HU205057B - Process for the adsorption separation of paraalkyl aromatic hydrocarbons by using fluoroaromatic adsorbent - Google Patents

Abstract

Sepn of p-dialkylaromatic hydrocarbons from mixts. with other isomers by effected by (a) contacting the mixt. with a Ba- and K-exchanged zeolite, esp. having a BaO/K2O molar ratio of 0.6-1.2, to adsorb the p-isomer and (b) desorbing the p-isomer with m- and/or o-difluorobenzene. - The zeolite is an X or Y zeolite with a BaO/K2O ratio of 0.9-1.1, and may be composited with an amorphous inorganic oxide. The process may be operated using a simulated moving-bed flow system (esp. countercurrent). Desorption is effected at 20-250 deg.C and a pressure sufficient to maintain liq. phase.

Description

The present invention relates to a process for the adsorption of hydrocarbons. The process comprises separating a dialkyl-substituted aromatic para-isomeric hydrocarbon from a nutrient mixture containing at least one more such dialkyl-substituted aromatic isomeric hydrocarbon. The process primarily separates para-xylene from a nutrient medium containing at least one additional xylene isomer in addition to para-xylene. The adsorbent used in the process is mainly zeolite and a certain fluorine-substituted aromatic hydrocarbon desorbent.

It is well known in the art of separation that crystalline aluminosilicates can be used to separate one hydrocarbon from another type of hydrocarbon. For example, the separation of normal paraffins from branched paraffins can be accomplished with 30,000-50,000 µm pore aperture type A zeolites. Such a separation process is described in U.S. Patent No. 2985589, 3, 201491. These adsorbents are separated on the basis of the physical size difference between the molecules such that smaller or normal hydrocarbons enter the cavities of the zeolite adsorbent and the larger or branched molecules remain outside.

The adsorbents used in the process for separating the various types of hydrocarbons may be either X or Y zeolites. For example, Nos. 3,626,020 and 3,663,638, 3,665,046, 3,668,666, 3,686,342, 3,700,744, 3,773,474, 3,897,109, 3,997,620 and U.S. Pat. No. 3,426,274 employs primarily zeolytic adsorbents to separate dialkyl substituted monocyclic aromatic para-hydrocarbons from other isomers, in particular para-xylene, from other xylene isomers.

U.S. Patent 3,707,550 discloses the use of, inter alia, silver or silver and potassium ion-exchanged X-zeolite for the separation of C8 aromatic isomers, but this separation is carried out only after conventional distillation in the ortho-xylene "separation zone".

Most of the processes described in the above patents use benzene, toluene or p-diethylbenzene as the desorbent. It has been previously recognized that the use of benzene as a low-boiling desorbent is detrimental to contact with the adsorbent (see U.S. Pat. No. 3,686,342, supra). When toluene is used as a desorbent for the separation of xylene isomers, the extract and raffinate can only be recovered at a loss because the boiling point of toluene is too close to the boiling point of the isomers to be separated. Therefore, it is desirable to use a material with a lower boiling point which satisfies the desirability of desorbent. Another part of the above patents discloses a process in which the separation of para-xylene from other xylene isomers utilizes a certain proportion of barium and potassium ion-exchanged zeolite. See, for example, U.S. Patent Nos. 3,663,638,3878,127, 3,878,129, 3,686,342, and 3,558,732. Japanese Patent Application No. 6,193,115 discloses the use of x-type zeolites, which have been ion-exchanged with metals such as barium, calcium, sodium, for the separation of aromatic isomers. Depending on the raw material mixture, the desorbent liquid may be benzene, toluene, chlorobenzene, 1-fluoroalkane, α, ω-dihaloalkane, halogen-substituted benzene, halogen-substituted toluene, e.g. benzene xylenes and / or ethylbenzenes, toluene trimethylbenzene ethyl toluenes and chlorobenzene to separate halogenated aromatic isomer mixtures.

U.S. Patent No. 4,584,424 discloses the separation of ethyl benzene from xylenes by adsorption on beta-zeolite. Although the preferred desorbent for this separation is para-diethylbenzene, the use of monohalobenzene, especially iodobenzene, is described herein.

It is an object of the present invention to provide a process for the separation of dialkyl-substituted aromatic para-isomeric hydrocarbons from a nutrient mixture containing at least one additional aromatic hydrocarbon isomer in addition to the para-isomer. In one preferred process, para-xylene is separated from the medium containing at least one other xylene isomer or ethyl benzene in addition to para-xylene. In the process, the medium is adsorbed under adsorption conditions to a crystalline aluminosilicate-containing adsorbent, a Y-type zeolite that contains Ba and K cations (0.6: 1) to (1.2: 1) on the exchangeable cation side of the adsorbent. ) preferably in an amount of 0.9: 1.1 molar ratio of barium oxide to potassium oxide. The resulting parisomer-containing adsorbent is then derivatized and the para-isomer is removed under desorption conditions with a suitable desorbent, ortho- or paradifluoro-substituted benzene, or a mixture thereof.

This description describes in detail the nutrient mixtures, the process flow diagram and the operating conditions.

The chromatogram shown in Figure 1A depicts p-xylene separated from xylene isomers and ethylbenzene by Bα and meta-difluorobenzene desorbent.

