitting antenna 17 is located on the axis of symmetry AA 'of the tube 25. The other antenna 18, which acts as a receiving antenna, is disposed near the inner wall of the tube 25.

The waveguide has a symmetry of revolution along the axis AA 'and is also symmetrical with respect to the plane containing the axes AA' and BB '(plane of the figure). The plasma is separated from the plug 48 and any protections

46, 47 by an electrostatic sheath assimilated to vacuum whose thickness is a few millimeters.

Here again, the surface waves generated by one of the antennas will be concentrated mainly in the vicinity of the interface 20 which separates the plasma 16 from the electrostatic sheath.

In this embodiment, such a surface wave will propagate radially from the transmitting antenna 17 located on the axis AA 'towards the edge of the radial waveguide where the receiving antenna 18 is disposed.

In this embodiment it is the plug 48 which ensures the electromagnetic continuity between the two antennas. In this mode, the cap has two functions:

• A sealing function of the tube which it closes the end,

• An electromagnetic continuity function.

The length d of antennas 17 and 18 may be about 5 millimeters. These antennas may again be an extension of the inner conductor of the semi-rigid coaxial cables 21, 23.

The outer diameter of rigid coaxial cables may be around

2 millimeters, and the distance L between the ends of the antennas approximately 7 millimeters. The diameter D of the dielectric tube 25 is about 2 centimeters. The cylinders 46, 47 of protection have a diameter slightly greater than that of rigid coaxial cables. Figure 8 is a view along the axis AA 'of the plane waveguide shown in the previous figure.

The plasma is separated from the tube 25 of outer diameter D by an electrostatic sheath 45 of thickness separated from the plasma by the interface 20. The protective dielectric 46 protecting the transmitting antenna 17 is located on the axis AA '.

The other dielectric protection element 47 is disposed at the edge of the radial plane guide near the inner wall of the tube 25. These two protections 46, 47 are fixed to the dielectric plug 48 which is pressed into the tube 25.

Measuring circuit

In order to determine the electron density with a plasma probe as presented above, it is necessary to use a measurement circuit located outside the reactor chamber, the block diagram of which is given in FIG. 9.

This circuit - as well as the various means it comprises and which will be described in detail later in this text - can be used with the two embodiments presented above for the plasma probe.

The purpose of this circuit is to measure the electromagnetic power transmitted between the two antennas 17, 18 to identify and measure the transmission frequency of the surface wave propagating in the waveguide formed by the dielectric cylinder 19 , the electrostatic sheath 45 and the plasma 16.

To accomplish this task the device must be able to perform the following four functions:

Supplying the transmitting antenna 17 with electrical energy coming from the source 22 in order to excite a surface wave at the input of the waveguide, said supply being made with a frequency that can be varied,

• measure the electromagnetic power picked up by the receiving antenna 18 at the exit of the waveguide,

• plot the captured power according to the frequency of the source 22, and deduce the local electronic density between the two antennas,

• protect the circuits from nuisances that could disrupt their operation.

These four functions are provided by four respective means that are a power supply circuit 22, a power meter 24, a display and processing circuit 40 and a protection circuit 33.

These four means will be described below (FIG. 9 illustrates a configuration corresponding to the second embodiment of the probe, but it is recalled that the circuit can naturally be implemented with all the embodiments of the probe).

• Power supply 22

This circuit delivers a sinusoidal voltage of frequency f whose power is constant whatever the frequency. The frequency of the voltage delivered by this circuit can in turn be modified in a controlled manner. It should be noted that this voltage may more generally be a periodic voltage from which a frequency characteristic can be varied in a controlled manner.

The power supply circuit 22 comprises a high frequency oscillator

35 whose output voltage is controlled so as to have a constant amplitude.

The sinusoidal output voltage of the oscillator 35 is amplified with a wide-band amplifier 36. The output of the amplifier 36 is connected to the transmitting antenna via the coaxial cable 21 which is inserted into the dielectric tube 25 . The oscillator 35 is controlled with a continuous control voltage which serves to adjust the frequency f. This frequency f is thus proportional to the control voltage applied, and after a preliminary calibration the measurement of this control voltage is equivalent to a measurement of the frequency f of the oscillator 35.

The minimum and maximum frequencies of the sinusoidal voltage delivered by the circuit 22 are f _m i _n and f _max, respectively, V _min and V _max are the corresponding control voltages.

In order to continuously scan the frequency interval between f _min and W, the oscillator 35 is connected to the output of a voltage ramp generator 37 which delivers a periodic sawtooth signal whose period is T.

In this way, the frequency of the supply delivered by the oscillator 35 is controlled to grow continuously over an interval T - this frequency thus varies according to successive cycles which follow one another as soon as an interval T has expired.

