Observation of universal thermopolarization effect in insulators Preprint: APS/123-QED

Shuichi Iwakiri, Yasumitsu Miyata, Takao Mori · 2026

Read Observation of universal thermopolarization effect in insulators Preprint: APS/123-QED by Shuichi Iwakiri, Yasumitsu Miyata, Takao Mori (2026) 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-2605-17224.txt.

Introduction

Iwakiri, Miyata and Mori spot thermopolarization in a wide set of insulators. That is electric charge split made by a heat gradient. Heat swell under a gradient makes strain gradients that split charge through flexoelectricity. The effect shows up in crystals, plastics and glass alike. The size of the response tracks heat swell, and thin samples or phase shifts boost it more than tenfold.

Catalog record

CREATORShuichi Iwakiri, Yasumitsu Miyata, Takao Mori
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE23041
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2605.17224

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Observation of universal thermopolarization effect in insulators Preprint: APS/123-QED Shuichi Iwakiri, Yasumitsu Miyata, Takao Mori Abstract Heat-to-charge conversion has traditionally been realized via the Seebeck effect in conductors and pyroelectricity in polar insulators. Here, we demonstrate that temperature gradients generate electrical polarization, namely thermopolarization, in a wide range of insulators through a thermomechanical pathway. We identify a mechanism where thermal expansion under a temperature gradient produces strain gradients that induce polarization via the flexoelectric effect. Using a device with an on-chip heater, we detect the heat-induced polarization in crystalline, polymeric, and amorphous systems, including MgO, Al 2 O 3 , MnO, mica, PET, PEN, polyimide, and soda-lime glass. The magnitude of the response exhibits a robust scaling with the coefficient of thermal expansion, which is reproduced by finite-element simulations. Furthermore, we identify two routes to enhance the response: reducing the sample thickness and exploiting structural instabilities such as glass and antiferromagnetic phase transitions, where more than an order-of-magnitude enhancement is observed. These results establish a symmetry-independent route for heat-to-charge conversion in insulators and provide a device-compatible platform for electrically probing lattice responses, with potential for enhancement in nanoscale systems such as two-dimensional materials. Introduction– Thermoelectricity, or heat-to-charge conversion, is a fundamental phenomenon in condensed matter physics with broad technological importance. In conductors, temperature gradients universally generate electrical currents through the Seebeck effect (Fig. 1 (a) left), which serves as a powerful probe of entropy and electronic correlations in quantum materials 20 ; 32 ; 28 ; 5 . In insulators, where charge transport is absent, the established route for heat-to-charge conversion is pyroelectricity (Fig. 1 (a) bottom right). Here, temporal temperature variations induce electrical currents via changes in polarization, which can either be spontaneous polarization or induced one by external symmetry breaking such as interfaces or strain 17 ; 22 ; 16 ; 43 ; 23 ; 9 . Beyond pyroelectricity, whether a temperature gradient can directly generate electrical polarization has been an active topic of research, often referred to as thermopolarization. Various forms of thermopolarization have been proposed, including electronic polarization responses in solids 30 ; 29 and ionic polarization in liquids 34 ; 7 ; 42 . A particularly fundamental mechanism for insulating solids was proposed by Tagantsev 35 , based on the flexoelectric effect, namely the generation of electrical polarization P_{i} by strain gradients \partial\varepsilon_{jk}/\partial x_{l} , expressed as P_{i}=\mu_{ijkl}\partial\varepsilon_{jk}/\partial x_{l} ( i,j,k,l=x,y,z ). The flexoelectric tensor \mu_{ijkl} defines the response, and this effect is symmetry-allowed in materials of any symmetry 37 ; 45 ; 24 ; 40 ; 41 . Then, when a temperature gradient is applied to a material, the nonuniform thermal expansion generates a strain gradient in it, creating a polarization via the flexoelectric effect (Fig. 1 (a) top right). Although the original derivation was formulated for ionic crystals, both thermal expansion and flexoelectricity are ubiquitous in solids, suggesting a universal thermopolarization in insulating materials. However, this route of heat-to-charge conversion has generally been considered negligible 18 except in certain ferroelectrics and quantum paraelectrics 21 ; 11 ; 12 ; 10 ; 38 ; 39 ; 36 ; 44 , and experimental investigation of its generality has been lacking. Therefore, whether thermopolarization universally exists in insulating solids, how it can be detected, and whether it can be enhanced remain open questions. In this