MXPA97003940A - Procedure for preparing d-arabi - Google Patents

Abstract

Process for the preparation of D-arabitol characterized in that it has the following steps: - the hydrolysis of a lactose solution, - the oxidation of the mixture of glucose and galactose as-ad sustained in a mixture of gluconic and galactonic acids, - the decarboxylation of this mixture of gluconic and galactonic acids in a mixture of D-arabinose and D-lixose, - the catalytic hydrogenation of this mixture of D-arabirose and D-lixose in D-arabit

Description

PROCEDURE FOR PREPARING D-ARABITOL The subject of the present invention is a process for the preparation of D-arabitol from mixtures of glucose and galactose. Advantageously, the present invention relates to a process for the preparation of D-arabitol from lactose hydrolysates. D-arabitol is an important synthesis intermediate for the manufacture of xylitol. Xylitol is a substitute for acariogen in sucrose and as such has increased success, especially in sugar-free sweets. The methods for obtaining D-arabitol and transforming D-arabitol into xylitol are described, for example, in the United States Patents.

Numbers: 5,096,820 and 5,238,826 of which the Applicant is a transferee. These methods consist in microbially oxidizing D-arabitol, in D-xylulose, in isomerizing this D-xylulose to D-xylose, then in catalytically hydrogenating D-xylose in xylitol. The raw material that allows to obtain the D-arabitol is in this case the glucose that is either directly fermented in D-arabitol, or oxidized in gluconic acid which is itself decarboxylated in D-arabinose which is finally hydrogenated catalytically in D-arabitol. These procedures suffer from the disadvantage of using only glucose as a raw material. Another method that makes it possible to obtain D-arabitol after transforming D-arabitol into xylitol is also described in the international patent application 93/19030. It consists of isomerizing at high temperature, under strong hydrogen pressure and under the action of a ruthenium-based catalyst, a solution of D-arabitol. About 20 percent of arabitol is thus isomerized in xylitol. The xylitol is extracted after this mixture by chromatography and the unreacted D-arabitol is again subjected to isomerization. The raw material that makes it possible to obtain D-arabitol is in this case, either the glucose that is oxidatively degraded in D-arabinonic acid which is catalytically hydrogenated in D-arabitol after the lactonization of the acid, either the galactose that it is oxidatively degraded in D-lixonic acid which is catalytically hydrogenated in D-arabitol, also after the lactonization of the acid. Such a process then has the advantage of being able to use the lactose hydrolysates as a raw material in the manufacture of D-arabitol, but suffers from the enormous disadvantage that the catalytic hydrogenation of the D-arabinonic or D-lixonic acids in D-arabitol is effected Extremely bad and with very poor yields. The reason for this, as indicated by this patent application, is that the salts of these pentetic acids are not hydrogenated and that it is necessary first to transform these into esters or lactones. However, even using the very expensive ruthenium catalysts, the hydrogenation performance of lactones is very poor. The use of more classic RANEY-based nickel catalysts does not allow, even using weight by weight with respect to the D-arabino-l-4-lactone subjected to hydrogenation, to obtain better yields. Another method of obtaining d-arabitol from lactose is described in the patent of the United States of America Number: 4,156,076. The process described in this patent consists in oxidizing lactose in lactobionic acid, then in hydroxylating this acid in a mixture of galactose and gluconic acid. The sugar and acid are then separated and only the separated gluconic acid is used to obtain D-arabitol by decarboxylation in D-arabinose after hydrogenation. The portion of the raw material is lost in the form of galactose or must be valorized in another way. All these procedures for obtaining D-arabitol do not give satisfaction then because the first cited suffer from the disadvantage of using relatively expensive raw material which is glucose and the second of them, suffer from low reaction yields that come to heavily tax the extremely interesting price of lactose, together with its great availability and the difficulty of valuing it in another form. The object of the present invention is then to