Cross-domain diagnosis for polymer electrolyte membrane fuel cell based on digital twins and transfer learning network

Existing research on fault diagnosis for polymer electrolyte membrane fuel cells(PEMFC)has advanced significantly,yet performance is hindered by variations in data distributions and the requirement for extensive fault data.In this study,a cross-domain adaptive health diagnosis method for PEMFC is proposed,integrating the digital twin model and transfer convolutional diagnosis model.A physical-based high-fidelity digital twin model is developed to obtain diverse and high-quality datasets for training diagnosis method.To extract long-term time series features from the data,a temporal convolutional network(TCN)is proposed as a pre-trained diagnosis model for the source domain,with feature extraction layers that can be reused to the transfer learning network.It is demonstrated that the proposed pre-trained model can hold the ability to accurately diagnose the various fuel cell faults,including pressure,drying,flow,and flooding faults,with 99.92%accuracy,through the effective capture of the long-term dependencies in time series data.Finally,a domain adaptive transfer convolutional network(DATCN)is established to improve the diagnosis accuracy across diverse fuel cells by learning domain-invariant features.The results show that the DATCN model,tested on three different target domain devices with adversarial training using only 10%normal data,can achieve an average accuracy of 98.5%(30%improved over traditional diagnosis models).This proposed method provides an effective solution for accurate cross-domain diagnosis of PEMFC devices,significantly reducing the reliance on extensive fault data.

Molecular dynamics for electrocatalysis:Mechanism explanation and performance prediction

Designing low-cost,high-performing electrocatalysts is key to green energy development,yet relying solely on the"synthesis-characterization"catalyst screening model is time-consuming and costly.There are two main applications for Molecular dynamics(MD)simulations in electrochemical reactions:explaining mechanisms and predicting performance,which play important roles in fabricating robust electrocatalysts.MD simulations of electrocatalysis include the adsorption and desorption of reactants,intermediates,and products in this review.The structural changes in active centers under various electric field states,the effects of alkali metal cations,common anions,and pH effects in the electrolyte on the electrocatalytic process are also discussed to reveal the reaction mechanism.Then the prediction of the catalysts performance in specific reaction using MD simulations are introduced.Finally,the optimization and challenges of MD techniques are discussed.

Bubble-like Fe-encapsulated N,S-codoped carbon nanofibers as efficient bifunctional oxygen electrocatalysts for robust Zn-air batteries

Nano Research volume 13,pages2175–2182(2020)Cite this article 208 Accesses 1 Altmetric Metrics details Abstract Efficient,robust and cost-effective bifunctional oxygen electrocatalysts for oxygen reduction reaction(ORR)and oxygen evolution reaction(OER)are of vital importance to the widespread utilization of Zn-air batteries.Here we report the fabrication of a bubble-like N,S-codoped porous carbon nanofibers with encapsulated fine Fe/Fe5C2 nanocrystals(∼10 nm)(FeNSCs)by a facile one-pot pyrolysis strategy.The novel FeNSC nanostructures with high Fe content(37.3 wt.%),and synergetic N and S doping demonstrate remarkable ORR and OER catalytic activities in alkaline condition.Particularly for ORR,the optimal FeNSC catalyst exhibits superior performance in terms of current density and durability in both alkaline and acidic media.Moreover,as catalysts on the air electrodes of Zn-air batteries,the optimal FeNSCs show a high peak power density of 59.6 mW/cm^2 and extraordinary discharge-charge cycling performance for 200 h with negligible voltage gap change of only 8%at current density of 20 mA/cm,surpassing its noble metal counterpart(i.e.Pt).The impressive battery stability can be attributed to favorable electron transfer resulting from appropriate graphitization of the bubble-like carbon nanofibers and thorough protection of Fe/Fe5C2 nanoparticles by carbon wrapping to prevent oxidation,agglomeration and dissolution of Fe nanoparticles during battery cycling.The present FeNSC catalyst,which is highly active,robust yet affordable,shows promising prospects in large-scale applications,such as metal-air batteries and fuel cells.

Ultrastable bimetallic Fe_(2)Mo for efficient oxygen reduction reaction in pH-universal applications

Iron-based nanostructures represent an emerging class of catalysts with high electroactivity for oxygen reduction reaction(ORR)in energy storage and conversion technologies.However,current practical applications have been limited by insufficient durability in both alkaline and acidic environments.In particular,limited attention has been paid to stabilizing iron-based catalysts by introducing additional metal by the alloying effect.Herein,we report bimetallic Fe_(2)Mo nanoparticles on N-doped carbon(Fe_(2)Mo/NC)as an efficient and ultra-stable ORR electrocatalyst for the first time.The Fe_(2)Mo/NC catalyst shows high selectivity for a four-electron pathway of ORR and remarkable electrocatalytic activity with high kinetics current density and half-wave potential as well as low Tafel slope in both acidic and alkaline medias.It demonstrates excellent long-term durability with no activity loss even after 10,000 potential cycles.Density functional theory(DFT)calculations have confirmed the modulated electronic structure of formed Fe_(2)Mo,which supports the electron-rich structure for the ORR process.Meanwhile,the mutual protection between Fe and Mo sites guarantees efficient electron transfer and long-term stability,especially under the alkaline environment.This work has supplied an effective strategy to solve the dilemma between high electroactivity and long-term durability for the Fe-based electrocatalysts,which opens a new direction of developing novel electrocatalyst systems for future research.

Nitrogen-doped carbon nanotube-encapsulated nickel nanoparticles assembled on graphene for efficient CO2 electroreduction

Exploring 3 D hybrid nanocarbons encapsulated with metal nanoparticles(NPs)are recently considered as emerging catalysts for boosting CO2 electroreduction reaction(CRR)under practical and economic limits.Herein,we report a one-step pyrolysis strategy for fabricating N-doped carbon nanotube(CNT)-encapsulated Ni NPs assembled on the surface of graphene(N/NiNPs@CNT/G)to efficiently convert CO2 into CO.In such 3 D hybrid,the particle size of Ni NPs that coated by five graphitic carbon layers is less than 100 nm,and the amount of N dopants introduced into graphene with countable CNTs is determined to 7.27 at%.Thanks to unique CNT-encapsulated Ni NPs structure and N dopants,the achieved N/NiNPs@CNT/G hybrid displays an exceptional CRR activity with a high Faradaic efficiency of 97.7%and large CO partial current density of 7.9 mA/cm2 at-0.7 V,which outperforms those reported metallic NPs loaded carbon based CRR electrocatalysts.Further,a low Tafel slope of 134 mV/dec,a turnover frequency of 387.3 CO/h at-0.9 V,and tiny performance losses during long-term CRR operation are observed on N/NiNPs@CNT/G.Experimental observations illustrate that the Ni NPs encapsulated by carbon layers along with N dopants are of great importance in the conversion of CO2 into CO with high current density.