EP1346201A1

EP1346201A1 - Device for ellipsometric two-dimensional display of a sample, display method and ellipsometric measurement method with spatial resolution

Info

Publication number: EP1346201A1
Application number: EP01989637A
Authority: EP
Inventors: Dominique Ausserre; Marie-Pierre Valignat
Original assignee: Centre National de la Recherche Scientifique CNRS; Universite Pierre et Marie Curie Paris 6
Current assignee: Centre National de la Recherche Scientifique CNRS; Sorbonne Universite
Priority date: 2000-12-18
Filing date: 2001-12-18
Publication date: 2003-09-24
Also published as: US20070188755A1; JP2008122394A; FR2818376B1; JP4163951B2; HK1065102A1; CN1275030C; US20040085537A1; JP4863979B2; FR2818376A1; WO2002050513A1; AU2810802A; KR100923271B1; CA2431943C; IL156480A0; PL363425A1; CN1489688A; AU2002228108B2; US7307726B2; IL193361A0; PL204813B1

Abstract

The invention concerns a device for ellipsometric two-dimensional display of a sample placed in an incident medium, observed between an analyser and a polarizer intersected by convergent light reflection, wherein the ellipsometric parameters of the assembly formed by the sample and a substrate whereon it is placed, are used. The substrate comprises a support and a stack of base layers and its ellipsometric properties are known. The ellipsometric properties of the substrate are such that variations of the sample ellipsometric parameters are displayed with contrast higher that the contrast produced in the absence of said substrate. The invention also concerns a display method and an ellipsometric measurement method with spatial resolution.

Description

TWO-DIMENSIONAL ELLIPSOMETRIC VISUALIZATION DEVICE OF A SAMPLE, VISUALIZATION METHOD AND ELLIPSOMETRIC MEASUREMENT METHOD WITH SPATIAL RESOLUTION

The present invention relates to a device for two-dimensional ellipsometric display of a sample, a method of visualization and a method of ellipsometric measurement with spatial resolution. It is particularly well suited for viewing in ellipsometric contrast or in interference contrast.

10 A sample receiving and reflecting light generally changes its polarization.

It is possible to use this property to visualize a sample or to characterize it by measuring its ellipsometric parameters generally designated by ψ and Δ.

15 On this subject, one could, for example, refer to the work

Azzam and Bashara published in 1979.

Initially, we sought to exploit the extinction of the Fresnel coefficient r _p at the Brewster angle to access a precise ellipsometric measurement of the parameters ψ and Δ (ellipsometry) or a

20 sensitive display fi ms very thin, especially on the surface of ^"water (Brewster angle microscopy).

Furthermore, we sought to illuminate an area of a sample under a single incidence and azimuth to measure the parameters ψ and Δ corresponding to this area.

In the context of the present invention, we are interested in a simultaneous exploitation of the parameters ψ and Δ for a certain number of points of a sample, each defined by their x and y coordinates. This is called a two-dimensional ellipsometric visualization or measurement of a sample.

In addition, in the context of the present invention, we are interested in samples of small dimensions observed, viewed or measured under an optical microscope in reflection. It can be traditional microscopy, differential interference contrast microscopy or fluorescence microscopy. This type of microscopic observations poses particular constraints in the measurement, on the one hand, where the objectives of microscopes have a significant numerical aperture which creates conditions of observations significantly different from the usual conditions of ellipsometric measurements, in which the beams so much lighting than measurement (or reflected beams) are generally collimated beams of small aperture and, on the other hand, where the lighting beams are most often distributed uniformly around the normal incidence, that is to say -to say in a range of angles of incidence not very favorable to ellipsometry.

Admittedly, visualization methods based on the use of an antireflective substrate have been proposed previously, but they are based on the “incoherent reflectivity” of the substrate. The substrates proposed above are therefore antireflective for non-polarized light or for polarized light with a direction of polarization constant with respect to the plane of incidence, which is incompatible with the use of a microscope. The principle is based on the minimization of the second member of the equation (E4).

where r _p and r _s are the complex reflection coefficients of each of the polarizations on the substrate concerned which implicitly depend on x and y, Φ _N (Θ, NP) being the normalized flux reflected for an angle of incidence θ, in non-polarized light.

Note that total extinction is only obtained for | r _p | = | r _s | = 0, which is an extremely restrictive condition, since the values of the two Fresnel coefficients are fixed. The condition of total extinction, | r _p + r _s | = 0, is much more flexible since it only results in a relationship between the two Fresnel coefficients,

r _P = -r _s E6 Anti-reflective substrates for polarized light have also been proposed to increase the performance of ellipsometers, but ellipsometry and optical microscopy have so far been considered irreconcilable. The object of the invention is therefore a two-dimensional ellipsometric display of an object of very small thickness invisible under an optical microscope under known observation conditions compatible with the use of a commercial optical microscope. Despite this, according to the invention, it is possible both to visualize the object and to measure its thickness and its index under a microscope.

To this end, the object of study is deposited on a particular substrate, the association of the object of study and the substrate forming the observed set that we call - the sample -. The substrate is designed in such a way that the object of study, although very thin, suffices by its presence to modify the appearance of the substrate, thus leading to the visualization of the object.

To this end, the substrate consists of a support covered with a stack of layers such that, on the one hand, the thickness e of the last layer satisfies the condition d ² / of ² [Ln | r _p + r _s |] = 0 and such that, on the other hand, the minimum of the quantity | r _p + r _s | over all the values of e is as low as possible.

Similarly, the presence of the object is sufficient under these conditions to modify measurably under an optical microscope the parameters ψ and Δ of the substrate, so that the optical characteristics of the object can be extracted from the measurement of the parameters ψ and Δ of the sample.

Thus, the substrate is designed in such a way that the sensitivity of the parameters ψ and Δ of the sample to a small perturbation of its constituent parameters is very large for low angles of incidence, therefore very different from the Brewster angle, while the visualization and measurement methods proposed are moreover designed in such a way that the radial geometry of the microscope is become compatible with the use of these ellipsometric characteristics.

In a preferred embodiment, using a differential interference microscope (DIC) (by means of a device inserted in the vicinity of the rear focal plane of the objective, for example a Nomarski device or a Smith device), the beam d lighting, linearly polarized along the azimuth φ = 0, is decomposed by the DIC device into two beams linearly polarized in the directions φ = 45 ° and φ = -45 ° and offset laterally with respect to each other d '' a small quantity Δd, the two wave planes associated with these two polarizations registering at the reflection on the sample phase variations due to the presence or inhomogeneities of the object, these phase variations transforming into variations of color or intensity after passing the return beam reflected in the DIC device, then in the cross analyzer with the polarizer. In this observation mode, the contrast of the object is optimized thanks to the adjustment of a compensator included in the DIC device. This adjustment consists in extinguishing the interference between the two beams reflected by the non-interesting regions of the sample by adjusting their phase shift at the level of the device where they interfere, i.e. at the level of the analyzer, the quality of this extinction conditioning the quality of the visualization. The mathematical condition of this extinction is the same as previously, namely r _p + r _s = 0. The condition of maximum sensitivity over the thickness e of the last layer of the stack in this observation mode is d ² / of ² [Ln | r _p + r _s |] = 0.