provide a method that allows to alleviate the aforementioned drawbacks and find a solution that valorises the lactose transforming it into D-arabitol. It has now been found, and this is the object of the present invention, a process for the preparation of D-arabitol characterized in that it has the following steps: the hydrolysis of a lactose solution, - the oxidation of the glucose mixture and of galactose thus obtained in a mixture of gluconic and galactonic acids, the decarboxylation of this mixture of gluconic and galactonic acids in a mixture of D-arabinose and D-lixose, the catalytic hydrogenation of this mixture of D-arabinose and D-lixose in D-arabitol. Although lactose is the preferred raw material in the process of the invention, mixtures of glucose and lactose or of galactose and lactose can also be used in this process. Likewise, certain vegetable hydrolysates of wood or straw, consisting almost exclusively of glucose and galactose, can also be used. The process of the invention thus has a first hydrolysis operation of a lactose solution, which allows this molecule to be cleaved in its two components and obtain an equimolar mixture of free glucose and galactose. The hemiacetalic function of the galactose that was bound in a glycoside bond with glucose is released by this hydrolysis and is then available for oxidation, in the same way as the hemiacetal function of the glucose that was already free in the lactose molecule. Without the separation of the components of the intermediate hydrolyzate, it is then linked to the stage of mixed oxidation of glucose and galactose, oxidation which transforms the hemiacetal functions of glucose and galactose into carboxylic functions. A mixture containing gluconic and galactonic acids is thus obtained. Always without separating the acids obtained, the process of the invention is then followed with the stage of decarboxylation of these acids. By cleavage of the carboxylic function and the restoration of a hemiacetal function on the central carbon structure thus cut out of acids, a mixture of D-arabinose and D-lixose is obtained. Finally, by the catalytic hydrogenation of this mixture of D-arabinose and D-lixose, and always without the separation of the two main components of the mixture, D-arabitol is obtained in the practically pure state since D-lixitol is synonymous with D-arabitol. The process of the invention is particularly interesting since it makes it possible to obtain D-arabitol with a very good reaction yield. Indeed, the entire lactose molecule is used, and this without ever having to proceed to some separation of the intermediates of the synthesis. Finally, another advantage and not of the minor ones of the process of the invention is that the catalytic hydrogenation step that has to be carried out on the hemiacetalic functions very easy to reduce in primary alcohols and this with the yields and selectivity close to 100% and with the help of cheap RANEY nickel catalysts, used in really catalytic amounts. It is therefore not necessary, with the method of the invention, to occupy the expensive ruthenium catalysts which are also contaminated with the lowest formic acid imprint, nor is it necessary to lactonize the acids that are formed intermediately or to use toxic and dangerous solvents as indicates the aforementioned international patent application WO 93/19030. The method of the invention thus makes it possible to value the entire lactose molecule at the best cost in a process that performs the preparation of D-arabitol. In accordance with the method of the invention, the lactose hydrolysis step is carried out either chemically or enzymatically. When the chemical route is used, it is preferred to use strong and non-oxidizing acids under the conditions of its use. Thus, hydrochloric or sulfuric acids are preferred while nitric acid is avoided as it leads to more complex reactions of lactose oxidation. The amounts of strong acid to be used are variable but remain catalytic. They are a function of the duration of hydrolysis, the temperature of this hydrolysis and the concentration of lactose in the solution to be hydrolyzed. Enzymatic hydrolysis of lactose can be carried out with the help of beta-galactosidase. The temperatures used in the case of enzymatic hydrolysis are weaker than those used when chemical hydrolysis of lactose is carried out. It will then be possible to use weaker concentrations of lactose to prevent any untimely crystallization of this sugar.