During a period T the voltage delivered by the generator 37 increases linearly with time, this voltage passing in a period

The period T determines the duration of the measurement of the electron density. This period is adjustable in order to perform a scanning V _min V _max (and hence f _m i _n f _max for the output frequency of the oscillator 35) in a time between, for example, 0.1 second and 1 second. The frequency range delimited by f _min and f _max is obviously a function of the value of the amplitude of the electronic density to be measured. The transmission frequency that is to be determined is related to the electronic density by the relation:

Thus, depending on the prior knowledge that one has of the range of values in which the density to be measured should be, one will adopt different values of f _min and f _max . For example, it can be stated that: • If the density to be measured is between 10 ⁸ electrons per cubic centimeter and 10 ¹⁰ electrons per cubic centimeter, then the frequencies fmin and f _m ax are 63 MHz and 630 MHz respectively.

• If the density to be measured is between 10 ⁹ electrons per cubic centimeter and 10 ¹¹ electrons per cubic centimeter then fmin and f _ma x frequencies of 200 MHz and 2 GHz should be used respectively.

• If the density to be measured is greater than 10 ¹² electrons per cubic centimeter then frequencies above 6 GHz (f _min = 6 GHz) should be used.

The frequency range to be used in the general case is between 50 MHz and 6 GHz - it is specified that for the purposes of this text the "high frequencies" are defined as frequencies above 50 MHz.

The gain of the amplifier 36 is adjustable to adjust the power of the sinusoidal signal applied to the transmitting antenna.

This power must be sufficiently high to allow detection of the surface wave by the receiving antenna 18 and the power meter 24.

At gas pressures less than 13 pascals the wave transmitted between the transmitting antenna and the receiving antenna is poorly absorbed by the plasma surrounding the probe. A power of 1 milliwatt is sufficient. When the pressure of the gas in the plasma reactor chamber is greater than 13 pascals the energy of the wave is dissipated during its propagation in the waveguide, and more power will be needed.

In addition, the power delivered must not exceed one watt so as not to disturb the discharge of the circuit components. • Power meter 24

The power meter 24 is used to measure the power picked up by the receiving antenna 18.

This power meter comprises a rectifier circuit 38. The rectifier circuit is connected to the receiving antenna 18 by the coaxial cable 23.

The rectifier circuit 38 has an input port and an output port. The output port delivers a DC voltage proportional to the power received at the input thanks to the rectification operated, for example, by a shottky type diode. This DC voltage is amplified using a voltage amplifier 39.

The circuit 38 must be able to measure powers of the order of one microwatt to several tens of milliwatts because in general the power captured by the receiving antenna 17 hardly exceeds 1% of the power delivered by the voltage source. If the electrical power delivered by the voltage source is of the order of milliwatt, then it is necessary to be able to measure a power of the order of microwatt (μ W).

The duration of the power measurement by the rectifier circuit 38 is much shorter than 0.1 second (the rectifying diode is not the circuit which sets the duration of the measurement of the electronic density.) The response time of the latter is much lower than the minimum frequency sweep time of the source, ie 0.1 s).

Therefore the duration of the measurement of the electron density is T. This is the time taken to perform the scan between f _mm and f _max .

• Display and processing circuit 40

This circuit 40 comprises two inputs powered respectively by:

• The output of the amplifier 39 of the power meter 24,

• The output of the ramp generator 36. The circuit 40 is responsible for displaying the power transmitted between the two antennas 17, 18 as a function of the frequency of the voltage source 35.

In addition, it analyzes the spectrum in order to deduce the transmission frequency of the surface wave to finally display the electron density. All or part of this device 40 can be realized by software means.

The voltage amplified by the amplifier 39 is proportional to the power of the wave picked up by the receiving antenna 18. The voltage delivered by the ramp generator 36 is in turn proportional to the frequency of the electromagnetic wave emitted by the transmitting antenna 16.

The frequency spectrum of the electromagnetic power transmitted between the two antennas is obtained by developing at the circuit 40 the DC voltage at the output of the amplifier 39 connected to the output of the rectifier circuit 38, as a function of the voltage delivered by the generator ramp 36.

• Protection circuit 33

This circuit serves to protect the circuits 22 and 24 described above against two sources of nuisance which are RF interference and continuous self-biasing.

RF Parasitaae

RF interference occurs when the probe is used in plasma created using an RF source (typically 13.56 MHz).

As the probe uses two antennas 17, 18 inserted in the plasma, the latter capture a small part of the power electromagnetic radiofrequency from the source 6 for generating the plasma.

The parasitic electromagnetic energy picked up by the transmitting antenna

17 is capable of disturbing the operation of the supply circuit 22. The energy picked up by the receiving antenna 18 can, if it is not identified, be read by the device as power from the transmitting antenna 17, this which would disturb the measurement of this power.

Plasmas also have the particularity of generating harmonics of the frequency of the plasma source 6. In the case of a source 6 excited at 13.56 MHz, therefore, also the power at 27.12 MHz, 40, 68 MHz etc.

