Field of the invention

The present invention relates to a high density multichannel confocal sensor, for use with confocal, chromatic confocal, or interferometry detection schemes. It also relates to a method for inspecting a sample or a substrate implementing said sensor. The field of the invention is, but not limited to, 2D- 3D inspection and metrology systems.

Background of the invention

Several techniques are known for bidimensional (2D) or tridimensional (3D) surface mapping and thickness measurements, for semiconductor or other industrial applications.

Among these techniques, several of them use a confocal detection scheme. Such confocal detection scheme comprises a light emitting aperture and a light collection aperture on one side of a focusing lens arrangement whose optical conjugates (or images) through that focusing lens arrangement are superposed. Such arrangement ensures that the light collected by the collection aperture corresponds essentially to the light issued from the emitting aperture which is reflected back by the sample at the focusing or conjugate point. By implementing such a detection scheme, lateral cross-talk between neighboring channels and collection of out-of-focus light may be avoided, for instance.

A confocal detection scheme may be used for instance for implementing an interferometric detection. The reflected light is spectrally analyzed on a spectral or color detector, and thicknesses or distances between layers may be retrieved by analyzing the resulting interference spectrum.

A confocal detection scheme may also be used for instance for implementing a confocal detection. In that case, the sensor, which preferably implements a high-NA focusing lens arrangement, is moved in height over the sample surface. The distance for which a maximum of intensity is detected on an intensity detector is representative of the local height of the surface. A confocal detection scheme may also be used for instance for implementing a chromatic confocal detection . The techniq ue relies on the use of a focusing lens arrangement with a chromatic element or lens with an enhanced chromatism, whose focal length depends strong ly on the optical wavelength . Each wavelength of the light crossing such lens is focused at a d ifferent d istance, or in a d ifferent focal plane. As previously explained, the chromatic lens is embedded in a confocal set-up with emitting or source apertures and col lection or detection apertures placed at confocal planes of the chromatic lens, so as to reject out-of-focus light. When a reflecting interface is placed in front of the chromatic lens, only the light with the wavelength whose focal plane corresponds to the position of the interface is transmitted by the detection aperture.

In order to obtain a 3D height information, detection is made by a spectral sensor or spectrometer, which comprises usual ly a d ispersing element and a sensor (CCD or CMOS) to acq uire the intensity spectru m of the l ig ht. The height (or d istance) of the interface relative to the chromatic lens is obtained by analyzing the intensity spectrum of the detected light.

The Chromatic confocal techniq ue al lows also obtaining an in-plane image intensity or g rey-level information (2D information) with an extended depth of focus. In that case, an intensity detector is used , which provides the intensity of the incident lig ht over the whole spectrum, without any spectral selectivity. The height information, which rel ies in the spectral content is lost, but the intensity detector provides an intensity (g rey-level) information correspond ing only to the light reflected on the sample at the wavelength focused on its surface . So an image with an optimal lateral resol ution may be obtained over a depth range given by the spectral d ispersion of the chromatic focusing lens arrangement.

Such set-ups allow measuring d istances on a single point at the time . Inspecting large surfaces may be very time-consuming . Acquisition speed may be improved by provid ing several measurement channels in paral lel .

It is for instance known to use fiber optics bundles such as described in the document US 2012/0019821 , which discloses a multichannel chromatic confocal system . One optical fiber modules d irect the l ig ht from the source towards the chromatic lens, and a second optical fiber module is coupled to the chromatic d ispersion objective so as to spatially filter the object l ig ht, thereby forming a filtered light. It is also known to use fiber couplers which d irect the light from the source towards the chromatic lens, and the reflected l ig ht towards the detectors.

Such configurations have the advantage that the source and detection apertures are the end of a measurement fiber.

However, the use of fiber optics elements has several d rawbacks, notably :

- they are difficult to use or manufacture with a very large number of channels (large fiber bundles) ;

- the optical fiber geometry (core to cladd ing ratio) set by the manufacturing requirements impose a l imit on the ach ievable channel density. The total useful area constituted by the fiber cores is dil uted by the total g lass area of the optical fiber. Th is, for a given surface to il luminate, red uces the lateral resol ution ;

- the manufactu ring of optical fibers and bund les impose constraints that makes difficult the matching of fiber cores pitch (separation between 2 beams) with the detection device pixels pitch ;

- d ue do their principle of operation using coupl ing of modes between fiber cores, the spl itters or couplers made with optical fibers have a coupling ratio which is very dependent with the wavelength, which may introd uce bias in the measurements.

