How to normalize heat detector measurements

Abstract

To normalize measurement values of a thermal detector that measures proximity between the thermal detector and a specimen.SOLUTION: A method includes a step of preparing a prove 10 having a probe body 12 and thereby defining a flat surface supporting a thermal detector. A detector surface of the thermal detector supports an electrical conductor having temperature-dependent electrical resistance. The method further includes the steps of: injecting a current into the electrical conductor; preparing an electronic circuit, connecting the electronic circuit to the electrical conductor, measuring electrical resistance by the electronic circuit, and providing output; and performing a mathematical normalization of the output by a first geometric dimension of the detector geometry, such that the output is normalized on the basis of the detector geometry.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to a method for testing electrical properties of a test specimen by contacting the probe with the test specimen, and is particularly directed to a microcantilever multipoint probe with a thermal detector. The present invention further relates to a system for testing electrical properties of a test specimen, a method for rescaling thermal detector measurements to compensate for geometric variations, and a method for determining thermal detector-test specimen proximity. [Background technology]

[0002] Some test specimens consist of a semiconductor wafer on which a thin planar continuous conductive film or a thin multi-layer stack is formed, for example, into a magnetic tunnel junction (MTJ).

[0003] Some specimens are semiconductor silicon wafers, with CMOS integrated circuits implemented, for example, with a large number of CMOS transistors. In this case, the specimens incorporate individual test pads where portions of the conductive film or stack under test can be dedicated for electrical characterization. These test pads may consist of dense arrays of finFET transistors.

[0004] Multipoint probe measurement and testing routines, such as four-point probe measurements, are also known as four-terminal sensing; electrical impedance measurement techniques use separate pairs of current-carrying and voltage-sensing electrodes (contact probes with pointed tips for contacting the specimen under test).

[0005] In performing resistance measurements to determine electrical properties of a test specimen, a test probe having one or more electrodes for establishing electrical contact with the test surface is brought into contact with the test surface.

[0006] An example of a micro four-point probe is disclosed in U.S. Patent No. 6,393,633, the disclosure of which is incorporated herein by reference. The probe disclosed in U.S. Patent No. 6,393,633 includes four cantilevers extending from the probe body.

[0007] The probe is part of a test device or system and has an actuator that can move the probe into contact with a test specimen placed within the test device.

[0008] Because the dimensions are on the μm or nm scale, it is important to have a controlled and precise landing of the tip (the tip of the contact probe) on the test specimen, so that the tip is at least within 200 nm, and preferably within ±50 nm, of the surface of the test specimen when the actuator stops.

[0009] Controlled landing will reduce electrode damage, reduce contamination of the test specimen with probe debris, and establish good, stable ohmic contacts.

[0010] When controlling the landing of the probe, a strain gauge type detector (mechanical contact detection) may be used, or electrical contact detection by measuring the electrical impedance between multiple contact probes may be used; for example, if the impedance between multiple contact probes decreases, it can be assumed that the contact probes have come into contact with the test specimen and that a current has flowed within the test specimen between at least two of the contact probes.

[0011] Each of these specimen landing or contact methods has its drawbacks; for example, a relatively large total footprint, which is due to the fact that in the case of strain gauge-based surface detection, many more cantilevers need to be in contact with the specimen surface compared to electrical detection. The larger footprint means that the specimen is more susceptible to contamination due to contact with the probe. On the other hand, electrical detection does not work on non-conductive (or oxidized) surfaces of the specimen, limiting its applicability.

[0012] An example of using heat to measure distance across an air gap is disclosed in U.S. Patent No. 6,393,633, which is incorporated herein by reference. However, the device in U.S. Patent No. 6,393,633 does not have a contact probe for measuring the electrical properties of a test specimen, and does not contact the test specimen itself. [Prior art documents] [Patent documents]

[0013] [Patent Document 1] European Patent No. 2293086 [Patent Document 2] U.S. Patent No. 7,186,019 Summary of the Invention [Problem to be solved by the invention]

[0014] OBJECTS AND SUMMARY OF THE PRESENT EMBODIMENT An object of the present invention is to alleviate at least some of the above-mentioned disadvantages. [Means for solving the problem]

[0015] As will become apparent from the description of the invention, the above objects and advantages, as well as many other objects and advantages, are achieved in accordance with a first aspect of the invention by the following means.

[0016] 1. A method for approaching a probe to a test specimen for testing an electrical characteristic of the test specimen, comprising: providing a probe having a probe body defining a flat surface for supporting a first cantilever and a first thermal detector; a first cantilever extends from the probe body in a coplanar relationship with the planar surface between a first proximal end supported by the planar surface and a first distal end opposite the first proximal end, the first cantilever having a cantilever surface supporting a first contact probe; a first thermal detector extending from the probe body in coplanar relation to the planar surface between a second proximal end supported by the planar surface and a second distal end opposite the second proximal end, the first thermal detector having a detector surface supporting an electrical conductor having a temperature dependent electrical resistance; Furthermore, establishing a temperature differential between the electrical conductor and the test specimen; injecting a current into the electrical conductor; providing an electronic circuit and connecting the electronic circuit to the electrical conductor and measuring the temperature dependent electrical resistance with the electronic circuit while moving a probe towards the test specimen; deactivating the probe when the temperature dependent electrical resistance reaches a first threshold indicative of approach or contact of the contact probe with the test specimen; It is a method having the following structure.