The proposed visualization method is therefore generally optimal for all observations under a microscope between crossed polarizer and analyzer, including when a DIC device is included in the microscope.

The invention therefore relates to a device for two-dimensional ellipsometric display of a sample, comprising an object, placed in an incident medium, observed between an analyzer and a polarizer crossed by reflection in convergent light, in which the ellipsometric parameters of the assembly formed by the object (4) and a substrate (8) on which it is placed, are used. According to the invention:

the substrate comprises a support and a stack of layers and that its ellipsometric properties are known,

- the ellipsometric properties of the substrate being such that the variations in the ellipsometric parameters of the sample due to the object are displayed with a contrast greater than the contrast produced in the absence of this substrate. The present invention also relates to the characteristics which will emerge from the description which follows and which should be considered in isolation or in all their technically possible combinations:

- the sample is illuminated through a wide aperture objective such as a microscope objective,

- the microscope is a differential interference contrast microscope,

- the microscope is a fluorescence microscope,

This embodiment is the most effective for viewing or detecting objects of nanometric dimensions. It is then a question of visualizing without solving. It allows in particular the visualization of all isolated filiform objects, that is to say distant by an amount greater than the lateral resolution of the microscope, whose length is greater than one micron (polymers, microtubules, collagen, bacteria, DNA , AN, carbon nanotubes, nanowires, etc.).

the thickness e of the layer of the stack in contact with the object is such that the complex reflection coefficients r _p and r _s of the substrate satisfy the condition d ² / of ² [Ln | r _p + r _s | ] = 0,

- the optical properties of the substrate are such that the minimum value taken by the quantity | r _p + r _s | over the set of values of e is as low as possible,

the device comprises a polychromatic light source, the device comprises a monochromatic light source, the support is made of silicon, More generally, the support is advantageously an absorbent medium, a metal or a semiconductor whose real part of the optical index is greater than 3.3.

- The stack consists of a single layer, This layer is advantageously mineral, consisting of a mixture SiO / Si0 ₂ in suitable proportions.

- the layer is a layer of silica,

the thickness of the silica layer is of the order of 1025 Å, the incident medium being air, the layer is a layer of magnesium fluoride,

the thickness of the layer of MgF ₂ is of the order of 1055 Å, the incident medium being air,

- the layer is a layer of polymer,

the layer is a layer of polymer, with an optical index approximately equal to 1.343, the incident medium being air,

the layer is a mineral layer, with an optical index approximately equal to 1.74, the incident medium being water,

the layer is a mineral layer, with an optical index approximately equal to 1.945, the incident medium being an oil with an optical index 1.5,

- the layer is discontinuous and formed of silica pads and index 1.343, of the same height defining the thickness of the layer and of cross-sectional dimensions much less than a micrometer, the incident medium being air, - the layer is a mesoporous or nanoporous organic or mineral layer with an index approximately equal to 1.343, the incident medium being air,

- the layer is a mineral airgel with an index approximately equal to 1.343, the incident medium being air, - the device comprises a microscope comprising an aperture diaphragm in the form of a longitudinal slot orientable around the axis of the microscope making it possible to restrict the light cone to a single plane of incidence in a chosen direction, the device comprises a microscope comprising an aperture diaphragm in the form of a ring limiting the lighting cone of the sample around an angle of incidence,

- the object is a thin film and the stack comprises a beveled layer s whose thickness varies monotonously in one direction

X along the surface.

This display method and device are compatible and advantageously superimposable on any scanning optical microscopy technique, on any invisible light optical technique (UV 0 or IR), on any spectroscopy technique, on any non-linear optical technique, any diffusion or diffraction technique, and all their combinations. They are in particular compatible with the fluorescence, micro-Raman, confocal microscopy, two-photon microscopy techniques, and all their combinations.

The implementation of the present invention with fluorescence microscopy is particularly advantageous. Indeed, the polarization of the light emitted by a fluorescent sample is often different from the polarization of the incident beam. The fluorescent marker therefore introduces a depolarization of the light to which the device of the invention is particularly sensitive. In addition, the extinction factor of the incident light specific to the device of the invention considerably reduces the noise accompanying the fluorescent signal. Finally, this implementation of the present invention with fluorescence microscopy makes it possible to recognize, among identical fluorescent objects, those of them which depolarize light, which corresponds to a very particular environment for molecules.

This implementation is particularly effective for the observation of surfaces immersed in a fluorescent medium. It is also very advantageous for reading the fluorescence signal of biochips, including the observation of hybridization kinetics. The invention also relates to a measurement method in which:

- the display device is cut parallel to the s direction X into two elements, - the thin film is deposited on one of these elements,

the two elements are placed between a crossed polarizer and an analyzer under a polarizing microscope lit in polychromatic light, so as to form fringes of colored interference on each of the elements,

- Measuring the offset of the fringes formed respectively in each of the elements to deduce the properties of the layer deposited on one of them.

The invention further relates to a device for viewing a sample as specified above, in which the substrate is the bottom of a Petri dish.

The invention further relates to a device for viewing a sample as specified above, in which the sample is a matrix multisensor, each pad or patch of the matrix possibly constituting the last layer of the stack. This multisensor can be a biochip with bacteria, viruses, antigens, antibodies, proteins, DNA, RNA, or chromosomes, the device then constituting a parallel reading device.

The invention also relates to a method for ellipsometric measurement of a sample with spatial resolution under a polarizing microscope forming an image of the sample in which:

- the sample is illuminated by a linearly polarized light beam through an aperture diaphragm, - the light reflected by the sample is analyzed by a polarizer-analyzer, characterized by the relative orientation φ of its direction of polarization compared to that of the polarizer,

- A modulation of the reflected intensity is ensured by the relative rotation of the polarization of the light beam and of the polarizer-analyzer.

According to this process:

- the aperture diaphragm of the lighting beam is a ring centered on the axis of the beam delimiting a single angle of incidence, - the average flux φM (x, y) reflected and its modulation amplitude φ _m (x, y) are measured simultaneously at each point of the image obtained from the sample,

- we treat the measurements φjyi y) and φ _m (x, y) to deduce simultaneously at each point of the sample two combinations of the ellipsometric parameters ψ (x, y) and Δ (x, y) and the reflection coefficient | r _s | ² (x, y) from the formulas:

l ι ι 1,,

- | r _s | ² (l + tan ”= φ _M and - | r _s | ² (tan ψ - 2tanψcosΔ) = φ _m

- we treat the measures φM (x, y) and φ _m (x, y) to deduce the combination sin (2ψ) cos of the only ellipsometric parameters ψ (x, y) and Δ (x, y) from the formula :

Optionally, in a measurement step:

- the orientation of the analyzer relative to the polarizer is fixed at a value different from π / 2 modulo π, - the aperture diaphragm of the lighting beam is a slit orientable around the optical axis of the microscope superimposed on a ring delimiting a single angle of incidence,

- the intensity of the reflected beam is measured for at least two different non non-redundant orientations of the slit, - these intensity measurements are processed from the relation:

- the values of the two ellipsometric angles ψ (x, y) and Δ (x, y) and those of the modules of the reflection coefficients | r _p | and | r _s |.

