Technical field of the invention

The present invention is applicable to contactless characterization of materials used in engineering, to measure thermal properties at microscopic ranges, particularly dilation, thermal conductivity, thermal diffusion coefficients and film thicknesses, the characterization of optical properties such as adsorption and dispersion spectrums of very small devices in the range of nanometers to micrometers and measuring of adsorption spectrums of very small particles (in the range of nanometers to millimeters) .

STATE OF THE ART AND PROBLEMS TO SOLVE

The determination of thermal parameters (dilation, thermal conductivity, thermal diffusion, etc.) of materials and devices has a major importance to predict performance in service thereof. Contactless measuring techniques allow detection without damaging the sample or system as the measurement during operation without disturbing the device. Also they allow remote measurements of parts not accessible by contact.

One of the most popular techniques is based on the periodic or transient disturbance of the system through electromagnetic heating and later or simultaneous measuring of the caused effects. Photothermal techniques that allow heating the sample with a pulsed or modulated laser beam and measuring of the temperature increase by infrared radiation are known in the art, see US 5.667.300 and 7.060.980 and published US patent applications 2002/0031164 and 2002/0011852. These techniques have the drawback of requiring infrared detectors that must be cooled and that prevent the use of optic microscopes for detection. Similar variations measure the reflection change of the sample using a second laser and measuring the reflected intensity that due to changes of temperature varies in time, allowing assessing said temperature change, (D. Rochais et. al., J. Phys. D: Appl . Phys . 38 (2005) 1498-1503; Jon Opsal et. al., J. Appl. Phys. 61(1), January 1, 1987; US 7.060.980). This technique has the drawing of requiring an important variation of reflectivity with temperature, and this depends on the wavelength of the measuring beam and changes from one material to the other. Another variation is to measure the deflection of a reflected beam on the sample that diverges from the heating beam. Deflection is a measure of the material deformation that depends on its expansion coefficient and temperature increase (Jon Opsal et. al . , Applied Optics Vol. 22 No. 20, 15 October 1983; Allan Rosencwaig et al, Appl. Phys. Lett. 46(11), June 1, 1985, Published US Patent Application 2004/0188602) . This method requires a very stable and precise alignment of both beams, as the amplitude and phase of the reflected signal vary with the relative position of the beams as reported by 0. E. Martinez et. al., Appl. Phys. B 90, 69-77 (2008). Alternatively, modulated dilation can be measured by an optical interferometer (US 6.756.591 Bl and 6.965.434 B2). This technique requires the highest mechanical stability with vibrations lower than a micrometer tenth.

The ratio between the dilation of the material due to periodic heating and the properties of the tested material, as thermal diffusion, absorption coefficient, thermal expansion coefficient and calorific capacity is known and has been described in the literature (0. E. Martinez et. al . , Appl . Phys . B 90, 69-77 (2008)) so, once established the material expansion, it can be obtained based on said data.

When measuring opaque film thicknesses there is a reference (A. Rosenwaig et al . Appl. Phys. Lett. 43, 166 (1983)) wherein said thickness is measured based on the deflection photometric signal amplitude according to the theory presented for films with diverse thermal behavior but similar elastic behavior as per Opsal and Rosenwaig ( J. Opsal et al, Thermal wave depth profiling: Theory. J. Appl. Phys. 53, p.4240, 1982.). It has also been determined thickness measuring the infrared emission of a sample being heated by an electromagnetic pulse (Published US Patent Application 2002/0094580) . The ratio between heating and the measured effect is similar to the present invention that is different in the way of detecting material changes.

To determine particle adsorption spectrums and devices in general systems based on spectrometers wherein extinction coefficients (absorption plus dispersion) without distinguishing between these two phenomena are used. To measure light adsorption by small particles there is a recent development that allow measuring in case of particles immersed in a liquid (US 6.756.591 Bl and 6.965.434 B2). It is based on the use of photothermal methods with interferometers on the detection part. Said technique cannot be used when the particle is on a dry environment, deposited on a surface or in case of a nanometric or micrometric device such as those used in photonics or microelectronics.

