Mesoporous adsorbents optimized to ensure in situ drying of a separated gas mixture in a VSA unit

Abstract

Use of a mesoporous adsorbent material in a VSA process for separating a gas mixture, said gas mixture comprising, in addition to the main constituents to be separated, water vapor to be stopped on a guard bed prior to the passage of said mixture over the main adsorbent mass, in order to protect said main adsorbent mass from humidity, said adsorbent material being characterized by:- a Henry constant relative to water measured at 30°C of between 6000 and 80000 mmol / bar / g, preferably between 8000 and 50000 mmol / bar / g,- an isosteric heat of adsorption of water in the Henry domain of between 60 and 80 kJ / mol,- an isosteric heat of adsorption of water greater than 40 kJ / mol, preferably between 50 and 65 kJ / mol for a relative humidity greater than 30%,- an average diameter pores between 2 and 8 nm, preferably between 2 and 5 nm. Abstract figure: Fig. 3

Description

Title of the invention: Mesoporous adsorbents optimized to ensure in situ drying of a separated gas mixture in a VS A unit

[0001] The present invention relates to the drying of a gas mixture by a guard bed prior to its separation by the main adsorbent mass into a first fraction comprising the most adsorbable compounds and a second fraction comprising the least adsorbable compounds. It relates to VSA type processes, i.e. adsorption separation processes using vacuum during regeneration. More specifically, it relates to the replacement of adsorbents conventionally used for this purpose for stopping water, namely activated alumina, silica gel, zeolite X, etc., by new, more efficient materials. The production of oxygen from ambient air is taken as an example, but the use of these new materials as a guard bed for stopping humidity is more general than this single application.

[0002] Generally speaking, a method implementing the adsorption phenomenon makes it possible to separate one or more molecules from a mixture containing them, by exploiting the difference in affinity of one or more adsorbents with respect to the constituent molecules of the mixture.

[0003] The processes implementing adsorption are of several types depending on whether the adsorbent is regenerable in situ or not. We can then distinguish 3 main families of adsorption processes: lost charge processes, TSA processes (Temperature Swing Adsorption) and finally PSA processes (Pressure Swing Adsorption).

[0004] In lost charge processes, a new charge is put in place when the one currently in use is saturated by impurities or more generally when it can no longer play its protective role sufficiently.

[0005] Differently, TSA or PSA type processes are cyclic processes continuously linking adsorption and regeneration phases. The cycle time Te is the duration at the end of which an adsorber returns to the same state as previously, for example at the start of an adsorption step.

[0006] In TSA-type processes, the adsorbent at the end of use is regenerated in situ, i.e. the impurities that have been stopped are removed so that the adsorbent recovers most of its adsorption capacity and can start a purification cycle again, the essential regeneration effect being due to a rise in temperature. Regeneration often takes place at low pressure, or possibly even under vacuum, which promotes the desorption of impurities. It is also common to push the heat front present in the adsorber at the beginning of the cooling phase to complete the desorption of impurities from the zone not yet regenerated. In these latter cases, the acronyms TPSA or TEPSA are sometimes found instead of TSA.

[0007] Finally, in PSA-type processes, the adsorbent at the end of the production phase is regenerated by the desorption of impurities obtained by means of a reduction in their partial pressure. This pressure reduction can be obtained by a reduction in the total pressure and / or by flushing with a gas free of or containing few impurities. This gas can in certain cases be external to the unit. Very generally, the regeneration of a PSA is incomplete and the adsorbent mass still contains a significant quantity of adsorbed constituents at the end of this stage. Wanting to eliminate all of these constituents by simply reducing the partial pressure would require the use of a high vacuum and / or very large quantities of elution gas incompatible with economical industrial operation.

[0008] It should be noted that a PSA type unit is not exempt from thermal effects. The adsorption and desorption of impurities locally create heating and cooling whose amplitude (beat) can be up to a few tens of °C when the adsorbed quantity is significant. There may also be a temperature profile between the inlet and the outlet with zones that are hotter and / or colder than the average. These thermal effects generally have a negative impact on regeneration. Pressure swing adsorption processes are used both to remove traces of impurities - for example with a content of less than 1% in the feed gas - and to separate mixtures containing tens of % of different gases. In the first case, we generally speak of purification (for example gas drying) and of fractionation in the second case (for example production of oxygen or nitrogen from atmospheric air).In the most complex cases, we can of course have purification and fractionation in the same unit. It is common to give more specific names depending on the pressure levels involved or the time needed for an adsorber to return to its initial point (cycle time): .

[0009] VSA processes in which adsorption takes place substantially at atmospheric pressure, preferably between 0.95 and 1.25 bar abs and the desorption pressure is lower than atmospheric pressure, typically 50 to 400 mbar abs.

[0010] MPSA or VPSA processes in which adsorption takes place at a high pressure above atmospheric pressure, typically between 1.3 and 6 bar abs, or even 10 bar abs, and desorption at a low pressure below atmospheric pressure, generally between 200 and 750 mbar abs.

[0011] The PSA processes themselves in which the high pressure is substantially higher than atmospheric pressure, typically between 3 and 50 bar abs and the low pressure substantially equal to or higher than atmospheric pressure, generally between 1 and 9 bar abs.