Optionally, in a complementary step: - the analyzer is fixed in an orientation not perpendicular to the polarizer, for example φ = 0,

the aperture diaphragm of the lighting beam is a slit orientable around the optical axis of the microscope superimposed on a ring delimiting a single angle of incidence,

- the reflected intensity is measured for the two orientations φ = 0 and φ = π / 2 of the slit,

- these intensity measurements are processed to obtain tanψ by taking the square root of their ratio according to the three formulas:

I = A; cos, 2 φ for φ = 0 modulo π

π

I ≈ AflrJ sin ² φ for φ = - modulo π 2

Optionally, in a measurement step:

the orientation of the analyzer relative to the polarizer is fixed at a value different from π / 2 modulo π,

- a modulation of the reflected intensity is ensured by the rotation of the diaphragm D around the optical axis,

- the average flux φ _M (x ₃ y) reflected and its modulation amplitude φm (x _> y) are simultaneously measured at each point of the sample,

- we treat the measures φM ( ₃ y) and φ _m (x ₅ y) to deduce the two ellipsometric angles ψ (x, y) and Δ (x, y) and the modules | r _p | and | r _s | reflection coefficients from the relationship:

Optionally, in a complementary step: - the orientation of the analyzer relative to the polarizer is fixed at φ = 0,

- a modulation of the reflected intensity is ensured by the rotation of the diaphragm D around the optical axis, - the average flux φ _M (x _. y) reflected and its amplitude are simultaneously measured at each point of the sample modulation φ _m (x, y),

- we treat the measures φ _M ( ₅ y) and φ _m (x, y) to deduce the two ellipsometric angles ψ (x, y) and Δ (x, y) and the modules | r _p | and | r _s | reflection coefficients from the relationship:

+

The invention also relates to a method for ellipsometric measurement of a sample with spatial resolution under a polarizing microscope forming an image of the sample, in which:

- the sample is illuminated by a linearly polarized light beam through an aperture diaphragm,

the light reflected by the sample is analyzed by a polarizer-analyzer, characterized by the relative orientation φ of its direction of polarization with respect to that of the polarizer,

- a modulation of the reflected intensity is ensured by the relative rotation of the polarization of the light beam and of the polarizer-analyzer, according to this method:

- the aperture diaphragm of the lighting beam is a disc centered on the axis of this beam,

- the average flux φ _M ( _> y) reflected and its modulation amplitude φ _m (x, y) are measured simultaneously at each point of the image obtained from the sample,

- we treat the measures φ _M (, y) and φ _m (x ₅ y) to deduce simultaneously at each point of the sample two combinations effective ellipsometric parameters ψ _ef (x, y) and Δ _efï (x, y) and the effective reflection coefficient | r _se ² (x _. y) from the formulas:

I l I 1 1 I

- r _s ² (1 + tan ² ψ _eff ) = φ _M and - r _s ² (tan ² ψ _eff - 2tanψcosΔ _eff ) = φ m

- we treat the measurements φ (, y) and φ _m y) to deduce the combination sin (2ψ) cos of the only effective ellipsometric parameters ψ _eff (x ₅ y) and Δ _eff (x, y) from the formula :

sin2ψ _eff cosΔ _eff = 1 - Φm

Φ M

The invention also relates to an ellipsometric measuring device under a microscope with lateral spatial resolution. According to this device:

- it has only one polarizer located between the lighting mirror and the sample on either side of the objective,

- It has a rotating slit in the plane of its aperture diaphragm, possibly superimposed on a ring diaphragm allowing the ellipsometric parameters of the sample to be extracted using at least three measurements for three different orientations of the slit and formula:

2L | 2 sin 2φ

I ≈ Af i fc rs tan ² ψ cos ⁴ φ + sin ⁴ φ - tan ψ cos Δ

applied to these three measurements, in which the parameters r _s , ψ and

Δ are possibly effective parameters derived from means over all the angles of incidence present:

According to this device:

- the polarizer and the analyzer have a fixed relative orientation,

- the aperture diaphragm is a hole or a ring,

- the image of the rear focal plane of the objective is formed in the focal plane of the eyepiece by a Bertrand lens,

- a CCD camera is placed in this plane,

- the intensity measurement obtained at each point of the CCD camera is used using the general formula:

in order to directly obtain all the ellipsometric parameters of the sample. According to this device:

- a modulation of the reflected intensity is obtained by a relative rotation of the analyzer and the polarizer,

- the aperture diaphragm is a hole or a ring,

- a CCD or possibly tri-CCD camera is placed in this plane,

- the measurement of the intensity obtained at each point of the CCD camera or possibly at each point and for each color component of the tri-CCD camera is used using the general formula:

φ) +

- a CCD camera is placed in this plane, - the measurement of the intensity obtained at each point of the camera

CCD is operated using the general formula:

in order to directly obtain all the ellipsometric parameters of the sample.

The camera is a tri-CCD color camera and the intensity measurement at each point is made and used for each of the colors. Advantageously, the object studied is placed on a substrate.

The thickness e of the layer of the stack in contact with the object is such that the complex reflection coefficients r _p and r _s of the substrate satisfy the condition d ² / of ² [Ln | r _p + r _s |] = 0.

Preferably, the object is placed on a substrate whose optical properties are such that the minimum value taken by the quantity jr _p + r _s | over all the values of e is as low as possible.

An embodiment of the invention will be described more precisely with reference to the accompanying drawings in which: - Figures 1 and 2 define the polarization parameters of the light p and s with respect to the propagation vector k and the orientation parameters in incidence and azimuth θ and φ of the rays in the optical system;

- Figure 3 shows the sample relative to the objective of the microscope;

- Figure 4 is a diagram of the polarizing microscope used according to the invention;

- Figure 5 is a schematic representation of the device direct thickness measurement according to the invention

- Figures 6A and 6B are a schematic representation of the display device of a multisensor implemented in certain embodiments of the invention. The description of the invention will be made using the notations of FIGS. 1 and 2, where p is the polarization vector of the light carried by a radius of angle of incidence θ on the sample.

Furthermore, the term “sample 1” denotes the assembly acting on the measurement. This sample is separated from objective 2 by an incident medium 3, it comprises, in order starting from the incident medium, a study object 4 (the one that we seek to visualize), a stack 5 of layers whose upper layer 6 is the layer in contact with the sample, and a support 7. The stack of layers and the support form the substrate 8. FIGS. 4A and 4B are representations of devices usable according to the invention; similar elements are represented there with the same reference numerals.

A sample 1 assumed to be plane and isotropic is therefore placed under an optical microscope operating in reflection. The microscope is provided with an objective 10 and a Kôhler type lighting, comprising at least two lenses 12 and 13 and an aperture diaphragm or pupil 11 conjugated by the lens 13 of the rear focal plane of the objective 10, represented by a dotted line in Figure 4A. The polarizer P polarizes the light directed towards the sample by the semi-reflecting lamina 15. The direction of the polarizer P serves as a reference. The light returned by the object is subjected to an analyzer A.

