Search

The heat dissipation of electronic devices has become an enormous challenge with the dramatic increasing of integrated transistors on a chip. Nanowires are viewed as promising materials for solving this problem. The accurate measurement of thermal diffusivity is a precondition for further applications. This paper presents a non-destructive method to measure the thermal diffusivity of a single supported nanowire, which is called dual-wavelength flash Raman (DFR) mapping method. This method has a 100 ps temporal resolution based on the minimum time delay between the heating and probing laser pulses and 50 nm spatial resolution based on the minimum step of the probing pulse. Compared with previously proposed concentric DFR method, the probing pulse can move along the nanwire length and the movable probing laser beam allowed temperature measurements at the most sensitive position,which improves the measurement accuracy by 50%. Moreover, instead of the normalizaiton temperature method in concentric DFR method, a temperature phase method was proposed for analyzing the temperature variation curves. The measurements on a single supported silicon nanowire verified the reliability of the DFR mapping method. One thousand virtual experiments showed that the uncertainty of this method can achieve ±5%.

A new method for measuring thermal conductivities of gases and liquids was established by demonstrating the measurement of five kinds of liquid and air. It uses a sensor named “micro-beam sensor” that is a ∼10-μm-long free-standing platinum membrane suspended across a trench on a silicon substrate and heated in a sample by DC. This method is unique in that it is a steady-state measurement but free from the effect of natural convection owing to the micrometer size of the sensor. Improving the method for precisely determining the temperature of the sensor and modifying the device from those used in our previous feasibility study, we successfully measured the thermal conductivity ranging from ∼0.03 to ∼0.6 W/(m⋅K) within 4% error.

Size dependence is one of the most important unique features of thermal conductivity in two-dimensional materials. Suspended single-layer graphene (SLG) provides a perfect platform for studying the size dependent phonon transport. Here we report measurement and theoretical analysis of heat conduction in suspended SLG as a function of width and temperature. The thermal conductivity of graphene was larger for wider SLG. This width effect was smaller at higher temperatures. In suspended SLG, the long wave-length phonons tend to be more scattered at the lateral boundaries of narrow SLG ribbon, in which the mean free path of phonons is close to the sample width. This behavior can be understood as a mode selectivity of phonon-boundary scattering for suspended SLG. The result revealed the unique width dependence of thermal conductivity in suspended SLG and provided useful guidelines for the future SLG-based thermal applications.

A new method is proposed for measuring the in-plane thermal conductivity of thin films using a free-standing micro-beam metallic sensor. The sensor is heated in a vacuum by DC to induce a temperature rise, which is determined from the electrical resistance of the sensor. The method consists of two protocols: measurement of a bare sensor before and after deposition of a sample film on its top surface. The thermal conductivity of the sample film is determined by comparing the measured temperature rise with that obtained through numerical analysis. This is based on the principle that the temperature rises of the sensor with and without a deposited film are different because of a difference in in-plane thermal conductance. In this study, measurement of a 20-nm-thick gold film was demonstrated by fabricating two platinum sensors of different widths. The measured thermal conductivities of the platinum sensors and gold film were significantly smaller than those of bulk materials. The relationship between the thermal and electrical conductivities was also discussed.

Raman-based thermal characterization is regarded as an invaluable tool in micro/nanoscale heat transfer research, offering exceptional contrast in conjunction with high specificity of noncontact and material specific temperature measurement compared with competing thermometry techniques at the micro/nanoscale. It has been extensively used to determine the thermophysical properties of micro/nanoscale materials. However, for commonly used steady state Raman methods, as two concluded main factors, the temperature coefficients of Raman properties and heating level (or optical absorption) will affect the accuracy of resulting temperature and thermal properties. In this review, we critically discuss the mechanism of Raman spectrum response to temperature and possible error factors in calibration and measurement. In addition, the influence of measurement setup is discussed, and possible technical solutions for improving the measurement accuracy are reviewed. Among noble developments in Raman-based thermal characterizations, the transient heat transfer analysis has been coupled to advances in noncontact Raman-based thermal measurement. By enhancing the temporal and spatial resolution in existing technical conditions, more efficient and accurate transient thermal properties measurement can be realized. Particular attention is paid to the so-called resolved Raman techniques for simultaneous measurement of multiple thermal properties at the micro/nanoscale. In particular, we critically review how these tools can reveal new insights into the complex energy transport processes in 2D semiconductors, which have been impossible to tackle using traditional tools. Considering its precision and sensitivity, there is still a large room for development of Raman techniques for the investigation of complex and coupled heat transfer process in low dimensions and new materials.

