Method for measuring atomic scale microscopic thermal conductivity

Abstract

A method for measuring an atomic scale microscopic thermal conductivity, said method being based on a focused ion beam-based sample preparation method and scanning transmission electron microscopy-electron energy loss spectroscopy. A temperature gradient is generated by locally heating a sample in a scanning transmission electron microscope; the temperature is measured in the electron microscope by means of electron energy loss spectrums; local heating in a nano-scale region of a sample is realized, and a stable temperature field having a large temperature gradient is constructed; and atomically resolved phonon spectrum detection and temperature measurement are realized in space. Local temperature information and a relative value of a microscopic thermal conductivity can be measured at an atomic scale, so that the quantitative analysis of the microscopic thermal conductivities of structures at different positions inside the sample can be performed; and the present invention has the advantages of accurate measurement, high spatial resolution, capability of studying nano-scale thermal transport and micro-zone thermal conductivity, etc., and is suitable for samples having complex structures and a plurality of material systems.

Description

A method for measuring atomic-scale microscopic thermal conductivity Technical Field

[0001] This invention relates to the field of thermal conductivity measurement technology, and more specifically to a method for measuring atomic-scale microscopic thermal conductivity. Background Technology

[0002] Thermal conductivity is one of the fundamental properties of materials, defined as the heat flux density per unit temperature gradient, describing a material's ability to conduct heat under non-equilibrium conditions. The unit for volumetric thermal conductivity is usually watts per meter per Kelvin (Wm). -1 K -1 The interfacial thermal conductivity is watts per meter per Kelvin (Wm). -2 K -1 Thermal conductivity provides crucial information about a material's thermal conductivity and heat transfer characteristics, playing a vital role in thermal management, heat dissipation design, and heat conduction applications. With the increasing integration and miniaturization of electronic devices, measuring the microscopic thermal conductivity of materials is even more critical for thermal management and optimizing internal heat transport. As device size shrinks and internal structural complexity increases, interface thermal resistance becomes dominant, making the measurement of interface microscopic thermal conductivity essential for designing heat dissipation systems, thermal interface materials, and thermal barrier materials. For novel temperature-sensitive power electronic devices, optimizing the microscopic thermal conductivity of heat dissipation materials can improve device performance and lifespan. For novel thermoelectric materials, understanding the internal microscopic thermal conductivity is crucial for optimizing thermoelectric performance. Traditional thermal conductivity measurement methods, such as heat flow metering and laser scintillation, can obtain the macroscopic thermal conductivity of materials, reflecting their average thermal conductivity, but lack spatial resolution and cannot obtain the microscopic thermal conductivity at specific locations. In recent years, many new measurement methods based on various microscopes and spectroscopy have been developed to obtain microscopic thermal conductivity, but these methods have certain limitations.

[0003] Photothermal Raman spectroscopy heats the sample at a laser spot by varying the laser power. The temperature at the spot is measured using the frequency shift of the Raman spectrum or the ratio of the Stokes peak to the anti-Stokes peak, thus deriving the thermal conductivity at that spot. This method can be used to measure the thermal conductivity of monolayer graphene. Its limitations include a lack of spatial resolution below 100 nanometers due to the diffraction limit of light, and the need to know the laser power absorbed at various points in the material, making it difficult to analyze complex systems. Time-domain thermal reflectometry (TDTR) uses a pump laser to heat the sample while simultaneously measuring the reflectivity of the surface metal. The relationship between reflectivity and temperature is used to obtain the temperature change over time, thereby calculating the thermal conductivity. This method can be used to determine the interfacial thermal conductivity of heterojunctions. However, this method requires sample sizes larger than micrometers, lacks spatial resolution, and is difficult to analyze the thermal conductivity at various points in complex material systems. Scanning thermal microscopy (SThM) uses a miniature resistance thermometer and heater at the end of the scanning probe to scan the temperature distribution on the sample surface and can also directly analyze the thermal conductivity of micro-areas by measuring the temperature drop of the thermal probe, achieving a spatial resolution of up to 50 nm. Its drawback is that it can only analyze the surface of the material, and the unknown contact thermal resistance between the probe and the surface cannot be ignored.

[0004] In summary, existing technologies cannot obtain microscopic thermal conductivity at the nanoscale and atomic scale, and are difficult to analyze multi-material systems. This is mainly due to the insufficient spatial resolution of characterization methods and the lack of an effective method for micro- and nano-scale processing of samples.

