Fractionation Method of Sucrose-Containing Solutions

The present invention relates to a method for separating sucrose and, optionally, a second dissolved component from solution. More particularly, the invention relates to a method in which a solution containing sucrose and other dissolved substances is first fractionated by a chromatographic simulated moving bed (SMB) process to yield a sucrose-enriched fraction and, optionally, a separate fraction enriched with a second component to be recovered. When two separate fractions are produced, the latter fraction is further fractionated chromatographically, either by a batch method or a simulated moving bed method. In a preferred embodiment, a beet-derived sucrose-containing solution is fractionated to yield a sucrose-enriched fraction and a fraction enriched with a second organic compound commonly present in beet-derived solutions, such as betaine, inositol, raffinose, galactinol, or serine and other amino acids.

It is known that sucrose and betaine are recoverable from molasses by chromatographic separation. U.S. Pat. No. 4,359,430 to Suomen Sokeri Oy, describes a chromatographic method for the recovery of betaine from molasses by a batch process in which diluted molasses is fractionated with a polystyrene sulphonate cation exchange resin in alkali metal form. This method achieves good separation of sucrose and betaine. This reference also discloses a method in which a betaine-enriched fraction obtained from a first fractionation is subjected to further chromatographic purification. The further purification step is said to be capable of separating other components from the betaine-enriched fraction. However, the dry solids content in the sucrose and betaine fractions obtained by this method is relatively low, therefore, large amounts of eluant water must be evaporated when recovering the sucrose and betaine from the respective fractions by crystallization.

Continuously operated chromatographic separation processes presently employ the SMB method, which method is used in a variety of different applications. The SMB method has a separating performance that is several times higher than that of the batch method, and also results in significantly lower dilution of the products or, conversely, lower consumption of eluant.

The SMB method may be carried out in either a continuous or a sequential mode. In both processes, the pattern of fluid streams is the same. These streams are (1) the supply of feed solution and eluant to the bed or beds, (2) the recycling of the liquid mixture between beds, and (3) the withdrawal of products from the beds. The flow rate for these flows may be adjusted in accordance with the separation goals, i.e., increased yield, purity, or capacity. In a continuous SMB process, which was first disclosed in the early 1960s in U.S. Pat. No. 2,985,589, all fluid streams flow continuously. Separation of sucrose by such continuous SMB methods has been described in international publication no. WO 91/08815 by the Amalgamated Sugar Company and in U.S. Pat. No. 4,990,259 to M. Kearney and M. Mumm and assigned to The Amalgamated Sugar Company.

In a sequential SMB process, some of the fluid streams do not flow continuously. Sequential SMB fractionation methods in which a sucrose fraction and a betaine fraction are recovered from beet molasses are disclosed in Finnish Patent 86,416 to Suomen Sokeri Oy, which corresponds to U.S. Pat. No. 5,127,957 and international publication no. WO 94/17213 to Suomen Sokeri Oy. German Offenlegungsschrift 4,041,414 to Japan Organo Co., which corresponds to British published application 2,240,053, also discloses a sequential SMB method by which several product fractions are recovered from sugarbeet molasses.

In the sugar industry, the important parameters in the fractionation of molasses to recover sucrose include the purity and yield of sugar, the separation capacity, and the eluant/feed ratio. A purity of 92% and a yield of 90% are the usual requirements for a sugar product. In order to increase the capacity to recover sugar from the process, the flow rates, which are generally higher in SMB processes than in batch processes, are increased. Along with the increase in the flow rate, however, a "flat tail" is produced in the sucrose elution profile. A "flat tail" means that the concentration of sugar has not peaked sharply, but, rather, has been considerably diluted with the increased volume of eluent. This is especially disadvantageous to the recovery of a second dissolved component. With respect of the recovery of sucrose and betaine, this effect is shown in the elution profiles presented in U.S. Pat. No. 4,359,430 and Finnish Patent 86,416, for example. In the course of obtaining a high sucrose yield, the betaine yield is diminished because part of the betaine is allowed to pass into the sucrose fraction wherefrom it is removed in the sucrose crystallization step. Likewise, if a high betaine yield is desired, considerable amounts of sucrose end up in the betaine fraction, thus diminishing the sucrose yield and considerably impairing the purity of the betaine fraction.