[0017] By close proximity, it is understood that the contact probe comes so close to the test specimen that it nearly hits it.

[0018] The threshold value may be selected based on empirical measurements so that the probe stops when it reaches a distance of 0 to 10 μm from the surface of the test specimen.

[0019] The probe is then moved a fixed distance towards the specimen and brought into physical contact with the specimen, which may be selected from the ranges of 0.5-8 μm, 0.5-6 μm, 0.5-4 μm, 0.5-2 μm, 1-7 μm, or 1-3 μm; depending on how far from the specimen the probe is stopped.

[0020] Alternatively, the threshold may be selected such that the probe contacts the test specimen when the threshold is reached, in which case it is not necessary to move the probe over the final distance to contact the test specimen.

[0021] Typically, contact occurs when the probe lands on the specimen and the tip of the contact probe touches the surface of the specimen.

[0022] However, in general, contact is not necessarily understood to mean that the contact probe is in physical contact with the test specimen; rather, it may be understood to mean that the contact probe is so close to the test specimen that capacitive coupling exists between the contact probe and the test specimen, such that the AC current in the contact probe can flow into the test specimen, i.e., the capacitive coupling is greater than 0 and less than 1 mF.

[0023] The probe may have a second cantilever extending from the probe body in coplanar relationship to the planar surface.The probe may have a second thermal detector extending from the probe body in coplanar relationship to the planar surface.

[0024] The temperature dependent electrical resistance may be measured negatively by measuring the voltage drop across the temperature dependent electrical resistance.

[0025] The first thermal detector may define a loop about the probe body; i.e., the first thermal detector may extend away from the probe body at a first point and return to the probe body at a second point (the thermal detector begins and ends on the probe body).

[0026] The temperature difference may be established by providing a heater to heat the conductor relative to the test specimen, i.e., by making the temperature of the conductor higher than the temperature of the test specimen. The reverse may also be true; i.e., by providing a heater to heat the test specimen relative to the conductor.

[0027] Alternatively, a cooler may be provided to cool the conductor relative to the test specimen; alternatively, a cooler may be used to cool the test specimen relative to the conductor.

[0028] An electric current injected into the conductor may be used to heat the conductor; i.e., the magnitude of the electric current may be such that the conductor is heated to a temperature above room temperature. The electric current may be an alternating current.

[0029] The current may be injected into the conductor by a current source, or a constant current may be supplied from the current source to conduct through the conductor.

[0030] The first threshold may be selected such that the contact probe-to-test specimen distance is less than a second threshold.

[0031] The first threshold may be selected to be a function of a statistical distribution that defines the variability between measurements made by a plurality of probes, each of which is constructed in accordance with the present invention.

[0032] According to a second aspect of the present invention, the above objects and advantages are achieved by the following means.

[0033] 1. A system for testing electrical properties of a test specimen, comprising: a probe having a probe body defining a flat surface for supporting a first cantilever and a first thermal detector; a first cantilever extends from the probe body in a coplanar relationship with the planar surface between a first proximal end supported by the planar surface and a first distal end opposite the first proximal end, the first cantilever having a cantilever surface supporting a first contact probe; a first thermal detector extending from the probe body in coplanar relation to the planar surface between a second proximal end supported by the planar surface and a second distal end opposite the second proximal end, the first thermal detector having a detector surface supporting an electrical conductor having a temperature dependent electrical resistance; Furthermore, a heater or cooler for establishing a temperature difference between the conductor and the test specimen; a power source interconnected with the conductor and configured to inject a current into the temperature dependent conductor; an actuator for moving the probe towards the test specimen; an electronic circuit interconnected with the electrical conductors for measuring the temperature dependent electrical resistance and comparing the temperature dependent electrical resistance to a threshold value; The system includes:

[0034] According to a third aspect of the present invention, the above objects and advantages are achieved by the following means.

[0035] 1. A method for normalizing a measurement value of a thermal detector that measures proximity between a thermal detector and a test specimen, comprising: providing a probe having a probe body defining a flat surface for supporting a thermal detector; the thermal detector extends from the probe body in coplanar relationship to the planar surface between a proximal end supported by the planar surface and a distal end opposite the proximal end; a detector geometry defined by the thermal detector having a first geometric dimension, the thermal detector having a detector surface supporting an electrical conductor having a temperature dependent electrical resistance; Furthermore, injecting a current into the electrical conductor; providing an electronic circuit and connecting the electronic circuit to the electrical conductor, the electronic circuit measuring the electrical resistance and providing an output; performing a mathematical normalization of the output with respect to a first geometric dimension such that the output is normalized with respect to the detector geometry; It is a method having the following structure.