Two RF filters 41, 42 are then used to reduce the intensity of the spurious signal picked up by the 13.56 MHz antennas and its harmonic frequencies. One of the filters 41 is arranged in series between the supply circuit

22 and the coaxial cable 21 connected to the transmitting antenna 17.

The other filter 42 is arranged in series between the power meter 24 and the coaxial cable 23 connected to the receiving antenna 18.

These filters are high-pass filters whose cutoff frequency is around 50 MHz. The signals delivered by the power supply circuit 22 are in the range 50 MHz-6 GHz. It is therefore necessary to ensure that the cutoff frequency of high pass filters remains below the frequency f _m in order to avoid filtering the measurement signals from the supply circuit. When the source 6 is a microwave source operating at 2.45 GHz it is necessary to filter this frequency using filters 41, 42 tape rejector or low-pass depending on whether the frequency of the source 6 is between f _m i _n and f _max or above:

• If the frequency of the source 6 is between f _n and f _m i _my χ, filters 41, 42 are band reject filters that eliminate the frequency of the source 6

(and its possible harmonics located between f _min and f _ma χ), • If the frequency of the source is above the frequency of the source 6 is between f _m i _n and f _ma χ, the filters 41 and 42 are low-pass filters allowing only frequencies lower than f _max .

Whatever the frequency used for the plasma source 6, this filtering must be taken into account in the interpretation of the spectra produced by the circuit 40.

Self-bias

Continuous self-bias is a DC voltage that occurs on antennas 17, 18 and ground in the presence of plasma when the antennas are not covered by an insulator.

This DC voltage can disturb the operation of the supply circuit 22 or power meter 24. It must be removed with circuits 43, 44 whose function is to block the DC current (DC blocker).

These circuits 43, 44 are constructed with capacitors arranged in series respectively between the filters 41, 42 and the coaxial cables 21, 23.

Elements of theory and modeling

We will in this section expose the main elements of theory and modeling implemented by the invention.

• Introduction

We will begin by recalling here some concepts used in the context of the invention.

A surface wave is an electromagnetic wave that can propagate at the interface between two media whose dielectric constants are of opposite signs. The usual insulating materials have relative dielectric constants ε greater than or equal to 1. The dielectric constant of the vacuum is equal to 1 and that of a conductor is negative.

It has already been recalled that in front of any surface exposed to a plasma, an area called an "electrostatic sheath" is formed, the typical thickness of which is a few millimeters.

There is therefore an electrostatic sheath 45 around the probe and in particular around the cylinder 19 and the antennas 17, 18. This zone is separated from the plasma 16 by an interface 20. The electrostatic sheath is a zone that is poor in electrons because the density electronics is much lower than that in plasma. We can thus assimilate this zone to a vacuum. The relative dielectric constant of an electrostatic sheath is thus approximately equal to 1.

The relative dielectric constant of a plasma for a wave at the frequency / is:

The magnitude f _p is called the electronic plasma frequency f _p

(or plasma frequency). This particular frequency is a function of the electronic density (n _e ). It is defined by:

2 _P ^ω s 2 n _e e ^Δ

ZK £ o ^m e

(e is the elementary electronic charge, ε ₀ is the permittivity of the vacuum and m _e is the mass of the electron). When the frequency f of the current generated by the source 35 is less than f _p , the relative dielectric constant of the plasma is negative, a surface wave is therefore likely to propagate at the interface between the electrostatic sheath and a plasma for frequencies f lower than the plasma frequency (f _p ).

The electron plasma frequency is therefore related to the electron density by the relation: f _p (GHz) = 8, 98 × ¹⁰ -6 γ _e (cm- ³ ). A density of 10 ¹⁰ electrons per cubic centimeter is common in the discharge plasmas therefore an electronic plasma frequency of the order of 1 GHz is also common.

An electromagnetic wave can penetrate a plasma and propagate in its volume only if the frequency of this wave is greater than f _p . Below this frequency a wave emitted outside of a plasma is reflected on its surface and does not penetrate deep. It is then a so-called evanescent wave in the plasma.

On the other hand, even for waves of frequency f less than f _p, surface waves can exist and propagate along the electrostatic plasma-sheath interface which then acts as a waveguide.

We can thus generate such surface waves by sending to a plasma an electromagnetic wave whose frequency is lower than f _P -

Most of the energy of this wave is reflected by the surface of the plasma, but some of this energy is transferred to a surface electromagnetic wave.

By arranging the source of the incident wave, ie a transmitting antenna, in the immediate vicinity of a plasma, the proportion of the energy that is transferred to the surface mode is increased - which makes it possible to detect more easily the surface wave.