FIG. 4B corresponds to the implementation of a differential interference contrast microscope (DIC), it comprises a polarizing element 16 which is either an ollaston biprism or a prism and a Nomarski compensator.

As is known elsewhere, it is also possible to replace the linear polarizations with circular polarizations.

So, instead of crossed polarizers and analyzers, we will find the semi-transparent mirror, a first polarizer, a quarter wave plate (λ / 4), the objective, the sample and then back again the objective, the λ / 4 plate, the polarizer mentioned above and the semi-transparent mirror.

In the case of the differential interference contrast microscope, we will then find the semi-transparent mirror, a polarizer, the polarizing element, a λ / 4 plate, the objective, the sample and then back to level the objective, the λ / 4 plate, the polarizing element, the polarizer mentioned above and the semi-reflecting mirror.

The angle of incidence of a ray is θ. The microscope is equipped with a linear polarizer and an analyzer located on either side of the sample on the light path. The lighting is episcopic and monochromatic. The analyzer is rotating and makes an angle φ with the polarizer. We measure the normalized reflected flux Φ _N as the ratio of the reflected flux and a reference flux. The reference flux is that which would be obtained on the same instrument adjusted in the same way in the absence of a polarizer and an analyzer with a hypothetical perfectly reflecting sample. The perfectly reflecting sample is defined by its complex Fresnel coefficients for the parallel (p) and perpendicular (s) polarizations as r _p = r _s = 1. For any angle::

cos2φ

Φ _N (θ, φ) = cos ² φ (r _p ² + | r _s | ² ) - r _p + r _s El

In the particular case where the polarizer and the analyzer are crossed (φ = π / 2), this formula is reduced to:

The second member of the formula (El) is directly interpretable. It is made up of two terms: The first, cos φ (| r _p | ² + | r _s | ² ), is the product of an extinction coefficient and a reflection coefficient in intensity which we will call a reflectivity . This reflectivity can be described as "inconsistent average reflectivity" because it would be obtained by ignoring the interferences between r _p and r _s , i.e. between the reflected parallel and perpendicular components, and by averaging over all possible azimuths φ, i.e. over all possible orientations of the plane of incidence relative to the direction of the polarizer. Reduced to its first term, equation (El) would therefore give the reflection obtained by inverting the order of the sample and the analyzer on the light path, the surface playing here only the role of an element. absorbent. This first term disappears completely when φ = π / 2: between crossed polarizer and analyzer, and in the absence of (de) polarizing elements, nothing happens.

The second term of equation (El) describes the interference between r _p and r _s . We call it "coherent reflectivity". It translates the depolarization of the incident beam by the surface, which transforms the linear incident polarization into an elliptical polarization. This ellipticity is different for each azimuth, that is to say for each plane of incidence defined by its angle φ with the direction of the polarizer, and this second term describes the average reflectivity which results therefrom for the conical geometry of the lighting. It disappears for φ = π / 4, where the contributions of all the azimuths compensate each other, and also for r _p = -r _s . It decreases the total reflectivity between parallel polarizer and analyzer and increases it when they are crossed.

The visualization technique, object of the present invention, directly exploits this second term. We choose φ = π / 2, and the second term of the equation (El) remains the only present. The extinction or, in a more elaborate version, the quasi extinction of the incoherent reflectivity is one of the foundations of the invention. What we call "coherent reflectivity" can also be called "ellipsometric reflectivity" since it results from the ellipticities (functions of the azimuth φ) of the reflected polarization. An expression equivalent to (El) is:

cos2φ

Φ _N (θ, φ) = - (r _p ² + | r _s | ² ) + r - E3

2

This expression makes it possible to compare the signal obtained in presence of polarizing elements to the signal obtained in the absence of polarizing elements, that is to say in non-polarized light, which is given by the first term alone. It will be noted:

Φ _N (θ ₅ NP) = ^ (| r _p | ² + | r _s | ² ) E4

In the presence of the polarizers, we still have experimental access to this quantity by imposing φ = π / 4, as shown in equation E3. For the visualization of the edge of a study object 4 having the form of a thin film placed on the surface, we use the intensities collected by observing the film and the bare surface which are denoted I _F and I _s . They are proportional to the corresponding standardized flows.

The contrast of the edge of the film is:

To properly visualize the film, you must maximize C and therefore make the ratio I _F / I _S maximum (I _s - 0, to tend towards a contrast of 1) or minimum (I _F - 0, to tend towards a contrast of - 1). So either turn off the surface or the film. Thus, a sensitive method is based on the one hand on a good extinction, and on the other hand on a selective extinction.

Our technique combines two extinction factors: i) the crossed or almost crossed polarizer and analyzer, ii) an anti-reflective substrate for this observation mode.

Equation (E3) highlights the dual nature of our extinction: the crossed polarizer and analyzer extinguish the first term of the second member, our antireflective substrate extinguishes the second. It can therefore be defined as an antireflective substrate for coherent reflectivity. This is the second foundation of our visualization technique. But a good extinction is not enough for a sensitive visualization. You must turn off I _F or I _s but not both at the same time. As the film we are viewing is very thin, and therefore its physical parameters do not disturb those of the bare surface, this means that the extinction must be critical. In other words, the extinction must be lost for a very small modification of the surface. This critical nature of the anti-reflective quality of our substrate is the third foundation of our visualization technique. The performance of a visualization method can be quantified by the contrast obtained when the film observed becomes extremely thin. In this case, I _F and I _s become neighbors and dl = I _F - I _{s is} similar to a differential element.

C can then be written:

where Δe is the thickness of the film which can be assumed for the circumstance of optical index identical to that of the upper layer, and where dl / de is the derivative of the intensity reflected by the bare substrate with respect to the thickness e of the last layer of the stack. In the case where the substrate is composed of a solid support covered with a single dielectric layer, e is therefore the thickness of this layer. The fact of taking an identical index for the film and for the last dielectric layer is not compulsory, but it simplifies the explanation and shows that our method does not exploit the reflection between the film and the substrate. The film is therefore considered here as a simple fluctuation in thickness of the upper layer.

We define the sensitivity of our technique in Angstroms ^"1 as the ratio of C to Δe:

The expression of r _p and r _s for a solid covered with a single layer is conventional (ref. AZZAM for example):

^{(k) "} T ¹ + ⁺ r ^{r ~} 01 (k) ^x r ^r 12 (k) ^e ^

with k = either s or p, depending on the polarization considered and

N e with β _L = 2π - - cosθ _{l 5} the index 1 referring to the layer, λ the index 2 to the support and the index 0 to the incident medium. This equation allows us to write:

σ

where σy and Ily represent respectively the sum and the product of r ^ ₎ and r _{ij (s)} . The sum σ is a periodic function of the optical thickness

Nie of period λ / 2. Its module | σ | generally presents two minima and two maxima per period. It is the same for Ln | σ |. The function | σ | being moreover a bounded function, it remains very regular and its derivative with respect to e is never very important. On the other hand, the function Ln | σ | diverges when | σ | tends towards zero and the sensitivity given by equation E7 becomes very important in absolute value on either side of the minimum when the extinction becomes total. The contrast is always negative to the left of a minimum and positive to the right. This is why we also designate the condition for obtaining a minimum by "contrast inversion condition".