This paper presents a dual-wavelength flash Raman (DFR) mapping method for characterizing the thermal diffusivity of suspended 2D anisotropic materials. Using one heating laser beam and another probing laser beam with different wavelength, the crystal orientation can be determined by steady-state temperature map. Then modulating the two continuous wave laser beams into periodic pulsed laser beams, the transient heating and cooling curves at various positions along principle axes can be measured by varying the delay time between two pulsed laser beams and changing the position of the probing laser spot. The temporal resolution can reach 100 ps while the spatial resolution can reach 50 nm which are significantly better measurement accuracies. Phase processing method and optimal measurement positions of various thermal diffusivities were presented to further reduce the uncertainty. The effect of the laser absorption coefficient can be eliminated and no other physical parameters are required unlike in existing methods. Feasibility analysis verified that this method is applicable to ultrathin 2D anisotropic materials and even for monolayer materials. When the probing spots are selected around the optimal positions, the measurement uncertainty can be within ±5%.

This paper presents a “dual-wavelength flash Raman (DFR) mapping method” for in-situ measuring the thermal diffusivity of suspended 2D nanomaterials. Using a periodical heating laser pulse to heat the sample, and using another different wavelength laser pulse as a probe to measure the temperature rises of the sample by its Raman band shifts, the temperature variations of the sample during the heating and cooling period can be determined by changing the time deviation between the heating pulse and the probing pulse. Through changing the position of the probing laser spot, the temperature variation curves at various positions can be measured. The sensitivity and uncertainty analysis shows the measurement accuracy of the mapping method can be significantly improved compared with it of the concentric DFR method. Furthermore, instead of normalization analysis in the DFR method, a specific definition of temperature phase is proposed to analyze the above-mentioned temperature variation curves for characterizing the thermal diffusivity. The uncertainty analysis with a hundred virtual experiments verified that phase analysis is more accurate than the normalization analysis of the same experimental data. When the uncertainty of the temperature measurement is ±2% and each temperature measurement repeated 4 times, with multi-position phase analysis, the uncertainty of the measured thermal diffusivity can be within ±5%.

In this paper, the optical absorption and heat transfer coefficients of a single suspended 4.3 μm diameter carbon fiber have been measured in the experiment. The test fiber was first heated locally by a laser beam and the temperature rise was measured from the shifts in the Raman G-band frequency, and then the laser heating was replaced by an electrical current heating, the electrical power was adjusted to have the same temperature rise. Since the electrical power can be measured accurately, the absorbed laser power can be extracted from one-dimensional heat conduction analysis and the heat transfer coefficient can be obtained as well. The present experimental results show that the optical absorption of 4.3 μm diameter carbon fiber is 39.62% and the heat transfer coefficient is 4150 Wm−2 K−1, which agrees well with the prediction of published experimental empirical formula.

Metallic nanofilms are of great importance in integrated circuit design and electronic devices. Understanding energy dissipation and transport in metallic nanofilms is essential to practical thermal management. The Wiedemann–Franz (WF) law states a precisely fixed ratio by which the electrons transport heat and charge, providing a basic rule to determine the thermal properties. Hitherto no bulk material has been known to violate the WF law. We report compelling evidence for the breakdown of the WF law in polycrystalline gold nanofilms at low temperatures, the Lorenz number increases notably with decreasing temperature. Our results show that the electrons dominate in heat transport at high temperatures, leading to a constant Lorenz number. While below 40 K, inelastic electron scattering at grain boundaries becomes significant and part of the electron energy is transferred to phonons. Correspondingly, the phonon thermal conductivity is increased and the WF law is violated. A detailed kinetic theoretical model has been developed to investigate several phonon scattering mechanisms in depth and matches well with the experimental results.