In the case of the transmission plasma probe object of the present invention, the transmitting antenna is inserted into the plasma however it is never in contact with the plasma because there always remains an electrostatic sheath between the antenna and the plasma. This transmitting antenna is connected to the piece 19 in order to force a surface wave to propagate around this piece along the electrostatic plasma-sheath interface. At the end of the piece 19 another antenna - receiving - is connected to capture a portion of the energy transported by the surface wave.

• Elements of theory

To exploit the probe according to the invention it is necessary to calculate the relationship between the transmission frequency / _fr and the electronic density that is to be measured.

The wavelength λ of a given wave varies with its frequency f. The relation between λ and f is called "dispersion relation".

The calculation of the dispersion relation makes it possible to determine the properties of a wave such as its propagation speed as well as the frequency range in which it exists. The dispersion relation of a given wave type in a medium depends on the composition of the medium and its geometry. In the case of the surface wave of a plasma it depends on the electronic density, the thickness of the electrostatic sheath and the dimensions of the probe. The operation of the probe according to the invention involves the measurement of the transmission threshold frequency (here also called simply "transmission frequency") which is the frequency above which a surface wave is transmitted in the waveguide between the two antennas. The calculation of the dispersion relation from a suitable theoretical model thus makes it possible to establish a relationship between the electronic density of the plasma surrounding the waveguide and the transmission frequency / _fr .

FIG. 10 shows the dispersion relations calculated for an exemplary transmission plasma probe according to the invention. The dimensions of this probe are those mentioned above with respect to an exemplary practical embodiment of the invention. The radius r of the probe is 1 millimeter, the thickness s of the electrostatic sheath is 0.72 millimeters and the density is 5.10 ⁹ electrons per cubic centimeter. Instead of representing λ as a function of FIG. 10 exposes ^{1% (} [ ^{+ s)} as a function of - £ - in order to have universal charts.

FIG. 10 comprises several curves because there exists for the surface wave generated between the antenna of the device different surface modes known as "angular modes" indicated by an index m. A given angular mode is characterized by a particular mapping of electric and magnetic fields.

It is observed that a surface wave - whatever its angular mode m - exists only between two particular frequencies. The theoretical calculation shows that these two frequencies are respectively equal to ^ 1 and f _p . The transmission frequency, f _tr , is thus equal to ^.

At frequencies above the plasma frequency f _p the energy is transported by a wave propagating in the plasma volume by reflecting on the walls or in the plasma.

At frequencies below the frequency f _p no energy, except that carried by the surface wave, is normally picked up by the receiving antenna. In practice, if the receiving antenna is too close to the transmitting antenna, the effect of the capacitive coupling between these antennas is not negligible and a current flows between the two antennas. This "parasitic" coupling is also done via plasma.

In this respect we will seek on the one hand to minimize the surface of the antennas in contact with the plasma and on the other hand not to bring too close antennas. In the absence of parasitic coupling between the two antennas, it is expected to measure for the transmitted wave a transmitted power spectrum as shown in FIG.

When the frequency of the voltage source is close to / _fr , the electric field of the wave and therefore its energy are located at the interface between the plasma and the electrostatic sheath. The electron density that is measured by determining / _fr is the edge density of the electrostatic plasma-sheath interface.

This frequency f _tr is independent of the thickness of the electrostatic sheath, the dimensions of the probe and the presence of dissipations which absorb the energy of the surface wave during its propagation.

The electric field of the wave accelerates the electrons and gives them energy. At very low ambient pressure - for example, pressures lower than 13 pascals, in the absence of collisions of electrons with the atoms or the neutral molecules of the gas, this energy is integrally restored to the surface wave, which facilitates its detection. .

But in the presence of collisions (at gas pressures greater than 13 pascals) the electrons give energy to the neutral molecules, this energy can not then be restored to the wave. The plasma thus becomes resistive. This resistivity also increases as the electron density decreases.

Increasing the resistivity of the plasma by increasing the pressure or reducing the electron density therefore has the effect of dissipating the energy of the surface wave, thereby attenuating the amplitude of the signal detected by the antenna.

Details on an example of realization

We will give below some additional information on an exemplary practical embodiment of the invention manufactured and used by the Applicant.

• Description of the reactor

Applicants have used a plasma reactor to implant a dielectric guide transmission plasma probe. The plasma is generated in this reactor by a capacitively coupled radiofrequency source.

The sealed chamber of the reactor is made mainly of stainless steel and aluminum in which a vacuum is created by two vacuum pumps placed in series. At rest the pressure in the reactor is about 10 ^-4 Pascal When gas is injected, the pressure in the chamber can be regulated to work at a constant pressure between 1 pascal and 133 pascals.

As illustrated in FIG. 12, which corresponds to a photograph taken from the outside by a reactor window, the plasma is created in the reactor between two plane and circular electrodes 12 cm in diameter, located opposite one of the other, and confined by a grid-shaped cylinder, connected to the ground.

The upper electrode, called RF electrode, is polarized using an RF voltage whose frequency is most often 13.56 MHz. Other frequencies such as 27.12 MHz or 40.68 MHz are also usable.