[0012] RPSA (Rapid PSA) processes for which the pressure cycle time is typically less than one minute.

[0013] URPSA (Ultra Rapid PSA) processes for which the duration of the pressure cycle is of the order of a few seconds maximum.

[0014] It can also be noted that the gaseous fraction used in a PSA process can correspond to the fraction produced at high pressure but also to the fraction extracted at low pressure provided that the constituent(s) sought are the most adsorbable in the mixture.

[0015] It should be noted that these various names (TSA, TEPSA, TPS A, RTS A PSA, VS A, VPSA, MPSA, RPSA, URPSA) are not standardized and that the limits are subject to variation depending on the authors. Other names exist in the literature highlighting this or that particular point of the process.

[0016] This document is concerned, according to our definitions, with VSA, MPSA, VPSA used to separate initially humid gas mixtures and comprising, in the same envelope hereinafter called adsorber, an adsorbent mass intended to stop the water located before the main adsorbent mass carrying out the separation of the gas mixture between a first fraction consisting of the most adsorbable constituents and a second fraction consisting of the least adsorbable constituents. The adsorbent recommended here to carry out the drying will be particularly suitable, but in a non-limiting manner, for cycles in which the High Pressure adsorption step and / or the regeneration step is of duration less than or equal to one minute. For simplification, the name VSA will henceforth be given to all of these units (VSA, MPSA, VPSA) concerned by this document.

[0017] The need to stop the water before the gas mixture to be separated enters the main adsorbent mass is a constraint well known to those skilled in the art. It concerns in particular cases where the humidity acts as a poison with respect to the main adsorbent due to its almost irreversible adsorption, as well as cases where it modifies the internal structure by causing for example a reduction in the mechanical resistance of the particles, or even by reducing the stability of the microcrystals themselves. Concerning the adsorbent mass intended here to stop the water and more generally the constituents considered as poisons for the main adsorbent mass, it is referred to indifferently as a guard bed or pretreatment bed.

[0018] It will be generally remembered that to treat by adsorption a gas mixture comprising on the one hand a poison for the main adsorbent, for example of the type acid gas, a constituent capable of polymerizing, reacting by making deposits, adsorbing irreversibly (here the case of H20) and on the other hand the main constituents (for example and in a non-limiting manner H2, CH4, CO, 02, N2) being separated into a more adsorbable fraction and a less adsorbable fraction, it is appropriate to use a guard bed intended to protect the main adsorbent mass, this guard bed being located in the same adsorber as the main bed and upstream of it.

[0019] In this general framework, the present document is only concerned with stopping the water vapor in a guard bed placed upstream of the main adsorbent mass in a process for separating a gas mixture by VS A, VS A being understood in the sense specified above.

[0020] A classic example of this problem concerns VSA 02. It is recalled that an oxygen VSA process is conventionally made up of two bottles each comprising a fixed bed (axial or radial depending on the geometry chosen for the adsorber) with generally an activated alumina bed ensuring the total elimination of water and partial elimination of CO2 and a LiLSX zeolite bed ensuring the N2 / O2 separation to produce a gas with an oxygen purity of the order of 93%. A 2 adsorber VSA cycle generally comprises, in addition to the intermediate balancing-type steps, 2 major steps: - an air supply phase comprising the pressurization phase and the adsorption phase at pressures between 1.1 and 1.5 bar abs - a regeneration phase comprising the depressurization phase and the vacuum elution phase whose minimum pressure can be between 0.5 and 0.35 bar abs at the end of the step.The first adsorbent bed ensuring air pretreatment must avoid contamination of the LiLSX with water. Extremely hydrophilic, a LiLSX zeolite bed exposed to water would see its initial fraction considerably lose its nitrogen adsorption capacity. Similarly, for certain sieves suitable for N2 / O2 separation, exposure to humid air causes a loss of crushing strength which can be critical for the process: the adsorbent could then disintegrate rapidly over time. The pretreatment step is therefore essential for the proper functioning of oxygen production. The pretreatment layer is generally made, as already indicated, of activated alumina (2 mm or 3 mm diameter beads) and / or sometimes a layer of 13X zeolite.

[0021] Without going into details, it can be seen that in general during the adsorption stage, the transfer of gas molecules to the adsorption sites takes place on only a portion of the adsorbent bed. Three zones are then distinguished: the upstream zone called the saturated zone where the adsorbent is completely loaded with constituent, the downstream zone which has not yet been in contact with the constituent in question and which is called the virgin zone and between the two, the zone where the transfer of the constituent contained in the gas takes place towards the adsorption sites that are still partially free, a zone which is indifferently called the adsorption front, material front, frontal zone, mass transfer zone or also called by the acronyms ZTM or MTZ (Mass Transfer Zone). This zone moves in the direction of gas flow. During its co-current movement, it generally has an inverted S shape both when plotting the composition of the constituent in the gas phase or in the adsorbed phase (adsorbate). We generally speak of the end of the front to define the interface between the front and the virgin zone.During regeneration, the adsorbed constituents are pushed back against the current towards the inlet of the adsorber and, in a somewhat similar way, for a given constituent, one end of the front is obtained at the end of regeneration, the end upstream of which (by reference to the direction of circulation of the gases in the adsorber during the regeneration stage) the constituent in question has been completely desorbed. Stable operation of the PSA corresponds to operation for which the extreme positions of the adsorption and regeneration fronts no longer change from one cycle to the next. There is then neither accumulation nor loss of molecules in the adsorbent mass over a complete cycle and what enters with the feed gas comes out in the production or the waste of the treatment unit.