[0036] The first geometric dimension may be represented by a measure of the temperature dependent electrical resistance at room temperature, i.e., by measuring the voltage drop across the electrical conductor and the current conducted through the conductor; when the current is held constant, the voltage drop is directly proportional to the electrical resistance.

[0037] The electrical resistance measurement may be normalized using a second geometric dimension of the thermal detector.

[0038] The second geometric dimension may be expressed as a measure of the difference between the temperature dependent electrical resistance at room temperature and its temperature dependent electrical resistance at a temperature above room temperature.

[0039] Electrical resistance measurements may be normalized using a third geometric dimension of the thermal detector, which is comprised of a thickness of the electrical conductor.

[0040] According to a third aspect of the present invention, the above objects and advantages are achieved by the following means.

[0041] A method for determining proximity between a thermal detector and a test specimen, comprising: providing a probe having a probe body defining a flat surface for supporting a thermal detector; the thermal detector extends from the probe body in coplanar relationship to the planar surface between a proximal end supported by the planar surface and a distal end opposite the proximal end; a detector geometry defined by the thermal detector having a first geometric dimension, the thermal detector having a detector surface supporting an electrical conductor having a temperature dependent electrical resistance; Furthermore, establishing a temperature differential between the electrical conductor and the test specimen; injecting an alternating current having a first frequency into the electrical conductor; providing an electronic circuit and connecting the electronic circuit to the electrical conductor, the electronic circuit measuring the electrical resistance while moving the probe towards the test specimen and providing an output; providing a filter having a cutoff frequency below a second harmonic of the first frequency and filtering the output with the high pass filter to provide a filtered signal; using the filtered signal as a measure of thermal detector-to-sample proximity; It is a method having the following structure.

[0042] The invention will now be explained in more detail by way of example of embodiments with reference to the very schematic drawings in which: [Brief description of the drawings]

[0043] [Figure 1] FIG. 2 is a perspective view of a probe for testing electrical properties of a test specimen. [Diagram 2] FIG. 2 is an enlarged view of part I in FIG. [Diagram 3] FIG. 13 shows an alternative embodiment in which a heat detector is provided with protrusions. [Figure 4] 13A-13C show alternative embodiments of the probe. [Diagram 5] FIG. 13 shows an alternative embodiment in which cantilevers are distributed on opposite sides of a single thermal detector. [Figure 6] FIG. 13 shows an alternative embodiment using a single cantilever. [Figure 7] 13A-13C show alternative embodiments of the probe. [Figure 8] FIG. 13 is a diagram showing the measurement results using five probes. [Figure 9] FIG. 9 is a diagram showing normalization results of the measurement results shown in FIG. 8. [Figure 10] FIG. 13 is a diagram showing the results of rescaling the measured electrical resistance of a conductor. [Figure 11] FIG. 11 is an enlarged view of the graph within the circled area II in FIG. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0044] The present invention will now be described more completely with reference to the accompanying drawings showing exemplary embodiments of the present invention. However, the present invention may be embodied in other forms and should not be construed as being limited to the embodiments described herein. Rather, these embodiments are provided to make the present disclosure consistent and complete, and to fully convey the technical scope of the present invention to those skilled in the art. Similar reference characters refer to similar elements throughout the text. Therefore, similar elements will not be described in detail in relation to the description of each drawing.

[0045] FIG. 1 shows a perspective view of the probe for testing the electrical characteristics of a test specimen.

[0046] The probe 10 includes a probe body 12, which may be fabricated from multiple semiconductor, metal, and / or dielectric layers. A first side of the probe body defines a substantially planar body surface; that is, as flat as the manufacturing process will allow. The probe body has a second side opposite the first side, and a front side between the first and second sides defining a front surface 32.

[0047] FIG. 2 shows an enlarged external view of part I in FIG.

[0048] The probe 10 comprises a cantilever 18. Seven further cantilevers are visible in Figure 2, and typically a probe can be equipped with a cantilever group consisting of 1-16 cantilevers.

[0049] Each cantilever carries a contact probe 24 that contacts a point area or small area on the surface of the test specimen; the contact establishes an electrical connection to that specimen.

[0050] Before measuring the electrical properties of a specimen, it may be necessary to burn off the oxide layer on the specimen. This can be done by applying a breakdown voltage (between the two contact probes) to the current that flows through the contact probes; a breakdown voltage is a voltage high enough to cause a breakdown of the dielectric oxide, e.g., greater than 5V.

[0051] The contact probes or electrodes consist of a conductive metal film deposited on the surface of the cantilever that faces the surface of the specimen under test when the test is performed. Usually, the metal film covers the entire width of the cantilever; this is because the film is deposited from above, whereas if it were deposited without a pattern, the film would cover the landings.

[0052] The cantilevers are made of a dielectric material to electrically insulate the contact probe from the probe body; this prevents current from flowing from one contact probe to another, even if the probe has multiple contact probes.

[0053] Although the probe shown in Figures 1 and 2 is a micro 8-point probe, in general a probe may have any number of contact probes; i.e., a probe may have four contact probes, thereby constituting a micro 4-point probe, or twelve contact probes, thereby constituting a micro 12-point probe.

