Adsorption Science and Technology

The following sections are included:

• Preface

• Contents

• CPEC Sponsors

Adsorption isotherms of water and ethanol on fluorinated, nitrogen-alloyed, and pristine pitch-based activated carbon fiber(ACF)s were measured in order to elucidate water adsorption mechanism in hydrophobic micropores. States of alloyed fluorine and nitrogen atoms were examined by XPS. Fluorine atoms are bound by covalent bonding, while the presence of the pyridinic nitrogen was observed. Fluorination of ACF lowered the adsorptivity for water and ethanol. On the contrary, the nitrogen alloying enhanced the adsorptivity for ethanol. The adsorption states of water and ethanol in micropores of ACF were examined by use of in situ X-ray diffraction. The electron radial distribution function (ERDF)s from X-ray diffraction patterns showed that water and ethanol molecules form highly ordered structures in micropores at the saturated filling. The in situ small angle X-ray scattering (SAXS) of water-adsorbed ACF indicated the presence of different physical processes upon adsorption and desorption, which can be associated with the adsorption hysteresis. These integrated information on water and ethanol adsorption on F or N-alloyed and pristine ACFs were summarized to understand the cluster-mediated filling of water in hydrophobic micropores.

Magnetic resonance imaging (MRI) techniques have been used to follow transient chromatographic profiles in a packed adsorbent column and intraparticle profiles in a single extrudate of molecular sieve adsorbent Preliminary experimental results are presented for several representative systems and the scope and limitations of the technique are discussed.

The following sections are included:

• Introduction

• Computer Simulation Algorithms

• Simulation Results

• Conclusions

• References

The following sections are included:

• Introduction

• π-Complexation

• Ag-Containing LSX Zeolite for Air Separation

• Olefin/Paraffin Separations By π-Complexation

• Concluding Remarks

• Acknowledgment

• References

Magnetic Resonance Imaging (MRI) was used to measure the concentration profiles of nitrobenzene inside an activated carbon monolith during an adsorption process. In a first experiment the starting point was a completely dry sample of activated carbon. Nitrobenzene vapour was led over the sample and the concentration profiles were measured. In the second experiment the starting point was a sample completely saturated with D₂O, causing the macropore diffusion of nitrobenzene to be much slower than in the first experiment. Clear nitrobenzene concentration profiles were obtained. With MRI information about the adsorption process is gained which cannot be obtained from conventional adsorption measurements that don't give a look inside the porous material.

The pressures at which capillary condensation takes place and at which the condensate vaporizes depend on the adsorption properties at the pore walls surface as well as on the width of the pore. An equation that expresses quantitatively these relations is used for the calculation of the capillary condensation of nitrogen at 77.3 K in porous adsorbents assuming that adsorption at the pore walls obeys the BET equation. The condensation pressure is plotted against the pore width for various values of the constant c of the BET equation for the two branches of the hysteresis loop. The condensation pressures obtained are higher than those predicted by the Kelvin equation with zero contact angle. The width of the hysteresis loop, and the deviation from the value given by the Kelvin equation with zero contact angle, increase with decrease of the pore width and decrease of the parameter c of the BET equation.

Isosteric heats of adsorption are calculated for activated carbons prepared by chemical activation of macadamia nutshell. A commercial activated carbon is also used for comparison. The isosteric heat is calculated directly from the experimental isotherm data of hydrocarbons and CO₂ and indirectly from the three equilibrium models of Sips, Toth, and Unilan fitted to experimental data. The heat values resulted from a model of Do and Do [4] (IHFL model), which takes into account the isosteric heat as a main source of heterogeneity of the solid, is also considered in the comparison. This investigation shows that although all equilibrium models fit the data well within the experimental error, the isosteric heat of these models do not agree and hence it can be used as additional criterion to choose the proper equilibrium model.

The adsorptive properties of a selection of carbons produced by the chemical treatment of lignite and peat have been investigated. Adsorption of phenol, basic red dye, copper ions and NO₂ by the carbons is assessed. Adsorption equilibrium isotherms were plotted to obtain Langmuir constants for each system. Adsorption capacity of each carbon depends on the method of activation. Gas sorption and mercury porosimetry were the methods used to study the micro, meso and macropore volumes, surface areas and pore radius distributions of the carbon samples.

Recent developments have given positive indications that biomaterials, in their native or modified states, may be used as alternatives to conventional adsorbent media. This work investigates egg shell membrane of a hen's egg as an adsorbent. This work specifically concentrates on the capacity of egg shell membrane to remove a reactive dye. The dye used was the commonly used reactive dye Levafix® Brilliant Red E-4BA. Using a batch system, the effect of initial concentration was investigated. The Langmuir, Freundlich and Redlich-Peterson equations were applied to the experimental data. The results showed that the biosorptive behaviour of egg shell membrane can be described well by the Langmuir isotherm. The Langmuir constants were determined to be Q_max = 101 mg/g and K_L = 330 l/mg. In addition, the influence of the agitation speed on this adsorption process was examined.

Experimental data on photoadsorption effects of O₂, H₂, and CH₄ under the irradiation with light (200-800 nm) on MCM-41 modified with V-oxide nanoparticles are presented. The particles were produced by chemical interaction of VOCl₃ vapors with dehydrated and partially dehydroxylated MCM-41 silica and subsequent hydrolysis with water vapors. The synthesis procedure is described in detail. In-situ IR spectroscopy was used to clarify some features of the synthesis process.

Unimpregnated activated carbons were studied as adsorbents of hydrogen sulfide. The samples were characterized using Boehm titration, thermal analysis, and sorption of nitrogen. It was found that the choice of unimpregnated carbon for application as H₂S adsorbents should be made based on parameters of its acidity such as number of acidic groups, pH of surface, or weight loss associated with the presence of acids. There are certain threshold values of these quantities which, when exceeded have a dramatic effect on the H₂S breakthrough capacity.

Sorption of water and methanol on two carbons of wood origin was studied. The isotherms were measured at low relative pressure (p/p₀ < 0.3) at temperatures close to ambient. Before the experiments, the detailed structural and chemical features of carbon surfaces were determined applying various physical and chemical methods. From the isotherms, the heats of adsorption were calculated using virial equation. The results showed differences in the uptake of water and methanol molecules at low relative pressure due to the different mechanisms of the adsorption processes. It was demonstrated that the effect of pore sizes and pore volume on sorption uptake is more pronounced in the case of methanol, whereas water sorption is governed mainly by surface chemistry.

Generating the mobile phase flow using shear-forces originating from a moving wall element allows to surpass the limitations on the mobile phase velocity, the channel diameter and the channel length which exist in pressure-driven chromatography, while all the beneficial characteristics can be retained. The proposed concept hence allows for unprecedented separation speeds and resolutions, and intrinsically leads to a miniaturisation of the equipment. To demonstrate the practical feasibility of the concept, the velocity distribution and the axial dispersion in a shear-driven channel with a flat rectangular cross-section have been investigated. We were able to move a tracer plug at speeds exceeding 2 cm/s through a 0.125 micron thick channel without the help from a pressure or a voltage gradient and without noting any significant peak broadening.

Experiments have been conducted to study the kinetics of Sumithion adsorption on rice field soil. The experiments were performed in a batch system using an erlenmeyer flask, magnetic stirrer and thermostat. The initial concentration of Sumithion used was 1.0141 ppm and the changes of its concentrations were analyzed using a FPD gas chromatography. Ratios of soil weight to water volume used (R_SW) were between 0.02 g soil/ml and 0.1 g soil/ml. Soil was taken from Donoharjo, Bantul, Yogyakarta at four different depths. Before the experiments, soil was dried, sterilized and analyzed its physical properties. Proposed mathematical model to describe adsorption process of Sumithion on rice field soil was in a good agreement with the experimental data. The correlation of equilibrium constant (K) with respect to temperature (T) for soil organic matters of 2.88% and for R_SW=002g soil/ml, is

Adsorption of N₂, O₂, Ar and CH₄ in zeolites Type Y and Mordenite, both in powder and pellet forms has shown occurrence of solid state ion exchange during pelletization using clay binders. Temperature programmed desorption of ammonia has shown depleted surface acidity as a consequence of solid state ion exchange phenomenon leading to lower catalytic activity in zeolite pellets.

A modification of the adsorption integral based on the Dubinin-Radushkevich model is proposed here to account for finite adsorbate molecular size and steric hindrance of the pore walls. The model is successfully applied to both gas and liquid phase adsorption, yielding more consistent and improved data interpretation in both cases. In particular for gas phase adsorption the model yields considerably improved predictions at high pressures, based on our application to the adsorption of various gases on Nuxit activated carbon. For the liquid phase our experimental data on the adsorption of flavour esters on Filtrasorb 400, Norit ROW 0.8 and ROX 0.8 is also found to be better represent by the proposed model, yielding more consistent and realistic parameters values. The ester molecular diameters, being largely unknown were also extracted from fits of the adsorption data, and were found to be consistent with estimates based on critical volumes.

A new isotherm is proposed here for adsorption of condensable vapors and gases on nonporous materials having a type II isotherm. The isotherm combines our recent molecular-continuum model in the multilayer region, with other widely used models for sub-monolayer coverage. The model is successfully tested using isotherm data for nitrogen adsorption on nonporous silica, carbon and alumina, as well as benzene and hexane adsorption on nonporous carbon. Based on the data fits, out of several different alternative choices of model for the monolayer region, the Freundlich and the Unilan models are found to be the most successful when combined with the multilayer model to predict the whole isotherm. The hybrid model is consequently applicable over a wide pressure range.

A new model for determining the pore size distribution of micro and mesoporous materials from gas adsorption isotherms has been proposed. The model uses the Dubinin-Rudushkevich (D-R) isotherm with Chen and Yang's correction at low pressures. For the mesopore region, a recent model of the authors which uses molecular-continuum model for multilayer region and the Unilan model for sub-monolayer region has been extended. The experimental adsorption data was inverted using regularization technique to obtain the pore size distribution. The family of model mesoporous adsorbent, MCM-41, was chosen for testing the present model. The present model was found to be successful in predicting the pore size distribution of pure as well as binary physical mixtures of MCM-41, with results in agreement with those from the XRD method.

The porous structure of active carbon modified by successive removal of external layers from a particle surface as produced by abrasion in a spouted bed and next by demineralization was investigated. The methods proposed improve both adsorptional and densimetric characteristics of modified samples and the methane storage capacity increases up to several percents over commercial carbon.

Commercial granulated activated carbon Norit R3ex after demineralisation with cone. HF and HCl acids was oxidised using cone, nitric acid at 353 K for 3 hrs and heated in vacuum at 423 K. From the oxidised carbon the appropriate powdered and core samples were obtained by the method in which the successive layers are removed from the granules by abrasion in a spouted bed. The abrasion treatment was carried out in two steps: elimination of about 33% and 67%wt. of the starting granules. The powdered samples were taken from the external and middle layer of the granules leaving behind granular core samples. In all carbon samples total oxygen contents, as well as surface functional oxygen groups were determined. Additionally the surface chemical structure of the samples was determined by FTIR spectroscopy. The excess adsorption isotherms of the selected cations (Fe³⁺, Co²⁺, Cu²⁺) on the obtained samples were determined at 298 K. Adsorption took place from aqueous solution with different pHs of corresponding nitrates. The relationship between the adsorption capacity and chemical structure of adsorbent surface was analysed and discussed

Isosteric heats of sorption measured directly by sorption isosteric and calorimetric methods are compared for nitrogen and oxygen on CaA and carbon dioxide on NaX zeolite. The concentration dependences of isosteric sorption heats obtained by these methods show similar shapes. Depending on concentration ranges, the heats from the isosteric method are somewhat smaller than those from calorimetry. The discrepancy amounts to 2 kJ/mol, at the most. Since the same materials were used for all experiments, the source of this disagreement remains unknown.