A particularly adapted enzyme is the beta-galactosidase LACTOZIMR marketed by NOVO. Whether it is carried out chemically or enzymatically, hydrolysis can be carried out batchwise or continuously but it is preferred to adjust the different parameters: duration, concentration of catalyst or enzyme, temperature, concentration of lactose. , so as to obtain a rate of lactose hydrolysis at least greater than 90 percent, preferably greater than 95 percent. These adjustments of the hydrolysis conditions are perfectly within reach of the technician who will be able to measure the hydrolysis rate in a simple way by means of the dosage of the reducing ends that appear, according to the BERTRAND method, for example, to mention only this method. According to the process of the invention, the sugars obtained are not isolated and the next step of oxidation of the mixture of glucose and galactose that is formed is directly linked in a mixture of gluconic and galactonic acids. In accordance with the process of the invention, this oxidation step can be conducted either chemically or microbiologically. The preferred chemical route in the process of the invention consists in oxidizing with the aid of air or oxygen in an alkaline medium and with the aid of palladium catalysts, the mixture of glucose and galactose obtained in the preceding step. A more particularly preferred method is that described in U.S. Patent Number: 4,845,208 of which the Applicant is also an assignee and which consists in using as an oxidation catalyst palladium fixed on active carbon and excited with bismuth. . It can also be considered in the process of the invention, to oxidize the glucose and galactose mixtures electrolytically or with the aid of hypobromite. When the procedure described in US Pat. No. 4,845,208 is employed, it is preferred to work with a concentration in sugars comprised between 10 and 40 percent, at a temperature between 25 and 60 ° C with an amount of palladium. expressed in terms of metal, comprised between 0.01 percent and 0.4 percent by weight of sugars. In these conditions where the oxygen supply is not limiting, the oxidation of the glucose and galactose mixture is completed in approximately 30 minutes to 5 hours. The preferred alkaline agent within the framework of the present invention is calcium carbonate or hydroxide. It is used to maintain a constant pH in the course of the oxidation reaction, preferably between 8.0 and 10.0. It is thus possible to proceed to the oxidation of the mixture of glucose and galactose by the microbial route. In this case, small quantities of mineral salts and nutritive elements are added to the aqueous solutions of the lactose hydrolyzate, which are comfortably added in the form of maceration liquor or yeast extract, for example. After sterilization of the culture medium thus obtained, a bacterial culture of a microorganism capable of oxidizing both glucose in gluconic acid and galactose in galactonic acid is sown. Microorganisms such as gluconobacter oxydans ATCC 19357 or ATCC 23773 are perfectly capable of oxidizing these two sugars in the corresponding aldonic acids. When these microorganisms are used, it is preferred to work at concentrations in sugars comprised between 100 and 250 grams per liter and at a temperature of 25 to 35 ° C. In conditions where the supply of oxygen to the culture medium is not a limiting factor, the oxidation of the glucose and galactose mixture is completed in approximately 15 to 30 hours. It is necessary to monitor, for bacterial oxidation to develop in satisfactory conditions, to maintain the pH at a value between 4.0 and 7.0. The preferred alkaline agent within the framework of the present inventionIt is also carbonate or calcium hydroxide. Whether the oxidation of the lactose hydrolyzate is chemically or microbially carried out, it is preferred to bring this up to the near disappearance of the reducing sugars and especially until the reducing sugars represent less than 10 percent and still more preferentially less than 5 percent. percent of the dry matter subjected to oxidation. At the end of this oxidation step, the reaction media is filtered at a temperature sufficient to maintain the calcium salts of the gluconic and galactonic acids in solution in order to eliminate either the catalyst or the microorganisms. To the knowledge of the Applicant, such a step of simultaneous oxidation of a mixture of galactose and glucose has never been described. According to the process of the invention, the calcium salts of the gluconic and galactonic acids thus obtained are not separated and the decarboxylation stage of these acids is directly linked to this mixture. Although this decarboxylation step could be conducted with the aid of hypochlorous acid, the method of the invention prefers to carry out the RUFF method using hydrogen peroxide.