It is an object of the invention to provide a confocal device for implementing detection schemes such as chromatic confocal, confocal and interferometry allowing implementation of a large number of channels.

It is an object of the invention to provide such confocal device allowing smal l pitch between the il luminating l ig ht spots delivered by the multiple channels on the measured substrate .

It is an object of the invention to provide a such confocal device allowing hig h lateral and hig h depth resolution (and/or extended depth range) for a confocal device with a large number of channels.

It is also an object of the invention to provide such confocal device with a large number of channels which red uces the manufacturing constraints. It is also an object of the invention to provide such confocal device that decouple the requirements of the illuminating stage and those of the detection stage with limited alignment elements and steps.

It is also an object of the invention to provide such confocal device with a large number of channels which is mechanically robust.

It is also an object of the invention to provide such confocal device with built-in alignment means to facilitate and automate its assembly.

It is also an object of the invention to provide such confocal device with optimal optical and metrological characteristics.

Summary of the invention

These objects are obtained with a multichannel confocal sensor comprising at least one light source, at least one focusing lens arrangement, and at least one optical detector,

characterized in that it further comprises:

- a first integrated optics circuit arranged for splitting a light beam coming from said at least one light source into a plurality of emitted light beams applied to a high-density array of emitting apertures,

- a second integrated optics circuit arranged for collecting on a plurality of collection apertures a plurality of reflected light beams from a sample to be inspected and for transferring said reflected light beams to said at least one optical detector,

- a beam splitter arranged (i) for directing said emitted light beams from said first integrated optics circuit to said inspected sample through said at least one focusing lens arrangement and (ii) for directing said reflected light beams from said measured sample through said at least one focusing lens arrangement into said second integrated optics circuit.

In an advantageous embodiment of the sensor according to the invention, the first integrated optics circuit comprises arrays of emission channel waveguides, which are achromatic or at least substantially achromatic within the spectral range of the light source, or the used spectral range. The emission channel waveguides may have an end forming the emitting apertures, directly (for instance on an edge of the first integrated optics circuit) or through some optics or lens elements.

The first integrated optics circuit may further comprise Y-junctions splitters, achromatic or at least substantially achromatic within the spectral range of the light source, or the used spectral range.

In another advantageous embodiment of the sensor according to the invention, the second integrated optics circuit comprises arrays of optical detection channel waveguides, achromatic or substantially achromatic within the spectral range of the light source, or the used spectral range.

The optical detection channel waveguides may have a first end forming the collection apertures, directly (for instance on an edge of the first integrated optics circuit) or through some optics or lens elements.

The optical detection channel waveguides and the optical emission channel waveguides may have a same spacing at, respectively, the collection apertures and the emitting apertures side.

In a preferred version of the invention, the optical detection channel waveguides have a larger diameter than those of the optical emission channel waveguides, at least at the collection aperture side.

The optical detection channel waveguides may have a larger angle acceptance at the collection aperture side than those of the optical emission channel waveguides at the emission aperture side.

The optical emission waveguides may notably comprise single mode waveguides, and the detection channel waveguides may comprise multimode waveguides, or waveguides with such behavior at least within the spectral range of the light source, or within the used spectral range.

The detection channel waveguides are preferably designed to match on a second end the aperture and/or pixel pitch of detection elements within an (or said at least one) optical detector, directly or in a conjugate plane (for instance through an imaging and/or spectrally dispersing element).

The detection channel waveguides may be arranged so that said at least one optical detector is out-of-sight of stray light coupled in the second integrated circuit between the detection channel waveguides at the collection apertures. The detection channels may for instance have a curved or an angled shape, so that their first and second ends are not within the same orientation. They may also have a "S" shape so that the first and second ends have the same orientation, but with a lateral deviation.

At least one of first and second integrated optics circuits may comprise planar waveguides on a planar substrate. Or in other words, the first and/or the second integrated optics circuits may be done using a planar waveguide deposition or inscription technique on a substrate.

The first and/or the second integrated circuit may be respectively done by combining and assembling several subpart integrated optics circuits, or several integrated optics circuits made on different substrate pieces (for instance if an integrated circuit is too large for being done on en single piece of substrate). The subpart circuit may be for instance glued together with the waveguides aligned.

In an advantageous embodiment, the sensor according to the invention may comprise first and second integrated optics circuits made on planar substrates arranged with an edge respectively facing a side of the beam splitter in the form of a beam splitter cube. The beam splitter cube may be of course of a square shape, or of any parallelepipedal shape (bar, ...)