[0005] Summary of the Invention

[0006] To address the shortcomings of existing technologies, the present invention aims to provide a method for measuring atomic-scale microscopic thermal conductivity, which is used to measure temperature distribution at the nanoscale and atomic scale, analyze microscopic heat transport, and obtain micro-region thermal conductivity and interfacial thermal conductivity.

[0007] The principle of this invention is to obtain the temperature gradient at various locations by measuring the local temperature distribution at the nanometer and atomic scales for a sample with a stable temperature gradient, and then obtain the local microscopic thermal conductivity according to Fourier's law of thermal conductivity.

[0008] To solve the above-mentioned technical problems, the technical solution provided by the present invention is as follows:

[0009] A method for measuring atomic-scale microscopic thermal conductivity includes the following steps:

[0010] S1. The sample to be characterized for microscopic thermal conductivity is pretreated to obtain a pretreated sample, and the macroscopic average thermal conductivity of a certain component of the pretreated sample is obtained.

[0011] S2. Using focused ion beam technology, the pretreated sample described in S1 is micro- or nano-fabricated to construct a thermally conductive path strip with microstructure or interface to be characterized. One end of the path strip is deposited with a circuit and connected to the electrode heating area to form a hot end, and the other end is connected to the substrate for heat dissipation to form a cold end, thus creating a transmission electron microscope sample with localized heating to generate a temperature gradient.

[0012] S3. The transmission electron microscope sample described in S2 is characterized in a transmission electron microscope using an in-situ heating system. A set current is applied to the sample to heat the hot end of the path strip and dissipate heat from the cold end, thereby constructing a stable temperature gradient on the path strip of the transmission electron microscope sample with the microscopic thermal conductivity to be measured.

[0013] S4. Use a scanning transmission electron microscope to collect the electron energy loss spectrum at each point of the path strip of the transmission electron microscope sample that generates the temperature gradient as described in S3, and obtain the phonon vibration spectrum at different positions of the sample path strip; use the obtained phonon vibration spectrum to calculate the temperature and temperature gradient at each point of the sample path strip that generates the temperature gradient based on the phonon peak intensity of the Bose distribution.

[0014] S5. Based on the inverse relationship between thermal conductivity and temperature gradient in the one-dimensional thermal conductivity model, obtain the relative value of the microscopic thermal conductivity of the sample path strip.

[0015] Preferred,

[0016] Specifically, S4 is:

[0017] The sample path bands were scanned point-by-point in a continuous space from the cold end to the hot end to acquire the electron energy loss spectrum. Background signals were removed from the electron energy loss portion to the right of the zero-loss peak and the electron energy gain portion to the left of the zero-loss peak to obtain the loss spectrum. and gain spectrum ;

[0018] The gain spectrum originates from the inelasticly scattered electrons that gain energy during the interaction with phonons in the sample material, corresponding to the annihilation process of phonons and equal to the phonon density of states. Multiply by the number of excited-state phonons The loss spectrum originates from inelastically scattered electrons that lose energy when interacting with phonons in the sample material, corresponding to the phonon generation process, and is the phonon density of states. Multiply by the sum of the phonon numbers of the excited state and the ground state. ;in It is the statistical average number of excited-state phonons, which increases with increasing temperature. Phonons follow a Bose distribution. It can be represented as

[0019] ;

[0020] in, It is the energy of the phonon, equivalent to the energy of electron loss or gain in the electron energy loss spectrum. It is Boltzmann's constant. It's temperature;

[0021] For the phonon spectrum collected at any location, its loss spectrum and gain spectrum There is a proportional relationship between them, a proportional function. With the temperature at that point Related, that is

[0022] ;

[0023] ;

[0024] Temperature-dependent residual spectra Theoretically, it should be zero. The temperature value corresponding to this point is obtained by least squares fitting.

[0025] Temperature is calculated by collecting phonon spectra at various points within a spatial range, thus obtaining the temperature distribution within a certain spatial range. The temperature gradient can then be calculated. .

[0026] Preferred,

[0027] Specifically, S5 is:

[0028] The sample path strips are treated according to a one-dimensional thermal conduction model, which is applicable to the one-dimensional Fourier heat transfer law.

[0029] ;

[0030] in, Heat flux density is the heat flux per unit cross-sectional area. It is thermal conductivity. This refers to the temperature gradient along the heat transfer path. Ignoring heat dissipation along the path and assuming a constant total heat flux, uniform strip width, and uniform thickness, and approximating a consistent heat flux density everywhere, the microscopic thermal conductivity at each location is inversely proportional to the temperature gradient at that location.

