Search

Abstract Understanding what is important and redundant within data can improve the modelling process of neural networks by reducing unnecessary model complexity, training time and memory storage. This information is however not always priorly available nor trivial to obtain from neural networks. There are existing feature selection methods which utilise the internal working of a neural network for selection, however further analysis and interpretation of the input features’ significance is often limiting. We propose an approach that offers an extension that estimates the significance of features by analysing the gradient descent of a pairwise layer within a model. The changes that occur with the weights and gradients throughout training provide a profile that can be used to better understand the importance hierarchy between the features for ranking and feature selection. Additionally, this method is transferable to existing fully or partially trained models, which is beneficial for understanding existing or active models. The proposed approach is demonstrated empirically with a study which uses benchmark datasets from libraries such as MNIST and scikit-feat, as well as a simulated dataset and an applied real world dataset. This is verified with the ground truth where available, and if not, via a comparison of fundamental feature selection methods, which includes existing statistical based and embedded neural network based feature selection methods through the methodology of Reduce and Retrain.

Abstract Transition metal carbyne complexes are of fundamental importance in carbon-carbon bond formation, alkyne metathesis, and alkyne coupling reactions. Most reported iron carbyne complexes are stabilized by incorporating heteroatoms. Here we show the synthesis of bioinspired FeMo heterobimetallic carbyne complexes by the conversion of C2H2 through a diverse series of intermediates. Key reactions discovered include the reduction of a μ-η2:η2-C2H2 ligand with a hydride to produce a vinyl ligand (μ-η1:η2-CH = CH2), tautomerization of the vinyl ligand to a carbyne (μ-CCH3), and protonation of either the vinyl or the carbyne compound to form a hydrido carbyne heterobimetallic complex. Mechanistic studies unveil the pivotal role of H2O as a proton shuttle, facilitating the proton transfer that converts the vinyl group to a bridging carbyne.

Abstract Photocatalytic redox reactions are important for synthesizing fine chemicals from olefins, but the limited lifetime of radical cation intermediates severely restricts semiconductor photocatalysis efficiency. Here, we report that Ag3PO4 can efficiently catalyze intramolecular and intermolecular [2 + 2] and Diels-Alder cycloadditions under visible-light irradiation. The approach is additive-free, catalyst-recyclable. Mechanistic studies indicate that visible-light irradiation on Ag3PO4 generates holes with high oxidation power, which oxidize aromatic alkene adsorbates into radical cations. In photoreduced Ag3PO4, the conduction band electron (e CB −) has low reduction power due to the delocalization among the Ag+-lattices, while the particle surfaces have a strong electrostatic interaction with the radical cations, which considerably stabilize the radical cations against recombination with e CB −. The radical cation on the particle’s surfaces has a lifetime of more than 2 ms, 75 times longer than homogeneous systems. Our findings highlight the effectiveness of inorganic semiconductors for challenging radical cation-mediated synthesis driven by sunlight.

Abstract Exploration of molybdenum complexes as homogeneous hydrogenation catalysts has garnered significant attention, but hydrogenation of unactivated olefins under mild conditions are scarce. Here, we report the synthesis of a molybdenum complex, [Cp*Mo(Ph2PC6H4S−CH = CH2)(Py)]+ (2), which exhibits intriguing reactivity toward C2H2 and H2 under ambient pressure. This vinylthioether complex showcases efficient catalytic activity in the hydrogenation of various aromatic and aliphatic alkenes, demonstrating a broad substrate scope without the need for any additives. The catalytic pathway involves an uncommon oxidative addition of H2 to the cationic Mo(II) center, resulting in a Mo(IV) dihydride intermediate. Moreover, complex 2 also shows catalytic activity toward C2H2, leading to the production of polyacetylene and the extension of the vinylthioether ligand into a pendant triene chain.

Abstract The assembly of cluster-based π-stacked porous supramolecular frameworks presents daunting challenges, including the design of suitable cluster building units, control of the sufficient C-H···π interactions, trade-off between structural dynamics and stability as well as understanding the resulting collective properties. Herein, we report a cluster-based C-H···π interaction-stacked porous supramolecular framework, namely, Cu12a-π, consisting of Cu12 nanocluster as a 6-connected node, which is further propagated to a dynamic porous supramolecular frameworks via dense intralayer C-H···π interactions, yielding permanent porosity. In addition, Cu12a-π can be transformed into cluster-based nonporous adaptive crystals (Cu12b-NACs) via ligand-exchange following a dissociation-reassembly mechanism. Moreover, Cu12a-π can efficiently remove 97.2% of iodine from saturated iodine aqueous solutions with a high uptake capacity of 2.96 g·g−1. These prospective results positioned at cluster-based porous supramolecular framework and enlighten follow-up researchers to design and synthesize such materials with better performance.

Abstract Metalloligands provide a potent strategy for manipulating the surface metal arrangements of metal nanoclusters, but their synthesis and subsequent installation onto metal nanoclusters remains a significant challenge. Herein, two atomically precise silver nanoclusters {Ag14[(TC4A)6(V9O16)](CyS)3} (Ag14) and {Ag43S[(TC4A)2(V4O9)]3(CyS)9(PhCOO)3Cl3(SO4)4(DMF)3·6DMF} (Ag43) are synthesized by controlling reaction temperature (H4TC4A = p-tert-butylthiacalix[4]arene). Interestingly, the 3D scaffold-like [(TC4A)6(V9O16)]11- metalloligand in Ag14 and 1D arcuate [(TC4A)2(V4O9)]6 - metalloligand in Ag43 exhibit a dual role that is the internal polyoxovanadates as anion template and the surface TC4A4- as the passivating agent. Furthermore, the thermal-induced structure transformation between Ag14 and Ag43 is achieved based on the temperature-dependent assembly process. Ag14 shows superior photothermal conversion performance than Ag43 in solid state indicating its potential for remote laser ignition. Here, we show the potential of two thiacalix[4]arene modified polyoxovanadates metalloligands in the assembly of metal nanoclusters and provide a cornerstone for the remote laser ignition applications of silver nanoclusters.

