NL1041245A

NL1041245A - A method to measure the specific resistivity of thin layer material without the need for a second surface contact.

Info

Publication number: NL1041245A
Application number: NL1041245A
Authority: NL
Inventors: Van Der Graaf Hendrik
Original assignee: Stichting Voor Fundamenteel Onderzoek Der Materie
Priority date: 2015-03-19
Filing date: 2015-03-19
Publication date: 2016-10-10
Also published as: WO2016148563A1; NL1041245B1

Abstract

According to the invention, a Micromegas is used for measuring properties of a layer (L) arranged on a carrier substrate (S), such as the layer resistivity (p) when the layer thickness (D) is known or the layer thickness (D) when the layer resistivity (p) is known, or such as the layer resistivity (p) when the relative dielectric constant (k) of the layer is known or the relative dielectric constant (k) of the layer when the layer resistivity (p) is known. With the Micromegas, a current is generated that causes a change in the surface potential of the layer, which in turn causes a change in the amplification factor of the Micromegas. With the relationship between amplification and potential being known, the voltage drop over the layer and the current through the layer are known, and hence the resistivity can be calculated if the layer thickness is known, or vice versa.

Description

ref.: P 2015 NL 005 TITLE: A method to measure the specific resistivity of thin layer material without the need for a second surface contact

FIELD OF THE INVENTION

The present invention relates in general to methods for measuring characteristics of thin layers. As an important characteristic of thin layers, the specific resistivity is mentioned, and the present invention will be specifically explained for this example, but the gist of the present invention is not limited to measuring the specific resistivity.

BACKGROUND OF THE INVENTION

In MEMS technology, layers with a specific resistivity p can be made. The specific resistivity can be influenced by doping, and by a variation of the composition of the layer material. For instance, insulating S13N4 can be made in a CVD process with p > 1015 Qm in a merging process of Silane (S1H4) and ammonia (NH3). With an excess of Silane, Silicon Rich Nitride (SRN) is formed that may be considered as Si-doped Silicon Nitride. A specific resistivity as low as 104 Qm is within reach.

It is desirable to be able to measure the actual specific resistivity of a layer. This is typically done by applying a current perpendicular to the layer: sensing the current magnitude and the voltage drop will allow to calculate a resistance value. However, this as such does not yet yield the specific resistivity: for calculating the specific resistivity, it is necessary to have the current applied evenly distributed over a certain surface area, and to know the size of that surface area.

In a prior art method, that only works when the layer is arranged on a conducting carrier, a contact electrode is brought into contact with the exposed outer surface of the layer. It is then assumed that the contact surface area of the contact electrode is equal to the cross-sectional area of the current distribution. However, for relatively thick layers and/or a relatively high value of p, the contact electrode must cover a large area with respect to the square of the layer thickness: a simple contact probe will not do. Further, it is difficult to assure that a solid contact electrode actually makes contact with the layer over the entire contact area: the probability is high that the contact actually is an undetermined plurality of point contacts with large areas of no or bad contact in between. The actual contact size and hence the measured value for the specific resistivity may then depend on the contact pressure.

In order to try to overcome these problems, it is known to use a liquid contact electrode, usually mercury, or alternatively an electrode can be deposited on the layer but this involves a risk of electron or hole injection. Using mercury involves all kinds of environmental hazards. Further: these methods are destructive: after the measurement has completed, the layer of which the specific resistivity is now known can not be used any more for its original purpose. So these methods can only be used to measure the results of varying process parameters: if the process parameters of a manufacturing process are maintained the same, one has to assume that specific resistivity of the resulting layers remains the same.

SUMMARY OF THE INVENTION

An object of the present invention is to overcome or in any case reduce the above problems. More particularly, an object of the present invention is to provide a non-destructive method allowing for accurately measuring the specific resistivity of thin layers, directly and in-situ, without the need to apply a contact electrode.

According to the invention, a Micromegas is used to generate a homogenous shower of electron avalanches, that is applied on the layer under investigation. This allows an accurate measurement of the total current, surface area, and voltage drop, without contact and without destruction of the layer.

BRIEF DESCRIPTION OF THE DRAWINGS

These and other aspects, features and advantages of the present invention will be further explained by the following description of one or more preferred embodiments with reference to the drawings, in which same reference numerals indicate same or similar parts, in which indications "below/above", "higher/lower", "left/right" etc only relate to the orientation displayed in the drawings, and in which: figure 1 schematically illustrates the design of a Micromegas; figure 2 is a schematic diagram of a measurement device based on the Micromegas technology in accordance with the present invention; figure 3 is a graph showing the results of a test measurement; figure 4 is a diagram schematically showing electron current; figure 5 is a graph showing some measuring results.

DETAILED DESCRIPTION OF THE INVENTION

First, with reference to figure 1, a description and explanation will be given of a Micromegas. It is noted that a Micromegas as such is a known device, albeit for other purposes, therefore the description will be kept brief.

