Search for: All records

Three fluorescent 5’-(p-hydroxyphenyl)pyridylthiazoles (HPPT) with different chelating groups at the 4’ position were synthesized and evaluated for their ability to detect transition metal ions in acetonitrile and aqueous buffers, based on changes in fluorescence intensity and intramolecular charge transfer (ICT). Both 4’-O-picolyloxy-HPPT (Pic-HPPT) and 4’-O-(o-carboxypicolyl)-HPPT (CPic-HPPT) respond strongly to Zn(II), Cd(II), and Pb(II) in CH3CN with a bathochromic shift in emission up to 68 nm, whereas 4’-O-carboxymethyl-HPPT (CM-HPPT) is unresponsive. Only CPic-HPPT responds to d10 metal ions in aqueous phosphate buffered solution (PBS, pH 7.4), attributable to the added chelating power of the ortho-carboxylate. CPic-HPPT forms a 2:1 complex with Zn(II) and a 1:1 complex with Cd(II) and Pb(II) in CH3CN, whereas a 1:1 complex forms with Zn(II), Cd(II), and Hg(II) ions in PBS. X-ray structural analysis of 1:1 Pic-HPPT–metal ion complexes reveals a planar tridentate binding motif with Zn(II) but a nonplanar tridentate geometry with the larger Cd(II) ion. Fluorescence titration of CPic-HPPT in PBS with Zn(NO3)2 established sub-micromolar sensitivity with a limit of quantitation at 50 nM. These results show that CPic-HPPT has promise as a fluorescent probe for d10 metal ions in physiologically relevant media. 

Pectin blended with cellulose nanofiber (CNF) sourced from wood pulp has excellent potential for modified atmosphere packaging (MAP), as demonstrated with refrigerated or sliced fruits enclosed in parchment coated with pectin–CNF composites. Addition of sodium borate (NaB) augments the antioxidant capacity of the composite, most likely through the generation of unsaturated pectic acid units. Packaging materials coated with pectin–CNF–NaB composites demonstrate better humidity regulation in refrigerated spaces over a 3-week period relative to uncoated controls (50% less variation), with improved preservation of strawberries as well as a reduction in the oxidative browning of sliced apples. Pectin–CNF films are both biorenewable and biodegradable as confirmed by their extensive decomposition in soil over several weeks, establishing their potential as a sustainable MAP material. Lastly, self-standing films are mechanically robust at 80% RH with tensile strength and toughness as high as 150 MPa and 8.5 MJ/m2 respectively. These values are on par with other bioplastic composites and support the practical utility of pectin–CNF composites in functional packaging applications. 

Trifluoroacetic anhydride (TFAA) reacts smoothly with low molecular weight carbohydrates and cellulose nanofibers (CNFs) under base-free conditions. Methyl α- and β-D-glucopyranoside were used as model compounds to optimize reaction conditions, which were then applied to lyophilized CNFs for surface modification. ATR-IR spectroscopy and powder X-ray diffraction were employed to characterize the modified CNFs. Trifluoroacetylation for 4 h yields a degree of substitution (DS) of 0.4 acyl groups per anhydroglucose unit while maintaining a crystallinity index near 50 %. DS values were quantified by gravimetry, acid–base titration after saponification, and a novel approach utilizing solution 19F NMR spectroscopy which offers greater accuracy than the other techniques. This study presents an efficient, base-free method for derivatizing carbohydrates as well as surface functionalization of CNFs with trifluoroacetyl groups, potentially expanding their application in fiber-reinforced thermoplastic composites. 

Introducing CF₃between two^tBu groups creates steric crowding that destabilizes the conformations of copper–pyrazolate tetramers. The impact of conformational behavior on luminescence is examined by low-temperature X-ray and photophysical studies. 

Bacteria subjected to antiseptic or antibiotic stress often develop tolerance, a trait that can lead to permanent resistance. To determine whether photodynamic agents could be used to counter tolerance, we evaluated three non-iron hemin analogs (M-PpIX; M = Al, Ga, In) as targeted photosensitizers for antimicrobial photodynamic inactivation (aPDI) following exposure to sublethal H2O2. Al-PpIX is an active producer of ROS whereas Ga- and In-PpIX are more efficient at generating singlet oxygen. Al- and Ga-PpIX are highly potent aPDI agents against S. aureus and methicillin-resistant strains (MRSA) with antimicrobial activity (3-log reduction in colony-forming units) at nanomolar concentrations. The aPDI activities of Al- and Ga-PpIX against S. aureus were tested in the presence of 1 mM H2O2 added at different stages of growth. Bacteria exposed to H2O2 during log-phase growth were less susceptible to aPDI but bacteria treated with H2O2 in their postgrowth phase exhibited aPDI hypersensitivity, with no detectable colony growth after treatment with 15 nM Ga-PpIX. 

