the current collector. Recently, progresses have been made in thick electrode architecture design by incorporating external magnetic fields and carbon templates for fast charge transfer kinetics, but the complicated producing processes and fragile electrode mechanical properties limit their ability for practical applications. [10][11][12][13][14][15] Fiber like carbon materials with large aspect ratio, such as carbon nanotubes (CNT), can significantly improve electrode mechanical strength and energy density due to its excellent electron conductivity and good compatibility to form continuous network with lower electrical percolation threshold. [16][17][18][19] Nonetheless, CNT is still constrained to complicated syntheses by expensive or low throughput methods which limits their application in bunch commercial products.Cellulose nanofiber (CNF) as an emerging biomass binder shows great potential in field of flexible and freestanding electrode fabrication due to its 1D nanostructure, excellent electrochemical stability, and robust mechanical property. [20][21][22][23][24][25][26][27][28][29][30][31][32][33][34][35][36][37][38] However, conventional CNFbased electrodes are characterized by low energy density owing to inadequate conductivity arising from the poor compatibility between CNF and conductive agents. Here, we report a conductive nanofiber network with decoupled electron and ion transfer pathways based on neutral carbon black (CB) nanoparticles and negatively charged CNF for high-loading thick electrode (up to 60 mg cm −2 ). This unique conductive CNF is achieved by a spontaneous electrostatic self-assembly technology as shown in Figure 1a. Microsize cellulose pulp was pretreated by 2,2,6,6-tetramethylpiperidine-1-oxyl (TEMPO) oxidization, which selectively oxidized the C6-hydroxyl group to a carboxyl group, leading to a strong negatively charged surface of the cellulose fibers. Negatively charged CNF was then obtained by disintegrating the microsized cellulose fiber down to the nanoscale by a probe sonication process for 1 h (Figures S1-S5, Supporting Information). Such negatively charged CNF can firmly absorb neutral CB nanoparticles by electrostatic attraction, forming a conformal conductive nanofiber. The conductive CNF can further assemble into an interconnected 3D network and tightly wrap the active electrode materials such as lithium iron phosphate (LFP) during the freeze-drying process (Figure 1b).Thick electrodes are appealing for high energy density devices but succumb to sluggish charge transfer kinetics and poor mechanical stability. Nanomaterials with large aspect ratio, such as carbon nanotubes, can help improve the charge transfer and strength of thick electrodes but represent a costly solution which hinders their utility outside of "lab scale production." Here, a conductive nanofiber network with decoupled electron and ion transfer pathways by the conformal electrostatic assembly of neutral carbon black particles on negatively charged cellulose nanofibers is reported. After integrating with ...

No abstract

We have successfully nickel doped a boron carbide ͑B 5 C͒ alloy film. The nickel doped boron-carbide ͑Ni-B 5 C 1ϩ␦) thin films were fabricated from a single source carborane cage molecule and nickelocene ͓Ni͑C 5 H 5) 2 ͔ using plasma enhanced chemical vapor deposition. Nickel doping transforms the highly resistive undoped film from a p-type material to an n-type material. This has been verified from the characteristics of diodes constructed of NiB 5 C 1ϩ␦ on both n-type silicon and p-type B 5 C. The homojunction diodes exhibit excellent rectifying properties over a wide range of temperatures.

No abstract

No abstract

No abstract

Traditional soft sensors may be ill-suited for batch processes because they cannot efficiently handle process nonlinearity and/or time-varying changes as well as provide the prediction uncertainty. Therefore, a novel adaptive soft sensor, referred to as online ensemble Gaussian process regression (OEGPR), is proposed for nonlinear time-varying batch processes. The batch process is first divided into multiple local domains using a just-in-time localization procedure, which is equipped with a probabilistic analysis mechanism to detect and remove the redundant local domains. Then the localized GPR models and probabilistic data descriptors (PDD) are built for all isolated domains. Using Bayesian inference, the posterior probabilities of any test sample with respect to different local domains are estimated and set as the adaptive mixture weights of local predictions. Further, the overall mean and variance of the predictive distribution of the target variable are estimated via the finite mixture mechanism. Additionally, the OEGPR method performs adaptation at two levels to handle time-varying behavior: (i) local GPR and PDD models; and (ii) the mixture weights. The effectiveness of the OEGPR approach is demonstrated through a simulated fed-batch penicillin fermentation process as well as an industrial fed-batch chlortetracycline fermentation process.

No abstract