The first and second integrated optics circuits further may advantageously include alignment optical channel waveguides.

The first and second integrated optics circuits and the beam splitter may be advantageously fixed permanently to each other, for instance by gluing them together.

According to some modes of realization, the focusing lens arrangement may comprise a chromatic lens with an extended axial chromatism, for focusing the light from the emitting apertures at different distances, depending on the wavelength, on the sample.

According to some modes of realization, the focusing lens arrangement may comprise only achromatic or essentially achromatic elements, for focusing the light from the emitting apertures on the sample.

According to some modes of realization, the focusing lens arrangement may comprise a reference plate, for generating a reference beam for use in interferometric detection modes. In all modes of realization, the focusing lens arrangement may advantageously comprise, or being constituted of, a telecentric lens system.

According to some modes of realization, the optical detector may comprise an intensity optical detector with an array or a matrix of detection elements or pixels, with pixels respectively receiving the light, over its whole spectrum, collected by a single collection aperture.

According to some modes of realization, said at least one optical detector may comprise a spectral optical detector with a filtering or spectrally dispersing element and an array or a matrix of detection elements or pixels, with pixels respectively receiving the light, over a subpart of its spectrum, collected by a single collection aperture.

According to another aspect of the invention, it is proposed a method for inspecting a sample, implemented in a multichannel confocal sensor according to the invention, comprises the steps of:

- emitting a light beam from at least one light source,

- splitting said light beam into a plurality of emitted light beams,

- directing said plurality of emitted light beams, through a focusing lens arrangement, onto said inspected sample,

- collecting reflected light beams from said inspected sample through said focusing lens arrangement, and

- directing said collected reflected light beams to at least one optical detector,

In the inspection method according to the invention

- the splitting step is implemented in a first integrated optics circuit,

- the collecting step is implemented in a second integrated optics circuit, and

- the light beam directing steps are implemented in a beam splitter. According to some modes of implementation, the method of the invention may further comprise steps of: Varying the height of the sensor relative to the sample; and using an achromatic focusing lens arrangement and an intensity optical detector, detecting intensity over the full spectrum of collected reflected light beams. In that case, the sensor and the method of the invention implements a confocal detection scheme. Accord ing to some other modes of implementation, the method of the invention may further comprise steps of: Using an achromatic focusing lens arrangement and a spectral optical detector, detecting spectra of collected reflected l ig ht beams. In that case, the sensor and the method of the invention implements a spectral interferometric detection scheme.

Accord ing to some other modes of implementation, the method of the invention may further comprise steps of: Using a chromatic focusing lens arrangement and a spectral optical detector, detecting spectra of collected reflected l ig ht beams. In that case, the sensor and the method of the invention implements a 3D chromatic confocal detection scheme .

Accord ing to some other modes of implementation, the method of the invention may further comprise steps of: Using a chromatic focusing lens arrangement and an intensity detector, detecting intensity over the ful l spectrum of collected reflected light beams. In that case, the sensor and the method of the invention implements a 2D chromatic confocal detection scheme.

So, the invention implements a multichannel confocal detection sensor core based on a first integ rated optics circuit, a second integ rated optics circuit and a beam spl itter.

A first integ rated optics circuit incl uding arrays of achromatic channel waveg uides is used to transmit a detecting l ig ht, usual ly broad band , to create a hig h-density array of beams or measurement spots th roug h a focusing lens arrangement. The hig h-density beam array covers a wide area on the measured substrate, while al lowing a high lateral resol ution .

A second integ rated optics circuit also incl uding arrays of channel wavegu ides col lects the reflected light from a measured sample and transfer the information to an optical detection stage.

This second detecting integ rated optics circuit decouples the constraints imposed on the first optics circuit channels by having the waveguide arrays adjusted to combine the functions of spatial filtering of the unfocused l ig ht, red ucing crosstal k between channels, maintain ing the lateral hig h-resolution set by the channel arrays from the first circu it and stil l matching the pitch and apertures requirements of the detection elements. The optical channel waveguide design and monolithic nature of the integrated circuits may be combined to minimize and automate the assembly and alignment steps to build a mechanically stable device.

Integrated optics circuits are designed and produced to meet the combined requirements of high throughput and high position and resolutions in confocal, chromatic confocal of interferometric measurements.

It implies a set of divergent constraints on the device design and assembly. One of them is that a high-density parallel measuring channel array is required to increase the measurement throughput. The array will then have a high count of measuring channels and minimum separation between them without compromising the resolution of the system by enabling crosstalk between the adjacent optical channels.