In summary, the rapid contrast inversions therefore correspond to the minima of | r _p + r _s | with respect to e, and the very rapid inversions of contrast are obtained when this minimum of | r _p + r _s | tends towards 0. Equation (E3) highlights the advantage of using a fluorescence microscope: in the presence of a fluorescence signal, the depolarized component of this fluorescence is added to the right member of equation (E3) without altering the extinction of the other two terms. The signal to noise ratio is therefore increased. This can also be transposed to a Raman signal. The invention also relates to an ellipsometric measurement method which can also operate without the need for a particular substrate being necessary:

The ellipsometric angles ψ and Δ are defined by:

i- = tanΨe ^jΔ E10 ^r s

Two equations chosen arbitrarily from the following four are sufficient to establish the correspondences which will be useful between reflectivities and ellipsometric parameters:

[r _P | ² + | r _s | ² = | r _s | ² (l + tan »El 1

| r _p + r _s | ² = | r _s | ² (l + tan ² ψ + 2 tan ψ cos Δ) El 2

| r _p - r _s | ² = | r _s | ² (l + tan ² ψ - 2 tan ψ cos Δ) El 3

IL prs ^* + rp ^* r „S = 2 | | rs I | ² tan Ψ cos Δ E 14

The first of these equations shows that the ellipsometric parameter ψ is accessible by measuring the incoherent reflectivity. Each of the other three shows that the determination of the second ellipsometric parameter, Δ, also requires the measurement of the coherent reflectivity (or of a combination of the two reflectivities). It is therefore by accessing the coherent reflectivity signal that we can determine ψ and Δ.

The measurement is carried out in two stages: i) The first is based on the rotation of the analyzer. The image of the sample is analyzed by a CCD camera or any other two-dimensional detector. Equation E3 shows that the reflected signal oscillates sinusoidally around the incoherent reflectivity with an amplitude | r _p - r _s | ² and a period π on the angle φ. Different procedures, requiring at least two measurements, make it possible to obtain two combinations of the three parameters | r _s | ² , tanψ, and cosΔ, for example | r _s | ² (l + tan ² ψ) and 2 | r _s | ² tanψcosΔ. This already makes it possible to determine the combination sin2ψcosΔ of the only ellipsometric parameters, but is not sufficient to determine Δ and ψ separately. ii) The second step requires breaking the radial symmetry of the lighting, which can be done in two ways: either by physically modifying the geometry of the aperture diaphragm, which must become a slit or a cross formed by two perpendicular slits, or an angular sector δφ (modulo π) of opening strictly less than π / 4, the apex of which coincides with the optical axis or the association of two or four identical angular sectors regularly arranged around the axis optics of the microscope, capable, like the analyzer, of rotating around the optical axis of the microscope, either by analyzing the intensity distribution present in a conjugate plane of the aperture diaphragm situated on the path of the reflected light, the microscope being in Koehler lighting. The microscope being equipped with a CCD camera to receive the image of the sample, this analysis can be carried out very simply by interposing a Bertrand lens between the objective and the pupil of the camera. It is therefore a conoscopic measurement. The advantage of this solution, which is easy to implement, is that the angle of incidence θ and the azimuth φ are, in the conjugate plane, two geographically separable parameters and that we can therefore access all of the function Φ _N (θ, φ, λ), λ designating the wavelength of the lighting. We can thus adjust the range of opening angles kept, explore the azimuth, or filter the lighting by digital means. This solution also makes it possible, in the absence of the Bertrand lens, to carry out the first step of the analysis simultaneously on several regions of a heterogeneous sample, and therefore to determine by a parallel measurement the quantity (sin2ψ cosΔ) ( x, y). For a complete analysis with the Bertrand lens, it is however necessary to select a homogeneous region of the sample by the use of a field diaphragm or a confocal geometry. This solution therefore does not allow the complete parallel analysis of the different points of the sample. The first solution on the other hand (diaphragm with rupture of radial symmetry), allows the total parallel analysis since one always keeps the image of the sample on the CCD camera.

The reflected intensity I when a very small angular sector sélection selects a particular azimuth φ on the lighting cone, provided by the equivalent of equations El to E4 is now:

Generally, this intensity is a periodic function of φ of period π and also includes terms of period π / 2.

If the relative orientation of the analyzer and the polarizer is fixed and: - if the slit has a uniform rotational movement around the optical axis at the frequency ω, the intensity reflected by each point of the sample is modulated and this modulation allows different combinations of quantities to be extracted | r _s | , ψ and Δ sought. For this, several techniques can be implemented, in particular photometric type techniques using time intensity averages and extremum amplitudes or synchronous detection techniques making it possible to compare the amplitudes and phases of the components of l 'intensity reflected at 2ω and 4ω; - if the orientation of the slit is manually adjustable, we can measure the intensities collected for several orientations of the slit, at least two, and deduce from the general formula above the values of different combinations of the parameters | r _s | , ψ and Δ, this which allows the values of these parameters to be determined completely;

- if the analyzer is driven by a uniform rotational movement around the optical axis, the signal I is modulated with a period π (on φ) and the measurement of I for different values of φ becomes more precise;

- if finally the analyzer and the slit are both animated with a uniform rotation, at different frequencies, the function I (φ, φ) can be fully recognized and the parameters | r _s | , ψ and Δ can be determined with great precision by a conventional numerical adjustment procedure with three parameters.

In the simplest particular case where the angle φ is fixed and where the measurement of I is carried out for the two orientations φ = 0 (modulo π) and φ = π / 2 (modulo π) of the slit, we obtain respectively :

Φ _N (θ, φ, φ = 0) = l / 2 | r _p | ² cos ² φ and Φ _N (θ, φ, cp = π / 2) = l / 2 | r _s | cos ² φ

It therefore suffices to take the root of the ratio of these two intensities to obtain the quantity tan ψ. This measurement combined with the previous two therefore makes it possible to completely determine | r _s | ² , ψ, and Δ, and therefore also | r _p | ² .

It should be noted that the determination of the only parameters ψ and Δ can be obtained by using only measured intensity ratios, and therefore does not require the use of a reference substrate.

An interesting particular case is that where Φ = 0 which corresponds to a parallel polarizer and analyzer and which can therefore be reduced to that of a single polarizer placed between the lighting mirror and the objective or even between the objective and the sample. The reflected intensity is written in this case:

I = A? ⁱ M cos ⁴ φ + | r _s | sin ⁴ φ tanψcosΔ j This shows that with a rotating diaphragm made up of a slit, a cross made up of two perpendicular slits crossing on their axis, of an angular sector whose apex is posed on the optical axis of the microscope and of amplitude azimuth less than 45 degrees, or the association of two or four angular sectors of the same type regularly arranged around the optical axis, which diaphragm is possibly superimposed on a ring to delimit a single angle of incidence, it suffices to perform three reflected intensity measurements with three different non-redundant orientations of the diaphragm to deduce all of the ellipsometric parameters of the sample. For example, in the case where the diaphragm is a slit or a very small angular sector identified by its orientation φ, the intensity reflected at each point of the image of the sample becomes:

∑ _! ≡ I (φ = 0) = A ² r _p = A ² | r _s | ² (l + tan ² ψ) for φ ≈O

1 ₂ ^≡ I (= ^) =? | r _s | ² for φ = π / 2

1 ₃ ≡ I (φ = -) = 2 A, | r _s | tan ψ cos Δ for φ = π / 4

So it suffices to calculate the ratio - to deduce tan ψ,

then the report deduce cosΔ.

