Inverse molecular design of alkoxides and phenoxides for aqueous direct air capture of CO2.

Aqueous direct air capture (DAC) is a key technology toward a carbon negative infrastructure. Developing sorbent molecules with water and oxygen tolerance and high CO2 binding capacity is therefore highly desired. We analyze the CO2 absorption chemistries on amines, alkoxides, and phenoxides with density functional theory calculations, and perform inverse molecular design of the optimal sorbent. The alkoxides and phenoxides are found to be more suitable for aqueous DAC than amines thanks to their water tolerance (lower pKa prevents protonation by water) and capture stoichiometry of 1:1 (2:1 for amines). All three molecular systems are found to generally obey the same linear scaling relationship (LSR) between [Formula: see text] and [Formula: see text], since both CO2 and proton are bonded to the nucleophilic (alkoxy or amine) binding site through a majorly [Formula: see text] bonding orbital. Several high-performance alkoxides are proposed from the computational screening. Phenoxides have comparatively poorer correlation between [Formula: see text] and [Formula: see text], showing promise for optimization. We apply a genetic algorithm to search the chemical space of substituted phenoxides for the optimal sorbent. Several promising off-LSR candidates are discovered. The most promising one features bulky ortho substituents forcing the CO2 adduct into a perpendicular configuration with respect to the aromatic ring. In this configuration, the phenoxide binds CO2 and a proton using different molecular orbitals, thereby decoupling the [Formula: see text] and [Formula: see text]. The [Formula: see text] trend and off-LSR behaviors are then confirmed by experiments, validating the inverse molecular design framework. This work not only extensively studies the chemistry of the aqueous DAC, but also presents a transferrable computational workflow for understanding and optimization of other functional molecules.

Machine learning-assisted search for novel coagulants: When machine learning can be efficient even if data availability is low.

Design of new drugs is a challenging process: a candidate molecule should satisfy multiple conditions to act properly and make the least side-effect-perfect candidates selectively attach to and influence only targets, leaving off-targets intact. The amount of experimental data about various properties of molecules constantly grows, promoting data-driven approaches. However, the applicability of typical predictive machine learning techniques can be substantially limited by a lack of experimental data about a particular target. For example, there are many known Thrombin inhibitors (acting as anticoagulants), but a very limited number of known Protein C inhibitors (coagulants). In this study, we present our approach to suggest new inhibitor candidates by building an effective representation of chemical space. For this aim, we developed a deep learning model-autoencoder, trained on a large set of molecules in the SMILES format to map the chemical space. Further, we applied different sampling strategies to generate novel coagulant candidates. Symmetrically, we tested our approach on anticoagulant candidates, where we were able to predict their inhibition towards Thrombin. We also compare our approach with MegaMolBART-another deep learning generative model, but exploiting similar principles of navigation in a chemical space.

Tailored Permeation Through ≈1 nm Thick Carbon Nanomembranes by Subtle Changes in Their Molecular Design.

Due to their nanoscale thickness (≈1 nm) and exceptional selectivity for permeation of gases, nanomembranes made of 2D materials possess high potential for energy-efficient nanofiltration applications. In this respect, organic carbon nanomembranes (CNMs), synthesized via electron irradiation-induced crosslinking of aromatic self-assembled monolayers (SAMs), are particularly attractive, as their structure can be flexibly tuned by choice of molecular precursors. However, tailored permeation of CNMs, defined by their molecular design, has not been yet demonstrated. In this work, it is shown that the permeation of helium (He), deuterium (D2) and heavy water (D2O) for CNMs synthesized from biphenyl-based SAMs on silver (C6H5-C6H4-(CH2)n-COO/Ag, n = 2-6) can be tuned by orders of magnitude by changing the structure of the molecular precursors by just a single methylene unit. The selectivity in permeation of D2O/D2 with an unprecedented value of 200 000 can be achieved in this way. The temperature-dependent study reveals a clear correlation between the molecular design and the permeation mechanisms facilitating therewith tailored synthesis of molecular 2D materials for separation technologies.

Lessons learned in application driven imaging agent design for image-guided surgery.

