Giant radiation heat transfer through micron gaps

Near-field heat transfer between two closely spaced media can exceed in orders radiation through the interface of a single black body. This effect is caused by exponentially decaying (evanescent) waves, which form the photon tunnel between two transparent boundaries or boundaries supporting surface waves. However, in the infrared range this huge energy transfer holds only in the case when the gap between two media is as small as a few tens of nanometers. We propose a different paradigm of the radiation heat transfer, which makes possible the strong photon tunneling (and therefore giant radiative heat transfer) for micron-thick gaps. For it the air gap between two media should be modified, so that evanescent waves are transformed inside it into propagating ones. This modification is possible without increasing the phonon thermal conductance between two media. The suggested structure containing arrays of nanotubes or nanowires is designed so that the photovoltaic conversion of the transferred heat will not be altered by these arrays.

Figure 1

(a) Illustration of the general problem formulation. (b) Interdigital arrangement of CNT allows us to get rid of the thermal conductance.

Figure 2

Transmission coefficient versus $k_x{a}_1$, calculated for $d=100$ nm. Blue curve corresponds to the vacuum gap, where the radiation heat transfer is supported by evanescent waves. Red (solid) curve shows transmission through the gap filled with CNTs, calculated using Green's-function method, black (dashed) curve corresponds to the effective-medium theory.

Figure 3

The same, as in Fig. 2, calculated for $d=1000$ nm.

Figure 4

Comparison between results of the effective-medium model (dashed black curve) and the GFM (solid red curve). Slow-wave factor versus normalized spatial frequency is calculated for $a=20$ nm.

Figure 5

Ratio $q_{\lambda }^K/q_{\lambda p}$, calculated for the vacuum gap (blue) and the gap filled with CNTs (red and black curves), $d=100$ nm (logarithmic scale). All parameters are the same as in Fig. 2. Red curve corresponds to the GFM, black curve corresponds to the EMM.

Figure 6

Ratio $q_{\lambda }^K/q_{\lambda p}$, calculated for the vacuum gap (blue) and the gap filled with CNTs (red and black curves), $d=1\mu$m (logarithmic scale). All parameters are the same as in Fig. 3. Red curve corresponds to the GFM, black curve corresponds to the EMM.

Figure 7

Ratio $q^K\left(\lambda \right)/q_p\left(\lambda \right)$, for fixed $K=0.5\pi /a$ versus $\lambda$, calculated for a vacuum gap $d=1\mu$m (dashed blue) and for the same gap $d=1\mu$m filled with CNTs (solid red). Logarithmic scale. Structure parameters are the same as in Fig. 6. Calculations using the EMM give the same result as those obtained with the GFM.