[0022] It is then understood that the efficiency of stopping a constituent of an adsorbent mass will depend both on its efficiency in stopping this constituent during the adsorption phase and on its efficiency in desorbing during the regeneration step.

[0023] The efficiency of the regeneration is characterized in particular by what is called the regeneration power (PF) of the PSA which is defined by the ratio of the gas volumes passing countercurrently through the adsorbent mass during a cycle compared to the gas volumes passing said mass cocurrently, the respective volumes being counted in real m3. By regeneration power (PF), we mean here its most complete definition, corresponding to the theoretical “Purge Factor”, that is to say that it is determined locally (here for example in the pretreatment adsorbent mass) and takes into account all the gas flows circulating at this level during all the stages of the cycle. For information, there is another global, simpler definition which only takes into account the feed gas during the adsorption stage (for the cocurrent) and the elution gas during the elution stage (for the countercurrent).This is not the definition used here, but depending on the cycles considered for the PSAs, this simplification can actually be a good approximation of the more rigorous formulation. The regeneration power (PF) must be greater than 1 and is often between 1.1 and 1.3 for a majority of PSAs.

[0024] It will be noted that an increase in the regeneration power will generally make it possible to reduce the adsorbent masses to be used and therefore the investment but to the detriment of the performance of the unit in terms of consumption. energy, extraction efficiency. Given this, the effective regeneration power is that required for - or optimized for - the main separation. The guard adsorbent mass, that corresponding to the water stop in the example cited, must then be sized according to this regeneration power. A useful adsorption capacity for this adsorbent mass is sometimes defined as equal to the quantity of constituent actually stopped per cycle per unit of adsorbent mass and also a useful capacity ratio by comparing this effective stopping capacity to the quantity of constituent likely to be theoretically stopped at saturation under the operating conditions. This generally low ratio, for example of the order of 30%, takes into account in particular the fact that the adsorbent is not completely regenerated as already explained as well as the fact that the mass transfer zone is by definition not saturated.It depends in a very complex way on many parameters characteristic of adsorption and desorption. Without going into details, we will remember that often what is favorable to adsorption is unfavorable to desorption, that is to say to regeneration and vice versa. For example, an isotherm of "favorable" shape with a downward curvature tends to shorten the frontal mass in the adsorption phase but to spread the regeneration or a strong affinity between constituent and adsorbent facilitates adsorption but slows down desorption. On the contrary, high kinetics will be able to improve both aspects, adsorption and regeneration, and is an important parameter to control.

[0025] It has been observed from experience since the commissioning of the first VS A comprising a guard bed to stop the humidity integrated in the adsorber, that the useful capacity of the adsorbent masses used for this purpose, whether activated alumina, zeolite X or A, silica gel or multi-beds formed from these adsorbents was particularly low and well below the usual values ​​that could be expected. This still results today in guard beds of large volume. For example, the volume of activated alumina or zeolite 13X in a VS A 02 represents practically 30% of the volume of LiLSX whereas it is only a question of stopping traces of water with adsorbents that are a priori very efficient for this purpose instead of carrying out, as on the main adsorbent, an O2 / N2 separation known to be difficult and covering almost the entire flow rate.

[0026] This low efficiency of the guard bed results in a large dead volume containing compressed air for nothing in the adsorption stage and nitrogen not evacuated at the end of regeneration. This has a negative impact on the oxygen extraction yield and therefore on energy consumption. In addition, this adsorbent mass located at the inlet of the adsorber creates additional pressure losses which degrade the specific energy of the main O2 / N2 separation.

[0027] There is therefore a definite interest in finding alternative adsorbents to activated alumina, silica gel, zeolite X or A in order to more effectively dry a gas mixture intended to be separated in a VS A type unit.

[0028] A solution according to the invention consists of using a mesoporous silica-based adsorbent material in a VS A process for separating a gas mixture, said gas mixture comprising, in addition to the main constituents to be separated, water vapor to be stopped on a guard bed prior to the passage of said mixture over the main adsorbent mass, in order to protect said main adsorbent mass from humidity, said adsorbent material being characterized by: - a Henry constant relative to water measured at 30°C between 6000 and 80000 mmole / bar / g, preferably between 8000 and 50000 mmole / bar / g, - an isosteric heat of adsorption of water in the Henry domain between 60 and 80 kJ / mole, - an isosteric heat of adsorption of water greater than 40 kJ / mole, preferably between 50 and 65 kJ / mole for a relative humidity greater than 30%, - an average pore diameter between 2 and 8 nm, preferably between 2 and 5 nm (20 and 50 angstroms), these characteristics relating to the powder (micro crystals) resulting from the synthesis of the adsorbent material, before shaping from this powder and by adding a binder of an adsorbent in an agglomerated form (agglomerated adsorbent).

