KR20010052680A - Separation of Epsilon Caprolactam from Isomers - Google Patents

Abstract

The present invention is 4-ethyl-2-pyrrolidinone (4-ethyl-2-pyrrolidinone), 5-methyl-2-piperdinone (5-methyl-2-piperdinone), 3-ethyl-2-pyrrolidinone One or more epsilon caprolactam isomers or octahydrophenazines selected from the group consisting of (3-ethyl-2-pyrrolidinone), and 3-methyl-2-piperdinone A method for separating epsilon caprolactam from a feed mixture consisting of and epsilon caprolactam, the method comprising contacting an adsorbent with the mixture under adsorption conditions, wherein the epsilon caprolactam is used to substantially exclude the epsilon caprolactam. Selectively adsorbing the isomer or octahydrophenazine, removing the nonadsorbed portion of the feed mixture from contact with the adsorbent, and recovering the high purity epsilon caprolactam It is broken. The epsilon caprolactam isomer or octahydrophenazine can be desorbed and recovered under desorption conditions. The method can be carried out batchwise, semibatch, or continuously using a moving bed or moving bed simulation technique.

Description

Separation of Epsilon Caprolactam from Isomers

Epsilon caprolactam is a useful intermediate in the manufacture of materials such as nylon 6. The methods used to prepare epsilon caprolactam are 4-ethyl-2-pyrrolidinone, 5-methyl-2-piperidinone, 3-ethyl-2-pyrrolidinone, and 3-methyl-2-py. It produces isomeric byproducts such as ferdinone, some of which have a vapor pressure difference of less than 1 ° C. at 10 mmHg (absolute atmospheric pressure) compared to caprolactam. Common separation methods, such as distillation, are not effective in such mixtures. Thus, it is desirable to separate epsilon caprolactam from their isomeric byproducts to provide high purity epsilon caprolactam.

The present invention provides at least one selected from the group consisting of 4-ethyl-2-pyrrolidinone, 5-methyl-2-piperidinone, 3-ethyl-2-pyrrolidinone, and 3-methyl-2-piperidinone. A method for separating epsilon caprolactam from a feed mixture consisting of additional hydrocarbons and epsilon caprolactam.

Summary of the Invention

The present invention on the one hand 4-ethyl-2-pyrrolidinone (4-ethyl-2-pyrrolidinone), 5-methyl-2-piperidinone (5-methyl-2-piperdinone), 3-ethyl-2- One or more epsilon caprolactam isomers or octahydrophenazines selected from the group consisting of 3-ethyl-2-pyrrolidinone, and 3-methyl-2-piperdinone A method for separating epsilon caprolactam from a feed mixture consisting of (octahydrophenazine) and epsilon caprolactam, the method comprising contacting an adsorbent with the mixture under adsorption conditions, in order to substantially exclude the epsilon caprolactam Selectively adsorbing the epsilon caprolactam isomer or octahydrophenazine, removing the nonadsorbed portion of the feed mixture from contact with the adsorbent, and recovering the high purity epsilon caprolactam Consists of steps. The epsilon caprolactam isomer or octahydrophenazine can be recovered by desorption under desorption conditions. The method can be carried out batchwise, semibatch, or continuously using a moving bed or simulated moving bed technologies.

The invention on the other hand consists of 4-ethyl-2-pyrrolidinone, 5-methyl-2-piperidinone, 3-ethyl-2-pyrrolidinone, and 3-methyl-2-piperidinone. A method for separating epsilon caprolactam from one or more epsilon caprolactam isomers selected from the group or mixtures of octahydrophenazine and oxylon caprolactam, wherein the method is used when the solution exceeds the solubility limit of epsilon caprolactam in solution. 4-ethyl-2-pyrrolidinone, 5-methyl-2-piperidinone, 3-ethyl-2-pyrrolidinone, and 3- which are mixtures having a composition of compositional region in which only caprolactam can be crystallized With one or more epsilon caprolactam isomers selected from the group consisting of methyl-2-piperidinone or non-eutectic mixtures of octahydrophenazine and epsilon caprolactam The oil phase is at conditions beyond the solubility limit of epsilon caprolactam to provide an initial solution and to form crystalline epsilon caprolactam containing relatively less epsilon caprolactam isomers or octahydrophenazine than were present in the initial solution. And maintaining the solution at a temperature above the eutectic point of the mixture.

Brief description of the drawings

1 shows the results of eluting caprolactam and its isomers, namely 4-ethyl-2-pyrrolidinone and 5-methyl-2-piperidinone, from a silicalite column using methanol as a solvent at 80 ° C. Drawing. Although isomers show tailing, caprolactam is completely separated from the isomers.

FIG. 2 shows the results of eluting caprolactam and its isomers from a silicalite column using 1% triethylamine and a methanol solvent at 80 ° C. FIG. Caprolactam is completely separated from the isomers. Treatment with base reduces tailing. The peak approaches zero. Base treatment combines with the acidic sites of silicalite to reduce chemical adsorption, reducing tailing.

FIG. 3 shows the results of eluting caprolactam and its isomers from a silicalite column using 1% triethylamine and a methanol solvent at 110 ° C. Caprolactam is completely separated from the isomers. Increasing the temperature will separate better.

4 is a diagram showing the result of eluting caprolactam and its isomers from a silicalite column using acetonitrile as a solvent at 120 ° C. Caprolactam is completely separated from the isomers.

FIG. 5 shows the results of eluting caprolactam and its isomers from a silicalite column using 1% triethylamine and acetonitrile solvent at 120 ° C. Caprolactam is completely separated from the isomers.

FIG. 6 shows the results of eluting caprolactam and its isomers from a silicalite column using acetonitrile as a solvent at 150 ° C. Caprolactam is completely separated from the isomers.

FIG. 7 shows the results of eluting caprolactam and its isomers from silicalite columns using methyl acetate as a solvent at 100 ° C. FIG. Caprolactam is completely separated from the isomers.

FIG. 8 shows the results of eluting caprolactam and its isomers from a silicalite column using 1% triethylamine and a methyl acetate solvent at 100 ° C. Caprolactam is completely separated from the isomers.

FIG. 9 shows the results of eluting caprolactam and its isomers from a carbon F400 column using methanol as a solvent at 80 ° C. FIG. Caprolactam elutes before isomers.

FIG. 10 is a phase diagram illustrating a phenomenon which occurs by carrying out the present invention when an aggregate is included.

11 is a reaction flow diagram when the desired species is non-retentive and melt crystallization is used.

12 is a reaction flow diagram when the desired species is non-retained and solution crystallization is used. Another solvent adaptation column was added as compared to FIG.

13 is a reaction flow diagram when the extract contains the desired species and solution crystallization is used. Compared to Figure 12 raffinate (raffinate) and extract flow was changed.

14 is a reaction flow diagram when the extract contains the desired species and melt crystallization is used.

15 is a graph showing the purity and fraction recovery results of caprolactam at various temperatures by solution crystallization using acetone as a solvent.

16 is a graph showing solubility of caprolactam in methanol.

FIG. 17 is a graph showing the results of caprolactam comparison with the starting solution of caprolactam in crystals collected using methanol as a solvent.

18 is a graph showing solubility of caprolactam in acetonitrile.

19 is a graph showing the results of caprolactam purity compared to the starting solution of caprolactam in crystals collected using acetonitrile as a solvent.

Detailed description of the invention

In one embodiment of the invention, the invention provides 4-ethyl-2-pyrrolidinone, 5-methyl-2-piperdinone, 5-methyl-2-piperdinone, 3- One or more epsilon caprolactam isomers selected from the group consisting of ethyl-2-pyrrolidinone, and 3-methyl-2-piperdinone; or A method for separating epsilon caprolactam from a feed mixture consisting of octahydrophenazine and epsilon caprolactam. The method comprises contacting a feedstock with an adsorbent consisting of activated carbon, molecular sieve carbon, molecular sieve, or zeolite that selectively adsorbs the epsilon caprolactam isomer or octahydrophenazine. The feed is then separated from the adsorbent and high purity epsilon caprolactam is recovered. The epsilon caprolactam isomer or octahydrophenazine can be desorbed and recovered under desorption conditions by a desorbent consisting of hydrocarbon or water. The individual components of the feed mixture and the desorbent have a boiling point difference of at least 5 ° C.

In another embodiment, the present invention consists of 4-ethyl-2-pyrrolidinone, 5-methyl-2-piperidinone, 3-ethyl-2-pyrrolidinone, and 3-methyl-2-piperidinone. One or more epsilon caprolactam isomers selected from the group or a method for separating epsilon caprolactam from a feed mixture consisting of octahydrophenazine and epsilon caprolactam, the method comprising contacting the mixture with an adsorbent under adsorption conditions Selectively adsorbing the epsilon caprolactam to substantially remove the epsilon caprolactam isomer or octahydrophenazine, removing the nonadsorbed portion of the feed mixture from contact with the adsorbent, and high purity epsilon under desorption conditions Recovering caprolactam. The method can be carried out batchwise, semibatch, or continuously using mobile bed or mobile bed simulation techniques.

Once the adsorptive separation step is complete, the optional step may further comprise crystallizing epsilon caprolactam from these solutions, which is to separate the higher purity form of epsilon caprolactam as described below. Crystallization may be used alone as described below. Other embodiments of the invention include details of feed mixtures, desorbents, flow diagrams, and operating conditions, all of which are described below for each of the various aspects of the invention.

The definitions of various terms initially used herein will be useful to clarify the operation, object, and advantages of the present invention.

A "feed mixture" is a mixture containing one or more raffinate components and one or more extraction components, such as epsilon caprolactam isomers and epsilon caprolactam, which are fed to the adsorbent in the present invention. "Feed stream" refers to the flow of the feed mixture to the adsorbent used in the process of the present invention.

A "raffinate component" refers to a compound or form that is less selectively adsorbed, whereas an "extract component" refers to a compound or form that is more selectively adsorbed by an adsorbent. In a preferred embodiment of the invention the epsilon caprolactam isomer is an extract component and the epsilon caprolactam is a raffinate component. "Raffinate stream" or "raffinate discharge stream" means a stream through which a raffinate component is removed from an adsorbent. The composition of the raffinate flow can vary from 100% complete desorbent (defined below) to 100% complete raffinate component. By "extract stream" or "extract output stream" is meant a stream through which the extract, which has been desorbed by the desorbent, is removed from the adsorbent. Similarly, the composition of the extract stream varies widely from 100% of desorbents to 100% of extract components. Although the present invention makes it possible to produce high purity extraction products (defined below) or raffinate products (defined below) at high recovery rates, the extraction components are not all adsorbed to the adsorbent and the raffinate component is adsorbent. It is not said that it is not adsorbed at all. Thus, small amounts of raffinate component may appear in extract streams and the like, and small amounts of extract component may appear in raffinate streams.

The extract and raffinate streams are further distinguished from the feed stream and from each other by the ratio of the concentrations of the extract component and the specific raffinate component, both of which appear in a particular stream. For example, the ratio of the concentration of more selective adsorbed epsilon caprolactam isomers to the less selective adsorbed epsilon caprolactam is highest in the extraction stream, then higher in the feed mixture, and lowest in the raffinate stream. will be. Similarly, the ratio of the less selective adsorbed epsilon caprolactam to the more selectively adsorbed epsilon caprolactam isomer ratios is highest in the raffinate stream and then higher in the feed mixture and lowest in the extraction stream. will be.

"Desorbent material" or "desorbent solvent" generally means a material capable of desorbing the extract component. "Desorbent stream" or "desorbent input stream" refers to the flow of desorbent material to the adsorbent. When the extraction stream and the raffinate stream contain desorbents, at least one part of the extraction stream, preferably at least one part of the raffinate stream, from the adsorbent will be passed to a separation means, generally a fractionator. At least one portion of the desorbent will be separated under separation conditions to produce an extract product and a raffinate product. "Extract product" and "raffinate product" refer to products produced in a process containing extract and raffinate components at higher concentrations than those in the extraction stream and raffinate stream, respectively.

Here, epsilon caprolactam is used in various fields such as the production of synthetic fibers (especially nylon 6), plastics, bristle, films, coatings, synthetic leathers, plasticizers and applicators, the synthesis of amino acids lysine or crosslinking agents for polyurethanes. Can be used.

Preferred processes include the pentenic acid salt route in epsilon caprolactam disclosed in US patent application Ser. No. 08 / 839,576, filed April 15, 1997 and epsilon disclosed in US patent application Ser. No. 08 / 843,340, filed April 15, 1997. Penthenol route of caprolactam and other routes of epsilon caprolactam disclosed in US patent application Ser. No. 09 / 094,651, filed June 15,1998. In addition, Beckman rearrangements of cyclohexanone oxime of epsilon caprolactam disclosed in U.S. Pat. Include. Epsilon caprolactam compositions can be prepared without the need to separate less stable intermediates, such as isomers of formyl valeric acid or salts or isomers of hydroxyhexanal, and without developing separation processes for less stable molecules. Thus, the separation becomes possible at a more desirable point in the process for producing epsilon caprolactam, thus improving the efficiency.

Some conversion processes for epsilon caprolactam include 4-ethyl-2-pyrrolidinone, 5-methyl-2-piperidinone, 3-ethyl-2-pyrrolidinone, and 3-methyl-2-piperidinone. The same boiling point also provides products comprising similar isomers of epsilon caprolactam. Another process for epsilon caprolactam is octahydrophenazine, cyclohexanone, cyclohexanone oxime, N-cyclohexylidenebutylamine, aniline, isovaleramide, valericamide, isocaproamide, gamma-methyl-gamma Provided are compounds containing similar boiling points or unreacted substances, such as valerolactam, caproamide, adipimine, N-butylacetamide, and methylcaprolactam. Thus, the conversion processes eventually need to remove these isomers, by-products or contaminants to produce high purity epsilon caprolactam. The quality of the polymerization grade is important because the separation and purification of epsilon caprolactam is mainly used for the production of nylon 6 by the polymerization process. Therefore, the presence of impurities is an important problem because of the sensitivity of the process to contaminants.

The present invention relates to certain epsilon caprolactams such as 4-ethyl-2-pyrrolidinone, 5-methyl-2-piperidinone, 3-ethyl-2-pyrrolidinone, and 3-methyl-2-piperidinone. The purification process is simplified by providing a simple and efficient method for removing epsilon caprolactam from the isomers. Thus, certain adsorbents may be selected from 4-ethyl-2-pyrrolidinone, 5-methyl-2-piperidinone, 3-ethyl-2-pyrrolidinone, and 3-methyl-2 in order to substantially exclude epsilon caprolactam. It has been found that it will adsorb epsilon caprolactam isomers such as pipedinone. Suitable feed mixtures for practicing the invention include epsilon caprolactam, 4-ethyl-2-pyrrolidinone, 5-methyl-2-piperidinone, 3-ethyl-2-pyrrolidinone, and 3-methyl-2 -Piperidinone alone or in combination. However, it is impractical to obtain a feed mixture that is completely free of by-products, and therefore suitable feed mixtures may contain small amounts of other hydrocarbons with similar boiling points that can be removed from extracts or raffinates. Typical sources for the feed mixtures according to the invention are the processes defined above. Preferred feed mixtures for use in the present invention are disclosed in US patent application Ser. No. 08 / 956,745 filed October 23,1997.

In order to separate epsilon caprolactam from the feed mixture according to the invention, the mixture is contacted with an adsorbent and other components in the mixture are relatively less adsorbed and removed from the interstitial voids between the adsorbent particles and the adsorbent surface, while epsilon capro The lactam isomer is more selectively retained adsorbed on the adsorbent. Adsorbents containing more selectively adsorbed epsilon caprolactam isomers are called "rich" adsorbents. The epsilon caprolactam isomer is then recovered from the rich adsorbent by contacting the desorbent with the rich adsorbent.

"Desorbent" or "desorption solvent" means any fluid material capable of removing the selectively adsorbed feed components from the adsorbent. Generally, in swing-bed systems in which the selectively adsorbed feed constituents are removed from the adsorbent by the refinery flow, desorbent selection is not very important and consists of gaseous hydrocarbons such as methane, ethane and the like. Desorbents or other forms of gas, such as nitrogen or hydrogen, may be used to effectively purify the adsorbed feed components from the adsorbent at high temperatures, reduced pressure, or both. In general, however, in adsorptive separation processes that operate continuously at substantially constant pressures and temperatures to maintain a liquid phase, reliable desorbents must be carefully selected to satisfy several categories.

First, the desorbent must be so strongly adsorbed that it must be replaced with the extract from the adsorbent at a suitable flow rate without completely preventing the extract from being replaced with the desorbent in the next adsorption cycle. Expressed in terms of selectivity (which will be described in more detail below), the adsorbent is preferably more selective to the extraction component relative to the raffinate component than to the desorbent relative to the raffinate component. Second, the desorbent must be compatible with certain adsorbents and certain feed mixtures. More specifically, they should not reduce or destroy the high selectivity of the adsorbent for the extract component relative to the raffinate component. Moreover, the desorbents that can be used in the process of the present invention should be those that are easily separated from the feed mixture that is delivered to the process.

After desorbing the extract component from the feed, the desorbent and extract component are generally removed from the adsorbent in the mixture. Likewise, one or more of the raffinate components are generally removed from the adsorbent in a desorbent mixture, and the purity of the extract product or of the raffinate product is without a method of separating at least one portion of the desorbent, such as distillation. It will not be very high. Thus, any desorbent used in the process of the present invention can be separated from the feed components in the extraction and raffinate streams by simple fractionation, thereby avoiding the boiling point of any individual components of the feed mixture so that the desorbents can be reused in the process. It can be expected to have substantially different boiling points. By "substantially different" is meant that the boiling point of the desorbent and the individual components of the feed mixture differ by at least about 5 ° C. The boiling range of the desorbent may be higher or lower than that of the feed mixture and is preferably lower. It has been found that the effective desorption material in the preferred isothermal, isostatic, liquid phase operation of the process according to the invention consists of hydrocarbons, especially aromatic and aliphatic hydrocarbons having up to 8 carbons or water.

Examples of desorbents useful in the present invention include methanol, ethyl ether, cyclopentane, acetone, methyl acetate, isopropyl ether, hexane, methyl cyclopentane, ethyl acetate, ethanol, benzene, cyclohexane, acetonitrile, propanol, propionitrile , Water, toluene, cycloheptane, ethylbenzene, p-xylene, and cyclooctane.

Certain features of the adsorbent are not necessary but highly desirable for the successful performance of the selective adsorption process. These properties include adsorption space for a constant volume of extract component per adsorbent volume, selective adsorption of the extract component in terms of raffinate component and desorbent, and sufficiently fast adsorption and desorption rate of the extract component from and to the adsorbent.

The space of the adsorbent to adsorb a particular volume of one or more extract components is of course a requirement; This spaceless adsorbent is unnecessary for adsorptive separation. Also, the larger the adsorbent space for the extract component, the better the adsorbent. Increasing the space of a particular adsorbent may reduce the amount of adsorbent required to separate the extracted constituents at a particular charge ratio of the feed mixture. Reducing the amount of adsorbent reduces the cost in the separation process required for a particular adsorptive separation. In any economically desirable life separation technique, it is important to maintain sufficient initial space for the adsorbent in practical applications.

The second essential characteristic of the adsorbent is the ability of the adsorbent to separate the components of the feed material; In other words, the adsorbent has adsorption selectivity (B) that can distinguish one component from another component. Relative selectivity is expressed not only to distinguish one feed component from another, but may also be expressed between any feed mixture component and desorbent. As used herein, selectivity (B) is defined as the ratio of two such components above the adsorption phase to the ratio of two components of the desorbable phase at equilibrium conditions.

Relative selectivity is represented by Equation 1:

Where C and D are the two components of the feed, expressed in volume percent, and the subscripts A and U represent the adsorptive and nonadsorbent phases, respectively. Equilibrium conditions are determined when the feed through one bed of adsorbent does not cause a change in composition after contact with the adsorbent bed. In other words, no mass transfer occurs between the nonadsorbed and adsorbed phases.

When the selectivity of two components approaches 1.0, the adsorption of one component to the other by the adsorbent is less prevalent; That is, they are not adsorbed to the same extent to each other. When (B) is greater than or less than 1.0, one component is adsorbed better than the other.

When comparing the selectivity of the C adsorbent to the component D, a value of (B) greater than 1.0 indicates that the component C adsorbs better in the adsorbent. A value of (B) less than 1.0 indicates that component D is better adsorbed so that component D remains richer in adsorption phase and component C in nonadsorption phase. Although it is theoretically possible to separate the extract component from the raffinate component even when the selectivity of the extract component to the adsorbent for the raffinate component is slightly above 1.0, it is preferred that the selectivity has a value of about 2 or more. Higher selectivity, such as relative volatility, facilitates separation. Higher selectivity reduces the amount of adsorbent used for separation. Ideally, the desorbent should have a selectivity of about 1 or less for all extract components so that all of the extract can be extracted in a class and all of the raffinate components can be completely separated into the raffinate stream.

A third important property is the rate of exchange of the extract components of the feed mixture, ie the relative desorption rates of the extract components. This feature is directly related to the amount of desorbent that must be used in the process for recovering the extract component from the adsorbent; Faster exchange rates can reduce the operating costs of the process by reducing the amount of desorbent needed to remove the extract. At fast exchange rates, desorbents have to be pumped less through the process and separated from the extraction stream for reuse in the process. As shown in the figure, the exchange rate of the desorbent and the extraction component is generally related to the width of the peak measured at half intensity.

The fourth important characteristic of the adsorbent is that the feed and desorption components are free of reactivity or catalysis that cause undesirable chemical changes. Zeolite materials are known to react with hydrocarbons, in particular olefins. In contrast to these chemically active adsorbents, the carbon adsorbent according to the invention is chemically inert to the components of the process stream.

Dynamic test apparatus are used to test various adsorbents and desorbents with specific feed mixtures to determine the adsorption space, selectivity, and exchange rate characteristics of the adsorbent. The apparatus consists of about 70 cm 3 of adsorbent chamber with inlets and outlets at opposite ends of the chamber. The chamber is enclosed in temperature control means and in addition a pressure regulating device used to operate the chamber at a predetermined constant pressure. The chromatographic analysis device can be used to analyze the elution stream “on-stream” leaving the adsorbent chamber attached to the outlet line of the chamber or the sample can be removed from the device for chromatographic analysis.

Pulse tests conducted using the apparatus and the following general procedure are used to determine selectivity and other data for various adsorption systems. By transferring the desorbent through the adsorbent chamber, the adsorbent is filled to equilibrate with the particular desorbent. At a convenient time, a pulse feed containing a known concentration of the particular epsilon caprolactam isomer completely diluted in the desorbent is injected. The desorbent flow is restarted and the feed isomer is eluted as in liquid-solid chromatography operation. The eluate can be analyzed by an on-stream chromatography apparatus and traces of the corresponding component peaks can be produced. Alternatively, eluted samples can be collected periodically and analyzed separately by gas chromatography.

From the information inferred from the chromatographic traces, the performance of the adsorbent can be determined by the spatial index of the extract component, the selectivity of the isomers as long as they are distinguished from others, and the desorption rate of the extract component by the desorbent. The spatial index can be characterized by the distance between the center of the peak of the component that is selectively adsorbed and any other known comparison point. This can be expressed as the volume cm 3 of the desorbent pumped during that time interval. The selectivity (B) of the extract component relative to the raffinate component can be expressed as the ratio of the distance between the extract component peak center and the trail peak edge to the distance between the raffinate component peak center and the trail peak edge. In general, the exchange rate between the desorbent and the extract component can be expressed as the peak width at half the intensity. The narrower the peak width, the faster the desorption rate. The desorption rate can be expressed as the distance between the disappearance of the extract component that has just been desorbed and the center of the trace peak. This distance is again the volume of desorbent pumped during that time interval.

Adsorbents used in the process of the present invention include any adsorbent that selectively adsorbs epsilon caprolactam to substantially exclude octahydrophenazine or selectively adsorbs epsilon caprolactam isomers to substantially exclude epsilon caprolactam. . Examples of suitable adsorbents are molecular sieves of zeolites such as activated carbon, molecular sieve carbon, and zeolites and non-zeolites such as molecular sieves. Suitable adsorbents include the adsorbents disclosed in the Chek-Othmer fourth edition (1996).

Activated carbon is a commercially available common material, such as Calgon Company's "Type F400" particle carbon, which was published in August 1978 by Calgon Brochure No. "PURASIV", Type PCBs, such as those disclosed in 23-108a, are activated carbon having a large micropore volume in the range of 15 to 20 microns in diameter penetrated by a large pore system larger than 1000 microns in diameter. "PURASIV" is a bead activated carbon made by spherical molten petroleum pitch followed by carbonization and activation.

As used herein, "molecular carbon" is not meant to be distinguished from a material called "activated carbon" but is used to not exclude any material effective for use in the present invention. For the purposes of the present invention, the two terms may be used in duplicate and may be used interchangeably. Particular molecular sieve carbons useful in the present invention are those having an average pore size of greater than about 5 kPa and less than 10 kPa, preferably between about 5.2 kPa and 8 kPa, more preferably between about 5.5 kPa and 6.5 kPa.

The adsorbent used in the present invention includes a zeolite molecular sieve such as zeolite and a non-zeolitic molecular sieve such as molecular sieve. Examples of zeolites useful in the present invention are LZ-10, LZ-20, 4A, 5A, 13X, 10X, Y, SK40, SK41, chabazite, faujasite, levynite, gas Gismondine, erionite, sodalite, analcime, gmelinite, harmotome, mordenite, epistillbite, Heurlandite, stilbite, edingtonite, mesolite, natrolite, squaleite, thomsonite, brusterite ( brewsterite, laumontite, philipsite, ZSM's (ZSM-5, ZSM-20, ZSM-12, ZSM-34) and the like. Examples of zeolites useful in the present invention are disclosed in US Pat. Nos. 3,702,886, 3,972,983, 3,832,449, 4,086,186 and 3,308,069.

Examples of molecular sieves useful in the present invention include silica molecular sieves such as silicalite (S115), as shown in US Pat. Nos. 4,061,724 and 4,073,865. Examples of another molecular sieve are crystalline microporous molecular sieve oxides based on the presence of aluminophosphate in the crystal structure, for example SAPO, MeAPO, FAPO, MAPO, MnAPO, CoAPO, ZAPO, MeAPSO, FAPSO, MAPSO, MnAPSO, CoAPSO And substances represented by abbreviations such as ZAPSO, ElAPO, ElAPSO, and the like. Such molecular sieves are disclosed in US Pat. Nos. 4,567,029, 4,440871, 4,500,651, 4,554,143, and 4,310,440.

Zeolite and molecular sieve adsorbents preferably have an average pore size of greater than about 5 mm 3 and less than 10 mm 3, preferably between about 5.2 mm and 8 mm 3, more preferably between about 5.5 mm and 6.5 mm 3. The adsorbent may of course have meso- and macro-pores together with the above preferred pore sizes. It has been discovered by the present invention that certain adsorbents can adsorb isomers of epsilon caprolactam for substantial exclusion of epsilon caprolactam.

The adsorbents may be in the form of granules or macrospheres with extrudate, aggregates, tablets, desired particle size, preferably about 4 to 60 mesh (US standard mesh). If the water is not a desorbent, if the moisture content of the adsorbent is low, there is an advantage of reducing the water contamination of the product.

The adsorbent can be used in the form of a dense fixed bed, which bed is in alternative contact with the feed mixture and the desorbent and in each case the process is only semi-continuous. In other embodiments, the desorbent may pass through one or more other beds in a set while the feed mixture may pass through one or more sets of adsorbent beds so that two or more stationary beds of adsorbent One set can be used with a suitable valve. The flow of feed mixture and desorbent may be up or down through the adsorbent in these beds. Any general apparatus used for stationary bed fluid-solid contact may be used.

However, mobile bed flow systems simulated with mobile beds have greater separation efficiencies than fixed bed systems and are therefore more desirable. In the mobile bed process simulated with a mobile bed, a continuous operation of the replacement and supply of the fluid flow and the replacement operation and retention which enables the continuous generation of the extraction flow and the raffinate flow occur continuously. Preferred embodiments of this process use those known in the art, such as simulated mobile bed countercurrent flow systems. In such a system, it is precisely the forward movement of the liquid passing multipoint below the adsorbent chamber that simulates the upward movement of the adsorbent contained in the chamber. U.S. Patent No. 2,985,589 discloses the principle and sequence of operation of such a flow system, published by DB Broughton at the 34th annual meeting of the Institute of Chemical Engineers, Tokyo, Japan, April 2, 1969, "Continuous Adsorptive Processing-A New". The paper entitled "Separation Technique" further describes the flow diagram of a simulated moving bed backflow process.

Another embodiment of a mobile bed simulation flow system suitable for use in the process of the present invention is a high efficiency mobile bed simulation process disclosed in US Pat. No. 4,402,832.

At least a portion of the extract discharge stream will be passed to the separation means, and it is believed that at least a portion of the desorbent can be separated under separation conditions to produce an extract product containing desorbed substances having reduced concentration. Although not essential to operating the process, it is preferred that at least a portion of the raffinate production stream is also passed to the separation means, and at least one portion of the desorbent may be reused in the process and the raffinate stream containing desorbed substances having reduced concentration. May be separated under separation conditions to produce a desorption stream. In the extraction and raffinate products, the concentration of desorbent will be less than about 5% by volume, more preferably less than about 1% by volume. In general, the separation means will be a sorting column whose design and operation are well known in the separation art.

Although liquid phase and gas phase operation can be used for many adsorptive separation processes, liquid phase operation is more likely due to higher yields of extractable products obtained in liquid phase operation than those obtained with gas phase operation, and because of lower temperatures, lower energy requirements, and smaller device sizes. This is preferred. Adsorption conditions will include a temperature of about 20 ° C. to 250 ° C., preferably 50 ° C. to 200 ° C., and sufficient pressure to maintain the liquid phase. Desorption conditions include the same temperature and pressure as used in adsorption conditions.

The size of units that can use the process according to the invention can vary from the test plant level to the commercial level, and flow rates can range from as little as a few cm 3 per hour to thousands of gallons per hour.

In another embodiment the present invention relates to epsilon caprolactam with 4-ethyl-2-pyrrolidinone, 5-methyl-2-piperidinone, 3-ethyl-2-pyrrolidinone, and 3-methyl-2-piper. Partially associated with a process for separating epsilon caprolactam from a mixture of one or more epsilon caprolactam isomers or octahydrophenazines selected from the group consisting of dinons, the process being involved in the solution when the solubility limit is exceeded in solution. 4-ethyl-2-pyrrolidinone, 5-methyl-2-piperidinone, 3-ethyl-2-pyrrolidinone, and 3-methyl-2-piper, which are mixtures having only the composition of the component system in which only crystallization occurs Providing an initial solution containing one or more epsilon caprolactam isomers or non- eutectic mixtures of octahydrophenazine and epsilon caprolactam selected from the group consisting of dinons, and an initial solution Maintain the solution at conditions above the solubility limit of epsilon caprolactam and at temperatures above the eutectic melting point of the mixture to form crystalline epsilon caprolactam containing relatively less epsilon caprolactam isomers or octahydrophenazine than were present at It consists of steps. Once the crystallization separation process is completed, further comprising adsorbing separation of epsilon caprolactam from their solution to separate epsilon caprolactam in a more pure form as described above. Adsorption may be used alone as described above.

Suitable solvents may be provided by using liquid epsilon caprolactam or by melting solid epsilon caprolactam (when melt crystallization is used). However, it is common for a suitable solution to consist of epsilon caprolactam dissolved in a suitable solvent (eg a solvent in which adsorptive separation is carried out). Any solvent that dissolves epsilon caprolactam can be used. Examples of suitable solvents are methanol, ethyl ether, cyclopentane, acetone, methyl acetate, isopropyl ether, hexane, methyl cyclopentane, ethyl acetate, ethanol, benzene, cyclohexane, acetonitrile, propanol, propionitrile, water, toluene , Cycloheptane, ethylbenzene, p-xylene, cyclooctane, ketones such as acetone, esters such as ethyl acetate, hydrocarbons such as toluene, nitrohydrocarbons such as nitrobenzene, and ethers such as tetra Hydrofuran (THF) and glyme]. Two or more solvent mixtures may be used to maximize the amount and purity of the desired epsilon caprolactam. The solution used may comprise other substances present in the crude reaction products (eg catalysts and by-products) of the epsilon caprolactam forming reaction. The concentration of epsilon caprolactam in solution will be limited by the solubility of epsilon caprolactam in the solvent. Examples of crystallization techniques useful in the present invention are described in US Pat. Nos. 5,430,194, and 5,675,022.

In the process of the invention, the solution containing epsilon caprolactam is maintained under conditions beyond the solubility limit of the desired epsilon caprolactam. Such conditions include adding a nonsolvent to the solution, removing any solvent from the solution, and preferably cooling the solution (including vacuum cooling of the solution). When these conditions are combined, they can be used to achieve the desired crystallization effect.

For crystallization by using solvent removal, it should be noted that if the solvent phase pressure is fixed, the given heat will increase the solution temperature until the solution boils. As heat continues to be applied, the solvent will evaporate and the solution will be saturated. At this point, the solution concentration will remain constant (Gibbs Phase Rule) and the solute (ie desired epsilon caprolactam) will precipitate with continued heating. Conversely, if the pressure on a saturated solution with increasing solubility increases with decreasing temperature, the solution temperature is lowered and the cooling will result in precipitation (crystallization) of the solute (ie desired epsilon caprolactam).

For crystallization using nonsolvent addition, such as hexane, the addition of a liquid to a saturated solution mixed with a solvent, however, causes the solute (ie, epsilon caprolactam) to precipitate (crystallize) if the solute has limited solubility therein. Will be

Although the description of the present invention mainly relates to crystallization by cooling as shown below, the present invention includes any conditions for obtaining the desired crystals.

If the mixture is a component system in which only epsilon caprolactam crystallizes when cooling the mixture solution, the present invention can be applied to separate epsilon caprolactam from a mixture containing isomers of epsilon caprolactam and epsilon caprolactam. Suitable mixtures include mixtures of epsilon caprolactam compound (shown in FIG. 10 described below) aggregates.

When the epsilon caprolactam to be isolated is an aggregate, the crystallization phenomena that occur when practicing the present invention are generally shown in FIG. 10, which is a phase diagram of X (i.e. Controlled by a factor. In FIG. 10, region (i.e., component system) I represents an unsaturated solution containing X and Y, and region (i.e., component system) II corresponds to where crystals of Y and a saturated solution containing X and Y coexist, (I.e., component system) III indicates where the crystals of X and the saturated solution containing X and Y coexist, and the region (i.e. component system) IV corresponds to the crystal mixture of X and Y. Curved separation regions (i.e. component systems) I and II are the solubility curves of Y, while curved separation regions (i.e. component systems) I and III are curves for phase equilibrium between solids X and corresponding solutions containing X and Y. . The curve intersection at point E with the solution having composition E saturated by solids X, solids Y, X and Y is the point of equilibrium. The points t _x and t _y are the melting points of the pure X and Y components, respectively.

When the unsaturated solution containing X and Y (represented by point A in FIG. 10) is cooled, the composition of the solution does not change, so the point representing the cooling solution moves down the vertical of the phase diagram (FIG. 10). With continued cooling, this vertical line intersects the solubility curve at point B on the region boundary corresponding to the separation of the Y crystals. As it continues to cool, only the component Y crystals separate and the solution depletes component Y, thus the composition of the solution moves along the solubility curve from right to left. For example, when cooling the solution below the temperature corresponding to point C, the mother liquor (melt or solution) with the crystals of composition F and the composition corresponding to point D Equilibrium in weight ratio. As the temperature continues to drop, the point representing the liquid phase (solution) moves along the solubility curve towards point E. Eventually, at the temperature corresponding to point G, the H crystals are in equilibrium with the composition E solution. Solution E is saturated with both components, and therefore the crystals of both components are separated from the liquid phase (solution) having a constant composition at a constant temperature t _e when heat is continuously removed. Thus, the temperature t _e is the lowest temperature at which single component crystals can be obtained from solution. In the initial solution A, the weight ratio of the maximum amount of Y crystals obtained to the E mother liquor is Given by the ratio of the segments. The point E is also called the eutectic point, the temperature t _E is the eutectic temperature, and the mixture of X and Y with the composition corresponding to the point E is a eutectic mixture.

In a component system in which only desired epsilon caprolactam is obtained by crystallization, crystallization proceeds using a solution containing a non- eutectic epsilon caprolactam mixture. During crystallization by cooling, the relative concentration of epsilon caprolactam, the constant of the solution temperature, the cooling rate, and the cooling temperature are adjusted to maintain the concentration of epsilon caprolactam in the zone in which only the desired epsilon caprolactam is crystallized. Thus, referring to FIG. 10, only the component Y must be adjusted so that the relative concentration of epsilon caprolactam is to the right of the eutectic concentration (E). During crystallization (when the concentration of Y in the solution is shifted left on the solubility curve towards the eutectic concentration, E), the appropriate concentration is maintained by stopping the crystallization before reaching the eutectic concentration and / or temperature.

If the composition is in a region where only the desired epsilon caprolactam crystallizes when the mixture is cooled, the epsilon caprolactam mixture useful in the process of the present invention may have any composition other than the composition in which the mixture is eutectic (ie, the mixture is non- eutectic). Can be. The use of non-eutectic mixtures is generally necessary because undesired isomers are crystallized from eutectic mixtures so much that they are unacceptable.

In one embodiment of the present invention, the solution containing epsilon caprolactam is cooled to be effective for crystallization of the desired epsilon caprolactam. First the crystal is produced at a temperature just below the initial crystal temperature, then the temperature of the solution can be raised slightly and then the temperature can be lowered again. This technique allows smaller crystals to dissolve again and larger crystals to grow larger, resulting in better crystals from solution. Crystallization temperature affects product purity and yield in that lower temperatures yield higher yields. Vacuum cooling can be used in the practice of the present invention.

In another embodiment of the invention, crystallization, if desired, is carried out by cooling in stages. That is, the initial solution of epsilon caprolactam can be cooled to the temperature at which the desired epsilon caprolactam is crystallized and maintained at that temperature until crystallization is complete. The crystals can then be filtered out of the remaining solution to produce a filtered fraction which can be cooled again to crystallize a larger amount of desired epsilon caprolactam. The cooling-crystal-filtration-cooling step can be repeated as desired. The advantage of the stepwise operation is that it increases the yield of desired epsilon caprolactam. It is desirable to remove some of the solvent between each cooling step.

In practicing the present invention, the crystallization of the desired epsilon caprolactam can be carried out using any conventional apparatus. A preferred device is the falling film crystallizer disclosed in US Pat. No. 3,621,664, which has a vertical wall surface (generally metal) that is cooled from the opposite wall surface. When the liquid phase (i.e. epsilon caprolactam solution) flows like a much smaller streamlike film that spreads over the wall, the flow rate when the liquid phase is a film fills the same flow rate, wet edges, and the entire cross section of the means, such as pipes. When separated much better than when obtained. This is because in the case of films the flow is vortex, but in other cases the flow has a Reynolds number of 1600 representing the laminar flow. The vortex flow in the falling film has a (1-3) / 10 millimeter thick lamina boundary layer with mass transfer caused by molecular diffusion while the boundary layer for full lamina flow is about 10 millimeters thick. When the crystallization rate is about 1 cm per hour as required in large scale processes, and the molecular diffusion coefficient in the liquid phase is about 10 ⁻⁵ cm 2 / sec, the true distribution coefficient equation disclosed in US Pat. No. 3,621,664 is close to optimal. Indicates that a distribution coefficient is obtained with the film flow; In other cases, the distribution coefficient is close to 1, indicating that no separation actually occurs. If you want to separate well in other cases, the Reynolds number must be high, which requires more flow and more power, especially in viscous liquids, which makes the operation uneconomical.

The desired epsilon caprolactam during crystallization can be well separated in the apparatus of US Pat. No. 3,621,664 even in the lamina zone if the curvature appearing on the film surface causes a mixing action. The layer thickness here is also (1-3) / 10 millimeters and therefore is well separated. The amount of liquid to be treated and the power consumed by the circulation pump are relatively small. In a preferred embodiment the cooled vertical wall of the crystallizer is a tube having the desired number of vertical, horizontal tubes, liquid introduced above the tube by a distributor which can flow down the tube inner wall like a film, and a refrigerant filling the jacket around the tube It is a bundle. The lower end of the crystallizer is connected with a tank for collecting the liquid phase.

Desired epsilon caprolactam crystals are generally formed on the inner surface of the falling film crystallizer. Crystals are removed by dissolving the crystals in a solvent (eg acetone) at a temperature below the melting point of the desired crystal to avoid substantial degradation of the desired crystal.

Two different arrangements of the U.S. Patent No. 3,621,664 can be used to crystallize according to the invention in industrial units. In one such arrangement, crystals appear on the outer surface of a heat exchanger consisting of thin, parallel tubes with bulkheads that produce a strong transverse flow of liquid phase. In another arrangement, crystals form on the outer surface of the horizontal pipe lattice and liquid phase flows down the lattice. In both arrangements the transverse flow near the pipes creates a vortex that provides a general mixing action, so the lamina boundary layer in each pipe is very thin. Similar results are obtained with cooled, in some cases heated short fins or gratings located within the flow which can cause the exposed cross flow.

Separation during crystallization will be improved by simply heating (in some cases cooling) the fluidized bed before entering the crystallizer in the preferred crystallizer. Such a device provides a smooth crystalline surface and avoids the growth of dendritic or uneven crystals with undesirable blockage of the mother liquor in the crystal layer.

Crystallization in a single device is conveniently carried out in such a way that a single crystallization is cyclically repeated, in which the desired component steps appear in the purest form in the above preferred crystallizing device and start with the highest concentration step of impurities. A small amount of mother liquor (i.e. epsilon caprolactam solution) on the surface of the crystallizer slightly contaminates the crystallization of the subsequent phases, and when one cycle ends and another cycle begins, the "pureest" phase to the "not purest" phase Going to does not affect separation.

The crystallization process can be carried out in a preferred crystallization apparatus in an inert atmosphere. The determination of the final stage can be further purified by distillation or partial melting and the less purified separation material is returned to the last stage. The surface on which crystallization takes place can be cooled by flowing a heat exchange medium over the surface opposite the crystallizer wall in a film form. This surface can be vertical, horizontal, or any angle in between.

The desired epsilon caprolactam crystals produced in the process of the present invention have much less other impurities than the epsilon caprolactam contained in the starting liquid. However, some of the other impurities may be present in the crystal due to incomplete release, retention, or occlusion of the solution from where the crystal is formed. Thus, the process of the present invention provides epsilon caprolactam with very high purity. By the process of the present invention it can be preferably obtained with a purity of 98% or more, more preferably 99% or more.

In embodiments that combine crystallization and adsorptive separation, the main advantages of this technique are that adsorption is best when the feed stream purity is about 10-95% epsilon caprolactam and that about 95-99.5% purity can be achieved. . Crystallization is best when the feed is as high as possible and a purity of about 99.5% or more can be achieved. Another advantage of the binding technique is that the epsilon caprolactam can be purified from solution if the concentration of epsilon caprolactam is less than or equal to the concentration of the eutectic composition. Thus, adsorption and crystallization can be used in their optimum range for a more economical process in the binding technique. In addition, in standard crystallization processes, the mother liquor is reworked or replaced to improve its yield. In the binding technique the mother liquor can be returned to the adsorption process for reuse. The bonding technique also allows for the use of higher purity and smaller adsorption columns, and smaller amounts of solvent as described above. The combination of crystallization and adsorptive separation may be effected with raffinate or extract. There is also an advantage in purifying the extract because the amount of solvent in the raffinate is generally less than the extract for purifying the raffinate. Thus, solvent recovery costs will be reduced and crystallization costs will be saved.

In a non-retentive process, ie when the desired species are not very strongly adsorbed relative to the unwanted species, the bonding process can be carried out by solution crystallization or melt crystallization. The advantage of melt crystallization is that one less distillation column is used. Of course, some chemical species crystallize out of solution much more easily than melts. As described above, Fig. 11 shows an example in which the desired kind is non-holding and melt crystallization is used. The feed is fed to a mobile bed simulation process where desorption solvent is supplied. Extracts containing unwanted species are sent to a distillation column (preferably a flash tank if the solvent has sufficient volatility) to recover the solvent. This solvent is recycled to the solvent feed stream. The raffinate containing the desired species is sent to a distillation column where the solution is recovered and returned to the solvent supply system. From the column (which is heated to keep the material fluid) the base is sent to the crystallizer and the desired product is further purified in the molten state. Crystals are removed from the product and the mother liquor is returned to the feed of the simulated moving bed.

As noted above, FIG. 12 shows an example when melt crystallization cannot be used and solution crystallization can be used. The flow chart was changed slightly and another solvent calibration column was added. It is basically the same as shown in FIG. 11 except that the raffinate containing the desired species is first sent to the crystallizer where the product crystallizes. The added column may of course also be positioned to adjust the amount of solvent in the stream to be fed to the crystallizer. The product is then dried in a distillation column while the mother liquor is sent to a distillation column and the solvent is recovered and the base from the solvent recovery column is returned to the feed to the simulation moving bed.

The following examples are provided to further refine the invention.

Example

Adsorption Example

Example 1 The first set was carried out with a solution containing 1% caprolactam and 1% 4-ethyl-2-pyrrolidinone (4-EP) in the solvent shown in Table 1 below. The resulting solution was analyzed as starting solution. 5 g of this solution were then added with 1 g of solid adsorbent, stirred at constant temperature overnight and then analyzed. Calculate the ratio of the amount of caprolactam adsorbed on the adsorbent surface and the amount of caprolactam in the solution, the amount of 4-ethyl-2-pyrrolidinone adsorbed on the adsorbent surface and the amount of 4-ethyl-2-pyrrolidinone in the solution It became. This ratio is known as the partition coefficient. A partition coefficient of 0.0 means that there is no caprolactam or 4-ethyl-2-pyrrolidinone adsorbed on the adsorbent surface. The temperature, solvent, adsorbent, and partition coefficients of caprolactam and 4-ethyl-2-pyrrolidinone are shown in Table 1 below. Adsorbents 13X, LZ-20, LZ-10, FS-115, S115 (silicalite), 4A, and 5A were obtained from Des Plains, UOP, Illinois. Adsorbent F400 was obtained from Pittsbergh, Calgon Corporation, Pennsylvania.

Temperature (℃) absorbent menstruum Distribution coefficient Caprolactam 4-EP 80 13X Cyclohexane ＞ 250 ＞ 250 80 LZ-20 Cyclohexane 154.299 177.072 80 LZ-10 Cyclohexane 32.379 37.774 80 F400 Cyclohexane 13.240 46.536 80 13X Acetone 6.043 2.932 80 LZ-20 Acetone 3.023 2.630 80 LZ-10 Acetone 1.899 1.919 80 LZ-20 Methanol 1.460 0.388 80 FS-115 Cyclohexane 1.142 69.596 80 F400 Methanol 0.864 1.541 80 S115 Acetone 0.554 1.808 80 LZ-10 Methanol 0.427 <0.01 80 FS-115 Methanol <0.01 <0.01 80 F400 Acetone <0.01 0.372 80 13X Methanol <0.01 <0.01 80 FS-115 Acetone <0.01 <0.01 80 S115 Cyclohexane <0.01 27670.0

Temperature (℃) absorbent menstruum Distribution coefficient Caprolactam 4-EP 60 13X Cyclohexane ＞ 250 ＞ 250 60 LZ-20 Cyclohexane 130.278 164.785 60 LZ-10 Cyclohexane 32.962 38.651 60 F400 Cyclohexane 17.122 63.333 60 13X Acetone 8.587 4.930 60 LZ-20 Acetone 4.396 3.910 60 FS-115 Cyclohexane 2.112 ＞ 250 60 LZ-10 Acetone 1.755 1.785 60 S115 Cyclohexane 1.637 ＞ 250 60 LZ-20 Methanol 1.443 0.741 60 F400 Methanol 1.424 1.989 60 LZ-10 Methanol 0.744 0.423 60 F400 Acetone 0.375 0.964 60 S115 Acetone 0.187 3.291 60 FS-115 Methanol 0.119 0.437 60 13X Methanol <0.01 <0.01 60 S115 Methanol <0.01 0.645 60 FS-115 Acetone <0.01 <0.01 60 4A Cyclohexane 0.181256 0.276112 60 4A Methanol 0.007061 0.002542 60 4A Acetone <0.01 <0.01 60 5A Cyclohexane <0.01 <0.01 60 5A Acetone <0.01 0.084599

As shown in Table 1A, small pore adsorbents, especially 4A and 5A, did not adsorb little or no 4-ethyl-2-pyrrolidinone. Thus, adsorbents having a pore size of 5 kPa units or less are not preferred for the selective adsorption of 4-ethyl-2-pyrrolidinone.

When larger adsorbents, especially 13X, LZ-10, and LZ-20, were used and acetone was used as the solvent, the partition coefficient of caprolactam and 4-ethyl-2-pyrrolidinone was much greater than 0.0. When methanol was used as a solvent, 4-ethyl-2-pyrrolidinone was not strongly adsorbed, but the adsorption coefficient difference between caprolactam and 4-ethyl-2-pyrrolidinone was small. When cyclohexane is the solvent, 4-ethyl-2-pyrrolidinone is very strongly adsorbed and caprolactam is less adsorbed. Thus, 4-ethyl-2-pyrrolidinone and caprolactam are adsorbed too strongly when indistinguishable or when at least about 10 GPa of voids are used.

When molecular sieves having pores of about 6 mm 3, in particular silicalite (S115) and FS115, caprolactam was not as strongly adsorbed as 4-ethyl-2-pyrrolidinone. In addition, when non-molecular bodies, especially carbon F400, are used, a large difference in the partition coefficient between caprolactam and 4-ethyl-2-pyrrolidinone is found, and caprolactam is adsorbed less strongly. This difference makes it possible to use adsorption as a separation means for the purification of caprolactam.

A second set of examples consisted of 1% caprolactam, 1% 5-methyl-2-piperidinone caprolactam (5-MP), and 1% 4-ethyl-2-pyrrolidinone in the solvents defined in Table 2 below. It was carried out with a solution containing (4-EP). The resulting solution was analyzed as starting solution. 5 g of this solution were then added with 1 g of solid adsorbent, stirred at constant temperature overnight and then analyzed. The ratio of the amount of caprolactam adsorbed to the adsorbent surface and the amount of caprolactam in the solution, the ratio of the amount of 4-ethyl-2-pyrrolidinone adsorbed to the adsorbent surface and the amount of 4-ethyl-2-pyrrolidinone in the solution, And the ratio of the amount of 5-methyl-2-piperidinone adsorbed on the adsorbent surface and the amount of 5-methyl-2-piperidinone in the solution was calculated as in the above case. The partition coefficients of temperature, solvent, adsorbent, and caprolactam, 4-ethyl-2-pyrrolidinone, and 5-methyl-2-piperidinone are shown in Table 2 below. Adsorbent S115 (silicalite) was obtained from Des Plains, UOP, Illinois. Adsorbent F400 was obtained from Pittsbergh, Calgon Corporation, Pennsylvania.

Temperature (℃) menstruum absorbent Boiling point (℃) Distribution coefficient Caprolactam 5 MP 4EP 22 Ethyl ether S115 34.6 0.557 2.803 5.218 60 Pentane S115 36.0 <0.01 4.981 4.169 60 Cyclopentane S115 49.3 2.170 10.612 18.893 60 Acetone S115 56.1 0.187 nd ^* 3.291 60 Methyl acetate S115 56.9 0.364 2.195 1.498 60 Methanol S115 64.5 <0.01 0.991 0.645 22 Methanol F400 64.5 1.756 3.256 2.155 60 Isopropyl ether S115 68.5 0.838 12.755 31.399 60 Hexane S115 68.7 2.329 8.505 17.366 60 Methylcyclopentane S115 71.8 2.159 9.149 19.259 60 Ethyl acetate S115 77.1 0.174 2.114 1.479 40 ethanol S115 78.3 <0.01 <0.01 <0.01 60 benzene S115 80.1 0.489 5.263 9.599 60 Cyclohexane S115 81.2 1.827 13.735 50.430 22 Cyclohexane F400 81.2 35.307 178.313 109.834 60 Isopropanol S115 82.2 <0.01 0.187 0.043 60 Acetonitrile S115 82.2 1.178 1.639 2 60 Propanol S115 97.1 <0.01 <0.01 <0.01 60 Propionitrile S115 98.1 61.636 82.309 70.288 60 water S115 100.0 <0.01 6.695 23.307 60 toluene S115 110.6 0.374 4.607 12.282 60 Cycloheptane S115 118.8 2.326 26.398 59.331 60 Ethylbenzene S115 136.1 1.551 6.116 19.871 60 p-xylene S115 138.5 1.257 3.755 13.372 60 Cyclooctane S115 151.1 1.567 15.215 57.949

* nd (not done) indicates not tested.

As shown in Table 2, the difference in partition coefficient between caprolactam isomer and caprolactam itself was lower than that of low molecular weight alcohol, that is, methanol. Pentane and ethyl ethers exhibited a low partition coefficient for caprolactam, while caprolactam isomers showed a higher partition coefficient. When water was used as the solvent, the difference in partition coefficient between caprolactam and its isomers was much better. Similar results were observed with activated carbon as adsorbent.

The following examples show that no solvent is needed for the adsorption portion of the process. Five grams of 90% caprolactam and 10% 4-ethyl-2-pyrrolidinone solution were contacted with one gram of S115 (silicalite) adsorbent and stirred overnight at 80 ° C. The next day the solution was analyzed and found that the S115 adsorbent adsorbed 4-ethyl-2-pyrrolidinone better than caprolactam. The results are shown in Table 3 below.

Caprolactam in solution 4-ethyl-2-pyrrolidinone in solution Starting solution 90.3% 9.7% 80 degrees Celsius (overnight) 92.1% 7.9%

Example A third set was run on a chromatographic column. A 4 foot long stainless steel cube column was used. The column is surrounded by a 1 inch tube through which water-propylene glycol (or silicone oil) flowed from a constant temperature bath for temperature control. The pump (FMI Corp.) flowed the desorption solvent through and to the column by a back pressure regulator at a space velocity of about 1 g / l / hr. The pressure was maintained such that the solvent was always liquid in the column. At the beginning of the test, the inlet of the column is with the valve open and 1% caprolactam, 1% 4-ethyl-2-pyrrolidinone and 5-methyl-2-piperidone in the solvent specified for each set of tests below. 2 g of solution consisting of About 10 mL was taken from the column outlet and put into gas chromatography for analysis.

The first set of tests was carried out on a 1/2 "diameter column containing about 90 g of S115 adsorbent. Methanol was used as the desorption solvent. The chromatographic results of operation at 80 ° C. are shown in FIG. A similar test was conducted except that methanol solution in which 1% triethylamine was dissolved in place of methanol was used.

Another set of tests was performed on a 3/8 "diameter column containing about 60 g of a S115 adsorbent again at a temperature of 80 ° C. Methanol in which 1% triethylamine was dissolved as the desorption solvent was used. It was also carried out at C. The chromatographic results are shown in Fig. 3. As the temperature is increased as shown in Fig. 3, separation may better occur.

Another set of tests was performed on a 3/8 "diameter column containing about 60 g of S115 adsorbent at 120 ° C. Acetonitrile was used as the desorption solvent. The chromatographic results are shown in Figure 4. As desorbent A similar test was conducted except that acetonitrile solution in which 1% triethylamine was dissolved in place of acetonitrile was shown in Fig. 5.

Another set of tests was performed on a 3/8 "diameter column containing about 60 g of S115 adsorbent at a temperature of 150 ° C. Acetonitrile was used as the desorption solvent. The chromatographic results are shown in FIG.

Another set of tests was carried out on a 3/8 "diameter column containing about 60 g of S115 adsorbent at a temperature of 100 ° C. Methyl acetate was used as the desorption solvent. The chromatographic results are shown in Figure 7. A similar test was conducted except that a methyl acetate solution in which 1% triethylamine was dissolved in place of methyl acetate, the chromatography results are shown in FIG.

Another set of tests was carried out on a 1/2 "diameter column containing about 80 g of carbon F400 adsorbent at a temperature of 80 ° C. Methanol was used as the desorption solvent. The chromatographic results are shown in Figure 9. As shown, activated carbon can be used as a suitable adsorbent.

Crystallization Example

The following examples are a combination of batch and semi-batch. In a batch operation, caprolactam and its isomers were placed in 4 g bottles, the contents of the bottles were heated to melt the samples, the bottles were placed at a constant temperature and crystals formed. Crystals and mother liquor samples were obtained.

Semi-batch operation is performed with a falling film crystallizer. The crystallizer used is disclosed in FIG. 6 of US Pat. No. 5,430,194, and is a 1 meter long column surrounded by a vertical tube having a kettle A, a jacketed column 1 inch in diameter. ) And means (D) for ejecting (ie circulating) liquid from the vessel to the film device (C) at the top of the falling film crystallizer. The jacket of the crystallizer was attached to supply the refrigerant E flowing with the falling film. That is, in the jacket, the falling film and the refrigerant flow down in a co-current manner. The crystallization device shown in FIG. 6 of U. S. Patent No. 5,430, 194 is similar in operation to that disclosed in U. S. Patent No. 3,621, 664.

1000 ml of concentrate prepared as described below was charged to vessel A of the falling film crystallizer. The concentrate in the vessel was circulated through column B at a constant temperature. During cooling, a small amount of heat was added to the vessel by means of a heating mantle (F) to prevent crystals from forming in the vessel. The recycled liquid was slightly cooled by circulating the refrigerant from bath (G) to cooler (H) to compensate for the heating. After crystallization was completed, the residual liquid in the vessel was emptied and the solids inside the crystallizer wall were removed partly by column B, which was slowly heated.

The purity of the crystal was measured by removing about 0.5 g of the crystal and mixing it with 1 g of water. This solution was then placed in HP 6890 gc / fid for analysis as follows. Gas chromatography was performed on a 30 mm long DB-1 column with a 5 micron film using helium as the carrier gas. The temperature program was started at 60 ° C. and held for 20 minutes by increasing to 275 ° C. at 20 ° C./min. Column pressure was maintained at 20 psi.

Crystallization processes can be divided into two types: solution crystallization and melt crystallization. Solution crystallization includes crystals produced from one solution of solute in solution. The driving force in this case is the solubility of the solute in the solvent. Melt crystallization occurs when no solvent is used. The driving force is the solute melting point and solubility of the solute in the predominant solute.

Solution crystallization

In solution crystallization a solvent must be selected. That is, particularly low temperatures (if possible) should be used to precipitate the crystals if the solute, ie caprolactam, dissolves too well in the solvent. If the solute hardly dissolves in the solvent, a large amount of solvent must be used to continue to circulate. In caprolactam crystallization, the solvent must be very volatile. This allows the solvent to be removed without much heating of the product caprolactam. If caprolactam is taken at the top of the distillation column, it will be oligomerized to reduce the purity and recovery of the product. The following examples use four solvents: acetone, methanol, acetonitrile, and cyclohexane. The four solvents are stable and do not decompose or react with the product caprolactam and are volatilized.

9 g of caprolactam and 1 g of 4-ethyl-2-pyrrolidinone were mixed with 10 g of acetone at room temperature. The mixture was cooled to 10 ° C. and about 0.1 g of crystal was removed from the mixture and analyzed. The cooling and subsequent sampling were repeated at 0 ° C, -10 ° C, and -20 ° C. The results are shown in FIG. In the first stage of crystallization, the purity increased from 90% to over 98% and 99.3% of caprolactam was obtained depending on the amount recovered.

90% caprolactam solution (acetone-free base) was added 4.5 g of caprolactam with 0.5 g of a mixture of 4-ethyl-2-pyrrolidinone (4EP) and 5-methyl-2-piperidinone (5MP) acetone. Prepared by addition to 5 g. This mixture was dissolved and left at the desired temperature to allow crystals to form. As described above, the crystal specimens were removed and analyzed. The results are shown in Table 4. The percentages in Table 4 are based on area calculations from gas chromatography.

Temperature (℃) % Recovery % Caprolactam % Caprolactam for 4EP % Caprolactam for 5 MP 22 (Departure solution) - 93.23 97.65 95.37 10 55 98.33 99.69 98.64 5 73 98.25 99.68 98.56 0 89 98.23 99.65 98.56

Purity increased from 93% (acetone free base) to more than 98% in the first stage. Furthermore, it recrystallized and washed the decision without any effort. These steps will lead to higher purity per step.

9 g of caprolactam and 1 g of 4-ethyl-2-pyrrolidinone in 10 g of acetone were mixed into a solution and placed in an enclosed container. The jacket temperature was adjusted to 0 ° C. by a water-propylene glycol barrel and the vessel was placed in an ultrasonic cleaning vessel. The starting materials and the obtained crystals were analyzed by gas chromatography and shown in Table 5.

% Caprolactam (acetone free base, gc area calculation) Starting material 90.7% decision 96.6%

In contrast to one large crystal on the bottle, crystal mixing appeared to have very small particles mixed by ultrasound. Thus, such crystallization can occur in the presence of ultrasound.

5 g of caprolactam and methanol (in various ratios) were added and the amount of crystals and mother liquor remaining after equilibration was measured. The results are shown in FIG. 16 shows that solubility is reduced when the temperature of the mixture is lowered. 16 also shows that samples with caprolactam of about 75-80% or less require very low temperatures to crystallize.

The three isomers, 4-ethyl-2-pyrrolidinone, 5-methyl-2-piperidone, and 4-methyl-2-piperidinone, in almost equal proportions, as one isomer mass, caprolactam solution Was prepared in 50-99% purity (methanol free base). The solution then contains 80% solute in 20% methanol. The solution was then cooled and the crystals recovered and analyzed. More than 88% caprolactam samples were collected at 0 ° C and samples with 91% caprolactam were collected at -20 ° C. Samples with 50% caprolactam were cooled to -60 ° C (dry ice-acetone bath) but the solution never became a solid-liquid solution. The result of comparing the purity of the collected crystals with the starting solution is shown in FIG. In all cases, the crystals had a higher purity than the starting solution.

