More Like this

1. An Adaptable Tough Elastomer with Moisture‐Triggered Switchable Mechanical and Fluorescent Properties

   https://doi.org/10.1002/adfm.201903543  

   Zhou, Xing ; Wang, Libing ; Wei, Zichao ; Weng, Gengsheng ; He, Jie ( July 2019 , Advanced Functional Materials)

   Smart materials with coupled optical and mechanical responsiveness to external stimuli, as inspired by nature, are of interest for the biomimetic design of the next generation of soft machines and wearable electronics. A tough polymer that shows adaptable and switchable mechanical and fluorescent properties is designed using a fluorescent lanthanide, europium (Eu). The dynamic Eu‐iminodiacetate (IDA) coordination is incorporated to build up the physical cross‐linking network in the polymer film consisting of two interpenetrated networks. Reversible disruption and reformation of Eu‐IDA complexation endow high stiffness, toughness, and stretchability to the polymer elastomer through energy dissipation of dynamic coordination. Water that binds to Eu³⁺ions shows an interesting impact simultaneously on the mechanical strength and fluorescent emission of the Eu‐containing polymer elastomer. The mechanical states of the polymer, along with the visually optical response through the emission color change of the polymer film, are reversibly switchable with moisture as a stimulus. The coupled response in the mechanical strength and emissive color in one single material is potentially applicable for smart materials requiring an optical readout of their mechanical properties.

    

   more » « less
2. Recent Advances in the TUH EEG Corpus: Improving the Interrater Agreement for Artifacts and Epileptiform Events

   Buckwalter, Grace Chhin ( December 2021 , Proceedings of the IEEE Signal Processing in Medicine and Biology Symposium (SPMB))

   Obeid, Iyad Selesnick (Ed.)