In the above references, the purity of the betaine fraction obtained by the process of German Offenlegungsschrift 4,041,414 is relatively good, 80.9% on a dry solids basis (d.s.), but the purity of the sucrose fraction, 87% d.s., is inadequate for the sugar industry. It can be concluded from the composition of the feed solution of Example 3 in the reference that the "thin juice" was demineralized prior to the SMB fractionation by the "KAAK method" (which refers to cation exchange-anion exchange-anion exchange-cation exchange as described in Sayama, K., Kamada, T., and Oikawa, S., Production of Raffinose: A New By-product of the Beet Sugar Industry, British Sugar plc, Technical Conference, Eastbourne 1992). Molasses produced by such a beet sugar process has a different composition from common molasses. Typically, beet molasses contains 1.5-3.5% by weight of raffinose and 3.5-6.5 weight % of betaine on a dry solids basis. Since the feed solution of Example 3 in German Offenlegungsschrift 4,041,414 has a raffinose content of 17.3% by weight and a betaine content of 12.2% by weight on a dry solids basis, it can be concluded, on the basis of the raffinose-to-betaine ratio, that roughly half of the betaine contained in common beet molasses was lost in the ion exchange treatment.

According to Finnish Patent 86,416, a purity as high as 70.9% d.s. for the betaine fraction was obtained, with 11.1% d.s. of sucrose present in the betaine-enriched fraction. However, the 86.8% purity of the sucrose fraction does not meet the requirements of the sugar industry. Similarly, the 47.5% purity of the betaine fraction reported in international publication no. WO 94/17213 is rather poor.

It is an object of the present invention to fractionate sucrose and a second desired organic component, such as betaine, inositol, raffinose, galactinol, serine, or other amino acids, from a solution so as to obtain higher yields with at least equivalent purity for sucrose.

It is also an object of the present invention to fractionate sucrose and betaine from a beet-derived sucrose-containing solution with a higher purity and a higher yield for betaine.

It is a further object of the present invention to provide an economical fractionation method, in terms of capacity and the eluant/feed ratio, that separates two components from a sucrose-containing solution with high yield and high purity as economically as the prior SMB methods for fractionating sucrose-containing solutions.

Accordingly, the present invention is a method for separating sucrose and, optionally, a second dissolved component from a sucrose-containing solution, preferably a beet-derived sucrose-containing solution, in which the solution is subjected to a first chromatographic fractionation by a SMB method to yield a sucrose-enriched fraction (hereinafter the first sucrose fraction) and a fraction enriched with the second dissolved component, and the resulting fraction enriched with the second component is subjected to a second chromatographic fractionation, to yield a second sucrose-enriched fraction (hereinafter the second fraction) and a second fraction enriched with the second dissolved component. The first fractionation may be carried out so that sucrose and the second dissolved component are enriched in the same fraction, or preferably, they are enriched in separate fractions.

FIG. 1 is a schematic diagram of a three-column sequential SMB process as described in Example 1 for separating sucrose and betaine from molasses;

FIG. 2 is a truncated schematic diagram of a fourteen-column continuous SMB process as described in Example 4;

FIG. 3 is a graph of the concentration gradient of the liquid circulating in the fourteen-column system shown in FIG. 2 over time;

FIG. 4 is a truncated schematic diagram of a fourteen-column continuous SMB process as described in Example 5;

FIG. 5 is a graph of the concentration gradient of the liquid circulating in the fourteen-column system shown in FIG. 4 over time;

FIG. 6 is a truncated schematic diagram of a fourteen-column continuous SMB process for separating sucrose and betaine from molasses as described in Example 6;

FIG. 7 is a schematic diagram of a three-column sequential SMB process for the second fractionation of betaine as described in Example 6; and

FIG. 8 is a graph of the concentration gradient of the liquid circulating in the fourteen-column system shown in FIG. 6 over time.