To the Applicant's knowledge, such decarboxylation stage, although already described, for both calcium gluconate and calcium galactonate (FLETCHER HG et al., Journal of the American Chemical Society, 1950, Vol. 72, p.4456) however it has not been described for a mixture of these two salts. The reason why the method of the invention satisfies the best of the use of the calcium salts of gluconic and galactonic acids and the RUFF process, is that this decarboxylation stage of these hexonic acids leads to the corresponding pentoses, carrying thus to the calcium carbonate very little soluble. It is then possible by simple filtration of the calcium carbonate to strongly reduce the mineral charge of the resulting pentose solution. The use of salts other than calcium salts is theoretically possible but is clearly more expensive and less practical. Also, the use of sodium hypochlorite instead and instead of the hydrogen peroxide used in the RUFF process would also make this mineral load even heavier. According to a preferred mode of the process of the invention, the decarboxylation step is thus conducted on a solution of calcium salts of the gluconic and galactonic acids. It is preferred to operate on solutions containing 100 to 400 grams per liter and preferably 200 to 300 grams per liter of anhydrous calcium hexonates and at a temperature of about 30 to 50 ° C. This decarboxylation reaction is catalysed by the ferric ions. In the process of the invention it is preferred to add to the solution of calcium hexonates, ferric sulfate at a rate of 1 to 5 percent anhydrous ferric sulfate with respect to the weight of the anhydrous hexonates. These ferric ions can, however, be provided in another form. Under agitation, hydrogen peroxide is then slowly added, preferably in the form of hydrogen peroxide of 30 percent strength, at a rate of 120 to 140 milliliters approximately for 100 grams of anhydrous hexonates. The addition of hydrogen peroxide is carried out at an expense such that the temperature of the reaction medium does not rise beyond 50 ° C. The reaction is usually achieved in 2 to 8 hours and its term is marked by the appearance of a purple color. The abundant precipitate of calcium carbonate formed is filtered, preferably after the reaction medium has been allowed to cool. This reaction medium is demineralized immediately on a battery of ion-exchange resins consisting of a strong cationic resin and a weak anionic resin, resins to which a strong weak anionic cationic resin and / or a mixed bed of strong cationic and anionic resins. The mineral charge of the filtered reaction media can thus be reduced by ion exclusion chromatography techniques on the cationic resins loaded in the calcium form before proceeding to the complete demineralization step carried out with the resins mentioned above. After these purification treatments which can also comprise a decolorization step with the help of charcoal, a colorless solution is obtained which contains almost exclusively D-arabinose and D-lixose. The last stage of the process according to the invention consists in catalytically hydrogenating this. D-arabinose and D-lixose solution without separating the two sugars. As already stated above, such hydrogenation carried over the hemiacetal functions is extremely easy to perform. This is another of the great advantages of the process of the invention with respect to the process disclosed in the international patent application WO 93/19030 where the hydrogenation is carried over the esterified carboxylic functions of D-arabinonic and D-lixonic acids. To the knowledge of the Applicant, such hydrogenation of a coarsely equimolecular mixture of D-arabinose and D-lixose is equally novel. However, the hydrogenation of such a mixture of sugars is carried out in accordance with the rules of the art which, for example, lead to the production of sorbitol from glucose. Ruthenium-based catalysts as well as RANEY nickel catalysts can also be used for this stage. It is however preferred to use RANEY nickel catalysts which are less onerous. In practice, 1 to 10 weight percent of catalyst is used with respect to the dry matter of sugar subjected to hydrogenation. The hydrogenation is preferably carried out on syrups whose dry matter is comprised between 15 and 50 percent, in practice close to 30 to 45 percent, under a hydrogen pressure of between 20 and 200 bars. It can be done continuously or discontinuously. When operating discontinuously, the hydrogen pressure used is generally between 30 and 60 bars and the temperature at which the hydrogenation takes place is between 100 and 150 ° C. It is also monitored to maintain the pH of the hydrogenation medium by the addition of soda or soda carbonate, for example, but without exceeding a pH of 9.0. This way of doing it allows to avoid the appearance of other pentitoles such as xylitol or ribitol that can appear consecutively to an alkaline isomerization of D-lixose or D-arabinose. The reaction is stopped when the content of the reaction medium in reducing sugars becomes less than 1 percent, preferably still less than 0.5 percent. After cooling of the reaction medium, the catalyst is removed by filtration and the D-arabitol syrup thus obtained is demineralized on the cationic and anionic resins. In this state, the syrups contain at least 90 percent of D-arabitol and it is easy to purify them by crystallization after concentration and cooling of the solutions. However, the purity of the D-arabitol solutions obtained by the method of the invention renders such purification not generally necessary when it is desired to pursue the transformation of D-arabitol into xylitol by the methods described in US Pat. of North America Numbers: 5,096,820 or 5,238,826 or also in the international patent application 93/19030.