Figure 1 schematically shows the basic design of a Micromegas. In origin, the Micromegas is a detector, i.e. a modern 2D variant of a Geiger-Muller detector. A carrier 1, typically a plate-shaped carrier, for instance a printed circuit board (PCB), has an upper surface with a plurality of anode segments 2. At a small distance above the anode segments 2, typically in the range of about 20-100 pm, a conductive microgrid 4 is arranged. The space between anodes 2 and grid 4 is indicated as amplification gap 3. The microgrid 4 basically has a thin planar shape, with openings that allow electrons to pass, as will be explained in more detail later. The microgrid 4 is typically embodied as a metal foil with holes. It would for instance also be possible to implement the microgrid as a grid of crossed wires. In the following, the microgrid 4 will also simply be indicated as "grid".

Above the grid 4, i.e. at the side of the grid 4 facing away from the anodes 2, a drift electrode 6 is arranged, at a distance which is not critical but which may typically be in the order of about 3-15 mm, with 10 mm being the distance used in a practical example. The space between drift electrode 6 and grid 4 is indicated as drift volume 5.

Reference numeral 7 indicates an external source of ionisation radiation P, for instance X-ray quanta. Reference numeral 8 indicates a source for applying high voltage potentials HV6 and HV4 to the drift electrode 6 and the grid 4, respectively, with respect to the carrier 1, which is held at ground potential. The potential of the anodes 2 is indicated as V2. In the basic Micromegas design V2 will be zero. The grid potential HV4 is always negative with respect to V2, and the drift electrode potential HV6 is always negative with respect to the grid potential HV4; with the drift electrode having the lowest potential, the drift electrode is also indicated as cathode.

In the drift volume 5, a drift electrical field in the order of 100 V/mm is set. In an embodiment where the spatial distance between the drift electrode 6 and the grid 4 is about 10 mm, such field is obtained by a potential difference between the drift electrode 6 and the grid 4 in the order of 1000 V. In the amplification gap 3, a strong electrical field in the order of 10 kV/mm is set. In an embodiment where the spatial distance between the grid 4 and the anodes 2 is about 50 pm, such field is obtained by a potential difference between the grid 4 and the anodes 2 in the order of 500 V.

In other words: V2 = 0V, HV4 = -500V, HV6 = -1500V.

The operation is as follows. The space above the anodes 2 is filled with a gas. When a high-energetic quantum or charged particle P passes the drift volume 5, ionization processes create one or more electron-ion pairs. The electrons created here are indicated as primary electrons pe. In the electric field in the drift volume 5, primary electrons drift down towards the grid 4 and (positive) ions drift upwards to the cathode 6. When a primary electron passes a hole in the grid 4 and enters the amplification gap 3, the strong electrical field in the amplification gap 3 accelerates it towards the anodes 2. The accelerated primary electron may ionize gas atoms or gas molecules, releasing a secondary electron, which is accelerated as well and may ionize the gas in its turn. Eventually, a single primary electron pe passing the grid 4 causes an electron avalanche A that reaches the anodes 2. The respective charge signals of the segmented anode 2 can be read out, possibly by applying a pixel chip, and thus it is possible to measure in 2D the position of the original incoming single electron.

In this respect, it is to be noted that the holes (or other openings) in the grid 4 have a pitch equal to the pitch of the anode segments 2, and that the grid 4 is I arranged such that the holes in the grid 4 are aligned with the centres of respective anode segments 2. It would also be possible to have more than one hole per segment. In an embodiment, the anode segments 2 are arranged in a square array of pitch 55 pm, and the holes in the grid 4 have an opening diameter of about 25 pm.

Essential for the operation of a Micromegas detector is the gas amplification process occurring in the avalanche gap 3. Assume that a single incoming X-ray quantum P creates NP electron-ion pairs; thus, NP indicates the number of primary electrons per quantum. Assume further that each primary electron passing the grid 4 causes an electron avalanche containing G electrons, with G indicating a gas I amplification factor or gas gain. It then follows that, per absorbed quantum, the total avalanche charge Q detected at the anode (i.e. integrated over all anode segments 2) can be expressed by the formula Q = NP x G x q (1) with q indicating the elementary charge (i.e. the charge of one electron).

For one single event, NP and G will have one single value. However, these values are not exactly fixed: they each show statistical fluctuation. Generally, the gain G will depend on the geometry of the detector, the applied gas mixture, and the potential difference over the amplification gap 3. In practice, the geometry of the detector is fixed and the applied gas mixture is usually selected for practical reasons i and can be considered to be constant. The inventor has realized that the average gain G can easily be varied by varying the potential difference over the amplification gap 3, which will be indicated as avalanche voltage VAv-

The Micromegas detector was primarily developed for use in experimental > physics, for detecting particles originating from for instance a particle accelerator.

The signal derived from the electrons is very fast, typically in the order of a few nanoseconds, and is particularly suitable for detecting the timing of the particle. Further, by monitoring which anode segments receive the avalanche signals, it is possible to reconstruct the path of high-energetic charged particles. In any case, for the Micromegas detector as known in the art, the particle source 7 is an external source.

Figure 2 is a schematic diagram of an exemplary measurement device 100, based on the Micromegas technology as described above. Reference numerals corresponding to the components in figure 1 are the same yet increased by 100; these components will not be described again. A carrier substrate S having a layer L is placed on an anode 102; for the purpose of the present invention, it suffices if the carrier substrate S is an integral substrate and the anode 102 is an integral anode, but the anode may consist of separated anode elements in order to allow for a current distribution to be measured. Part of the device 100 in this embodiment is a radiation source 107, which generates a continuous wide-angle beam of X-ray quanta, of which the intensity is known although this intensity is not critical.

