To develop new sensing films, there are a number of issues that need to be considered, which include, for example, 1) sensing unit (fluorophore) selection as it determines the effectiveness of sensing to the concerned analytes, 2) substrate selection since it could largely affect the sensing performances owing to screening or enriching the analytes from the medium, the so called substrate effect, [5][6][7] and 3) adlayer structure, another crucial but less concerned factor because it highly influences the sensitivity, selectivity, speed, and reversibility (3S + 1R) of a sensing process due to mass transferrelated reasons. [8] In fact, numerous fluorescent films have been developed by different fabrication methods, such as dropcasting [9] and spin-coating [10,11] , which are highly efficient but suffer from coffee-ring effect, [12][13][14] uncontrollable thickness and adlayer structures, etc. Langmuir-Blodgett and mono-layer self-assembled chemical techniques are another two kinds of popular ways to fabricate the sensing films. [15,16] The films fabricated in these two techniques could show much improved sensing performances mainly due to the unprecedented availability of the sensing units in the sensing. The films, however, are photochemically unstable, and the preparation is hard to scale up. Therefore, we proposed a molecular gel strategy to fabricate fluorescent films a few years ago, where the low-molecular mass gelators used are derivatives of sensing fluorophores. [17] The sensing films fabricated in the strategy show multiple advantages including efficient fabrication, improved photochemical stability, great porosity, and minimized substrate effect. However, the adlayer structures of the films are far from uniform, and the pore sizes could vary from sub-nanometers to tens or even hundreds of micrometers, which greatly weakens the size-related selectivity of the films in sensing. Even worse, fluorescent films with thickness in the nanometer range and with fine structures are hardly prepared in the way, and thereby, new strategies for fabrication of highperformance fluorescence sensing films need to be developed.Recently, we proposed a "mix design strategy" based on non-planar fluorophores, [18] where the capillary condensation and adsorption/desorption kinetics originated from the porous adlayer structures as well as micro-environment effectThe performance of a film-based fluorescent sensor is highly dependent on the innovative design of the sensing films aiming to elevate usability of sensing units, and enhance mass transfer of analytes within the fluorescent adlayer. Herein, a highly porous and uniform fluorescent nanofilm which is fabricated through an interfacially confined dynamic-covalent reaction between benzene-1,3,5-tricarbo-hydrazide and a mono-boron complex of 8-hydroxy-quinoline is reported. Remarkably, the thicknesses of the nanofilms prepared in this way can be modulated from tens to hundreds of nanometers. In addition, the nanofilms demonstrate much improved photochemical stability. Sensing p...