Process for the preparation of fructose

Abstract

Fructose is produced by the isomerization of glucose and/or mannose by a process comprising isomerizing the glucose or mannose in an alkali solution containing an aryl boric acid.

Description

The present invention relates to a process for the production of fructose by the isomerization of glucose or mannose in alkaline solution.

The base-catalyzed isomerization of glucose, mannose and fructose has been known for a long time; it is known by the name of its discoverers as the "Lobry de Bruyn-Alberda van Ekenstein rearrangement" (see Rec. trav. chim. Pays-Bas, 14, 203/1895 and 15, 92/1896). Since that time, the reaction has been thoroughly investigated by many workers. It has been ascertained that a large variety of bases, for example, sodium hydroxide, sodium carbonate, basic ion exchangers, ammonia and the like, catalyze the adjustment of the equilibrium but that the yield of fructose formed is only about 20-30% (see Gottfried and Benjamin, Ind. Eng. Chem., 44, 141/1952).

It is also known that in the case of the alkaline isomerization of glucose, the yield of fructose can be considerably increased when working in the presence of borates (see Mendicino, J.A.C.S., 82, 4975/1960). However, a disadvantage of this reaction is that only relatively dilute sugar solutions can be used and the high concentrations of the reagents employed make the working up of the solution for obtaining pure fructose uneconomical.

We have now found that the base-catalyzed isomerization of glucose and mannose to give fructose can be favorably influenced by using the process of the invention.

Essentially, the present invention comprises a process for the isomerization of glucose and/or mannose in alkaline solution, wherein the solution contains an aryl boric acid. Depending upon the alkali content, the temperature and the amount and nature of the aryl boric acid added, fructose yields of up to and beyond 80% can be obtained.

As the boric acid, there can be used, for example, phenyl boric acid, naphthyl boric acid, alkoxyphenyl boric acids, such as 4-methoxyphenyl boric acid, nitrophenyl boric acids or sulfonated phenyl or naphthyl boric acids. The process can be carried out especially advantageously when the aryl boric acid is a component of an insoluble organic polymer. Such insoluble polymers can be obtained, for example, by polymerizing a vinylphenyl boric acid or a derivative thereof with itself or with styrene or a similar compound or by fixing an aryl boric acid, via a further substituent, on to a reactive carrier material, for example, by fixing a sulfonylphenyl boric acid on to an anion exchanger resin.

The process according to the present invention can be carried out with the use of glucose solutions with a concentration of up to 30 or 40%; it is especially advantageous to use 20-25% glucose solutions. Instead of pure glucose, there can also be used appropriate crude starch hydrolyzates or invert sugar solutions from which a part of the fructose has been removed in known manner.

The reaction temperature used is usually between about 20.degree. and 80.degree.C and preferably between 45.degree. and 50.degree.C. In the case of higher temperatures, the decomposition of the fructose and of the aryl boric acids becomes more noticeable as a disturbing side reaction and, at comparatively low temperatures, the rate of the reaction is, as is to be expected, lower so that the reaction is uneconomic.

Since it is a base-catalyzed reaction, the rate of the reaction increases with increasing pH value but, on the other hand, the irreversible decomposition of the fructose is favored by high pH values. Consequently, pH values of between 10 and 14 have proved to be especially favorable because, on the one hand, the rate of formation of the fructose is great enough and, on the other hand, the irreversible decomposition of the fructose formed is not important. pH values between 11 and 12.5 are regarded as being optimal.

Our investigations have shown that, per mole of glucose, the reaction solution should contain about 1/2 to 1 mole of aryl boric acid in order to bring about a maximum formation of fructose. When using smaller amounts of aryl boric acid, the yield of fructose is impaired and, when using higher amounts of aryl boric acid, the rate of reaction drops very considerably, which is surprising and which naturally impairs the economy of the process.

The separation of the fructose formed from the reaction solution can be carried out in known manner, for example, by separation by means of an ion exchanger or by precipitation of the fructose from a neutral solution in the form of calcium fructosate. When the aryl boric acid is added to the reaction solution in the form of a component of an insoluble polymer, the fructose formed is preponderantly also fixed in complex form on the resin and can be removed with the resin from the reaction solution. By washing out the separated resin with dilute hydrochloric acid, there is obtained a fructose-enriched solution which can then be further worked up in known manner.

Although aryl boric acids are relatively stable under the conditions used for the process according to the present invention, the boric acids included in a polymeric resin are, nevertheless, partially decomposed at comparatively high temperatures. Consequently, an especially advantageous variant of the process according to the present invention is to carry out the isomerization of glucose at a temperature of about 30.degree. to 60.degree.C in solution in the absence of an aryl boric acid, subsequently to cool the solution and then to pass it through an aryl boric acid-containing polymer in which the fructose formed is almost completely absorbed, whereupon the fructose-impoverished solution can again be heated up and recycled to the reaction zone. When the resin used has a sufficient absorption capacity, repeated passage in this manner enables fructose yields to be obtained which are similar to those obtained when carrying out the isomerization in the presence of an aryl boric acid.