The radiation from source 107 continuously creates electron-ion pairs distributed over the entire drift volume 105, with the number of pairs created per unit time being substantially constant, at least on average. Thus, evenly distributed over the entire surface of the grid 104, primary electrons constantly pass the grid 104 and generate respective avalanches. Thus, evenly distributed over the entire surface of the layer L, a continuous shower of electron avalanches is generated, which the average number of avalanches per unit time being substantially constant. This corresponds to a constant current la homogeneously distributed over the layer L. The average number of primary electrons passing the grid 104 per unit time, and hence the current la, will inter alia depend on the mutual distance between the grid 104 and the source 107, indicated as the height of the source 107.

Reference numeral 108 indicates a window mask that shields off a portion of the layer L. Thus, only the portion of the layer L corresponding to the opening of the window mask 108 receives the electron avalanches. Where in the above the "entire" drift volume 105, grid 104, and layer L are mentioned, this is meant to indicate only the portion that projects on the opening of the window mask 108. The contour of the opening is not essential; it may for sake of convenience be for instance rectangular or circular. The cross-sectional area of this opening will be indicated as A.

The window mask 108 is useful for doing partial measurements on a portion of a layer L having a larger surface area. The size of the layer portion under investigation will then of course be equal to the size of the mask opening. It is noted that the mask may have a plurality of such openings. On the other hand, the window mask 108 may also be omitted, it is not essential for practicing the present invention.

In the following, the portion of the drift volume 105 corresponding to the size of layer L or the size of the window mask 108 opening, as the case may be, will be indicated as active portion.

Assume in figure 2 that the number of quanta absorbed per unit time in the active portion of the drift volume 105 is indicated as NQ. Referring to formula (1), if eventual losses are ignored, it will be clear that the constant (i.e. DC) current la that can be measured at the anode 102 can be written as la = NQ x Q = NQ x NP x G x q (2)

In an experimental embodiment of device 100, the source 107 was 55Fe, producing X-ray quanta of 6 keV. With the specific source used, NP was equal to about 230. Practical values for G are 1 k - 10 k.

The gas in the spaces 103 and 105 was a mixture of Argon and CO2 in a ratio of 90/10, at normal pressure and temperature.

Micromesh 104 was implemented as a 17 pm thick sheet of copper with holes having a diameter of 30 pm in a square pattern layout with a pitch of 60 pm. The width of the avalanche gap was 50 pm. The cross-sectional area A of the opening of the window mask 108 was 10 mm x 10 mm.

The voltage HV6 on the drift electrode 106 was set to -1000 V.

The voltage HV4 on the grid 104 was set to -380 V.

The potential at the anode 102 was set to 0 V (ground).

Figure 3 is a graph showing the results of a test measurement of charge pulses detected at the anode 2. The graph is a pulse height spectrum, in which the horizontal axis represents the pulse height, i.e. the charge of the respective charge pulses, while the vertical axis represents counts, i.e. the number of times a certain pulse height was measured. The large peak around a pulse height of channel 500 corresponds to the simultaneous arrival of about 230 electrons at the grid 104. The horizontal position of this ‘photo’ peak is a direct measure for the gain G, which in the test device was about 5000. It is noted that a smaller peak can be observed at a pulse height of about channel 240, and an even smaller peak at about channel 50, but these peaks, which are well-known to persons skilled in this art, are without relevance for the present discussion.

As already indicated above, the gain G depends on the potential difference VAv over the amplification gap 3, i.e. the potential difference between the grid 104 and the top surface of the layer L. With the knowledge of G, it is possible to calculate the potential at the top surface of the layer L, and, knowing the (zero) potential at the bottom surface of the layer L, it is possible to calculate the potential drop over the layer L. Several methods for doing this will be described below.

More specifically, it appears that the gain G can be written as ln 2.^- G = const · e vd (3) in which const indicates a constant, and VAv indicates the potential difference Vav over the amplification gap 3 (positive value), also indicated as "avalanche voltage".

Vd is indicated as a doubling voltage. From the above formula (3), it can easily be recognized that, if Vav is raised by the value of Vd, ceteris paribus, the gain G will double. In formula: ^ = 2 (4a) with Vx indicating any reference value of the avalanche voltage Vav-

Oppositely, if Vav is reduced by the value of Vd, ceteris paribus, the gain G will half. In formula: ^ = 0-5 (4b)

Per apparatus, the value of the doubling voltage Vd may differ, and the same may apply for different gas fillings or for different settings of the same apparatus. But in any case, the apparatus 100 will have a doubling voltage Vd, which can be measured and which can be considered to be an apparatus feature. The actual value of Vd can easily be measured by monitoring the anode current la as a function of the grid potential HV4, and can be visualized as a horizontal shift of the peak position in figure 3. In the experimental embodiment discussed here, it was found that Vd = 18 V.

The issues mentioned above can be summarised as follows: by measuring a change of the gain G, potential V2 of the top surface of the layer-under-test can be obtained.