Figures IB, IC show a chromatogram similar to Fig. IA, showing p-xylene separated under different conditions and at different adsorbent ion concentrations. For easier evaluation of Figures ΙΑ, IB, IC, 1D and 5A, 5B and 6A, no traces of ortho- and meta-xylene are present, otherwise these figures would be similar to Figures 2 and 3. Figures 1A-1D show separation of p-xylene at molar ratios of 0.63, 0.71, 0.98 and 5.17 BaO and K ₂ O, respectively.

The separation of p-xylene shown in Figure 2 was similar to the separation of p-xylene shown in Figure 1A, except that the molar ratio of BaO to K ₂ O was 1.2. The figure shows all the isomers present.

HU 205 057 Β

The separation of p-xylene in Figure 3 was similar to the separation of p-xylene in Figure 1A, except that the molar ratio of BaO to K ₂ O was 1.2 and the desorbent fluorobenzene was.

The separation of p-xylene shown in Figure 4 was similar to the separation of p-xylene shown in Figure 1A except that the molar ratio of BaO to K ₂ O was 1.05 and the desorbent was 30% v / v methadifluorobenzene. volt.

The separation of p-xylene shown in Figures 5A and 5B was similar to the separation of IC and p-xylene shown in Figure 1D, except that the molar ratio of BaO to K ₂ O was 1.13.

The separation of p-xylene shown in Figures 6A and 6B was similar to the separation of p-xylene shown in IC and Figure 1D, however, with a molar ratio of BaO to K ₂ O of 0.99.

The following is a description of the terms used herein to better understand the process and advantages of the process of the present invention.

The term "feed mixture" refers to a mixture containing one or more extracted and one or more raffinate components, and is fed to the adsorbent in the process of the present invention.

The term "feed stream" means a feed that passes through the adsorbent used in the process.

The term "extracted component" refers to individual components or to a specific group of components, such as aromatic isomers, which are more selectively adsorbed on the adsorbent than other components, while "raffinate component" refers to individual components or to a specific group of components, which are less selectively adsorbed compared to other components. In this process, the p-xylene isomer is the individual extracted component and the raffinate component is one or more other C 8 aromatic hydrocarbons.

The "raffinate stream" and the "az. "raffinate effluent leaving" means the flux of raffinate components leaving the adsorbent. The raffinate stream composition can vary from substantially 100% desorbent material to substantially 100% raffinate component.

The term "extracted stream" or "stripped extract stream" refers to the stripped material that is removed from the adsorbent by the desorbent (see definition below). The composition of the extracted material stream may also vary from substantially 100% desorbent material to substantially 100% extracted component.

The high purity of the extracted product (defined below) or raffinate product (defined below) can be obtained by the process of the present invention with high productivity, but it should be understood that an extracted product can never be completely adsorbed with the adsorbent and the raffinate component is not adsorbent. its attachment is not perfect either. Therefore, a small amount of raffinate component can be present in the extracted stream and similarly a small amount of extracted component may be present in the raffinate stream. The extracted and raffinate material streams differ from each other and from the feed stream in the concentration of the extracted component and the specific raffinate component present in the respective material stream. For example, the concentration of the more selectively adsorbed para-isomer in the extracted material stream relative to the less selectively adsorbed ortho and meta isomer or ethylbenzene is the highest, also high in the feed stream, and the lowest in the feed stream. Similarly, the concentration of the less selectively adsorbed meta or ortho isomer or ethylbenzene in the raffinate stream is the highest, also high in the feed stream, and lowest in the extracted stream.

The term "desorbent material" generally refers to a substance capable of desorbing the extracted component. The term "desorbent stream" or "effluent desorbent stream" refers to the desorbent stream flowing through the adsorbent. If the extracted stream and raffinate stream from the adsorbent contain desorbent material, at least a portion of the extracted stream and raffinate stream is preferably fed to a separator, generally a fractionator, whereby at least a portion of the desorbent and the extracted product and raffinate to recover product.

The terms "extracted product" and "raffinate product" mean average products obtained by the process in which the concentration of the extracted component or raffinate component is higher than that of the extracted stream or raffinate stream, respectively.

The term "selective pore volume" refers to the volume of adsorbent in which the extracted component is selectively adsorbed from the nutrient mixture.

"Non-selective cavity volume" refers to the volume of adsorbent that does not selectively retain the extracted component from the feed medium. This volume includes the cavities of the non-adsorbent portions of the adsorbent and the intermediate voids between the adsorbent particles.

Selective and non-selective pore volumes are generally determined in units of volume; their determination is important to determine the fluid velocity required to effectively execute a given amount of adsorbent process.

The nutrient mixtures for use in the process of the present invention contain a C 8 -C 9,10,11 or C 12 dialkyl-substituted aromatic para-isomeric hydrocarbon, at least one of which is a para-substituted dialkylbenzene such as ethylbenzene (C 9), diethylbenzene. (C10), cymol (isopropyltoluene) (C10), ethylisopropylbenzene (C11) and diisopropylbenzene (C12).

Most of the above p-substituted dialkylbenzene hydrocarbons are derived from alkylation processes such as

For example, a mixture of zymol isomers resulting from alkylation of toluene, but other sources exist and any such source may be used as starting material in the process of the present invention. Mixtures containing essentially para-xylene and other C 8 aromatic isomers are derived from reforming and isomerizing processes well known in the refining and petrochemical art. The isomeric components requiring separation may also be derived from well-known alkylation and dehydrocyclodimerization processes.