[0029] According to a variant, the adsorbent still considered in powder form (micro crystals) also has at least one of the following characteristics:

[0030] - a specific surface area greater than 600 m2 / g, preferably included between 900 and 1200 m2 / g,

[0031] - a mesoporous volume of between 0.6 and 1.0 cm3 / g,

[0032] - micro crystals of dimensions ranging from 0.1 to 50 microns, plus generally between 1 and 20 microns.

[0033] According to a variant, the agglomerated adsorbent, formed from the powder resulting from the synthesis to which a binder and additives are added, is in the form of balls, extrudates, rods, grains or pellets, this agglomerated adsorbent then having at least one of the following characteristics:

[0034] - presence of a binder in proportions of 5 to 30% by weight of the mass of adsorbent material (powder, micro crystals), this binder being essentially inert or partially zeolitized,

[0035] - an average diameter (balls, grains, etc.) or a length, i.e. the longest large dimension (sticks, extruded...) between 0.2 and 10mm, preferably between 0.5 and 5mm,

[0036] - a macropore porosity plus a mesopore porosity representing more than 70%, of preferably more than 80%, even more preferably more than 90% of the total porosity of the particle.

[0037] According to one variant, the adsorbent material is a silica-based mesoporous material (generally referred to in publications as “silica-based mesoporous” or “silicate mesoporous” material).

[0038] According to one variant, the adsorbent material is a mesoporous material obtained by reaction between an alkaline solution of quaternary alkylammoniums (surfactants) and silicates then, after reaction, by elimination of the surfactant.

[0039] According to one variant, the adsorbent material undergoes aluminum grafting and cationic doping, in particular with sodium, magnesium, potassium, calcium, lithium, barium or strontium cations.

[0040] According to one variant, the internal surface of the pores of the adsorbent material is functionalized by means of amine, carboxylic or phosphate groups.

[0041] The inventors observed by studying the water content profiles in the guard adsorbent mass installed upstream of the main LiLSX layer in a MPSA 02 type pilot unit representative of an industrial operation that these profiles did not correspond to the academic description previously made. This means that the classic configuration described above of a well-defined mass transfer zone advancing during the adsorption step and retreating during the regeneration step was not found. Instead, a wet adsorbent mass was observed changing very little during a complete cycle. The local quantity of water adsorbed then desorbed is very low, explaining the surprisingly high volume necessary to stop all the water upstream of the main bed. This phenomenon is true for both a guard bed composed of activated alumina or zeolite X even if the profiles are not similar.

[0042] It should be noted that the fact of having to install a large guard adsorbent mass, greater than what could be estimated, has been known since the very beginning of the commissioning of VSA 02. Such oversizing was for example already recommended by the Bayer Company to protect the CaX type adsorbents proposed for this application in the 1980s. These dimensions, which effectively made it possible in practice to protect the main adsorbent from any entry of humidity, have not fundamentally been called into question for various more or less justified reasons:

[0043] - low cost of activated alumina and then zeolite X compared to the adsorbent main which has always been an exchanged zeolite of more complex manufacture and implementing expensive products (lithium) with ecological constraints,

[0044] - secondary energy consumption taking into account initial productions relatively low oxygen levels of a few tens of t / d maximum.

[0045] Although the impact of the guard bed on performance has increased due to significant improvements in the main adsorbent (guard bed / main bed ratio now significantly higher), the use of small diameter activated alumina particles in response to the shortening of the cycle creating additional pressure drops, it seems that there have been no specific studies on this point despite the implementation of large capacity production units for which energy consumption was becoming an important factor. Optimization studies have focused very essentially on the performance of the main adsorbent, on the cycles implemented, on the machines used in the process and on energy management in reduced operation.

[0046] In order to search for adsorbents which would be more favorable than those conventionally used for the guard volume, the inventors focused on two points.

[0047] The first is to increase the adsorption and desorption kinetics at a given particle size. Laboratory tests show that kinetics is a sensitive parameter on the shape of the humidity profiles in a guard bed when rapid cycles are implemented. These results can be compared to those known in thermal on the operation of regenerators: longitudinal displacement between entry and exit of well-marked temperature fronts or simple local thermal respiration as a function of the duration of the heating / cooling steps. The first mode of operation with the advance and retreat of the fronts corresponds to the description given above of the movements of the frontal mass in adsorption then in regeneration. This can be understood in view of the similarities between the equations governing thermal or adsorption.It is then logical to attribute to insufficient kinetics in relation to the duration of the stages at least part of the phenomenon observed in the guard beds of the VSA.

[0048] By similarity, it appears that it is appropriate to obtain a substantial gain on this parameter to hope for a significant effect on the shape of the adsorption and regeneration profiles and thereby on the efficiency of the guard bed, a gain which could not be obtained by a simple improvement in the macroporosity of a conventional particle taking into account the progress already made in this field by adsorbent suppliers.

[0049] The second point of improvement retained concerns the optimal isotherm that the guard bed adsorbent should have. This is generally a theoretical problem often raised and generally resulting in hypothetical adsorbents having very high adsorption capacities, selectivities, and kinetics depending on the cycles involved.

[0050] [Fig.l] [Fig.l] represents the quantity adsorbed in % by weight, as a function of the relative humidity;

[0051] [Fig.2] [Fig.2] represents the pore or micelle diameter (in nanometers) in function of the number of carbon atoms in the alkyl chain of the surfactant; and

[0052] [Fig.3] [Fig.3] represents the effect of cations on the water adsorption isotherm of a mesoporous silica-based adsorbent.

