FI63065B - FOERFARANDE FOER UPPDELNING AV EN BLANDNING INNEHAOLLANDE EN KETOS OCH EN ALDOS I SINA BESTAONDSDELAR - Google Patents

Description

ΓΒΐ rt «ANNOUNCEMENTS8U 7 n, 'LBi (11) UTLÄGGNI NGSSKRIFT 6 3 0 6 5 2¾¾ c (45) v':; vnfty 11 34 2933 'w' (51) Kv.lk3 / IncCI.3 C 13 K 5 / 00

FINLAND —FINLAND (21) fwntt «» kemui —Mwwwseiimiig T7166T

(22) Htlcamlsptivi - AnaBitninpdif 25.05.77 (23) ΑΜαιρΜνλ — GIM | h «ud« g 25.05.77 (41) Tullut JulklMlul - Bllvlt offuntllf 28.11.77

Patent and sleigh ^ admired. / 44) Nihtivilcipanon j. kuLJulk * ™ p ™ .- ”'po

Patent- och regitterstyreleen '7 Anuektn uti «| d och utUkrHtwi pubiicend ^ ±. ± ^ .q <- (32) (33) (31) Pyydetty utuolkeui — B« | ftrd priorttet 27.05 · 76 27.O5.76, O6 .O7.76 USA (US) 690768, 690769, 702831 Proven — Styrkt (71) U0P Inc., Ten U0P Plaza - Algonquin & Mt. Prospect Roads, Des Plaines, Illinois, USA (72) Richard William Neuzil, Downers Grove, Illinois, James William Priegnitz, Elgin, Illinois, USA (7 * 0 Berggren Oy Ah (5 ^) Method for Ketosis and to subdivide the mixture containing aldose into its constituents - Förfarande för uppdelning av en blandning innehällande en ketos och en Aldos i sinä heständsdelar _____ _ ________

The present invention relates to the separation of monosaccharides by adsorption in a solid bed. More specifically, the invention relates to a process for separating one component from a mixture comprising ketose and aldose, which process comprises using an adsorbent formed by crystalline aluminosilicate which selectively adsorbs either ketose or aldose from a mixture containing them.

It is well known that crystalline aluminosilicates can be used to separate certain hydrocarbon species from their mixtures, such as to separate normal paraffins from branched chain paraffins and to separate olefinic hydrocarbons from paraffinic hydrocarbons. X and Y zeolites have been used in methods for separating individual hydrocarbon isomers. Thus, adsorbents containing X and Y zeolites have been used to separate alkyl trisubstituted benzene isomers (U.S. Patent 3,114,782), to separate alkyl tetrasubstituted monocyclic aromatic isomers (U.S. Patent 3,864,416), and to specific alkylated alkyls. (U.S. Patent 3,668,267). Perhaps the most widely used methods are those used to separate paraxylene from a mixture of C 8 aromatic compounds (U.S. Patents 3,558,730, 3,558,732, 3,626,020, 3,663,638 and 3,734,974).

2 63065

The present invention does not relate to the separation of hydrocarbons but more particularly to the separation of monosaccharides. We have found that adsorbents consisting of certain zeolites containing one or more specific cations at interchangeable cationic bonding sites have a selective adsorption capacity between ketose and aldose, while certain other cation exchange zeolites have a certain selectivity enabling the separation of ketose (or aldose) from a mixture containing ketose and aldose by selective solid bed adsorption. A specific embodiment of the method according to the invention relates to the separation of fructose from a mixture containing fructose and glucose.

Fructose is considered to be the most soluble and sweetest type of sugar. Compared to sucrose, which has a sweetness of 1.0, the relative sweetness of fructose is about 1.4 and that of glucose 0.7. It has been reported in the literature that one use of fructose in pure form is a source of calories for patients who need to be fed a blood vessel, whereas glucose is not suitable for this purpose. Although fructose is common in nature, methods for separating high-purity fructose are more difficult than methods for obtaining high-purity glucose. Namely, high purity glucose can be easily prepared by hydrolyzing starch with mineral acids at elevated temperature followed by purification and crystallization, while one method of obtaining high purity fructose involves hydrolysis of sucrose, separation of insoluble lime-fructose complex, acidification, acidification salts, separation of cationic and anionic impurities, concentration of the resulting solution and finally crystallization of fructose. Extensive attempts have been made to prepare fructose by hydrolyzing fructose-containing polysaccharides extracted from Jerusalem artichoke. Attempts have also been made to use several methods to separate glucose from invert sugar, leaving fructose, such as by forming insoluble benzidine derivatives of glucose and sodium chloride addition products of glucose, but these methods have not been practicable.

3 6 30 6 5

It is therefore an object of the present invention to provide a process for separating ketose and aldose from a mixture containing both of these ingredients in order to obtain ketosis and aldose products containing ketose and aldose in a higher concentration than in the starting material mixture. More specifically, it is an object of the invention to provide a process for the preparation of concentrates of fructose and glucose from a starting material mixture, such as an invert sugar solution, or from high fructose corn syrup containing the two ingredients.

The present invention is thus directed to a process for dividing a mixture containing ketose and aldose into its components by contacting the mixture with an adsorbent under adsorption conditions, then contacting the adsorbent containing the adsorbed component with water and recovering the desorbed component, contacting the mixture with aluminium. which for selectively adsorbing ketosis is (1) an X-zeolite at the cation-exchangeable positions containing at least one cation which is sodium, potassium, barium or strontium, or (2) a Y-zeolite at the exchangeable cation positions containing at least one cation at the cation-exchange site. ammonium, sodium, potassium, calcium, strontium, barium, or a combination thereof, or which for selectively adsorbing aldose is (3) X-zeolite containing potassium cations at interchangeable cationic positions.

The adsorption can be carried out at a temperature in the range of about 20 to 100 ° C and a pressure of about 1 to 17.5.

The method can be applied to the distribution of a mixture in which ketose and aldose are hexoses, in which case, for example, ketosis is fructose and aldose is glucose or dextrose.

The mixture to be treated may be, for example, an invert sugar solution or a high-fructose corn syrup.

The concentration of ketose and aldose in the raw material mixture can then be about 0.5-30% by weight, preferably 1-15% by weight.

The particle size of the X or Y zeolite may be about 0.25-1.20 mm, preferably about 0.30-1.60 mm.