Extensive studies of structure and energetic properties of the surface of the carbosils including from 0.8 to 35% carbon were undertaken. The structural characteristics of the adsorbents under investigation obtained by nitrogen adsorption and desorption at 77 K, p-nitrophenol adsorption from aqueous solutions and flow microcalorimetry were presented. The results of the studies obtained by the electron paramagnetic resonance EPR and transmission electron microscopy TEM methods were also given. The obtained data allowed to determine morphology and topography of carbon deposits in the studied adsorbents as well as the effect of carbon amount on the modified silica surface on the energetic properties of the obtained complex adsorbents. The general studies, particularly, those on a large scale are of significant importance for further effective usage of original properties of carbon-mineral adsorbents. Knowledge of the complex adsorbent structure including morphology and topography of carbon deposits, surface properties is still poor compared with their present and perspective applications.

A theoretical method is used to describe adsorption on the heterogeneous solid surfaces. The four layer model assuming that the anions and cations of the inert electrolyte (1:1) are adsorbed in two distinct layers, located at different distances from the surface is proposed. The theoretical development has been based on the 2-pK charging mechanism. This model includes a complete thermodynamic description (including that of the enthalpic effects) based on such a modified four layer model along with the comparison to the four layer model assuming that the oxide surface is geometrically and energetically homogeneous, as well to the triple layer model assuming that the oxide surface is an energetically heterogeneous one.

Phosphorus transport via macropore and attached particulates can enhance its leaching ability into drainage. This study was conducted to examine the subsurface transport of phosphorous in two lysimeters. The lysimeter surfaces were bare or planted with Bahia grass. The covered site had the higher macropore content. Water samples were collected at 1-h intervals and the subsurface flow (SSF) was collected at 10-min intervals. Twelve water samples from each of two storms on July 8, 1999 (event 1) and July 20, 1999 (event 2) were analyzed for phosphorus concentrations. The dissolved phosphorous was filtered with a 0.45 um membrane. The particulated phosphorous was not filtered. The results show that the particulated phosphorous concentrations were much larger than the dissolved concentrations in both sites and for both organic phosphorous and inorganic phosphorous. The accumulated phosphorous mass at the covered site was 2.1 to 9.0 times higher than at the bare site. The higher particulated phosphorous concentration at the covered site indicates that phosphprous attached to the macropore.

The results of calculations of adsorption properties of carbon single-walled nanotubes with typical diameters that are similar to the diameter of C₆₀ are submitted. We have carried out the quantum-mechanical calculations (using molecular mechanics, MNDO amd MNDO/PM3 methods) of zigzag-type (n,0) and armchair-type (n,n) single-walled nanotubes with adsorbed light (H, O, C, Cl) and metal (Li, Na) atoms, and compared the obtained results with those for the adsorption on graphite.

Many mesoporous materials have been prepared by the surfactant-assisted self-assemble of inorganic precursor. Most of such materials reported to date are amorphous in nature. If precise chemical structure can be build into the precursor, it may be possible to introduce certain crystallinity into the mesoporous framework by the same surfactant-assisted process. This proposal has been tested by preparing SiO₂ precursor with certain degree of zeolite structure before the incorporation of surfactant template. A mesoporous material was indeed produced that showed regular structure in nanometer scale by XRD, as well as double 5-ring molecular scale structure of by FTIR. The variation of properties with the preparation procedures will be discussed based on their XRD, ASAP and FTIR measurements.

Research is to make efforts on the modification of activated carbons surface by O₃, NaOH and by combining of NaOH and O₃ treatment. Results indicate that O₃ treatment significantly increases the specific surface area, pore volume and oxygen functional groups. Besides, NaOH treatment only increases pore volume and oxygen functional groups such as OH and COOH. Comparatively speaking, the formation of C=O functional group is few due to its formation energy. However, there is significant change when physicochemical properties of the activated carbon are treated by 3N NaOH. The result predicted from the stoichiometry of carbon treatment by NaOH is in accordance with from the amount of the surface acidity. Pore volume and amounts of oxygen functional groups significantly increases after activated carbon is treated by NaOH and by ozonation.

LPG (liquid petroleum gas) contains small amount of heavy oil. This heavy oil could be removed by adsorption with activated carbon fiber. Desorption of this heavy oil from the activated carbon fiber by supercritical propane was tried in this study.

The activated carbon fiber, adsorbed 1-nonadecene or lubricant oil from hexane solution, was prepared as model spent carbon. This spent carbon was packed in a column, which was set in a super critical fluid chromatograph with propane carrier. Reasonable adsorption and desorption amounts were observed to prove the possibility of this heavy oil removal process. Repeating operation of adsorption and desorption was successfully tried.

A new β-cyclodextrin phenyl isocyanate bonded stationary phase had been used to investigate the retention time, resolution, separation factor of fluoxetine under the different LC condition in the reverse phase. The optimal operation conditions were found. A base line separation of the enantiomers had been obtained. In the paper, the effect of mobile phase composition, ionic strength in the mobile phase, flow rate, pH and temperature on retention time, resolution and selectivity of R-FLU and S-FLU had been discussed. The experiment results show that these factors can all affect retention time and resolution, but have no or slight effect on the selectivity. The separation result is better than that reported before. When methanol-buffer ratio is 40:60, flow rate is 0.5 ml/min, the concentration of triethylamine in the mobile phase is 2%, pH is 5.0, and the temperature is 23°C, the retention time is 29.02 min and 25.72 min, respectively, the resolution is 1.130, the selectivity is 1.187.

A novel perfunctionalised cyclodextrin CSP immobilized onto aminized silica gel via a single ureido chemical bond was synthesized, exhibiting good enantiomeric separation abilities. A comparative study on the separation ability and separation-selectivity between the synthesized CSP column and the commercially available cyclodextrin-based column (CYCLOBOND I 2000 SN) was conducted. The results show that the ureido bonded CSP column depicted unique properties especially on the enantioseparation of β-adrenergic blockers and amine-containing racemic compounds which is complementary to the CYCLOBOND column which can separate acidic compounds more effectively.

Ash derived from palm oil wastes was examined for its ability to function as an adsorbent for heavy metal ions such as Cr(III) and Zn(II). Batch equilibrium studies demonstrated that the ash exhibited Langmuir maximum adsorption capacities for Cr(III) and Zn(II) of 251.6 and 126.7 μmol/g dry weight ash, respectively. Binary metal adsorption studies showed that the uptake of Zn(II) was significantly suppressed in the presence of Cr(III) while the binding of Cr(III) was moderately affected by the presence of Zn(II). In multiple cycles of adsorption and desorption studies bound Zn(II) was readily recovered from the ash by addition of 0.125% nitric acid. However, repeated exposure of the ash adsorbent to acidic conditions resulted in a drastic reduction in the adsorption capacity of the ash for Zn(II).

The effect of a double layered adsorber for the recovery of gasoline from the contaminated air emitted on gasoline storage tank and distribution facilities is investigated. Two different adsorbents, silica gel and activated carbon, are employed. A pilot unit consisting of an adsorptive buffer for the collection of contaminated air sporadically emitted and a PSA system was installed and tested. Contaminated air was fed to the adsorption buffer for a period of time, and subsequently the enriched hydrocarbon stream obtained during the regeneration of the adsorptive buffer was treated by the PSA system. Regeneration of the adsorbent was done by flowing air under vacuum. Approximately 15-25 liter of recovered gasoline per hour was recovered with the pilot scale PSA unit filled with 540 kg of adsorbent in total, and clean air containing light hydrocarbon less than 100 ppmv was emitted to the atmosphere.

An apparatus for measuring the adsorption equilibria and kinetics of gases at high pressures is described. A method of treatment of experimental data as well as examples of adsorption isotherms and kinetics are presented. Analysis of high - pressure adsorption isotherms at a number of temperatures for carbon dioxide, methane, nitrogen, oxygen and hydrogen on active carbon was carried out.

The investigations of adsorption from binary liquid mixtures on MCM-41 silicas are presented. Two mesoporous silica sieves were synthesized using ammonia as catalyst. n-Octadecylammonium bromide and n-dodecylammonium bromide were used as template reagents. The porosity of both silica samples was characterized by nitrogen adsorption-desorption method. The specific surface area was calculated using the BET equation. The parameters characterizing the structure of studied sorbents were estimated from adsorption data using the α_s method. For mesoporous MCM-41 samples the specific excess adsorption isotherms from methanol + benzene and benzene + cyclohexane mixtures were measured within the whole concentration range at 293 K using a static method. The adsorption isotherms from binary mixtures were interpreted in terms of adsorption model involving energetic heterogeneity of the studied systems.

Adsorption of benzene derivatives of various chemical properties from dilute aqueous solutions on one experimental activated carbon is investigated. The experimental isotherms were measured over a wide range of concentrations and solution pH, and for the constant value of ionic strength. The experimental systems were analyzed in terms of adsorption theory on energetically heterogeneous solids. The optimization isotherm parameters, i.e. equilibrium constants and dispersions of adsorption energies, were analyzed in terms of solute properties, e.g. number and character of functional groups.

The studies of adsorption from binary liquid mixtures on three activated carbons showing differences in their porous structure are presented. The chemical character of adsorbents was investigated by applying the potentiometric titration. The benzene adsorption-desorprion isotherms were used to estimate the porosity of carbon samples. The specific excess adsorption isotherms from benzene + 2-propanol and benzene + n-heptane mixtures were measured within the whole concentration range at 293 K. The liquid adsorption data were analyzed in terms of the theory of adsorption on energetically heterogeneous surfaces.

This paper considers the surface diffusion of propane, and n-butane in activated carbon. The flux measurements were obtained using a differential permeation apparatus, and the surface diffusion flux, and hence surface permeability, was obtained. The surface diffusivity and permeability increase very fast with loading. The increase in the permeability is not due to the surface heterogeneity. All theories in the literature exhibit a surface diffusivity increase with loading, but they all predict a decrease in the surface diffusion permeability, while experimental surface permeability shows an increase. In the light of failure of all models in the literature, we propose a new model here, where we follow the traditional approach of Darken, but allow for the activation energy for surface diffusion to be a function of loading.

This paper addresses the steady state diffusion and flow of benzene, ethanol and carbon tetrachloride in a commercial activated carbon over a complete range from low pressure to condensation region. Effects of temperature and particle length are also investigated. It is found that the total permeability is contributed mostly by Knudsen and surface diffusion at very low pressure, while at high pressures where capillary condensation is occurring, the surface diffusion and capillary condensate flow control the total flux through the particle. In terms of molecular weight and viscosity dependence, the permeability at low pressure is inversely proportional to the square root of molecular weight. While at high pressure it is approximately inversely proportional to over the region where the capillary condensation is occurring. A theory is presented in this paper to explain these observations.

A new method of studying the heat of adsorption is presented in this paper. In this method, a quantified heat supply is applied to a pellet, which is in equilibrium with a gas phase. As a consequence of the heating, the pellet temperature is increased and adsorbate is desorbed. Since the heat supply is controlled, the amount desorbed is a measure of the heat of adsorption. This is the opposite of common methods such as microcalorimetry, where heat released during adsorption is measured. In the experiment, the pellet is heated using a low voltage DC current. The temperature of the pellet and the pressure rise are monitored in the course of time. The obtained information is used to extract the heat of adsorption using energy and mass balance equations.

A theory for determining pore size distribution is presented in this paper. The effect of the pore size is accounted for by introducing the concepts of enhanced pore pressure and enhanced surface layering. In combination with the Kelvin equation, the theory is capable of modeling the adsorption in micro and mesopores indiscriminately. The technique is applied for nitrogen – activated carbon system at 77 K. The PSD results are comparable with those calculated using the DFT technique.

An approach to the adsorption equilibrium is developed on the basis of the condensation approximation: the micropore volume filling is regarded as an evolution of the two-dimensional condensation, which occurs on micropore walls at the critical condensation pressures.

A detailed simulation model was developed to investigate, in details, the effects of the important operating variables on the performance of a biofilter. It appears that full diffusion equations for pollutant transport in biofilms give rise to parameters, such as surface area and thickness of the biofilm, which are difficult to determine reliably. Theoretical findings related to the role of adsorption capacity of the support medium have been experimentally verified for biofiltration of MEK on compost and GAC. The model predictions, using independently measured equilibrium, mass transfer and reaction kinetic parameters, are in good agreement with the experimental results.

