Third DOE/BES Separations Research
Presentation Abstracts

Last Modified May 18, 1999

High Pressure and High Temperature Spectroscopic Studies of Supercritical Fluid Solutions
Clement R. Yonker

Understanding on a molecular level the intermolecular interactions underlying separations and extractions in supercritical fluids is important to extend these useful solvents to new separation technologies and chemical synthesis in addition to providing a basis for improving existing processes. The approach involves the use of techniques such as FT-IR, time-resolved FT-IR, and NMR spectroscopies.  These investigations in supercritical fluids result in an improved understanding of condensed phase thermodynamics and kinetics by bridging the gap between the gaseous and liquid states.  NMR relaxation measurements provide information about the rotational and translational motion of molecules in solution.  The T1 values of methanol as a function of pressure and temperature were investigated to address the role of hydrogen-bonding on solution relaxation rates. The effect of pressure and temperature on hydrogen-bonding in pure methanol was investigated at pressures up to 3500 bar and temperatures to 773K.  In a similar manner, high-pressure NMR was used to investigate the equilibrium phase behavior of two binary solvent systems, ethylene/methanol and propane/methanol.  The phase behavior was investigated as a function of pressure and temperature, with the molar composition of both phases being determined simultaneously.  The hydrogen-bonding behavior of methanol was determined in both the liquid and vapor phase.  High pressure NMR is an efficient method for obtaining vapor liquid equilibrium data and critical conditions of binary supercritical fluid solvent systems.  Kinetic studies with a capillary flow reactor and NMR detection was used to investigate deuterium substitution reactions under supercritical water conditions.  These in situ, real-time measurements under the extreme conditions of temperature (473K to 723K ) and pressure in supercritical water provided first-order rate constants for the H/D exchange process.  Time-resolved FTIR investigations of organometallic substitution reactions in supercritical fluids are being initiated and will be discussed.  The basic knowledge and new technologies generated from these research efforts related to the chemistry and physics of supercritical fluid solutions will find many ready applications in separations technologies, solvent substitution processes, carbon management and environmental remediation.

Intermolecular Forces between a Protein and a Hydrophilic Modified Polysulfone Film with Relevance to Filtration
Georges Belfort

Correlations between intermolecular forces and ultrafiltration measurements for a thin polysulfone film and membrane modified for increased hydrophilicity by graft polymerization of 2-hydroxyethyl methacrylate and a model protein (hen egg-white lysozyme, Lz) suggest that altering either the chemistry of the polymer surface or the solution conditions should lead to a minimization of protein adhesion and hence fouling for a specific protein/polymer combination.  Using the Surface Forces Apparatus, normalized forces were measured below, at and above the pI of Lz and compared with four different permeation fluxes from ultrafiltration experiments.  Simple linear correlations were obtained relating the normalized adhesion forces to the four different permeation fluxes .  Also, the amount of protein adsorbed onto the membrane from solution during filtration was linearly related to the adhesion force through the choice of solution pH.  The correlations imply that both protein-protein and protein-polymer interactions are important during ultrafiltration.  Therefore, either altering the chemistry of the polymer surface or the solution conditions could lead to a minimization of fouling for a specific protein/polymer combination.  The hydrophilic surface exhibited lower contact angles, reduced adhesion, reduced adsorbed amount and most importantly reduced protein fouling as compared with a hydrophobic surface.

Spectroscopic probing of transport and ion-exchange in living polymer films on silica
Mary Wirth

Controlled growth of living polymer films on silica is promising as a means of preparing advanced materials for large-scale extraction of ions from contaminated water, and as media for electrochromatographic analysis of ions.  Little is known about transport properties in these new media, giving no fundamental guidance to optimize exchange capacity and speed.  Single-molecule spectroscopy is shown in this presentation to enable the direct observation of ion-exchange, as well as diffusion between ion-exchange sites.  Charged organic dyes having high quantum efficiencies are used as the ions, and sulfonated, living polystyrene films covalently bonded to fused silica are used as the ion-exchangers media.  The film thickness and ionic strength are varied, revealing the relation between these factors and the film viscosity and capacity.

Hydrophobic Surface Forces and the Nature of Interfacial Water Structure
J. D. Miller

The interaction forces between smooth solid surfaces have been studied in great detail, with aid of the surface force apparatus and, more recently, the atomic force microscope. The methodologies for the study of interactions between essentially nondeformable surfaces are well established, and the results have been shown to be in reasonable agreement with the theoretical DLVO presentations. However, even for nondeformable surfaces, the hydrophobic attraction force, which is presumed to play a key role in bubble-particle interaction is not well understood. Most experimental studies have focussed on the symmetric interaction between solid surfaces (e.g. mica) that have been hydrophobized, usually by silanation or Langmuir-Boldgett deposition. In this case the range of the interaction has been shown to vary greatly in magnitude from 5 to 100 nm or more and to be critically dependent upon the type and mode of preparation of hydrophobic surfaces. Efforts are being made to further describe the hydrophobic surface force and the significance of interfacial water structure by AFM and spectroscopic measurements.

Anion Recognition: A New Tool for Enhancing Ion-Pair Extraction
Richard A. Sachleben, Jeffrey C. Bryan, and Bruce A. Moyer

Anion recognition is a relatively undeveloped area of host-guest chemistry. We have undertaken an effort to design and study specific anion-complexants for use in combination with cation extractants to solve the 'anion problem' in solvent extraction. Solvent extraction of cations using neutral ligands requires co-extraction of an anion to maintain charge neutrality. For hydrophilic anions, such as nitrate, unfavorable solvation contributes to weak extraction in the absence of additional measures. Conceptually, adding an anion-recognizing extractant to a solvent extraction system should increase ion-pair extraction. Towards this end, we have designed and synthesized a tripodal amide, N,N',N"-(2,5-bis-t-butylphenyl)benzene-1,3,5-tricarboxamide (1), which is capable of complexing nitrate anion through cooperative interaction between three appropriately oriented hydrogen-bonds. Extraction experiments show that while 1 does not by itself extract cesium nitrate, when used at a ratio of 2:1 with the known cesium-selective crown ether B424C8, cesium nitrate extraction doubles relative to extraction using B424C8 alone. NMR titrations of 1 with tetrabutyl ammonium nitrate in d4-DCE indicate complexation of the nitrate anion by 1 through H-bonding and provide evidence for the cooperative interaction of three hydrogen bonds. solid-state structural studies suggest that improved predisposition of the amide groups for hydrogen-bonding to nitrate may be achievable.

Unusual Solvent/Cesium Interactions: Structural Origins and Challenges for Separations

Jeffrey C. Bryan, Tatiana Levitskaia, Richard A. Sachleben, John D. Lamb, Bruce A. Moyer

Crown ether research has traditionally been directed toward understanding the relationship between the structure of the macrocycle and its ability to bind and transport guest species, especially alkali and alkaline earth metal cations.  Less understood is the fact that the resulting metal-ion crown complex can selectively interact with solvent molecules and counter anions.  In general, crown molecules incompletely fill the coordination sphere of the cation, allowing coordination of solvent molecules, anions, or other crown molecules.  Occasionally, the space occupied by the solvent molecule or anion is sufficiently well defined as to suggest a host-guest relationship between the metal-crown complex and the solvent or anion.  We have structurally characterized several cesium complexes with tetrabenzo-24-crown-8 which exhibit unprecedented metal-solvent interactions.  Specifically we have observed the first examples of 1,2-DCE, methylene chloride and h2- acetonitrile coordination to an alkali metal ion.  The structural origins of these unusual interactions will be discussed, as well as the challenges they present for solvent extraction studies.

Molecular Simulations of Osmosis and Reverse Osmosis

S. Murad

The method of molecular dynamics has been used to study osmosis and electro-osmosis in solutions that include electrolyte solutions in both  water and other polar solvents, such as methanol. These studies have been carried out using a method developed by us recently to study fluids confined by membranes. Our results have shown the significant role solvation forces play in reverse osmosis based separations. We have also found that external electric fields can be used to increase the rate of reverse osmosis based separations in a wide range of solutions -- a phenomena sometimes referred to as electro-osmosis. The overall objective of these studies has been to understand these phenomena at the fundamental molecular level.