It is to be realized that the above explanation does not only apply to doubling the gain G. For each ratio ζ, it will be possible to find a corresponding ratio voltage νζ such that the following formulas apply: ^ = ζ (5a) and G<^> - 1/ (5b\

G(Vx) * /ζ VOLV

According to one embodiment of the present invention, as will be explained later in more detail, the avalanche voltage VAv is set at a first value Vi, and the corresponding gain G(Vi) is measured. Then, the avalanche voltage VAv is increased (or reduced), always with the gain being monitored, until the avalanche voltage VAv reaches a second value V2 at which the corresponding gain G(V2) is twice as high (formula 4a) or half as high (formula 4b) or ζ times as high (formula 5a) or 1/ζ as high (formula 5b) as G(Vi). It then follows that IV2-V1I = Vd (formula 4a/4b) or |V2-V11 = \/ζ (formula 5a/5b). In fact, it is not necessary to change the avalanche voltage Vav until a predetermined ratio is reached for the gain: it is possible to just have two different settings V-i and V2, measure the corresponding gains G(Vi) and G(V2) and the resulting ratio ζ, and look up in for instance a predetermined look-up table the corresponding value νζ for ΙΝ^ΛΛΙ.

Varying the avalanche voltage VAv can be done by varying the potential V2 at the free surface of layer L while keeping the grid 104 voltage HV4 constant. Even if the exact value of that potential V2 is not known with certainty, from the above it should now be clear that it is possible to measure the variations of that potential V2. Put in different words: set the potential V2 at a first value V2(1) and set the potential V2 at a second value V2(2): maybe the absolute values V2(1) and V2(2) are not known, but the difference V2(1)-V2(2) can be measured.

According to another embodiment of the present invention, such difference V2(1 )-V2(2) can even be measured without having knowledge of any doubling voltage Vd.

Set the grid 104 voltage HV4 at a first value HV4(1) and set the potential V2 at a first value V2(1). The avalanche voltage VAv will now have a first value V-i. Measure the corresponding gain G(V-i).

While maintaining the grid 104 voltage HV4 at said first value HV4(1), set the potential V2 at a second value V2(2). The avalanche voltage VAv will now have a second value V2. Measure the corresponding gain G(V2).

While maintaining the potential V2 at the second value V2(2), change the grid 104 voltage HV4 to a second value HV4(2), in such a way that the measured gain again has the first value G(Vi).

It can then be deduced that the avalanche voltage VAv again has the first value V-i. From this it follows that V2(1 )-V2(2) = HV4(1 )-HV4(2).

In formula: G(V2(1), HV4(1)) = G(V2(1 )+AV, HV4(1 )+AV) (6)

It is noted that the entire measuring setup will include the anode 102, and will also include a measuring chamber in which the radiation source, the grids, the anode and the sample will be placed, which measuring chamber can be evacuated and filled with the desired gas. In a preferred embodiment, the grids 104 and 106 and the radiation source 107, and the possible window mask 108, are combined in one unit with a frame holding the grids 104 and 106 and the radiation source 107, and the possible window mask 108 in position. This unit, which will be indicated as the

Micromegas unit, can then be placed on a sample to be investigated in the measuring chamber.

FIRST EMBODIMENT A sample (substrate S + layer L) is placed on the anode 102 in the measuring chamber and the Micromegas unit is placed on top of the sample. A measuring amplifier for measuring charge pulses is connected to the anode 102 and/or to the grid 104. The output of the measuring amplifier can for instance be connected to a scope, or a multi-channel analysor. The chamber is sealed, and flushed with a suitable gas, in the current example Ar/C02 90/10. With reference to figure 3, a pulse height spectrum is obtained from the anode charge pulses. This allows an estimated value to be obtained for the gain G, in terms of mV of the scope readout, or a channel number of the photo-peak position.

Figure 4 is a diagram schematically showing the current of drifting primary electrons Ip and the anode current la. The anode current la can be measured directly, for instance by a multimeter. Although the current actually is composed of charge bunches (avalanches) initiated by primary electrons entering a grid hole, integrated over the entire anode and averaged over time the current appears as a real DC current through the layer L under test, due to the layer’s virtual capacitance. The current la is homogeneously distributed over the active area of the layer L if the active drift/conversion volume is homogeneously irradiated. In that case, the exposed surface of that active area of the layer L will be an equipotential plane having a potential V2 (see figure 2). Assuming that the bottom of the layer L is in good and contiguous contact with the grounded substrate S, the bottom of the layer L is an equipotential plane having a potential equal to zero.

It is now possible to express the resistance R of the layer L as R = V2/Ia (7)

It is important to realize that the potential V2 at the surface bordering the avalanche gap 103 is not equal to the anode 102 potential, because in such case there would be no current flow perpendicular to the layer. As a matter of principle, the potential V2 is not known. Basically, in prior art it has been attempted to measure V2 directly, with the problems as described. According to the invention, the potential V2 is varied by varying the vertical position of the radiation source 107, which will vary the radiation intensity and hence will vary NQ, while the amount of potential variation in V2 is established on the basis of the above considerations.

With a relatively large distance between the source 107 and the drift electrode 106, the irradiation rate is low and the current la through the layer L is low. The voltage drop over the layer L is now low. In first approximation, it is assumed that V2 = V102 = 0 (8) applies. The value of G now observed is indicated as GO.