In the reforming processes, petroleum is contacted with a platinum halide catalyst under defined conditions to produce C8 aromatic isomers. The reformate is then recovered, usually to concentrate the C8 aromatic isomers, in a fraction containing the C8 hydrocarbons which also contains the C8 aromatic isomers and the C8 non-aromatic hydrocarbons.

In xylene isomerization processes, a mixture of xylene lacking one or more isomers isomerized and the mixture thus obtained contains an approximately equilibrium C8 aromatic isomer and a C8 non-aromatic hydrocarbon.

The equilibrium composition of the xylene isomers and ethylbenzene at various temperatures is shown in Table 1 below:

Table 1: Equilibrium composition of aromatic carbon compounds

component C 8 aromatic isomer (mol%) Temperature (° C) 327 427 527 Ethylbenzene 6 8 11 Para-xylene 22 22 21 Meta-xylene 50 48 45 Ortho-xylene 22 22 23 100 100 100

The medium may contain small amounts of straight or branched paraffins, cycloparaffins or olefinic hydrocarbons. These materials should preferably be present in a minimal amount so that such materials, which are not selectively separated or adsorbed by the adsorbent, do not contaminate the product of the process. These impurities should preferably be present in an amount less than 20% by volume in the feed medium fed to the process.

Separation of the para-xylene from the medium containing para-xylene and at least one other C 8 aromatic hydrocarbon is effected by contacting the mixture with the adsorbent; during adsorption, para-xylene is more selectively adsorbed and retained in the adsorbent, while the other components are relatively less adsorbed and removed from the intermediate cavities between the adsorbent particles and the surface of the adsorbent. The adsorbent containing more selectively adsorbed para-xylene is referred to as a "rich" adsorbent - the adsorbent contains a large amount of selectively adsorbed para-xylene. The para-xylene is then recovered from the rich adsorbent by contacting the rich adsorbent with a desorbent.

From the other nutrients described above, the extracted fraction containing predominantly the para-isomer is recovered according to the procedure described above.

Desorption may also be effected essentially by the process and results described above.

The desorbent material may be a liquid capable of removing the nutrient component selectively adsorbed on the adsorbent. Removal of the selectively adsorbed nutrient component from the adsorbent is usually accomplished by blowing in a rotary bed apparatus; the selection of the desorbent is not very critical, the desorbent may include gaseous hydrocarbons such as methane, ethane or other gases such as nitrogen or hydrogen. Effective blowing of the adsorbed feed component from the adsorbent is accomplished at elevated temperature or under reduced pressure, or both. However, in adsorption separation processes which are generally continuous and under constant pressure and temperature to provide the liquid phase, the desorbent material must be selected according to some criteria. First, the desorbent material should replace the extracted material on the adsorbent at a suitable mass flow rate without itself being too adsorbed, thereby preventing the extracted component from being displaced by the subsequent adsorption cycle. In terms of selectivity, it is preferable (the issue of selectivity will be discussed in more detail below) if the adsorbent is more selective for the extracted component than for the raffinate component than for the desorbent material for the raffinate component. Second, the desorbent material should be compatible with the adsorbent and medium used. More specifically, the desorbent material cannot reduce or eliminate the critical selectivity of the adsorbent to the extracted component relative to the raffinate component. In addition, the desorbents used in the process of the present invention can be readily separated from the feed media fed to the process.

After desorbing the extracted nutrient components, the mixture of desorbent and extracted component is generally removed from the adsorbent. Along with the desorbent mixture, one or more raffinate components are removed from the adsorbent, so at least part of the desorbent material must be removed by, for example, distillation, otherwise

205 neither the extracted product nor the raffinate product will be sufficiently pure. Therefore, when using any substance as a desorbent, it is of primary importance that the average boiling point of the desorbent be substantially different from the boiling point of the medium so that the desorbent is simply separated from the nutrient component and raffinate stream in the extract and thus reused. The term "substantially different" means that the desorbent and the feed mixture have a boiling point difference of at least 5 ° C, preferably 40 ° C. The desorbent material may have a boiling point greater or less than its feed medium, but should preferably be lower due to economical operation.

In our work, it has been found that when the adsorbent is a barium and potassium cation exchanger, type X or Y zeolite in which the amount of barium and potassium cations is as described above (0.6: 1) to (1.2: 1) barium oxide and potassium oxide, ortho or meta-difluorobenzene or a mixture thereof may be used as an effective desorbent. The ortho- or meta-difluorobenzene compounds used in the process of the present invention are known materials and can be prepared by methods known in the art, such as those described in Kirk-Othmer, Vol. 10, 1980, p. 909-914. literature.

It has been previously recognized in the art that adsorbents have certain properties that are desirable or essential for successful implementation of selective adsorption separation processes. Such characteristics are: adsorption capacity, expressed in volume of the extracted component per volume of adsorbent; selective adsorption of the extracted component relative to the raffinate component and the desorbent material, the adsorbent having a sufficiently high adsorption rate on the adsorbent and desorption from the adsorbent.

