Thermal radiation between parallel objects separated by deep subwavelength distances and subject to large thermal gradients (>100 K) can reach very high magnitudes, while being concentrated on a narrow frequency distribution. These unique characteristics could enable breakthrough technologies for thermal transport control and electricity generation (for example, by radiating heat exactly at the bandgap frequency of a photovoltaic cell). However, thermal transport in this regime has never been achieved experimentally due to the difficulty of maintaining large thermal gradients over nanometre-scale distances while avoiding other heat transfer mechanisms, namely conduction. Here, we show near-field radiative heat transfer between parallel SiC nanobeams in the deep subwavelength regime. The distance between the beams is controlled by a high-precision micro-electromechanical system (MEMS). We exploit the mechanical stability of nanobeams under high tensile stress to minimize thermal buckling effects, therefore keeping control of the nanometre-scale separation even at large thermal gradients. We achieve an enhancement of heat transfer of almost two orders of magnitude with respect to the far-field limit (corresponding to a 42 nm separation) and show that we can maintain a temperature gradient of 260 K between the cold and hot surfaces at ∼100 nm distance.
Near-field heat transfer recently attracted growing interest but was demonstrated experimentally only in macroscopic systems. However, several projected applications would be relevant mostly in integrated nanostructures. Here we demonstrate a platform for near-field heat transfer on-chip and show that it can be the dominant thermal transport mechanism between integrated nanostructures, overcoming background substrate conduction and the far-field limit (by factors 8 and 7, respectively). Our approach could enable the development of active thermal control devices such as thermal rectifiers and transistors.
Photons are neutral particles that do not interact directly with a magnetic field. However, recent theoretical work has shown that an effective magnetic field for photons can exist if the phase of light changes with its direction of propagation. This direction-dependent phase indicates the presence of an effective magnetic field, as shown experimentally for electrons in the Aharonov-Bohm experiment. Here, we replicate this experiment using photons. To create this effective magnetic field we construct an on-chip silicon-based Ramsey-type interferometer. This interferometer has been traditionally used to probe the phase of atomic states and here we apply it to probe the phase of photonic states. We experimentally observe an effective magnetic flux between 0 and 2 pi corresponding to a non-reciprocal 2 pi phase shift with an interferometer length of 8.35 mm and an interference-fringe extinction ratio of 2.4 dB. This non-reciprocal phase is comparable to those of common monolithically integrated magneto-optical materials.
We demonstrate electrically driven nonreciprocity on a silicon chip. By achieving an indirect interband photonic transition, we show that the transmission coefficients between two single-mode waveguides become dependent on the propagation directions only in the presence of the electrical drive. Our structure is characterized by a nonsymmetric scattering matrix identical to a linear magneto-optical device.
We measure near field radiative cooling of a thermally isolated nanostructure up to a few degrees and show that in principle this process can efficiently cool down localized hotspots by tens of degrees at submicrometer gaps. This process of cooling is achieved without any physical contact, in contrast to heat transfer through conduction, thus enabling novel cooling capabilities. We show that the measured trend of radiative cooling agrees well theoretical predictions and is limited mainly by the geometry of the probe used here as well as the minimum separation that could be achieved in our setup. These results also pave the way for realizing other new effects based on resonant heat transfer, like thermal rectification and negative thermal conductance.
Optical isolators are essential components in optical networks and are used to eliminate parasite reflections that are detrimental to the stability of the optical systems. The challenge in realizing optical isolators lies in the fact that in standard optoelelectronic materials, including most semiconductors and metals, Maxwell’s equations, which governs the propagation of light, are constraint by reciprocity or time-reversal symmetry. As a result, standard optical devices on chip, including some of the passive metallo-dielectric structures recently explored for isolation purposes, all have a symmetric scattering matrix and therefore fundamentally cannot function as an optical isolator. In order to break this symmetry, traditional optical isolators rely upon magneto-optical effects, which require materials that are difficult to integrate in current micro-electronic systems. The breaking of this symmetry without the need for magnetic materials has been a long-term goal in photonics. Nonlinear structures for optical isolation have been explored. However, these typically provide optical isolation only within a particular range of operating power. Here we create a non-magnetic CMOS-compatible optical isolator on a silicon chip. The isolator is based on indirect interband photonic transition, induced by electrically-driven dynamic refractive index modulation. We demonstrate an electrically-induced non-reciprocity: the transmission coefficients between two single-mode waveguides become dependent on the propagation directions only in the presence of the electrical drive. The contrast ratio between forward and backward directions exceeds 30dB in simulations. We experimentally observe a strong contrast (up to 3 dB) limited only by our electrical setup, for a continuous-wave (CW) optical signal. Importantly, the device is linear with respect to signal light. The observed contrast ratio is independent of the timing, the format, the amplitude and the phase of the input signal.