Memory effects in time-modulated radiative heat transfer
R. Messina, P. Ben-Abdallah · 2025
Read Memory effects in time-modulated radiative heat transfer by R. Messina, P. Ben-Abdallah (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2510-19378.txt.
Introduction
Riccardo Messina and Philippe Ben-Abdallah build a theory of heat transfer by radiation. It covers many-body systems whose optical properties change over time. Heat exchange is usually treated at steady equilibrium. Time modulation breaks that picture. It adds memory effects and unsteady correlations.
Their formulas give heat currents for modulated systems. A Landauer-style picture covers pairwise exchange. It counts all frequency-changing channels. Near resonance, modulation boosts random heat motion across sidebands. The effect works like a tuned amplifier. It allows active control of heat flow by frequency, at the nanoscale.
Catalog record
| CREATOR | R. Messina, P. Ben-Abdallah |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 23 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2510-19378
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Memory effects in time-modulated radiative heat transfer R. Messina, P. Ben-Abdallah Abstract We develop a general theory of radiative heat exchange between dipoles with time-modulated optical properties. This framework extends fluctuational electrodynamics beyond equilibrium by incorporating nonstationary correlations and memory effects induced by temporal modulation. Closed-form expressions for the heat currents in modulated many-body systems are obtained, together with a generalized Landauer-like formulation of the pairwise exchanges, where the transmission coefficient accounts for all inelastic frequency-conversion channels. Near-resonant modulation redistributes and amplifies thermal fluctuations across Floquet sidebands, acting as a parametric amplifier of thermal radiation and enabling active, frequency-selective control of nanoscale heat transfer. pacs 44.40.+a, 78.20.N-, 03.50.De, 66.70.-f Controlling thermal radiation through temporal modulation of optical properties offers new opportunities to overcome the fundamental limits imposed by time-invariance and reciprocity. In a pioneering work Liberal , Vázquez-Lozano and Liberal developed a quantum framework for thermal emission in time-modulated solids, predicting radiation beyond the blackbody limit and even vacuum amplification. Subsequently, Yu and Fan Fan1 demonstrated that temporal modulation of optical properties in surface-polariton systems enables control of spatial coherence, paving the way for coherent thermal emission. In the slow-modulation regime, Latella et al. Latella ; Ma analyzed heat shuttling driven by periodic modulation of temperature or chemical potential, revealing enhanced transfer and negative differential thermal resistance. Messina and Ben-Abdallah Messina proposed radiative heat pumping via periodic modulation, while Ben-Abdallah and Biehs Biehs identified Berry-like geometric phases as a mechanism for nonreciprocal thermal relaxation. Using Floquet theory, Fernández-Alcázar et al. Alcazar showed that temporal modulation can induce directional heat flow between reservoirs at equal temperatures. Beyond the adiabatic limit, Fan and Yu Fan2 introduced a