Thus, the method of the present invention allows, by means of the particular combination of the stages claimed, known in its principles but never described in this order of chaining on the mixtures of sugars or aldonic acids that are considered, to be easily and economically valued the lactose transforming it into D-arabitol. Since no synthesis intermediate needs to be isolated in the process of this invention, D-arabitol is obtained with an excellent yield, close to 50 percent of the lactose used. The invention will be better understood by means of the following example and which has the sole objective of better illustrating the invention without wishing to reduce it to the embodiment described.

EXAMPLE: 1 / LACTOSE HYDROLYSIS A dry matter of 15 percent crystallized lactose is placed in solution in water under its commercial form of monohydrated crystals. This lactose solution is brought to the temperature of 40 ° C and its pH is adjusted to 6.5. 12,000 LAU units of beta-galactosidase LACTOZIMR 3000 L marketed by NOVO NORDISK DANEMARK are then added to one hundred grams of lactose dry matter, representing 4 milliliters of enzyme and allowed to hydrolyze for 8 hours. The rate of lactose hydrolysis is then 95 percent. The solution is then brought to the boil to denature the enzyme, after which it is purified by decolorization and demineralization. 2 / OXIDATION IN GLUCONIC AND GALACTONIC ACIDS A catalyst constituted by activated carbon is prepared on which palladium has been successively deposited after bismuth at a rate of 35 percent bismuth with respect to the palladium as indicated in Example 4 of the United States of America Number: 4,845,208 whose teachings are incorporated in the present description by this reference. The oxidation of the lactose hydrolyzate is then carried out, the dry matter of which has been lowered to 12% by following the dilutions caused by its purification. For this, proceed in an aerated, agitated vessel with a temperature of 35 ° C, equipped with a probe to measure the pH, with the help of 3 percent of catalyst obtained as indicated with respect to the hydrolysed dry matter of lactose. All along the reaction that takes place in 3 hours 30 minutes, the pH of the reaction medium is maintained at a value between 8.5 and 9.1 by the progressive addition of a milk of lime. At the end of the reaction which results in a stoppage of lime consumption, the catalyst is filtered. The content of reducing sugars in this reaction medium is 1.5 percent. 3 / D-ARABINOSA AND D-LIXOSE DESCARBOXILATION To the filtration of the calcium salts of the gluconic and galactonic acids thus obtained, a solution of ferric sulfate is added in order to obtain a concentration of 3% of anhydrous ferric sulfate with respect to to the dry matter of the calcium hexonates used. The temperature of this solution is stirred and set at 40 ° C, then hydrogen peroxide is slowly added at a concentration of 30 percent until a persistent purple-brown color is obtained from the reaction medium. This addition of the hydrogen peroxide was carried out in 6 hours and 130 milliliters of hydrogen peroxide had to be added for 100 grams of calcium hexonates put into practice. After cooling the reaction medium for a few hours, the calcium carbonate is filtered, then this filtrate is demineralized on a strong cationic resin regenerated in the form of hydrogen, then a weak anionic resin regenerated in the hydroxyl form.

This gives a virtually colorless D-arabinose and D-lixose solution, which is concentrated to 40 percent dry matter. 4 / HYDROGENATION IN D-ARABITOL In a stirred reactor with thermostat, this syrup of pentoses is introduced to which 5 percent of the catalyst was added to the nickel of RANEY, then this reactor is subjected to a hydrogen pressure of 50 bars and it is heated to 120 ° C. The course of this hydrogenation is maintained, the pH close to neutrality with the help of a soda carbonate solution. After 5 hours, the reducing sugar content of the reaction medium becomes equal to 0.15 percent and the reactor is then cooled. The catalyst is filtered after demineralizing the hydrogenated syrup over strong cationic and anionic resins. A colorless syrup containing 92 percent D-arabitol is obtained.

Claims (9)

1. A method of preparing D-arabitol characterized in that it has the following steps: the hydrolysis of a lactose solution, the oxidation of the glucose and galactose mixture thus obtained in a mixture of gluconic and galactonic acids, the decarboxylation of this mixture of gluconic and galactonic acids in a mixture of D-arabinose and D-lixose, the catalytic hydrogenation of this mixture of D-arabinose and D-lixose in D-arabitol.

2. The process for the preparation of D-arabitol according to claim 1, characterized in that the hydrolysis of the lactose solution is carried out chemically.

3. The process for the preparation of D-arabitol according to claim 1, characterized in that the hydrolysis of the lactose solution is carried out enzymatically.

4. The process for the preparation of D-arabitol according to claims 3 characterized in that the oxidation of the mixture of glucose and galactose is carried out with the help of air or oxygen, in an alkaline medium, and with the help of palladium catalysts.

5. The process for the preparation of D-arabitol according to claims 1 to 3, characterized in that the oxidation of the glucose and galactose mixture is carried out by microbial means with the help of microorganisms of the genus gluconobacter oxydans.

6. The manufacturing process of D-arabitol according to claims 1 to 5, characterized in that the gluconic and galactonic acids are neutralized in the form of calcium salts.

7. The process for manufacturing D-arabitol according to claims 1 to 6, characterized in that the decarboxylation of the mixture of gluconic and galactonic acids is carried out with the help of hydrogen peroxide.

8. The process for manufacturing D-arabitol according to claims 1 to 7, characterized in that the catalytic hydrogenation of the mixture of D-arabinose and D-lixose is carried out with the help of RANEY nickel.

9. The use of the D-arabitol obtained according to claims 1 to 8 for the manufacture of xylitol.