Of course, it is necessary that the adsorbent has the capacity to adsorb the specific volume of one or more extracted components; otherwise, the adsorbent is useless for adsorption separation. In addition, the better the adsorbent, the greater the adsorption capacity of any of the extracted components. Increasing the specific adsorbent capacity allows the amount of adsorbent needed to separate the extracted component at a given feed rate. The reduction in the amount of adsorbent required for a specific adsorption separation results in a reduction in the cost of the separation process. It is important that the good initial capacity of the adsorbent during the actual use in the separation process is maintained for a period desirable from an economic point of view.

The second important adsorbent property is the ability of the adsorbent to separate nutrient components; which means that the adsorbent has an adsorptive selectivity (B) for one component relative to the other component. Relative selectivity can be expressed not only by the selectivity of one component relative to the other component, but also by the selectivity of any of the nutrient components and the desorbent.

In this specification, the selectivity (B) is determined by the ratio of the volume ratio of the two components in the adsorbed phase to the volume ratio of the same two components in the non-adsorbed phase under equilibrium conditions.

The determination of relative selectivity is shown in equation 1 below.

1. Equation

Selectivity = B = volume by volume) in the adsorption phase (% C by volume /% by volume D) in the non-adsorption phase where C and D are the C and D components of the medium.

Steady state is achieved if the composition of the feed medium passing through the adsorbent bed is unchanged. This means that there is no transfer of material between the unabsorbed and the adsorbed phases.

When the selectivity of the two components reaches 1.0, the adsorbent does not adsorb more than the other; both components are adsorbed to the same extent or not. If B is less than or equal to 1.0, one component is more adsorbed than the other. Comparing the adsorbent selectivity for component C with that for component D, a value greater than B1 indicates that component C is more adsorbed than component D. A value of less than B1 means that component D is more adsorbed, and that the non-adsorbed phase is more abundant in component C and the adsorbed phase is more abundant in component D. Although it is theoretically possible to separate an extracted component from a raffinate component, even if the adsorption selectivity for the extracted component relative to the raffinate component is only greater than 1.0, it is preferable that this value be at or above 2.0. -t. Similarly to relative volatility, separation is easier with higher selectivity. Higher selectivity, however, permits the use of a smaller amount of adsorbent. Ideally, the selectivity of the desorbent material should be less than 1.0 or 1.0 for all the extracted components so that all the extracted components can be extracted and all raffinate components removed in the raffinate stream.

The third important feature is the exchange rate of the extracted component in the medium, i.e. the relative desorption rate of the extracted component. This characteristic is directly related to the amount of desorbent needed to recover the extracted component from the adsorbent. Faster replacement5

EN 205057 Β reduces the amount of desorbent material needed to remove the extracted component, and thus reduces the operating costs of the process.

At higher exchange rates, less desorbent material needs to be fed into the process and separated from the extracted material stream for reuse in the process.

A dynamic assay apparatus is used to determine the adsorption capacity, selectivity, and rate of exchange of various adsorbing and desorbing materials applicable to a particular medium. The apparatus comprises a chamber of approximately 70 cm ³ having inlet and outlet stubs on opposite sides. The chamber is equipped with temperature and pressure control units and the chamber is operated at constant pressure. A chromatographic analyzer is connected to the outlet of the apparatus, the so-called currents leaving the adsorption chamber. "On-stream". The so called pulsed assay and subsequent conventional procedure determine the selectivity and other properties of different adsorbent systems. The adsorbent is filled with a particular desorbent to equilibrium by passing the desorbent through the adsorbent chamber. After a suitable time, a known concentration of a non-adsorbable paraffinic hydrocarbon tracer component (e.g., n-nonane) and defined aromatic hydrocarbons are injected over several minutes diluted with a desorbent. The desorbent stream was eluted and the tracer as well as the aromatic isomers were eluted in the usual manner by liquid-liquid chromatography. The outgoing material flow is called a so-called. analyzed by on-stream chromatography to give the corresponding component peaks. The effluent samples may be collected periodically and subsequently analyzed on a separate gas chromatograph.

From the chromatogram information, the adsorbent capacity, the selectivity of one isomer relative to the other isomer, and the rate of desorption of the extracted component by the desorbent are determined using the capacity index number for the extracted component. The capacity index number is defined as the distance between the center line of the selectively adsorbed muscle peak envelope and the center of the tracer component peak envelope or any other known reference point.

It is expressed as the volume of desorbent expressed in cm ³ which was fed during this time. The selectivity (B) of an extracted component relative to the raffinate component is plotted against the centerline of the extracted component peak to the centerline (or reference point) of the trace component peak to the centerline of the raffinate component to the tracer component centerline. The rate of exchange of the extracted component with the desorbent is generally characterized by the difference between the width of the peak of the C 9 (trace) component measured at half height and the width of the peak of p-xylene (adsorbed component) (denoted in some places by W). If the peak width is narrower, the desorption rate is higher. The desorption rate can also be characterized by the distance between the centerline of the peak envelope curve of the tracer component and the disappearance point of the extracted component which has just been desorbed. This distance is also expressed as the volume of desorbent fed during this time.