Computer simulations have also been used for screening studies to determine the suitability of membranes for separating electrolyte solutions. As an example, we have carried out simulations to examine the feasibility of using thin silacalite (zeolite) membranes to separate aqueous electrolyte solutions.

Combining Spectroscopy and Chromatography in Analytical Chemistry
Edward S. Yeung

Chromatographic separation is a statistical process involving many repeated interactions between the molecules in a moving stream and an immobilized surface.  The standard picture is that molecules occasionally bind to the surface and become delayed relative to the bulk motion.  For the first time, images of individual protein molecules are recorded as they approach a fused-silica surface.  Charge interaction causes the molecules to be trapped in the interfacial liquid layer for tens of milliseconds.  This constitutes the direct verification of the statistical theory of chromatography.  Microscopic reversibility is conserved.  However, the molecules were not immobilized as portrayed in conventional models.  They are simply held near the surface by attractive forces and can diffuse freely within the interfacial layer.  The interaction distances are also found to be much longer than predicted by the electrostatic double-layer thickness.  The results imply that molecule/surface interactions are considerably more efficient than expected.  This is perhaps why in nature cell-surface receptors work so well in recognizing very low concentrations of target molecules.

Improved Separation of Metal Ions by Extraction Chromatography or Crown Ether-Synergized Ion Exchange
Mark L. Dietz (1), Renato Chiarizia (1), and Richard A. Bartsch (2)

Solvent extraction and ion exchange are among the most well-established methods for metal ion separations. Despite several advantages over competing techniques, however, each suffers from shortcomings that limit its utility. Solvent extraction, for example, is frequently regarded as too cumbersome; similarly, ion exchange typically lacks the selectivity required to deal with complex sample solutions.

Recent work in this laboratory has been directed at the development of metal ion sorbents combining the selectivity of an extraction process with the ease of handling of an ion exchange resin. One approach, referred to as extraction chromatography, involves the impregnation of an inert support with an appropriate extractant or extractant-diluent combination. An alternative approach involves the modification of the metal ion uptake properties of a conventional ion-exchange resin by addition of a water-soluble, ion-selective ligand to the aqueous phase. The result is frequently a synergistic enhancement of metal sorption and an improvement in the selectivity of the resin.

The Aptamer Approach to Chemical Separations
L. B. McGown

Aptamers are short oligonucleotides that are selected through combinatorial processes to bind to specific target molecules or macromolecules with high selectivity and binding affinity.  We are exploring the ability of various structural motifs associated with particular aptamer sequences to interact with non-target molecules for chemical separations. Our initial investigations focus on interactions of G-quartet, thrombin binding sequences with lanthanides and with PAHs.  Spectroscopic evidence supporting the potential of the G-quartet for these separations will be presented and initial work on attachment of aptamers to capillaries for CEC separations will be described.

Transitions in the Solvation Structure about Ionic Species in Supercritical Water Measured by X-ray Absorption Spectroscopy
John L. Fulton

Dramatic structural transitions occur in the first solvation shell of ionic species in water at temperatures just below or above the critical point (Tc = 375°C).  Most importantly, anions form contact pairs with cations.  This type of transition is driven in part by the strong electrostatic attraction between oppositely charged ions in an aqueous environment where the dielectric constant is dramatically lower due to the breakup of the water hydrogen-bonding. 

X-ray absorption fine structure (XAFS) spectroscopy can be used to precisely measure the atomic structure in the first solvation shell under hydrothermal conditions.  We used XAFS and two additional spectroscopic techniques to elucidate the detailed molecular structure about  Ni2+ and Cu2+solutions from 25°C to 525°C and 1 to 720 bar.  In one case, both Ni and Br XAFS spectra were acquired on a single solution and the structural parameters (types, numbers and distances) were extracted by analysis of both data sets using a single global model.  The derived results give a very high degree of confidence in the measured parameters.  Second, the x-ray absorption pre-edge peaks corresponding to the 1s-to-3d and to the 1s-to-4p electronic transitions were used to confirm the change in coordination and further to refine the symmetry structure of the new hydrothermal species.  Finally, crystal field spectra taken in the near-infrared region were used to confirm the symmetry structure derived from the XAS measurements. 

At room temperature, the octahedral Ni2+H2O)6 species persists at all salt concentrations.  This species is still prevalent at 325°C, but at higher temperatures it is replaced by four-coordinate structures.  Above 425°C, at moderate pressures up to 700 bar, the stable structures are a family of four-coordinated species (NiBr(H2O)3Br, NiBr2(H2O)23(H2)-Na) where the degree of Br- adduction and replacement of H2O in the inner shell depends upon the overall Br- concentration.  The most likely symmetry of these species is a distorted tetrahedron.  Thus, we report a definitive structural characterization of ionic species at high temperatures.  The results are of interest for predicting salt solubilities in hydrothermal processes including the separation of ionic species from aqueous waste streams.  They are also applicable to the areas of geochemistry and to power plant corrosion chemistry in which analogous chemical systems are found.

Single Particle Laser Ablation Time-of-Flight Mass Spectroscopy:
D. G. Imre

The construction of a Single Particle Laser Ablation Time of Flight Mass Spectrometer (SPLAT-MS) for in-situ size and chemical composition characterization of individual aerosol particles over a wide size range has been designed and is under way.  The ideal in-situ aerosol characterization research tool should be able to provide information on the size and chemical composition of individual particles. Recent applications of mass spectroscopies have shown great promise to this end. What makes this problem particularly difficult is the fact that for most applications aerosol can be as small as a few nm or as large as tens of microns. We are constructing an instrument that can be operated in either large or small particle detection modes. The larger particles (>500nm) are easily detectable but in most settings tend to be very rare. To detect this particle mode we use aerodynamic lens techniques to concentrate these rare particles and enhance their detection probability. Optical detection and sizing with synchronized ablation provide size and chemical composition.

The small aerosol operation mode ( 500 to 10nm size range) uses a different aerodynamic lens design and consequently pumping scheme. Particles 500 to 150nm can be optically detected and sized as in the large particle mode.  However, for particles smaller than 150nm it is no longer possible to synchronize the firing of the ablation laser with particle's arrival, based on particle detection by conventional elastic scattering. The solution to this problem has thus far been to randomly fire the ablation laser such that on the rare occasion of coincidence between laser and particle a spectrum is produced. This approach results in a very low duty factor.  This size limitation can be overcome by taking advantage of the difference in the ionization threshold between gas and particle. In particular, most carrier and atmospheric gases have ionization thresholds that are significantly higher than those of the condensed phase aerosol.  This difference can be exploited to generate a signal that signifies the presence of a particle. The choice of ionization radiation is crucial in order to maintain a high contrast between particle and the gas the wavelength must be selected, so that particles alone would be ionized.  Here a high flux rare gas VUV light source will be used. This light source is capable of delivering ~1016photonssec-1sr-1. When operating with He, this light source will definitely ionize both aerosol and surrounding gas (HeI - 21.2eV). But using the  Ar (ArI - 11.8eV), Kr (KrI - 10 eV) or Xe (XeI - 9.6eV) will allow us selectively ionize aerosols. The detection of a free electron indicates the presence of a particle and can be used to generate a trigger signal to fire the ablation laser and obtain a TOF-MS.  Alternatively, whole particle mass can be to determine using the TOF-MS directly.

At present this scheme will enable determination of particle size distributions, in whole particle mass analysis, or particle composition.  Once the system will be in operation we will explore various schemes to obtain simultaneously size and composition information about nanoparticles.  We will test the feasibility of deriving particle size from the total photoelectron signal or from the total ion signal or both. We will also explore the possibility of VUV elastic scattering for size and detection.  Here the use of HeII with a ~30nm wavelength which is rather close to the particle's size may prove to be a useful tool in distinguishing nanoparticles from gas phase molecules.

