Determination of toluene hydrogenation kinetics with neutron diffraction

Total neutron scattering has been used to follow the hydrogenation of toluene-d₈ to methylcyclohexane-d₁₄ over 3 wt% platinum supported on highly ordered mesoporous silica (MCM-41) at 298 K and under 150 mbar D₂ pressure. The detailed kinetic information so revealed indicates that liquid reorganisation inside pores is the slowest step of the whole process. Additionally, the results were compared with the reaction performed under 250 mbar D₂ pressure as well as with toluene-h₈ hydrogenation using D₂ at 150 mbar.

Machine learning classification of surgical pathology reports and chunk recognition for information extraction noise reduction

Background and aims: Machine learning techniques for the text mining of cancer-related clinical documents have not been sufficiently explored. Here some techniques are presented for the pre-processing of free-text breast cancer pathology reports, with the aim of facilitating the extraction of information relevant to cancer staging.

Materials and methods: The first technique was implemented using the freely available software RapidMiner to classify the reports according to their general layout: ‘semi-structured’ and ‘unstructured’. The second technique was developed using the open source language engineering framework GATE and aimed at the prediction of chunks of the report text containing information pertaining to the cancer morphology, the tumour size, its hormone receptor status and the number of positive nodes. The classifiers were trained and tested respectively on sets of 635 and 163 manually classified or annotated reports, from the Northern Ireland Cancer Registry.

Results: The best result of 99.4% accuracy – which included only one semi-structured report predicted as unstructured – was produced by the layout classifier with the k nearest algorithm, using the binary term occurrence word vector type with stopword filter and pruning. For chunk recognition, the best results were found using the PAUM algorithm with the same parameters for all cases, except for the prediction of chunks containing cancer morphology. For semi-structured reports the performance ranged from 0.97 to 0.94 and from 0.92 to 0.83 in precision and recall, while for unstructured reports performance ranged from 0.91 to 0.64 and from 0.68 to 0.41 in precision and recall. Poor results were found when the classifier was trained on semi-structured reports but tested on unstructured.

Conclusions: These results show that it is possible and beneficial to predict the layout of reports and that the accuracy of prediction of which segments of a report may contain certain information is sensitive to the report layout and the type of information sought.

Effect of type, form, and dosage of activators on strength of alkali-activated natural pozzolans

It is possible to synthesize environmentally friendly cementitious construction materials from alkali-activated natural pozzolans. The effect of the alkaline medium on the strength of alkali-activated natural pozzolans has been investigated and characterised. This paper highlights the effect of the type and form of the alkaline activator, the dosage of alkali and the SiO2/Na2O ratio (silica modulus, Ms) when using water–glass solutions and different curing conditions on the geopolymerisation of natural pozzolans. Activation of natural and calcined pozzolan for production of geopolymeric binder was verified by using Taftan andesite and Shahindej dacite from Iran as a solid precursor. The optimum range for each factor is suggested based on the different effects they have on compressive strength. The concentration of dissolving silicon, aluminium and calcium in alkaline solution, the formation of gel phase and the factors affecting this have been studied by using leaching tests, ICP–AES, and FTIR.

First-Principles Study of the Electronic and Magnetic Properties of Defects in Carbon Nanostructures

Understanding the magnetic properties of graphenic nanostructures is instrumental in future spintronics applications. These magnetic properties are known to depend crucially on the presence of defects. Here we review our recent theoretical studies using density functional calculations on two types of defects in carbon nanostructures: Substitutional doping with transition metals, and sp$^3$-type defects created by covalent functionalization with organic and inorganic molecules. We focus on such defects because they can be used to create and control magnetism in graphene-based materials. Our main results are summarized as follows: i)Substitutional metal impurities are fully understood using a model based on the hybridization between the $d$ states of the metal atom and the defect levels associated with an unreconstructed D$_{3h}$ carbon vacancy. We identify three different regimes, associated with the occupation of distinct hybridization levels, which determine the magnetic properties obtained with this type of doping; ii) A spin moment of 1.0 $\mu_B$ is always induced by chemical functionalization when a molecule chemisorbs on a graphene layer via a single C-C (or other weakly polar) covalent bond. The magnetic coupling between adsorbates shows a key dependence on the sublattice adsorption site. This effect is similar to that of H adsorption, however, with universal character; iii) The spin moment of substitutional metal impurities can be controlled using strain. In particular, we show that although Ni substitutionals are non-magnetic in flat and unstrained graphene, the magnetism of these defects can be activated by applying either uniaxial strain or curvature to the graphene layer. All these results provide key information about formation and control of defect-induced magnetism in graphene and related materials.