[0054] Each cantilever may support multiple contact probes.

[0055] Each cantilever has an L-shape and is therefore more flexible in three dimensions than, for example, a straight cantilever.

[0056] The distal end 30 of each cantilever (the tip constituting the tip) is located farthest from the probe body 12, and therefore there is a certain distance between the distal end and the probe body. The distance between the distal end and the probe body is typically within the range of 5 to 200 μm.

[0057] The proximal end of the cantilever is defined as the opposite or furthest from the distal end, i.e., closer to the probe body than the distal end.

[0058] Each cantilever is supported at its proximal end by a flat body surface and extends away from the probe body with a vector component along a direction perpendicular to the front surface 32 of the probe body 12 and along a cantilever plane that is parallel to (coplanar with) the flat surface of the probe body. When the cantilever contacts the specimen 22, it can be deflected out of the cantilever plane.

[0059] The axis extending through the proximal and distal ends 30 constitutes the cantilever axis. In Figure 2, the cantilever axis is at an angle greater than zero with respect to the normal to the front surface 32 of the probe body (in part due to the L-shape of the cantilever). The cantilever may also be straight, in which case the cantilever axis would extend parallel to the front surface normal.

[0060] The probe is equipped with a first thermal detector; this is visible to the left of the cantilever and extends parallel to it.

[0061] The thermal detector extends along a detector plane that is parallel to the planar surface (and coplanar with the planar body surface).

[0062] The thermal detector extends away from the front surface 32 in the same manner as the cantilever. In Figure 2, the thermal detector extends parallel to the normal to the front surface 32. However, the thermal detector may be at an angle within ±60° to the normal to the front surface 32 in the detector plane.

[0063] The heat detector is looped around the probe body; that is, it begins and ends on the probe body.

[0064] In Figure 2, the heat detector is formed by three straight segments: a first arm 34, a second arm 36 and a front segment 38. Two arms are perpendicular to the front surface 32 and the front segment is parallel to the front surface.

[0065] The proximal end of the first arm is supported by the planar surface, the proximal end of the second arm is supported by the planar surface, and the front segment 38 extends between the distal end of the first arm and the distal end of the second arm.

[0066] More than three segments may be used to form the loop; in a first alternative, the loop may be formed in a ring, thereby defining a geometry such as a circle, ellipse, etc. (but not spanning a full 360°). In a second alternative, the heat detector may be formed from a flat plate, the width and length of which may be at least as great as the coverage of the three straight segments.

[0067] The distal end of the thermal detector (that constitutes the detector tip) is at an anchor position with an anchor distance α relative to the probe body 12. The anchor distance (the length of the thermal detector) is illustratively smaller than the distance between the distal end of the cantilever and the probe body.

[0068] By way of consideration, the thermal detector may extend as far from the probe body as the cantilever 18; that is, the length of the thermal detector may be approximately the same as the length of the cantilever (effectively meaning that the length of the thermal detector does not deviate from the length of the cantilever by more than 10%).

[0069] The same considerations may allow the length of the thermal detector to exceed the length of the cantilever.

[0070] The anchor distance α may be less than 200 μm, for example, less than 180 μm, less than 160 μm, less than 140 μm, less than 120 μm, less than 100 μm, or even less than 50 μm.

[0071] The front segment 38 has an illustrative line width β, which may be in the range of 20 nm to 20 μm, for example 100 nm to 10 μm or 1 μm to 5 μm, and all segments of the thermal detector may be in this width range.

[0072] The thermal detector supports an electrical conductor 28, which is made up of a conductive metal film deposited on the surface of the thermal detector. The metal film typically covers the entire width of the thermal detector; that is, the width of the conductor (e.g., on the front segment) can correspond to the line width β.

[0073] This metal film may be made of the same material as the contact probe, such as nickel, tantalum, aluminum, platinum, ruthenium, titanium, cobalt, iron, tungsten, etc. The conductor has an electrical resistance, which is temperature dependent.

[0074] Each contact probe (metal film) is connected to a corresponding conductive line on the probe body 12. The conductive lines connect the contact probes to contact pads; i.e., the contact probe 24 on the first cantilever 18 is electrically connected to the first conductive line 16, which leads to the first contact pad 14 (the contact probes and conductive lines may be deposited as a continuous metal film that runs from the contact pad to the distal end of the cantilever).

[0075] Similarly, the electrical conductor 28 may be formed by depositing a metal film on the thermal detector that runs from a first contact pad on the probe body and back to a second contact pad; that is, the electrical conductor may be formed such that a loop involving the first contact pad on the probe body and the second contact pad on the probe body can be used as an electrical resistor in an electrical circuit having two terminals connected to the first and second contact pads.

[0076] The contact pads are for establishing an electrical connection between the probe and the test fixture; i.e., terminals on the test fixture contact the contact pads on the probe when the probe is placed in the test fixture. The test fixture also includes a power source, e.g., a current or voltage source, and one or more electronic circuits that transmit signals onto the thermal detector conductors and the contact probe. The electronic circuits may be implemented with one or more microprocessors.