When the source 107 is moved towards the drift electrode 106, NQ will increase, hence the current la through the layer L will increase, hence the potential drop over the layer L will increase by a certain voltage Vx. With the anode 102 potential and hence the potential of the bottom of layer L being kept constant, i.e. zero, the surface potential V2 of the layer L will become more negative. As a consequence, with the grid 104 potential being kept constant, the potential difference between the layer L and the grid 104 will decrease, in other words the avalanche voltage Vav over the avalanche gap 103 will reduce by said certain voltage Vx. And consequently, the gain G will reduce by a certain factor. With reference to figure 3, it is noted that the change of gain G can be observed as a shift of the peak position in the pulse height spectrum.

In the first embodiment, the source 107 is moved towards the drift electrode 106 until G = GO/2. With reference to formula (4b), we then know that V2 = Vd (9a)

SECOND EMBODIMENT

The second embodiment is identical to the first embodiment, with the exception that the source 107 is moved towards the drift electrode 106 over an arbitrary distance, the gain Gx is measured (see figure 3), and the ratio ζ = G0/Gx is calculated. With reference to formula (5b), we then know that V2 = V? (9b)

THIRD EMBODIMENT

The third embodiment is identical to the first embodiment, with the exception that the source 107 is moved towards the drift electrode 106 over an arbitrary distance. Then, the potential HV4 of grid 104 is made more negative by a value ΔΝ/4 and the resulting gain Gx is measured (see figure 3), until this potential HV4 reaches a value where G is found to be equal to GO. We then know that the avalanche voltage Vav over the avalanche gap 103 has returned to its original value, hence M2 = AV4 (9c)

In each of these three embodiments, with the voltage drop over the layer L and the current through the layer L now being known, the specific resistivity p can be expressed as p~- (10)

Ia-D with A indicating the surface area of the active portion of layer L and D indicating the thickness of layer L.

Hence it follows that specific resistivity p can be established if the layer thickness L is known, which can be measured with other known methods.

Conversely, it may be that the specific resistivity p is known, or assumed to be known, for instance by assuming that the same process parameters will result in the same specific resistivity p. In such case, the present invention allows to measure the layer thickness D in exactly the same procedure as described, and calculating ο = (11) la· p

FOURTH EMBODIMENT

Perpendicular to its layer surface, the layer L has a virtual capacitance C in parallel to a resistance R, with C = So k A/D and R = p-D/A, with k indicating the relative dielectric constant of the layer material, which is assumed to be known. Thus, the layer L has a relaxation time constant τ = RC = So k p, which does not depend on the active area A nor on the layer thickness D. The value of the time constant τ can be measured in a relaxation process.

First, the layer L is exposed to an intense radiation, causing a large current la that charges the layer L to a relatively high voltage V2 = Vc. As explained in the above, this will cause a reduction of the gain G. The precise value of Vc is not essential, but for instance Vc may be equal to V<j.

Then, at a certain time to, the radiation is stopped, either by removing the source 107 or, more conveniently, placing a radiation blocking element between the source 107 and the layer L. With time t, the layer L will now discharge according to said time constant τ, so that the potential V2 at the exposed surface of the layer L can be written as |V2| = |Vc|-exp(-T/t) (12)

Then, after a certain waiting time At, the radiation is applied again, for instance by briefly removing the radiation blocking element, and the pulse height spectrum is measured (see figure 3). The duration of the measurement, i.e. the removing of the radiation blocking element, should be sufficiently brief such that the inevitable recharging of the capacitance can be ignored. The gain G(At) is measured.

In advance, using any of the methods discussed above, it is possible to establish a relationship between G(At) and V2(At), for instance in the form of a lookup table. Using formula (12), the time constant τ is then calculated.

It is possible to repeat this measurement at regular time intervals t,. The obtained results V2(ti) can be fitted to formula (12). Or the logarithm of V2(tj)/Vc is noted against the corresponding time t, to obtain a linear fit for τ.

Alternatively, the formula G(t) / GO = exp(-V2(ti) ln2/Vd) (13) can be used to calculate V2(tj).

In either case, a value is obtained for the relaxation time constant τ.

Then, the specific resistivity p can be calculated according to p = T/(£0k) (14)

Conversely, it may be that the specific resistivity p is known, or assumed to be known, for instance by assuming that the same process parameters will result in the same specific resistivity p. In such case, the present invention allows to measure the relative dielectric constant k of the layer material in exactly the same procedure as described, and calculating k = τ / (ε0-ρ) (15)

An advantage of the fourth embodiment is that it is not necessary to know the layer thickness D.

EXPERIMENT A sample layer L was prepared of the material Si3N4, applied by CVD deposition to a thickness D of 1.02 pm on a Si substrate covered with a layer of aluminium. The relative dielectric constant k of the layer material was assumed to be equal to 5 (for p-Si3N4).

With the Micromegas device as described above, having Vd = 18 V, the sample was irradiated at low intensity in a gas atmosphere of Ar/CC>2 90/10. With reference to figure 3, a pulse height of 20 mV was recorded visually from an oscilloscope, as a representation of the gain G.

In accordance with the first embodiment, the source 107 was displaced towards the drift electrode 106 until the oscilloscope showed a reduced pulse height of 10 mV, indicating that the gain G was reduced by 50%. The anode current la was measured to be 15 nA at that moment.

Using formula (7), the resistance R of the layer L was calculated as R = 18/(15-10 9) = 1.2 Go.

Using formula (10), the specific resistivity p of the layer L was calculated as p = R A/D = 1.2-109 · 100-1 O'6 / (1.02-10"6) = 1.2-1011 Qm.

