CN105241918A - Low temperature thermal conductivity measurement method - Google Patents

Abstract

The invention relates to a low-temperature thermal conductivity measurement method, which can be used to measure the thermal conductivity of a solid substrate or film at normal temperature or low temperature. The present invention is based on the 3ω method of solid material thermal conductivity measurement, and improves it from the following aspects. By changing the thickness of the film on the substrate and measuring multiple times, the longitudinal thermal conductivity of the film is obtained by the differential method, which effectively avoids the deviation of the measured value caused by the boundary thermal resistance on the interface between the film and the substrate. By adding a preamplifier with extremely low linear deviation before the lock-in amplifier for measurement, and amplifying the signal by 10 times, the measurement accuracy is effectively improved. Fully automatic measurement is realized by controlling the digital potentiometer to adjust the magnification of the preamplifier by the single chip computer. By using the magnetic material permalloy, the measurement temperature range is extended to extremely low temperature.

Description

A method for measuring low temperature thermal conductivity

technical field

The invention relates to a method for measuring the thermal conductivity of a solid material, in particular to a method for measuring the thermal conductivity of a solid substrate or thin film at normal temperature or low temperature.

Background technique

The 3ω method is a method for measuring the thermal conductivity of a solid substrate or thin film. (Thermal conductivity measurement from 30 to 750K: the 3ω method, Cahill, David G., Review of Scientific Instruments, 61, 802-808 (1990)) The low-temperature thermal conductivity measurement method according to the present invention is based on the 3ω method.

The measurement principle of the 3ω method is: place a thin conductor wire on the surface of a solid dielectric material, apply a sinusoidal alternating current to the conductor wire through an external circuit, and measure the voltage signal at both ends of the conductor wire. Since the alternating current generates heating power on the conductor wire at twice the frequency of the source signal, the temperature of the conductor wire oscillates accordingly, and the resistance value oscillates at twice the frequency of the source signal, so the voltage signal at both ends of the conductor wire contains the source The ratio of the double frequency and triple frequency components of the signal reflects the amplitude of the temperature oscillation of the conductor wire. The specific quantitative relationship is:

The amplitude of the same-phase component of the source signal of temperature oscillation = (2/conductor wire resistance temperature coefficient) × (three times frequency signal amplitude/one time frequency signal amplitude).

The amplitude of the same-phase component of the source signal of temperature oscillation has a linear relationship with the logarithm of the source signal frequency. Change the source signal frequency to measure the temperature amplitude multiple times, and do a linear fit on the temperature amplitude-frequency logarithmic graph. The slope can be obtained. Thermal conductivity. The specific quantitative relationship is:

Temperature amplitude-frequency natural logarithmic slope=﹣conductor filament heating power/(2π×conductor filament length×substrate thermal conductivity).

The thin film material to be tested is grown on the substrate, and a thin conductor wire is placed on the film, and the measured temperature amplitude-frequency logarithmic graph is compared with the measurement graph of the substrate material without the thin film, from the difference The value gives the thermal conductivity of the film. The specific quantitative relationship is:

The temperature amplitude difference = heating power × film thickness / (conductor wire width × conductor wire length × film thermal conductivity).

The implementation method of the 3ω method is generally as follows: cut the material to be tested into a flat rectangular sheet with a side length of about 0.5-1cm and a thickness of about 400-600um, and define it with a metal mask or photolithography on the upper surface. Conductor wire shape, use physical vapor deposition to make a layer of metal conductor wire with a thickness of about 20-100nm and a width of about 5-40um. The material can be a metal with a large temperature coefficient of resistance, such as Au/Pt. The two ends of the conductor wire are respectively provided with two electrodes of a voltage terminal and a current terminal, and are connected with the measuring circuit. The measurement circuit is generally: a sinusoidal signal of a certain frequency is generated by a lock-in amplifier, which is connected in series with the fixed-value resistor and the conductor wire on the sample, and the voltage at both ends of the fixed-value resistor and the sample conductor wire is respectively amplified by two preamplifiers, and Adjust the magnification of the preamplifier on the fixed value resistor to make the double frequency components of the two signals equal, and finally input the two signals into the lock-in amplifier and measure the triple frequency signal differentially. Put the sample in the environment of the target temperature (such as in a refrigerator), measure it through the measuring circuit, get the temperature oscillation-frequency logarithmic graph, and do linear fitting to get the thermal conductivity of the sample. The heating power on the conductor wire is generally 0.5-2mW, and the frequency range of the source signal is about 1-3000Hz. The sample is generally placed in a 5Pa vacuum environment to reduce the influence of air convection.

The 3ω method is used to measure the thermal conductivity of solid substrates or thin films, which has the advantages of rapidity, easy access to samples, and low vacuum requirements. However, the existing 3ω method has the following disadvantages: when measuring the thermal conductivity of the film, the boundary thermal resistance on the interface between the film and the substrate cannot be deducted, which makes the measured value of the thermal conductivity of the film inaccurate; The temperature coefficient of resistance is small, and the triple frequency voltage signal is too weak, so that the system cannot perform effective measurement at extremely low temperature (<20K).