Molecular Modelling of Reversed Micellar Aggregates in Liquid/Liquid Extraction
Ronald Newman

Acidic organophosphorus extractants such as bis(2-ethylhexyl)phosphoric acid (HDEHP) and bis(2,4,4-trimethylpentyl)phosphinic acid (CYANEX 272) have received wide usage in liquid/liquid extraction. Traditionally, metal-extractant complexes in the organic phase under low loading (dilute) conditions have been characterized by slope analysis. On the other hand, as the loading increases, the composition of the metal-extractant complexes appears to change and aggregates form. Unfortunately, relatively little knowledge is known about the physicochemical nature of the metal-extractant aggregates, described as either polymers or reversed micelles in the technical literature, for which information is very difficult to obtain from direct experimental measurements in highly loaded systems. One approach, however, which has great potential for clarifying the molecular structure of metal-extractant aggregates is that of molecular modelling.

In this presentation it will be shown how molecular modelling techniques can be employed to improve the understanding of the liquid/liquid extraction of metal ions such as nickel (II) and cobalt (II) from aqueous to nonpolar organic phases. In particular, molecular modelling was employed to study the state of aggregation of the metal salts of HDEHP and its phosphonic and phosphinic acid analogs. The organization of the metal-extractant complexes, the nature of the intermolecular forces, and the orientation and location of water molecules in the metal-extractant aggregates which form under conditions simulating those of practical extraction processes were determined. The effects of extractant structure, metal ions, solvent type and water molecules on the aggregate structure were also examined. Both geometrical optimization and molecular dynamics were employed using HyperChem and SYBYL molecular simulation software. Molecular modelling shows that the metal -extractant aggregates are quasi-one-dimensional or rodlike reversed micelles in both alkane and aromatic solvents. Furthermore, the solubilized water molecules are not situated in the micellar core, as expected according to the classical model of reversed micelles, but instead are localized in channels within the surface of Ni(DEHP)2 reversed micellar aggregates, thereby confirming the recently proposed (ISEC '93) "open water-channel model" of reversed micelles. Although the nickel (II) salts of the phosphonic and phosphinic acid analogs also form rodlike reversed micelles, the surface of the micellar structures tends to be more lyophilic (or hydrophobic) in the order phosphoric < phosphonic < phosphinic because of structural differences associated with the amount of solubilized water which also decreases in the same order. Significantly, open water channels were not present in the surface of the rodlike reversed micelles of the Co (II) salt of HDEHP. Clearly, molecular modelling shows the dramatic effects that even minor changes in the extractant structure or metal ion substitution have on the reversed micellar structure of acidic organophosphorus reagents. These novel findings will be discussed from the perspective of extractability and selectivity in metal extraction processes.

Robin Rogers

Aqueous Biphasic Systems (ABS) form upon the admixture of certain water-soluble polymers or polymers and salts above critical concentrations, and thus represent liquid/liquid extraction systems whose nature is entirely aqueous.  These systems warrant consideration for the development of environmentally-benign wet extraction processes, a role which is well outside, and rather different, from their current rather limited application in biotechnological separations.  In support of this contention, it may be noted that strong similarities exist at the molecular, experimental, and design implementation levels of several apparently diverse systems which might be termed Wholly Aqueous Solvent Extraction (WASE) systems.  These include Cloud Point Extraction, Thermoseparating Polymer Systems (TSP), ABS, and Micellular Extraction.

In light of these similarities (which are often ignored in the literature unique to each area), the fact that each represents an alternative to traditional solvent extraction techniques employing volatile organic compounds (VOCs), and that each has the potential to be developed in the form of a solid phase analog (e.g., Aqueous Biphasic Extraction Chromatographic resins, ABEC, which can mimic ABS separations), the range of potential application in separation science broadens considerably.  However, apart from well known developments in the extraction of proteins in the biotechnology industry, there is little evidence of the widespread adoption of WASE technologies owing to a paucity of adequate design criteria available to the engineering community.  It seems likely that a valuable collection of polymer-based extractive processes which could represent a 'toolbox' of relatively environmentally-benign options in the design of extractive chemical processes, in fact appear as a bewildering array of disparate techniques.

ATR FT/IR Studies of Adsorbates at Liquid/Solid Interfaces using Silica Sol-Gel Films as Model Systems for Normal Phase Chromatography.
Joel Harris (Dion Rivera substituting for Joel Harris)

This work employs ATR FT-IR to study the structure of adsorbates at the liquid/solid interface of a high surface area silica sol-gel film.  The film is prepared by dip coating a germanium ATR element with a 0.3% silica sol suspension composed of silica 50 nm diameter particles.  This results in a film 400 nm thick that is stable even in hydrogen bonding solvents.  The film is used as model surface for normal phase liquid chromatography and allows direct structural evidence of the interactions that are responsible for retention of molecules on the silica surface to be acquired.  This information is generally not obtainable from chromatographic studies of the interface.  Adsorption studies of ethyl acetate have produced high quality, in situ infrared spectra.  Fitting the data using Langmuir isotherm models and classical least squares analysis to resolve component spectra reveals two distinct adsorption sites on the silica surface.  The spectra indicate that the sites of adsorption are likely free silanols and surface bound water.  Displacement studies of ethyl acetate with 2-propanol have produced spectra that show competition between adsorbates for the sites on silica.  These studies elucidate the nature of interactions between polar and hydrogen bonding molecules on the silica surface.

High performance membrane materials for gas vapor and liquid separations
W. J. Koros, D. R. Paul

This project focuses on understanding performance tradeoffs between conventional polymers, crosslinked polymers and molecular sieving materials for membrane-based separations.   While rigid molecular sieving carbons and zeolites resist swelling in aggressive environments, they have the potential to "plug" due to competition for sorption and transport pathways by co-permeating components and impurities in feeds.   Moreover, the high cost of forming molecular sieving membranes for large-scale applications encouraged us to assess alternative approaches such as covalently crosslinked polymers.  Such an option can benefit from existing economical processes for conventional uncrosslinked polymers.  By incorporating an additional appropriate precursor in the backbone, a reactive post-treatmentt allows increasing stability after primary membrane formation is completed.  The pros and cons associated with such a post-formation crosslinking approach have not been investigated systematically, and exploring this issue is an important part of our project.

Polyimides are ideal for such a systematic comparative study of the various membrane types, since samples synthesized for our crosslinking work can also be useful precursors for pyrolytic formation of carbon molecular sieves.  Both crosslinking and pyrolysis processes reduce the flexibility of the polymer matrix, but the two modifications create rather different morphological effects.  We are comparing performance changes caused by crosslinking in one case and pyrolyzing to a glassy carbon in the other case from nearly identical starting materials.    Fundamentally different responses for the two approaches result, since pyrolysis treatments create extended planar aromatic sheets, leading to actual slit-like pores that may serve as molecular sieving sites.  On the other hand, the crosslinked polymers retain much of their intrinsic polymer repeat unit nature, while minimizing detrimental swelling at higher feed pressures or temperatures. 

The evolution in membrane transport and physical properties during the transition from a starting uncrosslinked, unpyrrolyzed material to a carbon sieve membrane has been done for one of our samples already.  Work with a systematically more crosslinked analog is currently underway for comparison.  High resolution TEM and AFM have not provided much help in distinguishing the ultrafine morphological changes between samples. Modeling of the structure to draw detailed inferences from the x-ray results is also ambiguous, given the complex nature of the amorphous glassy polymer and glassy carbon solid states.  The traditional use of fractional free volume estimates, gas permeation and sorption studies, coupled with x-ray diffraction have, however, been more useful in sample characterization.

The following dianhydrides (underlined) and diamine monomers are being used:

Since beginning this new project in April 1998, we have obtained samples of the following six polyimides from these monomers, and measurements are underway to characterize them as outlined above.

6FDA-DAM (1:0-1:0)

6FDA:BPDA-DAM (1:1-1:0)

6FDA-BPDA-DAM (9:1-1:0) 6FDA-BPDA-DAM (8:2-1:0)

6FDA-DAM-DABA (1:0-9:1)

6FDA-DAM-DABA (1:0-8:2)

The "code" used to represent the above structures is simple.  First, underlined dianhydride monomers (6FDA & BPDA) are indicated in terms of the relative amounts of 6FDA:BPDA dianhydrides in the polyimide (e.g. 1:0 or 9:1, etc.).  This grouping is connected by a dash to the relative amounts of DAM and DABA diamino monomers in the polyimide (e.g. 1:0 or 9:1), resp.   For example, one of the combinations of the above monomers, 1:0-1:0, is a simple polyimide composed of equimolar 6FDA and DAM.   Another material, 1:1-1:0, is a copolyimide composed of equimolar 6FDA and BPDA dianhdyrides with only the DAM diamine present.  This sample has been used in the carbon molecular sieve evolution study mentioned above.  An unpyrolyzed 1:1-1:0 sample is currently being characterized, and results for the unpyrolyzed and fully pyrolyzed (to 800°C) samples will be compared at the workshop.

