Classifications

• • G—PHYSICS
  • G01—MEASURING; TESTING
  • G01B—MEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
  • G01B11/00—Measuring arrangements characterised by the use of optical techniques
  • G01B11/02—Measuring arrangements characterised by the use of optical techniques for measuring length, width or thickness
  • G01B11/06—Measuring arrangements characterised by the use of optical techniques for measuring length, width or thickness for measuring thickness ; e.g. of sheet material
  • G01B11/0616—Measuring arrangements characterised by the use of optical techniques for measuring length, width or thickness for measuring thickness ; e.g. of sheet material of coating
  • G01B11/0658—Measuring arrangements characterised by the use of optical techniques for measuring length, width or thickness for measuring thickness ; e.g. of sheet material of coating with measurement of emissivity or reradiation

Definitions

• the field concerned is that of the metrology of the thin layers used inter alia in the modern technologies of microelectronics, or even in other industrial fields such as coatings on metals (paints, varnishes, anodizing, ).
• These thin layers are developed from cleanroom processes that are used in the manufacture of electronic components.
• the thickness of these layers is generally between a few tenths and a few tens of microns. Thickness measurement represents a major problem, both during the production chain and in the end for the acceptance of the specifications of the production technology range. Thin film deposits cover a very wide range of functions required in the design of the components concerned.
• active layers that are at the heart of physics (for example semiconductors, electro-acoustics, electro-optics, etc.) and passive layers that are either conductors (metal deposits) ultimately ensuring interconnections between different active areas, either dielectrics that provide electrical isolation of different areas or other specific functions.
• the physicochemical nature of these thin layers is very broad, for example: they can be crystalline for the active layers (semiconductors silicon or III-V, piezoelectric ..), polycrystalline or amorphous for the insulants (oxides or nitrides), metallic (conductors) ...
• PVD Physical Vapor Deposition
• CVD Chemical Vapor Deposition
• Dispersion firstly concerns intra-wafer inhomogeneity but also plate-to-plate differences between production batches.
• the measurements performed by pointing at points distributed on the surface of the deposit and away from this reference are tainted thickness variations between the mechanical contact plane of the plate on the table and the lower level of the layer , that is to say, the thickness variation of the substrate (known by the acronym TTV corresponding to "Total Thickness Variation”) and also deformations of the geometry of the plate resting on three points on the table.
• TTV thickness variation of the substrate
• TTV Total Thickness Variation
• NDT non-destructive testing
• a first family of these NDT techniques brings together point-to-point probing processes that then allow displacement on an XY displacement table to reconstruct a 2D cartography of the thickness.
• the typical example is that of ultrasound probes.
• the temporal analysis between the pulses emitted and received in reflection or transmission between the two dioptres (surface and interface) makes it possible to go back to the local thickness of the layer.
• these systems are complex and expensive especially to achieve a resolution in the micron range, which requires the implementation of very high frequency acoustic probes (acoustic microscope).
• the weakness of these systems is the prohibitive scanning time and the cost of XY micrometric displacement tables, given the positioning accuracies sought.
• the present invention relates to a simple method of implementation, responding to the aforementioned problems and exploiting the emissivity of the thin layer whose spatial thickness is to be analyzed, this emissivity being revealed by the emission infrared in a temperature rise of said thin layer.
• the invention relates to a device for determining a spatial data set of thickness d (x, y) of a thin layer to be analyzed comprised in a thin layer (s) / substrate assembly, characterized in that it comprises:
• heating means for heating said thin layer so as to generate a set of infra-red radiations emitted energy W (x, y) dependent on the emissivity ⁇ [A, d (x, y) j of said layer, dependent on the emission wavelength ⁇ and the thicknesses of said layer d (x, y), said heating means being fed by control means;