To meet the growing demand for intraoperative molecular imaging, the development of compatible imaging agents plays a crucial role. Given the unique requirements of surgical applications compared to diagnostics and therapy, maximizing translational potential necessitates distinctive imaging agent designs. For effective surgical guidance, exogenous signatures are essential and are achievable through a diverse range of imaging labels such as (radio)isotopes, fluorescent dyes, or combinations thereof. To achieve optimal in vivo utility a balanced molecular design of the tracer as a whole is required, which ensures a harmonious effect of the imaging label with the affinity and specificity (e.g., pharmacokinetics) of a pharmacophore/targeting moiety. This review outlines common design strategies and the effects of refinements in the molecular imaging agent design on the agent's pharmacological profile. This includes the optimization of affinity, pharmacokinetics (including serum binding and target mediated background), biological clearance route, the achievable signal intensity, and the effect of dosing hereon.

Recent Progress of the Cathode Material Design for Aqueous Zn-Organic Batteries.

Aqueous zinc-ion batteries (ZIBs) have emerged as the most promising candidate for large-scale energy storage due to their inherent safety, environmental friendliness, and cost-effectiveness. Simultaneously, the utilization of organic electrode materials with renewable resources, environmental compatibility, and diverse structures has sparked a surge in research and development of aqueous Zn-organic batteries (ZOBs). A comprehensive review is warranted to systematically present recent advancements in design principles, synthesis techniques, energy storage mechanisms, and zinc-ion storage performance of organic cathodes. In this review article, we comprehensively summarize the energy storage mechanisms employed by aqueous ZOBs. Subsequently, we categorize organic cathode materials into small-molecule compounds and high-molecular polymers respectively. Novel polymer materials such as conjugated polymers (CPs), conjugated microporous polymers (CMPs), and covalent organic frameworks (COFs) are highlighted with an overview of molecular design strategies and structural optimization based on organic cathode materials aimed at enhancing the performance of aqueous ZOBs. Finally, we discuss the challenges faced by aqueous ZOBs along with future prospects to offer insights into their practical applications.

3,4-Dimethylenecyclobutene: A Building Block for Design of Macrocycles with Excited State Aromatic Low-Lying High-Spin States.

3,4-Dimethylenecyclobutene (DMCB) is an unusual isomer of benzene. Motivated by recent synthetic progress to substituted derivatives of this scaffold, we carried out a theoretical and computational analysis with a particular focus on the extent of (anti)aromatic character in the lowest excited states of different multiplicities. We found that the parent DMCB is non-aromatic in its singlet ground state (S0), lowest triplet state (T1), and lowest singlet excited state (S1), while it is aromatic in its lowest quintet state (Q1) as this state is represented by a triplet multiplicity cyclobutadiene (CBD) ring and two uncoupled same-spin methylene radicals. Interestingly, the Q1 state, despite having four unpaired electrons, is placed merely 4.8 eV above S0, and there is a corresponding singlet tetraradical 0.16 eV above. The DMCB is potentially a highly useful structural motif for the design of larger molecular entities with interesting optoelectronic properties. Here, we designed macrocycles composed of fused DMCB units, and according to our computations, two of these have low-lying nonet states (i. e., octaradical states) at energies merely 2.40 and 0.37 eV above their S0 states as a result of local Hückel- and Baird-aromatic character of individual 6π- and 4π-electron monocycles.

Rational Design of Coumarin-Based Hybridized Local and Charge-Transfer Blue Emitters for Solution-Processed Organic Light-Emitting Diodes.

Hybridized local and charge-transfer (HLCT) with the utilization of both singlet and triplet excitons through the "hot excitons" channel have great application potential in highly efficient blue organic light-emitting diodes (OLEDs). The proportion of charge-transfer (CT) and locally excited (LE) components in the relevant singlet and triplet states makes a big difference for the high-lying reverse intersystem crossing process. Herein, three novel donor (D)-acceptor (A) type HLCT materials, 7-([1,1'-biphenyl]-4-yl(9,9-dimethyl-9H-fluoren-2-yl)amino)-3-phenyl-1H-isochromen-1-one (pPh-7P), 7-([1,1'-biphenyl]-4-yl(9,9-dimethyl-9H-fluoren-2-yl)amino)-3-methyl-1H-isochromen-1-one (pPh-7M), and 6-([1,1'-biphenyl]-4-yl(9,9-dimethyl-9H-fluoren-2-yl)amino)-3-methyl-1H-isochromen-1-one (pPh-6M), were rationally designed and synthesized with diphenylamine derivative as donor and oxygen heterocyclic coumarin moiety as acceptors. The proportions of CT and LE components were fine controlled by changing the connection site of diphenylamine derivative at C6/C7-position and the substituent at C3-position of coumarin moiety. The HLCT characteristics of pPh-7P, pPh-7M, and pPh-6M were systematically demonstrated through photophysical properties and density functional theory calculations. The solution-processed doped OLEDs based on pPh-6M exhibited deep-blue electroluminescence with the maximum emission wavelength of 446 nm, maximum luminance of 8755 cd m-2, maximum current efficiency of 5.83 cd A-1, and maximum external quantum efficiency of 6.54 %. The results reveal that pPh-6M with dominant 1LE and 3CT components has better OLED performance.

