Isotope effect on radiative thermal transport

Isotope effects on heat conduction and convection have been known for decades. However, whether thermal radiation can be isotopically engineered remains an open question. Here, we predict over three-orders-of-magnitude variation of radiative heat flow with varying isotopic compositions for polar dielectrics at room temperature. We reveal this as an isotope mass effect which induces phonon line shift and broadening that in turn affect phonon-mediated resonant absorption both in the near and far field. In contrast, the isotope effect is negligible for metals and doped semiconductors which largely depend on free carriers. We also discuss the role of temperature with regard to surface mode excitation.

Figure 1

Isotope effects on the phonon properties and permittivity of cBN. (a) Modeled and measured (symbols) frequency shift of the zone-center optical phonons with varying fraction of ${}^{10}\mathrm{B}$, and the mass-fluctuation parameter $g$. (b) Permittivities of representative cBN.

Figure 2

Isotope effects on radiative thermal transport between bulk cBN at room temperature. (a) Total HTCs for representative pairs of isotope-engineered cBN as a function of gap size. Inset shows the magnitude of the isotope effect. (b) Spectral HTCs and (c) the corresponding LDOS.

Figure 3

Thin-film enhancement of the isotope effect. (a) Isotope effect at three typical gap sizes with varying cBN film thickness. (b), (d) Spectral HTCs of the symmetric and asymmetric pairs in the near and far field, respectively. (c), (e) Corresponding LDOS for the $s$ and $p$ polarizations.

Figure 4

Isotope effect for real materials supporting surface waves. (a) Polar dielectrics of varying damping factors. (b) Spectral HTC showing broadened SPhP peaks due to large damping. (c) Metals and semiconductors of different doping levels. All films are 1 nm thick.

Figure 5

Parametric study of the isotope effect. (a) $\eta$ map with respect to the damping factor and line shift. Select values are marked at $\mathrm{\Delta }{\omega }_{\mathrm{LO}}/{\omega }_{\mathrm{LO}}=2.8\%$. (b) Temperature dependence of $\eta$ for materials with different ${\omega }_{\mathrm{LO}}$ [50].