The adsorbent used in the process of the present invention is a crystalline aluminosilicate containing a barium and potassium cation of a defined molar ratio on the ion-exchange side. The present invention utilizes a lattice crystalline aluminosilicate in which aluminum oxide and silicon oxide tetrahedrons are bonded directly to each other in an open, three-dimensional mesh. The tetrahedron is cross-linked with oxygen atoms belonging to the molecule, and there are cavities between the tetrahedrons that previously contained water prior to partial or complete dehydration of the zeolite. Dehydration of the zeolite results in the crystals being converted to a molecular target. As a result, crystalline aluminosilicate is often referred to as "molecular filters" and their use in separation essentially depends on the size of the feed material molecules, for example, smaller normal paraffin molecules can be separated from larger isoparaffin molecules by specific molecular filters. However, the term "molecular sieve", which is otherwise widely used, is not entirely appropriate as the separation of the individual aromatic isomers is based on the difference in electrochemical attraction between the different isomers and the adsorbent, rather than on the size difference between the isomeric molecules.

Crystalline aluminum silicates in hydrated form are generally zeolites having the following general formula (1):

M ₂ / _n O: Al ₂ O ₃ : wSiO ₂ : yH ₂ O a

M is the cation equilibrating the electron valence of the tetrahedron, commonly referred to as the cation exchange moiety;

n is the valence of the cation;

w is the number of SiO ₂ molecules; and y is the number of water molecules.

The M cation may generally be a monovalent, divalent or trivalent cation or a mixture thereof.

It is generally known in the art that certain types of adsorption separation may employ zeolites X and Y. These are well known zeolites.

Hydrated or partially hydrated Z-type zeolites are characterized by the number of oxide molecules, as shown in the following compound (2):

(0.9: 0.2 / M _{2 n} O: Al ₂ O ₃ : / 2.5: 0.5 / SiO ₂ : y H ₂ O)

EN 205 057 Β in the formula

M is at least one cation having a valence of not more than 3;

n is the valency of the cation M; and y is 9, depending on the nature of M and the degree of hydration of the crystal.

As shown in formula (2), the molar ratio of silica to alumina is 2.5: 0.5.

M can be one or more cations, for example hydrogen, alkali metal or alkaline earth metal cations, and is commonly referred to as a cation exchange moiety.

Since the preparation of the X-type zeolites, they contain predominantly sodium cations and are therefore generally referred to as the sodium-type X zeolites.

Depending on the purity of the materials used in the preparation of the zeolite, the aforementioned other cations may be present, but they are only impurities.

Similarly, hydrated or partially hydrated Y-type zeolites are also characterized by the number of oxide molecules as shown in the following compound of formula (3):

(0.9: 0.2 / M ₂ / _n O: Al ₂ O ₃ : wSiO ₂ : y H ₂ O)

M is at least one cation having a valence of not more than 3;

n is the valence of the cation;

w is a number from about 3 to about 6; and y is 9, depending on the nature of M and the degree of hydration of the crystal. Thus, in Y-type zeolites, the molar ratio of SiO ₂ to Al ₂ O ₃ is from about 3 to about 6.

Like the X-type zeolites, there may be one or more different cations in the Y-type zeolites, but as a result of the production of the Y-type zeolites, these zeolites also contain predominantly sodium cations. Y-type zeolites containing predominantly sodium cation in the cation exchange region are called sodium-type Y zeolites.

The process of the present invention utilizes the Y-type zeolites described above. However, some of the cations in the factory-made zeolites are replaced with a certain amount of barium and potassium cations. The exchange of barium and potassium cations is such that the molar ratio of BaO to Κ ₂ θ in the finished zeolite is (0.6: 1) to (1.2: 1), preferably (0.9: 1) to (1.1: 1). ) be. The ion exchange can be accomplished in a single operation, with so many Ba and K cations, that the final zeolite contains the above-described amount of oxide molecule; or it may be accomplished by a series of sequential operations such that the amount of cations present in the finished zeolite is exactly the desired value. It has been found that the zeolite is suitable when the molar ratio of BaO to K ₂ O falls within the range described above; if this ratio is too high, the retention volume of p-xylene will be too high. The high retention volume results in the desorption time being too long. If the ratio falls below the lower range described above, the adsorbent becomes too selective for the desorbent and p-xylene cannot replace the desorbent in the next cycle.

Adsorption materials for use in separation processes generally have channels and cavities facilitating the entry of crystalline material dispersed in an amorphous material or inorganic material and serve as a binder for the powdered zeolite. Inorganic materials are generally silica, alumina and mixtures thereof. The binder facilitates the formation or agglomeration of otherwise fine powder crystalline particles. The particulate adsorbent thus formed may be in the form of an extruded, aggregate tablet, larger sphere, or in the form of granules having a desired particle size range, preferably between 250 and 1190 μτη (US standard: 16-60 mesh). The adsorbent should contain as little water as possible to minimize water contamination in the product.

The adsorbent can be used in the form of a compact, fixed bed that is alternately contacted with a nutrient mixture and a desorbent. In this case, the process is semi-continuous. Two or more static adsorbent beds connected to a suitable valve system may be used by passing the feed mixture through one or more adsorbent beds and the desorbent material through the other or more beds. The feed stream and the desorbent material stream may travel from the top down or the bottom down of the adsorbent beds. Any conventional apparatus used for liquid-solid contact may be used.