The RF power comes from a generator consisting of an RF signal source connected to a high power amplifier. In order to maximize the power delivered to the plasma, an impedance matching circuit is interposed between the RF electrode and the output of the amplifier.

A counter-electrode connected to ground is disposed above the RF electrode to prevent plasma initiation in this region above the upper electrode. The RF electrode is not visible in Figure 12 because it is hidden by the metal shield that covers it.

The lower electrode and the reactor walls are connected to ground. The distance between the electrodes is 3 centimeters. The plasma is confined radially by a conductive grid which is also connected to ground.

The injection of the gas into the chamber takes place through the upper electrode which is pierced with holes, it is an electrode called "shower head". The gas is evacuated to the vacuum pumps by four holes drilled in the walls of the reactor (two are visible in the background on the sides of Figure 12).

• Characteristics and implantation of the probe

The photograph of FIG. 13a is a general view from above of a photograph of the probe installed in the reactor - the upper part of the reactor having been removed for the capture of this photograph.

The probe is inserted into the volume occupied by the plasma through an orifice in the containment grid. This probe is constructed from RG-405 / U semi-rigid coaxial cables with a characteristic impedance of 50 ohms. The cables are inserted into an alumina tube 9.6 millimeters in diameter and 1.9 millimeters thick.

The photograph of Figure 13b is a close-up of the waveguide, taken from the view of Figure 13b.

The dipole antennas here have a length of 3 millimeters and the distance between the end of the antennas is 1 centimeter. These antennas are built by letting the inner conductor of the cables pass.

Insulation inside coaxial cables is teflon. A piece of this insulation is used to make the dielectric cylindrical part located between the two antennas. The protective layer is made with 0.05 mm thick Teflon tape wrapped around the cables. The thickness of the Teflon protective layer is about 0.2 millimeters. The seal between the reactor chamber and the alumina tube is provided by a gland and an O-ring.

A stopper built with a teflon cylinder and low pressure compatible glue is used to vacuum the coaxial cables. The ends of the cables that are out of the reactor are connected to SMA type coaxial connectors that are commonly used when transmitting low power signals in the 100 MHz - 6 GHz frequency range used by the plasma probe. transmission. These connectors are used to connect the probe to a measuring device.

• Measuring circuit

In the example described here, the probe is connected to a network analyzer. It is a commonly used measuring instrument for characterizing transmission of an electronic circuit. It has an internal high-frequency source which can fulfill the function of a power supply circuit as well as a device for measuring the power transmitted.

A PC-type computer is connected to the network analyzer via a GPIB port to acquire the transmitted wave spectra as data tables. The graphical study of the spectra drawn from the data tables makes it possible to deduce the frequency of transmission and therefore the electronic density. • Mention of some practical results

An argon plasma was produced in the capacitively coupled reactor described above with an RF voltage whose frequency is 40.68 MHz. The gas pressure is 5 pascals. By increasing the power delivered by the RF plasma generator, the electron density in the plasma is increased because the gas is further ionized by providing it with more energy.

FIG. 14a shows the frequency spectrum of the power received by the receiving antenna measured for a power, P _RF , of 50 W delivered by the RF generator.

The power of the microwave source of the power supply circuit connected to the transmitting antenna is constant and equal to 50 milliwatts regardless of the frequency. Less than 1 percent of this power is captured by the receiving antenna.

There is a low frequency "peak" pattern in the transmitted power spectrum.

As shown in FIG. 14b, this structure moves towards the high frequencies when the power delivered by the RF generator increases, that is to say when the electronic density increases. This peak is delimited by two frequencies f _b and f _h which are reported in FIG. 15 a as a function of the power delivered by the RF generator.

This peak is due to a surface wave that transports energy below the plasma frequency, ie in the lower part of the frequency spectrum.

To verify this, the ratio of two frequencies f _b and f _h which delimit this peak is plotted in FIG. 15b as a function of the power delivered by the RF generator.

It can be seen that the ratio f _b / fh is close to 1 / V2 ≈ 0.7 and that the shape of the spectra obtained (see FIGS. 14a and 14b) is close to what predicted by theory (see Figure 11). The two frequencies f _b and f _h identified in FIGS. 14a and 14b are thus respectively the transmission frequency, f _tr , and the plasma frequency, f _p .

Figure 16a shows a set of spectra measured at different pressures in an argon plasma. The RF power delivered by the RF generator is adjusted slightly around 50 W to maintain / _fr at the same value regardless of the pressure.

The ratio between the two frequencies f _b and f _h is here again 0.7, whatever the pressure. Figure 16b shows power spectra measured at 5 pascals at densities less than 5.10 ⁹ electrons per cubic centimeter.