In accordance with the fourth embodiment, the source 107 was moved close to the grid 106 such as to submit the sample to intense radiation. As a representation of the gain G, the oscilloscope showed a pulse height of 9 mV. Then the radiation was blocked, and the layer L was allowed to relax.

As a good approximation for the relaxation time constant t, the "half-life" time was taken for the gain to increase from 10 mV to 15 mV (oscilloscope reading); this time appeared to be 5 seconds.

Using formula (14), the specific resistivity p of the layer L was calculated as p = t / (£0 k) = 5 / (8.85-10’12 5) = 1.1-1011 Qm.

It can be seen that the results of the first and fourth embodiments are in good conformity with each other.

Another sample, also made of S13N4, having a thickness in the range 1-4 pm and a k of 5, was also charged using the fourth embodiment, and allowed to relax. After relaxation times of 250, 500, 1125,1750 seconds, the momentary gain was measured by briefly irradiating the sample. Figure 5 is a graph showing the measuring results as black squares. A relaxation fit curve was calculated, and is shown in the graph as a dotted line. A relaxation time τ was calculated to be 450 s as best fit.

Using formula (14), the specific resistivity p of the layer L was calculated as p= τ /(£0 k) = 450 /(8.85-1012 -5)= 1.02-1013 Qm.

Another sample, also made of S13N4, having a thickness in the range 1-4 pm and a k of 5, was also charged using the fourth embodiment, and allowed to relax. Even after a relaxation time as long as 12h, no noticeable change was observed in the gain. In conclusion, τ was estimated to be larger than 3-105 seconds, yielding an estimated specific resistivity p higher than 6-1015 Qm.

The relation between V2 and time, after charging up, is exponential. The relation, however, between gain G and time is double exponential. Datapoints are to be fitted according: G(t) = G0-exp(-ln2 · V2/VD), and V2 = V2’-exp(-t/ τ) where V2’ is the value of V2 at the moment of blocking the irradiation.

Summarizing, according to the invention a Micromegas is used for measuring properties of a layer L arranged on a carrier substrate S, such as the layer resistivity p when the layer thickness D is known or the layer thickness D when the layer resistivity p is known, or such as the layer resistivity p when the relative dielectric constant k of the layer is known or the relative dielectric constant k of the layer when the layer resistivity p is known. With the Micromegas, a current is generated that causes a change in the surface potential of the layer, which in turn causes a change in the amplification factor of the Micromegas. With the relationship between amplification and potential being known, the voltage drop over the layer and the current through the layer are known, and hence the resistivity can be calculated if the layer thickness is known, or vice versa.

It is noted that the invention has succeeded in providing methods capable of measuring the specific resistivity of virtually insulating layers. It is true that the fourth embodiment, relaxation half-life measurement, will take a long measuring time in such case, but the method of the invention allows this because there is no leakage of charge: neither through the edges of the area of passing current, nor via the layer surface due to the ultra dry gaseous environment and the electric avalanche field, pointed against electrons leaving the surface. The measuring setup thus provides for an extremely good electrical insulation. This makes it possible to accurately measure relaxation half-times in the order of a fortnight or longer, with an accuracy capable of distinguishing between specific resistivities of 1018 am and 1019 Qm. This in turn makes it possible to measure non-linear (non Ohmic) behaviour, since the electric field inside the layer is continuously known.

It is noted that the invention has succeeded in proposing methods and devices for applying a potential difference over a layer, and for measuring properties of such layer, without having to contact that layer.

It should be clear to a person skilled in the art that the present invention is not limited to the exemplary embodiments discussed above, but that several variations and modifications are possible within the protective scope of the invention as defined in the appending claims. For instance, instead of a radiation source for generating primary electrons, it is possible to liberate primary electrons from a photocathode, for instance a gold-plated photocathode that is irradiated by a UV LED or laser. The UV light source can be pulsed in operation. Such embodiment could offer the advantage that controlling the avalanche intensity can, instead of mechanically displacing the source, easily be effected by varying the output power of the UV light source.

Even if certain features are recited in different dependent claims, the present invention also relates to an embodiment comprising these features in common. Any reference signs in a claim should not be construed as limiting the scope of that claim.

Claims

Use of a Micromegas for measuring the properties of a layer (L) applied to a support substrate (S).

Use according to claim 1, for measuring the specific resistivity (p) of the layer when the layer thickness (D) is known.

Use according to claim 1, for measuring the layer thickness (D) when the resistivity (p) of the layer is known. Use according to claim 1, for measuring the resistivity (p) of the layer when the relative dielectric constant (k) of the layer is known.

Use according to claim 1, for measuring the relative dielectric constant (k) of the layer when the resistivity (p) of the layer is known.

Use according to claim 2 or 3, wherein use is made of the relationship "V2-A pD-Ia \" where A indicates the area of an active part of the layer, where V2 indicates the potential difference between the potential at the free layer surface and the potential at the interface between layer and substrate, wherein (Ia) denotes the total current perpendicular through said active layer portion.