Molecule auto-correction to facilitate molecular design.

Ensuring that computationally designed molecules are chemically reasonable is at best cumbersome. We present a molecule correction algorithm that morphs invalid molecular graphs into structurally related valid analogs. The algorithm is implemented as a tree search, guided by a set of policies to minimize its cost. We showcase how the algorithm can be applied to molecular design, either as a post-processing step or as an integral part of molecule generators.

Computational design and experimental confirmation of a disulfide-stapled YAP helixα1-trap derived from TEAD4 helical hairpin to selectively capture YAP α1-helix with potent antitumor activity.

Human Hippo signaling pathway is an evolutionarily conserved regulator network that controls organ development and has been implicated in various cancers. Transcriptional enhanced associate domain-4 (TEAD4) is the final nuclear effector of Hippo pathway, which is activated by Yes-associated protein (YAP) through binding to two separated YAP regions of α1-helix and Ω-loop. Previous efforts have all been addressed on deriving peptide inhibitors from the YAP to target TEAD4. Instead, we herein attempted to rationally design a so-called 'YAP helixα1-trap' based on the TEAD4 to target YAP by using dynamics simulation and energetics analysis as well as experimental assays at molecular and cellular levels. The trap represents a native double-stranded helical hairpin covering a specific YAP-binding site on TEAD4 surface, which is expected to form a three-helix bundle with the α1-helical region of YAP, thus competitively disrupting TEAD4-YAP interaction. The hairpin was further stapled by a disulfide bridge across its two helical arms. Circular dichroism characterized that the stapling can effectively constrain the trap into a native-like structured conformation in free state, thus largely minimizing the entropy penalty upon its binding to YAP. Affinity assays revealed that the stapling can considerably improve the trap binding potency to YAP α1-helix by up to 8.5-fold at molecular level, which also exhibited a good tumor-suppressing effect at cellular level if fused with TAT cell permeation sequence. In this respect, it is considered that the YAP helixα1-trap-mediated blockade of Hippo pathway may be a new and promising therapeutic strategy against cancers.

MolCFL: A personalized and privacy-preserving drug discovery framework based on generative clustered federated learning.

In today's era of rapid development of large models, the traditional drug development process is undergoing a profound transformation. The vast demand for data and consumption of computational resources are making independent drug discovery increasingly difficult. By integrating federated learning technology into the drug discovery field, we have found a solution that both protects privacy and shares computational power. However, the differences in data held by various pharmaceutical institutions and the diversity in drug design objectives have exacerbated the issue of data heterogeneity, making traditional federated learning consensus models unable to meet the personalized needs of all parties. In this study, we introduce and evaluate an innovative drug discovery framework, MolCFL, which utilizes a multi-layer perceptron (MLP) as the generator and a graph convolutional network (GCN) as the discriminator in a generative adversarial network (GAN). By learning the graph structure of molecules, it generates new molecules in a highly personalized manner and then optimizes the learning process by clustering federated learning, grouping compound data with high similarity. MolCFL not only enhances the model's ability to protect privacy but also significantly improves the efficiency and personalization of molecular design. MolCFL exhibits superior performance when handling non-independently and identically distributed data compared to traditional models. Experimental results show that the framework demonstrates outstanding performance on two benchmark datasets, with the generated new molecules achieving over 90% in Uniqueness and close to 100% in Novelty. MolCFL not only improves the quality and efficiency of drug molecule design but also, through its highly customized clustered federated learning environment, promotes collaboration and specialization in the drug discovery process while ensuring data privacy. These features make MolCFL a powerful tool suitable for addressing the various challenges faced in the modern drug research and development field.

Deep learning algorithms applied to computational chemistry.