The channel optical waveguides constituting the measuring channels must also be compatible with the specificity of a confocal detection system, either chromatic or interferometric: they need to guide, split and collect in an achromatic way the usually broadband light but also be compatible and maintain the power density required per measuring channel to enable a noise- free detection.

The optical waveguide channels need to optimize the collection of the reflected light and direct this light efficiently to the detection devices. In addition, increasing the resolution and the measuring channel count will lower the tolerance for aligning the different function elements.

In this invention, integrated optics circuits are designed and assembled to provide a solution to those constraints.

Due to the use of the integrated optical components and the optical waveguide technology, it is possible to make a device with a large number of measurement channels (such as a few hundred) which is very compact.

At least two optical integrated circuits and a beam splitter component are used to decouple the requirement on the channel optical waveguide designs and optimize the light coupled in and out of the optics circuits. The performances are set during the design stage of the optics circuit and channel waveguides and, due to the nature of the integrated optics technology, can be easily and cost-effectively maintained and reproduced in a production environment. Only a few elements have to be assembled and aligned. This is a significant difference and improvement over the optical fiber bundle approach.

A first integrated circuit is used to create the high count, high density measuring channels by splitting the light from a broadband source. Single mode channel waveguides are preferably used to build achromatic and low- loss power splitters. They emit light with small beam diameters (few microns) because of the diffraction limit achievable with single mode waveguide propagation and are distributed linearly to form a small pitch (from a few microns to a few tens of microns) arrays to maintain high density. A low numerical aperture (NA), also achievable thanks to the single mode waveguide propagation, make the emitted light easily managed by the focusing lens arrangement.

A second integrated circuit is built to accomplish the function of collecting the reflected light (inputs of the circuit) from a measured substrate interface that went through the focusing lens arrangement and direct it to a detection element (outputs of the circuit).

The optical channel waveguides of the second integrated circuit can be designed to match very accurately the positions of the optical channel waveguides made in the first optics circuit. They can be made with larger diameter and larger angle acceptance to guide multimode light and thus allow a better collection efficiency of the reflected light.

Stray light which is light coupled into the chip but not propagated by the optical channel waveguides can be eliminated thanks to curvature of the channel waveguides along their propagation direction. Moreover, the optical channel waveguide can be designed to match the apertures and pixel pitch of the detection elements.

The different circuits and the beam splitter can be aligned and fixed permanently relatively to each other to realize a mechanical stable device. This approach can be easily automated due to the mass production nature of the optical integrated technology.

It is to be noted that the arrangement of the invention with two separate integrated optics circuit and a beam splitter provides unique advantages over classical multichannel optics sensors of the prior art. As previously explained, the use of integrated optics circuit with planar waveguides allows higher density and number of measurement channels, and very easy and accurate adjustments of measurement channel pitch or separation compared to optical fiber based architectures. For instance, it is easy to build systems with several hundreds of channels, whereas fiber optics bundles with more than 100 fibers are difficult to manage. In addition, the spacing of planar waveguides is easy to adjust very accurately during the manufacturing of the component, while spacing of fiber in bundles require complex assemblies in V-grooves.

The arrangement of the invention has also unique advantages over sensors architectures based on couplers or Y-junctions with a single common optical aperture used as emitting aperture for the emission of the light and collection aperture for collecting the back-reflected light. In the invention, the emitting aperture and the collection aperture may be designed differently, and optimized.

For instance, it is preferable to implement a single mode waveguide for doing the emitting aperture. This ensure to produce a measurement spot at diffraction limit, with the best spatial coherence and a low numerical aperture best suited for the focusing optics. For the collection aperture, it is better to use a larger waveguide which allows collecting more light reflected from the object, on a larger numerical aperture, and which allows notably taking into account the enlargement of the spot due to the resolution of the optics. Such waveguide may then be multimode. Description of the drawings

The methods according to embodiments of the present invention may be better understood with reference to the drawings, which are given for illustrative purposes only and are not meant to be limiting. Other aspects, goals and advantages of the invention shall be apparent from the descriptions given hereunder:

- Fig. 1 illustrates a first mode of realization of the device of the invention,

- Fig. 2 illustrates a second mode of realization of the device of the invention, - Fig. 3a, Fig. 3b and Fig. 3c Illustrate some modes of realization of focusing lens arrangement.