Self-Assembled Ionophores: Novel and Efficient Method for Ion Separation
Jeffery T. Davis

Self-Assembled Ionophores: Novel and Efficient Method for Ion Separation.  

We fiind that "Self-assembled" ionophores made from lipophilic nucleosides coordinate alkali and alkaline earth cations with high affinity.  These nucleosides form cyclic, hydrogen bonded complexes that have cavities for ion complexation. Because synthesis is under thermodynamic control, ion-templated assembly via non-covalent synthesis is relatively efficient.  Some of these ionophores are Cs+ selective, making them unusual and potentially valuable compounds.  Our major goals are:

1) To understand structural and dynamic factors that control self-association and cation binding.

2) To design new self-assembled ionophores that selectively bind cations.

Our guanosine analogs self-associate and coordinate cations to give ionophores of defined geometry.  We use NMR, X-ray crystallography, and electrospray MS to characterize structure.  Ion binding thermodynamics have been determined using calorimetry.  Current problems include the dynamics of ligand and Cs+ exchange. Is there is a correlation between the complexes' thermodynamic and kinetic stablities? Our studies also have transport applications, as these lipophilic nucleosides move ions across liquid membranes.

At the Separations Workshop I am interested in better understanding the thermodynamics and kinetics of our system, and discussing ideas about new ionophores for cations and anions.

Probing Mesoscale Structures in Supercritical Carbon Dioxide
Hank D. Cochran

Supercritical carbon dioxide is a desirable solvent substitute, but its solvent properties for most polar and polymeric substances are poor. Surfactants can stabilize microdispersions of polar or polymeric substances in the cores of reverse micelles in supercritical carbon dioxide. Scattering experiments and molecular simulations can yield molecular-level understanding of these solute, surfactant, supercritical carbon dioxide systems. However, linking molecular-scale understanding to macroscopic (thermodynamic) models is very challenging because of the importance of structures and dynamics at the mesoscale (ca. 1-1000 nm). Bridging from the molecular scale to the mesoscale to the macroscale is a broadly challenging problem associated with almost all macroscopic phenomena which depend on the molecular nature of the substances involved.

Growth of Ultrathin Hyperbranched Polymer Membranes on Porous Substrates
Merlin Bruening

The objective of this work is to investigate relationships between ion transport and film chemistry in ultrathin, hyperbranched poly(acrylic acid) (PAA) membranes grafted onto highly permeable supports.  These PAA films have two properties that make them unique materials for studying transport.  First, their hyperbranched structure allows ultrathin PAA films to cover underlying pores without filling them.  Second, the high concentration of -COOH groups in PAA films permits derivatization of PAA films even after membrane formation.  Through derivatization, we will vary membrane chemistry (e.g., introduce fixed carriers or cross-linking agents) in order to control transport.  Recent FTIR external reflection spectra demonstrate both the hyperbranched growth of PAA on gold-coated porous alumina and the feasibility of derivatizing these membranes through amide formation.  Field-emission SEM and tapping-mode AFM images demonstrate that these films cover substrate pores without filling them.  Ion-transport experiments on these films are underway.  Preliminary results suggest that transport rates of sodium salts depend on the charge of the accompanying anion.

Note: The image referenced in this abstract is large (about 1MB). Link to it by Clicking here.

Novel Separation of Actinides and Fission Products Based on Molecular Imprinting and Ionic Liquids
Sheng Dai

The presentation consists of the following two parts:

  1. Molecular Imprinting on Sol-Gel Materials: Recently, the approach of imprinting organic polymers with neutral molecule and ion templates has shown promise in several areas of separation technology.  The idea behind imprinted materials is to combine the binding ability of specifically chosen functional groups or ligands for target substrates with shape- and size- selective cavities "imprinted" in a rigid polymer matrix to produce materials that will selectively bind target substrates with high affinity.  Our research at ORNL has combined the molecular imprinting with the use of sol-gel materials.  This development has resulted in a new class of molecularly imprinted microporous and ordered mesoporous materials with unprecedented selectivity factors for target templates (actinides and fission products) and fundamental insights into the ion-imprinting process.   We view these  sorbents as solid-state analogues to crown ether-type ligands that can be tailored for specific target templates.  The design principles illustrated by these results highlight opportunities for application in areas such as selective sorption,  chemical sensing, and catalysis.


  1. Fission Product Separation based on Room-Temperature Ionic Liquids: The discovery of crown ethers has led to a new class of ligands for alkaline and alkaline earth metal cations. Crown ethers and related macrocycles  have found wide application in the design of novel solvent extraction systems that are selective for fission products, such as Cs+ and Sr2+,  based on the sizes of the crown-ether rings.   The distribution ratios (DM) for the extraction process depend on two major factors: (a) the thermodynamic driving force for cation complexation by a crown ether and (b) the solvation of the cation and counter anion by the organic solvent.    The former factor is usually thermodynamically favored.  Difficulties in increasing the solvent extraction efficiency of conventional solvent extraction systems using crown ethers as extractants lie in the unfavorable transport of the cation and counter anions from aqueous phases to organic phases.    The inability of organic solvents to solubilize ionic crown-ether complexes and their counter anions is one of the main obstacles in improving the separation efficiency of fission products based on conventional solvent extractions.   This key deficiency associated with current extraction processes based on crown ethers prompted us to develop molten salt extraction media that could convert the solvation of ionic species into a more favorable thermodynamically process.   These ionic liquids are attracting increased attention world-wide because they promise significant environmental and health benefits.    Our experiments at ORNL found that ionic liquids containing crown ethers and other macrocycles exhibit unusually large distribution ratios for the selective extraction of Sr{2+ and Cs+ from aqueous solutions.

Manipulating and Interpreting Molecular Recognition for Separations of Polyaromatic Hydrocarbons Using Cyclodextrin Distribution Capillary Electrochomatography

Sepaniak, Michael J.; Fox, Shannon B.;  Culha, Mustafa

Neutral solutes can be separated based on modest binding with differentially migrating cyclodextrin (CD) running buffer additives in Cyclodextrin Distribution Capillary Electrochromatography (CDCE). In most cases, the CDs and solutes form inclusion complexes with binding constants that depend on the size, shape, and physiochemical properties of these species. Since the CDs interact with the solutes independently and with relatively rapid kinetics, capillary electrophoretic separation systems can be rationally designed involving specifically chosen combinations of CDs. The CDCE systems include one or more neutral and one or more charged CDs. Our efforts to understand the processes of molecular recognition that are operative in CDCE using molecular mechanics (MM) modeling and associated computationally assisted molecular visualization techniques will be a major aspect of this paper. Single isomer, derivatized, anionic CDs and native, neutral CDs are employed in separations of substituted polycyclic aromatic hydrocarbons, principally substituted naphthalenes. Inclusion complexation constants are experimentally determined, by monitoring the effects of CD concentration on capacity factors, and compared with interaction energies determined using various MM approaches. The influences that specific types of interactions (H-bonding, van der Waals, etc.) exert on molecular recognition are probed using naphthalenes with different types and positions of substitution. As an illustration, MOLCAD (MOLecular Computer Aided Design) Connolly surfaces with color coded H-bonding information are shown below for a single isomer charged CD and two isomers of dinitronaphthalene (DNN). The potential significance of geometric fit is indicated by the top view (on right). The complimentary H-bonding donor and acceptor surfaces for DNN and CD upon inclusion are more evident with the side view (on left). The 1,8-DNN can exhibit simultaneous H-bonding of both of its -NO2 groups with -OHs on the lip of the CD while the 2,7-DNN does not have this capability. This is consistent with MM data which indicates better electrostatic interactions for the 1,8-DNN isomer while the 2,7-DNN isomer owes it inclusion more to van der Waals interactions. It is hoped that this paper will generate discussion of certain separation challenges. Two possible general questions for discussion follow: (i) High selectivity is generally accomplished in separation science using systems with large analyte binding constants. In these cases, slow kinetics may seriously degrade efficiency (e.g., large chromatographic plate counts and peak capacities are not possible under these conditions). Is it possible to design systems with multiple weak binding interactions (e.g., with macrocycle reagents) that, in aggregate, provide custom selectivity while maintaining high efficiency? (ii) What are the current and future capabilities and limitations of molecular mechanics modeling in understanding and predicting the performance of dynamic separation systems (i.e., is the use of these computational methods just a pipe dream)?