Contents of the invention

The present invention solves the deficiency of the 3ω method of above-mentioned solid material thermal conductivity measurement from the following aspects, and it is improved:

1. Effectively avoid the measurement value deviation caused by the boundary thermal resistance on the interface between the film and the substrate;

2. Effectively improve the measurement accuracy;

3. Realized automatic measurement;

4. Extend the measurement temperature range to extremely low temperature (<20K).

Specifically, the present invention is a low-temperature thermal conductivity measurement method based on the 3ω method, which solves the above-mentioned deficiencies of the existing measurement methods through the following technical means, and improves it:

1. By changing the thickness of the film on the substrate and measuring it multiple times, the longitudinal thermal conductivity of the film is obtained by the differential method, which effectively avoids the deviation of the measured value caused by the boundary thermal resistance at the interface between the film and the substrate. Specifically, prepare two similar samples with different film thicknesses on the substrate, measure their temperature amplitude-frequency logarithmic plots, compare the two plots, and obtain the thermal conductivity of the film from the difference . The specific quantitative relationship is:

The temperature amplitude difference = heating power × film thickness difference / (conductor wire width × conductor wire length × film thermal conductivity).

Since the film-substrate boundary thermal resistance of the two samples is the same, only the film thickness is different, so when the difference between the two temperature amplitude-frequency logarithmic plots is made, the contribution of the boundary thermal resistance is eliminated, and only the difference caused by the film thickness is retained. The temperature amplitude difference, thereby measuring the longitudinal thermal conductivity of the film.

2. By adding a preamplifier with extremely low linear deviation before the lock-in amplifier for measurement, and amplifying the signal by 10 times, the measurement accuracy is effectively improved. Specifically, the triple frequency component to be measured in the measurement signal is very weak, even smaller than the amplitude of various noise and interference signals, so it is difficult to directly use a lock-in amplifier for accurate measurement. The present invention proposes a solution that firstly uses a preamplifier to amplify the signal. This amplification process requires the linear deviation of the amplifier to be extremely small (less than -90dB), so the magnification cannot be set too high. The present invention adopts the best magnification of 10 in the experimental test. The magnification of this magnification is realized by the self-made extremely low linear deviation preamplifier, which greatly improves the measurement accuracy and effectiveness.

3. Fully automatic measurement is realized by controlling the digital potentiometer to adjust the magnification of the preamplifier through the single-chip microcomputer. Specifically, the single-chip microcomputer controls the digital potentiometer to adjust its resistance value, and then adjusts the amplification factor of the preamplifier, so that the one-fold frequency signals output by the first two preamplifiers are exactly equal to each other and then in the last preamplifier The difference cancels out, allowing the measurement to be made. Other measurement system components such as refrigeration systems and electronic instruments can use the existing technology to program-controlled automatic operation, so that the entire system can realize program-controlled automatic measurement.

4. By using the magnetic material permalloy, the measurement of thermal conductivity at extremely low temperature is realized. Specifically, using the principle that the resistance of dilute magnetic alloys increases abnormally due to the Kondo effect at extremely low temperatures, the heating conductor wire can detect the temperature oscillations on the surface of the sample due to sinusoidal heating, and then measure its thermal conductivity. Extend the measurement range of this thermal conductivity measurement system to extremely low temperature.

By adopting the above technical means, the present invention improves the existing measurement method in many aspects, and the improved measurement system can achieve the following effects:

1. When measuring the thermal conductivity of the thin film, the measurement deviation caused by the boundary thermal resistance between the thin film and the substrate is eliminated.

2. Effectively improve the measurement accuracy (±5%).

3. Realized the program-controlled automatic measurement.

4. The measurement range is extended to extremely low temperature (4K-320K).

Description of drawings

Figure 1 is a schematic diagram of the overall structure of the sample device

Figure 2 is a photolithographic mask design diagram

Figure 3 is a structural diagram of the data measurement circuit

Figure 4 is an example of a temperature amplitude-frequency logarithmic graph

detailed description

The concrete implementation method of the present invention is as follows. It is mainly divided into two parts: device fabrication and data measurement. The following is the implementation steps of a typical measurement example.

1. Device fabrication

1) Sample acquisition: the sample material is prepared or cut into rectangular slices with a size of 5 mm×7 mm, and a thickness of about 0.4-0.6 mm.

2) Surface cleaning: put the sample into analytical pure trichlorethylene for 5 minutes, ultrasonic cleaning in analytical pure acetone for 5 minutes, ultrasonic cleaning in analytical pure isopropanol for 5 minutes, and blow dry with high-purity nitrogen.

3) Drying: The sample is baked on a hot plate at 100-200°C for 10-30 minutes.