Detailed description of the invention

It is well understood that the embodiments described hereinafter are in no way limitative. Variants of the invention can in particular be envisaged comprising only a selection of the features described below in isolation from the other described features, if this selection of features is sufficient to confer a technical advantage or to differentiate the invention with respect to the state of the prior art. This selection comprises at least one preferred functional feature without structural details, or with only one part of the structural details if this part alone is sufficient to confer a technical advantage or to differentiate the invention with respect to the state of the prior art.

In particular, all the described variants and embodiments can be combined if there is no objection to this combination from a technical point of view.

With reference to Fig . 1, a confocal device of the invention will be described .

According to a first mode of realization, it will be described in a chromatic confocal configuration.

In such configuration, the device of the invention comprises a broadband light source 14. The light source 14 may comprise, for instance, a thermal source (halogen for instance), an array of LEDs, or a supercontinuum light source for instance. The light source 14 is optically connected, for instance by an optical fiber 24, to a first optical integrated circuit 11.

This first integrated circuit 11 comprises a plurality of achromatic emission channel waveguides 12 and Y-junctions 28 guiding the light of the light source 14 towards an array of emitting apertures 29, made on an edge of the first optical integrated circuit 11 by terminating end of emission channel waveguides 12. The light exiting these emitting apertures 29 forms the optical measurement channels 13. Preferably, the emission channel waveguides 12 are single-mode within the spectral range of the source (used spectral range). They connect cascaded Y-junctions 28 which split the light of the source 14 between the emitting apertures 29. In Fig . 1 , only the first and the last stages of the cascade of Y-junctions is shown for sake of clarity. The emission channel waveguides 12 and the Y-junctions 28 are substantially achromatic within the spectral range of the source, which means notably that the coupling ratio of the Y-junctions 28 is little or not dependent on the wavelength .

After exiting the emitting apertures 29, the light of the optical measurement channels 13 passes through a beam splitter cube 22, and is directed to a focusing lens 10.

In the chromatic confocal configuration described, the focusing lens arrangement 10 may comprise, as illustrated on FIG . 3a, a collimating lens element 30 and a chromatic lens element 31.

The focusing lens arrangement 10 with the chromatic lens element 31 is designed according to well-known techniques so as to provide a strong chromatic aberration, allowing different optical wavelengths crossing the lens to be focused at different axial distances, i.e. distances along the optical axis of the lens, or along the Z axis as shown in Figure 1.

The light of the respective optical measurement channels 13 is focused by the focusing lens arrangement 10 on an array of respective measurement points 15, or more precisely, thanks to the chromatic lens element 31, along a measurement line 15 in the axial direction of the lens (z-axis in Fig . 1 ) depending on the wavelengths. The range at which the respective wavelengths of the light source 14 are focused defines the chromatic measurement range.

When an interface of an object 17 is present in the chromatic measurement range, the wavelengths focused on that interface for the respective optical measurement channels 13 are reflected and directed by the beam splitter 22 to collection apertures 18 defined by optical channel waveguides 19 fabricated in a second integrated optics circuit 20.

The light which is not focused on the object 17, or the wavelengths which are not focused on that object 17 are rejected thanks to the confocal configuration of the set-up. Emitting apertures 29 and collection apertures 18 are optical conjugate of the measurement points 15 for the chromatic focusing lens arrangement 10 only at wavelengths focused on the interface of the object 17.

The second integ rated optics circuit 20 comprises detection channel wavegu ides 19, with an entrance end on an edge of the second integ rated optics circuit 20 forming the emitting apertures 29. These detection channel wavegu ides d irect the collected reflected l ig ht towards optical detectors 25 facing an edge of the second integrated optics circuit 20.

In the chromatic confocal config uration, optical detectors 25 may comprise spectral detectors to create an intensity spectrum of the lig ht issued from each detection channel waveg uides 19.

Such spectral detectors may comprise a first lens for coll imating the incoming light issued from the detection channel wavegu ides 19, a d ispersing element such as prisms, a d iffraction array or a g rating for d ispersing ang ularly the d ifferent wavelengths of the incoming l ig ht, a second lens and a linear or matrixial detector such as a CCD for re-imag ing the d ispersed l ig ht so that d ifferent wavelengths are focused on d ifferent pixels of the sensor, and a pixel provides a measurement of l ig ht intensity at a specific wavelength or within a limited range of wavelength .

Spectral sensors may also comprise color sensors with color filters applied on an array or a matrix of pixels of a CCD or a CMOS sensor, so that a pixel provides a measurement of light intensity at a specific wavelength or within a limited range of wavelength .

In both cases, the intensity spectrum of the light is also obtained by col lecting the information on the pixels of the sensor. An intensity spectrum is produced for each detection channel waveg uide 19.