Recently, there has been a significant increase in the use of deep learning techniques in the molecular sciences, which have shown high performance on datasets and the ability to generalize across data. However, no model has achieved perfect performance in solving all problems, and the pros and cons of each approach remain unclear to those new to the field. Therefore, this paper aims to review deep learning algorithms that have been applied to solve molecular challenges in computational chemistry. We proposed a comprehensive categorization that encompasses two primary approaches; conventional deep learning and geometric deep learning models. This classification takes into account the distinct techniques employed by the algorithms within each approach. We present an up-to-date analysis of these algorithms, emphasizing their key features and open issues. This includes details of input descriptors, datasets used, open-source code availability, task solutions, and actual research applications, focusing on general applications rather than specific ones such as drug discovery. Furthermore, our report discusses trends and future directions in molecular algorithm design, including the input descriptors used for each deep learning model, GPU usage, training and forward processing time, model parameters, the most commonly used datasets, libraries, and optimization schemes. This information aids in identifying the most suitable algorithms for a given task. It also serves as a reference for the datasets and input data frequently used for each algorithm technique. In addition, it provides insights into the benefits and open issues of each technique, and supports the development of novel computational chemistry systems.

Design, synthesis and systemic acquired resistance of 2-benzothiadiazolylquinoline-4-carboxamides by COI1 based virtual screening.

Coronatine-insensitive 1 (COI1) has been identified as a target receptor of plant elicitor coronatine (COR). To discover novel plant elicitor leads, most of the potential molecules among 129 compounds discovered from the ZINC database by docking based virtual screening targeting COI1 were quinoline amides. On this lead basis, 2-benzothiadiazolylquinoline-4-carboxamides were rationally designed and synthesized for bioassay. All target compounds did not show significantly in vitro antifungal activity, compounds 4d, 4e and 4o displayed good in vivo systemic acquired resistance activity for Arabidopsis thaliana against Hyaloperonospora arabidopsidis isolate Noco2 with over 80% of inhibitory rate at the concentration of 50 µM. These results indicate that 2-benzothiadiazolylquinoline-4-carboxamides are promising plant elicitor leads for further study.

A comprehensive prediction system for silkworm acute toxicity assessment of environmental and in-silico pesticides.

The excessive application and loss of pesticides poses a great risk to the ecosystem, and the environmental safety assessment of pesticides is time-consuming and expensive using traditional animal toxicity tests. In this work, a pesticide acute toxicity dataset was created for silkworm integrating extensive experiments and various common pesticide formulations considering the sensitivity of silkworm to adverse environment, its economic value in China, and a gap in machine learning (ML) research on the toxicity prediction of this species, which addressed the previous limitation of only being able to predict toxicity classification without specific toxicity values. A new comprehensive voting model (CVR) was developed based on ML, combined with three regression algorithms, namely, Bayesian Ridge (BR), K Neighbors Regressor (KNN), Random Forest Regressor (RF) to accurately calculate lethal concentration 50 % (LC50). Three conformal models were successfully constructed, marking the first combination of conformal models with confidence intervals to predict silkworm toxicity. Further, the mechanism by analyzing structural alerts was summarized, and identified 25 warning structures, 24 positive compounds and 14 negative compounds. Importantly, a novel comprehensive prediction system was constructed that can provide LC50 and confidence intervals, structural alerts analysis, lipid-water partition coefficient (LogP) and similarity analysis, which can comprehensively evaluate the ecological toxicity risk of substances to make up for the incomplete toxicity data of new pesticides. The validity and generalization of the CVR model were verified by an external validation set. In addition, five new, low-toxic and green pesticide alternatives were designed through 50,000 cycles. Moreover, our software and ST Profiler can provide low-cost information access to accelerate environmental risk assessment, which can predict not only a single chemical, but also batches of chemicals, simply by inputting the SMILES / CAS / (Chinese / English) name of chemicals.

Synthesis and Antifungal Properties of 1,2,4-Triazole Schiff Base Agents Based on a 3D-QSAR Model.