The patent application publication (AR) N° 70.418 describes a technique to measure dilation based on the use of devices that determine the focus error. With a modulated heating beam and a measuring beam, the measuring system of focus error similar to the one used to focus CD reading systems allows to determine material dilation. This system differs from the present invention in that it requires a 4 quadrant detector to determine the out of focusing based on algebraic operations between said detectors. These operations must be performed at the modulation frequency of the heating beam or higher, and for a high space resolution (for example, for an aluminum sample in the limit of a commercial optical microscope the beam would have been modulated at more than 800MHz) . The present invention used a single detector that allows using common electronics in the field of optical communications for detecting at these high frequencies. Another substantial benefit of the present technology is that detection, heating and measuring lasers are automatically aligned by the same fiber optic, showing no inconveniencies in case of eventual dilation or accidental movement of the parts. In the above mentioned patent, even when a fiber optic to ensure alignment of both lasers is used, alignment of the four quadrant detector must be ensured and stabilized. US 7.230.708 describes a method for acquiring photothermal microscopic images based on detecting the protrusions generated by the local deformation produced by heating the modulated beam. In said patent, detection is performed by spatial dispersion of light due to the small protrusion by a sensor located out of axis, and not based on the disfocal measuring on axis of the reflected radiation of the measuring beam. In case of using a single detector in this configuration, protrusions induced by the heating beam cannot be distinguished from those derived from topography inhomogeneity of the sample. Said patent also proposes the use of multiple parallel detectors to sample different angles. This configuration has the benefit of requiring a very fast acquisition system, as if observed at microscopic ranges the heating beam must be modulated at very high frequencies (millions of Hertz) and therefore acquiring on each channel at even higher frequencies, which would be extremely expensive and voluminous if one also desires to increase the sensibility of the system using "lock-in" detection on each channel. The proposed configuration uses only one detection channel that allows working with only one detector and a single amplifier at very high frequencies. It also has to be noted for the purposes of this comparison what was herein mentioned about the sensibility of systems to alignment, as the proposed in US N° 7.230.708 should keep all detectors aligned. Confocal detection is a very popular microscopy technique for determining cross sections based on introducing a very small opening on the image plane of the sample to observe. Said opening has de property of mostly blocking the light from planes out of focus, providing in this manner only an image of the focal plane of the objective of the microscope used. This technique is very popular for detecting fluorescence in biology or to determine surface profiles on material. Confocal laser scanning microscopy, (See C.J.R. Sheppard and D. M. Shotton Bios Scientific Publishing Limited, 1997 ISBN 0387 91514 1) . In the present invention a scheme wherein the signal comes from an opening filtering but, conversely to the confocal system, it only appears when the opening is placed out of focus, therefore it becomes into an out of focus confocal detection or, more adequately, disfocal detection.

SUMMARY OF THE INVENTION

The present invention relates to a method and apparatus for measuring the protrusion generated on the surface of a sample due to the thermal dilation produced by heating with an electromagnetic radiation beam and based on said result determining the thermal diffusion, optical absorption, thermal dilation or other parameters derived from the same. The method is based on measuring the curvature of the surface that occurs due to the thermal dilation based on the transmission of a measuring beam reflected on the protrusion and through a focusing optical system and one opening located out of focus. The protrusion changes the focus position and therefore transmission through the opening, that being out of focus allows quantifying the deformation magnitude.

DESCRIPTION OF THE DRAWINGS

In order that the present invention is clearly understood and easily embodied, a preferred embodiment of the same is shown in the illustrative and non-limiting drawings that follow, wherein:

Figures 1A and IB are descriptive schemes of the method. Figure 1A shows the distribution of components before starting the heating beam. Figure IB shows the effect of heating on focusing the measuring beam.

Figure 2 is a scheme of a possible embodiment of the apparatus for measuring the thermal properties of opaque surfaces according to the present invention.

Figure 3 are graphics of the amplitude of the heating beam and the signal based on time indicating the characterizing parameters .