When the pressure increases (FIG. 16a) or when the electron density decreases (FIGS. 16b and 14), the peak of the surface wave is less and less discernible, which makes the determination of f _p more difficult. This is due to the increase in the resistivity of the plasma which has the effect of accentuating the attenuation of the surface wave. At constant incident power, if the attenuation increases then less power is captured.

On the other hand the detection of / _fr remains possible whatever the conditions.

This is the threshold frequency above which the power picked up by the receiving antenna increases significantly.

And the invention will be all the easier to implement that this threshold will be detected simply (that is to say in particular that the transmitted power will be important and that the transition will be abrupt).

Figure 17 shows a set of spectra measured under conditions close to those used in plasma-assisted material deposition applications (ie at a relatively high gas pressure of 65 pascals). The gas used is hydrogen and the measurements are carried out at 65 pascals. Hydrogen is a reactive gas often present in silicon or diamond deposition applications.

Although we no longer distinguish the peak of the surface wave at high pressure, a good estimate of the transmission frequency is given by the intersection between the ordinate axis and the best interpolating line above the spectrum. threshold.

Conclusion

The invention makes it possible to carry out density measurements by overcoming the drawbacks mentioned in the introduction to this text.

It will be added that the invention makes it possible to carry out such measurements in the context of an industrial process without disturbing this process. In particular the invention can be used to control the smooth running of an industrial process.

This is done by placing the probe at the edge of the plasma to minimize the disturbance caused by its presence. The probe can thus be inserted through a polarized electrode or through a wall of the reactor.

And the density measurements can be exploited in real time, in particular as follows:

• If the device detects a slight variation in the electron density, it can deliver to the RF generator a set point for modifying the radiofrequency or microwave gas excitation signal in order to correct this variation in density,

• If the device detects a sudden change in electronic density, it can generate an alarm to alert the operator to monitor the process. The invention also makes it possible to propose a device that makes it possible to characterize and adjust the characteristics of a plasma prior to an industrial process using this plasma.

In this case, the probe is used to adjust the various control parameters of the plasma reactor such as the gas pressure, the gas flow rate, the composition of the gas mixture or the electric power delivered by the plasma source in order to obtain the electron density. optimal for the intended application.

The applications concerned cover those known for plasmas. These include applications for plasma assisted deposition or etching or applications for which the plasma is used as a light source or as a device for treating gaseous effluents (depollution applications or as a thermonuclear fusion reactor) ). This preliminary characterization and source adjustment operation is carried out during the installation of the reactor or during its restart after a prolonged shutdown as part of a planned maintenance or following a failure.

Once this preliminary operation is performed, the probe can be removed to minimize the disturbance brought by the latter. It can also be stored to control the progress of the process as mentioned above.

In the event of a failure of the process, or of a problem detected by the probe if the latter is used during the process, the invention makes it possible to detect the source of the fault in order to remedy it. Once this operation is performed, the probe can still be used to characterize and adjust the reactor as described above.

Note that several probes can be distributed near the walls of the enclosure to have measurements in different regions of the enclosure. And it is also possible to provide a probe mounted in a sealed manner in the reactor, but can be moved in a controlled manner in the interior space of the reactor (having a knowledge of the location of the probe at any time). In this case we can use a single probe attached to a vacuum translator to move the probe parallel to the wall between two measurements. The interpretation of the measurements delivered by the probe at the different measurement locations will then make it possible to draw up a map of the distribution of the electronic density near the wall in question. Another advantage of the invention is that it makes it possible to verify, in a phase of characterization of the enclosure or of repair, the homogeneity of the electronic density in the region of the enclosure intended to receive a sample to be treated and this, with any plasma.

To do this, it is possible that: • several probes and measuring devices are distributed near the wall intended to receive, in normal operation, the sample to be treated, and / or that a probe is mounted in a mobile manner in the reactor.

A single measuring device can also be connected sequentially to several probes using a switching device. Such a switching device connects a single probe to the measuring device then disconnects and then connect to another probe.

The invention is particularly suitable for measuring low electron densities (10 ⁸ -10 ⁹ electrons per cubic centimeter) at gas pressures of several tens of pascals where the other techniques are inapplicable. This range of pressures and densities corresponds in particular to applications of plasma assisted deposits.

In addition to these deposition plasmas, the present invention also aims to be used in plasma assisted etching applications or in applications for which the plasma is used as a light source or as a device for treating gaseous effluents for remediation applications or as a thermonuclear fusion reactor.

In order to be able to be used in industrial plasma reactors having few openings on the outside, the invention also proposes a compact device which requires the presence of only one vacuum passage in the chamber to insert the probe therein. .

Of course, the present invention is susceptible to various modifications that will occur to those skilled in the art. In particular, the materials, dimensions, types of coaxial cables, and frequency range used as examples may be modified, in particular, depending on the plasma reactor to which the device is intended.

