Process for acid dehydration of sugar alcohols

Abstract

A process is described for the acid-catalyzed dehydration of a sugar alcohol, wherein the catalyst comprises a water-tolerant Lewis acid. In particular embodiments, the catalyst comprises a homogeneous water-tolerant Lewis acid, especially a homogeneous Lewis acid selected from the group consisting of bismuth (III) triflate, gallium (III) triflate, scandium (III) triflate, aluminum triflate, tin (II) triflate and indium (III) triflate. Such catalysts are effective for dehydrating both of sorbitol and the 1,4-sorbitan dehydration precursor of isosorbide, and bismuth (III) triflate particularly is beneficial for dehydrating mannitol to isomannide.

Claims

1. A process for producing an isohexide from a starting material of one or both of a hexitol and a monoanhydrohexitol, comprising contacting the starting material with from 0.005 mol percent to 0.1 mol percent of a homogeneous Lewis acid catalyst selected from the group consisting of bismuth (III) triflate, gallium (III) triflate, scandium (III) triflate, aluminum triflate, indium (III) triflate, tin (II) triflate and combinations of two or more of these, at a temperature of from 140 degrees to 160 degrees and over a period of from 1 hour to 3 hours, under reduced pressure and with continuous removal of water from the product mixture in the course of the dehydration.

2. The process of claim 1, wherein the starting material is sorbitol.

3. The process of claim 1, wherein the starting material is mannitol and the catalyst is bismuth (III) triflate.

Description

DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS

(1) A preferred process according to the present invention for dehydrating sorbitol involves mixing sorbitol with from 0.005 mol percent and greater of a water-tolerant Lewis acid, heating to at least 140 degrees Celsius, and carrying out the acid -catalyzed dehydration of sorbitol isothermally for an hour or longer under a reduced pressure to continuously remove water from the reaction. The water-tolerant Lewis acid is preferably one or more of bismuth (III) triflate, gallium (III) triflate, scandium (III) triflate, aluminum triflate, tin (II) triflate and indium (III) triflate, and while yields of isosorbide and the 1,4-sorbitan precursor of isosorbide obtained from these catalysts can be seen from the examples below to vary somewhat dependent on the catalyst used, the catalyst loading and reaction conditions of temperature and duration, it is expected that catalyst loadings of not more than 0.1 mol percent, temperatures of not more than 160 degrees Celsius and reaction times of not more than 3 hours will provide commercially acceptable yields of isosorbide. The crude product mixture may then be purified according to any of the known methods for doing so.

(2) As will be evident from the examples that follow, use of the preferred Lewis acids under these conditions provides a number of benefits, including enhanced yields of isosorbide and of the 1,4-sorbitan precursor of isosorbide as compared to the most effective Brnsted acid surveyed, namely, sulfuric acid, much reduced catalyst loadings for achieving a targeted yield of isosorbide, avoidance of the neutralization requirements posed by the conventional Brnsted acids before distillation of the crude product mixture and better color of the crystalline isosorbide distillates that may be realized.

(3) While some or all of these benefits are expected to be attainable in the dehydration of sugar alcohols generally (where sugar alcohols is understood to include partially dehydrated sugar alcohols such as, for example, monoanhydrohexitols from the partial dehydration of hexitols), the extent to which certain benefits or advantages are observed, the particular water-tolerant Lewis acid catalysts that prove most effective and the optimum process conditions for carrying out the Lewis acid-catalyzed dehydrations can be expected to vary somewhat from one sugar alcohol to the next. As an example, we found bismuth triflate to be particularly advantageous for catalyzing the dehydration of mannitol to isomannide. Those skilled in the art will be well able, in any event, to determine the optimum features of a process for dehydrating a particular sugar alcohol using a water-tolerant, Lewis acid catalyst as claimed herein by routine experimentation.

(4) The present invention is further illustrated by the following, non-limiting examples:

COMPARATIVE EXAMPLES 1-6

(5) For benchmarking the performance of the water-tolerant Lewis acid catalysts of the present invention, a number of Brnsted acids were evaluated for the acid-catalyzed dehydration of sorbitol. In each instance, a three neck, 250 mL round bottomed flask equipped with a magnetic stir bar was charged with 100 grams of sorbitol (0.549 mol), then was immersed in an oil bath set at 140 degrees Celsius. Once the sorbitol liquefied and attained an internal temperature of 140 degrees as determined by an internal temperature probe, a quantity (2 mol percent in all cases except for phosphoric acid, which was added at 5 mol percent) of the Brnsted acid in question was introduced by syringe through a rubber septum-capped neck. Under a reduced pressure of less than 5 torr, the reaction was then continued isothermally for 1 hour. After this time, the vacuum was broken, the crude product mixture was cooled and quenched with 50 percent aqueous sodium hydroxide, then was weighed and quantitatively analyzed by gas chromatography. The results, shown in Table 1, show that sulfuric acid was the most effective Brnsted acid of those surveyed for dehydrating sorbitol under the indicated conditions, though unidentified side products accounted for about 23 percent of the crude product mixture.