• a theoretical calibration curve d (V) giving the evolution of the thickness as a function of an electronic data (V) dependent on the characteristics of the optoelectronic detection means and the emissivity of a thin layer of reference of the same nature as said thin layer to be analyzed included in the same thin layer (s) / substrate; o determining all the spatial data of thickness of said thin layer to be analyzed d (x, y) from the data V (x, y) and of said calibration curve d (V);
• control means of said heating means generate a temperature rise of said thin layer of about one to three tens of degrees Celsius.
• the heating means comprise means for heating by Joule effect a support on which is positioned said thin layer (s) / substrate.
• said support is a plate with high thermal conductivity that can be aluminum.
• the heating means are heating means by electromagnetic induction.
• the heating means comprise an infra-red source that can be a laser.
• the optoelectronic detection means of said set of infra-red radiations comprise a CCD or CMOS sensor, sensitive in the infra-red domain.
• the optoelectronic detection means are coupled to an input / output power supply card of analog and / or digitized data.
• the device comprises optical means located between the thin layer (s) / substrate and the detection means, which may comprise an optical objective.
• all the thicknesses d (x, y) of the thin layer are of the order of a few tens of microns.
• the thin layer is a layer of dielectric material on the surface of a highly reflective layer that may be metallic, said highly reflective layer may have a thickness of the order of a few microns.
• the thin layer is a layer of dielectric material in contact with a dielectric substrate.
• the device comprises polychromatic detection means, which can operate in a wide spectral band of wavelengths that can be between about 8 ⁇ and 12 ⁇ .
• the processing means comprise an operation integrating ⁇ [d (x, y)] of the emissivity factor ⁇ [( ⁇ , d (x, y)] over said bandwidth.
• the thin layer is made of piezoelectric crystal, the substrate being made of silicon or quartz or sapphire or of another monocrystalline material with infra-red emissivity properties different from those of the upper layer in part. absence of a metallized interface, in any case in the opposite case (with an interstitial metal layer).
• the device of the invention may advantageously be used when the thin layer has been obtained by transfer of a substrate on a host plate by various bonding techniques (molecular adhesion, thermal diffusion of thin metallic layers, organic adhesives) and thinning by mechanical (lapping-polishing) or thermomechanical methods (separation of the postponed plate having undergone ad-hoc ion implantation by thermal shock and chemical mechanical polishing) or any other approach allowing a thickness layer between a fraction of a micron to be achieved and several tens of microns.
• various bonding techniques mocular adhesion, thermal diffusion of thin metallic layers, organic adhesives
• thinning by mechanical (lapping-polishing) or thermomechanical methods (separation of the postponed plate having undergone ad-hoc ion implantation by thermal shock and chemical mechanical polishing) or any other approach allowing a thickness layer between a fraction of a micron to be achieved and several tens of microns.
• the thin layer may be a layer of paint, or any other protective coating, the substrate being a metal, with as possible applications those of metal coatings concerning the numerous applications in the coating industry. metallurgy.
• the subject of the invention is also a device for determining the thickness mapping of a thin film, characterized in that it comprises:
• the means for rendering the thickness data being electronic display means for said thickness data so as to constitute a 3D map of said thickness data.
• FIG. 1 shows schematically the reflections involved at different interfaces in the case of a thin layer on the surface of a substrate, used in the principle implemented in the invention
• FIG. 2 illustrates the theoretical response of the emissivity of a thin layer as a function of its thickness with a metal interface and with a dielectric interface in the case of a monochromatic infra-red detection
• FIG. 3 illustrates the theoretical response of the emissivity of a layer in the case of a 10 ⁇ monochromatic infra-red detection, of a multi-spectral detection in the band (9.5 ⁇ ; ⁇ ) and multi-spectral detection in the band (8 ⁇ ; 12 ⁇ );
• FIG. 4 illustrates an example of a thickness mapping measurement device according to the invention
• FIG. 5 illustrates a block diagram of the measurement and processing chain carried out in a device according to the invention
• FIGS. 6a and 6b respectively illustrate an example of thickness mapping obtained with a device of the invention with a composite wafer 4 inches in diameter comprising a thin layer of lithium niobate (LN) on a sapphire substrate and a mapping obtained with a confocal probe from the same sample;