[0053] A more empirical approach consisted in making the hypothesis supported both by the shape of the water profiles in the current guard adsorbent masses and the isotherms represented in [Fig.l] that zeolite X had too high an affinity for water (resulting in a Henry constant and possibly an isosteric heat of sorption / desorption that were too high) making desorption difficult in the operating conditions of a VS A, whereas activated alumina had a Henry constant that was on the contrary too low and an isotherm of unfavorable shape (final curvature upwards) to be effective on short cycles of the order of a minute. The hysteresis phenomenon (quantity of water at equilibrium higher in regeneration than in adsorption) is also an unfavorable effect for alumina. It is largely linked to the size of the pores.

[0054] Typical adsorption isotherms of water on activated alumina and on zeolite X at room temperature are reproduced in [Fig.l]. It should be remembered that each adsorbent will have its own isotherm with some variations around its typical representation depending on the basic products, the manufacturing procedures, and even the degree of activation before measurement. Nevertheless, the conclusions of the comparison made cannot be called into question.

[0055] The figure on the left shows the entire isotherm for a relative humidity of 0% to 100% (saturation) on the abscissa, the ordinate corresponding to the quantities adsorbed in % weight. The figure on the right is a zoom on the zone of low relative humidity which in practice will correspond to the stopping of the last ppm of water contained in the gas mixture. We recall that the Henry constant is the slope at the origin of the isotherms and that the isosteric heat corresponds to the heat released by adsorption. The latter generally varies by a few tens of % depending on whether we are interested in the adsorption of the very first molecules in a monolayer on the wall (we then speak of isosteric heat in the Henry zone, therefore in the case where the water vapor pressure - or Relative Humidity - tends towards 0) or in the adsorption of molecules while the pore walls are already completely lined by a first layer, which corresponds to so-called multilayer adsorption.Methods for determining Henry's constant and isosteric heats are described later in the document.

[0056] This [Fig.l] highlights the great difference in behavior of activated alumina and zeolite X with respect to water adsorption.

[0057] From the values ​​in the literature and internal databases, Henry's constants were calculated for a large number of activated aluminas, silica gels and zeolites, in particular zeolite X, from various suppliers. As expected by the person skilled in the art, the results show, as indicated above, a certain dispersion due essentially to the material itself (raw materials, manufacturing processes), to its activation and partly to the precision of the available data. Since these are often measurements carried out on zeolite particles, a correction has been made to take into account a percentage of inert binder.

[0058] However, the very different ranges of Henry's constant between activated aluminas and silica gels on the one hand and zeolites X on the other hand are clearly seen. To visualize these differences, at this stage, a Henry's constant normalized to 1 was used for the lowest values ​​of Henry's constants of standard activated aluminas and for a temperature of 30°C.

[0059] We then obtain ranges of Henry's constant thus normalized from 1 to 3 for standard activated alumina, and from 1 to 2 for standard silica gel, but from 100 to 500 times more, and even more, for zeolite X. These values ​​reinforce the hypothesis that we are currently using either adsorbents that are not very effective in stopping traces of water, or adsorbents that have a very high affinity for water and are therefore not very regenerable in rapid VSA cycles. It is therefore very likely that an intermediate material with a normalized Henry's constant of between approximately 5 and 60 will have better efficiency with sufficient affinity for water in adsorption while being easily regenerable.

[0060] By standard activated alumina, standard silica gel, zeolite X, we mean the products marketed in large quantities for this type of drying application (VSA 02, purification at the head of a cold box, in particular cryogenic air separation) and not specific or laboratory products.

[0061] The determination of Henry's constant is known to those skilled in the art even if there are still developments on this subject and some differences in the procedures for obtaining it.

[0062] Here, we recommend the method described in FR 3 005 873, the main points of which are recalled below.

[0063] The Henry constant of a gas for a given adsorbent is defined as already indicated as the slope at the origin of the curve representing the quantity of gas adsorbed at a given temperature as a function of the adsorption pressure or, in the case of water as a function of the relative humidity, when this pressure or this relative humidity therefore tends towards 0. As it is not possible to carry out measurements under the strict assumption of zero pressure, a general procedure has been established to evaluate this value from a series of measurements approaching this assumption while remaining within the measurable range. It should be noted that the initial state of the adsorbents strongly impacts the quantity of adsorbed gases as well as the Henry constant. For this reason, the measurement of adsorption isotherms and the estimation of the Henry constant of a given gas are carried out after the activation or total regeneration of the adsorbent. The regeneration conditions depend strongly on the nature of the adsorbent and generally the samples are regenerated between 200°C and 450°C, depending on the type of adsorbents, under secondary vacuum (10-5 mbar) for 8 hours.About twenty adsorption points of the constituent under study, regularly spaced, are determined using the most suitable measuring devices, particularly for the partial pressure range of interest. A mathematical model is determined to link the quantities of adsorbed gas and the pressure, a model preferably inspired by well-known adsorption models and whose parameters are determined via a mathematical solver by minimizing the relative error between the measured values ​​and the calculated values. The slope at the origin of the equation thus obtained gives the value of Henry's constant. Also known methods make it possible to validate or determine the uncertainty on the retained value (robustness of the result in the case of adding measurement points, testing with different models).