The dynamic competitive adsorption of a new fiber ACF-P for binary organic vapor mixture was preliminary investigated by gas chromatography. The competitive adsorption behaviors on various ACFs used, in different dynamic adsorption conditions and for different binary organic vapor mixtures were compared in this paper. The results show that ACF-P, as with ACF-W, can remove both components of organic vapor mixture with about 99% of removal rate before first breakthrough point. In dynamic competitive adsorption, the more weakly adsorbable component is earlier to breakthrough and produces a rolling up in the breakthrough curve because the displacement and desorption of stronger adsorbate. The dynamic competitive adsorption ability of a vapor component in mixture is mainly related to its boiling point as well as its structure at some time. The preparation temperature of ACF-P and adsorption condition have effect on competitive adsorption.

In this study chloro- and nitrophenols were used as the test adsorbates, and modified, non-modified granular activated carbon (GAC), and macroreticular synthetic resin were used as the adsorbent. These adsorbents have carbonaceous structure, with different surface properties. The isotherms for all three systems were determined by the so-called batch bottle technique. The Highest Occupied Molecular Orbital (HOMO) electron density at the adsorption site was estimated by CAChE program, because it is suggested that the electron density will affect the adsorption isotherm [4]. A qualitative relationship between Freundlich 1/n and the HOMO density of adsorbent was found by combining the experimental and computational results for modified GAC. The value of Freundlich 1/n for all adsorbent systems including modified GAC was decreasing linearly with an increase in “adsorbent” HOMO density. The HOMO electron density of both the adsorbate and the adsorbent was one of the major factors that determined the Freundlich Exponent, 1/n, for phenolic adsorbates - carbonecious adsorbents systems.

This study investigated the use of powdered activated carbon (PAC) for the removal of hop compounds, particularly isomerised alpha acids (iso acids), from beer. The average iso acid adsorption capacities of the adsorbents in descending order were activated carbon manufactured from wood, coal, peat and coconut. Iso acid adsorption capacity was found to be strongly correlated with increasing pore volume and surface area in the mesoporous region. A good correlation between the adsorption of colour and iso acids from beer existed. Carbon treatment had little effect on the majority of beer quality indicators however, trace volatile compounds were adsorbed differentially, with both hydrophobicity and molecular size appearing to be important.

In this paper the adaptation of the classical Kelvin equation is discussed and applied to synthesised MCM-41 materials. Other analytical techniques as X-Ray Diffraction (XRD) and Transmission Electron Microscopy (TEM) are used to confirm the results. The correction made by Broekhoff-De Boer (BdB) is a correction for forces emanating from the walls of the adsorbent and results in increasing pore size. The results obtained by the nitrogen adsorption experiments using the BdB equation look promising. A good correlation is found between the results of the BdB equation and the other analytical techniques. Corrections up to almost 40% are found for mesopores in the pore size range up to 4 nm. A good similarity is also found between the corrected Kelvin equation and DFT incorporating cylindrical pore geometry.

We report both experimental measurements and molecular simulations of the melting and freezing behavior of simple fluids in porous media. The experimental studies are for carbon tetrachloride and nitrobenzene in controlled pore glass (CPG), Vycor and activated carbon fibers (ACF). Differential scanning calorimetry (DSC) was used to determine the melting point in these porous materials. In the case of nitrobenzene (which has a dipole moment), the structural information about the different confined phases was obtained by measuring the dielectric relaxation times using dielectric relaxation spectroscopy. Monte Carlo simulations were used to determine the shift in the melting point, T_m, for simple fluids in pores having repulsive, weakly attractive and strongly attractive walls chosen to model a range of porous materials from silica based pores to carbon based pores, in terms of energy of the fluid-wall interaction.

We have developed a model of porous carbon morphologies based on the Reverse-Monte—Carlo (RMC) technique. Using rigid carbon basal planes of random size and shape, we use the RMC method to systematically refine simulated carbon—carbon radial distribution functions to match experimental distributions that can be derived from small angle x-ray scattering and neutron scattering (SAXS and SANS). We offer a test of our procedure by comparing a RMC generated carbon structure to a hypothetical carbon structure generated by molecular dynamics. We show how structural characteristics like porosity, surface area and pore size distribution are essentially the same in both structures—validating our model procedure. Nitrogen adsorption isotherms were calculated for our model using grand canonical Monte Carlo techniques and evidence of isolated capillary condensation was observed in pore regions as low as 14 angstroms —a result not seen in slit-pore models of activated carbons.

Although many studies have reported the effects of confinement on the physical properties of fluids and mixtures, there seems to have been few fundamental theoretical or simulation studies of similar effects on chemical reaction equilibria. We report a study, based on classical Monte Carlo simulations, of the effects of confinement in carbon slit pores on the reaction 2NO⇔(NO)₂. We find that the confinement greatly increases the dimer yield, and that this effect is most pronounced in small pores and at low temperatures.

A simple method was developed to characterize solid adsorbents concerning their applicability in open sorption systems for the heating, cooling and heat storage in buildings. An appropriate adsorbent has to reach the temperature limit given by the heating/cooling system of the building. It has to provide a high energetic efficiency and a high energy density for storage applications. The criteria for the characterization can be calculated from the adsorption equilibrium of water vapor and the adsorbent as basic experimental data. An adsorption heat pump and a thermal energy storage (TES) have been installed. Experimental data of these systems were compared to the theoretically calculated values.

Atmospheric trace xenon (0.1 ppm) was enriched up to about 4000 times (400 ppm) by means of a 2-stage, 4-bed pressure swing adsorption (PSA) process equipped with enriching reflux and parallel equalization (PEQ), and utilizing MS-13X. Even for one stage, Xe enrichment exceeded that obtained from a single bed batch adsorption process by 40 times and from a two-bed PSA process equipped with stripping reflux by 16 times. Therefore, this new PSA process concept should prove to be very effective at enriching trace components.

Adsorbing colloid flotation is a separation technique with distinct advantages for removing trace elements from natural and industrial waters. These advantages are: rapid processing, simple design, small volume of final solution, low residual concentration, flexibility of operation, and moderate cost. In this method, the substance to be removed (e.g. metal ions) is adsorbed on colloidal particles or coprecipitated on a floe produced by the addition of a flocculating agent. A surfactant is then added to adsorb on the particle or floe thus rendering it hydrophobic. The metal-loaded particle is then removed by further adsorption at the surface of rising bubbles in a flotation system. Experiments were carried out by using cetyltrimethylammonium bromide (CTAB) as a cationic surfactant, alone or in combination with thiourea (a complexing agent) for removal of gold. Adding 10 mg/L of thiourea to an equal volume of CTAB instead of CTAB alone, increased the extraction to a dramatic degree. The flotation experiments were conducted at various pH values and air flow rates. Extraction of the gold into the foam layer was strongly dependent on the pH of the feed. A distribution coefficient (K_d) of 46,000 was achieved. The method was applied to a number of natural waters draining gold mines and a base metal mine in the Hauraki Goldfields region (Coromandel), New Zealand.

A model is proposed using the Ideal Adsorption Solution Theory (IAST) and micropore size distribution (MPSD) concept to describe the multicomponent adsorption equilibrium and kinetics of gases in activated carbon. In order to overcome the thermodynamic violation, the IAST instead of the extended Langmuir equation is used to calculate the local multicomponent adsorption equilibrium. The micropore size is related to the adsorbate-adsorbent interaction energy by the Lennard-Jones potential. The overall adsorption isotherm and the diffusion flux of the adsorbed species are the integrals of their corresponding local values over all micropore size distribution range accessible by the adsorbate molecules. The size exclusion effect is taken into account in the competition of different sized molecules for a given pore. The model predictions are tested with the adsorption kinetics data of binary gases on Norit activated carbon. The results are better than those predicted by a previously proposed multicomponent adsorption kinetics model also using MPSD concept but with the local multicomponent adsorption isotherm being calculated by the extended Langmuir equation.

Mesoporous high-surface-area activated carbons were prepared from coconut shells by a simultaneous physical and chemical activation technique that employs carbon dioxide and zinc chloride as the physical and chemical agents, respectively. The resultant activated carbons possess an N2-BET surface areas above 2000 m²/g, a total pore volumes as high as 1.9 cm³/g, and a mesopore content (ratio of mesopore volume to total pore volume) of over 70%. The average pore size ranges from 2 nm to 3.5 ran. Both the surface area and the mesopore content could be tuned by controlling the experimental parameters, viz. ZnCl₂ to coconut shell (Zn-C) ratio, duration of exposure to the carbon dioxide atmosphere and temperature of activation. In particular, Zn-C ratios above 1 yielded high surface areas and ratios above 2 resulted in high mesopore content.

The adsorption capacities of coconut shells-based microporous and mesoporous activated carbons were examined by uptake of phenol, 4-Cholophenol (4-CP), 4-nitrophenol (4-NP), methylene blue and erythrosine red from aqueous or ethanol solution. The results showed that the uptake of phenol and substituted phenols depends on the microporosity of the adsorbent and the hydrophobicity of the adsorbate. The higher the hydrophbicity and the micropore volume, the larger the uptake capacity. The adsorption is in the order of phenol < 4-CP < 4-NP. The adsorption of dyes depends on the mesoporosity of the adsorbent.

The desorption of phenol was studied by boiling water washing and subsequently thermal treating phenol-saturated activated carbon in nitrogen at temperatures from 300 to 700°C. The adsorption capacity could be recovered by about 90-110%.

In relation to the zero emissions research initiative in steel industry, raw materials of slag and sludge from different two steelmaking processes, the blast furnace slag(metal-rich) and surface-treatment sludge(oil-rich), were investigated to device a new useful material as adsorbent. A mixing of their species by a high temperature heat-treatment at 600° provided a new characteristic material usable as specific adsorbent for phosphate anion in liquid phase and the material was also applicable to volatile compounds such as toluene or acetic acid in gas phase. Also, their behaviors were investigated by thier adsorption isothermals and/or spectrometric analyses with some analytical tools such as Vis./UV. Spectrophotometer, FT-IR or SEM, and then thus obtained material provided an adsorption of Henry type to phosphate anion in liquid phase and that of Freundrich type to some non-polar compounds such as toluene in gas phase. A surface structure of the material was also discussed in relation to an adsorption ability of some tested compounds such as phosphate anion or toluene by SEM and FT-IR.

For the minimization of energy consumption, we studied pressure swing adsorption (PSA) with atmospheric pressure adsorption and vacuum pressure desorption (VPSA). Na-X, Na-Y, USY, silicalite, silica-gel, and alumina were evaluated as water vapor adsorbents, and Na-X was chosen as the adsorbent for the VPSA-dryer by a flow method. Following the selection of water vapor adsorbent, water vapor removal from wet air (relative humidity of 60% at 298 K) using a VPSA-bench scale apparatus (Na-X honey-comb, 21iters/tower, two towers) was investigated. Experimental results were obtained with dry air generation of 1,000ppm water vapor, loading capacity of 590-2, 100m³N/h/ton, electric power consumption of 30 wh/m³N, and minimum water vapor concentration of 100 ppm.

Progress in the large-scale production of carbon nanotubes has opened up possibilities for new applications in some areas of adsorption and catalysis. In this paper we will discuss the characterization of single and multiwalled nanotubes by nitrogen, butane and methane adsorption. Nitrogen adsorption data will be used to calculate pore size distributions of the nanotubes. The incorporation of carbon nanotubes into pitch to change the porosity of the resultant activated carbon will also be discussed. By altering the pore structure of the pitch-based carbon fibers from highly microporous to a wider size distribution of micro and mesopores, new applications of the fibers can be envisaged in areas of environmental science and catalysis.

A generalization of the Kelvin equation is derived analytically taking into account wall interaction potential, the dependence of surface tension on curvature, the position of the dividing surface for a cylindrical interface, and the dependence of molar volume of confined liquid on capillary pressure.

