Polaritonic Neuromorphic Computing Outperforms Linear Classifiers

Selective Excitation of Exciton-Polariton Condensate Modes in an Annular Perovskite Microcavity

Exciton-polariton systems composed of a light-matter quasi-particle with a light effective mass easily realize Bose-Einstein condensation. In this work, we constructed an annular trap in a halide perovskite semiconductor microcavity and observed the spontaneous formation of symmetrical petal-shaped exciton-polariton condensation in the annular trap at room temperature. In our study, we found that the number of petals of the petal-shaped exciton-polariton condensates, which is decided by the orbital angular momentum, is dependent on the light intensity distribution. Therefore, the selective excitation of perovskite microcavity exciton-polariton condensates under all-optical control can be realized by adjusting the light intensity distribution. This could pave the way to room-temperature topological devices, optical cryptographical devices, and new quantum gyroscopes in the exciton-polariton system.

Direct Writing of Room Temperature Polariton Condensate Lattice

Realizing lattices of exciton polariton condensates has been of much interest owing to the potential of such systems to realize analogue Hamiltonian simulators and physical computing architectures. Here, we report the realization of a room temperature polariton condensate lattice using a direct-write approach. Polariton condensation is achieved in a microcavity embedded with host-guest Frenkel excitons of an organic dye (rhodamine) in a small-molecule ionic isolation lattice (SMILES). The microcavity is patterned using focused ion beam etching to realize arbitrary lattice geometries, including defect sites on demand. The band structure of the lattice and the emergence of condensation are imaged using momentum-resolved spectroscopy. The introduction of defect sites is shown to lower the condensation threshold and result in the formation of a defect band in the condensation spectrum. The present approach allows us to study periodic, quasiperiodic, and disordered polariton condensate lattices at room temperature using a direct-write approach.

Controlling the Spatial Profile and Energy Landscape of Organic Polariton Condensates in Double-Dye Cavities

The development of high-speed, all-optical polariton logic devices underlies emerging unconventional computing technologies and relies on advancing techniques to reversibly manipulate the spatial extent and energy of polartion condensates. We investigate active spatial control of polariton condensates independent of the polariton, gain-inducing excitation profile. This is achieved by introducing an extra intracavity semiconductor layer, nonresonant to the cavity mode. Partial saturation of the optical absorption in the uncoupled layer enables the ultrafast modulation of the effective refractive index and, through excited-state absorption, the polariton dissipation. Utilizing an intricate interplay of these mechanisms, we demonstrate control over the spatial profile, density, and energy of a polariton condensate at room temperature.

Bose Condensation of Upper-Branch Exciton-Polaritons in a Transferable Microcavity

Exciton-polaritons are composite quasiparticles that result from the coupling of excitonic transitions and optical modes. They have been extensively studied because of their quantum phenomena and potential applications in unconventional coherent light sources and all-optical control elements. In this work, we report the observation of Bose-Einstein condensation of the upper polariton branch in a transferable WS2 monolayer microcavity. Near the condensation threshold, we observe a nonlinear increase in upper polariton intensity accompanied by a decrease in line width and an increase in temporal coherence, all of which are hallmarks of Bose-Einstein condensation. Simulations show that this condensation occurs within a specific particle density range, depending on the excitonic properties and pumping conditions. The manifestation of upper polariton condensation unlocks new possibilities for studying the condensate competition while linking it to practical realizations in polaritonic lasers. Our findings contribute to the understanding of bosonic systems and offer potential for the development of polaritonic devices.

Effects of Floquet Engineering on the Coherent Exciton Dynamics in Monolayer WS2

Coherent optical manipulation of electronic bandstructures via Floquet Engineering is a promising means to control quantum systems on an ultrafast time scale. However, the ultrafast switching on/off of the driving field comes with questions regarding the limits of the Floquet formalism (which is defined for an infinite periodic drive) through the switching process and to what extent the transient changes can be driven adiabatically. Experimentally addressing these questions has been difficult, in large part due to the absence of an established technique to measure coherent dynamics through the duration of the pulse. Here, using multidimensional coherent spectroscopy we explicitly excite, control, and probe a coherent superposition of excitons in the K and K' valleys in monolayer WS₂. With a circularly polarized, red-detuned pump pulse, the degeneracy of the K and K' excitons can be lifted, and the phase of the coherence rotated. We directly measure phase rotations greater than π during the 100 fs driving pulse and show that this can be described by a combination of the AC-Stark shift of excitons in one valley and the Bloch-Siegert shift of excitons in the opposite valley. Despite showing a smooth evolution of the phase that directly follows the intensity envelope of the nonresonant pump pulse, the process is not perfectly adiabatic. By measuring the magnitude of the macroscopic coherence as it evolves before, during, and after the nonresonant pump pulse we show that there is additional decoherence caused by power broadening in the presence of the nonresonant pump. This nonadiabaticity arises as a result of interactions with the otherwise adiabatic Floquet states and may be a problem for many applications, such as manipulating qubits in quantum information processing; however, these measurements also suggest ways such effects can be minimized or eliminated.