According to one embodiment of the invention, the process can be carried out as a stepwise process by (a) maintaining a net flow of liquid through a column containing X or Y zeolite in one direction, the column containing at least three series-connected zones, each operating separately, the end zones being connected by a continuous flow (b) maintaining an adsorption zone comprising an X or Y zeolite located between the upstream boundary of the liquid inlet to be treated and the downstream boundary of the raffinate outlet; (c) maintaining the column immediately upstream of the boundary formed by the zeolite; from the adsorption zone, a purification zone located between the upstream boundary formed by the extract outlet and the downstream boundary formed by said inlet of the liquid to be treated, (d) the column is maintained immediately upstream of the purification zone; a desorption zone located between the upstream boundary of the water inlet and the downstream boundary of said extract outlet; water is removed and the component selectively adsorbed to it in the previous section is displaced from the adsorbent in the zone, (g) an extract stream comprising water and selectively adsorbed component is removed from the desorption zone, (h) at least a portion of the removed extract stream is and forming a product stream comprising a selectively adsorbed component, (i) periodically passing in a column containing X or Y zeolite downstream of the liquid flow direction through the adsorption zone to be treated a liquid inlet, a raffinate outlet line, a water inlet and an extract outlet, to move different zones in the X or Y zeolite and to provide an extract 5 63065 effluent and a raffinate effluent.

At least a portion of the raffinate effluent can be passed to a separation device and separates at least a portion of the water to provide a raffinate product that contains less water.

Immediately upstream of the desorption zone, a buffer zone may be maintained between the downstream boundary formed by the water inlet and the upstream boundary formed by the raffinate outlet.

Other objects and embodiments of the invention include details regarding feed compositions, adsorbents, desorbents, and processing conditions, all of which are discussed below in the description of the invention.

The definitions used in this specification will be explained in more detail in connection with the operation, aims and advantages of the method.

A feed mixture is a mixture containing one or more extract components and one or more raffinate components which are separated by the process of the invention. The term "feed stream" means a feed mixture stream that passes through the adsorbent used in the process.

6 63065 An "extract component" is a compound or species of compounds that selectively adsorbs to an adsorbent, while a "raffinate component" is a compound or species of compounds that is selectively adsorbed. In this method, when ketose is the extract component, the aldose is the raffinate component and vice versa. The term "desorption material" generally refers to a material capable of desorbing an extract component. The term "desorption stream" or "desorbent input stream" means the flow with which the desorption material enters the adsorbent. "Raffinate stream" or "raffinate effluent" means a stream that removes a raffinate component from an adsorbent. The composition of the raffinate stream can range from substantially 100% desorbent to substantially 100% raffinate constituents. The term "extract stream" or "extract effluent" means a stream by which an extract material desorbed by an desorbent is removed from the adsorbent. Similarly, the composition of the extract stream can range from substantially 100% desorbent to substantially 100% extract components. At least a portion of the extract stream, and preferably at least a portion of the raffinate stream, is passed from the separation treatment to separation equipment, usually fractionation equipment, in which the desorbent is separated to obtain an extract product and a raffinate product. The terms "extract product" and "raffinate product" refer to products prepared by a process comprising the extract component and the raffinate component at higher concentrations, respectively, than those present in the extract stream and the raffinate stream. Although it is possible to prepare a high purity ketose product and an aldose product (or both) with a high recovery rate, it should be noted that the adsorbent never completely adsorbs the extract component and the adsorbent does not completely adsorb the raffinate component. As a result, varying amounts of raffinate component may be present in the extract stream, and similarly, varying amounts of extract component may be present in the raffinate stream depending on the treatment conditions used. The extract and raffinate streams are then further separated from each other and from the feed mixture using a series of concentrations of the extract component and the raffinate component in each particular stream. More specifically, for example, the concentration ratio of ketose to less selectively adsorbed aldose is lowest in the raffinate stream, followed by the highest in the feed mixture and highest in the extract stream. Similarly, the concentration ratio of the less selectively adsorbed aldose to the more selectively 7,63065 adsorbed ketose is highest in the raffinate stream, followed by the highest in the feed mixture, and lowest in the extract stream.

The term "selective pore volume" of an adsorbent refers to the volume of adsorbent that selectively adsorbs the extract component from the feed mixture. The term "non-selective pore volume" of the adsorbent means that volume of the adsorbent which does not selectively adsorb the extract component from the feed mixture. This volume volume comprises those pores of the adsorbent which do not have adsorption bonds and the pore volumes between the adsorbent particles. Selective pore volume and non-selective pore volume are generally expressed in volume amounts and are relevant in determining the appropriate flow rates of liquid required to be introduced into the treatment zone to provide effective treatment using a given volume of adsorbent. As the adsorbent flows into the treatment zone (described in more detail below) used in one embodiment of the present invention, its non-selective pore volume, together with its selective pore volume, transports the liquid to this zone. The non-selective pore volume is used to determine the amount of liquid that must flow into the same zone upstream of the adsorbent to displace the liquid in the non-selective pore volume. If the flow rate of the liquid entering the zone is lower than the non-selective pore volume of the adsorption material flowing into this zone, a net flow of liquid into the zone takes place in the adsorbent. Because this net flow is based on the liquid present in the non-selective pores of the adsorbent, in most cases it comprises less selectively attached feedstock components. The selective pore volume of the adsorbent may, in certain cases, adsorb portions of the raffinate material from the liquid surrounding the adsorbent, because in certain cases there is competition between the extract material and the raffinate material at the selective pore volume adsorption sites. If a large amount of raffinate material relative to the extract material surrounds the adsorbent, the raffinate material may be such that it is adsorbed by the adsorbent.

Such mixtures which can be fed in the process according to the invention are those containing ketose and aldose, and more particularly and preferably aqueous solutions of ketose and aldose. Although the feed mixture may contain more than one ketose and more than one aldose, usually the feed mixture will contain one ketose and one aldose each at concentrations of about 0.5-30% by weight and more preferably about 1-15% by weight. The method can be used to separate ketopentose from aldopentose, but more commonly it is used to separate ketohexose from aldhexose. Well known ketohexoses include fructose (levulose) and sorbose, well known aldohexoses include glucose (dextrose), mannose and galactose, while lesser known aldohexoses include gulose, houseose, allose, altrose and idose. Preferred hexose-containing mixtures to be treated are aqueous solutions of invert sugar formed by the hydrolysis of sucrose with acidic substances to equimolar amounts of fructose and glucose. Other preferred compositions to be treated are aqueous solutions which are high in corn syrup fructose (usually about 40-45%) and which have been prepared by enzymatic isomerization of glucose solutions.

The desorbent used in the process of the invention must meet several requirements. First, it must displace the extract component from the adsorbent at a reasonable mass flow rate without itself adsorbing so strongly that this adversely prevents the extract component from displacing the desorbed material in the next adsorption cycle. It should also be more selective for all extract components with respect to the raffinate component than it is as a desorbent with respect to the raffinate component. It must be compatible with the specified adsorbent and the specified feed mixture. It must not reduce or impair the critical selectivity of the adsorbent for the extract component and the raffinate component. It must be easily separable from the feed mixture as it passes for further processing, preferably by distillation. Since raffinate and extract products are nutrients intended for human consumption, desorbents must also be non-toxic. Finally, desorbents must be readily available at a reasonable cost. Water meets these requirements and is a suitable desorbent in the process of the invention.