Application of MCM-41 silicas and alkylsilyl-modified MCM-41 silicas as model adsorbents is discussed with a special emphasis on determination of the statistical film thickness in hydrophobic pores and the pore size as a function of the capillary condensation pressure for cylindrical pores. It is demonstrated that the use of the MCM-41-calibrated relations describing adsorption in hydrophobic pores allows one to dramatically improve accuracy and self-consistency of adsorption characterization of hydrophobic materials. To demonstrate the advantages of this new methodology, surface properties and pore structures of MCM-41 samples with different surface coverages of octyldimethylsilyl ligands are characterized.

Application of the adsorption potential distribution (APD) to calculate the average micropore width and the specific surface area is discussed. The monolayer capacity is evaluated from the amount adsorbed corresponding to the minimum between the monolayer formation and secondary micropore filling peaks on APD. Thus obtained monolayer capacity is used to calculate the specific surface area, which in turn allows one to evaluate the average micropore size from the micropore volume, assuming slit-like pore geometry. It is shown that low-pressure nitrogen isotherms and APDs for various activated carbons and activated carbon fibers with similar average micropore sizes are very analogous. This suggests that different microporous carbons may have similar microporous structures and similar widths of micropore size distributions.

Carbon nanotubes prepared by pyrolytic carbon deposition onto anodic aluminum oxide films were fluorinated in the temperature range from 323 to 473 K. Fluorine atoms are covalently bonded to carbon atoms on the internal surfaces of the nanotubes and the amount of attached fluorine increases with increasing fluorination temperature. The adsorptive properties of pristine and fluorinated nanotubes embedded in the template were investigated by means of adsorption and desorption isotherms of N₂ at 77 K. The isotherms for pristine and fluorinated samples were close to type IVisotherm, but the amount of adsorbed N2 molecules decreased with an increase in fluorination temperature. Although the α_s-plot for fluorinated samples had a good linearity at 0.6 < α_s < 1.5, a downward deviation from the linearity was clearly observed below α_s = 0.6. These results suggest that the fluorination of internal surfaces of the nanotubes provides the low surface energy for the cylindrical mesopore system of the tubes.

The relationship between the surface excess mass and absolute adsorbed amount for high pressure adsorption of supercritical gas is summarized in order to show new points proposed here. The present method determines the mean density of the adsorbed phase from the crossing point of the surface excess mass isotherm with the abscissa giving the volume of the adsorbed phase by iteration procedures. This approach was applied to high pressure methane adsorption on nonporous carbon black at 273-303 K. The obtained volume of adsorbed methane was great and the corresponding thickness of the adsorbed phase was in the range of 6.1 to 8.4 nm.

NH₃ adsorptivity of hollandite manganese oxide (H-HolMO), which consists of a [2X2] tunnel structure of [MnO₆] octahedral units, was examined using TG-DTA, isothermal adsorption, FT-IR, and TPD-MASS analyses. TG-DTA showed a gradual H₂O evolution with increasing temperature in the temperature range of 373 to 573 K. The H₂O/Mn mole ratios of the manganese oxide were 0.161, 0.082, and 0.035 for the sample pretreated in vacuo at 393,473, and 573 K, respectively. N₂ adsorption isotherms of the samples at 77 K showed no micropores irrespective of the pretreatment temperature, indicating that N₂ molecules cannot be inserted into the tunnel of H-HolMO. An NH₃ isotherm showed irreversible adsorption by chemisorption. The amount of NH₃ chemisorbed correlated well with the H₂O/Mn mole ratio. Also, XRD showed the expansion of the lattice by the NH₃ adsorption. These results indicate that NH₃ molecules are inserted and bind to lattice protons in the tunnel structure. Since IR analysis showed a main band at 1402 cm⁻¹, the adsorbed species was found to be NH₄⁺. Thus, a specific insertion of NH₃ into H-HolMnO was shown; NH₃ + H-HolMO → (NH₃)HolMO. TPD-MASS analysis showed that NH₃ is not only inserted but also reacts with H-HolMO in a redox process generating nitrogen and water; 2NH₃ + 6MnO₂ → 3Mn₂O₃ + N₂↑ + 3H₂O↑.

A method was developed to observe intracrystalline sorption processes in single crystallites of zeolite material. The change of the index of refraction in the probe material during uptake experiments is exploited for spatially resolved concentration profiles. Since as a feature of this method only two-dimensional data can be measured, additional efforts are made to calculate complete three-dimensional concentration profiles for special cases. The first results for the diffusion of methanol in zeolite A will be discussed. For this system, three-dimensional concentration profiles could be resolved.

The aim of this note is to briefly outline the physical principles and some technical aspects of traditional and new methods to measure gas adsorption equilibria in both single and multicomponent systems. The respective data are very important for the design of gas adsorption processes which are used more and more for both chemical engineering and environmental purposes [1, 2].

Based on a simple statistical model Langmuir's Adsorption Isotherm has been generalized to situations where interactions between adsorbed molecules (admolecules) on neighbouring sites no longer can be neglected but explicitly have to be taken into account. This is done by introducing the concept of “neighbourhood” of admolecules in the statistical derivation of the Langmuir Isotherm and specifying it by a simple phenomenological model. The resulting isotherm can be either of Type I, II or III depending on numerical values of the interaction parameters. Numerical examples will be given and discussed to a certain extent.

Sorption phenomena of swelling sorbents like polymers cannot be measured adequately by either gravimetric or volumetric methods, since the volume needed for buoyancy correction or dead space determination, respectively, depends not only on pressure and temperature of the sorptive gas but also on the amount of gas absorbed.

The oscillometric method proposed here allows to determine the inertia of mass by measuring the oscillating motion of a disk filled with the sorbent. By combining this method with gravimetric or volumetric measurements it is possible to determine simultaneously the change in volume of the disk's filling and the amount of gas absorbed in the sorbent. The sample can be either pulverous, pelletlike or of dense cylindrical shape.

Combined oscillometric-gravimetric measurements have been carried out for the system polycarbonate/CO₂ in the pressure range 0 < p < 6 MPa at T = 308 K and T = 318 K.

Sample data for the change in volume and the amount absorbed determined from these measurements are presented and discussed to a certain extend.

Optimum operating conditions of the H₂ PSA process from a typical cracked gas mixture (H₂/CO₂/CH₄/CO) using double layered beds of activated carbon, which removes CO₂ and CH₄, and zeolite 5A, which removes residual CH₄ and CO, is sought by means of a mathematical model. The model accounts for the mass and heat transfer resistance in the bed and adopts IAS theory to represent the multicomponent adsorption equilibrium. There is the optimum ratio of the blowdown time to adsorption time. It increases with the decrease of the purge quantity but does not depend on the height ratio of activated carbon to zeolite 5A. The purge quantity also has the optimum value, which decreases with the increase of the ratio of the blowdown time to the adsorption time. The optimum purge quantity depends on the height ratio of activated carbon to zeolite 5A.

The selective adsorption of a CH₂Cl₂ (1.5 mol%)/N₂ (98.5 mol%) mixture in AlPO₄-5 and MCM-41 was tested at pressures between 50-130 kPa and at a temperature of 318 K. AlPO₄-5 (pore diameter 0.76 nm) and MCM-41 (pore diameters 3.3 and 4.2 nm) showed selectivity for CH₂Cl₂ over N₂. These experimental results agreed qualitatively with GCMC computer simulations. The simulations indicated that MCM-41 is less selective for CH₂Cl₂ over nitrogen than AlPO₄-5.

Previous works in batch reactors have shown for activated carbon cloths quick mass transfer and high adsorption capacities against organic molecules present in water. In this study, activated carbon cloth is set up in a dynamic reactor to determine its adsorption performance for various micropollutants and operating conditions.

Experimental data of breakthrough curves for different conditions are generated using a pilot-unit to determine breakthrough times and adsorption capacities. Classical models set up for porous materials are not really useful for adsorption on activated carbon cloth filter. A neural network approach is then carried out in order to model experimental data by taking into account specific characteristics of cloths as input neurons. Equilibrium and kinetic parameters are also comprised in the input layer (pore diffusion coefficients and Freundlich constants). The relative influence of each input variable on the outlet concentration is analysed. Calculated and experimental data are compared, good predictions are obtained using this approach.

This paper presents adsorption-desorption cycles developed for the treatment of a mixture of ethyl acetate and ethanol vapours in air. A pilot unit was used to generate experimental data. The adsorption capacities are found about 100 mg.g⁻¹ for each compound at an inlet concentration of 3 g.Nm⁻³ and a mean flow velocity of 0.15 m.s⁻¹. The desorption step is performed by joule effect. The most important desorption parameters are the temperature, between 100 and 150 °C and the flow velocity of the nitrogen gas (0.001 m.s⁻¹) used to carry out the desorbed molecules. An experimental design methodology was used to assess the impact of these two operating conditions on the desorbate concentrations. In this case, the desorbate concentration values are 160 g.Nm⁻³ for ethyl acetate and 80 g.Nm⁻³ for ethanol. The ratio breakthrough time and desorption time is about four with a desorption yield up to 100 % between each cycle.

In the present paper, an experimental procedure is proposed to measure the total heat of sorption between a solvent and an absorbent. The method uses a thermogravimetric balance coupled with a differential scanning calorimeter. The calorimetric signal and the DTG signal are used in a simple model to determine a continuous expression of the total heat of sorption vs. the solvent mass content. The method is rapid and convenient for complex heterogeneous sorbents like foodstuff or sludge. Validation of the method for the couple microcrystalline cellulose - water is presented.

The adsorption dynamics of equilibrium separation beds and kinetic separation bed were investigated experimentally and theoretically through breakthrough experiments in the range of 4-16kg_f/cm² adsorption pressure and 4.5-9. 1LSTP/min flow rate. Using ternary hydrogen mixture (H₂/CH₄/CO; 60:30:10 vol.%) as feed gas, the experiments were performed on the equilibrium bed packed with activated carbon or zeolite 5A and the kinetic bed packed with carbon molecular sieve (CMS). To understand the adsorption dynamics of equilibrium and kinetic separation beds, a non-isothermal fix-bed adsorption model was used. The CMS bed showed similar characteristics of adsorption dynamics of the activated carbon bed. The roll-up of CO observed in the activated carbon bed and CMS bed. Also, there was a difference between mass transfer zones of CO and CH₄ in these beds. Therefore, these results brought temperature excursion twice. However, a small roll-up of CH₄ was seen in the zeolite bed without the temperature excursion.

Carbon whiskers/fibers with a diameter of approximate 1μ m and a maximum length of 20mm have been made successfully by thermal vapor deposition at 1000°C. A new recipe including nitrogen as carrier gas, benzene as precursor and ferric sulfate as the source of Fe catalyst were used for the formation of carbon whiskers on a ceramic substrate. A 2 cm. long substrate was coated with Fe catalysts by dipping-spining method using a 0.5M Fe aqueous solution. Drying process of the samples was completed at room temperature and then placed in an oven at 110°C prior to a thermal treatment at 1000°C in a quartz tubular reactor. The experiments were carried out using a mixture gas of 27% wt. vaporized benzene and 73% wt nitrogen with a flow rate of 13.9 ml/min for a deposition time of 5-90 min. Nano-scale hollow core carbon whiskers can be thereby generated. Characterization methods such as Scanning Electron Microscope (SEM), X-ray Diffraction (XRD), and Transmission Electron Microscope (TEM) and Thermogravimetric Analysis (TGA) were carried out to analyze the physical properties of the carbon whiskers.

Activated transport is one of the most important properties of microporous membranes which results in molecular sieving separation with very high selectivities. The flux J (mol.m ⁻².s⁻¹) through microporous membranes as a function of temperature can be derived from the isomeric heat of adsorption Q_st and the mobility energy E_m (kJ.mol⁻¹). In this work, high quality molecular sieving membranes have been successfully synthesised using sol–gel process to prepare weakly branched silica films. Maximum permselectivities calculated are 846 (He/CH₄), 130 (CO₂/CH₄) and 58 (H₂/CO₂). The effect of mobility energy on silica films is demonstrated by preparation methods such as single-step and two-step catalysed hydrolysis of tetraethoxysilane (TEOS) which leads to the production of weakly branched silica films.