The following Examples are given for the purpose of illustrating the present invention:

EXAMPLE 1

ISOMERIZATION IN THE PRESENCE OF PHENYL BORIC ACID

2.0 g radioactive marked D-glucose were added to a solution of 0.26 g sodium hydroxide, 1.15 g phenyl boric acid anhydride in 4.0 g water. After the addition of 0.39 ml of a 50% solution of sodium hydroxide, there was obtained a clear solution with a pH of 12.17 at 24.5.degree.C. The solution was placed in a glass flask under an atmosphere of nitrogen, further heated for 1.7 hours at 50.degree.C and thereafter stored at -15.degree.C until analyzed.

An aliquot thereof was diluted and determined according to the automated resorcinol method (see Yaphe et al., Anal. Biol. Chem., 13, 143/1965). The analysis gave a fructose content of 49.4% of theory.

Aliquots which had been chromatographed, either directly or after acidification with sulfuric acid, over a borate ion-charged anion exchanger gave a fructose yield of 46.7% of theory (also determined according to the resorcinol method).

A further aliquot was chromatographed over an anion exchanger in the borate form, the fructose-containing fractions were mixed with 5 g unmarked fructose and the total mixture freed from residual boric acid by means of "Borasorb" (Borasorb is a registered trademark for a boron absorbent resin available from Calbiochem, Los Angeles). In order to remove the residual glucose and mannose, the di-O-isopropylidene derivative of fructose was prepared according to the method of Bell (J. Chem. Soc., 1952, 3760) and purified by extraction with chloroform and recrystallization from petroleum ether. The radioactive determination corresponds to a yield of 44.5 % fructose.

EXAMPLE 2

Determination of the important reaction parameters for the isomerization of glucose to fructose

General process: A precise amount of an aryl boric acid anhydride was mixed in a reagent glass with a known amount of water and a known amount of sodium hydroxide. The solution or suspension obtained was mixed with a definite amount of glucose and stirred until solution was complete. The pH value of the solution was measured at ambient temperature and, when necessary, adjusted precisely by the addition of a small amount of sodium hydroxide solution. Parts of this solution were placed in small glass flasks which were closed under an atmosphere of nitrogen. The flasks were maintained for a definite period of time at a constant temperature and subsequently cooled to 0.degree. or to -15.degree.C and, at this temperature, stored until analyzed. The analyses were carried out according to methods described above in Example 1. The results obtained from the various experiments are summarized in the following Table I.

TABLE I

Effects of various reaction conditions on the isomerisation of glucose to fructose in alkaline solution in the presence of aryl boric acids in concentrated solutions

aryl boric acid I II III IV V VI __________________________________________________________________________ phenyl boric acid 1:1 12.05 25.1 50.0 57 120 1:1 12.16 25.7 37.0 55 340 1:1 12.16 25.7 50.5 50 90 1:1 12.16 25.7 61.5 50 15 1:1 10.95 26.5 50.0 34 720 1:1 -- 24.9 50.0 53 21 1:1 11.97 24.8 50.0 55 150 1:1 12.08 24.9 50.0 52 110 2:1 12.50 17.8 50.0 19 60 1:2 12.00 27.9 50.0 55 95 0:1 12.03 30.2 50.0 40 90 sulphonate phenyl boric acid 1:1 12.00 26.4 50.0 46 140 4-methoxy-phenyl boric acid 1:1 12.14 25.1 50.0 55 120 3-nitrophenyl boric acid 1:1 12.04 24.6 50.0 18 90 p-tolyl boric acid 1:1 12.00 3.6 50.0 68 240 1:1 12.00 9.0 50.0 63 250 1:1 12.00 14.4 50.0 61 220 1:1 12.00 23.4 50.0 56 250 __________________________________________________________________________ I = molar ratio of aryl boric acid to D-glucose II = pH of the solution III = initial concentration of D-glucose (% by weight) IV = temperature of the reaction in .degree.C. V = maximum yield of D-fructose (% by weight) VI = time of maximum yield in minutes

EXAMPLE 3

Isomerization of sugars in alkaline solutions in the presence of aryl boric acids in dilute solution

The results summarized in the following Table II were obtained by using the experimental procedures described in Example 2 (temperature 50.degree.C):

TABLE II __________________________________________________________________________ aryl boric starting ratio of sugar pH of % yield reaction acid material aryl boric content solu- of D- time in acid to in mole tion fructose min. sugar __________________________________________________________________________ phenyl boric D-glucose 1:1 0.02 12 73 270 acid D-mannose 1:1 0.02 12 53 270 D-fructose 1:1 0.02 12 77 270 D-glucose 1:1 0.02 13 66 120 D-mannose 1:1 0.02 13 65 120 D-fructose 1:1 0.02 13 51 120 D-glucose 2:1 0.02 12 40 210 D-mannose 2:1 0.02 12 36 210 D-fructose 2:1 0.02 12 83 210 D-glucose 1:1 0.125 12 70 240 73 480 D-glucose 1:1 0.125 11 17 600 21 1200 4-methoxy- D-glucose 1:1 0.02 12.1 81 230 phenyl boric acid 61 720 3-nitrophenyl D-glucose 1:1 0.02 12 81 280 boric acid 70 720 p-tolyl D-glucose 1:1 0.02 12.01 8 240 boric acid 1:1 0.02 11.0 7 500 1:1 0.02 13.0 77 130 __________________________________________________________________________

EXAMPLE 4

Production of poly-(4-vinylphenyl boric acid) resin

1.1 g iminodiethyl-4-vinylphenyl boric acid, 0.17 g divinyl-benzene/ethyl vinyl benzene (1:1), 0.76 g styrene and 0.045 g azo-bis-isobutyric acid nitrile were dissolved in 10 ml chloroform, deaerated by passing through nitrogen and heated in a closed flask for four hours at 70.degree.C. The gel obtained was dried and washed with 1N hydrochloric acid, until the wash water no longer gave a yellow coloration with periodate/pentane-2,4-dione. Subsequently, it was washed with distilled water and dried over phosphorus pentoxide. The yield of poly-(4-vinylphenyl boric acid) resin was 81%. The dried resin was sieved and the fractions of 60-120 mesh were used for the following experiments.

EXAMPLE 5

Isomerization of glucose and fructose in alkaline solution on poly-(4-vinylphenyl boric acid) resin

0.68 g of the resin prepared according to Example 4 was stirred at various temperatures in a thermostatically controlled vessel in an aqueous solution of glucose, fructose and sodium hydroxide, the solution having a pH of 12.0. Subsequently, the resin was separated off from the supernatant solution and the absorbed sugars were washed out with dilute hydrochloric acid. The supernatant solution and wash water were investigated for their content of glucose and fructose according to the cystein-sulfuric acid and according to the resorcinol method. The results obtained are summarized in the following Table III.

TABLE III __________________________________________________________________________ Isomerization and distribution of D-glucose and D-fructose on poly-(4-vinylphenyl boric acid) resins at pH 12 __________________________________________________________________________ expt. sugar content reaction temper- sugar content sugar content total of starting time in ature of supernatant of resin (mg) recov- solution (mg) minutes solution (mg) ery % __________________________________________________________________________ D-glu- D-fruc- glu- fruc- glu- fruc- cose tose cose tose cose tose __________________________________________________________________________ A 350 350 15 20.degree.C. 265 83 63 245 94 B 350 0 240 50.degree.C. 77 21 68 155 91.3 __________________________________________________________________________

EXAMPLE 6

Use of poly-(4-vinylphenyl boric acid) resin in a reactor

2 g of a poly-(4-vinylphenyl boric acid) resin prepared according to Example 4 were placed in a column with a diameter of 1.3 cm, which was thermostatically controlled at 19.degree.C. The inlet and outlet of the column were connected via pipes and an appropriate pump with a reaction circuit. The resin was first washed with 1N soda lye until the solution running off had a pH of 12.0. Subsequently, the total system was filled with a 1N aqueous sodium hydroxide solution to which 0.68 g D-glucose was added. With a pump capacity of about 0.23 ml per minute, the solution was pumped around the reaction circuit at various temperatures and subsequently, the content of glucose and fructose thereof was investigated. The resin was, after separation of the supernatant solution, washed out with hydrochloric acid and the fructose and glucose content of the wash water also investigated. The results of the various experiments are summarized in the following Table IV:

TABLE IV __________________________________________________________________________ Isomerization of D-glucose to D-fructose on poly-(4-vinylphenyl boric acid) resin in a reactor. __________________________________________________________________________ expt. D-fructose in D-fructose yield of temp. of reaction no. supernatant absorbed on D-fructose reaction time in solution (g) resin (g) (%) circuit hours __________________________________________________________________________ 1 0.161 0.224 56.5 37.degree.C. 22 2 0.219 0.348 56.7 37.degree.C. 46 3 0.018 0.342 52.8 50.degree.C. 6 4 0.652 0.406 42.3 37.degree.C. 21.5 __________________________________________________________________________

It will be understood that the specification and examples are illustrative but not limitative of the present invention and that other embodiments within the spirit and scope of the invention will suggest themselves to those skilled in the art.

Claims

What is claimed is:

1. Process for the production of fructose by the isomerization of a sugar selected from the group consisting of glucose, mannose and mixtures thereof comprising isomerizing said sugar in an alkali solution containing an aryl boric acid.

2. Process as claimed in claim 1, wherein the solution has a pH value of between 10 and 14.

3. Process as claimed in claim 2, wherein the solution has a pH value of between 11 and 12.5.

4. Process as claimed in claim 1, wherein the process is carried out in aqueous solution at a temperature of between 20.degree. and 80.degree.C.

5. Process as claimed in claim 4, wherein the process is carried out at a temperature of between 45.degree. and 50.degree.C.

6. Process as claimed in claim 1, wherein the aryl boric acid is a component of an insoluble organic polymer.

7. Process as claimed in claim 6, wherein the polymer is a polymer or copolymer of vinylphenyl boric acid or of a derivative thereof.

8. Process as claimed in claim 6, wherein the polymer is an ion exchanger resin loaded with a phenyl boric acid derivative.