
The future nanoelectronics development involves using the smaller- -and-smaller-sized circuit components based on the micro- and nanostructures. This causes a growth of the specific heat flows up to 100 W/cm2. Since performance of electronic devices is strongly dependent on the temperature there is a challenge to create the heat transfer models, which take into account the size effect and ensure a reliable estimate of the thermal conductivity. This is one of the crucial tasks for development of new generations of integrated circuits.
The paper studies heat transfer processes using the silicon thin films as an example. Thermal conductivity calculations are performed taking into account the influence of the classical size effect in the context of the Sondheimer model based on the solution of the Boltzmann transport equation.
The paper, for the first time, presents and considers the influence of various factors on the thermal conductivity of thin films, namely temperature, film thickness, polarization of the phonon waves (transverse and longitudinal), velocity and relaxation time versus frequency for the phonons of different wave types.
Based on the analysis, three models with different accuracy are created to estimate the influence of detailing processes under consideration on the thermal conductivity in a wide range of temperatures (from 10 K to 450 К) and film thickness (from 10 nm to 100 µm).
So in the model I for the first time in calculating thermal conductivity of thin films we properly and circumstantially take into account the dependence of the velocity and the relaxation time of phonons on the frequency and polarization. The obtained values are in a good agreement with available experimental data and theoretical models of other authors. In the following models we use few average methods for relaxation times and velocities, which leads to significant reduction in calculating accuracy up to the values exceeding 100%.
Therefore, when calculating the thermal conductivity for micro - and nanostructures it is necessary to employ a detailed process description of the phonons propagation for different energies (frequencies) and polarizations, and consider the real dispersion relations (velocities of their propagation) for each concerned temperature and thickness of the sample.
The above recommendations can be used to estimate the in-plane thermal conductivity of thin films when simulating the new structures in advanced semiconductor devices.