Fluorescent carbon dots (CDs) have been increasingly used in fluorescence detection and imaging based on their tunable fluorescence (FL) and resistance to photobleaching. However, the fast and reliable design of fluorescent CDs with specific optical properties involves a number of factors, such as the concentration of precursors, reaction time, and solvents. Therefore, it is usually considered difficult to design CDs with favorable optical properties. Herein, we report an extreme gradient boosting (XGBoost) model for guiding the fabrication of CDs with high FL intensity and tunable emission from p-benzoquinone (PBQ) and ethylenediamine (EDA) in different solvents at room temperature. Among a variety of studied machine learning models, XGBoost shows the best performance in the field of material synthesis, with a prediction coefficient of determination (R 2) higher than 0.96. The XGBoost model can effectively predict the optical properties of CDs, including the maximum FL intensity and emission centers. Guided by the XGBoost model, various green or blue fluorescent CDs with adjustable emission centers and solubility properties are designed and fabricated accurately. These CDs are successfully applied for Fe3+ detection, sustained drug release, whole-cell imaging, and poly­(vinyl alcohol) (PVA) film preparation. These results suggest the great potential of the combination of machine learning and CD synthesis as an effective strategy to help researchers realize accurate selection of reasonable CDs with individual customized properties to achieve different goals.

This work developed a multi-layer deep convolution neural network (DCNN) model for predicting the optical properties of carbon dots (CDs), including spectral properties and fluorescence color under ultraviolet irradiation.

The vibration of a cell membrane plays a key role in the regulation of cell shape and the behavior of cells. However, most existing approaches for the measurement of cell vibration require either exogenous modification or sophisticated techniques, and the main challenge lies in developing methods that can monitor membrane vibration of living cells directly. Herein, a noninvasive strategy based on ultrasmall quartz nanopipettes is introduced. With a tip size of less than 100 nm, nanopipettes can be spatially controlled for precision targeting of a specific location on the membrane of single living cells. Surprisingly, by employing a constant voltage, stable cyclic oscillations are observed from the continuous current versus time traces. The time‐domain current can be decomposed into two basic waves: the high‐frequency one indicates the local membrane vibration driven by the electro‐osmotic flow from the nanopipette, whereas the low‐frequency one indicates the natural frequency of the whole cell. This provides a simple but reliable method to test local and global membrane vibration of single living cells simultaneously with little damage, which provides a tool for the quantification of drugs, disease, or mutations of the cell structure.

The endoplasmic reticulum (ER) is crucial for the regulation of multiple cellular processes, such as cellular responses to stress and protein synthesis, folding, and posttranslational modification. Nevertheless, monitoring ER physiological activity remains challenging due to the lack of powerful detection methods. Herein, we built a two-stage cascade recognition process to achieve dynamic visualization of ER stress in living cells based on a fluorescent carbon dot (CD) probe, which is synthesized by a facile one-pot hydrothermal method without additional modification. The fluorescent CD probe enables two-stage cascade ER recognition by first accumulating in the ER as the positively charged and lipophilic surface of the CD probe allows its fast crossing of multiple membrane barriers. Next, the CD probe can specifically anchor on the ER membrane via recognition between boronic acids and o-dihydroxy groups of mannose in the ER lumen. The two-stage cascade recognition process significantly increases the ER affinity of the CD probe, thus allowing the following evaluation of ER stress by tracking autophagy-induced mannose transfer from the ER to the cytoplasm. Thus, the boronic acid-functionalized cationic CD probe represents an attractive tool for targeted ER imaging and dynamic tracking of ER stress in living cells.

H 2 O 2 is an essential signaling molecule in living cells that can cause direct damage to lipids, proteins, and DNA, resulting in cell membrane rupture. However, current studies mostly focus on probe-based sensing of intracellular H 2 O 2 , and these methods usually require sophisticated probe synthesis and instruments. In particular, local H 2 O 2 treatment induces cell membrane rupture, but the level of cell membrane destruction is unknown because the mechanical properties of the cell membrane are difficult to accurately determine. Therefore, highly sensitive and label-free methods are required to measure and reflect mechanical changes in the cell membrane. Here, using an ultrasmall quartz nanopipette with a tip diameter less than 90 nm as a nanosensor, label-free and noninvasive electrochemical single-cell measurement is achieved for real-time monitoring of cell membrane rupture under H 2 O 2 treatment. By spatially controlling the nanopipette tip to precisely approach a specific location on the membrane of a single living cell, stable cyclic membrane oscillations are observed under a constant direct current voltage. Specifically, upon nanopipette advancement, the mechanical status of the cell membrane can be sensibly displayed by continuous current versus time traces. The electrical signals are collected and processed, ultimately revealing the mechanical properties of the cell membrane and the degree of cell apoptosis. This nanopipette-based nanosensor paves the way for developing a facile, label-free, and noninvasive strategy to assay the mechanical properties of the cell membrane during external stimulation at the single-cell level.

Targeted delivery and labeling of single living cells in heterogeneous cell populations are of great importance to understand the molecular biology and physiological functions of individual cells. However, it remains challenging to perfuse fluorescence markers into single living cells with high spatial and temporal resolution without interfering neighboring cells. Here, we report a single cell perfusion and fluorescence labeling strategy based on nanoscale glass nanopipettes. With the nanoscale tip hole of 100 nm, the use of nanopipettes allows special perfusion and high-resolution fluorescence labeling of different subcellular regions in single cells of interest. The dynamic of various fluorescent probes has been studied to exemplify the feasibility of nanopipette-dependent targeted delivery. According to experimental results, the cytoplasm labeling of Sulfo-Cyanine5 and fluorescein isothiocyanate is mainly based on the Brownian movement due to the dyes themselves and does not have a targeting ability, while the nucleus labeling of 4′,6-diamidino-2-phenylindole (DAPI) is originated from the adsorption between DAPI and DNA in the nucleus. From the finite element simulation, the precise manipulation of intracellular delivery is realized by controlling the electro-osmotic flow inside the nanopipettes, and the different delivery modes between nontargeting dyes and nucleus-targeting dyes were compared, showcasing the valuable ability of nanopipette-based method for the analysis of specially defined subcellular regions and the potential applications for single cell surgery, subcellular manipulation, and gene delivery.

In this work, we design sticky, furry and fluorescent “wool-balls” based on p-benzoquinone/ethylenediamine polymer nanoclusters, which provide a convenient, fast labeling strategy for the imaging of cell surface sialic acids.

This work reports a single cell glycan labeling strategy based on nanopipettes, which provides a convenient, fast labeling method for bridging the gap between single cell manipulation and metabolic labeling.