The present invention is directed to a method for separating sucrose and, optionally, a second dissolved component from a sucrose-containing solution. The solution is subjected to chromatographic fractionation by a continuous or sequential SMB method where the two components are enriched in the same fraction or in separate fractions and either the single fraction or the second fraction is subjected to a second chromatographic fractionation in order to recover sucrose and a second component with an improved yield or purity.

In accordance with a preferred embodiment of the invention, the second sucrose fraction is combined with the sucrose fraction from the first chromatographic fractionation, and sucrose is recovered from the combined sucrose fraction thus obtained.

In accordance with another preferred embodiment of the invention, the second sucrose fraction is returned to the feed solution for the first fractionation. In this embodiment, sucrose is recovered from the first sucrose fraction.

Generally, the second dissolved component is recovered from the fraction obtained from the second fractionation, which is enriched with the second dissolved component. The term "second dissolved component" refers to organic compounds commonly present in beet-derived solutions, such as betaine, inositol, raffinose, galactinol, or serine and other amino acids. The second chromatographic fractionation, i.e., fractionation of the fraction enriched with the second dissolved component which is obtained from the first fractionation, may be performed either by a batch method or a SMB method.

The invention is particularly suitable for the recovery of sucrose and betaine from beet molasses. Therefore, the following description of the invention specifically refers to the recovery of sucrose and betaine, but the invention is not so limited. Instead of, or in addition to betaine, any other dissolved organic substance may be similarly recovered by adjusting the process conditions and parameters to suit the separation in question, which can be achieved easily by those skilled in the art.

With the method of the invention, the sucrose yield can be improved by up to about 10 percent compared with the SMB processes presently employed in the sugar industry. This improvement represents remarkable economic advantages in view of the large amount of the molasses used by the sugar industry for chromatographic separation. For example, in the United States, about 500,000 tons d.s. of molasses are currently used per annum.

The purity of the sucrose fraction produced by the method of the invention meets or exceeds the goal of about 92% set for industrial SMB methods.

With regard to betaine, the method of the invention can achieve yields as high as about 95%, which is a significant advantage over prior yields of about 30-70%, and a purity as high as about 95%, calculated on a dry solids basis, which is a substantial increase over purities of about 25-70% hitherto obtained.

The first chromatographic separation in the method of the invention may be carried out with prior art SMB methods and apparatus known to be suitable for the fractionation of molasses, such as those disclosed in U.S. Pat. No. 4,402,832 (continuous SMB method), Finnish Patent 86,416, and international publication no. WO 94/17213 (discussed above).

Also, the further fractionation of the betaine fraction produced in the first fractionation to yield a second sucrose fraction and a second betaine fraction may be carried out by known chromatographic separation methods and apparatus, such as the methods and apparatus disclosed in U.S. Pat. No. 4,359,430 (batch method), and Finnish Patent 86,416 and international publication no. WO 94/17213 (SMB methods).

In the continuous SMB method, all flows (supply of feed solution and eluant, recycling of liquid mixture, and withdrawal of product fractions) are typically continuous. The rates for these flows may be adjusted in accordance with the separation goals, namely, yield, purity, and capacity. There are normally 8 to 20 packed beds that are combined into a single continuous loop. The feed and product withdrawal points are shifted cyclically in the downstream direction in the packed bed. Due to the supply of eluant and feed solution, the withdrawal of products, and the flow through the packed bed, a dry solids profile is formed in the packed bed. Constituents having a relatively low migration rate in the packed bed are concentrated in the back slope of the dry solids profile, while ingredients having a higher migration rate are concentrated in the front slope. The points of introduction of the feed solution and eluant and the withdrawal points of the product or products are shifted gradually at substantially the same rate at which the dry solids profile moves in the packed bed. The product or products are withdrawn substantially from the front and back slopes of the dry solids profile. The feed solution is introduced substantially at the point where the composition of the cyclically moving dry solids profile is closest to the composition of the feed solution, and the eluant is introduced approximately at the point of minimum concentration of the dry solids profile. Part of the separated products are recycled due to the continuous cyclic flow, and only part of the dry solids is withdrawn from the packed bed during one sequence.