However, moving bed units or flow systems simulating moving beds have a much greater separation effect than fixed bed units and are therefore more advantageous. In a moving bed or simulated moving bed process, component retention and replacement operations are performed continuously, allowing continuous extraction and raffinate current generation, and continuous feed and replacement fluid flow. The process of the present invention preferably utilizes a counter-current simulated moving bed system known in the art. In this system, fluid flowing down several rays in the molecular sieve chamber simulates the upward movement of the molecular sieve. A more detailed description of the operation of counter-current simulated moving bed systems is provided in U.S. Patent No. 2,985,589, which is incorporated herein by reference, and in D.B. Broughton, 34th Annual Meeting of the Society of Chemical Engineers in Continuous Adsorptive Processing - A New Separation Technique, April 2, 1969.

A high efficiency, countercurrent, simulated moving bed method, also useful in the process of the present invention, is described in U.S. Patent 4402,832, which is incorporated herein by reference in its entirety.

The at least a portion of the extracted effluent stream is passed through a separator where at least a portion of the desorbent material is separated under conditions of separation.

It is separated from each other to reduce the amount of desorbent present in the extracted product. Advantageously, but not necessarily for the purpose of carrying out the process, at least a portion of the raffinate is fed to a separator, wherein at least a portion of the desorbent is separated under separation conditions to provide a raffinate product comprising reusable desorbent and less desorbent. The amount of desorbent present in the extracted product and in the raffinate product should generally be less than 5% by volume, preferably less than 1% by volume. The separating device is generally a well-known operating column.

Most adsorption separation processes employ both a liquid and a vapor phase system, but the process of the present invention preferably utilizes a liquid phase system because of the lower temperature requirement and the higher yield of the extracted product obtained by liquid phase operation. The adsorption separation process is carried out at a temperature in the range of 20 to 250 ° C, preferably 100 to 200 ° C, and at atmospheric pressures of up to 4238 kPa to maintain the liquid phase. Desorption is performed at a temperature and pressure appropriate to the adsorption process.

Equipment of various sizes may be used to carry out the process of the present invention. For example, it may be implemented in an experimental apparatus (see, for example, U.S. Patent No. 3,706,812) or in an industrial-scale apparatus, whereby the apparatus may operate at flow rates from a few cm ³ to a good several thousand liters.

The following examples are intended to illustrate the process of the present invention in more detail and, in particular, to illustrate the selectivity conditions.

Example 1

In this experiment, the so-called so-called. A pulse test was conducted to investigate the suitability of the process of the present invention for separating p-xylene (b.p. 138 ° C) from other C 8 aromatic hydrocarbons. The adsorbent used was Y-type zeolite which was ion-exchanged with barium and potassium cations as described below. A small amount of clay-based, amorphous binder was used to prepare the adsorbent. The zeolite contained substantially all of the potassium and barium cations on the cation exchange portion, and the amounts of cations are summarized in the table below. The cation exchange was carried out in 200 cm ³ volume of the adsorbent bed of 1.51 with 1M KCl was passed through 500cm ³ / hr at 50 ° C, followed by a 0.2M barium chloride solution was circulated through the bed for 4 hours at a flow rate of 1 liter / hour and then 1 liter of water at a flow rate of 500 cm ³ / hour. The adsorbent was then annealed at 300 ° C.

For each of the impulse tests, the column temperature was maintained at 150 ° C and the pressure maintained at 1136 kPa to maintain liquid phase operation. A gas chromatographic analyzer was connected to the column exit outlet stream to determine the composition of the outlet stream at a given time. For each of the assays, a medium containing 5% by volume of each xylene isomer, 5% by volume of ethylbenzene, 5% by volume of n-nonane as tracer and 75% by weight of desorbent was used. The desorbent used for Assays A, B and C contained 30% by volume of meta-difluorobenzene (m-BFB) and 70% by volume of C 7 -C-paraffin. The desorbent used for D was entirely meta-difluorobenzene. The assays were performed as follows: The desorbent was continuously introduced at a flow rate of 1.20 cm ³ / min. After a suitable time, the desorbent supply was stopped and the feed was introduced at a flow rate of 1 cm ³ / min for 10 minutes. The desorbent was then introduced onto the adsorbent column until all of the aromatic hydrocarbons were eluted from the column, as observed in the chromatogram of the material leaving this adsorption column.

The results of this assay are summarized in Table 2 below, and the chromatograms obtained are shown in Figures ΙΑ, 1B and IC. The figures show that the p-xylene curves are clearly and distinctly separated from the other curves, indicating good separation. Figure ID is included for comparison purposes, although separation can be observed, but the desorption time is long; this is due to the fact that the molar ratio of BaO to K ₂ O is above the range described above. The molar ratio of BaO to K ₂ O in the experiment of Figure 1D was 5.17: 1. The following data can be calculated from the curves:

AW (c ³ ): difference between the width of the p-xylene peak envelope and the half-height of the carbon-9 n-paraffin trace peak envelope; this characteristic is the rate of desorbent exchange;

the volume of p-xylene required to increase the concentration of p-xylene in the effluent stream from 10% to 90%; this characteristic is the degree of adsorption of p-xylene;

g ₀ is the volume of desorbent material required for hTJ to reduce the concentration of p-xylene in the effluent stream from 90 volumes to 10 volume%; this characteristic is a measure of the rate of desorption of p-xylene.