[0031] ;

[0032] It is the temperature difference between the two sides of the interface;

[0033] The temperature gradient at any point is calculated based on the phonon vibration spectrum. The microscopic thermal conductivity at these points is inversely proportional to the temperature gradient, thus yielding the relative values ​​of the microscopic thermal conductivity at each point.

[0034] Preferred,

[0035] Specifically, S2 is:

[0036] S2.1 Select a sampling area on the pretreated sample described in S1, and use an ion beam to etch pits with a depth of more than 15 μm above and below the sampling area. The vertical width of the two etched areas is between 10 μm and 20 μm, and the horizontal width exceeds the width of the sampling area. Use an ion beam to etch the upper and lower sides of the sampling area to make the upper and lower sides of the sampling area flat.

[0037] S2.2 Insert the probe, with the probe tip contacting the upper side of the sampling area. Deposit carbon at the probe tip to weld the probe to the sampling area, thus completely separating the sampling area from the sample.

[0038] S2.3 A groove slightly larger than the sampling area size is etched at the electrode position of the four-electrode substrate, and its depth is slightly less than the thickness of the sampling area. The sampling area is embedded in the groove. Carbon is deposited at the four corners of the sample in the sampling area so that it is connected to the four electrodes respectively. The sample in the sampling area is welded to the groove in the substrate. The carbon deposited on the probe tip is etched away to separate the probe tip from the sample in the sampling area. The excess carbon on the upper surface is thinned to make the upper surface of the sample flat.

[0039] S2.4 Select a set area and thin it vertically to make the thickness of the thin area less than 50nm. Etch path strips parallel to the heat transfer direction on the thin area. Deposit a large amount of carbon at one end of the thin area to connect with the substrate to form a heat dissipation channel. Deposit a thinner layer of carbon at the other end of the thin area as a heating resistor. Deposit a large amount of carbon at the electrode to concentrate the resistance in the heating circuit close to the lower part of the thin area.

[0040] Preferred,

[0041] Based on the relative value of the microscopic thermal conductivity obtained from S5, the proportional relationship between the microscopic thermal conductivity at different locations of the sample path strip is obtained; by querying or measuring the absolute value of the macroscopic thermal conductivity of any component of the sample, the absolute value of the microscopic thermal conductivity at that location is calibrated as a standard value, thereby obtaining the absolute values ​​of the microscopic thermal conductivity at all other locations.

[0042] Preferred,

[0043] Based on the obtained absolute values ​​of macroscopic thermal conductivity, the absolute values ​​of microscopic thermal conductivity and interfacial thermal conductivity at various points on the sample are calculated; the absolute value of interfacial thermal conductivity is expressed as...

[0044] ;

[0045] in, It is the bulk thermal conductivity of sample material one.

[0046] Preferred,

[0047] To address the issue of uneven thickness in the sample path strips, electron energy loss spectroscopy is used to calculate the thickness of the strips at various points along the sample path, and the microscopic thermal conductivity is corrected accordingly. The corresponding formula then becomes...

[0048] ;

[0049] in, The thickness at different locations is calculated by collecting plasmon spectra of the sample area within an electron microscope.

[0050] Preferred,

[0051] The sample pretreatment in S1 includes cutting the sample into small pieces that can be placed in a scanning electron microscope. Each piece is less than 1 cm in size, and then grinding and polishing it. The surface is then sprayed with carbon or gold to make it conductive.

[0052] Preferred,

[0053] The transmission electron microscope in S4 is equipped with a spherical aberration corrector and a monochromator, with a spatial resolution of over 60 pm and an energy resolution of less than 7 meV.

[0054] The beneficial effects of this invention are:

[0055] The method provided by this invention can measure local temperature information at the nanoscale or even atomic scale, thereby enabling quantitative analysis of the microscopic thermal conductivity of various structures at different locations within a sample. It has advantages such as accurate measurement, high spatial resolution, and the ability to study nanoscale heat transport and micro-region thermal conductivity, and is applicable to samples with complex structures and multiple material systems.