Mesoporous MCM-41 silica was modified using both inorganic and organic precursors and characterized by XPS, NMR and XRD. The surface properties and pore structure of the modified MCM-41 samples were characterized by measuring the adsorption isotherms of nitrogen, volatile organic compounds (n-hexane, benzene, acetone, and methanol) and water vapor. Modification using organic reagents was found to result in a significant loss in porosity and surface properties. With inorganic modifying reagents, the decrease in porosity was also observed, while the adsorption behavior was similar to that before modification. Surface modification using both organic and inorganic precursors can drastically enhance the hydrothermal stability of MCM-41 materials due to enhanced surface hydrophobicity (in the case of organic modifications) or increased pore wall thickness (in the case of inorganic modification).

The structure and properties of carbonized titania/silica gels containing CVD-titania and carbon deposit were investigated using adsorption and theoretical methods. It has been shown that the porosity and fractality of all the studied samples decrease with increasing amounts of TiO₂ and carbon covered the oxide surface not only in pores but also at the outer surface of particles. The influence of the carbon deposit on the changes in an accessible surface area of adsorbents is lower than that of titania due to the difference in the morphology of these phases and in the types of contacts between the coverage and the substrate particles.

A mathematical model for nonideal adsorption equilibria in multicomponent mixtures is derived and applied with good results for pure components as well as multicomponent mixtures using only pure component data. The multicomponent model is derived on the basis of vacancy solution theory and Non-Random-Two-Liquid (NRTL) theory. The NRTL equation is able to describe immiscible liquid systems which indicates that the mathematical model may be capable of describing strongly nonideal adsorption systems. The combination of the vacancy solution and NRTL theories accounts for adsorbent-adsorbate and adsorbate-adsorbate interactions by means of surface phase activity coefficients which depend on temperature, composition and spreading pressure. Fugacity coefficients in the bulk gas phase are calculated by the Second Virial Equation.

An adsorption isotherm equation for nonideal pure gas adsorption based on vacancy solution theory and the Non-Random-Two-Liquid (NRTL) equation is found to be useful for predicting pure component adsorption equilibria at a variety of conditions. The isotherm equation contains five parameters which are determined by fitting pure component experimental data obtained at different temperatures. The isotherm equation is evaluated successfully by means of pure component experimental data for different adsorbates (methane, ethane, propane, n-butane, ethylene, propylene, carbon dioxide) and different adsorbents (Nuxit-AL active carbon, 13X molecular sieves, BPL carbon and carbon molecular sieve 5A). In order to fully describe the pure component adsorption systems, spreading pressure and isosteric heat of adsorption are also calculated.

The adsorption of basic and acid dyes from aqueous solution onto natural zeolite has been studied using an agitated batch adsorber. Several factors have been studied: agitation, initial dye concentration and adsorbent mass. The adsorption isotherms parameters for Langmuir, Freundlich, combined Langmuir-Freundlich and Redlich-Peterson were determined using the adsorption data. The computer simulation programs based on the solid diffusion and combined solid diffusion and external film resistance models were developed by ISIM dynamic simulator. Liquid - phase concentration at particle surface was assumed to be time depended. These programs were used for fitting the kinetic experimental data for all investigated systems.

A new method, catalytic silane cracking method, was developed for controlling the effective pore opening of MFI-type zeolite. In this method, silane is cracked catalytically on active sites (acid sites) of MFI-type zeolite to produce methane and ethane, leaving coke containing Si atom on acid sites. This coked sample is calcined in air stream. By this procedure, a mono SiO₂ unit is deposited on an acid site. The pore diameter at the position of SiO₂ deposit is decreased to about the minimum molecular diameter of carbon dioxide. By this method, the amount adsorbed of benzene was largely decreased, but the amount of CO₂ was kept constant.

A numerical model that takes into account layering of different adsorbent particle sizes and adsorber bed geometry is used to examine the effects of isotherm nonlinearity on breakthrough behavior. Fixed-bed adsorber simulations were conducted for the adsorption of solutes from aqueous phase onto activated carbon with varying degrees of isotherm noltnearity. Differences in the breakthrough behavior of two different fixed-bed geometries, namely, reverse stratified tapered adsorber (RSTA) and the conventional stratified cylindrical adsorber (SCA) are explained using this approach. The simulations indicate that the RSTA yields longer breakthrough time than the SCA, and the difference in breakthrough times between the two is maximum at low concentrations for the linear isotherm, and decreases as the isotherm becomes more non-linear at higher concentrations. For both the RSTA and SCA, the MTZ sharpens as the isotherm becomes more non-linear.

A new model is proposed to a priori predict single component adsorption equilibria of different gases and vapors on activated carbons and graphite/graphitized carbon black. Emphasis is laid on the Henry region, where the interaction energy between adsorbate molecules and the adsorbent determines adsorption characteristics. It is shown that Henry coefficients can be predicted on the basis of the three-dimensional van-der-Waals equation for gases and the Hamaker constant of the solid, which is different for activated carbon or graphite/graphitized carbon black, respectively.

Adsorption now plays a key role in environmental protection engineering and is used to remove pollutants from industrial effluents. The removal of acid dye by adsorption onto activated carbon and the removal of copper ions by sorption onto bone char have been studied using an agitated batch adsorber. Equilibrium studies have been carried out and rate processes have been investigated to study the effect of initial solute concentration and adsorbent mass. A diffusional mass transport model has been developed to predict the concentration versus time decay curves. The basis of the model is the shrinking core theory [1,5], which was originally applied to liquid phase systems [2,3,4] and based on determining a single effective diffusivity, D_eff. The model applied in the present paper analyses the role of D_eff in terms of pore, D_p, and surface, D_s, diffusion.

The effect of equilibrium vapor-phase pressure onto freezing in nanopore is examined employing a molecular dynamics (MD) technique in a unit cell with imaginary gas phase. The MD simulations showed liquid-solid phase transitions, at a constant temperature, with the variation in the equilibrium vapor-phase pressure below saturated one Thus found solid-liquid coexistence line exhibited extraordinarily skewed shape, which implies the importance of the tensile effect on freezing in nanopores. The capillary effect on shift in freezing point was successfully described by a model based on the concept of pressure felt by the pore fluid.

Preliminary results of a feasibility study on the possibility of employing the hot exhaust gas to heat an adsorptive natural gas (ANG) storage cylinder during vehicle operation is presented. Energy is transferred by forced convection from the exhaust gas to the carbon bed in a jacketed reservoir. The results presented here for a regular carbon bed show that the adsorption tank can attain isothermal performance with perfectly feasible exhaust temperatures.

Among different isotherm equations presented for the adsorption of water vapor by AC, those of Dubinin-Serpinsky (D-S) and Talu-Meunier (T-M) are used to fit 107 sets of experimental data found in literature in an attempt to relate the parameters found in these equations to AC properties. The characteristics of adsorption isotherm and hystersis loop are studied foT different type of AC. The effects of temperature on the shape of adsorption isotherm, as well as the effects of pore structure and adsorption cycling on the hysteresis loop are presented. It was found that the isotherm of water adsorption usually has a hysteresis loop, which occurs at a medium to higher water relative pressure. The width of hysteresis loop depends on the structure of carbon and for carbon with very small micropores the hysteresis loop does no exist. For a given type of AC the hysteresis loop would disappear if it were subjected to successive adsorption-desorption cycles. Physical and chemical activation of carbon is considered.

Activated carbon was prepared by chemical activation of Kalimantan peat with ZnCl₂ and H₃PO₄ as activating agents. The effect of process variable on activation process such as concentration of the chemical agent, and activation temperature were studied. The adsorption of phenol from solution was used for characterisation of activated carbon. The best operating condition for production of activated carbon from Kalimantan peat by chemical activation were 50% concentration of ZnCl₂ solution, and activation temperature of 550°C.

Gas adsorption kinetics in porous activated carbon is modeled using a nonisothermal, nonisobaric Maxwell-Stefan model. Mass transfer is described by three mechanisms: pore diffusion (molecular and Knudsen diffusion), viscous flow and surface diffusion. To distinguish between the different types of transport in the porous particles two types of kinetic uptake measurements were performed; responses to carbon dioxide pressure steps and responses to concentration steps of carbon dioxide at constant feed pressure. The unknown transport parameter, the surface diffusion coefficient, was determined by fitting the model to both types of experiments.

The method of preparing mesoporous activated carbons useful for the adsorption of large molecules is desirable. The development of mesoporosity of activated carbons using catalytic effects of metal compounds has been previously reported. However, a non catalytic activation method which produces pores whose radii range between 2 ∼ 10 nm has not been reported as far as we are aware. The objective of present work is to propose a novel pre-treatment method for steam activation without gasification catalyst to improve the mesoporosity of activated carbons. In this article a calcium compound was mixed with raw material(PET), the samples were carbonized, and an acid treatment was then performed. We can completely wash the calcium compound out of the carbons by the acid treatment After this pre-treatment the carbon was activated by steam. Although the catalytic activation was not expected, we can prepare activated carbons that have large mesoporosity.

Capillary condensation hysteresis is studied by non-local density functional theory (NLDFT) and MC simulations with the example of nitrogen and argon adsorption in cylindrical pores of highly ordered MCM-41 materials. Good agreement is found between the theoretical predictions and experimental data. In sufficiently wide pores (>5nm in case of nitrogen) the NLDFT model predicts both the adsorption and desorption branches of the experimental isotherm.

The research is based upon a methodological foundation which has been developed over the last years in our laboratories. Along with water vapor adsorption isotherms and CHN analysis, we used a set of advanced magnetic resonance techniques, including low frequency Dynamic Nuclear Polarization (DNP) and EPR, both pulsed and cw, Pulsed-Field-Gradient (PFG) spin-echo NMR spectroscopy and relaxometry, X-ray scattering at small and large angels, SQUID magnetometry, as well as several independent methods to measure the electron spin-lattice relaxation times on the surface of chars, in order to address structural and molecular dynamic questions in a number of systems. Such a multi modality approach gives us new possibilities to study the basic chemical-physical aspects of solid-liquid interactions in porous media. A developed methodological approach has been applied to a new class of oxygen-sensitive paramagnetic carbon particles recently synthesized at the Illinois Research EPR Center (IERC) for use in magnetic resonance oximetry.

Adsorption and desorption isotherms of krypton on a mesoporous silica, MCM-41, were measured at twelve temperatures in the range from 80 K to 130 K using a newly designed cryostat. The isostehc heats of adsorption and desorption were calculated as a function of loading from the slopes of the interpolated adsorption and desorption isosteres. The isosteres in the pore filling regime showed no indication of a phase transition from fluid to solid in the experimental temperature range. The critical temperature for hysteresis was calculated to be 136.4 K.

The dynamics of the interaction between NH₃ and powder or granular commercial zeolites A, X, and Y were studied at 133 Pa and 400 K by the batch frequency-response (FR) technique. The residual water content was changed by evacuating the samples at different temperatures. Under the near-equilibrium conditions of the measurements macropore diffusion was found to be the rate-determining process in all of the beads. Sorption controlled the rate in the powders with the exception of K-A, where diffusion in the micropores was rate determining. It was found that the ammonia diffusivities are not influenced by the presence of adsorbed water in amount less than one molecule per K^* cation. The resonance peaks of Na− or Ca-A shifted to higher frequencies as the extent of dehydration increased. The FR method can be effectively used for most zeolites to characterize the dynamics of NH₃ sorption.

Equilibrium and dynamics of C₄ alkane sorption in H-ZSM-5 and Silicalite-1 were characterized by sorption isotherms and frequency-response (FR) spectra. At loading lower than 4 molecules/unit cell the isostructural samples showed similar adsorption properties: at comparable equilibrium conditions they adsorbed close to twice as much n-C₄ than i-C₄. The FR results suggested that at low loading the rate of intracrystalline diffusion controlled the rate of sorption. The transport diffusivity of the i-C₄ was about one order of magnitude lower than that of the n-C₄. Diffusivities were determined by the structure of the sorbent and the sorptive and were only slightly affected by the presence or absence of acid sites in the sorbent FR results obtained for a C₄ isomer mixture suggested that, at low loading, the diffusivity of the components is similar to those of the pure gases.