The use of claim 6, wherein the Micromegas comprises a plate-shaped drift electrode (106) and a plate-shaped microgrid (104) parallel to the drift electrode (106); wherein the substrate (S) is placed on an anode (102); wherein the Micromegas is placed over the layer (L) with the micro grid (104) facing the layer; wherein a first potential (0 V) is applied to the anode (102); wherein a second potential (HV4) is applied to the micro grid (104) negative with respect to the first potential (0 V); wherein a third potential (HV6) is applied to the drift electrode (106) negative with respect to the second potential (HV4); wherein primary electrons are generated in a drift volume (105) between the drift electrode (106) and the microgrid (104); wherein a primary electron that passes through the microgrid (104) will cause an electron avalan to the layer (L) under the influence of the electric field caused by the potential difference (Vav) across the gap (103) between the microgrid (104) and the layer (L); wherein the electron lawins together define the anode current (Ia); wherein a gain relationship is determined between a gain (G) and said potential difference (Vav) across the gap (103); wherein the generation speed of the primary electrons is varied to vary the anode current (Ia), which in turn causes a variation of the potential difference V2 across the layer (L), which in turn causes a variation of said potential difference (VAv) over the gap (103); wherein the gain (G) is measured; and wherein V2 is calculated from the measured gain (G) using said gain relationship.

Use according to claim 7, wherein: in a first step, the generation speed of the primary electrons is set to a low level to cause a relatively low anode current (Ia), which in turn causes a relatively low first value ( V2 (1)) of the potential difference V2 across the layer (L) and consequently develops a relatively high first value (Vi) of the said potential difference (Vav) across the gap (103) and consequently a relatively high first value (G1) of the gain (G); in a second step, the generation speed of the primary electrons is set at a high level to cause a relatively high anode current (Ia), which in turn causes a relatively high second value V2 (2) of the potential difference V2 to occur develops over the layer (L) and consequently a relatively low second value (V2) of said potential difference (VAv) across the gap (103) and consequently a relatively low second value (G2) of the gain (G); wherein the order of the first and second steps is irrelevant; wherein the ratio ζ = G1 / G2 is determined; wherein V2 (1) is assumed to be zero and where V2 (2) is assumed to be equal to N / ζ, where \ / ζ is a predetermined voltage that satisfies formulas 5a and 5b.

Use according to claim 7, wherein: in a first step, the generation speed of the primary electrons is set to a low level to cause a relatively low anode current (Ia), which in turn causes a relatively low first value ( V2 (1)) of the potential difference V2 across the layer (L) i and consequently develops a relatively high first value (Vi) of said potential difference (Vav) across the gap (103) and consequently a relatively high first value (G1) of the gain (G); in a second step, the generation speed of the primary electrons is set at a high level to cause a relatively high anode current (Ia), which in turn causes a relatively high second value (V2 (2)) of the potential difference V2 across the layer (L) and consequently a relatively low second value (V2) of said potential difference (Vav) across the gap (103) and consequently a relatively low second value (G2) of the gain (G); • where the order of the first and second steps is irrelevant; wherein the first and second steps are performed such that G1 / G2 = 2; wherein V2 (1) is assumed to be 0 and wherein V2 (2) is assumed to be equal to Vd, wherein Vd is a predetermined double voltage that satisfies formulas 4a and 4b.

The use of claim 7, wherein: in a first step, the generation speed of the primary electrons is set to a low level to cause a relatively low anode current (Ia), which in turn causes a relatively low first value to occur (V2 (1)) of the potential difference V2 across the layer (L) and consequently develops a relatively high first value (V ^ of said potential difference (Vav) across the gap (103) and consequently a relatively high first value (G1) of the gain (G); in a second step the generation speed of the primary electrons is set to a high level to cause a relatively high anode current (Ia), which in turn causes a relatively high second value (V2) (2)) of the potential difference V2 across the layer (L) and consequently develops a relatively low second value (V2) of said potential difference (VAv) across the gap (103) and consequently a relatively low second value (G2) of the gain (G); a third step, while the generation speed of the primary electrons is kept constant, the voltage (HV4) on the grid (104) is changed by an amount (AV) such that the gain has the first value (G1); wherein V2 (1) is assumed to be zero and wherein V2 (2) is assumed to be equal to AV.

The use according to claim 4 or 5, wherein use is made of the relationship k-p = t / ε0 wherein τ indicates the relaxation half-life of the self-decomposition of the layer (L); and wherein ε0 denotes the dielectric constant of vacuum.

The use according to claim 11, wherein the Micromegas is used to charge the layer (L), and wherein the value of τ is measured in a relaxation process.

The use of claim 12, wherein the Micromegas comprises a plate-shaped drift electrode (106) and a plate-shaped micro grid (104) parallel to the drift electrode (106); wherein the substrate (S) is placed on an anode (102); wherein the Micromegas is placed over the layer (L) with the micro grid (104) facing the layer; wherein a first potential (0) is applied to the anode (102); wherein a second potential (HV4) is applied to the microgrid (104), negative with respect to the first potential (0); wherein a third potential (HV6) is applied to the drift electrode (106), negative with respect to the second potential (HV4); wherein primary electrons are generated in a drift volume (105) between the drift electrode (106) and the microgrid (104); wherein a primary electron passing the microgrid (104) will cause an electron avalan to the layer (L) under the influence of the electric field caused by the potential difference (VAv) across the gap (103) between the microgrid (104) and the layer (L); wherein the electron lawins together define the anode current (Ia); wherein a gain relationship is determined between a gain (G) and said potential difference (VAV) across the gap (103); wherein in a charging step the generation speed of the primary electrons is set to a high level to cause a relatively high anode current (Ia), which in turn causes a relatively high value of the potential difference V2 to develop over the layer (L) and consequently a relatively low initial value (GO) of the gain (G); wherein the initial value (GO) of the gain (G) is measured; wherein the generation of the primary electrons is interrupted in a relaxation step; wherein after a certain relaxation time (t) the generation of the primary electrons is briefly resumed to measure the current value (Gt) of the gain (G); wherein the instantaneous value (V2t) of the potential difference (G2) across the layer (L) is calculated from the measured instantaneous value (Gt) of the gain (G) using said gain relationship; and wherein the relaxation half-life τ of the self-decomposition of the layer (L) is calculated from the relaxation time (t) and the instantaneous value (V2t) of the potential difference V2 over the layer (L), assuming exponential relaxation.