A separate CCD or CMOS array may be used for col lecting the l ig ht of each d ifferent detection channel waveg uide 19. But preferably, the light from al l or at least a plurality of detection channel wavegu ides 19 is d irected towards a CCD or CMOS matrix, the wavelengths or the colors being d ispersed along one d irection of the matrix.

With the chromatic lens element 31 , an interface of the object 17 present in the measurement range gives rise to a peak in the intensity spectrum around the wavelength focused at the corresponding axial position . The intensity spectrum is analyzed to obtain an axial d istance information, or the position of the interfaces or the surface of the object 17 within the measurement range . So, a 3D profile of the object 17 is obtained along the array of optical measurement channels 13. A 3D image can be obtained by scanning the surface of the object with the sensor.

In the chromatic confocal configuration, optical detectors 25 may also comprise intensity detectors for fast 2D imag ing . Such detectors comprise a CCD or a CMOS sensor (for instance) with an array or a matrix of pixels. As there is no d ispersing element, in that config uration a pixel of the sensor col lects the whole l ig ht over the ful l spectrum of the light issued from a detection channel waveg uide 19.

Preferably, the light from all or at least a plurality of detection channel wavegu ides 19 is d irected towards a same CCD or CMOS array (line sensor) or matrix.

So, a 2D intensity or g ray level line of the object 17 is obtained along the array of optical measurement channels 13, with an optimal lateral resol ution over an extended depth of focus. A 2D image can be obtained by scanning the surface of the object with the sensor.

In can be noted that, in the chromatic confocal config uration, optical detectors 25 can comprise simultaneously spectral detectors and intensity detectors, for instance on d ifferent optical measurement channels 13, for acq uiring simultaneously 2 D and 3D information .

The optical detectors 25 may be interfaced with the integ rated optics circuits 11 , 20 either d irectly, with the optical detector entrance or sensor facing the end of the detection channel waveg uides, or using known interfacing method such as fiber optics 24, V-g roove, butt-coupl ing, micro- lenses and/or tapers in waveg uides, in particular compatible with well automated methods and low coupl ing power loss.

Detection channel waveguides 19 are preferably desig ned for allowing multimode propagation . This allows to have them made with a larger cross- section area and larger acceptance ang le at the col lection apertures 18 so as to maximize the col lection of the reflected l ig ht from measured points 15 and also minimize the crosstal k d ue to reflected light captured by adjacent detecting channels. In the confocal detection scheme described, the detection channel waveguides 19 are further built with an angled (90 degrees) or curved design so as to direct any light not coupled into the detection channel waveguides 19 away from the optical detector 25. In other words, the detection channel waveguides 19 are arranged so that the optical detectors 25 are out-of-sight of the stray light coupled at the collection apertures 18 level .

It can be noted that the angles design of the detection channel waveguides 19 allows integrating optical detectors 25 on two sides of the second integrated optics circuit 20 as illustrated on Fig . 1. Of course, some other modes of realization may implement optical detectors 25 on a single side of the second integrated circuit 20.

As described above, the second (detection) integrated optics circuit 20 may be built and arranged so as to fulfill the following characteristics and requirements :

- match the detection channel waveguide 19 array pitch at its entrance

(or the collection apertures 18 array pitch) to the first circuit emitting channel waveguide 19 array pitch (or the emitting apertures 29 array pitch) of the first integrated circuit 11 ;

have multimode guides (larger dimensions and acceptance angle) for better light collection of the measuring light reflected from the substrate;

match the detection channel waveguide 19 array pitch at its exit to the pitch of the optical detector 25 pixels, ;

match the optical apertures of the optical detectors 25;

- make the detection channel waveguides 19 with a curvature to reject stray light not coupled into the detection channel waveguides 19. In particular, the detection channel waveguides 19 may be arranged on the detector side to match the pitch of the optical detector 25 pixels and the optical apertures of these optical detectors 25, or the entrance optics of these optical detectors 25.

For instance, when using intensity detectors, it is desirable to have the light issued from different detection channel waveguides 19 projected on different pixels of the optical detector 25 to avoid cross-talk. This may be achieved by providing detection channel waveguides 19 with a pitch at their exit side for instance correspond ing to twice the detector pixel pitch (to have one pixel over two used for cross-tal k isolation) . If the optical detector has an imag ing entrance stage, the detection channel wavegu ides 19 may be arranged with a pitch and a numerical aperture so as to match the detector pixel pitch in the same way throug h that imag ing stage .