The concept of the electric double layer capacitor was applied in order to remove typical ionic substances from water by adsorption on activated carbon, using electrostatic force. Ions are captured on the electrode of the opposite polarity when the direct current (up to 2 volts) is applied, and captured ions are released when electrodes are either short-circuited or discharged. Therefore, by changing the pattern of the power supply, concentration of ions in the outlet stream is changed periodically. Activated carbon fiber was used as an electrode. Typically, municipal water (300μS/cm) was passed at a flow rate of 0.1-3 ml/min/g-carbon. The degree of ion removal increased with increase in the voltage applied and the BET surface area of the activated carbon used, whereas it decreased with increase in the flow rate. In the cyclic operation, typical removal ratio was over 85% and 50-80% of the inlet flow was recovered in the product.

The amounts of O₂ adsorbed on zeolite SA and an activated carbon fiber (A10) at 77K increased with increasing magnetic fields up to 6T. Upon mixing an oxide-superconductor, YBa₂Cu₃O_x, with zeolite 5A or A10, the desorption of O₂ from the adsorbents (magnetodesorption) was induced along with an increase in the magnetic fields at 77K, because, presumably, a local magnetic-field gradient in the pores was produced by magnetic flux expelled from a superconductor owing to the perfect diamagnetism. These results demonstrate a possibility for a magnetic control in adsorption based on peculiar magnetic states of physisorbed oxygen in micropores.

The phase separation of various triethylamine/water mixtures (mole fraction of triethylamine X) in silica pores was examined by differential thermal analysis and H NMR as a function of the pore size (d). The phase-separation temperature in confined systems shifted by |ΔT| to a lower temperature than that in the bulk phase. |ΔT| increased from zero for small pores to a maximum with increasing d, followed by a decrease according to the inverse power of the pore size. The pore size, giving rise to the maximal |ΔT|, increased from 18 to 40 nm with increasing X from 0.1 to 0.4. This, d-dependence in |ΔT| corresponds very much to those of the self-diffusion coefficient (D) of water in a confined mixture at 298 K (one-phase region). The smaller is D, the larger is |ΔT|. This suggests that the translational molecular motion or the composition fluctuation as well as the selective adsorption of water should be related to the phase separation in confined spaces.

Experimental and modeling study on adsorptive distillation process for ethanol drying was undertaken. Ethylene glycol was adopted as entrainer and zeolite 4A was used for the adsorption of water. In the experiment, over 99wt% of ethanol was obtained with 95wt% of ethanol feed. Equilibrium plate theory was proposed for the modeling of the process. UNIQUAC model was used for the vapor-liquid equilibrium calculations and Gibbs adsorption equation was used for liquid-adsorbed phase equilibrium calculations. The equilibrium plate theory well represented the experimental measurement of temperature profile in the column and product purity.

Amorphous materials are typically characterised using nitrogen adsorption isotherms at 77K to obtain pore size distributions. However higher temperatures and different adsorbates could also be used. We have measured the adsorption isotherm of nitrogen at room temperature and 77K on the typical high surface area amorphous activated carbon AX21. We analyse these isotherms using density functional theory to obtain pore size distributions and explore whether they vary with temperature and compare with results obtained⁵ for carbon dioxide and methane. We find, for example, that the pore structures predicted for AX21 at 77K and 293K are quite different. The room temperature pore size distribution showing a unimodal micropore peak not seen at 77K. The room temperature pore size distribution is used to predict the selectivity of AX21 for methane over nitrogen at 293K between 0 and 1 bar.

The issue of global warming requires that prudent measures be taken to minimize the buildup of carbon dioxide in the atmosphere. Storage of CO₂ in coal mines is one way to solve this problem. Its benefits are multiple: (a) CO₂ sequestration, (b) recovery of valuable coalbed methane; (c) elimination of explosion hazards. There are two fundamental issues in the assessment of its feasibility and comparative advantages: (1) How much CO₂ can be stored and how much methane is displaced? (2) Which coalbeds are the prime candidates for CO₂ sequestration? This project was designed to begin to answer these questions. As expected, the uptakes found for the coals are lower than those on the high-surface-area activated carbons, but the trends are the same: higher uptakes and higher adsorption heats for CO₂ than for CH₄. The effect of the outgassing temperature is important, because most experimental data available in the literature were not obtained with CO₂ sequestration in mind and the potential sensibility of coal's pore structure to moisture content is well known. Results for three coals from the Argonne Premium sample bank are presented and discussed.

The adsorption of binary mixtures on heterogeneous surfaces is studied by computer simulation. The process is monitored by following the partial and total adsorption isotherms, adsorption energies, coverage fluctuations and differential heat of adsorption. These quantities are calculated by using Grand Canonical Monte Carlo simulation. Heterogeneous surfaces are represented by two kinds of sites, the so-called bivariate surface. Deep and shallow sites with energies ε_S and ε_D form l × l patches distributed at random or in a chessboard-like ordered domains on two-dimensional square lattices. The partial and total isotherms, as well as other thermodynamic quantities are very sensitive to the surface energy topography, particularly, when repulsive lateral interactions between adsorbed molecules are present.

The exact solutions for thermodynamic functions as well as for the collective diffusion coefficient of an interacting linear rigid adsorbate in a one-dimensional lattice are presented. Using this result and a connectivity ansatz we proposed a model of adsorption to analyze the deposition of non-interacting k-mers on heterogeneous two-dimensional substrates. Comparison with experimental isotherms of CO₂, N₂, and O₂ adsorbed on zeolites 5A and 10X are also presented.

The general basis for accurate entropy calculations of generalized lattice gases is presented. The me-thod relies upon the definition of an artificial Hamiltonian and thermodynamic integration. Entropy of interacting dimers and trimers in one and two-dimensional regular lattices are shown. Novel features of the coverage and temperature dependence of entropy in these systems are discussed.

This work shows a detailed dynamic model for a simulated moving bed (SMB), which takes into account non-instantaneous equilibrium at the solid-fluid interface. The theoretical model represents “n” fixed beds connected in a circuit with periodic shifting of boundary conditions. This corresponds to the advancement, from time to time, of the inlet (feed and eluent) and outlet (raffinate and extract) streams one bed ahead in the direction of fluid flow. The cases studied involved few columns per section, which is a typical case in many selective adsorbents with small particle size (<300 μm). Cases of instantaneous solid-fluid equilibrium were compared with those including increasing resistances to equilibrium attainment and different performance parameters were obtained. Experimental results for SMB separation of a chiral epoxide with cellulose triacetate (CTA) were well correlated with the proposed model using modified Langmuir isotherms. This model may also be useful for the adsorptive separation of proteins and macromolecules, where equilibrium of adsorption is usually not readily achieved.

The propylene/propane gas separation has been carried out with a fixed bed adsorption unit. High pressure feed and blowdown experiments completed with single component and kinetic experimental data and numerical simulations allowed us to define a five step Vacuum Swing Adsorption process designed to split an equimolar mixture of propylene/propane: I) Pressurization with feed, 60 s; II) High pressure feed, 120 s; III) High pressure purge with product, 120 s; IV) Co-current blowdown, 60 s and V) Counter-current vacuum blowdown, 120 s, where the enriched propylene product is obtained. The column packed with 13X zeolite pellets is fed with an equimolar feed of C₃ diluted to 50% with nitrogen at 5 bar and 423 K, while the product is withdrawn when the total pressure is lowered from 0.5 to 0.1 bar. After 15-20 cycles, the cyclic steady state condition is achieved and a propylene enriched stream of 98% mole relative to propylene/propane mixture, with 3.2% of nitrogen, a recovery of 19% (molar basis) with a productivity of 0.785 mol/kg.h are obtained.

The effects of compacting a highly microporous carbon, a graphite, and blends of the two carbons on CH₄ storage capacity, adsorption rate, and compaction properties were studied. Compaction was found to increase the uptake capacities and adsorption kinetics. High capacities, 97-119 ml CH₄ per ml carbon, were obtained with compacted, blended carbons when the methane adsorption rate was 9.3 e-5 g CH₄ per sec or better. The concentration of a more expensive microporous carbon component could be reduced by as much as 50% by volume in compacted blended samples while maintaining storage capacities. This effect has economic implications for practical gas storage. Additionally, the compaction properties of the blended carbons were significantly better than the compaction properties of the highly microporous carbon alone.

Starting from the fundamentals of the Statistical Rate Theory of Interfacial Transport, equations have been developed to calculate adsorption energy distribution from experimental TPD spectra. It is shown that, depending on the conditions at which TPD experiment is carried out, various equations are to be applied for that purpose.

The acid properties of mesoporous H-AlMCM-41 and H-FeMCM-41 catalysts were studied by temperature programmed desorption of ammonia. The results suggest the existence of four different types of acid sites. They are assigned to: (i) weak acid sites due to surface/terminal hydroxyl groups; (ii) moderate structural (Brønsted) acid sites associated with symmetrical tetrahedral Al(III) or Fe(III); (iii) strong structural (Brønsted) acid sites from a highly distorted tetrahedral Al(III) or Fe(III) and (iv) Lewis acid sites arising from tricoordinated Al(III) or Fe(III) in the framework. The t-butylation reaction of phenol over H-AlMCM-41 and H-FeMCM-41 showed high substrate conversion in the case of the former than the latter. This could be attributed to presence of large excess of type (ii) and type (iii) acid sites in H-AlMCM-41.

Transport of gases through porous materials that exhibit bimodal pore size distribution is determined by a combination of molecular, Knudsen and/or micropore diffusion. To isolate the mechanism of mass transport through such materials, transport rates of small alkane hydrocarbons were measured through single pellets of industrial adsorbents SA and 13X in a Wicke-Kallenbach set up. Experiments were performed with dry pellets as well as pellets with different levels of water loading to block the micropores present in the material. Transient responses were shifted due to water loading. Steady-state apparent effective diffusivity in 13X pellet was also affected. Macropore-micropore-macropore and micropore in series model was used to explain experimental observation.

Carbon monoxide adsorbents of high separation capacity and selectivity for industrial waste gas treatment have been prepared by solid state ion exchange technique. The suitable operating temperature range for the reversible chemical adsorption and diffusion process was optimized to be 358 ∼ 363 K. The actual adsorption capacity of CO was reached up to 46.6 ml/g. Influence of different ion exchange conditions of Cu(I) on adsorption and separation was also explored and discussed.

The structure of the interface formed by capillary condensation of a gas in a pore was investigated by means of the Grand Canonical Ensemble Monte Carlo method. The pore was represented by parallel plates in the z direction of the Monte Carlo cell but the adsorbate-wall interaction was truncated in the x direction to create a finite pore in order to facilitate the formation of an interface parallel to the wall-adsorbate surface. It is shown that at a low gas pressure the adsorbate consisted of essentially two high density layers adsorbed on each wall with the region in the middle of the cell consisting of bulk gas. By contrast at a higher gas pressure the gas condenses to form three adsorbed layers and a region of essentially bulk liquid in equilibrium with the gas in the cell centre with a well defined and stable liquid/gas interface.

Small angle scattering from a range of porous solids including coal, processed coal, glassy carbon, charcoal and porous catalyst supports indicates that the surfaces of the pore walls in these solids are not smooth i.e. that Porod's law is not obeyed. However, the small angle scattering curves for both X-rays and neutrons for many systems still obeys a simple power law dependence with scattering angle, as in Porod's law, but with a power law index α not equal to 4. Depending on the value of a this scattering behaviour may be interpreted as arising from different effects. Some porous solids have walls whose surfaces are fractaly rough i.e. exhibit self-similar surface structures, whilst some exhibit behaviour characteristic of mass fractals. Other systems, however, have been described as having pore walls which are not sharp i.e. possess fuzzy interfaces.