[0064] With the determination of the Henry constant by this method, a mesoporous silica-based adsorbent material is then selected, characterized by a Henry constant relative to H2O measured at 30°C of between 6000 and 80000 mmole / bar / g, preferably between 8000 and 50000 mmole / bar / g, corresponding to the range allowing for better initial adsorption than with activated alumina and better regeneration than with a zeolite X.

[0065] For the record, the Henry constants of standard aluminas and silica gels are around 1500 to 5000 mmole / bar / g while those of zeolites X are of the order of 250000 mmole / bar / g and often well beyond. As a first approximation, we can say that we are retaining an adsorbent whose Henry constant for water will be 2 to 3 times higher than that of a standard activated alumina and 5 times lower than a non-exchanged zeolite X (Na cation).

[0066] The variations in the initial isosteric heat of adsorption of water between the different adsorbents are much less than those relating to the Henry constant and vary in a range from about 50 to 90 KJ / mole, the minimum values ​​generally corresponding to the activated alumina and silica gel group and the higher values ​​to zeolite X.

[0067] Thus, with the same concern for compromise between adsorption and regeneration, a material with an initial isosteric heat of sorption, i.e. in the Henry domain, between 60 and 80 Kj / mole, will be selected. Outside the Henry domain, in particular for a relative humidity RH greater than 30%, a material with an isosteric heat greater than 40 kj / mole, preferably between 50 and 65 kj / mole, will be retained.

[0068] It is recalled that isosteric heat is the energy released by the action of intermolecular forces of attraction between water molecules and the surface of the pores (or required to break these same forces during regeneration). These forces must be sufficient and not excessive both at very low partial pressures corresponding to the stopping of the last traces of water in the gas mixture, as well as at partial pressures or relative humidities close to those of the inlet gas mixture.

[0069] A person skilled in the art knows how to determine the isosteric heat of sorption at a given coverage of the surface of the adsorbent material. Without wishing to go into the details of the equations, it can be calculated from the experimental measurement of the adsorption isotherms by applying the Van't Hoff equation or from the Clausius-Clapeyron equation by adopting the hypothesis that the gas phase behaves like an ideal phase and that the volume of the adsorbed phase is negligible compared to that of the gas phase. It is this latter method, known for correctly estimating the isosteric heat of adsorption as a function of the loading of the surface of the material, for a pure body, which is used here, in particular within the framework of the values ​​retained for the invention. The isosteric heat in the Henry zone corresponds to the coverage by a monolayer of molecules.The initial value, i.e. when the load tends towards zero, can be estimated by a method identical to that used for Henry's constant, namely the limit value of the function correlating the values ​​determined as a function of relative humidity. The variations in isosteric heat are relatively small and therefore this method is particularly well suited.

[0070] As has already been pointed out, a significant increase in kinetics can only be obtained by a change in the nature of the porosity, i.e. by moving towards mesoporous materials having a majority of pores of a size greater than or equal to 2 nanometers favorable to the circulation of molecules.

[0071] More precisely, it can be estimated that an optimum between kinetics and adsorption capacity concerns mesoporous materials having pores of relatively small diameter in the range covered by this name (mesopore ranging from 2 to 50 nm) with a significant impact on the kinetics but having sufficient internal surface area. It is recalled that for a given pore volume, crystals having pores of smaller diameter develop more internal surface area. The adsorbent selected will therefore have a porosity in the range of 2 to 8 nm, preferably 2 to 5 nm, i.e. 20 to 50 Angstroms, i.e. at least twice the size of current adsorbents. At the low pressures used in VSAs, this particularly modifies the diffusion mechanisms and their relative importance.

[0072] It is recalled that nanoporous materials consist of an amorphous or crystalline structure with empty spaces, which can be essentially cylindrical or cage-like. Most nanoporous materials belong to three broad categories: microporous, mesoporous and macroporous.

[0073] Microporous materials—such as MOFs, zeolites, activated alumina, the majority of activated carbons, silica gels, some amorphous glasses—have narrow pore size distributions in the range of 0.5 to 2 nm, offering significant affinities to gas adsorption due to the confinement effect but at the same time presenting limits to the diffusion of species in these pores. In contrast, macroporous materials with pore sizes between 50 and 1000 nm, such as porous polymer beads, allow easy access to the internal pores at the expense of adsorption capacity and selectivity. These drawbacks have led in recent years to the development of mesoporous materials, having an intermediate pore size range, between 2 and 50 nm.This class of materials is characterized by a narrow pore size distribution, relatively high specific surface areas (typically greater than 600 m2 / g) and the possibility of performing structure / wall modifications with functionalization strategies to modulate the energy of the interaction of the pore walls with gas molecules. These strategies include the introduction of cations into the structure and / or the grafting of functional groups, both methods allowing the modification of the electric fields at the wall.

[0074] Several pore topologies are possible for mesoporous materials. Cylindrical topologies, such as MCM-48, AMS-6, MCM-41, SBA-15 and NFM-1, have a uniform pore diameter and have application potential in catalysis and adsorption. In contrast, cage-like solids, such as FDU-1, SBA-1 and AMS-8, consist of spherical or ellipsoidal cages that are connected in 3D by smaller connection windows or cages and are used to control the mass transfer of active agents such as in drugs.