• LN lithium niobate
• FIGS. 7a and 7b respectively illustrate examples of thickness mapping obtained with a device of the invention, with as a sample: a layer of paint on the surface of a steel plate and patterns made by different stacks of adhesive tapes simulating different thicknesses of coatings on metal as illustrated in Figure 7b.
• the present invention uses the known principle of thermography. Thermography is a technique in full development thanks to the very important progress of infra-red detectors worn for several decades for the needs of the defense (optronics) and more recently for the diagnosis of thermal losses in the context of energy savings of buildings and more broadly, the limitation of greenhouse gas emissions.
• n optical index
• the properties of the materials differ from this ideal situation with an emissivity factor of less than unity (ideal case).
• R the energy reflection factor of the body with incident radiation coming from outside
• Radiometry systems for remote temperature measurement have been developed since the development of photodetectors in the infrared domain and associated reading electronics. Imaging was then made possible with the advent of matrix detectors. The entire infra-red spectral range can be covered given the abundance of many modern photo-detection materials. Cameras with resolution better than one-tenth of a degree are common with imaging matrices of several hundreds of thousands of pixels, that is to say with a submillimetric spatial resolution for objects of one hundred millimeters of side, typically for the wafer dimension of current microelectronic materials (3 to 12 inches in diameter).
• the discontinuous nature at the boundary of two materials with different thermal properties is exploited (conduction and radiative emissivity), this concerns in particular the situation near the interface of the deposition of a thin layer with its substrate. If the layer remains thin vis-à-vis this disruption rupture at the interface, the radiative emission balance is strongly dependent on its thickness, this especially as the layer is thin. In this abnormal situation with regard to homogeneous body equations, it becomes possible to analyze the thickness of the layer as a function of the radiation balance measured by an infra-red imaging system.
• R represents the energy reflection of this structure composed of two diopters trapping an absorbing layer that behaves like a dissipative interferometer.
• the layer behaves like an interferometer with multiple internal reflections of the Fabry-Perot type.
• the incident beam E passes through the first diopter (upper surface of the thin layer) between the layer 1 of air (index n a ⁇ 1) and layer 2 (index n b ).
• the optical reflection coefficient r ab is perfectly defined by the indices n a and n b according to the well-known laws of Fresnel. If the incident optical amplitude is E, the amplitude of the first reflected beam is E r :
• the transmitted beam is t a .E, t a being connected to r ab by the conservation of energy:
• the beam is again a transmission value t in the air.
• the emerging beam E r2 is therefore:
• This outgoing beam E 3 also accumulates the same absorption and phase delay of value e "2 (j (p + ad) and we therefore have:
• the total reflected optical amplitude Er is obtained by summation of all the consecutive reflections E n , which makes it possible to define the total reflection coefficient:
• the overall reflection coefficient of this layer is obtained by using the following geometric series which is decomposed into amplitude and phase:
• n ⁇ °°, FX -2 «(3 ⁇ 4 ⁇ «.
• E rn concern a sequence on the reflection and transmission coefficients resulting from the Fresnel laws on the two dioptres as explained previously.
• the balance sheet also including the phase terms results in a simple expression of the reflection coefficient p:
• this thin multilayer structure behaves as a reflective opaque "gray body" of emissivity ⁇ :
• the infra-red emission is therefore very dependent on this thermal emissivity factor ⁇ (, ⁇ ) which depends on the thickness of the layer and than the emission wavelength in the emission band of the black body at the temperature T (in K).
• the invention thus takes advantage of the high sensitivity of the infrared emission with the thickness of the thin layer under certain conditions of stacking optical indices according to the law e (d).
• This high sensitivity makes it possible to detect the luminance variation as a function of the thickness even in the case of an infra-red radiation spectrum emitted for very low temperatures.