687 g of about 99.3% caprolactam with 0.7% mixture of the remaining three isomers, 4-ethyl-2-pyrrolidinone, 5-methyl-2-piperidone, and 4-methyl-2-piperidone And 172 g solution of methanol were placed in the falling film crystallizer as described above. The mixture was heated to 28 ° C. in a vessel and a pump heat exchanger. The temperature of the falling film slowly cooled to 0 ° C. when crystals began to form on the inner surface of the crystallizer wall. After the crystals had grown to the extent that they started blocking the flow to the tube, the crystals were recovered by removing the vessel and slowly heating the crystallizer wall to about 75 ° C. As the crystals melted and flowed down the column, they were collected in vials for analysis. Table 6 shows the caprolactam concentration in the crystals and the concentration of caprolactam (methanol free base) in the starting solution.

% Yield specimen water Start 0.992531 1.09% First 0.995983 3.33% 2nd 0.997721 16.00% 3rd 0.998889 5.43% 4th 0.999715

The purity of caprolactam was increased from 99.3% to about 99.9% (weight average).

As in methanol, the first step here was to produce a caprolactam solubility curve with acetonitrile. This is shown in FIG. 18. 18 shows that acetonitrile does not dissolve caprolactam as quickly as methanol.

The three isomers, 4-ethyl-2-pyrrolidinone, 5-methyl-2-piperidone, and 4-methyl-2-piperidinone, in almost equal proportions, as one isomer mass, caprolactam solution Was prepared in 50-99% purity (acetonitrile free base). The solution then contains 70% solute in 30% acetonitrile. The solution was then cooled and the crystals recovered and analyzed. More than 88% caprolactam samples were collected at 0 ° C and samples with 91% caprolactam were collected at -20 ° C. Samples with 50% caprolactam were cooled to -60 ° C (dry ice bath). The result of comparing the purity of the collected crystals with the starting solution is shown in FIG. In all cases, the crystals had a higher purity than the starting solution.

700 g of about 99.3% caprolactam with 0.7% mixture of the three remaining isomers, 4-ethyl-2-pyrrolidinone, 5-methyl-2-piperidone, and 4-methyl-2-piperidone And a 300 g solution of acetonitrile were placed in a falling film crystallizer as described above. The mixture was heated to 45 ° C. in a vessel and a pump heat exchanger. The temperature of the falling film slowly cooled to 0 ° C. when crystals began to form on the inner surface of the crystallizer wall. After the crystals had grown to the extent that they started blocking the flow to the tube, the crystals were recovered by removing the vessel and slowly heating the crystallizer wall to about 75 ° C. As the crystals melted and flowed down the column, they were collected in vials for analysis. Table 7 shows the caprolactam concentration in the crystals and the concentration of caprolactam (acetonitrile free base) in the starting solution.

% Yield specimen water Start 0.992889 0.35% First 0.996594 5.11% 2nd 0.997589 26.73% 3rd 0.998982 1.77% 4th 0.999799

The purity of caprolactam was increased from 99.3% to about 99.9% (weight average).

504.2 g of about 99.3% caprolactam with 0.7% mixture of the remaining three isomers, 4-ethyl-2-pyrrolidinone, 5-methyl-2-piperidinone, and 4-methyl-2-piperidone And a 504.2 g solution of cyclohexane were placed in a falling film crystallizer as described above. The mixture was heated to 80 ° C. in a vessel and a pump heat exchanger. The temperature of the falling film slowly cooled to 20 ° C. when crystals began to form on the inner surface of the crystallizer wall. After the crystals had grown to the extent that they started blocking the flow to the tube, the crystals were recovered by removing the vessel and slowly heating the crystallizer wall to about 75 ° C. As the crystals melted and flowed down the column, they were collected in vials for analysis. Table 8 shows the caprolactam concentration in the crystals and the concentration of caprolactam (cyclohexane free base) in the starting solution.

% Yield specimen water Start 0.993468 0.36% First 0.996411 19.8% 2nd 0.99682 17.52% 3rd 0.996131 7.47% 4th 0.998258

The purity of caprolactam was increased from 99.3% to about 99.7% (weight average).

Melt crystallization

10 g of caprolactam and 1 g of 4-ethyl-2-pyrrolidinone were mixed together at 75 ° C. in a 4 g bottle to form a liquid mixture. The liquid mixture was slowly cooled to about 50 ° C. until it was completely solid and then slowly heated to 66 ° C. until the solid was a mixture of solid crystals and liquid (mother liquor). Crystals were removed and analyzed. The results are shown in Table 9.

% Caprolactam (acetone free base, gc area calculation) Starting solution 90.9 wt% decision 96.0 area%

Crystals produced by the melting method showed a substantial increase in purity. Thus, caprolactam can be purified by melting.

Approximately 99.4% caprolactam solution 845 with a 0.6% mixture of the three remaining isomers, 4-ethyl-2-pyrrolidinone, 5-methyl-2-piperidone, and 4-methyl-2-piperidone g was placed in the falling film crystallizer as described above. No solvent was used in this experiment. The mixture was heated to 90 ° C. in a vessel and a pump heat exchanger. The temperature of the falling film slowly cooled to 25 ° C. when crystals began to form on the inner surface of the crystallizer wall. After the crystals had grown to a degree that would stop the flow to the tube, the crystals were recovered by removing the vessel and slowly heating the crystallizer wall to about 75 ° C. As the crystals melted and flowed down the column, they were collected in vials for analysis. Table 9 shows the caprolactam concentrations in the crystals and the caprolactam concentrations in the starting solution.

% Yield specimen water Start 0.994178 0.12% First 0.996443 0.26% 2nd 0.991256 8.16% 3rd 0.995472 23.51% 4th 0.998602 9.53% 5th 0.998834

The purity of caprolactam was increased from 99.4% to about 99.8% (weight average).

Although the present invention has been described by the specific embodiments described above, the scope of the invention is therefore not to be construed as limiting, but rather the invention covers the general scope as set forth above. Various modifications and embodiments can be made within the scope of the present invention.

Effects of the Invention

The present invention provides at least one selected from the group consisting of 4-ethyl-2-pyrrolidinone, 5-methyl-2-piperidinone, 3-ethyl-2-pyrrolidinone, and 3-methyl-2-piperidinone. It has the effect of providing a process for separating epsilon caprolactam in a feed mixture consisting of additional hydrocarbons and epsilon caprolactam.

Claims (10)

Adsorbent and 4-ethyl-2-pyrrolidinone, 5-methyl-2-piperdinone, 3-ethyl-2-pi under adsorption conditions One or more epsilon caprolactam isomers or octahydrophenazines selected from the group consisting of 3-ethyl-2-pyrrolidinone, and 3-methyl-2-piperdinone; contacting a feed mixture consisting of octahydrophenazine) and epsilon caprolactam, selectively adsorbing the epsilon caprolactam isomer or octahydrophenazine to substantially exclude the epsilon caprolactam, and contacting the feed mixture from contact with an adsorbent. Removing the non-adsorbed portion, and recovering the high purity epsilon caprolactam. The method of claim 1 wherein the adsorbent is selected from activated carbon, molecular sieve carbon, molecular sieve, and zeolite. The method of claim 1, further comprising recovering the epsilon caprolactam isomer or octahydrophenazine by desorption under desorption conditions. The method of claim 1, wherein the separation is performed by a simulated moving bed flow scheme. 5. The method of claim 4, wherein the simulated moving bed configuration uses countercurrent flow or co-current flow. The method of claim 1, further comprising solution crystallizing or melt crystallizing the high purity epsilon caprolactam. Under adsorption conditions, the adsorbent is selected from the group consisting of 4-ethyl-2-pyrrolidinone, 5-methyl-2-piperidinone, 3-ethyl-2-pyrrolidinone, and 3-methyl-2-piperidinone. Contacting a feed mixture consisting of one or more epsilon caprolactam isomers or octahydrofenazine and epsilon caprolactam, selectively adsorbing the epsilon caprolactam to substantially exclude the epsilon caprolactam isomer or octahydrophenazine Removing the nonadsorbed portion of the feed mixture from contact with the adsorbent, and recovering the high purity epsilon caprolactam by desorption under desorption conditions. 8. The method of claim 7, further comprising solution crystallizing or melt crystallizing the high purity epsilon caprolactam. 4-ethyl-2-pyrrolidinone, 5-methyl-2-piperidone, which is a mixture having a composition of the compositional region in which only epsilon caprolactam crystallizes when the solubility limit of epsilon caprolactam in solution is exceeded , One or more epsilon caprolactam isomers selected from the group consisting of 3-ethyl-2-pyrrolidinone, and 3-methyl-2-piperidinone or non-eutectic consisting of octahydrophenazine and epsilon caprolactam Providing an initial solution containing the mixture, and beyond the solubility limit of epsilon caprolactam to form crystalline epsilon caprolactam containing relatively less epsilon caprolactam isomers or octahydrophenazine than were present in the initial solution. Maintaining the solution at a degree and at a temperature above the eutectic point of the mixture. One selected from the group consisting of 4-ethyl-2-pyrrolidinone, 5-methyl-2-piperidinone, 3-ethyl-2-pyrrolidinone, and 3-methyl-2-piperidinone A method for separating epsilon caprolactam from the above-mentioned epsilon caprolactam isomer or a mixture consisting of octahydrophenazine and epsilon caprolactam. 10. The method of claim 9, further comprising adsorbing and separating the solution.