It is known that certain properties of the adsorbent are necessary for the successful use of a selective adsorption method. Such features include e.g. the following: the adsorption capacity of a given volume of the extract component per volume of adsorbent and also a sufficiently high adsorption and desorption rate. Another necessary property is the ability of the adsorbent to separate the components of the mixture to be treated, i. that the adsorbent has a selective adsorption capacity (B) over one component over another. By selectivity (B) used in this specification is meant the ratio of the two components of the adsorbed phase to the ratio of the same two components in the non-adsorbed phase under equilibrium conditions. Relative selectivity is shown in Equation 1 below.

Equation 1

Selectivity - (E) = (volume * C / volume% D) K

(volume% C / volume% D) u wherein C and D are the two components of the substance to be treated, calculated as volume%, and A and U denote adsorbed and non-adsorbed phases, respectively. If the selectivity between the two components approaches 1.0, the adsorbent will not adsorb one component to the other. When comparing the selectivity of the adsorbent with respect to component C over component D, a value (B) greater than 1.0 indicates that the adsorbent mainly adsorbs component C. If the value (B) is less than 1.0, this indicates that component D is mainly adsorbed, leaving more component C in the non-adsorbed phase and more component D in the adsorbed phase. Ideally, the selectivity of desorbents should be about 1 or less for all extractants so that all extractants can be extracted in one category and all raffinate components clearly remain. remaining in the raffinate stream. Although it is theoretically possible to separate the extract component from the raffinate component when the selectivity of the adsorbent for the extract component relative to the raffinate component is greater than 1, it is preferred that this selectivity approach 2. As the relative volatility, the greater the selectivity separation.

A dynamic test apparatus is used to test various adsorbents using a given feed mixture and desorbent to measure the selectivity and rate of adsorption of the adsorbent. Such equipment may consist of an adsorption chamber with a volume of about 70 ml, with inlet and outlet points at opposite ends, a temperature and pressure control device to ensure a predetermined constant pressure, and analytical equipment such as refractometers, polarimeters and chromatographic devices. to perform a quantitative or qualitative determination of several components. The pulsation test performed with this equipment can be used to determine the selectivity and other important values of different adsorption systems. The adsorbent is filled to equilibrium using a specified desorbent material. At an appropriate time, a feed pulse containing a specified concentration of analyte and a specified ketose or aldose, or both, all diluted with a desorbent, is injected over several minutes. The flow of desorbent is stopped and the analyte and ketose and aldose are eluted by liquid-solid chromatography. The effluent is analyzed to determine the peaks of the corresponding constituents formed.

Based on the data obtained from this experiment, the activity of the adsorbent based on the pore volume, the capacity number of the extract component, the selectivity of the second component with respect to the other component and the desorption rate of the extractant component of the desorbent can be calculated. The capacity number of the extract component can be described as the distance between the center of the peak of the extract component and the peak of the analysis component or some other reference point. It is expressed as the volume in cubic centimeters of desorbent pumped during this period, represented by the distance between the peaks. By selectivity of the extract component (B) for the raffinate component is meant the distance between the peak of the extract component and the centers of the analyte peak (or other reference point) compared to the corresponding distance between the raffinate component peak and the analyte peak centers. The rate of exchange of the extract component with the desorbent can generally be characterized by the width of the peak arcs in the middle of them. The narrower the width of the peak, the greater the desorption rate. The desorption rate can also be defined as the distance between the center of the peak arc of the analyte and the center of the freshly desorbed extract component. This distance is again the volume of desorbent pumped during this period.

11 6 30 6 5

Adsorption systems can also be studied by actual experiments in a continuous countercurrent liquid-solid contact device, as described, for example, in U.S. Patents 2,985,589 and 3,706,812. Further details of the adsorbent research technique can be obtained from A.J. deRosset'in, R.W. Neuzil'in, D.J. Korous and D.H. Rosback, "Separation of Cg Aromatics by Adsorption," presented at the American Chemical Society, Los Angeles, California, March 28-April 2, 1971.

Adsorbents used in the process of the invention include crystalline aluminosilicate structures in which alumina and silica tetrahedra are bonded together to form an open three-dimensional network structure. Crystalline aluminosilicates are often referred to as "molecular sieves", especially if the separation they perform depends mainly on differences in the size of the molecules treated. the method according to the invention does not, the term "molecular sieves" is, however, particularly suitable since the separation of ketose from an aldose obviously depends on the other side of the differences in electrochemical attraction between the adsorbent and the ketose and an aldose, and on the other side of the adsorbent rather than the physical size differences in the molecules.

In hydrated form, the crystalline aluminosilicates generally comprise zeolites of the formula I

M2 ^ n0: Al2O3: wSiO2: yH2O I

wherein "M" is a cation that balances the electric valence of the aluminum-centered tetrahedron and is generally referred to as an alternating cationic bond, "n" is the valence of the cation, "w" means moles of SiO 2 and "y" means moles of water. The cation "M" may be monovalent, divalent or trivalent or a mixture thereof.

Zeolite X in hydrated or partially hydrated form can be represented by molar oxides, as shown in formula II:

(0.9 + 0.2) M 2 /, nO: Al 2 O 3: (2.5 + 0.5) SiO 2: yH 2 O II

63065 12 wherein "M" is at least one cation having a valence not greater than 3, n is the valence of "M", and the value of y is always about 9 depending on the quality of "M" and the degree of hydration of the crystal. As shown in formula II, the SiO 2 / Al 2 O 3 molar ratio of the X zeolite is 2.5 + 0.5.

The cation "M" may be one or more of various cations, such as a hydrogen cation, an alkali metal cation or an alkaline earth metal cation, or any other selected cation, and is generally referred to as an exchangeable cationic position. In the preparation of X-zeolite, the cation "M" is usually predominantly sodium and the zeolite is therefore referred to as sodium X-zeolite. However, depending on the purity of the reactants used in the preparation of the zeolite, other cations mentioned above may also be present as impurities.

Y-zeolite in hydrated or partially hydrated form can be represented in the same way as molar oxides as shown in formula III: (0.9 + 0.2) M2 / n °: A12 ° 3: wS iO2: yH2 ° 111 where "M" is at least one cation having a valence not greater than 3, "n" is the valence of "M", "w" is greater than about 3 to about 6, and "Y" is always about 9 depending on the quality of "M", and In Y-zeolites, the molar ratio of SiO 2 / Al 2 O 3 may thus be about 3 to 6. The cation "M" may be one or more of several cations, as in the case of the X-zeolite, but in the preparation of the Y-zeolite the cation "M" is also usually Y-zeolite, which contains mainly sodium cations at the positions of the exchangeable cations, is therefore in this context called sodium Y-zeolite.