When using spectral detector, it is desirable to have the light issued from d ifferent detection channel waveg uides 19 and spectral ly d ispersed projected on d ifferent rows of pixels of the optical detector 25, to avoid cross ^¬ talk. This may also be achieved by provid ing detection channel waveg uides 19 with a pitch at their exit side for instance correspond ing to twice the detector pixel pitch (to have one row of pixel over two used for cross-talk isolation) . Of course, as previously, the imag ing entrance stage of the optical detectors 25 has to be taken into account.

The sensor of the invention further comprises means for control and data processing such as computer, microcontrol ler and/or FPGA modules 23, connected to the optical detectors 25.

The first and second integrated optics circuits 11 and 20 are al ig ned relatively to each other and to the focusing lens 10, and are made to stay permanently in close contact with the beam splitter 22.

In particular, in some modes of real ization, the first and second integ rated optics circuits 11 and 20 are bounded or g l ued to the beam spl itter 22 so as to form a single piece or monol ithic arrangement. A method for doing the permanent connection is to use an optical adhesive in or around the optical paths. A mechanically robust and optically stable device is then manufactu red .

Al ignment optical chan nel waveguides 26, 27 can be added respectively to the first integ rated optics circuit 11 and to the second integ rated optics circuit 20 in order to provide alignment means. Those alignment optical channel wavegu ides 26, 27 can be coupled permanently or temporarily to optical fibers g uid ing the light from a pilot l ig ht and/or to a mon itoring device (power meter or camera for examples) .

For instance, a first set of al ig nment waveg uides 26 is used to alig n the first integrated optics circuit 11 in the focal plane (or an imag ing plane) of the focusing lens 10. These first alig nment wavegu ides 26 are arranged on both sides of the emission channel waveguides 12 and have an emitting aperture located close to the emitting apertures 29 of these emission channel waveguides 12.

In a first step of alignment, light is injected in these first alignment waveguides 26 towards the focusing lens arrangement 10 and a reference object 17 or mirror placed at a reference position or a conjugate plane of the focusing lens arrangement 10. Reflected light is collected back in these first alignment waveguides 26. The relative positioning of the focusing lens arrangement 10 and the first integrated optics circuit 11 is adjusted by optimizing this coupling .

Then, a second set of alignment waveguides 27 on the second integrated optics circuit 20 is used to align this second integrated optics circuit 20 relative to the first integrated optics circuit 11. These second alignment waveguides 27 have collection apertures 18 designed to be in confocal arrangement with the emitting apertures of the first alignment waveguides 12 through the focusing lens arrangement 10.

In a second step of alignment, light is injected in the first alignment waveguides 26 towards the focusing lens arrangement 10 and a reference object 17 or mirror placed at a reference position or a conjugate plane of the focusing lens arrangement 10. The second integrated optics circuit 20 is then positioned so as to optimize the reflected light coupled in the second alignment waveguides 27.

At that stage, the relative alignment of the first integrated optics circuit 11, the second integrated optics circuit 20, the beam slitter cube 22 and the focusing lens arrangement 10 is achieved . All pieces can be fixed together.

In a last step, optical detectors 25 can be aligned relative to the optical detection channel waveguides 19 using the light propagated in these detection channel waveguides 19.

In addition to providing a mean to optimize the performance of the confocal device, the alignment optical channel waveguides 26, 27 can also be used to automate the device assembly process.

The first and second integrated optics circuits 11 , 20 may be done with several techniques to produce optical waveguides, which are areas with a higher index of refraction embedded in a transparent substrate with a lower index of refraction . For instance, the integ rated optics circuits may be done using an ion exchange process on a g lass substrate. Such ion exchange occurs between a g lass substrate and a molten salt bath when these are suitably broug ht into contact. This phenomenon local ly increases the optical index of the g lass by mod ifying its composition .

A thin metal layer is deposited on a glass substrate. Openings with a size of a few microns to a few tens of microns, with desig ns corresponding to waveg uides, Y-junctions and other components, are opened in the metal lic layer using a classical photol ithog raphy technique.

A one-step or two-steps ion-exchange process may be implemented to create the waveg uides on, or below, the g lass surface.

The first step consists in d iffusing at hig h temperature ions such as silver or potassium ions into the glass wafer using a molten salt bath, so as to create surface waveg uides.

Then, if applicable (for instance with silver ions), an electric field may be applied for moving the ions and thus the waveguides deeper into the glass.

The integ rated optics component may also be done using techn iq ues involving deposition of layers of doped sil ica or other materials on wafers for constituting waveguides. The deposition steps usual ly involve CVD techniques.