Large language models trained for safety and harmlessness remain susceptible to adversarial misuse, as evidenced by the prevalence of "jailbreak" attacks on early releases of ChatGPT that elicit undesired behavior. Going beyond recognition of the issue, we investigate why such attacks succeed and how they can be created. We hypothesize two failure modes of safety training: competing objectives and mismatched generalization. Competing objectives arise when a model's capabilities and safety goals conflict, while mismatched generalization occurs when safety training fails to generalize to a domain for which capabilities exist. We use these failure modes to guide jailbreak design and then evaluate state-of-the-art models, including OpenAI's GPT-4 and Anthropic's Claude v1.3, against both existing and newly designed attacks. We find that vulnerabilities persist despite the extensive red-teaming and safety-training efforts behind these models. Notably, new attacks utilizing our failure modes succeed on every prompt in a collection of unsafe requests from the models' red-teaming evaluation sets and outperform existing ad hoc jailbreaks. Our analysis emphasizes the need for safety-capability parity -- that safety mechanisms should be as sophisticated as the underlying model -- and argues against the idea that scaling alone can resolve these safety failure modes. 

Large language models trained for safety and harmlessness remain susceptible to adversarial misuse, as evidenced by the prevalence of "jailbreak" attacks on early releases of ChatGPT that elicit undesired behavior. Going beyond recognition of the issue, we investigate why such attacks succeed and how they can be created. We hypothesize two failure modes of safety training: competing objectives and mismatched generalization. Competing objectives arise when a model's capabilities and safety goals conflict, while mismatched generalization occurs when safety training fails to generalize to a domain for which capabilities exist. We use these failure modes to guide jailbreak design and then evaluate state-of-the-art models, including OpenAI's GPT-4 and Anthropic's Claude v1.3, against both existing and newly designed attacks. We find that vulnerabilities persist despite the extensive red-teaming and safety-training efforts behind these models. Notably, new attacks utilizing our failure modes succeed on every prompt in a collection of unsafe requests from the models' red-teaming evaluation sets and outperform existing ad hoc jailbreaks. Our analysis emphasizes the need for safety-capability parity -- that safety mechanisms should be as sophisticated as the underlying model -- and argues against the idea that scaling alone can resolve these safety failure modes. 

Nonaqueous dispersions of graphene nanoplatelets (GrNPs) can be used to prepare thin films and coatings free of surfactants, but typically involve polar organic solvents with high boiling points and low exposure limits. Here we describe the mechanochemical exfoliation and dispersion of GrNPs in volatile aprotic solvents such as ethyl acetate (EtOAc) and acetone, which rank favorably in green solvent selection guides. GrNPs in powder form were exfoliated with solvent on a horizontal ball mill for 48 hours then sonicated at moderate power, to produce suspensions in excess of 300 µg/mL with minimum loss of dispersion stability over 7 weeks at room temperature. Atomic force microscopy (AFM) of individual particles indicate a median thickness and lateral width of 8‒10 layers and 180 nm, respectively. GrNP films can be deposited by conventional airbrush equipment with a dry time of seconds and applied as layers and coatings that enhance the reproducibility and performance of electronic devices. We demonstrate the utility of spray-coated GrNPs as contact layers for low-cost electrochemical sensing with improvements in intrabatch reproducibility, and as conformal coatings on metal heat sinks with enhanced rates of heat dissipation. 

Colloidal Ag particles decorated with Fe 3 O 4 islands can be electrochemically or photochemically activated as inverse catalysts for C(sp 2 )–H heteroarylation. The silver–iron oxide (SIO) particles are reduced into redox-active forms by cathodic charging at mild potentials or by short-term light exposure, and can be reused multiple times by magnetic cycling without further activation. A negative shift in the reduction peak is attributed to an overpotential produced by surface Fe 3 O 4 which separates residual Ag ions or clusters from bulk silver. The catalytic efficiency of SIO is maintained even with acid degradation, which can be countered simply by adding water to the reaction medium.