   The Temple University Hospital EEG Corpus (TUEG) [1] is the largest publicly available EEG corpus of its type and currently has over 5,000 subscribers (we currently average 35 new subscribers a week). Several valuable subsets of this corpus have been developed including the Temple University Hospital EEG Seizure Corpus (TUSZ) [2] and the Temple University Hospital EEG Artifact Corpus (TUAR) [3]. TUSZ contains manually annotated seizure events and has been widely used to develop seizure detection and prediction technology [4]. TUAR contains manually annotated artifacts and has been used to improve machine learning performance on seizure detection tasks [5]. In this poster, we will discuss recent improvements made to both corpora that are creating opportunities to improve machine learning performance. Two major concerns that were raised when v1.5.2 of TUSZ was released for the Neureka 2020 Epilepsy Challenge were: (1) the subjects contained in the training, development (validation) and blind evaluation sets were not mutually exclusive, and (2) high frequency seizures were not accurately annotated in all files. Regarding (1), there were 50 subjects in dev, 50 subjects in eval, and 592 subjects in train. There was one subject common to dev and eval, five subjects common to dev and train, and 13 subjects common between eval and train. Though this does not substantially influence performance for the current generation of technology, it could be a problem down the line as technology improves. Therefore, we have rebuilt the partitions of the data so that this overlap was removed. This required augmenting the evaluation and development data sets with new subjects that had not been previously annotated so that the size of these subsets remained approximately the same. Since these annotations were done by a new group of annotators, special care was taken to make sure the new annotators followed the same practices as the previous generations of annotators. Part of our quality control process was to have the new annotators review all previous annotations. This rigorous training coupled with a strict quality control process where annotators review a significant amount of each other’s work ensured that there is high interrater agreement between the two groups (kappa statistic greater than 0.8) [6]. In the process of reviewing this data, we also decided to split long files into a series of smaller segments to facilitate processing of the data. Some subscribers found it difficult to process long files using Python code, which tends to be very memory intensive. We also found it inefficient to manipulate these long files in our annotation tool. In this release, the maximum duration of any single file is limited to 60 mins. This increased the number of edf files in the dev set from 1012 to 1832. Regarding (2), as part of discussions of several issues raised by a few subscribers, we discovered some files only had low frequency epileptiform events annotated (defined as events that ranged in frequency from 2.5 Hz to 3 Hz), while others had events annotated that contained significant frequency content above 3 Hz. Though there were not many files that had this type of activity, it was enough of a concern to necessitate reviewing the entire corpus. An example of an epileptiform seizure event with frequency content higher than 3 Hz is shown in Figure 1. Annotating these additional events slightly increased the number of seizure events. In v1.5.2, there were 673 seizures, while in v1.5.3 there are 1239 events. One of the fertile areas for technology improvements is artifact reduction. Artifacts and slowing constitute the two major error modalities in seizure detection [3]. This was a major reason we developed TUAR. It can be used to evaluate artifact detection and suppression technology as well as multimodal background models that explicitly model artifacts. An issue with TUAR was the practicality of the annotation tags used when there are multiple simultaneous events. An example of such an event is shown in Figure 2. In this section of the file, there is an overlap of eye movement, electrode artifact, and muscle artifact events. We previously annotated such events using a convention that included annotating background along with any artifact that is present. The artifacts present would either be annotated with a single tag (e.g., MUSC) or a coupled artifact tag (e.g., MUSC+ELEC). When multiple channels have background, the tags become crowded and difficult to identify. This is one reason we now support a hierarchical annotation format using XML – annotations can be arbitrarily complex and support overlaps in time. Our annotators also reviewed specific eye movement artifacts (e.g., eye flutter, eyeblinks). Eye movements are often mistaken as seizures due to their similar morphology [7][8]. We have improved our understanding of ocular events and it has allowed us to annotate artifacts in the corpus more carefully. In this poster, we will present statistics on the newest releases of these corpora and discuss the impact these improvements have had on machine learning research. We will compare TUSZ v1.5.3 and TUAR v2.0.0 with previous versions of these corpora. We will release v1.5.3 of TUSZ and v2.0.0 of TUAR in Fall 2021 prior to the symposium. ACKNOWLEDGMENTS Research reported in this publication was most recently supported by the National Science Foundation’s Industrial Innovation and Partnerships (IIP) Research Experience for Undergraduates award number 1827565. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the official views of any of these organizations. REFERENCES [1] I. Obeid and J. Picone, “The Temple University Hospital EEG Data Corpus,” in Augmentation of Brain Function: Facts, Fiction and Controversy. Volume I: Brain-Machine Interfaces, 1st ed., vol. 10, M. A. Lebedev, Ed. Lausanne, Switzerland: Frontiers Media S.A., 2016, pp. 394 398. https://doi.org/10.3389/fnins.2016.00196. [2] V. Shah et al., “The Temple University Hospital Seizure Detection Corpus,” Frontiers in Neuroinformatics, vol. 12, pp. 1–6, 2018. https://doi.org/10.3389/fninf.2018.00083. [3] A. Hamid et, al., “The Temple University Artifact Corpus: An Annotated Corpus of EEG Artifacts.” in Proceedings of the IEEE Signal Processing in Medicine and Biology Symposium (SPMB), 2020, pp. 1-3. https://ieeexplore.ieee.org/document/9353647. [4] Y. Roy, R. Iskander, and J. Picone, “The NeurekaTM 2020 Epilepsy Challenge,” NeuroTechX, 2020. [Online]. Available: https://neureka-challenge.com/. [Accessed: 01-Dec-2021]. [5] S. Rahman, A. Hamid, D. Ochal, I. Obeid, and J. Picone, “Improving the Quality of the TUSZ Corpus,” in Proceedings of the IEEE Signal Processing in Medicine and Biology Symposium (SPMB), 2020, pp. 1–5. https://ieeexplore.ieee.org/document/9353635. [6] V. Shah, E. von Weltin, T. Ahsan, I. Obeid, and J. Picone, “On the Use of Non-Experts for Generation of High-Quality Annotations of Seizure Events,” Available: https://www.isip.picone press.com/publications/unpublished/journals/2019/elsevier_cn/ira. [Accessed: 01-Dec-2021]. [7] D. Ochal, S. Rahman, S. Ferrell, T. Elseify, I. Obeid, and J. Picone, “The Temple University Hospital EEG Corpus: Annotation Guidelines,” Philadelphia, Pennsylvania, USA, 2020. https://www.isip.piconepress.com/publications/reports/2020/tuh_eeg/annotations/. [8] D. Strayhorn, “The Atlas of Adult Electroencephalography,” EEG Atlas Online, 2014. [Online]. Availabl 

   more » « less
3. Atomistic mechanisms of adhesion and shear strength in graphene oxide-polymer interfaces

   Jin Y. Choi, Xu Zhang ( January 2021 , Journal of the mechanics and physics of solids)