Based on our previous research, a 3D-QSAR model (q2=0.51, ONC=5, r2=0.982, F=271.887, SEE=0.052) was established to predict the inhibitory effects of triazole Schiff base compounds on Fusarium graminearum, and its predictive ability was also confirmed through the statistical parameters. According to the results of the model design, 30 compounds with superior bioactivity compared to the template molecule 4 were obtained. Seven of these compounds (DES2-6, DES9-10) with improved biological activity and readily available raw materials were successfully synthesized. Their structures were confirmed through HRMS, NMR, and single crystal X-ray diffraction analysis (DES-5). The bioactivity of the final products was investigated through an in vitro antifungal assay. There was little difference in the EC50 values between the experimental and predicted values of the model, demonstrating the reliability of the model. Especially, DES-3 (EC50=9.915 mg/L) and DES-5 (EC50=9.384 mg/L) exhibited better inhibitory effects on Fusarium graminearum compared to the standard drug (SD) triadimenol (EC50=10.820 mg/L). These compounds could serve as potential new fungicides for future research. The interaction between the final products and isocitrate lyase (ICL) was investigated through molecular docking. Compounds with R groups that have a higher electron-donating capacity were found to be biologically active.

Design, Synthesis and Biological Evaluation of Novel Phenyl-Substituted Naphthoic Acid Ethyl Ester Derivatives as Strigolactone Receptor Inhibitor.

Strigolactones (SLs) are plant hormones that regulate several key agronomic traits, including shoot branching, leaf senescence, and stress tolerance. The artificial regulation of SL biosynthesis and signaling has been considered as a potent strategy in regulating plant architecture and combatting the infection of parasitic weeds to help improve crop yield. DL1b is a previously reported SL receptor inhibitor molecule that significantly promotes shoot branching. Here, we synthesized 18 novel compounds based on the structure of DL1b. We performed rice tillering activity assay and selected a novel small molecule, C6, as a candidate SL receptor inhibitor. In vitro bioassays demonstrated that C6 possesses various regulatory functions as an SL inhibitor, including inhibiting germination of the root parasitic seeds Phelipanche aegyptiaca, delaying leaf senescence and promoting hypocotyl elongation of Arabidopsis. ITC analysis and molecular docking experiments further confirmed that C6 can interact with SL receptor proteins, thereby interfering with the binding of SL to its receptor. Therefore, C6 is considered a novel SL receptor inhibitor with potential applications in plant architecture control and prevention of root parasitic weed infestation.

Applying Molecular Modeling to the Design of Innovative, Non-Symmetrical CXCR4 Inhibitors with Potent Anticancer Activity.

The identification of new compounds with potential activity against CXC chemokine receptor type 4 (CXCR4) has been broadly studied, implying several chemical families, particularly AMD3100 derivatives. Molecular modeling has played a pivotal role in the identification of new active compounds. But, has its golden age ended? A virtual library of 450,000 tetraamines of general structure 8 was constructed by using five spacers and 300 diamines, which were obtained from the corresponding commercially available cyclic amines. Diversity selection was performed to guide the virtual screening of the former database and to select the most representative set of compounds. Molecular docking on the CXCR4 crystal structure allowed us to rank the selection and identify those candidate molecules with potential antitumor activity against diffuse large B-cell lymphoma (DLBCL). Among them, compound A{17,18} stood out for being a non-symmetrical structure, synthetically feasible, and with promising activity against DLBCL in in vitro experiments. The focused study of symmetrical-related compounds allowed us to identify potential pre-hits (IC50~20 µM), evidencing that molecular design is still relevant in the development of new CXCR4 inhibitor candidates.

Revolutionizing Molecular Design for Innovative Therapeutic Applications through Artificial Intelligence.

The field of computational protein engineering has been transformed by recent advancements in machine learning, artificial intelligence, and molecular modeling, enabling the design of proteins with unprecedented precision and functionality. Computational methods now play a crucial role in enhancing the stability, activity, and specificity of proteins for diverse applications in biotechnology and medicine. Techniques such as deep learning, reinforcement learning, and transfer learning have dramatically improved protein structure prediction, optimization of binding affinities, and enzyme design. These innovations have streamlined the process of protein engineering by allowing the rapid generation of targeted libraries, reducing experimental sampling, and enabling the rational design of proteins with tailored properties. Furthermore, the integration of computational approaches with high-throughput experimental techniques has facilitated the development of multifunctional proteins and novel therapeutics. However, challenges remain in bridging the gap between computational predictions and experimental validation and in addressing ethical concerns related to AI-driven protein design. This review provides a comprehensive overview of the current state and future directions of computational methods in protein engineering, emphasizing their transformative potential in creating next-generation biologics and advancing synthetic biology.