[0077] An actuator in the test fixture holds the probe and moves it relative to a test specimen that is also placed in the test fixture.

[0078] During the measurement with the probe, an alternating current with a constant amplitude is injected into the conductor 28 by a current source connected to the conductor; for example, the current may be a sinusoidal wave with a fundamental frequency (first harmonic) selected in the range 1 Hz to 10 kHz, for example 1 kHz, and with a certain phase. Alternatively, the current may be a direct current.

[0079] Instead of a current source, a voltage source may be connected across the conductor, thereby causing a current to flow through the conductor.

[0080] This current is intended to heat the conductor 28 to a temperature above room temperature, and the heating can be accomplished with a current having an amplitude selected within the range of 10 μA to 10,000 μA, for example 2000 μA; i.e., the power delivered to the electrical resistance of the conductor can be dissipated as heat.

[0081] Room temperature in the context of this application refers to a temperature between 15°C and 30°C, for example between 20°C and 30°C, between 20°C and 25°C or between 25°C and 30°C.

[0082] Alternatively, the test specimen may be heated to a temperature higher than that of the electrical conductor.

[0083] Near room temperature, the resistivity of a metal normally increases with increasing temperature. If the temperature T does not vary excessively, a linear approximation is usually used: R(T) = R0(1 + c(T-T0)), where c is the temperature coefficient of resistance, T0 is a fixed reference temperature (usually room temperature), and R0 is the resistance at temperature T0.

[0084] The probe is then moved towards the test specimen; this is initially at room temperature.

[0085] The probe may be moved towards the specimen at an angle to the specimen surface; that is, the probe may be moved at an angle of ±90° to the normal to the specimen surface. Alternatively, the probe may be moved at an angle of 45°±40° to the specimen surface (e.g., 45°±30° or 45°±15°) and not moved along the specimen but parallel to the specimen surface.

[0086] When the probe reaches the specimen, the dissipation of heat generated in the electrical conductor changes. When the heat detector is far from the specimen surface, heat is mainly dissipated into the probe body through thermal conduction along the heat detector structure. The closer the heat detector is to the specimen surface, the greater the proportion of the generated heat is dissipated through the air into the specimen itself. The specimen can be considered a constant temperature heat sink.

[0087] By observation, as the probe approaches the test specimen, the temperature of the conductor decreases (the current injected into the conductor or the voltage applied to the conductor remains constant), which means that the electrical resistance of the conductor decreases.

[0088] Using electronic circuitry in the test equipment, the electrical resistance of the conductor can be measured as the probe moves towards the test specimen using a Wheatstone bridge with the resistance of the conductor being one of four resistances used in the bridge.

[0089] Alternatively, a voltmeter may be used to measure the voltage drop across the conductor (when the current is held constant, the voltage measurement is a measure of the electrical resistance).

[0090] If a constant voltage is applied to the conductor using a voltage source, the measured current in the conductor is a measure of the electrical resistance of the conductor 28.

[0091] The measure of electrical resistance (the output provided by a measuring circuit included in the electronic circuit) is used as a measure of how close the probe is to the test specimen; it is compared to a threshold value τ. When the measure of electrical resistance reaches the threshold value, the probe is stopped.

[0092] Alternatively, an equivalent measure, such as a change in a measure of electrical resistance, or a change in voltage drop, may be used as a measure of how close the probe is to the test specimen.

[0093] The threshold value is selected so that the contact probe is near the test specimen when the threshold value is reached and therefore only needs to move a fixed distance the remainder of its travel towards the test specimen.

[0094] The threshold value may alternatively be selected such that when the threshold value is reached the contact probe contacts the test specimen.

[0095] When the contact probe contacts the test specimen, an electrical connection will be established between the contact probe and the test specimen.

[0096] The thermal detector-to-sample surface distance should be reproducible to within 300 nm, preferably within 50 nm, from landing to landing and from probe to probe.

[0097] The threshold may be determined through empirical measurements or by determining it as a function of the difference between the length of the thermal detector (anchor distance) and the length of the cantilever (the distance between the distal and proximal ends of the cantilever).

[0098] Alternatively, the test specimen may be heated while the current in the conductor is kept at a value, i.e., a value that does not heat the conductor so much that the test specimen is heated; i.e., the current may be, for example, 100 μA or less.

[0099] As the probe is moved towards the test specimen, heat dissipated from the specimen will heat the conductor, causing the conductor's electrical resistance to rise, which can be measured by the electronic circuitry and used as a measure of how close the heat detector is to the test specimen;

[0100] By the same token, the test specimen may be cooled to a temperature below room temperature to provide a temperature differential between the thermal detector and the test specimen.

[0101] FIG. 3 shows an alternative embodiment in which the heat detector is provided with a protrusion.

[0102] The projection 40 extends along the length of the thermal detector, so that the distal end of the thermal detector (which now corresponds to the distal end of projection 40) extends as far from the probe body 12 as the cantilever; i.e., the distal end of the thermal detector is aligned with the distal end of the cantilever. In this way, the thermal detector and the cantilever contact the test specimen at approximately the same time; depending on whether the probe is tilted somewhat, the thermal detector can then be used as a strain gauge to determine the probe-to-test specimen distance.

[0103] Additionally or alternatively, the protrusions may allow the heat detector to be used as a contact probe; that is, the heat detector may have a dual or triple purpose.

[0104] An alternative embodiment of the probe is shown in FIG.

[0105] Because the probe includes a second thermal detector 42 in addition to the first thermal detector 26 and the cantilever, it is able to identify potential tilt of the probe, i.e. whether the tip (distal end) of the cantilever is parallel to the surface of the test specimen, thereby reducing the risk that one of the outer cantilevers will penetrate deeper into the test specimen than desired.

[0106] The cantilever is located between the first and second thermal detectors.

[0107] The second thermal detector supports a second electrical conductor in the same manner as the first thermal detector described above; that is, a second electrical conductor having a temperature dependent electrical resistance forms a loop involving a third contact pad on the probe body and a fourth contact pad on the probe body so that it can be used as an electrical resistor in an electrical circuit. The presence of the second thermal detector doubles the number of contact pads on the probe body.

[0108] Ideally, the two thermal detectors are made as similar as possible, ie, with the same anchor points / distances and line widths.

[0109] FIG. 5 shows an alternative embodiment in which the cantilevers are distributed on opposite sides of a single thermal detector.

[0110] FIG. 6 shows an alternative embodiment in which a single cantilever is used.

[0111] The single cantilever is constructed of an electrically insulating membrane 44, for example an oxide membrane, and supports a first contact probe 24. Seven other contact probes are shown next to the first contact probe. Membrane 44 is shown supporting electrical conductors 28 of the thermal detector, although electrical conductors 28 may be supported by a second membrane separate from the first membrane.

[0112] The first contact probe has a distal end that extends to the membrane edge 46, i.e., furthest from the probe body. The other contact probes also extend to the membrane edge.

[0113] The distal end of the heat detector 26 has a distance from the membrane edge 46, but may extend all the way to the membrane edge.

[0114] The membrane allows the contact probes to be deposited with smaller line widths; i.e., the cantilevers can be closer together without breaking each individual contact probe due to their narrower cantilever width. The distance between the contact probe tips along their first edges may be less than 1 μm.

[0115] The conductor 28 on the membrane 44 is shown as a loop surrounding the probe body 12 .

[0116] An alternative embodiment of the probe is shown in FIG.

[0117] Membrane 44 is supported by cantilevers 18. A total of eight cantilevers are used to support the membrane, as depicted in FIG.

[0118] The cantilever and membrane form a loop around the probe body making the structure more flexible compared to the probe shown in FIG. 6 where the membrane is supported by the probe body.

[0119] The first cantilever 18 supports a contact probe that extends onto the membrane all the way to the membrane edge 46 .

[0120] The two central cantilevers each support a metal film that extends onto and is integrated onto the membrane to form electrical conductors 28 that form a loop around the probe body. Electrical conductors 28 may have a temperature dependent electrical resistance and may be used as thermal detectors in the same manner as described above with respect to Figures 1-6.

[0121] The metal films are joined on the membrane at a distance from the membrane edge. From the joining point, a projection 40 extends towards the membrane edge parallel to the contact probe 24; i.e., the projection 40 is integrated into the conductor 28. The projection 40 may perform the function of a strain gauge and may be used to determine the probe-to-sample distance. In addition, the conductor 28 may be used as a contact probe together with other contact probes; i.e., this conductor has a triple purpose, as described in connection with FIG. 3.

[0122] The contact probes can be placed closer together because of membrane 44. Because conductors 28 also act as contact probes, the probe in Figure 7 has a total of seven contact probes, which run parallel and close together on the membrane (closer than they would be if the contact probes were supported by individual cantilevers).

[0123] Figure 8 shows the measurement results using five probes.

[0124] Each probe corresponds to one of the probes described in connection with any of Figures 2 to 7, for example the probe described in connection with Figure 2. The measurement routine is also carried out as described in connection with Figure 2.

[0125] The five probes have differences due to manufacturing tolerances; that is, the anchor position (anchor distance α from the distal end of the thermal detector to the probe body), the line width β of the conductor at the distal end of the thermal detector, and the thickness of the conductor 28 vary from probe to probe (the probe dimensions are in the μm range, which means that it is difficult to manufacture each probe with exactly the same anchor distance and line width).

[0126] The thermal detectors in probes 1, 2 and 3 have a conductor 28 made of Ni and 90 nm thick.

[0127] The thermal detectors provided in probes 4 and 5 have a conductor 28 made of Ni and 110 nm thick.

[0128] During the measurement, a conductor in the thermal detector is heated by an electric current as each probe is moved towards the test specimen. Each graph shows the measured electrical resistance as a function of the distance to the test specimen. As the probe approaches the test specimen, the electrical resistance decreases.

[0129] These five graphs are offset from each other due to the different geometric dimensions of the five probes, and the different sensitivity of each probe, which is depicted by the different slopes of the tangents (in Figure 8 the tangents are shown near the 0 m distance for probes 3 and 5).

[0130] Figure 9 shows the normalized results for each of the measurements shown in Figure 8. In this description, the words "normalize" and "rescaling" are used interchangeably.

[0131] Compared with the graph in FIG. 8, the graph in FIG. 9 has a smaller offset and a smaller difference in tangents.

[0132] During normalization, the electrical resistance measurements R(T), denoted as Ω in Fig. 8, are normalized with respect to the geometric dimensions of the detector geometry, i.e., the anchor distance and the line width. The electrical resistance measurements may also be normalized with a single parameter.

[0133] The offset between the graphs in FIG. 8 is a function of line width and metal (Ni in this case) film thickness, and the sensitivity is a function of anchor distance, line width and metal film thickness.

[0134] The linewidth of the thermal detector of each probe may be measured as a physical dimension in meters, or alternatively, may be expressed as an electrical resistance measured at room temperature with a current small enough not to measurably change the temperature of the thermal detector (the voltage drop across a conductor at room temperature, taken as value V0).

[0135] The anchor distance of the thermal detector of each probe may also be measured as a physical dimension in meters, or alternatively, may be expressed as the difference between the electrical resistance measured at room temperature and the electrical resistance measured at a temperature above room temperature (e.g. 5°C or 10°C above room temperature, depending on the current passing through the thermal detector).

[0136] For example, in a first measurement, a first current having a first amplitude (eg, 100 μA) is injected into the electrical conductor and the equivalent electrical resistance of the electrical conductor is measured.

[0137] In a second measurement, a second current having a second amplitude (e.g., 1000 μA) is injected into the conductor and the equivalent electrical resistance of the conductor is measured (the voltage drop across the conductor at the higher temperature, which we will call the value V T The second current causes the temperature of the conductor to increase above room temperature.

[0138] The difference between these two measurements is then calculated as ΔV = V0 - V T By using this, the anchor distance can be expressed.

[0139] Using the measurements V0 and ΔV, a measure of the electrical resistance of the conductor can be rescaled, for example, by dividing it by the square of each output. y=f(R,V0,V T )=R(T) / (V0 2 ΔV 2 )

[0140] The electrical resistance rescaling value y is a function of the electrical resistance measurement, the electrical resistance measured at room temperature, and the electrical resistance at temperatures above room temperature.

[0141] Alternatively, the electrical resistance rescaling value may be calculated using the difference between the electrical resistance measurement R(T) and R', where R' is the electrical resistance measured at a distance far enough away from the specimen that the electrical resistance of the conductor is not affected by the proximity of the specimen. R' in Figure 8 would be the electrical resistance obtained from a measurement made as far away from the specimen as possible, for example 100 μm.

[0142] In this case y=f(R,V0,V T )=(R(T)-R') / (V0 2 ΔV 2 ) It becomes.

[0143] The electrical resistance rescaling value y given above, i.e., that shown in FIG. 9, is a function of the difference between the electrical resistance measurements R(T) and R′, the electrical resistance measured at room temperature, and the electrical resistance at temperatures above room temperature.

[0144] Compared to the graphs in FIG. 8, this rescaling brings the measurements closer together; that is, the offset between the graphs in FIG. 9 is smaller, and the sensitivity differences between the five probes are also smaller.

[0145] FIG. 10 shows the rescaling results of the electrical resistance measurements of the conductor.

[0146] Compared to the rescaled graph in FIG. 9, the graph in FIG. 10 has an even smaller offset and a smaller difference in tangents.

[0147] During rescaling, the electrical resistance measurement is rescaled (the two are multiplied) by the thickness ω of the electrical conductor, i.e. the thickness of the metal film layer deposited on the thermal detector, which may be known from the deposition process or determined by electrical measurements.

[0148] The normalized electrical resistance value y1 of the conductor can be rescaled using the thickness ω by, for example, multiplying it by the square of the thickness. y1'=f(R,V0,V T ,ω)=R(T) / (V0 2 ΔV 2 )ω 2

[0149] That is, the electrical resistance rescaling value y1' is a function of the electrical resistance measurement, the electrical resistance measured at room temperature, the electrical resistance at temperatures above room temperature, and the thickness of the electrical conductor.

[0150] In another example of rescaling, the electrical resistance rescaled value y is rescaled by the square of the thickness ω of the electrical conductor, i.e. the thickness of the metal film layer deposited on the thermal detector (the two are multiplied together). y'=f(R,V0,V T )=(R(T)-R') / (V0 2 ΔV 2 )ω 2

[0151] That is, the electrical resistance rescaling value y' shown in FIG. 10 is a function of the difference between the measured electrical resistance and R', the electrical resistance measured at room temperature, the electrical resistance at temperatures above room temperature, and the thickness of the conductor.

[0152] Compared to the graphs in FIG. 9, this normalization brings the measurements closer together; that is, the offset between the graphs in FIG. 10 is smaller, and the sensitivity differences between the five probes are also smaller.

[0153] FIG. 11 shows an enlarged view of the graph within the circled area II in FIG.