[0075] The most developed synthesis of mesoporous materials is done by mixing alkaline solutions of quaternary alkylammoniums (surfactants) and silicates. The resulting nanomaterials are similar to liquid crystals or assemblies of micellars. After removal of the surfactant, generally by calcination, the inorganic phase has porosity with variable pore sizes but determined, in particular between 2 and 5nm, these different porosities can be created using surfactants of variable chain length as indicated in [Fig.2] where the diameter of the porosities before calcination (black dots and black straight line in solid line) and after calcination (white dots surrounded by black and black dotted line) shown on the ordinate is a linear function of the number of atoms of the surfactant.

[0076] High alkalinity syntheses produce materials with wall thicknesses as low as 0.5 nm, corresponding to a silica bilayer. After removal of the surfactant, also called a "structuring element" or "template," the material exposes a highly reactive surface that will tend to rearrange to a more stable configuration by migration and recondensation of silicates. This rearrangement, which results in a decrease in specific surface area and a broadening of the pore size distribution, is enhanced by increasing temperature and the presence of water vapor. A stable configuration is obtained for a wall thickness between 1 and 2 nm.

[0077] As an example, one of the most studied mesoporous materials is MCM-41.

[0078] Its purely silicic network is electronically neutral following the charge (+4) of the Si and four (- 1) charges of oxygen atoms. However, the substitution of another element such as aluminum atom affects the charge density of the structure. As a result, purely silicic MCM-41 loses neutrality when Si +4 cations are replaced by Al+3 cations and the whole becomes consequently negatively charged. The distribution of silicon and aluminum tetrahedral elements in the framework is generally governed by Loewenstein's rule. Thus, according to this principle, Al-MCM-41 consists of alternating silicon and aluminum tetrahedra which imposes an overall negative charge. The negatively charged lattice is balanced by charge compensating cations like Na+, Mg++, K+, Ca++, Ba++, Li+, Sr++ in the system.

[0079] As regards mechanical strength, the crushing mechanism of, for example, MCM-41s subjected to uniaxial pressure is in all respects similar to that of macroscopic honeycomb structures. In the case of well-structured solids, with wall thicknesses greater than 1 nm, the measured yield strength - which varies with the square of the ratio between the wall thickness and the pore size - corresponds to that of the microcrystals of usual adsorbents. By applying the shaping procedures known and used for zeolites, for example, it will be possible to obtain particles of millimetric dimension that are perfectly usable in adsorption processes, in particular in VSAs. These agglomeration procedures intended to obtain, for example, beads or rods are discussed below.

[0080] More generally, there are families of mesoporous materials other than MCM-41 also having uniform cylindrical pores or regular geometric section of the hexagonal type, for example of adjustable size, also having the possibility of incorporation by post-treatment of heteroatoms such as aluminum in the structure of the silica which implies as we have seen the addition of charge compensating cations whose specificities (charge / ionic radius) will make it possible to modulate the energy of the interaction of the water molecules with the walls of the pores of the materials. More precisely, the cationic doping of a mesoporous material which has undergone aluminum grafting allows the modulation of the Henry constant of adsorption of water in this material.Selective cationic doping can in fact lead in particular to an increase in the water adsorption capacity at low relative humidity, i.e. to obtain a Henry constant higher than that of activated alumina or silica gel.

[0081] [Fig.3] taken from literature data is only intended to illustrate the effect of different cations (Li, Na, Mg) on ​​the shape of the isotherm and in particular at low partial pressure on an MCM type material and does not constitute a limiting selection of materials for the application envisaged in this document. This effect of the nature of the cations on the adsorption of water is not limited to this particular material but is more general with more or less marked effects.

[0082] Indeed, by choosing a mesoporous material as a guard bed, a large number of levers are available to obtain the adsorbent with the right adsorption / desorption compromise, based in particular on the values ​​of the Henry constant and the isosteric heats and a largely improved kinetics. These levers concern the initial synthesis which, after removal of the template, leads to the basic material (composition, shape, volume and dimension of the pores, etc.), the surface density of the aluminums grafted to the surface of the pores, the nature of the charge compensating cations to modulate the interaction energy (isosteric heat at the Henry region and the Henry constant KH) as well as possibly the degree of functionalization of the internal surface of the pores with charged groups.

[0083] In order to be effective, among the large number of possible materials having the chosen characteristics (Henry's constant, isosteric heats, pore diameter), the adsorbent material selected will have a high specific surface area at least of the order of conventional adsorbents, i.e. greater than 600 m2 / g and preferably between 900 and 1200 m2 / g. The technique used to determine the specific surface area of ​​an adsorbent is well known to those skilled in the art (BET method). Similarly, it will have a high pore volume greater than the majority of conventional adsorbents and preferably between 0.6 and 1 cm3 / g. The mesoporous volume is determined by the Barrett-Joyner-Halenda method (known as the BJH method).