This example illustrates:

- how the measurement of three intensities with three different orientations of the slit makes it possible to determine the totality of the ellipsometric parameters by using only intensity ratios, therefore without additional calibration;

- how a modulation of the reflected intensity which includes these three measurements but also others makes it possible to obtain the same information with increased precision; - how you can make an ellipsometer under an optical microscope or under a binocular magnifier using a single polarizer and a rotating slot.

In the presentation of the measurement method, we used equations valid for a single angle of incidence θ. Like | r _s | ² , ψ, and Δ depend on θ, we can: either access these quantities for a single angle θ using an annular aperture diaphragm, or access quantities averaged over a range of angles of incidence [θ _m i _n , θ _max ], with most often 0. The same formulas apply to effective quantities, assigned below an “eff” index defined from means on θ. It is therefore advisable to ask:

ef

s leff = <the

cosΔ _eff

to write the reflected intensity as

The measurement of I therefore makes it possible to determine the effective quantities and, in particular, the ellipsometric angles ψ _eff and Δ _eff , which can be compared with values calculated to deduce therefrom. properties of the object or sample, as is conventionally practiced with ellipsometric angles with single incidence.

In fact, the interest of the method is above all to be able to perform ellipsometric measurements under a microscope in order to combine ellipsometric measurement and imaging. We must therefore consider that the natural geometry of the lighting is a cone of light around the normal. However, the ellipsometric parameters vary little for low incidences. This is also why ellipsometry is only a sensitive technique at high angles of incidence. The counterpart in our process is that the average which is carried out on the cone of lighting scrambles little the exploitable signal. In the absence of an optimized substrate, the drawback of ellipsometric measurement under a microscope is that its sensitivity is poor. But in the presence of the optimized substrates as we propose them, the sensitivity of the measurement to the physical parameters of the sample becomes excellent again, comparable in fact to that of traditional measurements around the Brewster angle, while the signal used remains little sensitive to the angle of incidence. This is due to the fact that between crossed polarizer and analyzer, the extinction of the coherent reflectivity is always total in normal incidence, so that only the non-zero incidences participate in the construction of the signal which we exploit. With the conditions of a good extinction for non-zero incidences, the extinction is good on all the incidences of the lighting cone. It is possible to optimize the thickness of the last layer for any material.

Indeed, we have seen that the function | σ (e) | = | r _p + r _s | always has more or less pronounced minima, which correspond to the contrast inversion conditions. The contrast is therefore zero for these particular values of e. Being also periodic and continuous, it reaches a minimum on the left of these values and a maximum on the right. It is therefore always possible to choose the thickness e so that one of these extrema is reached. Whatever the nature of the substrate, it is therefore possible to optimize the thickness of the dielectric layer by calculating jσ (e) |. It becomes particularly interesting when you approach critical conditions.

The critical compositions of the substrates are defined by the existence of a solution to the equation | σ (e) | = 0. A critical substrate has a layer thickness close to a solution of this equation. A solution of the equation | σ (e) | = 0 necessarily corresponds to a minimum of | σ (e) |. It is therefore a thickness of inversion of the contrast. The smallest of these values, e _c , of course plays a special role. The other thicknesses of contrast inversion are then given by e _c , k = e _c + KNιλ / 2. From E9, the equation we are discussing is:

σ ₀ ι + σ ₁₂ (l + π ₀ ι) e- ^2jβl + σ ₀ ιll ₁₂ e- ^4jβl = 0

E15

In the case of a stack reduced to a single layer, the values e _c of e are the solutions of the equation El 5 taken from E9:

σ ₀ ι + σι ₂ (l + π ₀ ι) z + σ ₀ ιπ ₁₂ z ² = 0 E16

It has two complex solutions zl and z2 which are functions of the indices of the incident medium, of the layer, of the complex index of the substrate, and of the angle of incidence Θ0 (or in an equivalent manner of the angle refracted in the layer, θl). Critical conditions are reached when the modulus of one of these two solutions is equal to 1. This problem is fairly simple to explore numerically. Analytically, it is possible to develop each of the terms up to order 4 in θl because each depends only slightly on the angle close to the normal. We can thus find solutions "by hand". In practice, the two extreme media are often imposed and it is the index and the thickness of the layer that must be determined. We therefore draw the contrast as a function of the thickness for some arbitrary indices and we note a monotonic variation of the contrast. So you just have to move in the direction where things are improving until they start to get worse. From there, we continue the exploration around the best value in refining changes in the index. We can also use the numerous results published in the literature for a single layer in terms of ψ and Δ. The situations sought correspond simultaneously to:

tanψ = 1 and Δ = π E17

The solutions that we have found numerically are fairly well approximated by the empirical formula:

Particularly interesting results are obtained by producing a silicon substrate covered with a single layer meeting the following parameters, the lighting being preferred monochromatic of wavelength λ = 540 nm and the opening angle of the lighting cone being assumed 30 degrees:

where N ₀ is the index of the ambient medium, N ₃ is the index of the substrate support, N ₂ is the index of the layer and e its thickness.

The optimal thickness e is a linear function of λ but is not not proportional to λ. For observations in air, = 0.2. δλ

The layers of indices 1.74 and 1.945 can be produced by numerous methods, such as PECVD deposits. The layers of index 1.345 are more difficult to produce. They can be formed of a hydrogel, an airgel, a polymer, or be heterogeneous, for example formed of studs of constant thickness and very small dimensions. It can also be a solution in water, sugar, salt, polymer ...

A particularly advantageous visualization method of the optical thickness (Ni x ej) of a very thin film can be carried out with the substrate used in the invention in the following manner shown in FIG. 5.

A layer 21 of variable thickness in the form of a bevel is deposited on a support 20 (FIG. 5A, FIG. 5B).

This substrate 20 is then cut so as to obtain two identical elements 22, 23 (23 not shown is then identical to 22) (FIG. 5C).

One of these elements is then covered with the thin film 24 to be studied (FIG. 5C).

Then these two elements are then observed under a microscope lit in white light with a disc-shaped pupil, after having positioned these two elements relative to each other, in their initial relative position, using a mark, notch or wedge 25.

Fringes 26, 27 are then observed in white light, respectively on each of these elements and their relative offset M allows the properties of the layer deposited on one of these elements to be measured.

The invention is particularly suitable for viewing elements contained in multisensors. A chemical or biological multisensor (= biochip) consists of a support 30 on which are deposited pellets 31 (= spots) each formed from a different layer capable of selectively fixing each a different species to be recognized within a mixture liquid (biochip) or gaseous (artificial nose), and forming between them a matrix of ordered elements along the surface. Each pellet has a surface of a few square microns and often a molecular thickness.