ObjectivesTo develop: (1) two validated risk prediction models for coronavirus disease-2019 (COVID-19) positivity using readily available parameters in a general hospital setting; (2) nomograms and probabilities to allow clinical utilisation.MethodsPatients with and without COVID-19 were included from 4 Hong Kong hospitals. The database was randomly split into 2:1: for model development database (n = 895) and validation database (n = 435). Multivariable logistic regression was utilised for model creation and validated with the Hosmer-Lemeshow (H-L) test and calibration plot. Nomograms and probabilities set at 0.1, 0.2, 0.4 and 0.6 were calculated to determine sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV).ResultsA total of 1330 patients (mean age 58.2 ± 24.5 years; 50.7% males; 296 COVID-19 positive) were recruited. The first prediction model developed had age, total white blood cell count, chest x-ray appearances and contact history as significant predictors (AUC = 0.911 [CI = 0.880-0.941]). The second model developed has the same variables except contact history (AUC = 0.880 [CI = 0.844-0.916]). Both were externally validated on the H-L test (p = 0.781 and 0.155, respectively) and calibration plot. Models were converted to nomograms. Lower probabilities give higher sensitivity and NPV; higher probabilities give higher specificity and PPV.ConclusionTwo simple-to-use validated nomograms were developed with excellent AUCs based on readily available parameters and can be considered for clinical utilisation.

Abstract Organic room temperature phosphorescence (RTP) in water has attracted much attention recently for its potential biological applications. However, it remains a formidable challenge to achieve efficient RTP from pure organic compounds in aqueous phase due to the dramatic deactivation of triplet excited states in water and the poor water dispersibility of large organic particles/crystals. Represented herein is covalent incorporation of a pure organic monochromophore in silica nanoparticles (SiNPs) featuring fluorescence and bright phosphorescence in aqueous solution. The covalent bonding of organic phosphors in polysiloxane framework was found to show excellent water dispersibility, at the same time suppress the non‐radiative deactivation of triplet excited states especially from water, thus leading to high phosphorescence quantum yields (up to 22%) and long lifetimes (up to 3.5 ms) in aqueous phase. More strikingly, oxygen‐insensitive fluorescence as internal reference and oxygen‐dependent phosphorescence as oxygen indicator from the organic chromophore in the porous SiNPs realized ratiometric hypoxia detection with ultrasensitivity (KSV = 449.3 bar−1).

Controlling the structural transformations of metal clusters is challenging. Here, the authors synthesize a silver nanocluster and study the structural changes that take place upon reaction with bipyridine derivatives.

The preparation of chiral monolayer-protected metal clusters is interesting for their potential applications in a variety of fields, including catalysis. Here, the authors synthesize chiral Ag14 nanoclusters in an all-achiral environment, and decipher the origin of chirality at the molecular level; the solvent choice is key to achieve homochiral crystallization.

Identifying factors that affect research productivity is critical to both the scientific community and policy-makers. This topic is especially useful for developing countries like Vietnam where such studies are scarce with limited original data. This paper, through a manual data collection process that yields the profiles of 406 Vietnamese social scientists with publications in Scopus-indexed journals in 2008-2017, uses the ordinary least squares method to analyse the effects of two factors on research productivity. It adds to the literature by showing the extent to which (i) work environment ('universities' and 'research institutions'); (ii) collaboration affect the adjusted research productivity of social scientists. Contrary to the usual belief, university-affiliated authors in Vietnam turned out to have higher research productivity than institution-affiliated peers. International collaboration could boost research output, although this effect is insignificant among the high-performing authors. The paper also suggests some policy implications for Vietnam facing challenges in science management.

Objectives: To develop: (1) two validated risk prediction models for coronavirus disease-2019 (COVID-19) positivity using readily available parameters in a general hospital setting; (2) nomograms and probabilities to allow clinical utilisation. Methods: Patients with and without COVID-19 were included from 4 Hong Kong hospitals. The database was randomly split into 2:1: for model development database (n = 895) and validation database (n = 435). Multivariable logistic regression was utilised for model creation and validated with the Hosmer–Lemeshow (H–L) test and calibration plot. Nomograms and probabilities set at 0.1, 0.2, 0.4 and 0.6 were calculated to determine sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV). Results: A total of 1330 patients (mean age 58.2 ± 24.5 years; 50.7% males; 296 COVID-19 positive) were recruited. The first prediction model developed had age, total white blood cell count, chest x-ray appearances and contact history as significant predictors (AUC = 0.911 [CI = 0.880−0.941]). The second model developed has the same variables except contact history (AUC = 0.880 [CI = 0.844−0.916]). Both were externally validated on the H–L test (p = 0.781 and 0.155, respectively) and calibration plot. Models were converted to nomograms. Lower probabilities give higher sensitivity and NPV; higher probabilities give higher specificity and PPV. Conclusion: Two simple-to-use validated nomograms were developed with excellent AUCs based on readily available parameters and can be considered for clinical utilisation.