[0056] Attached Figure Description

[0057] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0058] Figure 1 is a schematic diagram of a method for localizing sample heating within a scanning transmission electron microscope; Figure 1a shows an in-situ heated sample rod used to carry the sample and introduce current into the sample through internal wires and probe electrodes; Figure 1b shows a four-electrode substrate with four independent electrodes and a slot capable of carrying a miniature sample, with the inset showing the distance between the electrodes; Figure 1c is a schematic diagram of localizing sample heating, in which the sample is welded and fixed to the slot in the middle of the four-electrode substrate and connected to the electrodes, and local heating is achieved by applying current to two of the electrodes; Figure 1d is a schematic diagram of a sample structure that can achieve localized heating to build a stable temperature gradient.

[0059] Figure 2 shows a scanning electron beam (and ion beam) image of the local heating sample preparation process based on focused ion beam technology; the operations include moving and tilting the sample stage, deposition (white dashed box), etching and thinning (black cross box), and moving the manipulator probe.

[0060] Figure 3 is a schematic diagram of the principle of measuring temperature and interfacial thermal conductivity based on scanning electron energy loss spectroscopy; Figure 3a is a scanning transmission electron microscope image of a strip-shaped thin region containing an interface in a heterojunction sample, with black text indicating the interface location, the positions of the two different materials, and the heating on the right and cooling on the left; Figure 3b is a series of electron energy loss spectra collected in the white dashed box area in 3a, with the spectral lines from bottom to top corresponding to the white arrows in 3a from left to right, the dashed line being the zero-loss peak representing the background signal of transmitted electrons, the signal on the left being the electron energy gain spectrum, and the signal on the right being the electron energy loss spectrum; Figure 3c is a schematic diagram of a one-dimensional strip heat transfer model, which simplifies the heat conduction on the sample structure; Figure 3d is the spatial distribution of temperature in the heat transfer direction under ideal conditions corresponding to the one-dimensional strip model.

[0061] Figure 4 shows the experimental data of measuring the interfacial thermal conductivity of the aluminum nitride / silicon carbide heterojunction interface at different temperatures in the example; Figures 4a and 4b are temperature distribution diagrams including the interface region, with the horizontal and vertical axes representing locations and colors representing temperatures; Figure 4c is a graph of temperature versus space.

[0062] Figure 5 is a flowchart of the method of the present invention.

[0063] Detailed Implementation

[0064] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the following embodiments are given for illustrative purposes only and are not intended to limit the scope of the present invention. Those skilled in the art can make various modifications and substitutions to the present invention without departing from its spirit and essence.

[0065] This invention provides a method for measuring atomic-scale microscopic thermal conductivity. The method involves locally heating a sample within a scanning transmission electron microscope to generate a temperature gradient, thereby measuring the microscopic thermal conductivity at the atomic scale. It is mainly based on a sample preparation method using a focused ion beam and electron energy loss spectroscopy based on scanning transmission electron microscopy.

[0066] Specifically, the sample is micro- or nano-fabricated to construct an elongated heat-conducting path containing microstructures or interfaces to be studied. A circuit is deposited at one end of the path and connected to electrodes for heating, while the other end dissipates heat, creating a stable temperature gradient along the path. Using scanning transmission electron microscopy (STEM) electron energy loss spectroscopy, the phonon vibrational spectra at different locations on the sample are obtained. The temperature at each location is then calculated based on the phonon Bose distribution, thus obtaining the local temperature gradient with sub-nanometer spatial resolution. Since the sample is in a high vacuum, it can be approximated as an adiabatic boundary condition, and the heat flow along the heat-conducting path is consistent throughout. According to Fourier's law of thermal conductivity, the local thermal conductivity is inversely proportional to the temperature gradient at that location; therefore, the relative values ​​of the thermal conductivity at different locations can be obtained. By combining this with the known thermal conductivity of a component along the heat-conducting path or the overall average thermal conductivity as a reference, the absolute value of the microscopic thermal conductivity can be derived. Since the spatial distribution map of temperature provides all the information except for the heat flux value under steady-state heat transport, the method provided by this invention can measure local temperature information at the nanometer or even atomic scale, thereby enabling quantitative analysis of the microscopic thermal conductivity of various structures at different locations inside the sample.

[0067] This invention has the advantages of accurate measurement, high spatial resolution, and the ability to study nanoscale heat transport and micro-area thermal conductivity. It is applicable to samples with complex structures and multiple material systems.

[0068] Example 1: Measurement of interfacial thermal conductivity of silicon carbide / aluminum nitride heterojunction

[0069] The method of the present invention includes the following steps:

[0070] A. Pre-treat the sample, including cutting it into small pieces, grinding and polishing it, and attaching it to the sample stage with conductive adhesive.