Directional Emission from Electrically Injected Exciton-Polaritons in Perovskite Metasurfaces

We present a new approach to achieving strong coupling between electrically injected excitons and photonic bound states in the continuum of a dielectric metasurface. Here a high-finesse metasurface cavity is monolithically patterned in the channel of a perovskite light-emitting transistor to induce a large Rabi splitting of ∼200 meV and more than 50-fold enhancement of the polaritonic emission compared to the intrinsic excitonic emission of the perovskite film. Moreover, the directionality of polaritonic electroluminescence can be dynamically tuned by varying the source-drain bias, which induces an asymmetric distribution of exciton population within the transistor channel. We argue that this approach provides a new platform to study strong light-matter interactions in dispersion engineered photonic cavities under electrical injection and paves the way to solution-processed electrically pumped polariton lasers.

Optically trapped room temperature polariton condensate in an organic semiconductor

The strong nonlinearities of exciton-polariton condensates in lattices make them suitable candidates for neuromorphic computing and physical simulations of complex problems. So far, all room temperature polariton condensate lattices have been achieved by nanoimprinting microcavities, which by nature lacks the crucial tunability required for realistic reconfigurable simulators. Here, we report the observation of a quantised oscillating nonlinear quantum fluid in 1D and 2D potentials in an organic microcavity at room temperature, achieved by an on-the-fly fully tuneable optical approach. Remarkably, the condensate is delocalised from the excitation region by macroscopic distances, leading both to longer coherence and a threshold one order of magnitude lower than that with a conventional Gaussian excitation profile. We observe different mode selection behaviour compared to inorganic materials, which highlights the anomalous scaling of blueshift with pump intensity and the presence of sizeable energy-relaxation mechanisms. Our work is a major step towards a fully tuneable polariton simulator at room temperature.

Only-train-electrical-to-optical-conversion (OTEOC): simple diffractive neural networks with optical readout

Machine learning hardware based on optical diffraction is emerging as a new computing platform with high throughput and low latency. The current all-optical diffractive deep neural networks often suffer from complex optical configuration, lack of efficient optical nonlinear activation, and critical alignment between optical layers for system integration. The opto-electronic diffractive neural networks can partially address these issues by shifting some computation load, e.g., nonlinear activation and adaptive training, to the electronic domain. However, these hybrid networks require extra optical-to-electrical conversion that inevitably slows the overall process down. Here, we propose a simple opto-electronic diffractive neural network with just one optical layer enabled by a standard phase-only spatial light modulator. The proposed system can classify images by optical readout and does not need to collect the light distribution for subsequent electronic computation. The nonlinear function is intrinsically integrated in the essential encoding process from the electronic input to the modulated wavefront of light. Thanks to its simplicity, the system can reach high classification accuracy without calibration and can be reconfigured by updating the weights without changing or moving any physical component. We believe this technology brings diffractive neural networks a step closer to building realistic optics-based neurocomputers.

Self-Hybridized Exciton-Polaritons in Sub-10-nm-Thick WS2 Flakes: Roles of Optical Phase Shifts at WS2/Au Interfaces

Exciton–polaritons (EPs) can be formed in transition metal dichalcogenide (TMD) multilayers sustaining optical resonance modes without any external cavity. The self-hybridized EP modes are expected to depend on the TMD thickness, which directly determines the resonance wavelength. Exfoliated WS₂ flakes were prepared on SiO₂/Si substrates and template-stripped ultraflat Au layers, and the thickness dependence of their EP modes was compared. For WS₂ flakes on SiO₂/Si, the minimum flake thickness to exhibit exciton–photon anticrossing was larger than 40 nm. However, for WS₂ flakes on Au, EP mode splitting appeared in flakes thinner than 10 nm. Analytical and numerical calculations were performed to explain the distinct thickness-dependence. The phase shifts of light at the WS₂/Au interface, originating from the complex Fresnel coefficients, were as large as π/2 at visible wavelengths. Such exceptionally large phase shifts allowed the optical resonance and resulting EP modes in the sub-10-nm-thick WS₂ flakes. This work helps us to propose novel optoelectronic devices based on the intriguing exciton physics of TMDs.