In addition, although a reactor generating the plasma was used in the embodiment example by means of a capacitively coupled radio frequency generator, the invention applies regardless of the mode of excitation of the gas. it is continuous, radiofrequency or microwave.

Claims

1. A device for measuring electronic density in a plasma reactor, said plasma being distributed in a plasma space contained in the reactor, characterized in that it comprises:

Means (35) for generating a periodic electrical voltage associated with a so-called transmission frequency,

Control means (37) for controllably varying said transmission frequency; at least one transmitting antenna (17) connected to said emission voltage generating means for emitting surface waves on the surface of the transmitter; plasma space,

At least one receiving antenna (18) for receiving said surface waves emitted by the transmitting antenna (s), electromagnetic continuity means (19) between the transmitting antenna (s) and the receiving antenna (s), to form a continuous transmission sheath corresponding to the interface between an electrostatic sheath and the plasma between each transmitting antenna and at least one receiving antenna, the boundary (20) of said sheath. electrostatic plasma generator for conveying the surface waves emitted by each transmitting antenna to at least one receiving antenna,

Means (24) for detecting the surface waves received by the receiving antenna (s), processing means connected to said control means and to said detection means, for:

determine a transmission threshold frequency corresponding to an emission frequency above which the surface waves are emitted with a substantial power, and deducing from said transmission threshold frequency an electronic density in the plasma.

2. Device according to the preceding claim characterized in that the electromagnetic continuity means are arranged to prevent the plasma enters between a transmitting antenna and an associated receiving antenna and ensure the continuity of the electrostatic sheath between the two antennas.

3. Device according to one of the preceding claims characterized in that for each transmitting antenna and a receiving antenna associated said transmission sheath is located in the continuity of the electrostatic sheaths respectively associated with said transmitting antenna and said receiving antenna.

4. Device according to one of the preceding claims characterized in that said continuous transmission sheath is obtained by controlling the following parameters:

• sizing of each transmitting and receiving antenna, • arrangement of each transmitting antenna, with respect to each receiving antenna intended to receive the surface wave emitted by said transmitting antenna.

5. Device according to one of the preceding claims characterized in that said electromagnetic continuity means correspond to a dielectric continuity element (19).

6. Device according to one of the preceding claims, characterized in that said processing means comprise a spectral analysis device or selective frequency capable of analyzing the energy electromagnetic received by each receiving antenna as a function of the transmission frequency.

7. Device according to one of the preceding claims characterized in that said means of electromagnetic continuity (19) comprise an electrostatic continuity structure between each transmitting antenna and each receiving antenna intended to receive the surface wave emitted by said transmitting antenna.

8. Device according to the preceding claim characterized in that said electrostatic continuity structure is made of a dielectric material or insulator.

9. Device according to one of the preceding claims characterized in that the device comprises an emitting antenna disposed opposite an associated receiving antenna and coaxially with said receiving antenna, so that the space between the ends of the two antennas is substantially centered on the common axis of the two antennas.

10. Device according to the preceding claim characterized in that said electrostatic continuity structure is a cylinder (19) disposed coaxially between a transmitting antenna and an associated receiving antenna.

11. Device according to one of claims 1 to 8 characterized in that the device comprises a transmitting antenna disposed non-coaxially to an associated receiving antenna, the ends of the transmitting antenna and the receiving antenna being situated substantially opposite one another. the other.

12. Device according to the preceding claim characterized in that said electrostatic continuity structure is a piece which blocks the space between the respective ends of the transmitting antenna and the receiving antenna so as to prevent the plasma occupies said space .

13. Device according to one of the four preceding claims characterized in that the transmitting antenna and the receiving antenna are carried by the same support.

14. Device according to one of the preceding claims characterized in that the antennas are isolated from the plasma.

15. Device according to one of the preceding claims characterized in that the antennas are dipolar.

16. Device according to one of the preceding claims, characterized in that it comprises:

Two dipole antennas (17, 18) inserted into the plasma (16), each disposed at a first end of a respective coaxial cable (21, 23) whose second end opens out of the enclosure (1) a plasma reactor;

A dielectric tube (25) in which the coaxial cables (21, 23) are inserted; • a cylindrical piece (19), made of a dielectric material, disposed between and in contact with two dipole antennas (17, 18) in which the latter are inserted over their entire length; said cylindrical piece for constituting between the two antennas an electrostatic sheath whose interface (20) for separation with the plasma is continuous between the two antennas,

An axisymmetric coaxial assembly (17, 18, 19) formed by the dielectric cylinder (19) and rectilinear dipole antennas (17, 18) whose axes coincide with the axis of symmetry of this assembly;

17. Device according to the preceding claim characterized in that:

The device also comprises a plug (48) made of a dielectric material plugging the end of the dielectric tube (25) in its portion exposed to the plasma (16),

> a first straight dipole antenna (17) being disposed on the axis of the tube (25) protruding from the tube (25) at its end located inside the enclosure (1), passing through the plug ( 48)

a second rectilinear dipole antenna (18) being disposed parallel to the axis of the tube (25) close to its inner wall, protruding from the tube (25) at its end located inside the enclosure (1) passing through the plug (48),

The device further comprising:

a part (46) made of a dielectric material into which the antenna (17) is inserted,

> a part (47) made of a dielectric material into which the antenna (18) is inserted.