[0071] Depending on the research and measurement objectives, the details of sample preparation may vary. In this example, the sample is aluminum nitride directly grown on a silicon carbide substrate, where the silicon carbide is a single-crystal 4H phase. The bulk thermal conductivity is known (obtained from the literature), so only a small sample needs to be prepared for subsequent micro / nano fabrication using a focused ion beam. For other samples with unknown thermal conductivity, it is necessary to measure their macroscopic average thermal conductivity.

[0072] B. Using focused ion beam (FIB) technology to prepare samples that can be locally heated to create a temperature gradient and are suitable for scanning transmission electron microscopy.

[0073] This step is performed within the Helios G4 instrument, which integrates ion beam and scanning electron microscopy. The instrument allows for operations such as sample stage movement and tilting, carbon and metal deposition, probe insertion and movement, and probe rotation. The specific implementation method is as follows:

[0074] B1. Select an area of ​​approximately 25μm × 5μm (hereinafter referred to as the sampling area) at an appropriate location on the bulk sample. Deposit approximately 2μm of amorphous carbon on the surface of this area, as shown in the white dashed box in Figure 2a. Etch pits with a depth exceeding 15μm above and below the sampling area using an ion beam. The vertical width of the two etched areas should be between 10μm and 20μm, and the horizontal width should exceed the width of the sampling area by 25μm. Continue etching with an ion beam above and below the sampling area to make the upper and lower sides of the sampling area flat. The etching location is shown in the black box in Figure 2a.

[0075] B2. Cut the sampling area from the top side with an ion beam, cutting through the sample edge along the black box in Figure 2b, so that the left, right and bottom of the sampling area are separated from the block sample, but leave a small section on the right side uncut.

[0076] B3. Insert the probe and move it to contact the upper side of the sampling area with the tip, as shown in Figure 2c. As shown in Figure 2d, deposit carbon at the tip to firmly weld the probe to the sampling area, and cut off the small section on the right side of the sampling area to completely separate it from the block sample. Finally, manipulate the probe to remove the small block sample from the sampling area.

[0077] B4. As shown in Figure 2e, grooves slightly larger than the size of the small sample (25μm × 15μm) are etched at the electrode positions on the four-electrode substrate, with a depth slightly less than the thickness of the small sample. As shown in Figure 2f, the probe is manipulated to embed the small sample into the groove. As shown in Figure 2g, carbon is deposited at the four corners of the sample to connect them to the four electrodes respectively, while firmly welding the sample to the substrate groove. Then, the carbon deposited on the probe tip in the previous step is etched away to separate the probe tip from the sample. Finally, as shown in Figure 2h, the excess carbon on the upper surface is thinned to make the upper surface of the sample smooth.

[0078] B5. As shown in Figure 2i, a region approximately 6 μm wide is selected and thinned vertically to a thickness of less than 50 nm for scanning transmission electron microscopy (STEM) observation. Slits parallel to the heat transfer direction are etched into this thin region to create a strip structure, as shown in Figure 2j. Simultaneously, a large amount of carbon is deposited at one end of the thin region (above Figure 2j) to connect it to the substrate, forming a heat dissipation channel. A thinner layer of carbon is deposited at the other end of the thin region as a heating resistor. Furthermore, a large amount of carbon is deposited at the electrodes to concentrate the resistance in the heating circuit below the thin region.

[0079] C. Using an in-situ heating system and scanning transmission electron microscopy (STEM) electron energy loss spectroscopy, a temperature gradient is constructed and detected on the strip sample to obtain the interfacial temperature difference, thereby calculating the interfacial thermal conductivity.

[0080] The in-situ heating system consists of an in-situ heating rod, connecting wires, a control power supply, and supporting control software. It can apply different currents to the sample inside the electron microscope and monitor changes in voltage and resistance. Using a scanning transmission electron microscope (Nion U-HERMES200) equipped with a spherical aberration corrector and monochromator, a spatial resolution of up to 60 pm is achieved, enabling atomic resolution. The energy resolution can reach below 7 meV, used to obtain the lattice vibration spectrum of the sample, i.e., the phonon spectrum. Since phonons follow a Bose distribution, the temperature at a given point can be calculated based on the intensity of the phonon peak. The specific implementation method is as follows:

[0081] C1. Mount the sample into the in-situ heating rod and characterize it inside the scanning transmission electron microscope (SEM). Apply an appropriate current to the sample using a software-controlled power meter and wait for the sample to reach a steady state within the microscope where it no longer drifts; at this point, the temperature of each part no longer changes with time. Since one end of the strip-shaped thin region on the sample is connected to the heating area and the other end is connected to the substrate for heat dissipation, a stable temperature gradient has been generated within the space containing the thermal conductivity of the interface to be measured.