Figure 4 are two graphics of the amplitude of the signal (A) and the phase of the signal (B) based on the modulation frequency divided by the critical frequency.

Figure 5 is a graphic of the amplitude of the signal based on the out of focusing in relation with the confocal configuration .

Figure 6 is a graphic of the phase of the signal based on the frequency when the substrate is covered by a layer of another material. The different graphics correspond to different material thicknesses. Figure 7 shows the preferred embodiment of the apparatus using fiber optics to combine the heating and measuring lasers and to configure the opening of the disfocal detection.

Figure 8 is a graphic with the results obtained on the amplitude of the signal with the method of the invention when testing a sample of fused quartz with a 30nm chromium coating.

Figure 9 is a graphic with the results obtained on the phase delay of the signal in the assay of Figure 8.

Figure 10 is a graphic that shows how diffusion data is recovered based on a numeric adjustment according to the data of Figures 8 and 9.

DETAILED DESCRIPTION OF THE INVENTION

The method of the present invention measures the protrusion generated on the surface of a sample due to the thermal dilation produced by heating with an electromagnetic radiation beam and based on said measuring determining the thermal diffusion, optical absorption, thermal dilation or others parameters derived thereof. Figure 1 shows the method. Figure 1A shows a possible distribution of the components, wherein the sample to be analyzed 40 is illuminated by a laser or other measuring or detection electromagnetic radiation beam 10 that is focused on the surface of the sample by an adequate optics, in this case a beam divider 33 and one lens 32 located at a distance L2 of the sample, though it can be any suitable combination of optical elements that perform this function. The beam is reflected on the surface of the sample and directed through a collection optical system that in this view is comprised by lenses 32 and 31 and is focused on a plane 60 at a distance L0 of the focusing system. Lines 11 indicate the path of two rays representative of said measuring or detection beam. An opening of dimensions similar to the ones of the focused measuring beam is located at a distance LI of the lens. Said distance is different (longer or shorter) from the distance L0 corresponding to is confocal location, so that part of the energy carried by the reflected beam is obstructed on said opening. A detector 52 located behind the opening determines the transmitted power that will be maximum power in case the opening is located on the plane 60 confocally. Location out of focus is an essential part of this method of disfocal detection. The method is completed when starting a heating beam 9 as shown in Figure IB and producing a protrusion of a radius R by thermal dilation on the sample. The heating beam should be centered in relation with the measuring beam. The protrusion is spontaneously produced as heating is inhomogeneous . On the center of the beam the intensity is higher and therefore heating is higher and consequently dilation is higher.

The side projection of the protrusion depends on the size of the beam and modulation frequency, as when modulating at a low frequency, heat has more time to laterally propagate before the beam is turned off. Therefore the dimensions and delay to produce the protrusion depend on the modulation frequency, the size of the beam and the thermal diffusion of the sample under study. The measuring beam reflected is out focused by this protrusion and remains focused on a plane different from the original plane 60 as indicated by the representative rays 12. If the opening was behind the plane 60 (L2 higher than L0) the signal transmitted by the opening 50 and captured by the detector 52 increases. Conversely, if the opening was located in front (L2 lower than L0) a signal reduces. This signal variation allows measuring the curvature acquired by the surface based on solving the propagation of beams by optical systems conventionally.