   Combining experimental and computational studies of nanocomposite interfaces is highly needed to gain insight into their performance. However, there are very few literature reports, combining well-controlled atomic force microscopy experiments with molecular dynamic simulations, which explore the role of polymer chemistry and assembly on interface adhesion and shear strength. In this work, we investigate graphene oxide (GO)-polymer interfaces prevalent in nanocomposites based on a nacre-like architectures. We examine the interfacial strength resulting from van der Waals and hydrogen bonding interactions by comparing the out-of-plane separation and in-plane shear deformations of GO-polyethylene glycol (PEG) and GO-polyvinyl alcohol (PVA). The investigation reveals an overall better mechanical performance for the anhydrous GO-PVA system in both out-of-plane and in-plane deformation modes, highlighting the benefits of the donor-acceptor hydrogen bond formation present in GO-PVA. Such bond formation results in interchain hydrogen bond networks leading to stronger interfaces. By contrast, PEG, a hydrogen bond acceptor only, relies primarily on van der Waals inter-chain interactions, typically resulting in weaker interactions. The study also predicts that water addition increases the adhesion of GOPEG but decreases the adhesion of GO-PVA, and slightly increases the shear strength in both systems. Furthermore, by comparing simulations and experiments, we show that the CHARMM force field has enough accuracy to capture the effect of polymer content, water distribution, and to provide quantitative guidance for achieving optimum interfacial properties. Therefore, the study demonstrates an effective methodology, in the Materials Genome spirit, toward the design of 2D materials-polymer nanocomposites system for applications demanding mechanical robustness. 

   more » « less
4. Emerging interface materials for electronics thermal management: experiments, modeling, and new opportunities

   https://doi.org/10.1039/C9TC05415D  

   Cui, Ying ; Li, Man ; Hu, Yongjie ( January 2020 , Journal of Materials Chemistry C)

   Thermal management is becoming a critical technology challenge for modern electronics with decreasing device size and increasing power density. One key materials innovation is the development of advanced thermal interfaces in electronic packaging to enable efficient heat dissipation and improve device performance, which has attracted intensive research efforts from both academia and industry over the past several decades. Here we review the recent progress in both theory and experiment for developing high-performance thermal interface materials. First, the basic theories and computational frameworks for interface energy transport are discussed, ranging from atomistic interface scattering to multiscale disorders that contributed to thermal boundary resistance. Second, state-of-the-art experimental techniques including steady-state and transient thermal measurements are discussed and compared. Moreover, the important structure design, requirements, and property factors for thermal interface materials depending on different applications are summarized and exemplified with the recent literature. Finally, emerging new semiconductors and polymers with high thermal conductivity are briefly reviewed and opportunities for future research are discussed. 

   more » « less
5. Influence of the Graft Length on Nanocomposite Structure and Interfacial Dynamics

   https://doi.org/10.3390/nano13040748  

   Genix, Anne-Caroline ; Bocharova, Vera ; Carroll, Bobby ; Dieudonné-George, Philippe ; Chauveau, Edouard ; Sokolov, Alexei P. ; Oberdisse, Julian ( February 2023 , Nanomaterials)

   Both the dispersion state of nanoparticles (NPs) within polymer nanocomposites (PNCs) and the dynamical state of the polymer altered by the presence of the NP/polymer interfaces have a strong impact on the macroscopic properties of PNCs. In particular, mechanical properties are strongly affected by percolation of hard phases, which may be NP networks, dynamically modified polymer regions, or combinations of both. In this article, the impact on dispersion and dynamics of surface modification of the NPs by short monomethoxysilanes with eight carbons in the alkyl part (C8) is studied. As a function of grafting density and particle content, polymer dynamics is followed by broadband dielectric spectroscopy and analyzed by an interfacial layer model, whereas the particle dispersion is investigated by small-angle X-ray scattering and analyzed by reverse Monte Carlo simulations. NP dispersions are found to be destabilized only at the highest grafting. The interfacial layer formalism allows the clear identification of the volume fraction of interfacial polymer, with its characteristic time. The strongest dynamical slow-down in the polymer is found for unmodified NPs, while grafting weakens this effect progressively. The combination of all three techniques enables a unique measurement of the true thickness of the interfacial layer, which is ca. 5 nm. Finally, the comparison between longer (C18) and shorter (C8) grafts provides unprecedented insight into the efficacy and tunability of surface modification. It is shown that C8-grafting allows for a more progressive tuning, which goes beyond a pure mass effect. 

   more » « less