time-dependent dyadic Green’s function formalism that enables active control of radiative heat transfer—including enhancement, suppression, and reversal—and later proposed photonic refrigeration through thermal photon up-conversion Fan3 . Despite these advances, a comprehensive framework for describing strongly coupled, many-body, and far-from-equilibrium modulated systems is still lacking. In this Letter, we develop such a framework by extending fluctuational electrodynamics to time-modulated many-body systems and generalizing the fluctuation–dissipation theorem beyond equilibrium. We consider a set of N finite-size thermal emitters at temperatures T_{i} ( i=1,2,\dots,N ) immersed in a thermal bath at temperature T_{\mathrm{b}} . The objects are assumed small compared to their thermal wavelength, so they can be described as electric dipoles. We further assume that their polarizabilities can be actively modulated in time. The spectral component of each dipole moment can be decomposed as \mathbf{p}_{i}(\omega)=\mathbf{p}^{\mathrm{(fl)}}_{i}(\omega)+\mathbf{p}^{\mathrm{(ind)}}_{i}(\omega)+\delta\mathbf{p}^{\mathrm{(ind)}}_{i}(\omega), (1) where \mathbf{p}^{\mathrm{(fl)}}_{i}(\omega) is the fluctuating part of dipolar moment due to thermal fluctuations, whereas \mathbf{p}^{\mathrm{(ind)}}_{i}(\omega) and \delta\mathbf{p}^{\mathrm{(ind)}}_{i}(\omega) correspond to the induced parts of the dipolar moment, due to the surrounding emitters and thermal bath, induced by the polarizability \alpha_{i} and its time-modulated part \delta\alpha_{i} , respectively. The first two terms correspond to the usual terms in the fluctuational-electrodynamics theory of many-body systems pba2011 ; messina2013 ; biehs1 . On the other hand, the third term results from the external modulation of the polarizabilities. This driven term is related to the local electric field by an expression of the following form Monticone ; Mirmoosa \delta\mathbf{p}^{\mathrm{(ind)}}_{i}(\omega)=\varepsilon_{0}\int d\Omega\,\delta\alpha_{i}(\Omega,\omega-\Omega)\mathbf{E}^{\mathrm{(loc)}}_{i}(\Omega), (2) where \delta\alpha_{i} is polarizability kernel which depends on the external driving of optical properties. Here, \mathbf{E}^{\mathrm{(loc)}}_{i}(\Omega) is the local field taking into account the self-radiated part by the particle under the external modulation, which reads \mathbf{E}^{\mathrm{(loc)}}_{i}(\Omega)=\mathbf{E}_{\mathrm{b}i}(\Omega)+\mu_{0}\Omega^{2}\sum_{j=1}^{N}\mathds{G}_{0}(\mathbf{r}_{i},\mathbf{r}_{j},\Omega)\mathbf{p}_{j}(\Omega), (3) where \mathbf{E}_{\mathrm{b}i}(\Omega) is the field at the location of dipole i coming from the thermal bath while the second term denotes the field resulting from all dipoles, including the i -th dipole and \mathds{G}_{0}(\mathbf{r}_{i},\mathbf{r}_{j},\Omega) is the propagator in vacuum. It is worthwhile to note that unlike \mathbf{\delta p}_{i}(\omega) , \mathbf{p}^{\mathrm{(ind)}}_{i}(\omega) is defined from the surrounding dipoles and from the field radiated by the external bath, i.e. \begin{split}\mathbf{p}^{\mathrm{(ind)}}_{i}(\omega)&=\alpha_{i}(\omega)\Bigl(\frac{\omega}{c}\Bigr)^{2}\sum_{j\neq i}\mathds{G}_{0}(\mathbf{r}_{i},\mathbf{r}_{j},\omega)\mathbf{p}_{j}(\omega)\\ &\,+\varepsilon_{0}\alpha_{i}(\omega)\mathbf{E}_{\mathrm{b}i}(\omega).