Neuromorphic System for Edge Information Encoding: Emulating Retinal Center-Surround Antagonism by Li-Ion-Mediated Highly Interactive Devices

Center-surround antagonism, a key mechanism in the retina, contributes to the encoding of edge contrast rather than of the overall information on a visual image. Here, a neuromorphic system consisting of multiple ionic devices is built, where each device has a lithium cobalt oxide channel arranged on a common lithium phosphorus oxynitride electrolyte. Because of the migration of Li ions between the channels through the electrolyte, the devices are highly interactive, as is seen with retinal neurons. On the basis of the excitation of single devices and device-to-device inhibition, the system successfully emulates the antagonistic center-surround receptive field and the Mach band effect in which perceived contrast is enhanced at the edges between dark and bright regions. Furthermore, a two-dimensional array system is simulated to implement edge detection for real images. This scheme enables computer vision tasks with simple and effective operations, owing to the intrinsic properties of the materials employed.

Perovskite semiconductors for room-temperature exciton-polaritonics

Lead-halide perovskites are generally excellent light emitters and can have larger exciton binding energies than thermal energy at room temperature, exhibiting great promise for room-temperature exciton-polaritonics. Rapid progress has been made recently, although challenges and mysteries remain in lead-halide perovskite semiconductors to push polaritons to room-temperature operation. In this Perspective, we discuss fundamental aspects of perovskite semiconductors for exciton-polaritons and review the recent rapid experimental advances using lead-halide perovskites for room-temperature polaritonics, including the experimental realization of strong light-matter interaction using various types of microcavities as well as reaching the polariton condensation regime in planar microcavities and lattices.

Exciton-Photonics: From Fundamental Science to Applications

Semiconductors in all dimensionalities ranging from 0D quantum dots and molecules to 3D bulk crystals support bound electron-hole pair quasiparticles termed excitons. Over the past two decades, the emergence of a variety of low-dimensional semiconductors that support excitons combined with advances in nano-optics and photonics has burgeoned an advanced area of research that focuses on engineering, imaging, and modulating the coupling between excitons and photons, resulting in the formation of hybrid quasiparticles termed exciton-polaritons. This advanced area has the potential to bring about a paradigm shift in quantum optics, as well as classical optoelectronic devices. Here, we present a review on the coupling of light in excitonic semiconductors and previous investigations of the optical properties of these hybrid quasiparticles via both far-field and near-field imaging and spectroscopy techniques. Special emphasis is given to recent advances with critical evaluation of the bottlenecks that plague various materials toward practical device implementations including quantum light sources. Our review highlights a growing need for excitonic material development together with optical engineering and imaging techniques to harness the utility of excitons and their host materials for a variety of applications.

Neuromorphic Binarized Polariton Networks

The rapid development of artificial neural networks and applied artificial intelligence has led to many applications. However, current software implementation of neural networks is severely limited in terms of performance and energy efficiency. It is believed that further progress requires the development of neuromorphic systems, in which hardware directly mimics the neuronal network structure of a human brain. Here, we propose theoretically and realize experimentally an optical network of nodes performing binary operations. The nonlinearity required for efficient computation is provided by semiconductor microcavities in the strong quantum light-matter coupling regime, which exhibit exciton-polariton interactions. We demonstrate the system performance against a pattern recognition task, obtaining accuracy on a par with state-of-the-art hardware implementations. Our work opens the way to ultrafast and energy-efficient neuromorphic systems taking advantage of ultrastrong optical nonlinearity of polaritons.

Demonstration of Self-Starting Nonlinear Mode Locking in Random Lasers

In ultrafast multimode lasers, mode locking is implemented by means of saturable absorbers or modulators, allowing for very short pulses. This occurs because of nonlinear interactions of modes with well equispaced frequencies. Though theory predicts that, in the absence of any device, mode locking would occur in random lasers, this has never been demonstrated so far. Through the analysis of multimode correlations we provide clear evidence for nonlinear mode coupling in random lasers. The behavior of multiresonance intensity correlations is tested against the nonlinear frequency matching condition equivalent to the one underlying phase locking in ordered ultrafast lasers. Nontrivially large correlations are clearly observed for spatially overlapping resonances that sensitively depend on the frequency matching condition to be satisfied, eventually demonstrating the occurrence of nonlinear mode-locked mode coupling. This is the first example, to our knowledge, of an experimental realization of self-starting mode locking in random lasers, allowing for many new developments in the design and use of nanostructured devices.