18. Device according to claim 16 characterized in that the device comprises:

Two rectilinear dipole antennas (17, 18), each disposed at a first end of a respective conductor (21, 23) whose second end opens out of the enclosure (1) of a plasma reactor, External means (22) for the enclosure (1) for feeding a first antenna (17) with a sinusoidal voltage of high frequency delivered by a source (25), whose frequency is adjustable, and connected to said second end of the conductor (21) whose first end is associated with said first antenna, • external means (24) to the enclosure (1 ) for measuring the power picked up by the second antenna (18) and connected to said second end of the conductor (23) whose first end is associated with said second antenna,

Means (41, 43) arranged in series on the conductor (21) associated with said first antenna for subtracting from the sinusoidal voltage component the frequency of the source generating the plasma and a possible DC voltage component on an internal electrode of said conductor (21) associated with said first antenna, • means (42, 44) arranged in series on the conductor (23) associated with said second antenna for subtracting from the sinusoidal voltage component the frequency of the source generating the plasma and a any DC voltage component q on an internal electrode of said conductor (23) associated with said second antenna,

A display and processing device (40) for developing the power measured by the device (24) as a function of the frequency of the source (22), and to deduce therefrom the threshold frequency above which the sensed power increases. significantly to determine the value of the electron density of the plasma (16) located around the antennas (17,18).

19. A method for measuring electronic density in a plasma reactor, said plasma being distributed in a plasma space contained in the reactor, characterized in that it comprises the steps of: • Establish between at least one transmitting antenna located in said plasma and at least one receiving antenna separate from said transmitting antenna also located in the plasma a continuous electrostatic sheath, • Generating at the interface (20) between said electrostatic sheath and the plasma a a surface wave propagating from said transmitting antenna to said receiving antenna, by applying to said transmitting antenna a periodic electrical voltage whose frequency - said so-called emission frequency - is controlled, • acquiring the electromagnetic signal received by said transmitting antenna receiving and quantifying its power,

• vary the transmission frequency while continuing the acquisition of said electromagnetic signal received by the receiving antenna, to detect a transmission frequency corresponding to a transmission frequency above which the power of said signal received by the receiving antenna increases significantly,

• Deduce from the value of said transmission frequency a density of electrons in the plasma near the transmitting antenna and the receiving antenna.

20. Method according to the preceding claim, characterized in that the establishment of said electrostatic sheath is obtained by arranging between each transmitting antenna and each associated receiving antenna a structure of electromagnetic continuity.

21. Method according to one of the two preceding claims, characterized in that the density is obtained by the relation n _e (cm- ³ ) = 2.5x \ 0 ^w f J (GHz), with: • n _e = density /

• f _tr = - _j = = transmission threshold frequency

J p 2π

P ε _o m _e e = elementary electronic charge,

• ε ₀ = permittivity of the vacuum

• m _e = mass of the electron.

22. Method according to one of the three preceding claims, characterized in that said antennas are dipole antennas.

23. Method according to one of the four preceding claims characterized in that said antennas are arranged at the end of a tube (25), made of a dielectric material, and inserted into the plasma (16) through a wall the enclosure (1) of a plasma reactor.

24. Method according to one of claims 19 to 22 characterized in that said antennas are arranged in a plane, separated by a plane substrate consisting of a dielectric material, and inserted into the plasma (16) through a wall of the enclosure (1) of a plasma reactor, or flush with the reactor wall in order to be in contact with the plasma by disturbing it as little as possible.

25. The method according to one of the six preceding claims, characterized in that the method comprises determining, from the measurement of the power transmitted between two antennas (17, 18), the minimum frequency, "". , for which significant electromagnetic power is carried by a surface wave guided by the interface (20) between the plasma (16) and the sheath electrostatic (45).

26. Method according to the preceding claim, characterized in that said significant electromagnetic power corresponds to a microwatt.

27. Method according to one of the eight preceding claims, characterized in that the method comprises:

Supplying a transmitting antenna (17) with a high frequency sinusoidal voltage delivered by a source (22) whose frequency of the delivered signal can be modified,

Measuring the electromagnetic power picked up by a receiving antenna (18) disposed near the transmitting antenna, as a function of the frequency of the source (22).

28. Method according to the preceding claim characterized in that the method comprises continuously and increasing the variation of the frequency of said sinusoidal voltage, while measuring the electromagnetic power picked up by another antenna (18) to determine the frequency above from which the power picked up by the antenna (18) and due to the propagation of a surface wave, increases significantly.