DEPARTMENTAL SEMINAR - Department of Chemical and Biomolecular

Document Sample
DEPARTMENTAL SEMINAR - Department of Chemical and Biomolecular Powered By Docstoc
					                               DEPARTMENTAL SEMINAR
National University of Singapore
4 Engineering Drive 4 Singapore 117576
Tel: (65) 6516 2186    Fax: (65) 6779 1936

TOPIC                 Adsorption and Diffusion in Nanopores
                      R. KRISHNA
                      Van ‘t Hoff Institute for Molecular Sciences, University                            of
                      Amsterdam, 1018 WV Amsterdam, The Netherlands
                      Dr Jiang Jianwen

DATE                  14, 15 & 16 Oct 2008 ( Tuesday, Wednesday & Thursday)

TIME                  2.30 – 4.00 p.m.

VENUE                 E5-02-32 , Faculty of Engineering, National University of
                      Singapore NUS Campus Map & NUS: Faculty of Engineering

SYNOPSIS                  The design and development of many emerging separation and catalytic process
                      technologies require a proper quantitative description of adsorption and diffusion of
                      mixtures of different species within nano-sized pores of materials such as zeolites,
                      metal-organic frameworks (MOFs), titanosilicates (such as ETS-4, ETS-10), carbon
                      molecular sieves (CMS), and carbon nanotubes (CNTs).                For example, the
                      development of a zeolite membrane process for separating CO 2 from CH4 relies on
                      an accurate description of both adsorption and diffusion within zeolites at high
                      molecular loadings. In zeolite catalyzed isomerization and cracking processes, the
                      selectivity and product slate depends on subtle diffusional effects. The application of
                      MOFs for storage and separation applications requires fundamental insights into the
                      precise location, and movement, of molecules within the framework.

                          Molecules are first adsorbed, and then transported within nanopores. The
                      diffusion process involves hopping from one sorption site to another. There is an
                      associated energy barrier to this hopping process; diffusion is an activated process.
                      Molecules can only hop to a site that is vacant, and therefore occupancy plays an
                      essential role. The hopping rate, and the related diffusivity, depends on the
                      molecular loading. The proper description of the loading dependence of the diffusivity
                      is essential. The loading dependence of the diffusivity is dictated by several factors,
                      which include topology and connectivity, degree of confinement of guest molecules
                      (how tight is the fit?), molecule-molecule and molecule-host interactions (including
                      charge effects). For LTA zeolite that consists of cages separated by narrow windows
                      there is a significant reduction in the diffusion energy barrier with increased loading
                      of diffusant. In FAU zeolite (cages with large windows) the energy barrier is hardly
                      altered with molecular loading. CuBTC has two types of cages, and two types of
                      windows; transport in CuBTC has hybrid diffusion characteristics, having features of
                      both FAU and LTA. In one-dimensional channels of AFI, MOR, MTW, TON zeolites
                      and in carbon nanotubes and there is an increase in the diffusion energy barrier with
                      increased loading. Also, for some guest-host combinations we have single-file
                      diffusion. In some cases the pure component sorption isotherm provides a clue to the
                      understanding of the description of the loading dependence of the diffusivity. If, for
                      example, the sorption isotherm exhibits inflection behavior (perhaps due to second
                      order phase transitions), this is also reflected in the loading dependence of the
                      diffusivity. How strongly a molecule is adsorbed at a site influences the rate at which
                      it can hop to neighboring site. Configurational factors such as branching influences
                      sorption, and therefore diffusion on the basis of “leg room” considerations. The
                      length of molecule in relation to the periodicity of the potential energy field of the
                      restraining wall could yield non-monotonous molecular length effects due to
                      commensurate – incommensurate situations.
                A molecule can hop back to a site it has recently vacated and the jumps are
            therefore correlated. The proper description of the correlation effects is the key to
            the description of mixture diffusion. Correlations have the effect of slowing down
            “faster” molecules and speeding up “tardy” ones. The stronger the correlation, the
            higher are the speeding-up and slowing-down processes. The degree of correlation
            depends inter alia on the connectivity and pore size. The channel intersections of
            MFI and ISV zeolites, for example, serve as traffic lights for molecular transport.
            Correlations effects are virtually absent for inter-cage transport in LTA zeolite
            because the window is narrow and allow passage of one molecule only. In carbon
            nanotubes the correlation effects are so strong that individual species in a mixture
            traverse the tube at nearly the same rate.

                Turning to quantification of diffusion, we have to reckon with both self- and
            transport (or Fick) diffusivities. Thermodynamic effects need to be factored out of
            Fick diffusivities in order to obtain easier-to-interpret Maxwell-Stefan (M-S) (also
            called corrected) diffusivities. The Onsager formulation is particularly useful for
            extracting diffusivity data for mixtures from MD simulations of mean square
            displacements. There is no unique best formulation of diffusion: Fick, Maxwell,
            Stefan, Einstein, and Onsager will need to co-exist and are all needed. Correlation
            effects are reflected in the differences between self and M-S diffusivities. Mixture
            diffusion experiments (and simulations) are not amenable to simple interpretation; we
            need pure component adsorption and pure component diffusion data. Separation
            process development must be carried out in a systematic manner. I will also have
            something to say about transport in mesopores where Knudsen transport, along with
            surface diffusion and viscous transport, also come into play. In many cases
            molecular (MC, MD) simulations are invaluable adjuncts to experiments.

BIOGRAPHY   Krishna is a Professor at the University of Amsterdam since 1990. He is a Chemical
            Engineering graduate of the University of Bombay, and hold a PhD from the
            University of Manchester. He gained industrial experience with the Shell group of
            companies in The Netherlands and later was Director of the Indian Institute of
            Petroleum. He has published two text books on mass transfer, one of which has
            been translated into Chinese. He has more than 300 publications and several
            patents. His research interests include molecular simulations of adsorption and
            diffusion in nanoporous materials, multiphase reactor hydrodynamics, distillation,
            reactive distillation. Currently his h-index is 41.

                     ALL            ARE             WELCOME

                 Please visit our website for more details,