In this paper the mechanism of heat transport in metallic nanofilms under ultra-short pulsed laser heating is examined theoretically and experimentally. In order to easily understand the non-equilibrium heat transport in metallic nanofilms the study of heat transport behavior is first carried out in dielectrics. The analyses indicate that there may be two kinds of wave phenomena in dielectrics subjected to a periodic surface temperature. One is the thermal wave governed by the C–V model based hyperbolic equation and the other is the diffusive wave governed by the Fourier model based parabolic equation. According to the hyperbolic two step model for non-equilibrium heat transport, such two kinds of wave phenomena can also occur simultaneously in the metallic nanofilms under pulsed laser heating, where the diffusive wave is induced by the electron temperature oscillation at the surface due to the non-equilibrium between electrons and lattices. Unlike the propagation speed of the thermal wave, the propagation speed of the diffusive wave depends not only on the medium properties but also the period of the temperature oscillation at the boundary. Hence, the propagation speed of the diffusive wave in the electron gas may be of as high as 10 6 m s −1 , when the laser pulse duration is less than 1 ps. A transient thermoreflectance (TTR) system has been built to measure the transient electron temperature responses caused by the femtosecond laser heating and a pump–probe technique is used to ensure the femtosecond temporal resolution in the experiments. Different from the commonly used front heating-front detecting (FF) method for measuring the material properties, a rear heating-front detecting (RF) method is applied, so that measuring the propagation speed of heat becomes available. The non-equilibrium heat diffusion model is used to fit the measured normalized electron temperature profiles of 27.2 nm, 39.9 nm and 55.5 nm Au films. The best-fitted coupling factor G basically agrees with the theoretical value 2.3 × 10 16 W m −3 K −1 . The propagation speed of the diffusive wave in the electron gas can be obtained by comparing the measured delay time of peak electron temperatures of Au films with different thicknesses. The average propagation speed of the temperature oscillation or diffusive wave in Au films for the range of thickness from 27.2 nm to 55.5 nm is equal to 8.1 × 10 5 m s −1 , which is close to the value predicted by the non-equilibrium heat diffusion model.

Based on Einstein’s mass–energy relation, the equivalent mass of thermal energy or heat is identified and referred to as thermomass. Hence, heat conduction in carbon nanotubes (CNTs) can be regarded as the motion of the weighty phonon gas governed by its mass and momentum conservation equations. The momentum conservation equation of phonon gas is a damped wave equation, which is essentially the general heat conduction law since it reduces to Fourier’s heat conduction law as the heat flux is not very high and the consequent inertial force of phonon gas is negligible. The ratio of the phonon gas velocity to the thermal sound speed (the propagation speed of thermal wave) can be defined as the thermal Mach number. For a CNT electrically heated by high-bias current flows, the phonon gas velocity increases along the heat flow direction, just like the gas flow in a converging nozzle. The heat flow in the CNT is governed by the electrode temperature until the thermal Mach numbers of phonon gas at the tube ends reach unity, and the further reduction of the electrode temperature has no effect on the heat flow in the CNT. Under this condition, the heat flow is said to be choked and temperature jumps will be observed at the tube ends. In this case the predicted temperature profile of the CNT based on Fourier’s law is much lower than that based on the general heat conduction law. The thermal conductivity which is determined by the measured heat flux over the temperature gradient of the CNT will be underestimated, and this thermal conductivity is actually the apparent thermal conductivity. In addition, the heat flow choking should be avoided in engineering situations to prevent the thermal failure of materials.