The multisensor is used in the following way: it is brought into contact with the mixture that we want to analyze. Each patch 31, 32 ... fixes the species which it can recognize when it is present in The mixture. In situ or after rinsing, we then look at which pads 32 have loaded and which pads 31 have remained empty to know the composition of the mixture. A fixed species creates an additional thickness, visible at the level of the stud, the position of the studs 31, 32 in the matrix informing us of the nature of the recognized species. This step is the step of reading the multisensor.

Our microscopy method is sensitive enough to tell the difference between an empty pad and the same charged pad, in many types of multisensors. It therefore provides a simple, direct and parallel reading method for all multisensors.

In a preferred example, we describe its implementation on a particular type of multisensor: biochips. They include, for example DNA chips, antibody chips, bacteria chips, virus chips, chromosome chips, protein chips, etc.

In the example of DNA chips, each pad consists of a molecular layer of identical oligonucleotides capable of hybridizing with and only with their complementary strand. The analyzed DNA is cut into strands of suitable length, amplified by the PCR technique, which means that each strand is replicated a large number of times, then put in solution in contact with the chip. The recognized strands are fixed by the corresponding pads.

Our method makes it possible to recognize loaded tablets. To this end, the pellets whose thickness is regular and known, are taken as elements of the multilayer building, so that the support + multilayer + patch or spot assembly constitutes an optimized substrate of very high sensitivity. Under these conditions, the presence of additional strands after hybridization is easily detected by the change in intensity or color that it causes in the observation of the pellet by our visualization process. The quantity of material present on a pellet can also be quantitatively evaluated by our measurement process.

Claims

1. Device for two-dimensional ellipsometric display of a sample, comprising an object, placed in an incident medium, observed between an analyzer and a polarizer crossed by reflection in convergent light, in which the ellipsometric parameters of the assembly formed by the object (4) and a substrate (8) on which it is placed, are used, characterized in that:

- the substrate comprises a support and a stack of layers and that its ellipsometric properties are known, 0 - the ellipsometric properties of the substrate being such that the variations of the ellipsometric parameters of the sample due to the object are displayed with a contrast greater than the contrast produced in the absence of this substrate.

2. Device for two-dimensional display of a sample 5 according to claim 1, characterized in that the sample is illuminated through a wide aperture objective such as a microscope objective.

3. Device for two-dimensional display of a sample according to claim 2, characterized in that the microscope is a 0 microscope with differential interference contrast.

4. Device for two-dimensional display of a sample according to claim 2, characterized in that the microscope is a fluorescence microscope.

5. device for two-dimensional display of a sample 5 according to any one of claims 1 to 4, characterized in that the thickness e of the layer of the stack in contact with the object is such that the reflection coefficients complexes r _p and r _s of the substrate satisfy the condition d ² / of ² [Ln [r _p + r _s |] = 0.

6. Device for two-dimensional display of a sample 0 according to claim 5, characterized in that the optical properties of the substrate are such that the minimum value taken by the quantity | r _p + r _s | over all the values of e is as low as possible.

7. Two-dimensional display device of a sample according to one of claims 1 to 6, characterized in that it comprises a polychromatic light source.

8. Device for two-dimensional display of a sample according to one of claims 1 to 6, characterized in that it comprises a monochromatic light source.

9. Device for two-dimensional display of a sample according to one of claims 1 to 8, characterized in that the support is made of silicon.

10. Device for two-dimensional display of a sample according to one of claims 1 to 9, characterized in that the stack consists of a single layer. o

11. Two-dimensional display device of a sample according to claim 10, characterized in that the layer is a layer of silica.

12. Two-dimensional display device for a sample according to claim 10, characterized in that the thickness 5 of the silica layer is of the order of 1025 Å, the incident medium being air.

13. Device for two-dimensional display of a sample according to claim 10, characterized in that the layer is a layer of magnesium fluoride. 0

14. A device for two-dimensional display of a sample according to claim 10, characterized in that the layer is a layer of polymer.

15. Device for two-dimensional display of a sample according to claim 10, characterized in that the layer 5 is a polymer layer, with an optical index approximately equal to 1.343, the incident medium being air.

16. A device for two-dimensional display of a sample according to claim 10, characterized in that the layer is a mineral layer, with an optical index approximately equal to 0.74, the incident medium being water.

17. Device for two-dimensional display of a sample according to claim 10, characterized in that the layer is a mineral layer, with an optical index approximately equal to 1.945, the incident medium being an oil with an optical index 1.5.

18. Device for two-dimensional display of a sample according to claim 10, characterized in that. the layer is discontinuous and formed of silica studs and of index 1.343, of the same height defining the thickness of the layer and of cross-sectional dimensions much less than a micrometer, the incident medium being air.

19. Two-dimensional display device of a sample according to claim 10, characterized in that the layer is an organic or mineral mesoporous or nanoporous layer with an index approximately equal to 1.343, the incident medium being air.

20. Two-dimensional display device for a sample according to claim 10, characterized in that the layer is a mineral airgel with an index approximately equal to 1.343, the incident medium being air.

21. A device for two-dimensional viewing of a sample according to one of claims 2 to 20, characterized in that it comprises a microscope comprising an aperture diaphragm in the form of a longitudinal slot orientable around the axis of the microscope making it possible to restrict the light cone to a single plane of incidence in a chosen direction.

22. Device for two-dimensional display of a sample according to one of claims 1 to 19, characterized in that it comprises a microscope comprising an aperture diaphragm in the form of a ring limiting the lighting cone of the sample around an angle of incidence.

23. A device for two-dimensional display of a sample according to claim 1, characterized in that the object is a thin film and the stack comprises a bevelled layer whose thickness varies monotonously in a direction X along the surface.

24. Measuring method using the device of claim 23, characterized in that:

- the display device is cut parallel to the direction X into two elements, - the thin film is deposited on one of these elements, - the two elements are placed between a crossed polarizer and an analyzer under a polarizing microscope lit in polychromatic light, so as to form fringes of colored interference on each of the elements, - the offset of the fringes formed is measured respectively in each of the elements to deduce the properties of the layer deposited on one of them.

25. A device for two-dimensional display of a sample according to any one of claims 1 to 23, characterized in that the substrate is the bottom of a Petri dish.

26. Two-dimensional display device _. of a sample according to any one of claims 1 to 23, characterized in that the sample is a matrix multisensor.

27. Device for parallel reading of a matrix multi-sensor according to claim 26, characterized in that each pad or patch of the matrix constitutes the last layer of the stack.

28. Device for parallel reading of a matrix multisensor according to claim 26, characterized in that the multisensor is a biochip with bacteria, viruses, antigens, antibodies, proteins, DNA, RNA, chromosomes.