[0082] C2. As shown in Figure 3a, the electron energy loss spectrum is collected by scanning point by point within a certain spatial range including the interface. Figure 3b shows a series of electron energy loss spectra collected from the continuous space from the cold end to the hot end. The background signal is removed from the electron energy loss portion to the right of the zero-loss peak and the electron energy gain portion to the left of the zero-loss peak to obtain the loss spectrum. and gain spectrum The gain spectrum originates from the inelasticly scattered electrons that gain energy when interacting with phonons in the material, corresponding to the annihilation process of phonons, and is equal to the phonon density of states. Multiply by the number of excited-state phonons The loss spectrum originates from inelastically scattered electrons that lose energy when interacting with material phonons, corresponding to the phonon generation process and representing the phonon density of states. Multiply by the sum of the phonon numbers of the excited state and the ground state. .in It is the statistical average number of excited-state phonons, which increases with increasing temperature. Since phonons follow a Bose distribution, It can be represented as

[0083] ;

[0084] in, It is the energy of the phonon (equivalent to the energy of electron loss or gain in the electron energy loss spectrum). It is Boltzmann's constant. It is temperature. For a phonon spectrum collected at any location, its loss spectrum... and gain spectrum There is a proportional relationship between them, a proportional function. With the temperature at that point Related, that is

[0085] ;

[0086] ;

[0087] Temperature-dependent residual spectra Theoretically, it should be zero. Therefore, the temperature value corresponding to the point is obtained by least squares fitting. The temperature is calculated by collecting the phonon spectrum of each point in the spatial range, and the temperature variation curve in a certain spatial range is obtained. Ideally, it should be similar to Figure 3d.

[0088] C3. As shown in Figure 3c, the sample in the experiment can be treated according to a one-dimensional heat conduction model, which is applicable to the one-dimensional Fourier heat transfer law.

[0089] ;

[0090] in, Heat flux density is the heat flux per unit cross-sectional area. It is thermal conductivity. This refers to the temperature gradient along the heat transfer path. Ignoring heat dissipation along the path, the total heat flux can be assumed to be constant. Furthermore, since the width of the strip-shaped path is constant and its thickness is considered uniform, the heat flux density can be approximated as being uniform everywhere. Therefore, the thermal conductivity at each location is inversely proportional to the temperature gradient at that location.

[0091] ;

[0092] in, It is interfacial thermal conductivity. This represents the temperature difference between the two sides of the interface. Based on the temperature distribution obtained in the previous step, the temperature gradient at any point can be calculated. The microscopic thermal conductivity at these points is inversely proportional to the temperature gradient, allowing us to obtain the relative magnitude of the microscopic thermal conductivity at each point. Combined with the volume thermal conductivity obtained in step A, the absolute values ​​of the microscopic thermal conductivity and interfacial thermal conductivity at each point can be calculated. In Figure 3d, the absolute value of the interfacial thermal conductivity can be expressed as...

[0093] ;

[0094] in, It is the bulk thermal conductivity of material 1.

[0095] As a preferred embodiment, in step C3, the heat flux density can be corrected using thickness information at different locations of the sample, which is suitable for cases where the sample thickness is uneven. The corresponding formula becomes...

[0096] ;

[0097] in, The thickness at different locations can be directly calculated by collecting the plasmon spectrum of the sample area in an electron microscope.

[0098] The interfacial thermal conductivity of the aluminum nitride / silicon carbide (AlN / SiC) heterojunction interface at different temperatures was measured using the method of this invention. Sample preparation was consistent with steps A and B, and experimental procedures were consistent with step C1. Applying different currents to the heating end yielded steady-state heat transfer at different temperatures and temperature gradients. The temperature distribution within a 400 nm region near the interface was detected by scanning electron energy loss spectroscopy, using the same method as described in step C2. The results for 5 mA and 8 mA currents are shown in Figures 4a and 4b, respectively, and the temperature variation along the heat transfer path is shown in Figure 4c. Figure 4a corresponds to the curve below Figure 4c, with the interface temperature at approximately 460 K and a temperature difference of 478 K - 448 K = 30 K across the interface. The temperature gradient of silicon carbide on the left is (442K-425K) / 140nm = 1.06E8 K / m. According to the literature (J. Appl. Phys. 113, 053503 (2013)), the bulk thermal conductivity of silicon carbide along the heat flow direction at this temperature is approximately 110 W / (mK). Based on the calculation formula in step C3, the interfacial thermal conductivity at this temperature is calculated to be 390 W / (mK). 2 K). The curve in Figure 4b corresponds to the curve in Figure 4c. The temperature at the interface is about 540K, and the temperature difference between the two sides of the interface is 559K-534K=25K.