Note: The image referenced in this abstract is large (about 0.6MB). Link to it by Clicking here.

Looking Beyond LB Films For Gas Separations
Steven L. Regen

A long-range goal that we have been trying to reach for the past decade has been to use single Langmuir-Blodgett (LB) monolayers as barriers for gas transport. Our motivation for such work stems from the inverse relationship that exists between the flux of a gas across a membrane (F) and the thickness of that membrane (l). Thus, thin membranes are more productive than thick ones, provided that defects can be minimized. Most recently, we have succeeded in fabricating novel composite membranes that use a single LB monolayer as a separating layer. These membranes have been prepared by using poly[1-(trimethylsilyl)-1-propyne] (PTMSP) as support material, {alpha}-cyclodextrin/sodium phenoxide or {beta}-cyclodextrin/sodium phenoxide as a gutter layer, and a single monolayer of a calix[6]arene as the primary barrier for transport.

Two fundamental questions that we now seek to answer, which have important practical implications, are: (i) Can analogous composites be fabricated by use of self-assembly methods? (ii) Can the permeation selectivity of such composites be improved by cross-linking? The specific approaches that we are now pursuing, in order to address these questions, will be briefly discussed.

Evidence of The Radial Heterogeneity Of Chromatographic Columns.  Origin and Possible Remediation
Georges Guiochon

Chromatographic beds are made by consolidating packing materials under stress in a cylindrical column.  Packing materials are made of porous silica particles with a narrow size distribution and a spheroidal shape.  Their consolidation process is dynamic and proceeds at a velocity which decreases constantly, without ever reaching the dense theoretical limit.  Various experiments demonstrate the radial heterogeneity of the columns packed by all known methods.

Particles cannot pack densely and homogeneously because (1) their kinetic energy is negligible compared to their potential energy, so they can hardly move relatively to each other; and (2) there is friction and surface interactions between them and with the wall.  The internal angle of friction depends on the packing material used.  It was 36o for Zorbax.  The resultant of all the friction forces between particles is not perpendicular to the column wall, hence the friction at the wall.  The friction angle at the wall is 21o for Zorbax.  From this angle and the known compressibility of the packing material, it is possible to derive the distribution of stress in the packed bed.  This distribution is uneven, some regions being under a high mechanical stress, where consolidation is stronger and particle breakage may take place.  The consequences of wall friction will be discussed.

Spatial Control of Chemical Potentials: An Avenue to Controlled 2-Dimensional Transport
Paul W Bohn

Molecular motion (transport) is an ubiquitous process of fundamental importance to molecular separations.  Work in our laboratory seeks to understand, then control, molecular motion on the nanometer length scale, through the judicious use of molecular assemblies.  Experiments focus on novel ways to fabricate the assemblies and high spatial resolution ways to study their behavior.

Adsorption at the Au-liquid interface for Au films with thicknesses between the percolation threshold (d ~ 5 nm) and the electron mean free path (d ~ 80 nm) results in significant changes in the measured resistivity of the Au film.  Surprisingly, the resistivity turns out to be an inherently high signal-to-noise ratio experiment.  We, thus conjectured that injecting mA currents, such as is done in the resistivity measurement, could be used to vary the electrochemical potential -- relative to a solution phase reference couple -- laterally across the surface of the Au electrode.  To test this idea we examined C8H17SH in a combined SPR/4-point probe resistance-electrochemistry cell specially constructed for the purpose.  Injecting sufficient current causes the potential to vary by ca. 500 mV across the electrode surface, sufficient to straddle the oxidative adsorption - reductive desorption regions for the C8 mercaptan.  In successive experiments the center potential was moved anodic, thereby increasing the surface area subject to oxidative adsorption. SPR images reveal that using this strategy a surface chemical potential gradient is formed, the position and distribution of which can be controlled by simply changing the center potential and magnitude of the injected current. 

The marriage of laterally-varying electrochemical potential and potential-dependent adsorption-desorption raises several interesting possibilities: (a)  for the first time it is possible to create externally controlled chemical potential gradients, (b) because the solution-end of the adsorbing mercaptan can be synthetically elaborated, chemical potential gradients can be converted to surface energy gradients, and (c) the characteristics of the gradient can be altered, by changing the magnitude of the current injected (size of the potential window), {Delta}V = IR, or the potential relative to the solution reference (position of the potential window).

The Behavior of Complex Systems in Supercritical Fluid Solutions
Frank V. Bright

Many reactions and separations are especially challenging when carried out using conventional liquid solvents.  For example, a reaction solvent may exhibit a high boiling point and be hard to remove from the products, the solvent may be difficult to reuse and/or recycle, it may be expensive to purchase, or impractical to dispose of without incurring substantial costs and/or adversely affecting the environment or personnel.  Moreover, slow mass transfer in liquids limits many extractions, separations and reactions.  These problems are exacerbated further because of the Montreal Protocol.  As a result, there is clear political, economic, environmental, and energy-related motivation to replace traditional liquids solvents with effective yet environmentally friendly (i.e., green) alternatives.

Supercritical fluids exhibit liquid-like densities and gas-like mass transfer, which make them appealing solvents for extractions, chemical reactions, and separations.  In addition, the physicochemical properties of a supercritical fluid are a strong function of pressure.  Thus, supercritical fluids represent completely tunable solvents because one can adjust solvent strength, density, etc. by simply controlling system pressure and temperature.  This work in our laboratories focuses on the following areas:

  1. Determining how the local microenvironment surrounding a solute is affected by the presence of an interface. This work focuses on addressing questions of solute solvation when such is part of a sample matrix and is key to issues like extraction.


  1. The molecular-level interactions that occur between polymers and supercritical CO2.  This effort centers on understanding the specific types of fluid-polymer interactions that occur in neat and co-solvent modified CO2.
  1. Quantifying the tail-tail and segmental dynamics of polymers (e.g., poly(dimethyl)siloxane) dissolved when in neat and co-solvent modified CO2.  This effort centers on understanding how polymer segment and tail mobility and tail accessibility can be modulated in fluids.  This has ramification in areas ranging from the use of polymers to control fluidity and the systheis of polymers in fluids.


  1. Determining the local environment surrounding ions dissolved in supercritical H2O.  This work aims to develop reasonably simple tools for measuring the behavior of ions dissolved in water in the critical region and for testing concepts related to solvent-separated and contact ion pairing.
  1. Determining the behavior of and exploiting reverse micelle systems formed in supercritical CO2.  These effort center on understanding the chemistry (e.g., pH, solute mobility) within the water pool, the transport of key materials into and out of the micelle core, and the ability of these micelles to serve as nanoreactors for a host of biologically-based reactions.


Development of In-Situ Raman Spectroscopy to Monitor Metal-Ion Complexation at Liquid/Solid Interfaces
Rory Uibel*and Joel Harris

A fiber-optic system was developed for exciting and collecting Raman scattering signals of metal-ion complexation with silica-immobilized 8-hydroxyquinoline under flow-rate and pressure conditions typical of chromatographic separations.  The extraction of metal ions from aqueous sources onto immobilized ligands is critical to strategic metal recovery and environmental clean-up initiatives.  In-situ vibrational spectroscopic observations of metal complexation by silica-immobilized ligands can assist in the development of these applications. Raman spectra were collected with a bifurcated fiber-optic bundle inserted into the side of a column which had been packed with silica immobilized 8-hydroxyquinoline.  Reactions of this reagent with different metal ions were studied by using flow injection methods to control the solution conditions and the exposure of the reagent to the metal ion samples, providing insight into the chemical interactions that affect the equilibrium behavior.  The results of this study included:  (1) the observation of a 1:1 metal-to-ligand ratio; (2) a comparison of surface equilibrium constants to free solution values; (3) differences between the free solution and immobilized reagent can be related to changes in surface potential of the silica gel.  Future plans for this system include in-situ monitoring of solid-phase extraction of metal ions by a variety of different immobilized ligands.