Here we first briefly review previous work in this area and then show how to derive general scattering laws for all the above cases which lead to a power law scattering curve. Two methods are used one being physically motivated whilst the second is a general mathematical analysis showing what conditions have to be met in order for the scattering laws to be rigorously obeyed.

Secondly we describe a newly developed Synchrotron based ultra small angle X-ray scattering apparatus of the Bose-Hart design for measurements with scattering vectors of magnitude between about 0.001 to 1.0 nm⁻¹. We then discuss an application of this instrument to investigate the evolution of the pore structure of glassy carbon as a function of its formation temperature.

The basic properties of a series of cationic zeolites have been investigated by means of FTIR and TPD coupled with mass spectrometer techniques using ammonia as probe molecule. Our TPD results demonstrate clearly that the adsorption strength of ammonia on these cationic zeolites is strongly dependent on the Lewis acidity of counter-ions, indicating that the interaction between ammonia and cationic zeolites is dominated by an electrostatic interaction between counter-ions and the lone electron pair on nitrogen atoms. However, the variation of the shift value in wavenumber of vibration bands of ammonia adsorbed on different cationic zeolites revealed by FTIR can not be explained only by this electrostatic interaction. The existence of a supplementary interaction between the negatively charged oxygen atoms of framework with the hydrogen atoms of ammonia is postulated.

The location of benzene in KL zeolite upon coadsorption of ammonia and CH₃NH₂ has been investigated and the interaction strength of ammonia, methylamine and benzene with KL has been evaluated by means of infrared spectroscopy. The adsorption behaviour of benzene in KL zeolite has been correlated with the average negative charge of oxygen atoms, counter-ions, zeolite structure and the interaction strength of ammonia, methylamine and benzene with KL zeolite. It has been also observed that besides the benzene adsorption on counter-ions, the 12R windows occurring in KL, which are not the preferential adsorption sites for benzene in the presence of benzene alone or upon coadsorption of ammonia, become favourable adsorption sites for benzene upon coadsorption of methylamine. The adsorption of benzene on 12 R windows becomes possible because of a strong deformation of KL zeolite framework upon coadsorption of methylamine.

The electronic structure of activated carbons is discussed by using the semi-empirical molecular orbital (MO) method. Micrographites, which contain several benzene rings, have been used as models for the surfaces of activated carbons. The electronic structure of micrographites and the adsorptive interactions between micrographites and adsorbates (water and its cation and its anion, aniline and its cation and its anion) in gas phase have been estimated by PM3 method. It is found that there is a tendency that the values of the adsorptive interaction energies increase with the size of micrographites. It is also found that the adsorptive interactions for anions are stronger than those for the cations and the neutral molecules. This is because the interactions between micrographites and anions are chemisorptive rather than physisorptive. It is also found that the chemisorptive interactions between micrographites and anions are still strong even when the solvent (water) is present in the system.

Experiments were conducted to investigate the adsorption behavior of phenolic compounds, namely phenol and chlorophenol, onto activated natural zeolite. Adsorption of both compounds is studied in a binary component system and the results are compared with that of single component. The experiment was conducted in equilibrium batch system at various temperature and various ratio of solution to solid. Competition effect was studied by adding a certain volume of aqueous solution containing certain concentrations of phenol and para chloro phenol (PCP) (1:1) into certain amount of zeolite. The mixture was then agitated and kept at certain temperature. It is indicated that adsorption capacity of zeolite for phenol at binary component system is much lower than that at single component system. This is in contrast to the capacity for PCP, which shows higher at binary component system. Adsorption capacity for phenol and PCP at binary component system increases by increasing the temperature (30-40°C). It is indicated that chlorophenol was the preference compound to be adsorbed compared with phenol. The Langmuir competitive model describes quite well such a binary component adsorption.

RF hydrogels were synthesized by the sol-gel polycondensation of resorcinol with formaldehyde and RF cryogels were prepared by freeze drying of the hydrogels with t-butanol. The cryogels were characterized by nitrogen adsorption and density measurements and their porous properties were compared with those of the aerogels prepared by supercritical drying with carbon dioxide. RF cryogels were mesoporous materials with large mesopore volumes > 0.58 cm³/g. Although surface areas and mesopores volumes of the cryogels were smaller than those of the aerogels, the cryogels were useful precursors of mesoporous carbons. Aerogel-like carbons (carbon cryogels) were obtained by pyrolyzing RF cryogels in an inert atmosphere. The carbon cryogels were mesoporous materials with high surface areas > 800 m²/g and large mesopore volumes > 0.55 cm³/g. When pyrolyzed, micropores are formed inside the cryogels more easily than inside the aerogels.

Mesoporous silica (MPS) and aluminosilicates (MPAS, Si/Al=5-49 in molar ratio) with uniformly sized mesopores between 22 and 25 Å were synthesized by a rapid room-temperature method which basically consisted of a liquid phase reaction for 3 h and air-calcination at 823 K for 6 h. The specific surface area (Sa) of MPS was over 1200 m² g⁻¹ and the Sa value tended to decrease with increasing the Al content. The adsorption isotherm of SO₂ at 298 K on MPS was almost linear up to high relative pressure (p/p₀≈0.85), indicating that the interaction between SO₂ and MPS is weak. The incorporation of Al in the silicate framework was effective to improve the SO₂ adsorption property. The present results have proven that the mesoporous silica and aluminosilicate are potential adsorbents of SO₂ with good reversibility at room temperature.

A new candidate surface area and pore volume reference material is proposed. A mesoporous gel-derived silica monolith has been investigated by nitrogen sorption and mercury intrusion porosimetry. Specific surface area (165.5 m²g⁻¹) specific pore volume (0.986 cm³g₋₁) and mean pore radius (119Å) have been monitored for up to twelve months by nitrogen sorption. No significant changes occurred following either storage or repeated usage. The pore structure remained unaltered despite multiple mercury intrusion cycles up to high pressure. This remarkable stability and durability confers properties that compare favorably with those of commercially available reference materials. The primary advantage of the gel-silica monolith is the elimination of errors associated with powder sampling.

An approach based on statistical thermodynamics is developed for mixed gas adsorption. The model takes into account energetic heterogeneity of zeolite and molecular rearrangement in a cavity, which leads to decrease of internal energy and entropy of mixtures. The information of the heterogeneity of the adsorption field in a cavity is evaluated from pure component isotherms. The approach allows predicting selectivity and total amount adsorbed without any empirical parameters. It is shown that all experimental points are to be in the range restricted by two curves, which correspond to ideal mixture and to the limiting case of non-ideal mixture, respectively. The model has been confirmed by analysis of highly non-ideal systems oxygen – nitrogen and oxygen – carbon monoxide – zeolite 10X at 144 K.

Equilibrium adsorption of cyclopentane, benzene and their mixtures on activated carbon AC from the statistical thermodynamics viewpoint is considered. The approach suggested allows determining of state of binary mixture adsorbed on the level of a separate micropore. It is shown that excess internal energy and entropy are negative due to rearrangement of molecules in non-uniform adsorption field. The excess Helmholtz free energy is positive at low loading of a micropore and becomes negative near to saturation limit. The negative departure from Raoult's law in the adsorption phase behavior is proved to be caused by decreasing of the excess Helmholtz free energy with increasing of total number of molecules in a micropore rather than its negative value.

To establish the nature of oxygen transport into chars burned under Regime II conditions; measurements of gas diffusion coefficient into partially burned chars were conducted using a non-isobaric gas chromatographic technique. The chars used included two low volatile bituminous chars, Norwich Park and Collinsville and one brown coal chars. The chars were prepared by conducting the combustion in an entrained flow reactor between 900- 1300 K and comparisons were made with chars burned under Regime I in a fixed bed between 500-800K. The experimental results were compared with coefficient predicted from Wheeler's equation based on the assumption of Knudsen diffusion. The results suggest gas transport occur mainly by activated diffusion. This mode of transport appearing more as Knudsen diffusion at the elevated temperatures. Analysis of the surface area suggest the development of fine pores under Regime II condition, which may suggest that the activated diffusion is accompanied or is due to surface migration of oxidant at the elevated temperatures.

The structure of coal char and its development during combustion has been a key interest among workers in this area. In particular the importance of the role of the pore structure relative the chemical reactivity on the observed rates. This paper reports on the pore development of three Australian coal chars with combustion under high (Regime II) conditions. The chars were prepared from a swelling low volatile bituminous coal (Norwich Park), a non-swelling low volatile bituminous coal (Collinsville) and a brown coal (Loy-Yang). The variation in the reactivity of these three Australian coal chars has been estimated in a fixed bed under Regime I conditions. The effect of the pore structure and development on the global reactivity of the chars burnt under Regime II conditions were examined. The results suggest a significant role of pore structure in determining the observed global rates of chars under Regime II conditions. The study also provides results suggesting possible role of surface migration of oxidant in the development of fine micropore under Regime II conditions

Two new pillaring techniques to synthesize Al, Zr and Fe-pillared interlayered clays are proposed in order to overcome the problems encountered with the conventional direct pillaring procedure, such as interstratiflcation of pillared and unpillared layers and the non-ideal stacking of clay aggregates. The first method is based on the pre-adsorption of template amines prior to the pillaring. The amines influence in a positive way the pillar distribution and the parallel orientation of clay sheets, the latter being extremely important in case of small laponite sheets. This leads to higher micropore volumes and larger adsorption capacities for N₂ and O₂. The second new procedure of self-assembly is also discussed allowing the synthesis of homogeneous Al-pillared laponite films on a Au-support. For this method, a close control on the pillaring is possible due to the stepwise character of the process. The films are tested for their adsorption capacity of six volatile organic compounds.

In this contribution, the synthesis and characterization of a new type of Al-grafted porous clay heterostructure (PCH) is discussed. The support PCH consists of small saponite clay plates, bound by a templated silica matrix. This 3-dimensional solid has a unique pore structure, with a high BET surface area (999 m²/g), microporosity (0.306 cc/g) and mesoporosity (0.633 cc/g). Based on the ‘Molecular Designed Dispersion’ method, a grafting of the PCH surface with aluminium oxide species is performed. The reaction involves the interaction between the aluminium acetylacetonate Al(acac)₃ complex and the silanol surface groups of the PCH, followed by a temperature treatment. The acidic properties of the Al-modified PCH are evaluated by the adsorption of acetonitrile-d₃(CD₃CN), revealing a strong Bronsted acidity.

Two techniques are used to modify the cubic MCM-48 structure for use in catalytic applications: an in situ incorporation and a post-synthesis modification. The direct hydrothermal synthesis, using the vanadyl sulfate pentahydrate, yields crystalline V-containing MCM-48 structures with high surface areas and narrow pore size distributions. The second technique, the molecular designed dispersion method, creates VO_x species on the surface of the MCM-48 support, while the crystallinity and porosity properties are maintained. The coordination environment of the V-ions of the incorporated and supported VO_x-MCM-48 materials was investigated by spectroscopic techniques. FTIR-PA measurements reveal the appearance of VOH bonds. UV-VIS diffuse reflectance spectra show that the VO_x centers have a different geometry depending on the Modification technique used and on the vanadium content.

The use of dimethyldichlorosilane as a coupling agent for the grafting of VOx structures on the MCM-48 surface, produces a material that is simultaneously hydrophobic (immiscible with water) and very active (all V-centers are accessible). The VOx surface species are grafted by the Molecular Designed Dispersion of VO(acac)2 on the silylated surface, followed by calcination in air at 450°C.

These hydrophobic MCM-48 supported VOx catalysts are stable up to 500°C and show a dramatic reduction in the leaching of the V-centers in aqueous media. Also the structural stability has improved enormously. The crystallinity of the materials does not decrease significantly, not even when the samples are subjected to a hydrothermal treatment at 160°C and 6.1 atm. pressure.

Synthesis of iron oxide nanoparticles in and on the zeolite-Y structure has been achieved. The influence of the iron solution concentration on the iron oxide formed and the role of the iron on the zeolite structural damage were evaluated by AAS, infra red analysis and XRD. The result indicated that the iron cations play a role in zeolite structural damage by bonding with the zeolite framework. The iron oxide partly dispersed on the external surface of zeolite-Y can be removed by dissolution using EDTA as a media/solvent, and so the remaining iron is assigned to be encapsulated inside the zeolite. The small oxide nanoparticles located in the zeolite cages exhibited a significant blue shift in the UV-Vis absorption spectra, in contrast to the respective bulk oxide, due to the quantum size effect.

This paper investigates the ability of a detailed numerical model of multiple-layer, non-isothermal, bulk gas pressure swing adsorption processes to match experimental data over a range of operating conditions. The experimental apparatus is a 2-bed, 8-step VSA process for air separation using multiple layers of alumina and CaX zeolite. Input parameters are determined independently of the VSA system and no parameters in the model are adjusted to improve the fit to experimental data. A limited range of system purities and pressure swings are examined to illustrate the ability of the numerical model to capture overall performance. In general good agreement is obtained between the model and the experimental data although accurate modelling requires careful attention to details that influence the thermal behaviour of the process. The numerical model can also represent the unusual temperature profiles characteristic of multiple-layered beds. Overall, while good agreement can be obtained between model and experimentation, accurate design must still rely on experimental work. The model is, however, sufficiently accurate to permit approximate optimisation studies to be performed.

Control of pressure and vacuum swing adsorption systems is challenging due to the unsteady state nature of the processes and the high degree of system non-linearity. System purity, flow rates and pressure must be regulated to within small tolerances to ensure compliance with customer requirements. Traditionally, industry has applied multiple, individual PID controllers to pressure swing plants to provide the required control. However, control problems in the field have arisen due to the fact that the plants operate in a batch-like manner and are highly non-linear resulting in PID loop interaction and instability. This paper describes the experimental operation of the oxygen VSA process, with the PID scheme implemented on a pilot scale VSA plant and shows the open and closed-loop response of system purity, flow rate and process pressures to changes in external and internal disturbances. Limitations of the PID scheme are highlighted.

The effects of active components, promoters, and reforming reaction on structure variation of catalysts for carbon dioxide reforming with methane are extensively investigated. It is found that the active component and promoters disperse in different sites on the support. The obvious decrease of the mean micropore size of the samples indicates that the impregnated metal species maybe mainly locate on the inner surface of the micropore of the support and that the reforming active sites also locate in these micropore metal sites. It is also interesting to note that the pore distribution and surface areas of the samples vary after the reforming reaction. The addition of the alkaline oxide promoter prohibits the metal dispersion on the support and inappropriate addition amount of the promoters can result in sharp decrease of the metal dispersion.

In order to meet the challenges of the raw materials becoming heavier, the USY zeolites are further treated by citric acid to enlarge the secondary mesopore amount. The N₂ adsorption isotherms of the samples were measured at 77K, which have been respectively treated by BET equation, t-Plot calculations, BJH method and Horvath-Kawazoe equation. The results show that the pore system of the modification zeolites is developed, especially the mesopores are obviously increased. The micropore distribution is more concentrate.

Infrared spectra are reported of Pt/Al₂O₃ exposed to CO and H₂ at different conditions. Two broad bands appeared in the spectra at 2180 and 2550cm⁻¹ due to the catalyst Pt/Al₂O₃ itself and became stronger while increasing the temperature. Adsorption of CO on Pt/Al₂O₃ gave an infrared band at 2085cm⁻¹, and when H₂ was adsorbed, methane was formed due to reaction of H₂ and the pollutant CO, while increasing the temperature, above 673K for 1 hour or so, when the pollutant CO had almost disappeared during the reaction, a band at 2060cm⁻¹ appeared due to H₂ adsorption on Pt/Al₂O₃.

The usage of the diffuse reflectance infrared spectroscopy (DRIFTS) allows the measurement of the radiation reflected by the sample surface. Moreover, a DRIFTS chamber provides an in-situ control of temperature, pressure and atmosphere compositions. A triphenylmethane dye, Acid Blue 9 (AB9), was taken as the reactant for the photodegradation study. During the reaction, a DRIFTS reactor was used to monitor chemical reactions on the solid surface. Time resolved IR spectrum was first collected and interpreted. Peaks were assigned and then the reaction mechanism was proposed. In this experiment, the structure of the benzene sulfonate is hydrous and therefore, it can be easily attached on the hydroxylated TiO₂ surface. The surface of anatase particles was covered with the OH⁻ and H₂O at room temperature. This type of catalyst when illuminated by the UV light can directly photodegrade adsorbed compounds by holes and hydroxyl radicals. The remarkable decrease was notified in the cyclic conjugated C=N stretching at 1580cm⁻¹, the C-H bending at 1407 and 1390cm⁻¹ from the ethyl group. The bonds of –CH₂– and –C₂H₅ attaching to the nitrogen atom were likely to be attacked by the holes and OH radicals. Consequently, the color of AB9 was found getting fade quickly since the conjugated structures such as C=N and C=C bonds were destructed by the photo-redox process. A significant peak absorbed at 1580cm⁻¹ was chosen as the basis for analysis. Apparently, the dry air system (1.5g/kg absolute humidity) obviously has less decomposition rate than that in the air-equilibrated sample.

Equilibria and kinetics for adsorption of CO₂ on one strongly and two weakly basic bidispersed anion exchange resins have been investigated by breakthrough curve measurements. Experiments were performed under dry and humid conditions. Intraparticle diffusivities were determined by matching the experimental response curves to an appropriate theoretical model. The CO₂ uptakes on the weakly basic adsorbents were more than 5 times higher that that on the strongly basic adsorbent. At a humidity of 20 %, considerably higher equilibrium uptakes compared to the dry state could be observed for all adsorbents. Surface diffusion in the microparticles showed to be the rate controlling transport process in case of the weakly basic adsorbents. For the strongly basic adsorbent, the rate-limiting step could not be identified. However, in this case, the diffusional transport process proved to be much faster than in the other two adsorbents.

A new method is proposed to improve the adsorption properties of basic, macroreticular (MR) anion exchange resins by forming a material composite with chitosan. In this study, one strongly basic and one weakly basic MR resin were used as matrixes for the new composites. Chitosan was accumulated in the macropores of the resins, and breakthrough curves for adsorption of CO₂ on the new adsorbents were measured. The resulting equilibrium and kinetic data were compared with those of the original resins. The composite of strong basic MR resin and chitosan showed a more than 2 times higher adsorption capacity for CO₂ than that of the pure resin. In the same time, the mass transport in the adsorbent slowed down due to a narrowing of the pores by deposited chitosan. In the case of the weak basic MR resin, it seems the accumulated chitosan blocked pores on the surface of the microparticles, resulting in a lower uptake of CO₂ in comparison with the original resin.

The possibility of recovering phosphate ions using a strongly basic ion exchanger was investigated. An OH-type DIAION SA10A was used to measure the adsorption equilibria of H₃PO₄, H₂PO₄⁻, HPO₄²⁻, and PO₄³⁻ ions. The results appeared technically feasible. In all the systems, the experimental adsorption isotherms show a high amount of phosphate adsorption. Moreover, differences in the adsorption behaviors were observed. These differences were highly influenced by the pH of the equilibrated solutions. The ion-exchange reaction models were then proposed. By the mass action law, theoretical equations for the adsorption isotherms were derived and the unknown experimental equilibrium constants were determined. The proposed models were able to correlate the experimental data well and able to explain the peculiarities in the adsorption isotherms.

From the chemical shifts and longitudinal relaxation times obtained by H- and F-NMR measurements at a state of low adsorption, and from the simulation for adsorption isotherms using adsorptive data, it is possible to conclude that the F atom of the adsorbed CHCIF2 molecule locates near Na and the H atom locates far from Na on the surface of the cage wall of NaY-5.6 zeolite. The model obtained here is consistent with the model previously proposed using FT-IR, shown in figure 2. Moreover, the best results of the curve-fitting simulation for the adsorption isotherms are also obtained using Hill's equation.

The selective catalytic reduction (SCR) of NOx was studied at 323-423 K, which is relatively lower temperature than the conventional SCR method by a V₂O₅/TiO₂-anatase catalyst. Transition metal oxides supported on pitch based Active Carbon Fibers (ACF) were investigated as the catalytic component. Three transition metal oxides, Fe₂O₃, CO₂O₃, and Mn₂O₃ were tested as catalytic components, and Mn₂O₃ showed the best activity among the three metal oxides. As support material, ACF showed the highest effect among three support materials including Granular Active Carbons (GAC), andγ-Al₂O₃. The activity of a Mn₂O₃/ACF catalyst increased in proportion to the reaction temperature, and reached 92% as NO conversion at 423 K. ACF contribute to the production of highly dispersed metal oxide particles and provide a large surface area for gas contact.

In this research, comparing the adsorption characteristics of activated carbon and activated carbon with ZnO were studied. The adsorption characteristics, such as adsorption capacity, adsorption rate constant, isothermal curve, and breakthrough curve for a single adsorbate and binary mixtures on adsorbents were investigated. The modified Wheeler equation, IAST, and Nader Vahdat model were used to predict the breakthrough curves of the binary components (MEK and TOL) from single-component adsorption data. From adsorption tests, the adsorption capacities on AC-ZnO increased by 4 - 6 % than the adsorption capacities on AC for a single adsorbate, MEK or toluene. Comparing for the adsorption breakthrough curves of MEK and TOL in the binary system, the breakthrough of the adsorption bed for MEK occurred early. The reason is that MEK is a weak adsorbate in the binary system. When activated carbon was added with ZnO, the total adsorption capacities of adsorbent for VOCs in the binary system were also increased. The breakthrough of the adsorption bed for VOCs on activated carbon occurred slightly earlier than that on activated carbon with ZnO.

The absolute adsorption was formulated by a strategy of isotherm transformation. As consequences, the density of adsorbate in the adsorbed phase was calculated; isotherms with maximum were well modeled; monolayer adsorbate was prompted for the adsorption system studied.

The CO₂ adsorption data may show more than one section in the DR plot if samples had been over-activated. Each section represents a range of pore sizes. The whole DR plot provided an information of the PSD of the sample. This PSD was compared with that obtained from a PSD-oriented isotherm model. The effect of activation condition was discussed basing on the PSD of samples.

Porous structure with interconnected pores of desired dimensions are the prime characteristics of activated carbons. Vegetations or Bio-mass could be prospective precursors for activated carbons having structural integrity since natural precursors have unique fibrous-cellular morphology giving rise to monolithic porous carbons. In our laboratory we have studied pyrolysis behavior of four Biomass materials: banana stem, bagasse, babbool and castor oil plant. The chemical analysis of the raw materials show varying amount of carbon and silica. The silica content varies from 8-30 %. The pyrolysis behavior of these materials has been studied using thermogravimetric analysis which exhibits different pyrolysis regimes depending on the nature of raw material. The carbon yield varies from 16 to 50 %. These materials have been pyrolysed to 650°C. Structure of the chars have been studied by SEM and correlated with the original cell structure. Adsorption behavior of the carbonized materials as such has been studied by liquid adsorption using methylene blue as adsorbate.

Desiccant cooling is a good alternative for air conditioning in tropical climates. It permits to reduce consumption of fossil fuel and of chlorofluorocarbons (CFC). The thermal energy to regenerate the desiccant can be provided from solar energy or heat waste. This paper presents a dynamic study of moist air dehumidification in view of its use in an air conditioning process by evaporative cooling in tropical climates. A reduced prototype has been built in Guadeloupe to study dehumidification of ambient air through a fixed compact bed of silica gel. Desiccant storages were filled with 10 kg of dry silica gel with a mean diameter of 3 mm. Solar plates collectors (7 m²) has been performed with a balancing water tank. The compact bed system working in an intermittent process, gave the foreseen cooling power of 500 W. A rotary compact model has been built and tested with 3.5 kg of desiccant and the specific cooling power was about 5 times higher with a lower regeneration temperature.

The following sections are included:

• Author Index