Dihydroxy aromatics in coal tar undergo rapid coupling reactions with each other and with some coalderived species during coal pyrolysis and liquefaction. However, only limited information on the structure of dihydroxy aromatics is known. In this study, Fourier transform ion cyclotron resonance mass spectrometry (FT-ICR MS) and gas chromatography-mass spectrometry (GC-MS) were used to characterize the dihydroxy aromatics in a low-temperature pyrolyzed coal tar and its distillate and residue fractions. Negative-ion electrospray ionization (ESI) FT-ICR MS spectra of coal-tar-derived samples had low molecular masses and narrow mass ranges. Negative-ion ESI is extremely selective for the dihydroxy aromatics present in coal tar. The O 2 class species had the highest ion intensity among the species in the negative-ion FT-ICR mass spectrum for each sample. Dihydroxy compounds with an aromatic core structure of benzene, indan, biphenyl (and/or acenaphthene), and naphthalene were identified in the distillate fractions. Dihydroxy compounds in the residue fraction were dihydroxy fluorenes and phenanthrenes (and/or anthracenes). The O 2 class species with relatively high molecular masses were likely dihydroxy acenaphthylenes and/or hydroxy dibenzofurans but could not be distinguished from each other because these compounds have the same molecular mass.

Adaptive gradient methods like AdaGrad are widely used in optimizing neural networks. Yet, existing convergence guarantees for adaptive gradient methods require either convexity or smoothness, and, in the smooth setting, only guarantee convergence to a stationary point. We propose an adaptive gradient method and show that for two-layer over-parameterized neural networks -if the width is sufficiently large (polynomially) -then the proposed method converges to the global minimum in polynomial time, and convergence is robust, without the need to finetune hyper-parameters such as the step-size schedule and with the level of over-parametrization independent of the training error. Our analysis indicates in particular that over-parametrization is crucial for the harnessing the full potential of adaptive gradient methods in the setting of neural networks.

A systematic analysis is presented of the B s →K form factor f (q 2 ) in the whole range of momentum transfer q 2 , which would be useful in analyzing the future data on B s →K decays and extracting ͉V ub ͉. With a modified QCD light cone sum rule ͑LCSR͒ approach, in which the contributions cancel out from the twist 3 wave functions of the K meson, we investigate in detail the behavior of f (q 2 ) at small and intermediate q 2 and the nonperturbative quantity f B* g B*B s K ͑f B* is the decay constant of B* meson and g B*B s K the B*B s K strong coupling͒, whose numerical result is used to the study q 2 dependence of f (q 2 ) at large q 2 in the single pole approximation. Based on these findings, a form factor model from the best fit is formulated, which applies to the calculation on f (q 2 ) in the whole kinematically accessible range. Also, a comparison is made with the standard LCSR predictions.A study on heavy-to-light exclusive processes plays a complementary role in the determination of the fundamental parameters of the standard model ͑SM͒ and in the development of QCD theory. At present, an important task in the SM is to extract the Cabibbo-Kobayashi-Maskawa ͑CKM͒ parameter ͉V ub ͉. Recently, a new QCD factorization formula ͓1͔ has been proposed for nonleptonic B decays and has been applied to discuss the phenomenology of B→, K, and D. This approach, however, is not adequate to extract ͉V ub ͉ from the relevant data, for the effects of the long distance QCD are anyway difficult to control in exclusive nonleptonic processes. Semileptonic B decays into a light meson, induced by the b→u transition, are regarded as the most promising processes suitable for such a purpose. It is the QCD lightcone sum rule ͑LCSR͒ approach ͓2͔ that can deal effectively with this type of decay channel. For a detailed description of this method, see Ref. ͓3͔. However, a problem to be solved is how to control the pollution by higher twist, especially twist 3 wave functions, which are not well understood and whose influence on the sum rules is considerable in most cases. In Refs. ͓4,5͔, an improved LCSR approach has been worked out, to eliminate twist 3 wave functions and enhance the reliability of sum rule calculations, and has been applied to reexamine heavy-to-light form factors in the region of momentum transfer 0рq 2 рm b 2 Ϫ2m b ⌳ QCD , where the operator product expansion ͑OPE͒ goes effectively in powers of small light-cone distance x 2 .Most previous works ͓3,4,6 -8͔ are devoted to discussing B→, semileptonic transitions within the context of LCSR, with the aim to extract ͉V ub ͉. A study on B s →K semileptonic processes is equally important. As compared with the case B→,, however, the B s →K form factors are more difficult to evaluate, for SU͑3͒ breaking corrections to the twist 3 wave functions of K meson have not been inves-*Email address: rxj@ytu.edu.cn † Mailing address.

No abstract

No abstract

Red-green-blue polymer laser emission is achieved in a free-standing membrane device consisting of three distributed feedback cavities. The polymer membrane is fabricated via interference lithography and a simple lift-off process. Multilayer structures can be assembled by cascading several polymer membranes. Thus optically pumped, simultaneous, red-green-blue laser emission is obtained from a three-layer cascaded membrane structure. This simple and low-cost fabrication technique can be used for compact, integrated laser sources.

An ultrathin plasmonic random laser is fabricated by a simple lift off process, which consists of a free-standing polymer membrane embedded with silver nanoparticles. Low threshold random lasing is observed when the 200-nm-thick membrane device is optically pumped, due to the strong plasmonic feedback and high-quality waveguide confinement provided by the silver nanoparticles and the polymer membrane, respectively. The free-standing polymer membrane is very flexible and transplantable, which can be attached to an optical fiber end face to achieve random lasing. This fabrication technique provides a promising way to realize plasmonic random lasing on surfaces with arbitrary shapes.

No abstract