Optical computing holds promise for high-speed,energy-efficient information processing,with diffractive optical networks emerging as a flexible platform for implementing task-specific transformations.A challenge,howev...Optical computing holds promise for high-speed,energy-efficient information processing,with diffractive optical networks emerging as a flexible platform for implementing task-specific transformations.A challenge,however,is the effective optimization and alignment of the diffractive layers,which is hindered by the difficulty of accurately modeling physical systems with their inherent hardware imperfections,noise,and misalignments.While existing in situ optimization methods offer the advantage of direct training on the physical system without explicit system modeling,they are often limited by slow convergence and unstable performance due to inefficient use of limited measurement data.Here,we introduce a model-free reinforcement learning approach utilizing Proximal Policy Optimization(PPO)for the in situ training of diffractive optical processors.PPO efficiently reuses in situ measurement data and constrains policy updates to ensure more stable and faster convergence.We validated our method through both simulations and experiments across a range of in situ learning tasks,including targeted energy focusing through a random diffuser,image generation,aberration correction,and optical image classification,demonstrating in each task better convergence and performance.Our strategy operates directly on the physical system and naturally accounts for unknown real-world imperfections,eliminating the need for prior system knowledge or modeling.By enabling faster and more accurate training under realistic experimental constraints,this in situ reinforcement learning approach could offer a scalable framework for various optical and physical systems governed by complex,feedback-driven dynamics.展开更多
In this work,we present a top-down method for the preparation of 2D MOF nanosheets with cavity structures.The pro-ligand 25,26,27,28-tetrakis[(carboxyl)methoxy]calix[4]arene was elaborately selected,and a layered MOF ...In this work,we present a top-down method for the preparation of 2D MOF nanosheets with cavity structures.The pro-ligand 25,26,27,28-tetrakis[(carboxyl)methoxy]calix[4]arene was elaborately selected,and a layered MOF with cavity structures was constructed.The large molecular skeleton and cup-shaped feature of the calix[4]arene caused large layer separations and weak interlayer interactions among the 2D layers,which enabled the layered MOF to be readily delaminated into ultrathin 2D MOF nanosheets.Owing to the cup-shaped feature of the calix[4]arene,there are permanent cage-like cavities loaded on the as-prepared MOF nanosheets.By decorating oxygen-containing functional groups(carboxyl and ether groups)in the cage-like cavities,the resultant Cu-MOF nanosheets showed excellent adsorption performance for Pb^(2+).The intimate contact and sufficient interactions on the exposed surface areas of Cu-MOF nanosheet resulted in ultrahigh adsorption selectivity and anti-interference ability for Pb^(2+),together with an outstanding Pb^(2+)uptake capacity of 738.65 mg g^(-1),which were obviously better than those of its 3D precursor.The possible adsorption mechanism was systematically investigated by the investigations of zeta potential,FT-IR,XPS,and DFT calculations.This study opens the door to achieving ultrathin MOF nanosheets with cavity structures,which would well expand the applications of MOF nanosheets.展开更多
文摘Optical computing holds promise for high-speed,energy-efficient information processing,with diffractive optical networks emerging as a flexible platform for implementing task-specific transformations.A challenge,however,is the effective optimization and alignment of the diffractive layers,which is hindered by the difficulty of accurately modeling physical systems with their inherent hardware imperfections,noise,and misalignments.While existing in situ optimization methods offer the advantage of direct training on the physical system without explicit system modeling,they are often limited by slow convergence and unstable performance due to inefficient use of limited measurement data.Here,we introduce a model-free reinforcement learning approach utilizing Proximal Policy Optimization(PPO)for the in situ training of diffractive optical processors.PPO efficiently reuses in situ measurement data and constrains policy updates to ensure more stable and faster convergence.We validated our method through both simulations and experiments across a range of in situ learning tasks,including targeted energy focusing through a random diffuser,image generation,aberration correction,and optical image classification,demonstrating in each task better convergence and performance.Our strategy operates directly on the physical system and naturally accounts for unknown real-world imperfections,eliminating the need for prior system knowledge or modeling.By enabling faster and more accurate training under realistic experimental constraints,this in situ reinforcement learning approach could offer a scalable framework for various optical and physical systems governed by complex,feedback-driven dynamics.
基金supported by the Natural Science Foundation of Shandong Province(ZR2021MB106 and ZR2021QB090)the Major Projects of Natural Science Research in Universities of Jiangsu Province(20KJA150002)the Huaishang Talent Program of Huaian.
文摘In this work,we present a top-down method for the preparation of 2D MOF nanosheets with cavity structures.The pro-ligand 25,26,27,28-tetrakis[(carboxyl)methoxy]calix[4]arene was elaborately selected,and a layered MOF with cavity structures was constructed.The large molecular skeleton and cup-shaped feature of the calix[4]arene caused large layer separations and weak interlayer interactions among the 2D layers,which enabled the layered MOF to be readily delaminated into ultrathin 2D MOF nanosheets.Owing to the cup-shaped feature of the calix[4]arene,there are permanent cage-like cavities loaded on the as-prepared MOF nanosheets.By decorating oxygen-containing functional groups(carboxyl and ether groups)in the cage-like cavities,the resultant Cu-MOF nanosheets showed excellent adsorption performance for Pb^(2+).The intimate contact and sufficient interactions on the exposed surface areas of Cu-MOF nanosheet resulted in ultrahigh adsorption selectivity and anti-interference ability for Pb^(2+),together with an outstanding Pb^(2+)uptake capacity of 738.65 mg g^(-1),which were obviously better than those of its 3D precursor.The possible adsorption mechanism was systematically investigated by the investigations of zeta potential,FT-IR,XPS,and DFT calculations.This study opens the door to achieving ultrathin MOF nanosheets with cavity structures,which would well expand the applications of MOF nanosheets.