The volume defined at S _a which is required to increase from% volume to 90% volume is hereinafter referred to as the "breakthrough slope of adsorption". The higher the values of S _d and S _a , the slower the desorption and the adsorption respectively. If the desorption slope S _{d is} greater than _the adsorption slope S _d , then

The selectivity for benzo-p-xylene relative to the desorbent is less than 1, which is unfavorable to the adsorption process. In an ideal system, the selectivity for p-xylene relative to the desorbent is slightly greater than 1 and the retention volume 5 of the p-xylene peak envelope is small. The selectivity (B) of p-xylene relative to the desorbent can also be expressed by _the quotient S _d / S.

For example, from Table 2, the selectivity of p-xylene in Examples IC, IA, and IB can be calculated as follows:

Bib - θ) 76

The figures obtained show that the selectivity for p-xylene is advantageous in the IC experiment, while it becomes unfavorable in the experiment 1A and IB. From these data it can also be observed that as the molar ratio of barium oxide and potassium oxide increases, the desorbent becomes weaker (i.e. retention volume increases).

ΪΊ ⁹⁰ _-Seated Lemma *] - =) 90 a Yen?] —Iio

Table 2

Results of impulse examination of Ba-K-Y faujasites at different Ba and K ratios

Example 1 1A JHA IC ID desorbent 30% m-DFB 100% m-DFB nC ₉ Half Width (HW) 11.5 9.94 12.3 12.9 Selectivity (B) B p / eb 3.27 3.53 3.32 4.64 B w / w 4.04 4.47 4.13 5.22 B p / o 3.40 3.64 3.56 5.16 P-Yil meadow. vol. cm ³ 16.4 17.0 34.7 78.9 ÁW cm ³ 0.04 1.3 6.9 32.5 Properly? 14.6 10.8 10.8 6.1 7Ί ⁹⁰ –J10 ^crn3 8.0 8.3 14.5 30 SiO ₂ 60.6 60.9 60.7 59.0 A1 ₂ O ₃ 19.6 18.6 18.7 18.0 K ₂ O 8.8 8.6 6.9 2.2 BaO 9.1 10.0 11.0 18.5 Na ₂ O 0.72 0.53 0.68 0.58 BaO / K ₂ O molar ratio 0.63 0.71 0.98 5.17

HU 205057 Β

Example 2

The IC experiment was performed except that the temperature was lowered to 125 ° C. The results obtained changed only marginally, as shown in Table 3.

Table 3

Bp / e 3.37 B w / w 4.21 B p / o 3.70 WCC 7.6 n-CgH.W.cm ³ 12.00 p-xylene meadow. Vol. cm ³ 34.0 BaO / K ₂ O molar ratio 0.98

The Ba-K-Y adsorbent and the 30 vol% desorbent were identical to those used in Example IC.

Example 3

The experiment was carried out under the conditions used in Example 1 and using the same materials except that the molar ratio of barium oxide to potassium oxide was 1.2. N-heptane containing 30% m-DFB-t (bp. 82 ° C) (nC ₇₎ were used Deszorbensként.

The B selectivity for xylene isomers was as follows:

B p / eb = 3.02

B w / w = 3.94

B p / o = 3.00

5 ~ 90 g-OO -11.7 cm ³ ; λ = 16.7 cm ³ —-w —'io

The retention volume of p-xylene is 22.8, indicating an appropriate desorption time. The width of the n-nonane peak at half height is 11.9 cm ³ . AAW (described above) is 1.1 cm ³ . Composition of adsorbent:

SiO _{2 by} weight 60.6 A1 ₂ O ₃ % by weight 19.1 BaO% by weight 13.1 K ₂ O% by weight 6.7 Na ₂ O% by weight 0.6 BaO / K ₂ O (molar ratio) 1.2

The results obtained are also shown in Figure 2, where it can be observed that p-xylene can be separated from the other isomers if the molar ratio of BaO to K ₂ O is 1.2: 1.

Example 4

The impulse test described in Example 3 was repeated except that 100% by volume of fluorobenzene (b.p. 85 ° C) was used as desorbent. The B selectivity values for p-xylene are as follows:

B p / eb = 2.25

B w / w = 2.80

B p / o = 3.50

The retention volume of p-xylene is 25.7 cm ³ . The width of the n-nonane peak at half height is 12.3 cm ³ . The AAW value is 18.3 cm ³ . The molar ratio of barium oxide to potassium oxide is 1.2: 1. Breakthrough slope:

³ = 7.9 cm ⁴ .

—Lio g- | 90

The desorption slope: j = 46.4 cm ³ .

5Jio

The result of the impulse test can also be seen in Figure 3.

The selectivity thus obtained is similar to that obtained with toluene desorbent assays.

Example 5

The impulse test described in Example 3 was repeated, except that the desorbent used was n-heptane containing 30% m-DFB and the Ba-KY zeolite containing 1.05: 1 molar BaO-K ₂ O was

SiO _{2 by} weight 60.8 A1 ₂ O ₃ % by weight 18.8 Bao% by weight 11.8 K ₂ O% by weight 6.9 Na ₂ O% by weight 0.63

The B selectivity values for p-xylene are as follows:

B p / eb = 2.72

B w / w = 3.85

B p / o = 3.40

The retention volume of p-xylene is 39.6 cm ³ . The width of the n-nonane peak at half height is 10.1 cm ³ and the AAW value is 6.71 cm ³ . The result of the test can also be seen in chromatogram 4.