The integ rated optics component may also be done using direct inscription techniques. For instance, the wavegu ides may be done by modifying locally the index of refraction of a sol-gel substrate or a polymer resin with a UV laser beam, by photo-polymerization .

A Y-junction as used in the invention may comprise a first waveguide which is en larged progressively in a tapered zone (or a taper) end ing in two branching waveguides. The tapered zone is preferably smooth enough to allow an ad iabatic transition with a spread of the spatial modes of the g uided light issued from the first waveg uide, which are coupled in the branching waveg uides.

Such arrangement has the advantage of being essential ly achromatic over a broad spectral range. So, a fixed spl it ratio (for instance 50/50) may be obtained over such broad spectral range. In is to be noted that such achromatic behavior is not achievable with classical fiber based couplers or Y-splitters which rely on a propagation mode coupl ing scheme with a strong dependence on the wavelength .

With reference to Fig . 2, the first integrated optics circuit 11 , the second integ rated optics circuit 20 and the beam spl itter 22 may be arranged so that the first integ rated circuit 1 1 and the second integ rated optics circuit 20 are facing orthogonal faces of the beam spl itter 22 (as in the config uration of Fig . 1 ), but are also arranged so that their waveguide planes are orthogonal to each other (while in the arrangement of Fig . 1 the waveguide planes of these integrated circuit are parallel to each other or in the same plane) .

Or in other words, in the arrangement of Fig . 2, the first integrated circuit 1 1 and the second integ rated optics circuit 20

Such configuration has the advantage of req uiring a much smal ler beam splitter cube 22, which may be for instance in the shape of an elongated bar.

The device of the invention has been described for a chromatic confocal configuration .

As previously explained, it may also be used to implement other detection schemes, some examples of which are explained below. All what has been previously explained, in particular in relation with Fig . 1 , Fig . 2 and Fig . 3a is applicable to these other modes of realization, except for the d ifferences which are explained below.

Accord ing to another mode of real ization, with reference to Fig . 3b, the device of the invention may implement a confocal detection scheme. In that case :

- The light source 14 may be a monochromatic (laser) or a broadband source;

- The focusing lens arrangement 10 may comprise achromatic lenses 33 with a hig h NA for focusing the measurement beam on the surface of the object 17;

- The optical detectors 25 may comprise intensity detectors. The device may further comprise a Z-scanning stage 34 to move relatively the sensor and the object 17 along the Z axis. The height of the surface of the object 17 may then be obtained at the measurement points from the position of the Z-scanning stage 34 for which a maximum of intensity is obtained on the intensity optical detectors 25.

Accord ing to another mode of real ization, with reference to Fig . 3c, the device of the invention may implement an interferometric detection scheme based on a low-coherence interferometry technique.

Accord ing to a first variant, a low-coherence interferometry techniq ue in the spectral domain may be implemented . In that case :

- The light source 14 may be a broad band source;

- The focusing lens arrangement 10 may comprise achromatic lenses 35 for focusing the measurement beam on the surface of the object 17, and optionally a reference plate 32 of the interferometer;

- The optical detectors 25 may comprise spectral detectors.

In that config uration , the intensity spectrum of the l ig ht as obtained by the spectral optical detectors 25 is modu lated by the interferences between beams reflected by the reference plate 32, and/or interfaces of the object 17. The heig ht between the reference plate 32 and the object 17, and/or the thickness of layers of the object 17 may be obtained from that intensity spectrum using well-known techniques.

Accord ing to a second variant, a low-coherence interferometry technique with a swept source may be implemented . In that case :

- The lig ht source 14 may be a swept laser;

- The focusing lens arrangement 10 may comprise achromatic lenses 35 for focusing the measurement beam on the surface of the object 17, and optionally a reference plate 32 of the interferometer;

- The optical detectors 25 may comprise intensity detectors;

In that configuration, an intensity spectrum of the light as obtained by sweeping the freq uency of the light source 14 over a freq uency range and acquiring the light intensity for each freq uency with the intensity l ight detectors 25. As previously, the height between the reference plate 32 and the object 17, and/or the th ickness of layers of the object 17 may be obtained from that intensity spectrum using well-known techniq ues. While this invention has been described in conjunction with a number of embodiments, it is evident that many alternatives, modifications and variations would be or are apparent to those of ordinary skill in the applicable arts. Accordingly, it is intended to embrace all such alternatives, modifications, equivalents and variations that are within the spirit and scope of this invention.