Ultra-Thin Film Composite Membranes by Surface Modification of Mesoporous {gamma}-Alumina Substrates
J.D. Way and K.C. McCarley

Composite organic/inorganic gas separation membranes have been fabricated by surface modification of commercially available mesoporous {gamma}-alumina substrate using silane coupling agents such as octadecyl-trichlorosilane and heptadecafluoro-1,1,2,2-tetrahydrodecyl-1-trichlorosilane.  Based on ellipsometry measurements, our hypothesis is that the composite membrane contains a very thin, {approx}10 nm selective polymer layer deposited on the surface of the ceramic ultrafilter.

The permeance of condensable gases is strongly enhanced by surface diffusion of adsorbed species.  After modification, the alumina membranes exhibit reverse selectivity where a larger, heavier molecule such as butane permeates faster than a smaller, lighter penetrants such as H2, CO2, CH4or N2.  We have measured pure gas selectivities as high as 48 for n-C2H2/N2 and 20 for n-C4H10/CH2.  Mixed gas selectivities observed for n-C4H10/i-C4H10 as high as 7.8 suggest that the selective polymer layer also exhibits shape and size selectivity.  Less-interacting gases exhibit Knudsen diffusion acting in parallel with surface flow.  Mixed gas selectivities of 5.7 observed for CO2/N2 for the alkyl silane modified membrane are evidence of the difference in the sorption of the two penetrants with the selective layer.  Gas permeation measurements with the fluorosilane modified membrane are currently in progress.  We expect that the order of permeation for the fluorosilane membrane may be significantly different from the results obtained with the octadecylsilane membrane due to the high solubility of CO{sub 2} in fluorinated polymers.

Enhancement of Transport Processes in Multiphase Systems
D. W. DePaoli, V. F. de Almeida, and C. Tsouris

This presentation will begin by highlighting the results of studies focused on the application of electromagnetic fields to dramatically improve transport rates in multiphase processes such as solvent extraction and distillation, as well in other areas, including biotechnology. Secondly, this presentation will discuss scientific challenges involved in understanding multiphase processes, and modern ways of facing these challenges. Finally, we will project some important implications of the research for industrially relevant processes.

Electric fields can be effectively applied to generate fine droplets/bubbles of electrically nonconductive fluids into more conductive liquids (Fig. 1), extending the application set of electrically driven processes. Electrically driven microbubble formation in water and organic liquids can be characterized as having two clearly defined regimes. First there is a lower-voltage regime where interfacial electric stresses dominate, decreasing bubble size and increasing pressure. Secondly, a higher-voltage regime exists, which is dominated by electohydrodynamic (EHD) flows, causing pumping, spraying, and mixing. Recent experiments have shown that bubble formation dynamics in these systems exhibit the classic signs of a period-doubling bifurcation to chaos with increasing applied potential (Fig 2). This discovery has several important practical implications that will be discussed. It suggests that the developing tools of nonlinear science provide keys to understanding complex phenomena in multiphase systems. EHD flows can be employed for rapid mixing at the microscopic scale (Fig. 3). This capability has been demonstrated by the production of homogeneous submicron particles (Fig. 4) by sol-gel reactions under fast reaction conditions. Conventional sol-gel processing yields undesirable products.

Figure 1. Inverse electrostatic spraying of air in deionized water.

Figure 2. Pressure traces for formation of air bubbles in glycerol from an electrified capillary, showing period-doubling with increasing potential. Applied potentials (from left): 0, 5, 7.5, and 10 kV.

Figure 3. EHD mixing of butanol in deionized water. Conditions: butanol - 0.8 mL/min, water - 50 mL/min, tube i.d. = 7.5 mm.  

Figure 4. Hydrous zirconia particles produced by EHD mixing. Size scale is 1 micrometer.  

Most multiphase processes, with or without applied fields, employ flows with exceedingly complex dynamics. Consequently, even though many industrially important separation processes are mature in business terms, their design remains largely empirical or based on groundless assumptions. Difficulties in substantially advancing the basic understanding of multiphase flows and associated transport phenomena stems from lack of simple models able to capture complex and collective phenomena ranging from the microscale to the continuum, for instance, electro-spraying of droplets and bubbles.

Recent developments in computing capabilities of teraflops distributed parallel computers, modern algorithms, and emerging simple models of complex physics are promising means of advancing the understanding and application of multiphase flows. Notably, the recent development of flow modeling in porous media by the lattice-Boltzmann method stands out as a candidate for studying cross-cutting applications in separations. For instance, microscale hydrodynamics of packed beds can benefit from these novel tools to bring distillation to a scientifically mature technology with improved economics.

Significant challenges are anticipated in modeling multiphase separation systems subjected to electric and magnetic fields due to inherent nonlinear microscale complexity. These systems will require state-of-the-art computing power and further advances in parallel algorithms to resolve representative microscale phenomena. In addition, efforts to integrate macroscale modeling and microscale simulations need to be developed in order to simulate realistic processes. For instance, suitable homogenization techniques need to be elaborated further in order to combine modeling at multiple scales. Coupled experimental and computational studies are key to the development and verification of these new approaches. Notwithstanding the challenges, it is anticipated that a broad class of separations (e.g., distillation, extraction, absorption, gas scrubbing, biofiltration) and other important systems (e.g., rheology of suspensions ranging from blood to radioactive slurries) will be positively impacted by evolving multi-scale modeling of multiphase flows.

Separating Products from Xerogel Encapsulated Enzymes with CO2
Ted Eyring

The feasibility of catalyzing reactions using xerogel encapsulated enzymes is well known.  The enzyme is introduced into a sol-gel at room temperature, and the mix cures to a damp glass (xerogel) in a couple of weeks without experiencing elevated temperatures.  The catalytic activity of the enzyme is thus preserved, and because of its size, the enzyme cannot escape from the porous glass matrix whereas smaller substrate and product molecules readily enter and exit the xerogel.  Supercritical carbon dioxide(sc CO2) is a suitable solvent for the introduction of cholesterol and the removal of product cholesterone from a xerogel in which cholesterol oxidase has been encapsulated.  The sc CO2 is a particularly attractive solvent in this instance because by releasing pressure, one achieves a clean separation of solvent and cholesterone.  There are, however, many interesting questions that remain to be answered:  Does the sc CO2 dry out the xerogel rapidly (to a water-free aerogel) thus precluding a high number of reaction turnovers?  How much is the enzyme-catalyzed reaction slowed down in sc CO2 compared to other supercritical solvents (that are less attractive from the environmental contamination point of view)?  Can unclouded xerogel monoliths be routinely produced that lend themselves to quantitative spectroscopic examination in sc CO2 How rapidly does the porosity of a xerogel monolith or thin film change with time thus possibly precluding repeated use?

Development of Ion-Selective Polymer-Supported Reagents
Spiro D. Alexandratos

The immobilization of ligands on to crosslinked polymers is an important route for preparing reagents that can be used for the ion-selective removal of radioactive or otherwise toxic metal ions from aqueous solutions, whether as industrial process streams, stored wastes, or water in the environment.  Polystyrene remains an important support on to which ligands can be covalently bound due to its ability to undergo both electrophilic and nucleophilic substitution reactions.  A wide range of ligands with varying levels of selectivity have been studied.  Such ligands include polyethyleneimine, which is selective for Cu(II) over Ni(II), (2-aminoethyl)piperazine, which has a high sorption capacity for Au(III), thiacrown ethers, which are selective for Ag(I), and phenols, which have a high Cs(I) affinity.  Our BES-funded research has led to the development of a number of ion-selective reagents with phosphorus-based ligands.