The feed and withdrawal points are shifted cyclically by using feed and product valves located along the packed bed, typically at the upstream and downstream end of each section of the packed bed. (The requisite valves and feed and withdrawal equipment are part of the apparatus). If product fractions of very high purity are desired, short phase times and multiple sections of packed beds must be employed.

In the sequential SMB system, not all flows (supply of feed solution and eluant, recycling of liquid mixture, and withdrawal of products) are continuous. Yet the shifting of the dry solids profile or profiles moving cyclically in the system is continuous. The flow rate and the volumes of the different feeds and product fractions may be adjusted in accordance with the separation goals, i.e., yield, purity and capacity.

During the feeding phase, a feed solution, and, optionally, also an eluant during a simultaneous eluting phase, is introduced into predetermined packed beds, and, simultaneously, one or more product fractions are withdrawn. During the eluting phase, eluant is introduced into a predetermined packed bed or beds and, during the feeding and eluting phases, one or more product fractions are withdrawn.

During the recycling phase, essentially no feed -solution or eluant is supplied to the packed beds and no products are withdrawn. A forward flow is maintained in a fixed direction in a system comprising at least two columns, and the products are recovered during a multi-step sequence comprising the above phases. One column may equal a single packed bed or a portion of a packed bed. In the latter configuration, the columns that make up a packed bed are referred to as "sectional packed beds."

During the feeding phase, feed solution is introduced into a column and a corresponding quantity of any product fraction is withdrawn at a point which may be located either in the same column as the feed point (in which case the other columns in the system may be, for example, in the eluting or recycling phase) or in a different column from that of the feed point, which column is connected in series (optionally through other columns) with the column into which the feed is introduced. During the recycling phase, the liquid in the columns, along with its dry solids profile or profiles, is recycled in a loop comprising one, two or several columns. In the eluting phase, eluant is introduced into the column and a corresponding amount of product fraction(s) is (are) withdrawn from the same or a downstream column.

As stated previously, a detailed description of these sequential SMB processes applied to the recovery of sucrose and betaine from beet molasses is provided in Finnish Patent 86,416 and international publication no. WO 94/17213; these processes may be employed in the method of the present invention to carry out both the first and the second fractionation.

By moving the resin bed in a column into which the feed solution is introduced counter-currently to the liquid flow direction of the dry solids profile, an actual moving bed system can be achieved. It is self-evident that results very similar to those achieved with a simulated moving bed can be obtained with such an actual moving bed.

In the method of the invention, preferably a gel-type strong cation exchanger (e.g., "Dowex", Finex" or "Purolite") is employed as the packing material for the columns, and it is preferably in sodium and/or potassium form. The packing material is preferably equilibrated to the tonic form of the feed solution prior to the fractionation.

The dry solids content of the beet-derived sucrose-containing solution to be fed to the chromatographic separation is typically 20-80 g/100 g, preferably 40-70 g/100 g. The solution is heated to 40.degree.-95.degree. C., preferably 65.degree.-85.degree. C., prior to being supplied to the separation process.

The elution phase employs mainly water and/or very dilute aqueous solutions (having a dry solids content less than 8% by weight, preferably less than 1% by weight). The eluant has a temperature of 40.degree.-95.degree. C., preferably 65.degree.-85.degree. C.

The dry solids content of the betaine fraction obtained from the first fractionation is adjusted prior to the second fractionation to about 20-50 g/100 g for batch separation or, typically, to 20-80 g/100 g, and preferably 40-70 g/100 g, for SMB separation.

Sucrose can be recovered from the sucrose fraction by methods commonly used in the sugar industry, such as by crystallization or as a syrup, or as liquid sugar subsequent to purification. Betaine is recovered from the second betaine fraction by crystallization, for example, as described in U.S. Pat. No. 4,359,430.

To optimize the sucrose and betaine yields and purity, the pH of the feed solution also may be adjusted. It is generally adjusted prior to the second fractionation to the range 6.5-12, and preferably between 9.5-11.5.

The following examples illustrate the method of the invention in the context of fractionating beet molasses to recover sucrose and betaine. It is not intended that the invention be limited by this description of the preferred embodiments.

A chromatographic apparatus as schematically shown in FIG. 1 was employed. The apparatus comprised three columns 1-3 connected in series, fluid conduits 4-7 connecting the columns, a molasses container 8, a water/eluant container 9, a molasses feed conduit 10, an eluant feed conduit 11, a recycle pump 12, a molasses feed pump 13, an eluant feed pump 14, heat exchangers 15-17, product fraction withdrawal conduits 6, 18-20, 48 and 49 and valves 21-47.

The columns were packed with a strong cation exchanger resin known in the trade as Finex CS 11 GC.TM., manufactured by Finex Oy. The resin had a polystyrene/divinylbenzene backbone and was activated with sulphonic acid groups; the mean bead size (in Na+ form) was about 0.38 mm. The resin had a DVB content of 5.5%. Prior to the test, the resin was regenerated to sodium form; during the fractionation it was equilibrated by cations from the feed solution.

The feed solution was beet molasses wherefrom calcium was precipitated by adding sodium carbonate (pH about 9); the calcium carbonate precipitate was removed by filtration.

Fractionation was performed using the following seven-step sequence:

Step 1: Feed solution 10 was introduced (feeding phase) into column 1 through valve 22 at a flow rate of 80 l/hr., and a residue fraction was eluted from the downstream end of same column 1 through conduit 48. Simultaneously, eluant 11 was supplied (eluting phase) to column 2 through valve 26 at a flow rate of 25 l/hr., and a sucrose fraction was eluted from column 3 through conduit 6.

Step 2: The liquid in the columns was recycled (recycling phase) in the loop formed by all columns at a rate 120 l/hr.

Step 3: Eluant 11 was introduced into column 1 through valve 23 at a rate of 120 l/hr. and, simultaneously, a betaine fraction was eluted from column 3 through conduit 6.

Step 4: Eluant 11 was introduced (eluting phase) into column 1 through valve 23 at a flow rate of 120 l/hr., and a second residue fraction was eluted from the downstream end of column 2 through conduit 49.

Simultaneously, eluant was supplied (eluting phase) to column 3 through valve 29 at a flow rate of 55 l/hr., and a second betaine fraction was eluted from the downstream end of the same column through conduit 6.

Step 5: Same as step 2.

Step 6: Eluant 11 was introduced into column 1 through valve 23 at a flow rate of 120 l/hr., and a third residue fraction was eluted from the downstream end of column 3 through conduit 6.

Step 7: Same as step 2.

After the sequence was carried out to completion, the process control program returned to step 1. By repeating this sequence five to seven times, the system was equilibrated. The method proceeded in a state of equilibrium, and the process of the separation process was monitored with a density meter, a meter for optical activity, a conductivity meter, and the separation was controlled by a microprocessor whereby precisely defined volumes and flow rates of feeds, recycled liquid and product fractions were controlled employing quantity/volume measuring means, valves and pumps.

In this method, a sucrose fraction from column 3, two betaine fractions from column 3, and one residue fraction from each column were withdrawn. The betaine fractions were combined, as were the residue fractions.

Analyses of the feed solution and the product fractions withdrawn during one sequence after an equilibrium was reached are presented in Table 1, where the percentages of the different components are given as percent by weight dry solids basis.

The sucrose yield into the sucrose fraction was 90.1% and the betaine yield into the combined betaine fraction 58.70%.

The apparatus and test conditions described in Example 1 were employed. The procedure was also similar to that of Example 1, except that the fraction volumes were adjusted. The adjustment caused a higher purity, but lower yield, for sucrose and a lower purity, but higher yield, for betaine than obtained in the first fractionation in Example 1. Subsequent to evaporation, the resulting betaine fraction was subjected to refractionation by a similar sequential SMB process with different conditions. The sucrose fraction obtained from the second fractionation was combined with the sucrose fraction from the first fractionation, and the residue fractions were likewise combined.

Analyses of the feed solutions and the product fractions withdrawn during one sequence after an equilibrium was reached are presented in Table 2, where the percentages of the different components are given as percent by weight dry solids basis.

The sucrose yield from the first fractionation was 89.4% and the betaine yield was 89.9%. The total sucrose yield, calculated from the combined sucrose fraction, was 92.6% and the total betaine yield, calculated from the betaine fraction obtained from the second fractionation, was 88.2% The second fractionation afforded remarkable improvement of the sucrose yield and betaine purity. In addition, the betaine yield improved significantly as compared with Example 1.

The method described in Example 2 was essentially followed, but the effect of the pH of the feed solution for the second fractionation (which solution had been obtained from the betaine fraction from the first fractionation) was studied, performing the second fractionation in such a way that (a) the pH of the feed solution was not adjusted, and, hence, the pH was 10.2, (b) the pH of the feed solution was adjusted with hydrochloric acid to 9.5, and (c) the pH of the feed solution was adjusted with NaOH to 11.2.

Analyses of the feed solution for the second fractionation (i.e. further separation of the betaine fraction) and the product fractions withdrawn during one sequence after an equilibrium was reached are presented in Table 3, where the percentages of the different components are given as percent by weight dry solids basis.

The yields from the second fractionation in the above cases (a), (b) and (c) were as follows:

(a) sucrose 57.3%, betaine 95.4%

(b) sucrose 59.6%, betaine 96.8%

(c) sucrose 51.9%, betaine 96.8%.

As will be seen from the results, the pH of the feed solution affects the purity and yield of sucrose and betaine. The pH may be adjusted in accordance with the economical optimum.

The test apparatus comprised 14 columns connected in series, each having a diameter of 0.2 m and each containing a packed bed having a height of 0.85 m. FIG. 2 shows a schematic diagram of the test apparatus.

The columns were packed with a polystyrene-based cross-linked (5.5% DVB) strong cation exchanger having a mean bead size of 0.32 mm. The packing material was equilibrated with feed solution and was predominantly in potassium and sodium form.

Water, as eluant, was introduced into the column system through conduit 50 at a flow rate of 83.5 l/hr. Feed solution was introduced through conduit 51 to each column through feed valves 52-65 at a flow rate of 13.5 l/hr. for 150 seconds. The feed conduits were rinsed with eluant (30 s, 13.5 l/hr.) subsequent to the introduction of the feed solution. The flow rate of the product fraction through valves 66-79 was adjusted to 21 l/hr., which produced a by-product flow rate of 76 l/hr. The by-product fraction was withdrawn through springbiased valves 80-93 securing the desired pressure for the system. An average recycle rate of 300 l/hr. was maintained. In practice, this rate varies according to the change of the relative positions of the feed and product valves along the recirculation loop. The points of introduction of the feed solution and eluant and the withdrawal points of the product fractions were shifted downstream one column each successive step at intervals of 180 seconds.

Initially, the system was filled with a higher feed flow rate and lower eluant flow rate. Once the system was filled, the flow rate setpoints stated above were used to run the system until an equilibrium had been established.

Samples were taken at two-minute intervals via a sampling valve placed in the recirculation loop. The concentration gradient shown in FIG. 3 was drawn on the basis of an analysis of the samples. In addition, the feed solution and the product and by-product fractions were analyzed. The results are shown in Table 4, where the percentages of the differen components are given as percent by weight on a dry solids basis.

Molasses was fractionated by the continuous SMB method, wherein the fourteen-column system of Example 4 was modified in such a way that it was possible to withdraw three product fractions: sucrose, betaine, and by-product fractions. FIG. 4 shows a schematic diagram of the test apparatus. The flow rate of the sucrose fraction was adjusted to 21 l/hr. and the flow rate of the betaine fraction to 18 l/hr. The feed rate of the eluant through conduit 94 was 90.5 l/hr., and the feed flow rate through conduit 95 was 13.5 l/hr. Hence, the flow rate of the by-product fraction through conduit 96 was 65 l/hr.

The betaine fraction was concentrated to a dry solids content of 55% and fed to a separation system comprising two columns connected in series. The columns had a diameter of 0.2 m, and the packed bed in each column had a height of 0.85 m. The packing material was the same as in Example 4.

The betaine fraction was further fractionated using a batch process, supplying 2.6 liters of feed solution (55% by weight on a dry solids basis) to the upstream end of the first column. The feed was repeatedly introduced at intervals of 60 minutes. Elution was performed at a flow rate of 30 l/hr. The following fractions were withdrawn from the bottom of the column:

Fraction 1: By-product 8.6 liters

Fraction 2: Recycle fraction 2 liters (introduced into the column prior to the actual feed)

Fraction 3: Product solution 2.6 liters

Fraction 4: Recycle fraction 1.4 liters (introduced into the column subsequent to the actual feed)

Fraction 5: Betaine fraction 5.0 liters

Fraction 6: Elution recycling 10 liters

In the test example, the betaine separation had more capacity than the single SMB-column system with respect to betaine fraction produced.

The filling and equilibration of the column system, sampling, and analyzing of the samples were performed as above. The concentration gradient from the first continuous SMB separation is shown in FIG. 5. The results are shown in Table 5, where the percentages of the different components are given as percent by weight on a dry solids basis.

As can be seen from the results, the yield of sucrose increased from 87.6% to 91.8% and the purity of sucrose increased from 87.2% to 87.8%. With this modification, betaine was recovered with a yield of about 35% and a purity of 88.3% The low betaine yield is a result of the continuous SMB method in which the feed flow was uninterrupted, and thus a considerable portion of the betaine was lost in the sucrose fraction. By increasing the eluant flow rate and increasing the flow rate of the betaine fraction proportionately, the betaine yield is predicted to increase up to about 50-60%.

The continuous SMB process disclosed in Example 4 was converted into a sequential method in such a way that the columns of Example 4, referred to as sectional packed beds herein, were interconnected in sequence to form a four-column system in which two columns were formed by sectional packed beds 1-3 and 4-6, and two columns by sectional packed beds 7-10 and 11-14. Thus, the system comprised two columns having a total sectional packed bed height of 2.55 m each, and two columns having a total sectional packed bed height of 3.4 m each. FIG. 6 shows a schematic diagram of the apparatus.

Fractionation was performed sequentially by the following eight-step sequence:

Step 1: 15 liters of feed solution was introduced into sectional packed bed 1 at a flow rate of 75 l/hr., and a by-product fraction was withdrawn from sectional packed bed 10. 20 liters of eluant was introduced into sectional packed bed 11 at a flow rate of 100 l/hr., and a sucrose fraction was withdrawn from sectional packed bed 14.

Step 2: 8 liters of liquid was recycled at a flow rate of 100 l/hr. in the loop formed by all columns.

Step 3: 12 liters of eluant was introduced into sectional packed bed 1 at a flow rate of 120 l/hr. and a by-product fraction was withdrawn from column sectional packed bed 3. Simultaneously, 12 liters of eluant was supplied to sectional packed bed 4 at a flow rate of 120 l/hr., and a betaine fraction was withdrawn from sectional packed bed 14.

Step 4: 14 liters of eluant was introduced into sectional packed bed 1 at a flow rate of 120 l/hr., and a betaine fraction was withdrawn from sectional packed bed 14.

Step 5: 8 liters was recycled at a flow rate of 100 l/hr. in the loop formed by all columns.

Step 6: 10 liters of eluant was introduced into sectional packed bed 1 at a flow rate of 100 l/hr., and a by-product fraction was withdrawn from sectional packed bed 14.

Step 7: 4 liters of eluant was introduced into sectional packed bed 1 at a flow rate of 120 l/hr., and a by-product fraction was withdrawn from sectional packing material bed 14.

Step 8: 12 liters of eluant was introduced into sectional packed bed 7. The profile was shifted by way of recirculation to sectional packed bed 1, and a by-product fraction was withdrawn from sectional packed bed 6.

The betaine fraction was concentrated to a dry solids content of 55% and introduced into a separation system comprising three columns. FIG. 7 shows a schematic diagram of the apparatus. The columns had a diameter of 0.2 m, and the packed bed in each column had a height of 0.85 m. The packing material was the same as in Example 4.

Fractionation was performed sequentially by the following eight-step sequence:

Step 1: 2 liters of feed solution was introduced into column 1 at a flow rate of 60 l/hr., and a by-product fraction was withdrawn from column 2. 2.7 liters of eluant was supplied to column 3 at a flow rate of 80 l/hr., and sucrose fraction was withdrawn from column 3.

Step 2: 1.5 liters of feed solution was supplied to column 1 at a flow rate of 60 l/hr., and a sucrose fraction was withdrawn from column 3.

Step 3: 1.5 liters was recycled at a flow rate of 60 l/hr. in the loop formed by all columns.

Step 4: 3 liters of eluant was introduced into column 1 at a flow rate of 60 l/hr., and a betaine fraction was withdrawn from column 3.

Step 5: 1.8 liters of eluant was introduced into column 1 at a flow rate of 54 l/hr., and a by-product fraction was withdrawn from column 1. Simultaneously, 4 liters of eluant was supplied to column 2 at a flow rate of 120 l/hr., and a betaine fraction was withdrawn from column 3.

Step 6: 3 liters was recycled at a flow rate 60 l/hr. in the loop formed by all columns.

Step 7: 1.5 liters of eluant was introduced into column 1 at a flow rate of 60 l/hr., and a by-product fraction was withdrawn from column 3.

Step 8: 3 liters was recycled at a flow rate of 60 l/hr. in the loop formed by all columns.

With this procedure, the betaine separation has double the capacity of the first separation stage with respect to the amount of the betaine fraction produced. Thus, it was not attempted in this test to optimize the sequence with respect to capacity and energy consumption, but good yields and purities were pursued. This resulted in low fraction concentrations. It is obvious to those skilled in the art that, on an industrial scale, optimization is realized on an economic basis, thus the optimum values may vary from the values disclosed herein.

The filling and equilibration of the column system, sampling, and analyses of the samples were performed similarly as in Example 4. The concentration gradient from the output of sectional packed bed 14 in the first continuous SMB separation is shown in FIG. 8. The results are shown in Table 6, where the percentages of the different components are given as percent by weight on a dry solids basis.

As can be seen from FIG. 8, significantly better separation of betaine from sucrose is achieved compared to the fully continuous method of Example 5. Table 6 shows that with substantially similar column loads, the sequential method also yields a considerably higher purity of 92.3%, for the sucrose fraction than the 87.2-87.8% for the fully continuous method. Double separation permits the first fractionation to be performed with a relatively low sucrose yield, e.g., 84.1%, thus realizing the need for a high separation capacity and low evaporation. Double separation increases the sucrose yield to 91.0%. The betaine yield may easily be increased to 82.8%, and with a higher eluant quantity and column capacity, the betaine yield may exceed 90%.