If the located heating means, preferably a laser, is modulated in time at a controlled frequency, it produces the periodic dilation of the material at the modulation frequency. Said dilation conventionally depends on the power absorbed, spatial dimensions of the heating means and the frequency of heating as described in the state of the ar. Said dilation produces a protrusion on the heated zone, as it is higher at the center heating beam than at the periphery of the same. At the time of the maximum dilation of the material, the maximum departure of the focus plate rearward will take place (away from the original image plane) . If the opening was retracted rearward, material dilation increases the transmission through the opening. When the material shrinks, transmission decreases. If the opening is moved on the opposite direction, closer to the sample, the transmission variation reverts, increasing when shrinking and decreasing when dilating. The method of the present invention is based on this signal produced by the disfocal location of the opening. Figure 2 shows a possible embodiment of the present invention. Through a fiber coupling system 19 two lasers are injected in a fiber optic 20. Both lasers are in this embodiment collinearly at the outlet of the fiber. An optical system 300 produces an image of the fiber outlet on a plane very close to the surface of the sample. This scheme shows a possible embodiment of said optical system 300 consisting of two lenses, a lens 31 that collimates the outlet beam of the fiber and another beam 32 that focuses it close to the surface of the sample. Distances LI of the fiber to the optical system and L2 of the optical system 300 to the sample 40 allow adjusting the level of out of focusing of the system. One of the beams acting on the sample, known as heating beam, produces a protrusion by thermal dilation with a curvature radius R. The second beam, known as testing, preferably of a wavelength different from that of the heating beam, is reflected on the surface of the protrusion and returns to the fiber but out focused in relation with the beam reflected on the undeformed surface. Lines 11 indicate that return path for two representative rays of said measuring beam reflected on the sample before dilating (without heating) that after covering the optical system 300 is focused close but not on the same surface of the fiber outlet 20. Due to the out of focusing the measuring beam is not optimally coupled on the fiber. Dotted lines 12 indicate the return path of the beam in the presence of a protrusion produces by heating, that due to the out of focusing produced by the protrusion and the out of focusing of the original alignment provides a change on the coupling magnitude on the fiber. In the case shown coupling increases, and the amplitude of said coupling depends on the level of out of focusing and the curvature radius R of the protrusion. The portion reinjected on the fiber of the measuring beam emerge the other end 50 and is detected by a detector 52 after filtered by a filter 51 that prevents the heating beam reinjected remaining portion passing. Changes on the curvature radius of the protrusion produce changes on the amplitude detected by 52.

If the heating beam is modulated at a frequency f, the protrusion will increase and decrease at the rate of heating, oscillating at the same frequency f. However, due to the thermal inertia of the system, there will be a delay between said oscillation and this delay will be evidenced on a delay of the oscillation detected by the detector 52. Figure 3 shows a temporal sequence on the graphic A representing the oscillation of the heating beam (or a component at a frequency f if heating is periodic but not harmonic) . The cycle period of heating beam is

T=l/f

Graphic B of Figure 3 shows the cycle followed by the amplitude collected by detector 52. Peak to peak amplitude will depend on the modulation frequency, the size of the heating beam, the intensity of the heating beam and the measuring beam and the properties of the material including thermal diffusion, thermal expansion coefficient, density, calorific capacity and optical reflectivity at the heating and testing wavelength. The signal depends on the frequency only through a function which amplitude is shown in Figure 4A. The critical frequency f ₀ is expressed as follows:

. . , _ thermal diffusion

Critical frequency =

beam area

The delay Dt (figure 3) between the signal of the detector and the heating cycle or, equivalently, its phase delay φ according to

cp=2* 7i;*Dt*f

will only depend on the modulation frequency, the size of the heating beam on the sample and the thermal diffusion of the material. The graphic of Figure 4B shows the way said phase depends on these parameters, therefore it is possible to determine the diffusion based on the measured phase, the modulation frequency and the heating beam area.

The signal amplitude depends on the level of initial out of focusing as shown in Figure 5 that shows the amplitude of the signal (in arbitrary units) based on the out of focusing related to the confocal position measured in units of focus depth (LI if the fiber moves or L2 if the sample moves) . There is an out of focus displacement LI or L2 that maximizes the signal and consequently provides the best embodiment of the system. On the contrary, it is noted that the signal amplitude is null in case of confocal configuration.

The above expression is valid in case of a homogeneous sample. In more complex cases, it is necessary to solve the heat diffusion equation to determine the thermal parameter values better suited to the experimental result. Such a case is the presence of a layer of thickness d on a given substrate. Figure 6 shows the phase of the measured signal based on the frequency divided by the critical frequency of the substrate for different layer thicknesses measured in units of the radius of the heating beam. It is clearly observed in this case how the phase measuring at different frequencies allows to recover the thickness data of the layer included the thermal properties of said layer.