This presentation will give an overview of the state-of-the-art in polymer-supported reagents as well as summarize two of our latest projects.  We have developed a technique to produce high-stability solvent-impregnated resins (SIRs).  This technique is particularly important because there are many solvent extractants known and their ionic selectivities well understood.  Attempts have been made to sorb these extractants into polystyrene beads and then use them to

complex metal ions.  The finite solubility of the extractant, however, eventually causes loss of the extractant into the water.  Our research has led to a method whereby each polymer bead can be encased in a semi-permeable membrane that allows the polystyrene to retain the extractant.  Metal ions can travel through the membrane to be complexed by the extractant within the bead and the resulting complex remains held within the bead.  The high-stability SIRs have been shown to be regenerable.  The second area of our research that we will summarize will be immobilized calixarenes.  Calixarenes are important ion-selective molecules whose mmobilization now allows for a wide range of ion-complexing applications.

Adsorption of Halocarbons in Nanoporous Materials: Current Status and Future Challenges
Mellot, C. F. ; Eckert, J.;  and Cheetham, A. K.

A variety of environmental issues are motivating our work on the behavior of chloro- and fluoro-carbons in a variety of zeolites. Our aim is to understand the energetics and dynamics of halocarbon adsorption under equilibrium conditions by probing the influence of parameters such as the Si/Al ratio, the cation content and the sorbate loading.

In the first stage we have developed a new forcefield for chlorocarbon-type molecules in zeolites, where short-range non-bonding and long-range electrostatic interactions, as well as guest-guest interactions, are considered. The structure and the energetics of chloroform binding sites in NaY zeolite have been studied by energy-minimization calculations and were compared with inelastic neutron scattering and Raman measurements on the same system [1]. Three components of the total host-guest interactions are revealed: (i) short range interactions between chlorine and framework oxygens (ii) electrostatic interactions between chlorine atoms and accessible Na ions and (iii) hydrogen bonding with the framework oxygens.

Monte Carlo simulations on the adsorption of the two model chlorocarbons, chloroform and trichlorethylene, in the three faujasite-type zeolites - NaX, NaY and siliceous faujasite - have been performed (Si/Al = 1.2, 3.0 and infinity) and compared with calorimetric determinations on the same systems [2,3]. Excellent agreement provided validation of our forcefield for both saturated and unsaturated chlorocarbons, successfully capturing both the host composition dependence and the coverage dependence of chloroform and TCE adsorption heats in all zeolites. In all cases, the heats of adsorption are enhanced according to the basicity and the cation content of the host (siliceous Y < NaY < NaX), underlining the importance of the dipolar nature of the sorbate

Our current and projected work involves (i) synchrotron X-ray and neutron diffraction studies of the structures of some model systems, (ii) the use of maximum entropy methods to analyse such data, (iii) molecular dynamics simulations on these systems, (iv) NMR studies of the molecular motion, (v) the extension of our work to other classes of zeolites (e.g. silicalite). All of these studies present interesting challenges which will discussed in detail at the workshop.

[1] C.F. Mellot, A.M. Davidson, J. Eckert, A. K. Cheetham, J. Phys. Chem, 1998, 102, 2530

[2] C.F. Mellot, A. K. Cheetham, S. Harms, S. Savitz, R. J. Gorte, A. L. Myers, J. Amer. Chem. Soc. 1998, 120, 5788. ,

[3] C.F. Mellot, A. K. Cheetham, S. Harms, S. Savitz, R. J. Gorte, A. L. Myers, Langmuir 1998, 14, 6728.

[4] C.F. Mellot, A. K. Cheetham, J. Phys. Chem. in press.

Chromatographic Stationary Phase Structure-Function Relationships: Understanding Conventional Stationary Phases and Exploring Unconventional "Designer" Stationary Phases
Pemberton, Jeanne E.

Studies utilizing Raman spectroscopy are underway in this laboratory to understand the degree of conformational order/disorder in conventional alkylsilane-based stationary phases for reverse-phase liquid chromatography (RPLC). Raman spectroscopy is a powerful tool for elucidating the conformational structure of such systems given its relative insensitivity to spectral interference by silica or water and the accessibility to vibrational information from ca. 100 to 4000 cm-1. Several "conformational indicators" have been identified in the Raman spectra of alkane-based systems that are used to quantitatively assess the level of conformational order in such systems and the extent to which this order changes upon perturbation by environmental effects such as temperature, solvent, presence of solute and exposure to pressure. The results from such studies provide direct, definitive evidence for conformation in these systems that provides insight into the chemical forces responsible for retention and separation of mixtures of solutes. Certain results obtained to date have been somewhat surprising in light of the "conventional wisdom" of RPLC retention. For example, as shown in Figure 1, solvent studies show a surprising lack of difference in  conformational order of octadecylsilane stationary phases in water, methanol or acetonitrile despite the relatively profound influence that such solvents may have on separation efficiency for a given class of solutes. The current state of temperature and solvent studies will be presented as well as the initial results from the presence of solutes on conformational order.

Results from the above studies have led us to consider ways to control and optimize the degree of conformational order/disorder in chromatographic stationary phase materials. Toward this end, we have synthesized a series of sol-gels modified with alkylsilanes that possess varying degrees of conformational order in the alkyl portion. One Raman spectral conformational indicator for a series of such systems is shown in Figure 2. The conformational order in these systems ranges from almost completely liquid-like to totally crystalline based on the numerical values of this Raman conformational indicator. These new materials have been coated as thin films onto silica particulate supports to create a series of "designer" stationary phases.  Hopefully, preliminary results of the use of these systems in RPLC experiments will be presented. Ultimately, we believe that such chemically engineered stationary phase systems should provide a much more definitive assessment of the conformational characteristics desirable for RPLC on different classes of molecules.

Fundamentals of electric-field enhanced separations
Osman Basaran

A central problem in electroseparations is the dispersion of one phase, in the form of drops, in a second phase.  Drop formation is a prototypical free boundary problem and as such lies at the frontier of separation science, fluid physics and interfacial phenomena.  In the past couple of years, significant progress has been made in this program to elucidate certain aspects of drop formation in both the absence and presence of electric fields using a two-pronged attack based on computation and experiment.

Novel algorithms based on the finite element method (FEM) have been developed to model the dynamics of  drop formation, including breakup, in the absence of electric field.  First, the theoretical predictions have been shown to agree with experimental measurements with better than 1% accuracy at any value of the operating parameters (Wilkes et al. 1999). By contrast, simulations based on competing techniques such as the volume of fluid (VOF) method either yield about 10% accuracy for low viscosity systems (Zhang 1999) or fail outright for high viscosity systems (Delametter 1998).  With the new FEM algorithm, we have been able to predict new phenomena such as interface overturning and occurrence of non-periodic flow states which had heretofore proved elusive.

The FEM algorithm described above has also been extended to predict drop formation in the presence of electric field when the drop liquid is either a perfect conductor or a perfect insulator.  Once again, we have been able to predict phenomena such as micro-dripping (Cloupeau and Prunet-Foch 1990, 1994) and switch of breakup sequences (Zhang and Basaran 1996) which had heretofore remained without explanation.

Ultra-fast visualization techniques have also been introduced to study drop breakup phenomena occurring in time scales as short as tens of nanoseconds.  Such a visualization capability allows probing of dynamics of small drops of interest in MEMS-type applications and getting insights into ever-present satellites and sub-satellites.

Studies have also been started to investigate mass transfer effects, for example as in the transport of surface-active species.  New discoveries vis-a-vis the role played by Marangoni stresses have been reported in a publication which is now in press (Ambravaneswaran and Basaran 1999).

 A number of challenges and opportunities await us in the next 2-5 years.  These include theoretical and experimental characterization of various electrohydrodynamic (EHD) jetting phenomena, the effects of finite conductivity, the dispersion of drops of complex liquids (containing surfactant and polymeric additives), and the dispersion of drops when forced by complex flows and electric fields.

The talk will highlight a number of past accomplishments and future challenges and conclude with a series of ultra high-speed experimental visualizations and computer animations of computed predictions.

It is both exciting and rewarding that the studies being conducted in this program are applicable to diverse areas outside of separations.  Some of these include biochip processors, spray coating, and crop spraying, among others. Evidence of such applicability is provided by the diverse interactions of the PIs group with research laboratories of several large and small companies on research activities that owe their existence to this BES program on electro- separations.