[0154] Each probe is stopped when it reaches a threshold (a value chosen in the range of -1.000 to -2.000, in this case τ=-1.600).

[0155] As shown in the close-up, even with normalization, the measurements from the five probes are not necessarily identical, and the probes stop at different positions / distances to the sample. No matter what threshold is used to stop the probes, there will be a corresponding difference (error ε) between the probe that stops closest to the sample (probe 1) and the probe that stops furthest from the sample (probe 5). At a threshold τ=-1.600, the difference is 200 nm.

[0156] To determine what threshold the error ε must be below a certain (second) threshold, one can perform many measurements with many probes. Since it is not desirable to have the probe go too deep into the test specimen, some error (e.g., 200 nm) can be tolerated.

[0157] That is, the threshold value τ for determining when to stop the probe can be determined as a function of the error ε, for example, by empirical experiments using the five probes described above; that is, when the error ε is below a certain value (for example, 200 nm, 100 nm, or 50 nm), the electrical resistance of the conductor, i.e., the electrical resistance measured when the error is below that value, can be used as the threshold value for when to stop the probe in the actual measurement of the test specimen to be measured. In the empirical experiment shown in Figure 11, the threshold value determined is τ = -1.600 as described above.

[0158] Once the rescaling has been performed, it can be used on all thermal detectors with the same design, i.e. the same geometry, the same dimensions, e.g. anchor distance and line width within manufacturing tolerances, the same material, etc. If the thermal detector has protrusions, the geometry of the protrusions must be the same (within manufacturing tolerances).

[0159] Once a thermal detector has been calibrated, i.e. the thermal signal vs. distance relationship has been measured, all other thermal detectors of the same design can be calibrated with the benefit of this rescaling. The thermal signal is the measured resistance of the conductor R(T), an example of which is shown in FIG.

[0160] That is, and importantly, the calibration makes it possible to know the distance to the surface of the test specimen for any given thermal signal measured for all thermal detectors of that design.

[0161] As an alternative measurement procedure for the temperature dependent component of the electrical resistance of the electrical conductor, high order harmonic filtering is proposed.

[0162] In this method, the electrical resistance of the conductor 28 is measured by means of an alternating current injected into the conductor; that is, the current may be a sine wave having a fundamental frequency (first harmonic) selected in the range 1 Hz to 10 kHz, for example 1 kHz, and exhibiting a certain phase.

[0163] The electrical current can be used to heat the conductor, or by other means, to establish a temperature difference between the conductor and the test specimen, as described above.

[0164] The electrical resistance of the conductor is then measured using an electronic circuit, such as a Wheatstone bridge or a voltmeter.

[0165] The electrical resistance measurement is then filtered with a high-pass filter having a cutoff frequency lower than the second harmonic of the first harmonic; i.e., the first harmonic is suppressed relative to the higher harmonics. Alternatively, the electrical resistance measurement can be filtered with a band-pass filter centered on a particular higher harmonic (e.g., the second harmonic) so that the selected harmonic component is amplified relative to the first harmonic.

[0166] The filtered signal can then be used as a measure of the thermal detector-sample proximity: as the probe is moved towards the sample, the probe is stopped when the filtered signal reaches a threshold value, which can be determined empirically as described above.

[0167] Below is a list of reference numbers used in the detailed description of the invention and in the drawings referenced in the detailed description of the invention. [Explanation of symbols]

[0168] α anchor distance, β thermal detector line width, ω conductor thickness or metal film thickness, 10 probe, 12 probe body, 14 contact pad, 16 conductive line, 18 cantilever, 22 test specimen, 24 contact probe, 26 thermal detector, 28 conductor, 30 distal end, 32 front surface, 34 first arm, 36 second arm, 38 front segment, 40 protrusion, 42 second thermal detector, 44 membrane, 46 membrane edge, 48 detector end.

Claims

1. 1. A method for determining proximity between a thermal detector and a test specimen, comprising: providing a probe having a probe body defining a flat surface for supporting a thermal detector; the thermal detector extends from the probe body in a coplanar relationship with the planar surface between a proximal end supported by the planar surface and a distal end opposite the proximal end; a detector geometry defined by the thermal detector having a first geometric dimension, the thermal detector having a detector surface supporting an electrical conductor having a temperature dependent electrical resistance; Furthermore, establishing a temperature differential between the electrical conductor and the test specimen; injecting an alternating current having a first frequency into the electrical conductor; providing an electronic circuit and connecting the electronic circuit to the electrical conductor, the electronic circuit measuring the electrical resistance while moving the probe towards the test specimen and providing an output; providing a filter having a cutoff frequency below a second harmonic of the first frequency and high pass filtering the output to provide a filtered signal; using the filtered signal as a measure of the proximity between the thermal detector and the test specimen; The method according to claim 1,

2. 2. The method of claim 1, wherein the filter is provided in the frequency domain by transforming the output from the time domain to the frequency domain and reducing the amplitude of the output having frequencies below a cutoff frequency of the output relative to the amplitude of the output having frequencies above the cutoff frequency of the output.

3. The method of claim 1 , wherein the filter is a high-pass filter.

Images