Designing salt stress-resilient crops: Current progress and future challenges.

Excess soil salinity affects large regions of land and is a major hindrance to crop production worldwide. Therefore, understanding the molecular mechanisms of plant salt tolerance has scientific importance and practical significance. In recent decades, studies have characterized hundreds of genes associated with plant responses to salt stress in different plant species. These studies have substantially advanced our molecular and genetic understanding of salt tolerance in plants and have introduced an era of molecular design breeding of salt-tolerant crops. This review summarizes our current knowledge of plant salt tolerance, emphasizing advances in elucidating the molecular mechanisms of osmotic stress tolerance, salt-ion transport and compartmentalization, oxidative stress tolerance, alkaline stress tolerance, and the trade-off between growth and salt tolerance. We also examine recent advances in understanding natural variation in the salt tolerance of crops and discuss possible strategies and challenges for designing salt stress-resilient crops. We focus on the model plant Arabidopsis (Arabidopsis thaliana) and the four most-studied crops: rice (Oryza sativa), wheat (Triticum aestivum), maize (Zea mays), and soybean (Glycine max).

High-Voltage Catholyte for High-Energy-Density Nonaqueous Redox Flow Battery.

Redox flow batteries (RFBs) with high energy densities are essential for efficient and sustainable long-term energy storage on a grid scale. To advance the development of nonaqueous RFBs with high energy densities, a new organic RFB system employing a molecularly engineered tetrathiafulvalene derivative ((PEG3/PerF)-TTF) as a high energy density catholyte was developed. A synergistic approach to the molecular design of tetrathiafulvalene (TTF) was applied, with the incorporation of polyethylene glycol (PEG) chains, which enhance its solubility in organic carbonate electrolytes, and a perfluoro (PerF) group to increase its redox potential. When paired with a lithium metal anode, the two-electron-active (PEG3/PerF)-TTF catholyte produced a cell voltage of 3.56 V for the first redox process and 3.92 V for the second redox process. In cyclic voltammetry and flow cell tests, the redox chemistry exhibited excellent cycling stability. The Li|(PEG3/PerF)-TTF batteries, with concentrations of 0.1 M and 0.5 M, demonstrated capacity retention rates of ~94 % (99.87 % per cycle, 97.52 % per day) and 90 % (99.93 % per cycle, 99.16 % per day), and the average Coulombic efficiencies of 99.38 % and 98.35 %, respectively. The flow cell achieved a high power density of 129 mW/cm2. Furthermore, owing to the high redox potential and solubility of (PEG3/PerF)-TTF, the flow cell attained a high operational energy density of 72 Wh/L (100 Wh/L theoretical). A 0.75 M flow cell exhibited an even higher operational energy density of 96 Wh/L (150 Wh/L theoretical).

Ladder-Type Redox-Active Polymer Achieves Ultra-Stable and Fast Proton Storage in Aqueous Proton Batteries.

A ladder-type rigid-coplanar polymer with highly ordered molecular arrangement has been designed via a covalent cycloconjugation conformational strategy. Benefitting from the extended π-electron delocalization in the highly aromatic ladder-type polymeric backbone, the prepared polymer exhibits fast intra-chain charge transport along the polymeric chain, realizing extraordinary proton-storage capability in aqueous proton batteries.Affordable and safe aqueous proton batteries (APBs) with unique "Grotthuss mechanism," are very significant for advancing carbon neutrality initiatives. While organic polymers offer a robust and adaptable framework that is well-suited for APB electrodes, the limited proton-storage redox capacity has constrained their broader application. Herein, a ladder-type polymer (PNMZ) has been designed via a covalent cycloconjugation conformational strategy that exhibits optimized electronic structure and fast intra-chain charge transport within the high-aromaticity polymeric skeleton. As a result, the polymer exhibits exceptional proton-storage redox kinetics, which are evidenced by in-operando monitoring techniques and theoretical calculations. It achieves a remarkable proton-storage capacity of 189 mAh g-1 at 2 A g-1 and excellent long-term cycling stability, with approximately 97.8 % capacity retention over 10,000 cycles. Finally, a high-performance all-polymer APB device has been successfully constructed with a desirable capacity retention of 99.7 % after 6,000 cycles and high energy density of 56.3 Wh kg-1.