Figure 7 shows the best complete embodiment of the system including the combination system of two lasers within the fiber. A heating laser 9 and a measuring laser 10 coupled to respective fibers are combined by means of a coupling multiplexor or 2x2 coupler 14 similar to those used on optical communication systems but adapted to the wavelengths of the lasers used. Beam samples can be taken with couplers 15 to monitor operation. A circulator 16 or a 1x2 coupler takes the return on the fiber and sends it to a detector 52 connected to a lock-in amplifier 53 or other electronic filter. Saud lock-in amplifier or filter detects the component of the signal at the desired frequency to determine the phase and consequently the parameters of interest as previously described. Lasers 9 and 10 are controlled by the supplies 61 that are modulated by function generators 62 at different frequencies fl and f2. Alternatively they can be externally modulated with modulators. A fiber optic 20 connects the lasers to the optical system 300 comprising in this case a microscope adaptor system 310 that generates an image of the fiber outlet on an inlet plane of the microscope such as, for example, the camera port, and a microscope 320 that generates a new image of the fiber close to the plane of the sample as shown in Figure 1. Sample 40 is scanned transversally with a microscope motorized slide 330. The whole assembly is controlled by a computer 500 that registers the signal obtained with the lock-in amplifier 53, the modulation frequency f and the position of the sample to compose the desired phase information based on the frequency and position to obtain a complete map of the properties measured on the sample. Alternatively the beam can be scanned on the sample instead of moving the sample by an optical scanner using conventional techniques in the field of scattering microscopy.

Figure 8 show the results of a test conducted with the configuration described by Figure 7. It shows the amplitude of the signal based on the frequency (at a range from 50Hz to 20.000Hz) and the out of focusing on the sample (with +/- 6mm displacements) . The sample used is a fused quartz substrate covered with a thin layer of chromium of a thickness of 30nm. The layer acts as a heat absorbent, transmitted to the substrate without disturbing the measurement thanks to its extremely thin thickness (more than 100 times smaller than the size of the beam) . The heating beam used had a wavelength of 980nm and a power of lOmW. El measuring beam is of 1.550nm of wavelength and lOmW of power. The size of the beam on the sample is of 20 micrometers. The signal is measured with an infrared photodiode typical of optical communications and a lock-in amplifier Stanford Research model 830DSP. Figure 9 shows the phase delay for these conditions. It is observed the phase decay φ with the modulation frequency f and, also, the existence of optimal positions for the out of focusing. In fact, the signal undergoes a minimum in the condition of confocal focus (z=0mm in figures 8 and 9) and has a maximum with an out of focusing of about of 2mm, which is consistent with what was expected according to Figure 5.

To compute the thermal diffusion it is enough to measure the signal based on the frequency for a fixed out of focusing (it is preferred but not necessary to use the optimum that is the one providing the highest signal) . From one of said curves the parameter "critical frequency" is adjusted in order to obtain the best adjustment of the theoretical curve illustrated in Figure 4. Figure 10 shows one of these adjustments that for the sample of fused quartz resulted in:

Critical frequency = 635Hz

And based on a measurement of the beam size (for example with ca CCD camera) diffusion D is obtained as follows:

D = critical frequency x area of the beam = 0,88mm ² /s

The upper curve of Figure 10 corresponds to the error obtained in the adjustment based on the assumed critical frequency, indicating the value of the critical frequency that minimizes the error. The other two curves are the amplitudes and phases measures based on the frequency and their respective adjustments for the optimum value of critical frequency. In an alternative embodiment of the method, instead of using a fiber optic to spatially filter the beam reflected, a small opening is used as indicated in the description of the method (Figure 1) .

If on a substrate of known thermal properties surface or embedded particles of much reduced sizes are added (micrometric or nanometric range) and in case said particles absorb the radiation of the heating beam, they will produce heating and consequent dilation of the substrate, that can be measured by the method described. In this way, by varying the incident wavelength of the heating beam and measuring the deformation of the surface it is possible to determine the particle absorption spectrum.