The use of claim 13, wherein the instantaneous gain is measured at multiple relaxation times.

Use of a Micromegas for generating a potential difference over a layer (L) applied to a support substrate (S).

Use according to any of the preceding claims, wherein primary electrons are generated with a radiation source.

The use according to claim 16, wherein the generation speed of the primary electrons is varied by moving the radiation source in a direction perpendicular to the layer surface.

Use according to any of the preceding claims, wherein primary electrons are generated with a photocathode irradiated by UV radiation.

The use of claim 18, wherein the excitation rate of the primary electrons is varied by varying the UV radiation.

A method for measuring properties of a layer (L) applied to a conductive support substrate (S), wherein a Micromegas is used according to any of the preceding claims.

The method of claim 20, wherein the layer (L) can have a specific resistivity of higher than 1015 to, even higher than 1018,

The method according to claim 20 or 21, for measuring non-linear (non-Ohms) behavior of the specific resistivity p of the layer (L).

A method for applying a potential difference over a layer (L) applied to a carrier substrate (S), wherein a Micromegas is used according to any of the preceding claims.

Measuring device (100) for measuring the properties of a layer (L) applied to a conductive support substrate (S), which device comprises a controlled chamber with therein: - an anode (102) for placing the substrate thereon; - a grid (104) arranged parallel to the anode; - a drift electrode (106) arranged parallel to the grid; - a drift volume (105) between the grid and the drift electrode; - a reinforcement gap (103) between the layer (L) and the grid; means for establishing a controlled gas atmosphere in the chamber; means for creating electron ion pairs in the drift volume; voltage supply means for applying different voltages (HV6, HV4, 0) to the drift electrode, the grid and the anode, respectively, the voltage supply means being adapted to adjust the potential difference between the drift electrode and the grid such that the electrons of said electron ion pairs in the drift volume drift toward the grid, and wherein the voltage supply means are adjusted to adjust the potential difference between the grid and the anode so that electrons passing through the grid and entering the gain gap are accelerated toward the layer (L) to create an electron avalanche.

The measuring device of claim 24, further comprising a window mask (108) for shielding a portion of the layer (L) and having an opening for defining an active portion of the layer (L) with a defined area (A) ).

The measuring device according to claim 24 or 25, wherein said means for creating electron-ion pairs in the drift volume comprises a radiation source (107).

The measuring device of claim 26, wherein the radiation source is movable in the direction perpendicular to the upper surface of the anode (102).

A measuring device according to any of claims 24-27, wherein said means for creating electron-ion pairs in the drift volume comprises a photoelectrode.

29. Measuring device according to any of claims 24-28, wherein said means for creating electron-ion pairs in the drift volume comprise a UV source.

The measuring device of claim 29, wherein the output power of the UV source is controllable.

A voltage generating device for applying a potential difference over a layer (L) applied to a conductive support substrate (S), which device comprises a controlled chamber with therein: - an anode (102) for placing the substrate thereon; - a grid (104) arranged parallel to the anode; - a drift electrode (106) arranged parallel to the grid; - a drift volume (105) between the grid and the drift electrode; - a reinforcement gap (103) between the layer (L) and the grid; means for establishing a controlled gas atmosphere in the chamber; means for creating electron ion pairs in the drift volume; voltage supply means for applying different voltages (Hv6, Hv4, 0) to the drift electrode, the grid and the anode, respectively, the voltage supply means being adapted to adjust the potential difference between the drift electrode and the grid such that the electrons of said electron ion pairs in the drift volume drift towards the grid, and wherein the voltage supply means are adapted to adjust the potential difference between the grid and the anode so that electrons passing through the grid and entering the gain gap are accelerated towards the layer (L) to create an electron avalanche.

The voltage generating device according to claim 31, further comprising a window mask (108) for shielding a portion of the layer (L) and having an opening for defining an active portion of the layer (L) with a defined area (A) ).

The voltage generating device according to claim 31 or 32, wherein said means for creating electron-ion pairs in the drift volume comprises a radiation source (107).

The voltage generating device of claim 33, wherein the radiation source is movable in the direction perpendicular to the upper surface of the anode (102).

A voltage generating device according to any of claims 31-34, wherein said means for creating electron-ion pairs in the drift volume comprises a photoelectrode.

A measuring device according to any of claims 31-35, wherein said means for creating electron ion pairs in the drift volume comprises a UV source. i

The measuring device of claim 36, wherein the output power of the UV source is controllable.