29. Process for ellipsometric measurement of a sample with spatial resolution under a polarizing microscope forming an image of the sample, in which:

a modulation of the reflected intensity is ensured by the relative rotation of the polarization of the light beam and of the polarizer-analyzer, characterized in that:

- the aperture diaphragm of the lighting beam is a ring centered on the axis of the beam delimiting a single angle of incidence, - the average flux φ _M (x ₅ y) reflected and its modulation amplitude φ _m (x, y) are measured simultaneously at each point of the image obtained from the sample,

- we treat the measurements φ _M (x, y) and φ _m (x, y) to deduce simultaneously at each point of the sample two combinations of the ellipsometric parameters ψ (x, y) and Δ (x, y ) and the reflection coefficient | r _s | ² (x, y) from the formulas:

1 1 | ι

- | r _s | ² (l + tan ² ψ) = φ _M and - | r _s | ² (tan ² ψ - 2tanψcosΔ) = φ _m

- we treat the measures φi _v r (x ₅ y) and φ _m (x, y) to deduce the combination sin (2ψ) cos of the only ellipsometric parameters ψ (x, y) and Δ (x, y) from the formula:

φ _m = φ _M (l - sin (2ψ) cosΔ)

30. A method of ellipsometric measurement of a sample with spatial resolution under a polarizing microscope forming an image of the sample, in which: - the sample is illuminated by a beam of light linearly polarized through an aperture diaphragm ,

- the light reflected by the sample is analyzed by a polarizer-analyzer, characterized by the relative orientation φ of its direction of polarization with respect to that of the polarizer, - modulation of the reflected intensity is ensured by the relative rotation of the polarization of the light beam and of the polarizer-analyzer, characterized in that, in a measurement step:

the aperture diaphragm of the lighting beam is a slit orientable around the optical axis of the microscope superimposed on a ring delimiting a single angle of incidence, the intensity of the reflected beam is measured for at least two different non non-redundant orientations of the slit,

- we treat these intensity measurements from the relation:

31. A method of ellipsometric measurement of a sample with spatial resolution under a polarizing microscope according to claim 30, characterized in that, in a complementary step:

- the analyzer is fixed in an orientation not perpendicular to the polarizer, for example φ = 0, - the aperture diaphragm of the lighting beam is a slit orientable around the optical axis of the microscope superimposed on a ring delimiting a single angle of incidence,

- we measure the intensity reflected for the two orientations φ = 0 and φ = π / 2 of the slit, - we process these intensity measurements to obtain tanψ by taking the square root of their ratio according to the three formulas:

I = A ² cos φ for φ = 0 modulo π

_τ A 21 \ • 2 ι π

I = A, r _s on φ for φ = - modulo π

tanψ =

32. Process for ellipsometric measurement of a sample with spatial resolution under a polarizing microscope forming an image of the sample, in which:

a modulation of the reflected intensity is ensured by the relative rotation of the polarization of the light beam and of the polarizer-analyzer, characterized in that, in a measurement step:

the orientation of the analyzer relative to the polarizer is fixed at a value different from π / 2 modulo π, - modulation of the reflected intensity is ensured by the rotation of the diaphragm D around the optical axis,

- the average flux φ _M (x, y) reflected and its modulation amplitude φ _m (x, y) ₅ are simultaneously measured at each point in the sample

- we treat the measures φ _M (x, y) and φ _m (x, y) to deduce the two ellipsometric angles ψ (x, y) and Δ (x, y) and the modules | r _p | and | r _s | reflection coefficients from the relationship:

I

33. A method of ellipsometric measurement of a sample with spatial resolution under a polarizing microscope according to claim 31, characterized in that, in a complementary step:

the orientation of the analyzer with respect to the polarizer is fixed at φ = 0, a modulation of the reflected intensity is ensured by the rotation of the diaphragm D around the optical axis,

- the average flux φ _M (x, y) reflected and its modulation amplitude φ _m (x, y) ₅ are simultaneously measured at each point in the sample - we treat the measures φ] y _t (x ₃ y) and φ _m (x, y) to deduce the two ellipsometric angles ψ (x, y) and Δ (x, y) and the modules | r _p | and | r _B | reflection coefficients from the relationship:

34. Measuring method according to any one of claims 30 to 32, characterized in that the measurement of claim 29 is associated therewith. 0 35. Process for ellipsometric measurement of a sample with spatial resolution under a polarizing microscope forming an image of the sample, in which:

- the sample is illuminated by a linearly polarized light beam through an aperture diaphragm, 5 - the light reflected by the sample is analyzed by a polarizer-analyzer, characterized by the relative orientation φ of its direction of polarization with respect to that of the polarizer,

a modulation of the reflected intensity is ensured by the relative rotation of the polarization of the light beam and of the 0 polarizer-analyzer, characterized in that:

- the average flux φM (x, y) reflected and its modulation amplitude φ _m (x, y) are measured simultaneously at each point of the image 5 obtained from the sample;

- we treat the measurements φM (x, y) and φ _m (x, y) to deduce simultaneously at each point of the sample two combinations of the effective ellipsometric parameters ψ _e ff (x, y) and Δ _efï ( x, y) and o coefficient of effective reflection | r _is _ff. ² (xy) from the formulas:

- | r _s | ² (l + tan ² ψ _efr ) = φ _M and - | r _s | ² (tan ² ψ _eff - 2tanψcosΔ _eff ) = φ _m - we treat the measurements ^ _M (x, y) and φ _m (x, y) to deduce the combination sin (2ψ) cos from the only effective ellipsometric parameters ψ _eff (x ₃ y) and Δ _eff (x, y) from the formula:

sin2ψ _eff cosΔ _eff = 1 - Φm

Φ M

36. Device for ellipsometric measurement under a microscope with lateral spatial resolution, characterized in that:

applied to these three measurements, in which the parameters r _s , ψ and Δ are possibly effective parameters obtained from means over all the angles of incidence present:

e

leff A

tanψ _eff = ^• leff s leff cosΔ _eff

37. Ellipsometric measuring device under a microscope, characterized in that: 5 - the polarizer and the analyzer have a fixed relative orientation,

- the aperture diaphragm is a hole or a ring,

- a CCD camera is placed in this plane, 0 - the measurement of the intensity obtained at each point of the camera

CCD is operated using the general formula:

5 in order to directly obtain all of the ellipsometric parameters of the sample.

38. Device for ellipsometric measurement under a microscope according to claim 37, characterized in that:

- the aperture diaphragm is a hole or a ring,

- a CCD or possibly tri-CCD camera is placed in this plane,

- the measurement of the intensity obtained at each point of the CCD camera or possibly at each point and for each color component of the tri-CCD camera is used using the general formula: 0 -l-

in order to directly obtain all the ellipsometric parameters of the sample.

39. Device for ellipsometric measurement under a microscope according to claim 36, characterized in that:

CCD is operated using the general formula:

in order to directly obtain all the ellipsometric parameters of the sample.

40. Device according to any one of claims 37 to 39, characterized in that the camera is a tri-CCD color camera and that the measurement of the intensity at each point is made and used for each of the colors.

41. A method of ellipsometric measurement of a sample with spatial resolution under a polarizing microscope according to claims 29 to 35, the object studied being placed on a substrate comprising an object, characterized in that the thickness e of the layer of the stack in contact with the object is such that the complex reflection coefficients r _p and r _s of the substrate satisfy the condition d ² / of ² [Ln | r _p + r _s |] = 0.

42. A method of ellipsometric measurement of a sample with spatial resolution under a polarizing microscope according to any one of claims 29 to 35, characterized in that the object is placed on a substrate whose optical properties are such that the value minimum taken by the quantity | r _p + r _s | over all the values of e is as low as possible.