work, we experimentally demonstrate that temperature gradients generate electrical polarization in a wide range of insulating materials through the flexoelectric effect. Using an on-chip heater to create temperature gradients and the resulting strain gradients, we induce polarization in centrosymmetric crystalline, polymeric, and amorphous systems. The magnitude of the response scales with the coefficient of thermal expansion, which is quantitatively reproduced by finite-element simulations. We further identify two pathways for enhancing thermopolarization: reducing the sample thickness and exploiting structural instabilities such as glass and antiferromagnetic phase transitions. These results establish an insulating analogue of the Seebeck effect and provide a route to probe materials previously regarded as electrically inactive. Figure 1: (a) Comparison of various mechanism of heat-to-charge conversion. Left: the Seebeck effect. Top right: thermopolarization (this work). Bottom Right: pyroelectricity. (b) Demonstration of the voltage generation across a glass sample with applied temperature gradient. Results– Figure 1 (b) shows a demonstration of thermopolarization in a soda-lime glass substrate. An on-chip heater and metallic electrodes are fabricated on the top and bottom surfaces of a 1-mm-thick substrate. The heater and the top detector electrode have the same geometry as those in Fig. 2 (a), which will be discussed in detail later. The sample is mounted on a room-temperature stage, which fixes the bottom boundary condition. When a current is applied to the heater, a temperature gradient is generated across the substrate. As a result, the temperature gradient develops both in-plane and out-of-plane, giving rise to corresponding strain gradients. In our setup, the detected response is dominated by the out-of-plane polarization generated via the flexoelectric effect, which can be expressed as P_{z}\sim\mu\frac{\partial}{\partial z}\left(\varepsilon_{xx}+\varepsilon_{yy}\right)\sim\mu\alpha\nabla T, (1) where P_{z} is the out-of-plane polarization, \varepsilon_{xx} and \varepsilon_{yy} are the in-plane strains, \alpha is the coefficient of thermal expansion (CTE), and \nabla T denotes the temperature gradient. We also approximate the flexoelectric tensor as \mu_{ijkl}\simeq\mu_{zjkz}\simeq\mu , under the assumption that, for the present geometry, the measured response is predominantly governed by the out-of-plane polarization. This approximation is justified by considering that the strain gradients vary primarily along the thickness direction, while the leading strain components are in-plane. We apply an AC current I=I_{\mathrm{heater}}\sin\omega t with \omega/2\pi=1~\mathrm{kHz} , which produces Joule heating and a temperature gradient \nabla T(t)\propto I_{\mathrm{heater}}^{2}\sin^{2}(\omega t)\propto 1-\cos(2\omega t) . Then, the thermopolarization is detected as a second harmonic voltage across the sample. We indeed observe such voltage generation as shown in Fig. 1 (b), demonstrating the presence of thermopolarization. While the voltage generation marks a signature of thermopolarization, for a more quantitative and systematic analysis of thermopolarization, it is preferable to measure the current induced in the detector by the temporal variation of the polarization, since the voltage across the substrate is strongly influenced by the specific capacitance characteristics of the device. Figure 2 (a) shows the device consisting of an on-chip heater and a metallic detector wire on a soda-lime glass. The heater is again driven by an alternating current I=I_{\mathrm{heater}}\sin\omega t . Under such modulation, the change in the induced polarization rather than the polarization itself generates a screening current in the detector electrode, \frac{dP}{dt}=\frac{dP}{dT}\frac{dT}{dt}\propto\frac{d}{dT}\left(\mu\frac{\partial\varepsilon_{xx,yy}}{\partial z}\right)\frac{dT}{dt}. (2) Since the temperature follows T(t)\propto I_{\mathrm{heater}}^{2}\sin^{2}(\omega t) , its time derivative becomes \frac{dT}{dt}\propto I_{\mathrm{heater}}^{2}\sin(2\omega t) . We observe such a current generation by measuring the second-harmonic current response I_{\mathrm{2\omega}} . Figure 2 (b) shows that both the in-phase ( X ) and out-of-phase ( Y ) components of I_{2\omega} increase quadratically with I_{\mathrm{heater}} , in agreement with the above discussion. Unless otherwise noted, all data are obtained at \omega/2\pi=1~\mathrm{kHz} , and the substrate is placed on a metallic stage which is kept at room temperature. The heater resistance R is typically 150 \Omega . In addition, we conduct multiple control experiments to exclude spurious effects such as capacitive coupling and thermally-stimulated current (see END MATTER). We also find that the lock-in phase and its frequency dependence encode the thermal origin of the generated signal. In

Read the complete text