4) Glue rejection 1: Place the sample on a glue rejection table, add 1-2 drops of MicroChemPMGISF6 photoresist on the surface, and shake the glue at 4000r/min for 60s.

5) Baking 1: Baking on a hot plate at 170°C for 5 minutes.

6) Glue rejection 2: place it on the glue rejection table, continue to drop 1-2 drops of AZ1500 photoresist on the surface, and shake the glue at 5000r/min for 60s.

7) Baking 2: Baking on a hot plate at 100°C for 60s.

8) De-edge photoresist: Wipe the edges and corners of the photoresist on the surface of the sample with acetone of analytical grade to dissolve the lifted part of the photoresist.

9) Exposure: use the photolithographic template shown in Figure 2 as a light shield, and expose the sample for 11s with a mercury lamp with a power of 300W.

10) Post-baking: put the sample on a hot plate at 140°C and bake for 120s.

11) Developing: Put the sample in AZ300MIF developer for 120s, put it in deionized water for 60s, put it in MicroChemDeveloper101 developer for 90s, put it in deionized water for 60s, and dry it with high-purity nitrogen.

12) Evaporation: Clean the surface of the sample with oxygen ions for 5 minutes under a vacuum of 20-30mTorr, and then perform physical vapor deposition under a vacuum of 2E-6mbar. 3nm Ti is deposited first, followed by 20nm Au.

13) Remove the photoresist: Put the sample into MicroChemRemoverPG and soak it at 60-150°C for 10-30min until the photoresist on the surface falls off. Dry the sample with high-purity nitrogen.

14) Dotted line: the sample is pasted on the copper sample holder, and the electrode of the sample conductor wire is connected to the pin of the sample holder with a thin aluminum lead wire. The overall structure of the sample device is shown in Figure 1.

2. Data measurement

1) Installation: The sample holder is installed in the refrigerator, and the measurement leads are connected. Vacuumize the cooling chamber to 5E-2mbar.

2) Connection circuit: Connect the measurement circuit as shown in Figure 3, the magnification of the preamplifier connected to the sample is set to 1, the magnification of the preamplifier connected to the fixed value resistor is variable, and the magnification of the preamplifier connected to the lock-in amplifier is The magnification was set to 10. The frequency range of the signal output by the lock-in amplifier is 1.7Hz-3kHz, and the heating power on the sample conductor wire is 0.5mW/mm.

3) Program-controlled data acquisition: the refrigerator cools down to the temperature to be measured, and the data acquisition program controls the output of sinusoidal alternating current with different frequencies within the range of 1.7Hz-3kHz to the sample conductor wire. Use the lock-in amplifier to measure the double-frequency signal on the sample conductor wire, and then the program automatically adjusts the magnification of the preamplifier at both ends of the fixed value resistor until the double-frequency signal measured by the lock-in amplifier is the smallest, and uses the lock-in amplifier to collect The magnitude of the triple frequency signal on the sample conductor wire. For each frequency, 200 data points are measured within 20 s and the mean and variance are calculated. Calculate the temperature amplitude at different frequencies, and make a temperature amplitude-frequency logarithmic graph as shown in Figure 4.

4) Thermal conductivity fitting: the slope fitting is performed on the linear portion of the temperature amplitude-frequency logarithmic graph, and the thermal conductivity of the sample material at this temperature is calculated.

For thin film samples, the basic steps for thermal conductivity measurement are the same as above, that is, two samples with different film thicknesses are grown, device fabrication is carried out according to the above step 1, and data measurement is carried out according to steps 1)-3) in step 2. Subtract the two temperature amplitudes measured and plot the difference with respect to frequency. The thermal conductivity of the film can be calculated by taking the average of the flatter section of the graph.

For samples that need to measure thermal conductivity at extremely low temperatures, the magnetic material permalloy can be selected as the conductor wire material, and the measurement can be carried out according to the above steps. The thermal conductivity of materials at very low temperatures can be successfully measured due to the anomalous increase in electrical resistance of magnetic materials at very low temperatures.

Claims (4)

1. the low-temperature thermal conductivity measuring method based on 3 ω methods, thermal conductivity under normal temperature or low temperature can measure solid substrate or film, it is characterized in that: by changing the thickness repetitive measurement of film on substrate, obtain longitudinal thermal conductivity of film with difference method, effectively prevent the measured value deviation that the thermal boundary resistance in film and substrate interface causes.

2. low-temperature thermal conductivity measuring system according to claim 1, is characterized in that: being amplified by signal by increasing a prime amplifier before measurement lock-in amplifier, effectively improve measuring accuracy.

3. low-temperature thermal conductivity measuring system according to claim 1, is characterized in that: the method being regulated prime amplifier enlargement factor by Single-chip Controlling digital regulation resistance, achieves all automatic measurement.

4. low-temperature thermal conductivity measuring system according to claim 1, is characterized in that: by adopting magnetic material permalloy, measurement warm area is expanded extremely low temperature.