Carbon Dioxide Based Solvents for Waste Minimization
Keith P. Johnston

Environmentally benign carbon dioxide-based solvent formulations may be utilized to replace toxic organic solvents for chemical processing. CO2-based solvents offer exciting new opportunities in chemical manufacturing involving heterogeneous reactions (including polymerization), solvent free coatings, extraction of heavy metals including radioactive compounds from soils and wastewater, polymer processing, and separations processes including cleaning and purification. A fundamental knowledge of colloid and interface science for CO2 based systems is being developed to design surfactants for microemulsions, emulsions, and latexes. The effects of various surfactants on the interfacial tension between water or organics and carbon dioxide are reported along with measurements of colloid formation and stability. Fundamental thermodynamic properties including the solvation of the surfactant tail by carbon dioxide, the adsorption of the surfactant at the interface, the balance between the interactions of the surfactant with each phase, and the interactions between droplets are examined and used to interpret the phase behavior, droplet size and stability in microemulsions, emulsions and latexes. We have formed novel water-in-CO2 and CO2-in-water emulsions that are stable for hours. The mechanisms of stabilization and destabilization of these colloids are being characterized with lattice fluid self-consistent field theory and computer simulation to interpret the experimental results and to plan new experiments.

Carbon Dioxide Based Solvents for Waste Minimization
Gary E. Maciel (Nancy E. Levinger and Sally Sutton, Co-PIs)

The overall goal of this project is the development of detailed information on the chemical/physical states (reactions, interactions, mobilities) of specific ketones, aromatics and chlorohydrocarbons with major soil constituents (clays, humic substances, silica) and, to some degree, with whole soils. The main theme areas of the project are: 1) behaviors of organic pollutants adsorbed on/in soil components; 2) photo-assisted decomposition of adsorbed chlorohydrocarbons; 3) interactions of Cu2+ in humic materials; and 4) the clay-humic complex.

Regarding theme area 1), our experience to date is that, in the absence of microbiological activity or photochemical assistance, the chemical fates of typical organic pollutants in soil (components) seem to be a very long life. The following key issues/questions arise as the work progresses: How useful are shorter term, higher-temperature studies in simulating longer-term processes at ambient temperatures? How efficient and how specific are pollutant decompositions in soils by microbiological mechanisms? Are such mechanisms so important (efficient) that abiotic processes are not worth studying?

The main purpose of theme 2) is to simulate and characterize the photo-assisted decomposition of organic pollutants at the soil/air interface under the influence of sunlight. Preliminary results indicate the "complete" decomposition of trichloroethylene adsorbed on a variety of substrates (montmorillonites, silica, humic acid) over a period of days under near-UV light. In shorter periods of time, toxic intermediates are formed. To what degree is this behavior, at least for some of the substrates, due to trace amounts of the known, highly-efficient photo-activator, TiO2? Probable answer: not significantly (to be tested by the use of highly purified substrates). To what degree will the susceptibility of Cl2C=CHCl to photo-assisted decomposition under these conditions carry over to other chlorohydrocarbon pollutants, or indeed to organic pollutants in general? To what extent can sunlight-assisted decomposition of organic pollutants impact on the total soil/pollutant/groundwater scenario? This appears to be a highly complex problem, with different reaction rates and intermediates characteristic of each specific pollutant (and perhaps each substrate).

The Cu2+ ion is well known to function as a catalytic center in a variety of processes in organic chemistry and is a major contaminant in soils and sediments associated with certain marine environments. In what specific ways is Cu2+ "complexed" to humic materials? Does this Cu2+ complexation bring about an "autocatalysis" in chemical transformations (decompositions) of humic materials? Or, does Cu2+-humic complexation simply sequester Cu2+, which can then be available to impact micro-organisms that may be associated with humics?

In searching for the precise nature of the often postulated "clay-humic complex," one is faced with trying to observe such linkages, which are presumably present in relatively small concentrations, in the presence of overwhelming concentrations of intra-clay and intra-humic covalent linkages. Hence, a highly relevant question/problem is: How can one reduce (ideally eliminate) the presence of humic and clay components that are merely physically associated with each other in a soil from those humic and clay components that are bonded by specific clay-humic linkages in the soil, without disrupting those clay-humic linkages of primary interest? This is the key issue which is at the heart of our NMR-based strategies for elucidating any clay-humic linkages that may exist.

Microscopic Insights into the Architecture of Organic Monolayer Films
M. Porter

This presentation explores strategies for the characterization of the structure and reactivity of organized monolayer films at liquid-solid interfaces. Such interfaces play a critical role in several technologically significant areas, including catalysis, lubrication, friction and wear, and chemical analysis. Advances in these areas are, however, generally hindered by limitations in fundamental insights into factors that influence the microscopic origins of interfacial reactivity and by the lack of methods that probe such interfaces at a molecular level. This two-part presentation describes some of our most recent results from investigations of model interfaces (e.g., gold-bound alkanethiolate monolayers) that address the challenges in realizing such advancements. The first part of the presentation will describe our recent findings in the use of scanning probe microscopy (SPM) as a tool to map the chemical and physical properties of surfaces by friction- and adhesion-based techniques. Results that demonstrate the ability to manipulate as well as to monitor in situ the rate of the base hydrolysis of an ester monolayer will be discussed. The ability to detect subtle differences in the observed friction of such materials that arise from differences in the spatial orientation of end groups is also demonstrated and possible origins of these observations examined. Extensions of SPM to gain additional insights into interfacial architectures will also be described. The second part of the presentation will discuss our continuing investigations on the use of electrochemical techniques for probing the nature of the sulfur-gold interactions of these systems. We will describe the use of conventional as well as thin-layer spectroelectrochemical techniques to examine possible products from the chemisorption process (e.g., the fate of the sulfhydryl hydrogen). Comparisons of thiol-, disulfide-, and sulfide-derived systems will also be presented.

Highly Fluorinated Polyphenylene Derivatives as Gas Separations Membranes
Valerie Sheares

A new high performance material, poly[[1,1'-biphenyl]4,4'-diyl[2,2,2,-trifluoromethyl)ethylidene]](PDTFE), was prepared by Ni(0)-catalyzed coupling polymerization of 2,2-bis(p-chlorophenyl)hexafluoropropane. The resulting high molecular weight polymer is amorphous and soluble in common organic solvents including tetrahydrofuran, chloroform, and acetone. The solubility leads to ease of preparation, characterization, and processing of PDTFE. Excellent thermal properties were exhibited with a glass transition temperature of 266 °C and 10% weight loss values of 549 °C and 557 °C in nitrogen and air, respectively. The flame retardance properties showed that the material meets the Federal Aviation Administration's current criteria for heat release capacity, total heat release, and char yield for flame retardant materials. Additionally, colorless, transparent, creasable films were cast from chloroform. The film formation made the gas permeability measurements possible. PDTFE has a rather high oxygen permeability coefficient of 120 x 10-10cm3 (STP)cm/cm2s x cm Hg). Additionally, the O2/N2selectivity of 2.9 and the carbon dioxide/methane selectivity of 13.8 at 35 °C are 73% of their calculated upper bound values according to the Robeson equation. The combination of facile synthesis and an excellent property profile make this a unique phenylene-based high performance polymer that will lead to a series of novel fluorinated high temperature gas separations membranes.

Supramolecular Recognition in Separation Technology
Andrew Goshe and Brice Bosnich

There are numerous current methods for the separation of cations and anions, the least developed of which is the use of supramolecular assemblies constructed with transition metal ions. Such receptors should be relatively simple to construct and should be capable of incarcerating anions and cations and, in some cases, the complete ion pair provided the receptor has sites capable of recognizing both the anion and cation. Further, these receptors should also be capable of hosting neutral and charged organic molecules. Current research is directed at developing a number of receptor types which will incorporate square planar metal complexes. The basic elements of these receptors consist of an approximately eight angstrom spacer, chelators and linkers. The spacers and chelators are joined to give spacer-chelators.

With these structural elements receptors resembling squares, trigonal prisms and cubes could be constructed using square planar metals. Our current work has concentrated on using gold (III) complexes. A commercially available spacer was used to incorporate a tridentate chelator precursor which was prepared in high yield. The spacer-chelator forms gold (III) complexes from which molecular squares are prepared using a 4,4’-dipyridyl linker.

The molecular square, as yet, has not been fully characterized but it appears that it will incarcerate chloride ions inside the square by weak interactions with cofacial gold atoms.