Those cations located at the positions of the exchangeable cations in the zeolite can be replaced by other cations by well known ion exchange methods, for example by contacting the zeolite or zeolite-containing base material with an aqueous solution of a cation or soluble salt of such cations. After the desired ion exchange has taken place, the zeolite is removed from the aqueous solution, washed and dried to the desired water content. By such methods, sodium cations and any other cations which may be present at alternating positions as impurities in sodium-X or sodium-Y zeolites can be partially or mainly replaced by other cations.

13 '63065

The term "base material" as used herein means an X and Y zeolite-containing material that can be used to prepare the special adsorbents described below. The zeolite is usually present in the base material in amounts ranging from about 75 to 98% by weight, based on the volatile matter-free composition of the base material. Volatile-free compositions are generally understood to mean the base material after it has been calcined at 900 ° C for the purpose of removing volatiles. The remainder of the base material is generally an amorphous material, such as silica, alumina, or a silica-alumina mixture, or various compounds, such as clay, which material is present in intimate admixture with small particles of zeolite. This amorphous material may be an excipient in the preparation of the X or Y zeolite (e.g., by intentionally performing incomplete purification of the zeolite during its preparation), or may be added to a relatively pure X or Y zeolite, but in either case is intended to act as a binder and help form or to agglomerate the hard crystalline particles of the zeolite. The base material is usually in the form of particles, such as extrudates, aggregates, tablets, pills, macrospheres or granules, prepared by grinding any of the above substances to the desired particle size. The particle size of the adsorbent used in the process of the invention is preferably in the range of about 0.25 to 1.2 mm and most preferably about 0.3 to 0.6 mm. Examples of suitable base materials that can be used in preparing the adsorbents used in the process of the invention include "Molecular Sieves 13X" and "SK-40", both of which are manufactured by Linde Company, Tonawanda, New York. The former product contains X-zeolite and the latter Y-zeolite.

It has been found that the X and Y zeolites containing said cations at the positions of the alternating cations meet the above requirements and are therefore suitable for use in this process. Some of the suitable zeolites contain essentially only a certain type of cation at the positions of the variable cations, while some others are substantially completely exchanged using selected cation pairs. The zeolite can be considered to be substantially completely exchanged when the residual sodium content of the zeolite after ion exchange is less than about 2% by weight of Na 2 O. In particular, we have found that the adsorbents formed by X-zeolite, which contain sodium cations, barium cations or strontium cations at the interchangeable cation positions, all have selectivity for ketosis over aldose. Other similar adsorbents with selectivity are X-zeolites which contain barium and potassium as the cation pair or barium and strontium at the interchangeable cationic positions. The zeolite X adsorbent having barium and potassium at the interchangeable cationic positions preferably has a weight ratio of barium to potassium in the range of about 1: 1 to 100: 1, and more preferably in the range of about 1: 1 to 10: 1. The zeolite X adsorbent having barium and strontium at alternating cationic positions preferably has a weight ratio of barium to strontium in the range of about 1: 1 to 15: 1, and more preferably about 5: 1 to 15: 1.

On the other hand it has been found that the X zeolite contains potassium changing the positions of the cation is suitable for use in the method of the invention due to its selectivity, with respect to the aldose compared to ketosis. Such an adsorbent can be prepared using substantially complete "Molecular Sieves 13X" (Na-X zeolite) anion exchange with potassium cations (usually using KCl solution), washing the treated material with water to remove excess ion exchange solution, and drying the adsorbent to less weight loss. about 10% by weight, by annealing (LOI) at 900 ° C.

It has further been found that Adsorbents which form a Y zeolite and which contain, at alternating cationic positions, at least one cation selected from the group consisting of ammonium, sodium, potassium, calcium, strontium, barium and combinations thereof are suitable for use in the present invention. in the method due to their selectivity for ketosis over aldose. These zeolites are exchanged substantially completely with the selected cation or cations. A very preferred adsorbent is a base material containing Y zeolite and an amorphous material with calcium cations at alternating cationic positions.

It is surprising that it is not possible to know in advance the suitability of an adsorbent for use in the process according to the invention. The adsorbents formed by many X- and Y-zeolites and amorphous material do not in fact have selectivity for ketose or aldose and are therefore not suitable for use in the process according to the invention. For example, a Y zeolite containing ammonium cations at variable cation positions has a selectivity for ketose over aldose, but a Y zeolite with hydrogen cations at variable cation positions has no selectivity for ketose or aldose.

Such an X-zeolite with potassium at the variable cation positions appears to be unique among X-Zeolites in its ability to selectively adsorb aldose over ketosis. An adsorbent formed by either a cesium-exchanged X or Y zeolite has no selectivity for aldose or ketosis. Potassium-exchanged Y zeolite, in contrast to potassium-exchanged X zeolite, has selectivity for ketosis over aldose. Adsorbents consisting of X-zeolites with either barium or strontium or barium and potassium cations at variable cation positions have selectivity for ketose over aldose, while X-zeolite adsorbents with magnesium at either cationic or cationic positions have either calcium there is no selectivity for aldose or ketosis.

With respect to Y-zeolite adsorbents having Group IIA metals of the Periodic Table of the Elements at variable cationic positions, all those containing calcium, strontium or barium have selectivity over ketose over aldose, but magnesium-containing Y zeolites have no selectivity for it. aldose. Of suitable adsorbents formed by Y zeolites having Ca, Sr or Ba cations at alternating cationic positions, we have found that the adsorbent formed by Y zeolite and having Ca cations at these positions is superior to those adsorbents which contain Sr or Ba cations. The reason why some Adsorbents are suitable for use in the process of the invention and some are not is not entirely clear.

6 30 6 5 16

The adsorbent can be used in the form of a dense solid layer which is brought into alternating contact with the mixture to be treated and the desorption materials. In the simplest embodiment of the invention, the adsorbent is used in the form of a single static layer, whereby the process is only semi-continuous. According to another embodiment, two or more permanent layers can be used which are in contact with each other via suitable valve devices so that the mixture to be treated is passed through one or more adsorption layers, while desorption materials can pass through one or more layers in series. The flow of the mixture to be treated and the desorbents can take place either upwards or downwards through the desorbent. Any conventional equipment used in the static layer to bring the liquid and solid into contact can be used.

However, a countercurrent moving bed or similar countercurrent moving system has a much higher separation efficiency than a fixed adsorption bed system and is therefore more advantageous. In processes using a moving bed or the like, the adsorption and desorption treatments take place continuously, which enables the continuous production of both extract and raffinate streams and the continuous use of feed and desorbent streams. The handling principles and various steps of a simulated moving bed countercurrent system are set forth in U.S. Patent 2,985,589. In such a system, the successive downward movement of multiple liquid inlets in the adsorption chamber mimics the upward movement of the adsorbent containing the chamber. Only four input lines are active at any given time; feed stream / desorbent feed stream, raffinate effluent and extract effluent. Simultaneously with the simulated upward flow of this solid adsorbent, there is a flow of the liquid that fills the pore volume of the adsorbent in the dense bed. By maintaining countercurrent contact, a downward flow of liquid in the adsorption chamber can be provided by a pump. As the active fluid inlet moves through the series, i. from the top of the chamber to its bottom, the chamber recirculation pump moves through different zones that require different flow rates. A programmed flow control device may be used to provide and control these flow rates.

6 30 6 5

The active liquid inlets efficiently divide the adsorption chamber into separate zones, each with a different function. In this embodiment of the process according to the invention, it is generally necessary for three separate treatment zones to be present in order to carry out the process, although in a few cases a fourth zone may also be used.

Adsorption zone 1 is defined as the adsorbent located between the inlet stream of the substance to be treated and the effluent stream of raffinate. In this zone, the feedstock is in contact with the adsorbent, the extract component is adsorbed and the raffinate stream is removed. Since the general flow through zone 1 occurs from the feed stream entering the zone to the raffinate stream leaving the zone, the flow in this zone is considered to be downstream when performing treatment from the feed point to the raffinate effluent stream.

Immediately upstream of the liquid flow, zone 1 has a purification zone, zone 2. The purification zone is formed by an adsorbent located between the extract discharge stream and the incoming substance feed stream. The basic treatments in zone 2 are the displacement of the non-selective pore volume of the adsorbent for the possible raffinate material that has migrated to zone 2 when the adsorbent is changed to this zone, Purification is accomplished by directing a portion of the extract stream material leaving zone 3 to zone 2 at the boundary of this zone as viewed upstream, i.e., the extract effluent stream to effect displacement of the raffinate material. The material flow in zone 2 occurs downstream of the extract effluent into the inlet of the substance to be treated.

Immediately upstream of zone 2 with respect to the flow of liquid in zone 2 is the desorption zone, i.e. zone 3. The desorption zone is formed by an adsorbent located between the inlet of the desorbent and the outlet stream of the extract. The purpose of the desorption zone is to allow the desorption material that enters i. .

In a few cases, a buffer zone, zone 4, may be used. This zone, which can be considered as an adsorbent located between the raffinate effluent and the desorbent inlet, if used, is immediately upstream of the liquid stream to zone 3. Zone 4 can be utilized to recover its desorption. which was used in the desorption step because part of the raffinate stream removed from zone 1 may enter zone 4 and displace the desorption material in this zone therefrom out of the desorption zone. Zone 4 contains sufficient adsorbent so that the raffinate material in the raffinate stream exiting zone 1 to zone 4 can be prevented from entering zone 3 and contaminating the extract stream removed from zone 3. In the case that the fourth zone is treated, the raffinate stream introduced into zone 4 must be carefully monitored so that the flow directly from zone 1 to zone 3 can be stopped when there is a sufficient amount of raffinate material in the raffinate stream passing from zone 1 to zone 3 so that the effluent does not become effluent.

Cyclic arrangement of inlet and outlet flows through the solid bed of adsorbent can be accomplished by utilizing a piping system in which valves are used sequentially to cause supply and inlet flows to alternate, thereby allowing fluid to flow countercurrent to the solid adsorbent. Another treatment method in which countercurrent flow of solid adsorbent to liquid flow can be achieved involves the use of a rotary poppet valve in which the inlet and outlet streams are in contact with this valve, and the lines through which the substance to be treated, extract, desorbent and raffinate are discharged in the same direction through the adsorption layer. Both the piping arrangement and the poppet valve arrangement are known in the art. Certain types of rotary poppet valves that can be utilized in this process are disclosed in U.S. Patent Nos. 3,040,777 and 3,422,848.

19 63065

It should be noted that at least a portion of the extract effluent must flow to separation devices in which at least a portion of the desorbent is removed to provide a product containing a smaller amount of desorbent. It is preferred that at least a portion of the raffinate effluent be subjected to separation devices in which at least a portion of the desorbent is removed to provide a desorbent stream that can be reused in the treatment and to provide a raffinate product containing a smaller amount of desorbent. The usual separation device is a fractionation column.

Although both liquid and vapor phase treatments can be used in a variety of adsorbent separation treatments, liquid phase treatment is more preferred in the process of the invention due to the lower temperatures required and the higher yields of extract obtained. Adsorption conditions include a temperature range of about 20 to 200 ° C, with a range of 20 to 100 ° C being most preferred, and a pressure range of about atmospheric pressure to about 35 atmospheres, wherein a range of about atmospheric pressure to about 17.5 atmospheres is required to provide a liquid phase. cheaper. Desorption conditions include the same temperature and pressure ranges as adsorption conditions.

The following examples illustrate the unique selectivity inherent in the process of the invention.

Example 1 This example shows the retention volume and selectivity results obtained by pulse experiments using eleven adsorbents, one comprising A zeolite, the other Y zeolite and nine X zeolites. More specifically, the adsorbent formed by zeolite A was a "Linde 5A" molecular sieve (calcium ion exchange, zeolite A), the zeolite Y was prepared by performing substantially complete ion exchange in "Linde SK-40" with potassium, and the adsorbents formed by zeolite X were " Linde 13X "molecular sieves with substantially complete ion exchange with the cations of metals K, Cs, Mg, Ca, Sr, Ba, Ba + K and Ba + Sr. The particle size range of all adsorbents was approximately 0.4-0.8 mm (Na-X = non-exchanged 13X screens).

20 6 3 0 6 5 The pulse test apparatus and method used have been described previously. The adsorbents were tested in a helical column equipped with a 70 ml heating apparatus maintained at 55 ° C and 4.4 atmospheres (abs.) And using pure water as the desorption material. The treatment sequence for each experiment was as follows: The desorbent (water) was continuously passed through a column containing the adsorbent at a nominal volume flow per hour (LHSV) of about 1.0. At the appropriate time, the flow of desorbent was stopped, a 4.7 ml sample of 10% by weight of fructose in water was injected into the statue via a sampling device, and the flow of desorbent was restarted. The sugar leaving was found to be continuous using a refractometer analyzer and the peak curve was plotted. A second sample containing 10% by weight of glucose was treated in the same manner. A saturated aqueous solution of benzene was similarly injected while acting as an analyte from which the interparticle volume of the adsorbent could be determined. Thus, three peaks were developed from each adsorbent tested, one for glucose, one for fructose, and a third for benzene. The glucose retention volume was calculated by measuring the distance from time zero, i.e. from the reference point to the center of the glucose peak, and subtracting from this the distance corresponding to the volume between the parts of the adsorbent obtained by measuring the distance from the same reference point to the center of the benzene peak. When a few adsorbents were used, both the fructose and glucose peaks were substantially on top of the benzene peak, indicating that both inono-saccharides were relatively non-adsorbed when the adsorbent in question was used in the presence of water. The selectivity of the adsorbent for fructose over glucose is the quotient obtained by dividing the retention volume of fructose by the corresponding volume of glucose. The results of these experiments are shown in Table 1 below.

63065 21

table 1

Selectivities of different adsorbents for fructose compared to glucose

Test Adsorbent Fructose Glucose retention Selective retention volume, ml Sep (B) volume, ml 1 Na-X 7.1 5.0 1.42 2 KX 11.9 21.6 0.55 3 KY 10.8 4 .9 2.21 4 Cs-X Both were relatively unadsorbed

5 Mg-X

6 Ca-X

7 Ca-A

8 Sr-X 8.0 1.3 6.15 9 Ba-X 27.1 9.6 2.82 10 Ba-KX 16.4 7.5 2.19 11 Ba-Sr-X 21.3 4, 2 5.0

The adsorbents used in Experiments 1-4 were three X-zeolites and one Y-zeolite, each containing cations of Group IA metals of the Periodic Table of the Elements at alternating cationic positions. The K-X adsorbent used in Experiment 2 had a selectivity of 0.55 (relative to fructose to glucose) and, as a result, actually had selectivity to glucose over fructose. The Na-X adsorbent used in Experiment 1 (selectivity 1.42) and the KY adsorbent used in Experiment 3 (selectivity 2.21) both had selectivity for fructose over glucose, while the Cs-X adsorbent used in Experiment 4 had no relative selectivity for either of these substances. .

The adsorbents used in Experiments 5, 6, 8 and 9 were X-zeolites containing cations of Group IIA metals of the Periodic Table at alternating cationic positions, while the adsorbent used in Experiment 7 was calcium-exchanged A-zeolite. Both fructose and glucose were relatively non-adsorbed using Mg-X, Ca-X and Ca-A adsorbents in the presence of water in Experiments 5, 6 and 7, but Sr-X and Ba-X adsorbents used in Experiments 8 and 9, respectively, had selectivity for fructose. compared to glucose. Although not definitively defined, it is assumed that adsorbents containing X zeolites and having a group HA cation at the 63065 22 living cationic positions generally become less acidic as they move down from Section 3 to Section 6 in the Periodic Table of Elements when selecting a Group IIA cation. Thus, Adsorbents containing Ca- or Mg-substituted X-zeolites are unsuitable for use in the process of the invention because they are too acidic, while Adsorbents containing Ba- or Sr-substituted X-zeolite are suitable for use in the process of the invention because they are less acidic.

The adsorbents used in Experiments 10 and 11 were X-zeolites containing cation pairs Ba and K and Ba and Sr, respectively, at the exchangeable cationic positions. The Ba-K-X adsorbent used in Experiment 10 had selectivity for fructose over glucose, whereas the K-X adsorbent used in Experiment 2 did not, but the selectivity of the Ba-K-X adsorbent was not as high as that of the Ba-X adsorbent used in Experiment 9. The selectivity of the Ba-Sr-X adsorbent used in Experiment 11 for fructose-glucose was lower than that of the Sr-X adsorbent used in Experiment 8, but higher than that of the Ba-X adsorbent used in Experiment 9.

Example II

To ensure that fructose could be separated from a mixture containing a solution of fructose and glucose containing 20% by weight of both fructose and glucose in water, a test was performed on a column of 440 ml of adsorbent containing barium-exchanged X-seo compound in such a manner. in a column with an inner diameter of 1.27 cm and a height of 2.28 m. The adsorbent was the same as used in Experiment 9 of Example I and the same processing temperature and pressure as in Example I were also used in this case. The desorbent was first passed with water through the adsorbent, then a dose of the substance to be treated was injected, and then the flow of desorbent was recirculated. The effluent was analyzed for both refractive index and polarity, and this combination was used to quantify rather than qualitatively the amounts of fructose and glucose in the effluent. The larger sample volumes used in these analyzes resulted in a higher column use than in Example I.

The results obtained in this example as well as the results of Experiment 9 of Example I (using the same adsorbent) are shown in Table 2.

63065 23

Table 2

Selectivity comparison using Ba-X adsorbent

Experiment Fructose Glucose Selectivity retention volume, retention volume, (B) ml ml

Example II 105.0 35 3.0

Example I

test 9 27.1 9.6 2.82

The selectivity obtained by treating fructose and glucose together was essentially the same as that obtained when treated separately.

Example III

This example shows the widths and retention volumes of glucose and fructose and the selectivity of fructose for glucose and water obtained by performing pulsation experiments using 10 different adsorbents. Of these ten adsorbents, one consisted of X-zeolite and nine of Y-zeolites. More specifically, the adsorbent formed by the X-zeolite was a "Linde 13X" molecular sieve that was substantially completely exchanged with Ca cations, and the nine adsorbents formed by the Y-zeolite were "Linde SK-40" products that were substantially completely ion-exchanged. using hydrogen, ammonium, Na, K, Cs, Mg, Ca, Sr and Ba cations. These ten adsorbents are hereinafter referred to as zeolite adsorbents NH 2 -Y, H-Y, Na-Y, K-Y, Cs-Y, Mg-Y, Ca-Y, Ca-X, Sr-Y and Ba-Y, respectively. The particle size of all these adsorbents was approximately 0.4-0.8 mm.

The adsorbents were tested in a 70 ml heating coil column maintained under the same conditions as in Example I and using the method of Example I, except that after treatment with 10% by weight glucose, a period of deuterium oxide was injected. Deuterium oxide has a different refractive index than water and thus deuterium oxide can be detected by refractometer in the same way as sugars. For each adsorbent, four peaks were tested, one developed for glucose, the other for fructose, the third for deuterium oxide, and the fourth for benzene. Retention values for glucose, fructose and water as well as deuterium oxide were obtained by the method described in Example I. The results of these experiments are shown in Table 3.

63065 24

The NH4-Y zeolite adsorbent used in Experiment 1 had a good selectivity for 6.5 fructose over glucose and a moderate, albeit somewhat low, selectivity for 0.71 fructose over water. The most preferred selectivities of the desorbent over the extract component are about 1.0-1.5 so that the extract component can easily displace the desorbent from the adsorbent in the adsorption zone while still allowing the extract component to be removed together with appropriate amounts of desorbent material from the adsorbent. The H-Y zeolite used in Experiment 2 had no selectivity for fructose or glucose in the presence of water as both eluted simultaneously. Both the Na-Y zeolite adsorbent used in Experiment 3 and the KY zeolite adsorbent used in Experiment 4 had a selectivity for fructose over glucose, although less than that obtained with the NH 4 -Y adsorbent, but the Cs-Y-zeolite adsorbent used in Experiment 5 did not. there was no selectivity for either of these. The selectivities of fructose with respect to water using Na-Y and K-Y zeolite adsorbents were also less than 1.0.

The Mg-Y zeolite used in Experiment 6 and the Ca-X adsorbent used in Experiment 8 had no selectivity between glucose and fructose because both eluted simultaneously. The Sr-Y zeolite and Ba-Y zeolite used in Experiments 9 and 10, respectively, each had significant selectivity for fructose, but also had the highest selectivity for fructose over water of all the adsorbents tested, indicating that higher amounts of desorbent (water) were required to desorb. -uutekomponentti. The best overall utility as measured by the pulse test was obtained for the Ca-Y zeolite used in Experiment 7. This adsorbent had the best selectivity between fructose and glucose, the ideal selectivity between fructose and water, and peak widths that showed relatively fast transfer values. For these reasons, Ca-Y zeolite is the most preferred adsorbent in the process of the invention.

6 30 6 5 25

•B

to p vo m

O I '' vo oo av P

o - * · »*» ^ * · po -POO -PP Ρ P Pm ® A p cd rd rd <d ^ 3 to>>>> p φ · Η Ρ Ή Ή 10 A> 3 3 3 3> ι - μ -Ρ ρ -Ρ>, Η Ρ Ρ Ρ · Η to ο ο ο ο Ρ \ 3 3 3 3 · Η Ρ Ή Ρ * -! Ή CO Ρ Ο) Φ φ 0) ρ O to m mo O vo ΐ "G -POO» Ρ - «Ρ Ή * · Ή -" Ο) a Ρ Ο ό to ^ pmto to o to vn m ό φ a α ο ο ο ρ ο λ ρ 3 3 ο οο ο: td J) MH Ρ Ρ P Ρ G 10 A ϋ A AA * 3 3 3 3 G Ρ Μ ρ ρ • H Ρ Ρ Ρ Ρ

•B

CO; Π3 Cfd Ή CCd Ή! Id · Η

o Ρ P Ρ P Ρ P Ρ P

o Ρ P p p P (0 P (0 P CO

4J gto (Uto Φ Φ Φ Φ φ Φ AO tl) to to 3 * · O Ρ tO Γ "O -H -H -H -H (NP p CN 00 p tO-P» tO -H * «(J 10 Hi * M (tl '*

vp 3 A oo 0 3 ^ 00 O A 0 A <N O * '"O VO

33 O A o Ρ O Ρ Ρ O Ρ Ρ P

ρ> ρ a -rt a td a m a 3

φ rd A 3 (d 3 G 3 3 3 C

Ό P P G P (d P rd 1 — I (d

> 1 -H O 'to O'EjO'S 0'S

> 1 ρ e <0 to S

to O -h! <0 to: rd to: rd to: td to • p -p to Ai to a: Ai a:> po φ Φ Φ Φ • pco to to to to m ρ φ a eo h Is * ( N en vo ρ ρ 3 * * · »* · -« O Ai φ lp fp P (NP CN P "

Ai Φ AU

Ai P 3 Φ P to

3 G

rd G Φ G

H Φ Φ Φ P

ΌΡ-p tn en o tNoorr r- m n * cn pjdPO '· * · - - - - φ 3: rd o m αν o n * 00 r- 00 1 O cn G to λ; pm <n m m m <n m ρ1 n1 p Ai rd G to 3

O Φ NP P

ρ Ό Φ £ A

Ρ Φ> ft> Φ «* Ρ P rd O rd ρ P CO τη G P to Ό PrdOOO τ O vo r» m 00 rd G A Ό O * * «.k» *. > * *

Ρ -h p A σν p t ^ Ooo r "vo 1 OP

G to p o 3 cm <n cn m <N m cn mm

Φ O ft A P

BOO

0) A

P 3 <d p 1 p en o

P P

P rd P

W -m ft * Ρ Φ I S *> r> · X> r> t

OG · * <X I * I I II II

(OP SI rdltO O '(d rd ptd Ό td 2® 2 A U g O U to® <Φ

O pm m p vo vo Γ "00 av O

A P

63065 26

Example IV

This example illustrates the ability of the method of the invention to separate ketosis from aldose when working using a preferred embodiment that utilizes a continuous, simulated moving bed countercurrent system. In particular, this example illustrates the experimental results obtained using a synthetic 16.5% by weight mixture of fructose and glucose in water using a barium-exchanged X-zaolite adsorbent having a particle size of approximately 0.4-0.8 mm and water as the desorbent. in a semi-factory scale experimental device (described in U.S. Patent 3,706,816). Briefly, the apparatus had 24 adsorption chambers, each with a volume of about 18.8 ml. The adsorption chambers were connected in series by means of connecting pipes of small diameter relative to each other and using a rotary valve with separate piping. The valve had inlet and outlet points that led to the flow of treatment medium and desorbent into the chambers and extract and raffinate flows out of the chambers. By treating the rotary valve and maintaining the specified pressure differences and flow rates in the various tubes leading into and out of the chamber series, a simulated counterflow was obtained. The adsorbent remains in place while the liquid flows through the chambers connected in series in such a way that, viewed from any position in the adsorption chambers, it forms a continuous flow in the opposite direction. The treatment of the rotary valve was carried out periodically so as to allow a new treatment to take place in the layers of adsorbent located between the active supply and discharge points of the rotary valve. The rotary valve was connected to supply and discharge lines through which liquids flowed into and out of the treatment. The rotary valve had a treatment feed line through which the treated mixture containing extract and raffinate was passed, an extract stream discharge line through which the desorbent was passed as a mixture with the extract component, a desorbent feed line through which the desorbent was passed, and a raffinate through which the desorbent was passed. In addition, a rinse aid supply line was used to intentionally flush out the components from the lines that had previously contained the material to be treated and which subsequently contained a raffinate or extract stream. The rinsing agent used was desorption material, which then left the apparatus as part of the extract and raffinate effluent streams. Raffinate and extract effluents were collected and analyzed for fructose and glucose concentrations of 27,63065 using chromatographic analysis, but no attempt was made to remove desorbent. The yield of fructose was determined by calculating the amount of fructose "lost" in the raffinate stream, determining this amount per unit time as the percentage of fructose fed to the apparatus, and subtracting this percentage from 100%. The operating pressure of the experiments was 10.2 atmospheres and the processing temperatures in Experiments 1 and 2 were 50 and 75 ° C, respectively. The fructose purity (as a percentage of total sugars) and fructose yield of the extract effluent are shown in Table 4 below.

Table 4

Experiment Extract stream, fructose- Yield of fructose,% purity,% 1 80 60 2 80 71

The results of Experiments 1 and 2 do not necessarily represent the optimums that can be obtained.

Example V

In this example, the procedure of Example 4 was essentially repeated, but using Ca-Y zeolite as the adsorbent to separate ketose from aldose. In particular, this example shows the results obtained by treating an aqueous solution of corn syrup using Ca-Y zeolite as described in Example 3 as an adsorbent and deionized water as a desorbent.

The raw material was treated as a 50% sugar solution in water. Its solids content was 52% glucose, 42% fructose and 6% higher saccharides. The treatment temperature was 60 ° C. The purity of the fructose of the extract inlet stream (as a percentage of the total amount of sugars present) and the yield of fructose are shown in Table 5. (The pressure used was 10.5 atmospheres).

28 63065

Table 5

Experiment Extract stream, fructose Yield of fructose,% purity,% 1 97 10 2 94 49 3 92 65 4 87 83 5 84 88 6 80 90

For the sake of clarity, the analyzes of the extract and raffinate streams in a fructose purity curve, namely at a fructose content of 85%, are shown in Table 6 below.

Table 6

Extract and raffinate stream analyzes at 85% yield

Extract stream Raffinate stream% fructose 88.3 10.7% glucose 11.7 79.5% higher saccharide traces 9.8% sugars 14.9 13.3

The results of the above experiments do not necessarily represent the optimum values that can be obtained.

Example VI

This example shows the peak widths and retention volumes of xylose, which is aldopentose, and ribulose, which is ketopentose, and the selectivity for ribulose over xylose and water, which values were obtained by performing a pulse test using a Ca-Y faujasite adsorbent. The grain size of the adsorbent was about 0.4-0.8 mm.

3

The adsorbent was tested on a 70 cm helical column under the same conditions as in Example I and using the same method as in Example I, except that sucrose was used as the analyte and the helical column was kept at 60 ° C. The feed pulse used in this example contained 2% each of ribulose, xylose and sucrose and 20% D_0 with the remainder being water. The volume of the pulse was 10 cm.

29 63065

The results of the pulse test are shown in Table 7.

Table 7 for tip points, ml Retention volume, ml Selectivity (B)

Precipitate Ksy- Ribu- Ksy- Ribu- Ribu- / Ksy- Ribu- / Water rose lodge lodge lodge lodge lodge lodge lodge 10.4 10.0 20.4 2.20 38.8 17.6 3.2

As shown in Table 7, the Ca-Y faujasite adsorbent had good selectivity for 17.6 ribulose compared to xylose. The selectivity for 3.2 ribulose over water indicates that a relatively larger amount of desorbent would be required to extract the ribulose of the extract component than when the selectivity of the extract component over the desorbent is in the preferred range of about 1.0-1.5.

Claims (11)

30 6 30 6 5 A method for dividing a mixture containing ketose and aldose into its ingredients by contacting the mixture with an adsorbent under adsorption conditions, then contacting the adsorbent containing the adsorbed ingredient with water and thus recovering the desorbed ingredient, characterized in that the mixture is contacted with crystalline aluminum. for selective adsorption of ketosis is (1) an X-zeolite at the interchangeable cation positions containing at least one cation which is sodium, potassium, barium or strontium, or (2) a Y-zeolite at the interchangeable cation positions containing at least one cation which is ammonium , sodium, potassium, calcium, strontium, barium, or a combination thereof, or which, to selectively adsorb aldose, is (3) X-zeolite containing potassium cations at interchangeable cationic positions. Process according to Claim 1, characterized in that the adsorption is carried out in a temperature range from about 20 to 100 ° C and at a pressure of about 1 to 17.5 ata. Process according to Claim 1 or 2, characterized in that the ketose and aldose are hexoses. Process according to Claim 3, characterized in that the ketose is fructose and the aldose is glucose or dextrose. Process according to Claims 1 to 4, characterized in that the mixture to be treated is formed by an invert sugar solution. Process according to Claims 1 to 4, characterized in that the mixture to be treated consists of high-fructose corn syrup. Process according to Claim 5 or 6, characterized in that the concentration of ketose and aldose in the raw material mixture is about 0.5 to 30% by weight, preferably 1 to 15% by weight. 6 30 6 5 31 Process according to Claims 1 to 7, characterized in that the particle size of the X or Y zeolite is about 0.25 to 1.20 mm, preferably about 0.30 to 0.60 mm. A method according to claim 1, characterized in that (a) a net flow of liquid in one direction is maintained through a column containing said X or Y zeolite, the column comprising at least three series-connected zones, each operating separately, the end zones being connected by a continuous circulating flow (b) maintaining an adsorption zone comprising an X or Y zeolite located between the upstream boundary of the liquid inlet to be treated and the downstream boundary of the raffinate outlet; (c) maintaining the column immediately upstream; from the adsorption zone, a purification zone located between the upstream boundary of the extract outlet and the downstream boundary of said liquid inlet to be treated, (d) a desorption zone immediately upstream of the purification zone is maintained in the column; is located between the upstream boundary formed by the water inlet and the downstream boundary formed by said extract outlet, (e) the liquid to be treated is introduced into the adsorption zone and selectively adsorbs the second, - or displacing from the Y-zeolite the component selectively adsorbed to it in the previous section, (g) removing an extraction stream comprising water and the selectively adsorbed component from the desorption zone, generating a product stream comprising a selectively adsorbed component, (i) periodically passing in a column containing X or Y zeolite downstream of the liquid flow direction through the adsorption zone the point of entry of the liquid to be treated, r an affinity outlet line, a water inlet point, and an extract outlet point, to move the different 63065 32 zones in the X or Y zeolite and to provide an extract outlet stream and a raffinate outlet stream. A method according to claim 9, characterized in that at least a part of the raffinate outlet stream is passed to a separation device and at least a part of water is separated therein to obtain a raffinate product containing less water. A method according to claim 9, characterized in that a buffer zone is maintained immediately upstream of the desorption zone, located between the downstream boundary formed by the water inlet and the upstream boundary formed by the raffinate outlet.