[0084] The materials being produced in powder form with micro crystals of dimensions ranging from 0.1 to 50 microns, more generally between 1 and 20 microns with variable size distributions depending on the synthesis, it is appropriate to make them into millimetric type particles which can be used in industrial processes. The agglomeration processes, of the type used for conventional zeolites, consist of forming a paste using a suitable binder, generally based on clay such as kaolin, halloysite, attapulgite, sepiolite, montmorillonite, bentonite...one or more additives can also be added. The additives are preferably organic, for example methyl celluloses, polycarboxylic acids, surfactant molecules..., intended to facilitate the handling of the paste formed by the micro crystals and the binder by modifying the rheology, the adhesive power or to give the final agglomerates satisfactory properties, in particular macroporosity and resistance to crushing. The porosity relative to the binder should obviously not compromise the gains provided by the mesoporous material. Newly developed high kinetic binders containing, for example, polymers can be used for this purpose. The measurement of the mechanical resistance to crushing of the particles is conventionally determined with devices developed for this purpose, for example, the “Grain Crushing strength” type marketed by Vinci Technologies, according to the corresponding ASTM standards or equivalent procedures.

[0085] The agglomerated form of the material, formed from the powder resulting from the synthesis, the binder and the additives as usable in a VS A process can be in the form of balls, extrudates, rods, grains, pellets, etc. These particles then have the following characteristics: - a binder in proportions of 5 to 30% by weight of the mass of adsorbent material (powder, micro crystals), this binder being essentially inert or partially zeolitized, - an average diameter (balls, grains, etc.) or a length, i.e. the largest dimension (rods, etc.) between 0.2 and 10 mm, preferably between 0.5 and 5 mm, - a macropore porosity plus a mesopore porosity representing more than 70%, preferably more than 80%, even more preferably more than 90% of the total porosity of the particle.

[0086] It is recalled here that there are several methods for determining pore volumes and pore sizes, from simple intrusion procedures (mercury, N2, Ar, He, etc.) and their associated treatments (BET, BJH) to sophisticated methods such as Density Functional Theory (DFT). which allows for example to go back to the size and distribution of mesopores, or possibly even with software which allows to define in detail the characteristics of the crystalline structure. Added to this is direct observation via electron microscopes (SEM, TEM) associated with image processing.

[0087] Mesopores are characterized here by their diameter. While many materials do indeed have cylindrical pores, there are some with different geometries, such as hexagons. Here, the pore section is considered to be a circle of the same surface area, and in these cases, we also speak of diameter. This is a common assimilation (cylindrical pore model) which also gives a result close to that of the circle whose circumference would be equal to the perimeter of the polygon.

[0088] In addition to the description of the material recommended for stopping water in a VSA, it is interesting to address the means currently available to those skilled in the art, on the one hand, to restrict the number of candidates and, on the other hand, to select the one or ones corresponding to the best compromise between adsorption and regeneration, this latter selection being a priori dependent on the VSA process envisaged (high pressure, low pressure, temperature, cycle, regeneration power).

[0089] Molecular modeling / simulation, which is the set of techniques used to simulate the behavior of atoms, ions, molecules in a given environment, is increasingly used when the actual experiment is costly, long or dangerous. Here, it allows a screening of the hundreds of theoretical possibilities resulting from the potential ranges of composition, dimension, post-treatment and makes it possible to obtain a theoretical isotherm for the materials meeting the selected criteria. This type of approach is, for example, used for the study of MOFs which also include a very large number of possible arrangements.

[0090] From the moment when isotherms are established for materials having a Henry constant, isosteric heats, pore dimensions entering into the criteria defined in this document, the latter are tested for the envisaged application via adsorption process simulation software. The person skilled in the art actually uses today dynamic simulation software for cyclic adsorption processes. This simulation software is based on a division in the space of the adsorber (into slices called meshes) and in the time of the cycle and carries out successive sequential calculations until a stable state is found verifying material and thermal balances as well as all the imposed constraints.

[0091] The joint use of molecular modeling and dynamic simulation of adsorption processes makes it possible a priori to determine the few best candidates to test for the envisaged application and to considerably limit the sample manufacturing and the number of pilot tests. The screening of potential materials will begin with the structure corresponding to the family of MCMs having undergone aluminum grafting and cationic doping, a structure which can lead to adsorbents corresponding to the selection criteria.

[0092] For economic reasons, it will most likely be appropriate to retain a single material suitable for the different VSA cycles, at least for a given application (VSA 02, CO, CO2, CH4, etc.).

Claims

Claims

1. Use of a mesoporous silica-based adsorbent material in a VS A process for separating a gas mixture, said gas mixture comprising, in addition to the main constituents to be separated, water vapor to be stopped on a guard bed prior to the passage of said mixture over the main adsorbent mass, in order to protect said main adsorbent mass from humidity, said adsorbent material being characterized by: - ​​a Henry constant relative to water measured at 30°C of between 6000 and 80000 mmol / bar / g, preferably between 8000 and 50000 mmol / bar / g, - an isosteric heat of adsorption of water in the Henry range of between 60 and 80 kJ / mol, - an isosteric heat of adsorption of water greater than 40 kJ / mol, preferably between 50 and 65 kJ / mol, for a relative humidity greater than 30%, - an average pore diameter between 2 and 8 nm, preferably between 2 and 5 nm,these characteristics concerning the powder resulting from the synthesis of the adsorbent material, before shaping from this powder and by adding a binder of an adsorbent in an agglomerated form.,

Images