\end{split} (4) It is important to note that the spectral convolution in Eq. ( 2 ) introduces a memory-like effect. Indeed, the dipolar moment induced at the frequency \omega by the modulation of optical properties depends on the local field undergone by the particle over the full spectral range. Combining Eqs. ( 1 ), ( 2 ) and ( 4 ), we see that the the spectral components of dipolar moments are solution of the system of Fredholm’s integral equation \begin{split}&\mathbf{p}_{i}(\omega)-\alpha_{i}(\omega)\Bigl(\frac{\omega}{c}\Bigr)^{2}\sum_{j\neq i}\mathds{G}_{0}(\mathbf{r}_{i},\mathbf{r}_{j},\omega)\mathbf{p}_{j}(\omega)\\ &\,-\int d\Omega\,\mathbf{\delta\alpha}_{i}(\Omega,\omega-\Omega)\Bigl(\frac{\Omega}{c}\Bigr)^{2}\sum_{j=1}^{N}\mathds{G}_{0}(\mathbf{r}_{i},\mathbf{r}_{j},\Omega)\mathbf{p}_{j}(\Omega)\\ &=\mathbf{p}^{\mathrm{(fl)}}_{i}(\omega)+\varepsilon_{0}\alpha_{i}(\omega)\mathbf{E}_{\mathrm{b}i}(\omega)\\ &\,+\varepsilon_{0}\int d\Omega\,\mathbf{\delta\alpha}_{i}(\Omega,\omega-\Omega)\mathbf{E}_{\mathrm{b}i}(\Omega).\end{split} (5) This equation forms the foundation of radiative heat transfer in time-modulated many-body systems. It can be written in a compact block-matrix form and solved perturbatively to arbitrary order in \delta\alpha , allowing calculation of dipole correlations and the resulting thermal emission spectrum. By distinguishing in the integral term the contribution coming from the surrounding dipoles (i.e. \mathbf{r}_{i}\neq\mathbf{r}_{j} ) from the self contribution of each particle and by rewriting the previous equation into a block integral equation we get \begin{split}&\mathds{A}(\omega)\mathbf{p}(\omega)-\int d\Omega\,\mathds{K}(\Omega,\omega-\Omega)\mathbf{p}(\Omega)=\mathbf{S}(\omega)+\delta\mathbf{S}(\omega),\end{split} (6) where we have set the (3\times N) -dimension dipolar moment \mathbf{p}(\omega)=(\mathbf{p}_{1}(\omega),\dots,\mathbf{p}_{N}(\omega))^{T} and the source terms \mathbf{S}(\omega)=(\varepsilon_{0}\alpha_{1}(\omega)\mathbf{E}_{\mathrm{b}1}(\omega)+\mathbf{p}^{\mathrm{(fl)}}_{1}(\omega),\dots,\varepsilon_{0}\alpha_{N}(\omega)\mathbf{E}_{\mathrm{b}N}(\omega)+\mathbf{p}^{\mathrm{(fl)}}_{N}(\omega))^{T} and \delta\mathbf{S}(\omega)=\varepsilon_{0}\bigl(\int d\Omega\,\mathbf{\delta\alpha}_{1}(\Omega,\omega-\Omega)\mathbf{E}_{\mathrm{b}1}(\Omega),\dots,\int d\Omega\,\mathbf{\delta\alpha}_{N}(\Omega,\omega-\Omega)\mathbf{E}_{\mathrm{b}N}(\Omega)\bigr)^{T} , while the block matrices \mathds{A} and \mathds{K} are defined as follows \begin{split}\mathds{A}_{ij}(\omega)&=\delta_{ij}\mathds{1}+(1-\delta_{ij})\alpha_{i}(\omega)\Bigl(\frac{\omega}{c}\Bigr)^{2}\mathds{G}_{0}(\mathbf{r}_{i},\mathbf{r}_{j},\omega),\\ \mathds{K}_{ij}(\omega,\omega^{\prime})&=\delta\alpha_{i}(\omega,\omega^{\prime})\frac{\omega^{2}}{c^{2}}\Bigl[\delta_{ij}\mathds{G}_{0}(\mathbf{r}_{i},\mathbf{r}_{j}\to\mathbf{r}_{i},\omega)\\ &\,+(1-\delta_{ij})\mathds{G}_{0}(\mathbf{r}_{i},\mathbf{r}_{j},\omega)\Bigr],\end{split} (7) where, according to Ref. Lakhtakia , the propagator in vacuum asymptotically reads \mathds{G}_{0}(\mathbf{r},\mathbf{r}^{\prime}\to\mathbf{r},\omega)=G_{0}(\omega)\mathds{1}=\Bigl(-\frac{c^{2}}{3\omega^{2}V}+i\frac{\omega}{6\pi c}\Bigr)\mathds{1}, (8) where V is the particle volume. The dipolar moments can be calculated in a perturbative