She. example

The impulse test described in Example 3 was repeated, except that a 30% v / v ortho-difluorobenzene and 70% v / v nheptane desorbent was used and the molar ratio of BaO to K ₂ O was 1.05.

The B selectivity values for p-xylene are as high as for the meta-difluorobenzene assay. The following selectivity values B were obtained:

B p / eb = 2.4

B m / z = 3.23

B p / o = 4.45

The 32.2 cm ³ retention volume of p-xylene indicates a moderate desorption time. The width of the n-nonane peak at half height is 10.1 cm ³ . The AAW value is 9.7 cm ³ .

Example 7

The impulse test described in Example 3 was repeated, except that the molar ratio to BaO K ₂ O was 1.13 and in Example 7A 30% m-DFB was used and in Example 7B 100% m-DFB was used. . Figures 5A and 5B show that with the same molar ratio of BaO and K ₂ O,

HU 205 057 Β

The B selectivity increased with the increase of m-DFB from 30% to 100% by volume.

The results obtained are summarized in Table 4:

s | 90

Table 4

Example VIIA VIIB BaO / K ₂ C molar ratio 1.13 1.13 desorbent 30 vol% m-DFB 100% v / v mDFB B p / e 2.50 3.95 B w / w 3.63 4.46 B p / o 3.21 3.82 p-xylene meadow. vol. 47.89 cm ³ 17.97 cm ³ AW 10.84 cm ³ -0.7 cm ³ nC ₉ HW 9.64 cm ³ 11.8 cm ³ RPO the -40 9.32 cm ³ 7.2 cm ³ ri9o - £ Ji0 14.5 cm ³ 7.2 cm ³

Table 5

Example VIIIA VIIIB, BaO / K ₂ O molar ratio 0.99 0.99 desorbent 30 vol% m-DFB 100% v / v mDFB B p / e 2.86 3.84 B w / w 3.77 4.18 B p / o 3.44 3.94 p-xylene meadow. vol. 37.17 cm ³ 11.44 cm ³ AW 6.02 cm ³ -2.39 cm ³ nC ₉ HW 9.91 cm ³ 13.05 cm ³ the ⁹⁰ 10 9.05 cm ³ 8.6 cm ³ É. 90 10 12.3 cm ³ 10.1cm ³

Example 8

The impulse test described in Example 7 was repeated, except that a molar ratio of 0.99: 1 BaO to K ₂ O was used. The results obtained are shown in Table 5 below and in the accompanying Figures 6A and 6B. Figure 6A shows that the m-xitol in the nutrient mixture left before the ethyl benzene fraction, but in the experiment of Figure 6B, the m-xylene did not separate from the ethyl benzene and / or o-xylene, these components were removed together away from the equipment.

Comparison of Figure 6A in Example 8 with Figure 5A in Example 7 and Figure 5B in Figure 6B shows that as the molar ratio of BaO to K ₂ O increases, the desorption of m-DFB decreases as it is from the retention volume of p-xylene. (str. vol.) is also obvious. For example, the retention volume at 0.99 was 37.17 cm ³ (Example 8A, Figure 6A) and at 1.13 at 47.89 cm ³ (Example 7, Figure 5A) at 30% m-DFB. Similarly, in addition to 100% by volume of m-DFB, the retention volume of 0.99 cm ³ ratio of 11.44, 1.13 and 17.97 cm molar ratio of ^{the third} In addition, the adsorbent p-xylene has a B selectivity greater than 1, relative to the desorbent.

Example 9

The pulse test described above was used to determine the potency of barium potassium faujasite to separate another medium. The test consisted of a mixture of p-diethylbenzene (p-deb) meta-diethylbenzene (m-deb) and orthodiethylbenzene (o-deb). The molar ratio of Ba-KY adsorbent BaO to K ₂ O was 0.99: 1. The binder described in Example 1 was used to prepare the adsorber. Other test conditions were substantially similar to those used in Example 1. The adsorbent material was n-heptane containing 30% v / v m-difluorobenzene. The B selectivity for desorbent A was 1.09. The results obtained are shown in Table 6 and in the accompanying Figure 7.

Table 6

BaO / K ₂ O molar ratio 0.99 desorbent 30% m-DFB in n-C7 B w / w 2.30 B p / o 3.78

HU 205057 Β

Claims (3)

PATENT CLAIMS A continuous process for separating para-isomers of dialkyl-substituted aromatic hydrocarbons having a total of 8 to 12 carbon atoms from a feed mixture containing the para-isomer and at least one other isomer of the aromatic hydrocarbon by crystalline zeolite adsorbed with barium and potassium ions and exchangeable cationic sites. by desorption with a halogenated aromatic hydrocarbon, characterized in that the feed mixture is contacted with a Y-type zeolite adsorbent containing a barium oxide and a potassium oxide molar ratio of 0.6: 1 to 1.2: 1, and then the para-isomer is The adsorbent containing the adsorbent is contacted in a known manner with a desorbent containing meta-difluorobenzene or ortho-difluorobenzene or optionally containing paraffinic hydrocarbon. 2. A process according to claim 1, wherein the adsorption and desorption is carried out at 20-250 ° C and at a pressure sufficient to maintain the liquid phase. 3. A process according to claim 1 wherein the molar ratio of barium oxide to potassium oxide (0.9: 1) is: