Evaluation of the thermal oxidative stability of polyimides via TGA techniques

Thermogravimetric analysis (TGA) has been used for many years to evaluate polymer thermal stability. The objective of this study is to determine if weight-loss curves from TGA and isothermal TGA (IGA) can be used to determine degradation activation energies and thus rank the thermal stability (TS) and thermooxidative stability (TOS) for selected polyimides. Two high-temperature stable addition-cured polyimides and two aromatic condensation polyimides, all four containing fluorinated connecting linkages in the dianhydride monomers, were compared. Three TGA kinetic methods (Coats/Redfern, Ingraham/Marier, Horowitz/Metzger) were used to determine the activation energy for decomposition in air. The results were then used to rank polyimide stability compared to more traditional rankings based on long-term isothermal air aging weight-loss (IWL) studies and thermal decomposition temperatures (T_d) from TGA data. Use of TGA coupled to a Fourier transform infrared (TGA–FTIR) spectrophotometer allowed for the simultaneous identification and relative quantification of evolved decomposition products (CO₂, CO, ArNCO, and CHF₃) of the four polyimides degraded in air or nitrogen. Isothermal TGA–FTIR (IGA–FTIR) was also done in air to determine the relative rate of product evolution at a constant temperature. Activation energies using TGA and IGA data were determined and then compared with IWL values for the degradation of the polyimide to examine for correlations of real-life thermal oxidative aging to accelerated aging techniques. The Coats/Redfern method and T_d were found to best reproduce stability rankings of those from long-term, high-temperature IWL studies. Together, they may provide a time-saving technique to evaluate polyimide thermal oxidative stability. © 1999 John Wiley & Sons, Inc. J Polym Sci A: Polym Chem 37: 3943–3956, 1999     

Approaching Adversarial Example Classification with Chaos Theory

Adversarial examples are one of the most intriguing topics in modern deep learning. Imperceptible perturbations to the input can fool robust models. In relation to this problem, attack and defense methods are being developed almost on a daily basis. In parallel, efforts are being made to simply pointing out when an input image is an adversarial example. This can help prevent potential issues, as the failure cases are easily recognizable by humans. The proposal in this work is to study how chaos theory methods can help distinguish adversarial examples from regular images. Our work is based on the assumption that deep networks behave as chaotic systems, and adversarial examples are the main manifestation of it (in the sense that a slight input variation produces a totally different output). In our experiments, we show that the Lyapunov exponents (an established measure of chaoticity), which have been recently proposed for classification of adversarial examples, are not robust to image processing transformations that alter image entropy. Furthermore, we show that entropy can complement Lyapunov exponents in such a way that the discriminating power is significantly enhanced. The proposed method achieves 65% to 100% accuracy detecting adversarials with a wide range of attacks (for example: CW, PGD, Spatial, HopSkip) for the MNIST dataset, with similar results when entropy-changing image processing methods (such as Equalization, Speckle and Gaussian noise) are applied. This is also corroborated with two other datasets, Fashion-MNIST and CIFAR 19. These results indicate that classifiers can enhance their robustness against the adversarial phenomenon, being applied in a wide variety of conditions that potentially matches real world cases and also other threatening scenarios.     

Multimodal Biomedical Implant with Plasmonic and Simulated Body Temperature Responses

This work presents a novel nanoparticle-based thermosensor implant able to reveal the precise temperature variations along the polymer filaments, as it contracts and expands due to changes in the macroscale local temperature. The multimodal device is able to trace the position and the temperature of a polypropylene mesh, employed in abdominal hernia repair, by combining plasmon resonance and Raman spectroscopy with hydrogel responsive system. The novelty relies on the attachment of the biocompatible nanoparticles, based on gold stabilized by a chitosan-shell, already charged with the Raman reporter (RaR) molecules, to the robust prosthesis, without the need of chemical linkers. The SERS enhanced effect observed is potentiated by the presence of a quite thick layer of the copolymer (poly(N-isopropylacrylamide)-co-poly(acrylamide)) hydrogel. At temperatures above the LCST of PNIPAAm-co-PAAm, the water molecules are expulsed and the hydrogel layer contracts, leaving the RaR molecules more accessible to the Raman source. In vitro studies with fibroblast cells reveal that the functionalized surgical mesh is biocompatible and no toxic substances are leached in the medium. The mesh sensor opens new frontiers to semi-invasive diagnosis and infection prevention in hernia repair by using SERS spectroscopy. It also offers new possibilities to the functionalization of other healthcare products.     

Raman and FTIR imaging of lung tissue: bronchopulmonary sequestration

Bronchopulmonary sequestration (BPS) involves abnormal pulmonary tissue lacking trancheobronchial connection and having a separate vascular supply that arises from the thoracic or abdominal aorta. For the first time, BPS is studied by Raman and Fourier‐transform infrared (FTIR) imaging. Cryosections were mounted on calcium fluoride substrates from tissue samples of two BPS patients. Raman images were collected at a step size of 66–100 µm to assess the whole tissue section, and at a lower step size of 10 µm to resolve details in selected areas. FTIR images were collected at resolutions of 63 and 4 µm per pixel. Data sets were segmented by cluster analyses, and the spectra of each cluster were compared. The spectra revealed higher red blood cell content and lower collagen and lipid content in BPS than in an additional tissue section from the BPS margin. Comparison with previous Raman and FTIR images of congenital cystic adenomatoid malformations (CCAM) and normal lung tissue suggested that the marginal tissue sample of the BPS patient contains CCAM. We conclude that CCAM and BPS are biochemically distinct, which enable the differential diagnosis of lung malformations by specific vibrational spectroscopic fingerprints. Copyright © 2008 John Wiley & Sons, Ltd.     

Blends Based on Polystyrene and Statistical Aromatic Copolycarbonates

The compatibility of polystyrene (PS) with aromatic copolycarbonates containing bisphenol A (BPA) and tetramethyl bisphenol A (TMBPA) was investigated. The simple prevision scheme developed by Sonja Krause was employed to evaluate the effect of the copolymer molecular structure on the miscibility with polystyrene. These prevision data were used to select copolycarbonates of potential interest. Statistical copolycarbonates (CPC) containing different BPA/TMBPA molar ratios were synthesised by polycondensation reaction between a mixture of the two monomers and phosgene. PS/polycarbonates blends, prepared by casting from chloroform solution, were studied with differential scanning calorimetry and optical microscopy to evaluate the components compatibility, which increases with the TMBPA copolymer content.