way using the ansatz \mathbf{p}(\omega)=\sum_{n=0}^{\infty}\mathbf{p}^{(n)}(\omega), (9) where \mathbf{p}^{(n)}(\omega) is of order n in the perturbation \delta\alpha . By inserting this expression into Eq. ( 6 ) we get for the unmodulated system \mathbf{p}^{(0)}(\omega)=\mathds{A}^{-1}(\omega)\mathbf{S}(\omega) , whereas for order 1 and any n>1 we have \begin{split}\mathbf{p}^{(1)}(\omega)&=\mathds{A}^{-1}(\omega)\Biggl[\int d\Omega\,\mathds{K}(\Omega,\omega-\Omega)\mathbf{p}^{(0)}(\Omega)+\delta\mathbf{S}(\omega)\Biggr]\\ \mathbf{p}^{(n)}(\omega)&=\mathds{A}^{-1}(\omega)\int d\Omega\,\mathds{K}(\Omega,\omega-\Omega)\mathbf{p}^{(n-1)}(\Omega).\end{split} (10) These recursive relations allow us to write a general expression relating the n^{\mathrm{th}} order ot the dipolar moment to the dipolar moments {\mathbf{p}}^{(0)} of the unmodulated system and to the field \mathbf{E}_{\mathrm{b}} associated to the external bath by the following expression \mathbf{p}^{(n)}(\omega)=\mathds{O}^{(n)}(\omega)\mathbf{p}^{(0)}+\mathds{O}^{(n-1)}(\omega)\delta\mathbf{S}, (11) where we have defined the integral operator \begin{split}{\mathds{O}}^{(n)}(\omega)\mathbf{f}&\equiv\int d\Omega_{1}\,\mathds{P}(\omega,\Omega_{1})\int d\Omega_{2}\,\mathds{P}(\Omega_{1},\Omega_{2})\times\dots\\ &\,\times\int d\Omega_{n}\,\mathds{P}(\Omega_{n-1},\Omega_{n})\mathbf{f}(\Omega_{n}),\end{split} (12) being \mathds{P}(\omega,\omega^{\prime})=\mathds{A}^{-1}(\omega)\mathds{K}(\omega^{\prime},\omega-\omega^{\prime}) . The cross-correlation functions of dipolar moments in the modulated system can thus also be formally expressed as perturbative expansions over the modulation order, generalizing the standard correlations in a static system. In particular, the background field and intrinsic dipole fluctuations obey the usual fluctuation-dissipation relations biehs1 ; messina2013 ; pba2024 . The explicit expressions, including the component-wise structure and all higher-order terms, are developed in detail in the Supplementary Material SupplMat . Figure 1: Dipolar corrections of order n in the scalar approximation with (a) a sinusoidal variation of \delta\alpha of period \tau=2\pi/\Omega and with (b) a rectangular variation \delta\alpha(\omega,t) with the same period ( \delta\alpha(\omega,t) is equal to one when \tau\in[0,\tau/2] and zero elsewhere and it is decomposed in 20 Fourier modes). Here we take \omega_{0}=10^{15}\,\mathrm{rad}\cdot\mathrm{s}^{-1} , \gamma=5\cdot 10^{13}\,\mathrm{s}^{-1} , \alpha_{0}=0.001 , \tilde{p}=0.01 and the temperature is T=300\, K. Here, the modulation frequency coincides with the resonance frequency of the Lorentzian (i.e. \Omega=\omega_{0} ). Inset: temporal variation of \delta\alpha . To describe how the polarizability modulations impact the dipolar moments, we consider below a simple scenario in which \delta\alpha is a \tau -periodic function ( \tau=2\pi/\Omega ) and thus can be written as \delta\alpha(\omega,\omega^{\prime})=\delta\hat{\alpha}(\omega)\underset{n}{\sum}\beta_{n}\delta(\omega^{\prime}+n\Omega), (13) where \beta_{n} are the Fourier components of the temporal variation of \delta\alpha(\omega,t) . We start from the case of a single scalar dipole. By employing Eq. ( 11 ) and neglecting the external bath, the dipolar correction of order m\geq 1