[0099] The temperature gradient of silicon carbide on the left is (534K-509K) / 160nm = 1.56E8 K / m. The bulk thermal conductivity of silicon carbide along the heat flow direction at this temperature is approximately 80 W / (mK). Based on the calculation formula in step C3, the interfacial thermal conductivity at this temperature is calculated to be 500 W / (m). 2 The results conform to the general rule that interfacial thermal conductivity increases with increasing temperature (K).

[0100] This invention achieves localized heating of samples within a micrometer-scale region, constructing a stable temperature field with a large temperature gradient (hundreds of Kelvin per micrometer), through sample preparation and in-situ heating schemes using localized heating within a scanning transmission electron microscope (STEM) to generate a temperature gradient. It also enables atomically resolved phonon spectroscopy detection and temperature measurement, revealing that temperature abrupt changes at heterojunction interfaces actually occur at a scale within 2 nm. This method is sufficient to measure the relative value of microscopic thermal conductivity at the atomic scale, and combine this with macroscopic measurements of bulk thermal conductivity (see references) to obtain the absolute value of microscopic thermal conductivity.

[0101] The contents not described in detail in this specification are existing technologies known to those skilled in the art.

[0102] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method of measuring microscale thermal conductivity at atomic scale, characterized by, The method comprises the following steps: S1. Preparing a sample to be characterized for microscale thermal conductivity, and obtaining a macroscopic average thermal conductivity of a component of the sample; S2. Using a focused ion beam technique to micro-nano process the sample prepared in S1, and constructing a thermal conduction path strip of a microstructure or interface to be characterized, one end of the strip being deposited with a circuit and connected to an electrode heating area to form a hot end, and the other end being connected to a substrate to dissipate heat to form a cold end, thereby preparing a transmission electron microscope sample for local heating to generate a temperature gradient; S3. Characterizing the transmission electron microscope sample prepared in S2 in a transmission electron microscope using an in-situ heating system, applying a set current to the sample, heating the hot end of the path strip, and dissipating heat at the cold end, thereby constructing a stable temperature gradient on the path strip of the transmission electron microscope sample to be characterized for microscale thermal conductivity; S4. Collecting electron energy loss spectra of the path strip of the transmission electron microscope sample prepared in S3 using a scanning transmission electron microscope, and obtaining phonon vibration spectra of different positions of the path strip of the sample; and calculating temperatures and temperature gradients of different positions of the path strip of the sample prepared for a temperature gradient according to a phonon peak intensity of a Bose distribution of the obtained phonon vibration spectra; S5. Obtaining a relative value of microscale thermal conductivity of the path strip of the sample according to a relationship that thermal conductivity is inversely proportional to a temperature gradient in a one-dimensional heat conduction model.