Robust and efficient single-pixel image classification with nonlinear optics

We present a hybrid image classifier by feature-sensitive image upconversion, single pixel photodetection, and deep learning, aiming at fast processing of high-resolution images. It uses partial Fourier transform to extract the images' signature features in both the original and Fourier domains, thereby significantly increasing the classification accuracy and robustness. Tested on the Modified National Institute of Standards and Technology handwritten digit images and verified by simulation, it boosts accuracy from 81.25% (by Fourier-domain processing) to 99.23%, and achieves 83% accuracy for highly contaminated images whose signal-to-noise ratio is only -17dB. Our approach could prove useful for fast lidar data processing, high-resolution image recognition, occluded target identification, and atmosphere monitoring.

Ultralow Threshold Polariton Condensate in a Monolayer Semiconductor Microcavity at Room Temperature

Exciton-polaritons, hybrid light-matter bosonic quasiparticles, can condense into a single quantum state, i.e., forming a polariton Bose-Einstein condensate (BEC), which represents a crucial step for the development of nanophotonic technology. Recently, atomically thin transition-metal dichalcogenides (TMDs) emerged as promising candidates for novel polaritonic devices. Although the formation of robust valley-polaritons has been realized up to room temperature, the demonstration of polariton lasing remains elusive. Herein, we report for the first time the realization of this important milestone in a TMD microcavity at room temperature. Continuous wave pumped polariton lasing is evidenced by the macroscopic occupation of the ground state, which undergoes a nonlinear increase of the emission along with the emergence of temporal coherence, the presence of an exciton fraction-controlled threshold and the buildup of linear polarization. Our work presents a critically important step toward exploiting nonlinear polariton-polariton interactions, as well as offering a new platform for thresholdless lasing.

Nonlinear Parametric Scattering of Exciton Polaritons in Perovskite Microcavities

Comparing with pure photons, higher nonlinearity in polariton systems has been exploited in various proof-of-principle demonstrations of efficient optical devices based on the parametric scattering effect. However, most of them demand cryogenic temperatures limited by the small exciton binding energy of traditional semiconductors or exhibit weak nonlinearity resulting from Frenkel excitons. Lead halide perovskites, possessing both a large binding energy and a strong polariton interaction, emerge as ideal platforms to explore nonlinear polariton physics toward room temperature operation. Here, we report the first observation of nonlinear parametric scattering in a lead halide perovskite microcavity with multiple polariton branches at room temperature. Driven by the scattering source from condensation in one polariton branch, correlated polariton pairs are obtained at high k states in an adjacent branch. Our results strongly advocate the ability to reach the nonlinear regime essential for perovskite polaritonics working at room temperature.

Creating and concentrating quantum resource states in noisy environments using a quantum neural network

Quantum information processing tasks require exotic quantum states as a prerequisite. They are usually prepared with many different methods tailored to the specific resource state. Here we provide a versatile unified state preparation scheme based on a driven quantum network composed of randomly-coupled fermionic nodes. The output of such a system is then superposed with the help of linear mixing where weights and phases are trained in order to obtain desired output quantum states. We explicitly show that our method is robust and can be utilized to create almost perfect maximally entangled, NOON, W, cluster, and discorded states. Furthermore, the treatment includes energy decay in the system as well as dephasing and depolarization. Under these noisy conditions we show that the target states are achieved with high fidelity by tuning controllable parameters and providing sufficient strength to the driving of the quantum network. Finally, in very noisy systems, where noise is comparable to the driving strength, we show how to concentrate entanglement by mixing more states in a larger network.

Single-pixel pattern recognition with coherent nonlinear optics

In this Letter, we propose and experimentally demonstrate a nonlinear-optics approach to pattern recognition with single-pixel imaging and a deep neural network. It employs mode-selective image up-conversion to project a raw image onto a set of coherent spatial modes, whereby its signature features are extracted optically in a nonlinear manner. With 40 projection modes, the classification accuracy reaches a high value of 99.49% for the modified national institute of standards and technology handwritten digit images, and up to 95.32%, even when they are mixed with strong noise. Our experiment harnesses rich coherent processes in nonlinear optics for efficient machine learning, with potential applications in online classification of large-size images, fast lidar data analyses, complex pattern recognition, and so on.