• the first case concerns a highly reflective interface, for example metallic, with a layer having an index n b of 2 (r ab -0.9); the answer shows an increasing and periodic law damped of the emissivity tending towards the unit for the high thicknesses (> 100 ⁇ ). The answer is all the more sensitive as the thickness is small.
• the preponderant factor on the sensitivity is the reflection coefficient r bc at the interface, the ideal function being the case presented of a reflection close to the unit, which is the case of a metal interface. Without metal layer, the reflection coefficient r ga is even higher than the index difference between the layer of n b and n c of the substrate will be high;
• the surface reflection coefficient r ab is 0.1; the law ⁇ () does not show an increasing rise starting from the zero emissivity for the very thicknesses low.
• the variations tend towards those of interferences of thin layers, the behavior tending towards that of a Fabry-Pérot cavity for a null absorption and the equalization of the coefficients of reflection with the two diopters.
• the monochromatic response of the emissivity law ⁇ () shows a more or less periodic behavior related to the interferometric behavior of the layer, a behavior that is very sensitive to interface properties, making it difficult to exploit this type of detection. for a thickness measurement.
• the computation of the W (d) wide-band infrared energy thickness dependence for a thin film consists of integrating the luminance law L (, d) over the wavelength range of the system detection, that is:
• FIG. 3 illustrates the filtering effect obtained on the oscillating behavior of the response and shows the evolution of the emissivity as a function of the thickness with the following parameters:
• the substrate may be glass, quartz, ...
• the principle of the invention can thus be implemented with a device such as that illustrated in FIG. 4 and uses the thermographic analysis of a layer 10 which in this example is heated by the rear of its substrate 1 1 on a heating plate 12 brought to a temperature slightly higher than that of the ambient.
• an infra-red detector 20 disposed vertically of the heated plate, is then able to record information relating to the different radiated energies W (x, y) materialized by vertical corrugations, themselves dependent on different spatial thicknesses of thin layers.
• the infra-red detector 20 records a set of spatial information V (x, y) related to the different infra-red radiation from the layer 10 analyzed.
• This set V (x, y) is treated so as to restore a set of thicknesses d (x, y).
• the aim is to be able to raise the temperature of the layer to be measured on its substrate.
• This substrate is in the most general way a wafer of the microelectronics but other substrates can be envisaged in a broadening of the field of the applications (one can consider the measure of homogeneity of thickness of industrial coatings on objects of the type plates but also of different shapes).
• the required temperature is only a few tens of degrees above ambient temperature because of the adoption of the detection principles proposed by the invention, typically between thirty and a hundred degrees Celsius.
• the choice of a temperature close to 50 ° C seems a good compromise between the sensitivity and the accuracy of the measurement, freeing itself from parasitic radiation from the environment at room temperature. It also promotes the speed of steady temperature of the sample and the preservation of the sample.
• the Applicant has established that a homogeneity of the heat source better than 0.5 ° C is sufficient for a precision on the thickness of the a few %. Furthermore stability of the same order during the measurement is also a criterion of importance to ensure reliability.
• the plate is heated by Joule effect and stabilized by thermostat, for example with stabilization of the temperature referenced on a setpoint by a metrological sensor (thermocouple, thermistor ..).
• a metrological sensor thermocouple, thermistor ..
• This means is more particularly adapted to the heating of a substrate having a flat surface back which can ensure a good heat exchange contact with the top of the heating plate.
• the accuracy can go down to a few tenths of a degree for temperatures below 50 ° C, thus respecting the required specifications (previously defined).
• the homogeneity depends very much on the nature of the plate and the thermal contact between heating source and sample.
• the heating means may also use an oven, means by high frequency electromagnetic induction, an infra-red lamp, or even a laser ....
• These means and methods are more suitable for a rapid rise and high temperature (flash heating) in specific contexts, for example on an industrial line.
• the use of a low-temperature oven oven with a transparent optical port in the infra-red detector strip is a solution for the measurement of thin coatings on objects that do not have a flat rear surface. In certain performance restrictions, the measurement of layers covering three-dimensional objects is also conceivable.