2. The method according to claim 1, wherein S4 is specifically: which can be expressed as The sample path strip is continuously spatially scanned point by point from the cold end to the hot end, electron energy loss spectrum is collected, background signals are removed from the electron energy loss part on the right side of the zero loss peak and the electron energy gain part on the left side of the zero loss peak respectively, and loss spectrum is obtained and gain spectrum ； where the gain spectrum is derived from the inelastic scattering of electrons in the transmission electron microscope that gain energy from the interaction with phonons in the sample material, corresponding to the annihilation of a phonon state, equal to the phonon density of states in terms of excited state phonon number and the loss spectrum comes from inelastic scattering of electrons losing energy when interacting with phonons in the sample material, corresponding to the process of phonon creation, the phonon density of states multiplying the sum of the phonon numbers of the excited state and the ground state ; wherein is the statistical average of the number of excited phonons, which increases with temperature, phonons obeying a Bose distribution, is a temperature; ； wherein, is the energy of the phonon, equivalent to the energy of the electron loss or gain in the electron energy loss spectrum, is the Boltzmann constant, is related to, that is, for the phonon spectrum collected at any position, its loss spectrum and gain spectrum proportional relationship, proportional function temperature at the point Theoretically, it should be zero, and the temperature value corresponding to the point is obtained by least square fitting. ； ； Residual spectrum with temperature dependence 3. The method according to claim 2, wherein S5 is specifically: The temperature distribution in a certain space range is obtained by calculating the temperature of each point in the space range according to the phonon spectrum of the point , and the temperature gradient can be calculated 。 The path strip of the sample is processed according to a one-dimensional heat conduction model, and is applicable to a one-dimensional Fourier heat conduction law is a temperature gradient in a heat conduction path direction; heat dissipation in the heat conduction path is ignored, the total heat flow is considered to be constant, the width of the strip path is constant, the thickness is considered to be uniform, and the heat flux density is considered to be uniform everywhere, so that the microscale thermal conductivity of each position is inversely proportional to the temperature gradient at the position is a temperature difference on both sides of an interface; ； wherein J is the heat flux, which is the heat flow per unit cross-sectional area; is the thermal conductivity, The microscale thermal conductivities at the points are inversely proportional to the temperature gradient, and the relative values of the microscale thermal conductivities at different positions are obtained. ； wherein is the interfacial thermal conductance, 4. The method according to claim 1, wherein S2 is specifically: Determination of temperature gradient at any point from phonon vibration spectrum S2.1 Selecting a sampling area on the sample prepared in S1, etching pits with a depth of more than 15 μm above and below the sampling area using an ion beam, the width of the two etching areas being 10 μm to 20 μm, and the left and right widths being greater than the width of the sampling area; and etching above and below the sampling area using an ion beam to make the upper and lower sides of the sampling area flat; S2.2 Inserting a probe, the probe needle being in contact with the upper side of the sampling area, depositing carbon at the needle position to weld the probe and the sampling area, and completely separating the sampling area from the sample; ​ ​ ​ S2.3 Etching a groove slightly larger than the size of the sample area on the electrode position of the four-electrode substrate, the depth of which is slightly smaller than the thickness of the sample area, embedding the sample area into the groove, depositing carbon on the four corners of the sample area to make it connected with the four electrodes respectively, welding the sample area in the groove of the substrate, etching the carbon deposited on the probe tip to separate the probe tip from the sample area, and etching the excess carbon on the upper surface to flatten the sample surface; S2.4 Selecting a set area to thin from top to bottom, making the thin area less than 50 nm thick, etching a path strip parallel to the heat transfer direction on the thin area, depositing a large amount of carbon on one end of the thin area to connect with the substrate base to build a heat dissipation channel, depositing a thin carbon on the other end of the thin area as a heating resistor, and supplementing a large amount of carbon at the electrode to concentrate the resistance in the heating circuit close to the lower part of the thin area.

5. The method of claim 3, wherein the relative value of the micro thermal conductivity obtained in S5 is used to obtain the proportional relationship of the micro thermal conductivity at different positions of the sample path strip; and the absolute value of the micro thermal conductivity at any position of the sample is calibrated by taking the absolute value of the macro thermal conductivity of any component of the sample as a standard value, so as to obtain the absolute values of the micro thermal conductivities at all other positions.

6. The method of claim 5, wherein the absolute values of the micro thermal conductivity and the interface thermal conductivity at different positions of the sample are calculated by combining the obtained absolute value of the macro thermal conductivity, and the absolute value of the interface thermal conductivity is expressed as: k = k0(1 - e -d / k0) where k is the absolute value of the interface thermal conductivity, k0 is the bulk thermal conductivity of the sample material, d is the thickness of the sample path strip, and e is the base of the natural logarithm.

7. The method of claim 6, wherein for the case of uneven thickness of the sample path strip, the thickness of the sample path strip at different positions is calculated by using electron energy loss spectroscopy, and the micro thermal conductivity is corrected, and the corresponding formula becomes: k = k0(1 - e -d / k0) where k is the absolute value of the micro thermal conductivity, k0 is the bulk thermal conductivity of the sample material, d is the thickness of the sample path strip at different positions, and e is the base of the natural logarithm.

8. The method of claim 1, wherein the sample pretreatment in S1 includes cutting the sample into small pieces that can be placed into a scanning electron microscope, each of which has a size of less than 1 cm, grinding and polishing, and spraying carbon or gold on the surface to make it conductive. ； wherein 9. The method of claim 1, wherein the transmission electron microscope in S4 is equipped with a spherical aberration corrector and a monochromator, with a spatial resolution of more than 60 pm and an energy resolution of less than 7 meV. ​ ​ ； wherein, ​ ​ ​ ​ ​

Images