Effect of salinity on petroleum biodegradation

Summary The biodegradation of Ashtart crude oil by a mixed bacterial community (EH1) isolated from a marine sediment was investigated in varying concentration of sodium chloride (0 to 2 mol/l). Each fraction of Ashtart crude oil was biodegraded after a 30 day incubation period for a NaCl concentration equivalent to the one of seawater (0.4 mol/l). Saturates were more readily degraded than aromatics. The amount of oil degraded increased initially with increasing salt concentrations to a maximum level for 0.4 mol/l NaCl concentration. Thereafter the amount of oil degraded decreased with increasing salt concentrations. The polar fraction increased for the 0.4 to 1.4 mol/l NaCl concentration range. Asphaltenes were only slightly degraded (10%) for a NaCl concentration equivalent to the one of seawater (0.4 mol/l). NaCl concentrations seemed to affect the relative biodegradation of pristane and phytane. GC, UV fluorescence and FTIR techniques were used to study the evolution of each fraction of Ashtart crude oil.     

Influence of the Crystallinity of Silver Nanoparticles on Their Magnetic Properties

The magnetic properties of noble-metal nanoparticles are a puzzling phenomenon, tentatively often explained as a size effect or a ligand effect. Many experimental studies performed to date have attempted to vary these readily available parameters without reaching a definitive conclusion. In an attempt at better understanding the role of core crystallinity on these magnetic properties, we have compared the behavior of silver nanoparticles, which were either single-crystalline or multi-twinned, of almost identical sizes and with the same ligand coating. Our results indicate that single-crystalline nanoparticles tend to behave as classical paramagnetic materials, whereas multi-twinned ones exhibit a combination of para- and ferro-magnetic behaviors. Our hypothesis is that lattice defects within the core bear magnetic moments which couple through conduction electrons, with dipolar interactions also playing a local and macroscopic role.     

THE INTERACTION OF 8-AZIDOETHIDIUM AND tRNA

Abstract Optical spectroscopy and proton nuclear magnetic resonance spectroscopy have been used to investigate the non-covalent and covalent binding of 8-azidoethidium (8-azido-3-amino-5-ethyl-6-phe-nyl-phenanthridium) with transfer RNA (tRNA) molecules. The nuclear magnetic resonance measurements demonstrate that ethidium and ethidium monoazide bind to the same unique binding site on tRNA molecules. However, the optical studies demonstrate that the photochemical reaction between ethidium monoazide and tRNA generates at least three different photoproducts. One of the covalent photoproducts has fluorescence properties which mimic those of intercalated, non-covalently bound ethidium, but it is produced only in about 10% yield. The major photoproduct(s) is(are) virtually non-fluorescent. The ethidium monoazide photochemistry with tRNA molecules is obviously complex and this may complicate its use as a photolabel of polynucleotides.     

Derived Azumaya algebras and generators for twisted derived categories

We introduce a notion of derived Azumaya algebras over ring and schemes generalizing the notion of Azumaya algebras of Grothendieck (Le groupe de Brauer. I. Algèbres d’Azumaya et interprétations diverses. Dix Exposés sur la Cohomologie des Schémas, pp. 46–66, North-Holland, Amsterdam, 1968). We prove that any such algebra B on a scheme X provides a class ϕ(B) in . We prove that for X a quasi-compact and quasi-separated scheme ϕ defines a bijective correspondence, and in particular that any class in , torsion or not, can be represented by a derived Azumaya algebra on X. Our result is a consequence of a more general theorem about the existence of compact generators in twisted derived categories, with coefficients in any local system of reasonable dg-categories, generalizing the well known existence of compact generators in derived categories of quasi-coherent sheaves of Bondal and Van Den Bergh (Mosc. Math. J. 3(1):1–36, 2003). A huge part of this paper concerns the treatment of twisted derived categories, as well as the proof that the existence of compact generator locally for the fppf topology implies the existence of a global compact generator. We present explicit examples of derived Azumaya algebras that are not represented by classical Azumaya algebras, as well as applications of our main result to the localization for twisted algebraic K-theory and to the stability of saturated dg-categories by direct push-forwards along smooth and proper maps.     

Champs affines

The purpose of this work is to introduce a notion of affine stacks, which is a homotopy version of the notion of affine schemes, and to give several applications in the context of algebraic topology and algebraic geometry. As a first application we show how affine stacks can be used in order to give a new point of view (and new proofs) on rational and p-adic homotopy theory. This gives a first solution to A. Grothendieck’s schematization problem described in [18]. We also use affine stacks in order to introduce a notion of schematic homotopy types. We show that schematic homotopy types give a second solution to the schematization problem, which also allows us to go beyond rational and p-adic homotopy theory for spaces with arbitrary fundamental groups. The notion of schematic homotopy types is also used in order to construct various homotopy types of algebraic varieties corresponding to various co-homology theories (Betti, de Rham, l-adic, ...), extending the well known constructions of the various fundamental groups. Finally, just as algebraic stacks are obtained by gluing affine schemes we define $$ \infty $$-geometric stacks as a certain gluing of affine stacks. Examples of $$ \infty $$-geometric stacks in the context of algebraic topology (moduli spaces of dga structures up to quasi-isomorphisms) and Hodge theory (non-abelian periods) are given.     

Non-parametric clustering over user features and latent behavioral functions with dual-view mixture models

We present a dual-view mixture model to cluster users based on their features and latent behavioral functions. Every component of the mixture model represents a probability density over a feature view for observed user attributes and a behavior view for latent behavioral functions that are indirectly observed through user actions or behaviors. Our task is to infer the groups of users as well as their latent behavioral functions. We also propose a non-parametric version based on a Dirichlet Process to automatically infer the number of clusters. We test the properties and performance of the model on a synthetic dataset that represents the participation of users in the threads of an online forum. Experiments show that dual-view models outperform single-view ones when one of the views lacks information.