• the device of the invention also comprises optical means involved in the detection of infra-red emissions.
• optical imaging block of the sample that makes it possible to project the image in the field.
• infrared of the surface of the layer to be measured on a matrix sensor. It is essentially an optical lens that can be optionally equipped with a zoom option to refine the measurement on a specific area of the layer.
• Other optical components can also be used in addition, such as, for example, wavelength filters that can be used to adapt the performance of the measurement to various specific specifications.
• the device of the invention also comprises an optoelectronic detection and image acquisition subsystem.
• this subsystem comprises a base component of matrix type sensitive to power flows in the infra-red domain.
• CCD or CMOS complementary metal-oxide-semiconductor
• CMOS complementary metal-oxide-semiconductor
• To this matrix sensor is associated a power supply card and input / output commands and analog data (video signal) and / or digitized.
• Each pixel P (i, j) of the matrix sensor delivers an electric voltage V (i, j) in response to the radiation energy flux focused by the optical subsystem from the infrared emission W (x, y) in each point of coordinates (x, y) of the layer to be measured.
• An electronic amplification chain is connected to each of these voltage sources.
• the global gain balance k of the chain is the product of the optoelectronic response of the detection element by the gain of this amplification chain:
• V (x, y) k. W (x, y)
• video frames are available in analog or digital forms.
• the complete frame of an image corresponds to a spatial matrix of the voltages V (x, y).
• the radiative flux W (x, y) emitted by each point of the thickness layer is expressed as a function of the thickness d (x, y) by: ⁇ [ ⁇ , ( ⁇ , ⁇ )] is the emissivity factor representative of the thickness of the layer at each point (x, y).
• ⁇ is Stephan's constant and ⁇ [ ⁇ ( ⁇ , ⁇ )] the emissivity factor at each point (x, y) of the thickness layer d (x, y).
• the matrix V (x, y) is therefore expressed by:
• V (x r y) ⁇ . ⁇ *. ⁇ [ ⁇ (. ⁇ , ⁇ )]
• the subsystem also includes a system for controlling and acquiring images of the data V (x, y).
• a computer equipped with camera control cards and software serves as a human-machine interface.
• the extraction from the camera of the matrix type data files V (x, y) is easily performed via a usual fast link so as to process the data collected.
• the device of the invention comprises for this purpose means of postprocessing for determining the matrix of thicknesses d (x, y):
• the determination of the matrix of thicknesses d (x, y) results from the application of the inverse transformation ⁇ ⁇ ⁇
• the application of the theoretical parameters is not exploitable because of the approximations of the modeling and the too many external parameters.
• the solution is to use the determination of a d (V) calibration curve by experimentation on a thin reference layer.
• a calibration curve d (V) appropriate for each type of stack must be implemented, and for a given amplification chain k.
• FIG. 5 provides a block diagram of the measurement and processing chain carried out in a device according to the invention, making it possible to synthesize the various means and steps implemented.
• the thin reference layer 10c makes it possible to define the calibration law d (V).
• the results obtained from the thickness mapping of the thin layer to be analyzed 10 are determined from the information V (x, y) collected at the detector 20, and transferred for display on a screen 30.
• T (x, y) ⁇ ( ⁇ , ⁇ , ⁇ ) x To, where To is the true temperature of the layer.
• the Applicant has shown that, under the conditions of thermal conduction of the tested parts (typically wafers and layers of thin microelectronic materials ( ⁇ 1 mm)), the inhomogeneity of this true temperature that can result from thickness variations is very weak, less than one thousandth of a millimeter. These true temperature variations are negligible compared to the fictitious temperature variations related to the variations of emissivity ⁇ [ ⁇ ( ⁇ , ⁇ )] observed and demonstrated with the variations of thickness of the layer and its dioptres realizing a "gray body"reflective;
• a concrete experimental case validates the concept of the invention. This is the measurement of the stack which consists of a ferroelectric crystal layer of lithium niobate LiNb0 3 of inhomogeneous thickness between 15 and 65 microns on a layer of fine metal (1 ⁇ ) itself deposited on a reference substrate in polished thick glass with a diameter of 3 inches, parallelism of a few microns and flatness of the AR face better than a micron.
• An infrared mapping measurement is performed according to the principle of Figure 4 with the following materials and protocols: a heating plate is raised to 40 ° C (accuracy +/- 1 ° C), followed by waiting for the stationary thermal regime for 2 minutes; measurement and acquisition of the thermography are performed using a FLIR SC325 series camera; in parallel, a reference measurement is performed with a mechanical probe, the plate being supported on a marble. The reference of the measurement (zero) is made by pointing to an edge of the flat of the glass substrate not covered by the LiNbO 3 layer and the different thickness values are then pointed at different equitably distributed locations on the surface of lithium niobate. . The surfaces are all in the state of optical polish. Thickness mapping is extrapolated from this spatial distribution of thicknesses measured at the mechanical probe;
• the comparison between the temperature mapping measured by thermography and the mapping of the thicknesses measured at the mechanical probe shows an almost perfect correspondence of the distributions of the different equi-distant zones. Different tests show identical results.
• the correlation between the thickness and the measurement with the infrared camera is perfectly proved.
• the sensitivity of the temperature with the thickness is of the order of 0.3 ° C per ⁇ for the small thicknesses and 0.15 ° C / ⁇ for the thick ones.
• the thickness measurement accuracy is of the order of 0.4 ⁇ in the range of thicknesses from 0 to 30 ⁇ then 1 ⁇ of 30 to 60 ⁇ thick.
• a finer calibration is quite possible with a chromatic confocal optical probe instrumentation for example. Nevertheless, these systems remain slow because of the need for mechanical X, Y scanning of the surface and the investment is expensive.
• the Applicant has also mapped the thickness of a thin layer of LiNbO 3 on sapphire.
• Figure 6a illustrates this mapping.
• This example relates more precisely, the case of a composite wafer 4 inches LiNb0 3 thinned up to a few tens of microns on sapphire substrate.
• the reference measurement of the thickness mapping is performed on a high-end commercial confocal probe as illustrated in FIG. 6b and a very good correlation is obtained between the two types of measurement.
• the experimental case mentioned above is only a particular case of a structure resulting from the same technological family of elaboration of thin crystalline layers obtained by a wafer bonding technique of a monocrystalline active wafer ( LiNb0 3 in the case cited above) on a passive wafer (glass) with a metal interlayer.
• the active wafer is then thinned down to a few tens of microns.
• LiTa0 3 lithium tantalate
• the Applicant has also tested the feasibility of mapping the thicknesses of a wafer 4 inches undergoing thinning, that is to say in a state of frost that excludes any other optical process. This possibility of almost instantaneous measurement is very interesting for the control and the catching up of parallelism during slimming in a production objective.
• the Applicant has also carried out tests to validate the device concept according to the invention from a paint layer on the surface of a steel plate as shown in FIG. 7a and patterns made by different stacks of ribbons. adhesives simulating different thicknesses of coatings on metal as illustrated in Figure 7b.
• the sensitivity of the measurement in the standard operating conditions of the camera used is sufficient to allow a measurement within the accuracy limits described above.
• the presence of the metal layer exacerbates the discontinuity of the radiative heat dissipation regime due to its very low emissivity vis-à-vis the upper layer (ratio of more than 100).
• the physical explanation of this radiative behavior is well modeled and validated experimentally: it involves a Fabry-Perot type interference mechanism of the radiation emitted in the layer and trapped by multiple reflections between the metal layer (reflection close to the unit) and the upper diopter layer / air.

Landscapes

• Physics & Mathematics (AREA)
• General Physics & Mathematics (AREA)
• Length Measuring Devices By Optical Means (AREA)

Applications Claiming Priority (2)

Publications (1)

Family

ID=46826845

Family Applications (1)

Country Status (3)

Families Citing this family (1)

Citations (3)

Family Cites Families (3)

• 2012
  • 2012-07-26 FR FR1257234A patent/FR2993972B1/fr active Active
• 2013
  • 2013-07-26 WO PCT/EP2013/065827 patent/WO2014016416A1/fr active Application Filing
  • 2013-07-26 EP EP13753590.2A patent/EP2877812A1/de not_active Withdrawn

Patent Citations (3)

Non-Patent Citations (1)

Also Published As

Similar Documents

Legal Events