(6) TABLE-US-00001 TABLE 1 % Iso- 1,4- 2,5- sorbitol sorbide sorbitan sorbitan Account- conver- yield yield yield ability Acid pKa sion (mol %) (mol %) (mol %) (wt %) Sulfuric 5 100.00 67.72 0.00 9.30 76.95 p-toluene- 2.8 100.00 22.83 54.90 9.30 87.00 sulfonic Methane- 1.9 100.00 18.20 59.30 8.70 86.20 sulfonic Oxalic 1.25 12.88 0.00 3.62 1.48 92.22 Betaine HCl 1.84 14.30 0.00 4.40 1.33 91.14 Phosphoric 2.14 72.00 5.18 55.98 3.15 92.31

COMPARATIVE EXAMPLE 7 AND EXAMPLES 1-6

(7) The same experimental setup, procedure and conditions were used as in Comparative Examples 1-8, except that 0.1 mol percent of various Lewis acids (for Examples 1-6) or 0.1 mol percent of sulfuric acid (for Comparative Example 7) was used. The results were as shown in Table 2, follows:

(8) TABLE-US-00002 TABLE 2 Isosorbide 1,4-sorbitan 2,5-sorbitan Account- % sorbitol yield (mol yield (mol yield (mol ability Acid conversion %) %) %) (wt %) Bi(OTf).sub.3 83.49 7.05 69.78 6.44 100.00 In(OTf).sub.3 86.11 7.70 66.77 6.99 98.40 Sc(OTf).sub.3 94.73 12.93 75.02 7.97 100.00 Ga(OTf).sub.3 95.13 12.50 72.36 7.76 99.64 Sn(OTf).sub.3 53.10 2.38 43.28 3.67 99.17 Al(OTf).sub.3 80.04 5.76 65.51 5.84 98.73 Sulfuric 62.97 3.39 54.86 4.42 100.00

COMPARATIVE EXAMPLE 8 AND EXAMPLES 7-12

(9) The same experimental setup, procedure and conditions were used as in Comparative Example 7 and Examples 1-6 (0.1 mol percent of catalyst), except that the reaction was continued for 2 hours at 140 degrees Celsius after introduction of the catalyst, as opposed to 1 hour. The results are shown in Table 3:

(10) TABLE-US-00003 TABLE 3 Isosorbide 1,4-sorbitan 2,5-sorbitan Account- % sorbitol yield (mol yield (mol yield (mol ability Acid conversion %) %) %) (wt %) Bi(OTf).sub.3 99.53 23.73 61.06 7.67 94.98 In(OTf).sub.3 98.79 19.20 66.41 7.74 100.00 Sc(OTf).sub.3 99.58 25.68 56.64 8.77 96.43 Ga(OTf).sub.3 99.88 31.59 49.67 7.39 92.09 Sn(OTf).sub.3 94.06 12.70 73.49 8.03 100.00 Al(OTf).sub.3 100.00 29.40 53.87 8.00 94.10 Sulfuric 83.63 6.33 68.34 6.23 98.87

COMPARATIVE EXAMPLE 9 AND EXAMPLES 13-18

(11) The same experimental setup, procedure and conditions were used as in Comparative Example 8 and Examples 7-12 (0.1 mol percent of catalyst), except that the reaction was continued for 3 hours at 140 degrees Celsius after introduction of the catalyst, as opposed to 2 hours. The results are shown in Table 4:

(12) TABLE-US-00004 TABLE 4 Isosorbide 1,4-sorbitan 2,5-sorbitan Account- % sorbitol yield (mol yield (mol yield (mol ability Acid conversion %) %) %) (wt %) Bi(OTf).sub.3 99.86 32.45 51.09 7.33 92.67 In(OTf).sub.3 100.00 44.85 36.27 8.24 90.15 Sc(OTf).sub.3 100.00 49.35 32.28 9.01 88.40 Ga(OTf).sub.3 100.00 67.20 3.96 6.96 79.37 Sn(OTf).sub.3 100.00 24.02 66.37 7.90 100.00 Al(OTf).sub.3 100 00 47.13 31.85 7.70 88.98 Sulfuric 100.00 25.30 60.22 5.60 91.88

COMPARATIVE EXAMPLE 10 AND EXAMPLES 19-24

(13) The same experimental setup and procedure were used as in previous examples, except that the reaction temperature was increased to 160 degrees Celsius, and the reaction was continued for 1 hour after introduction of the acid catalyst being evaluated (again at 0.1 mol percent). Results were as shown in Table 5:

(14) TABLE-US-00005 TABLE 5 Isosorbide 1,4-sorbitan 2,5-sorbitan Account- % sorbitol yield (mol yield (mol yield (mol ability Acid conversion %) %) %) (wt %) Bi(OTf).sub.3 100.00 62.02 3.83 7.34 76.21 In(OTf).sub.3 100.00 68.40 8.74 7.89 83.63 Sc(OTf).sub.3 100.00 32.10 46.21 8.15 89.13 Ga(OTf).sub.3 100.00 64.62 5.07 6.23 77.71 Sn(OTf).sub.3 97.95 17.95 64.64 9.68 97.35 Al(OTf).sub.3 100.00 48.01 28.48 7.64 84.05 Sulfuric 100.00 49.78 26.35 8.70 88.92

COMPARATIVE EXAMPLE 11 AND EXAMPLES 25-30

(15) The acids were evaluated at a lower catalyst load of 0.05 mol percent, the lower temperature of 140 degrees Celsius and with a reaction time of two hours, with the results shown in Table 6 as follows:

(16) TABLE-US-00006 TABLE 6 Isosorbide 1,4-sorbitan 2,5-sorbitan Account- % sorbitol yield (mol yield (mol yield (mol ability Acid conversion %) %) %) (wt %) Bi(OTf).sub.3 98.77 22.25 68.13 8.47 100.00 In(OTf).sub.3 92.22 11.16 72.98 7.80 100.00 Sc(OTf).sub.3 94.92 15.76 67.33 6.91 95.87 Ga(OTf).sub.3 97.72 15.83 67.27 8.83 92.98 Sn(OTf).sub.3 56.84 2.70 49.76 3.97 100.00 Al(OTf).sub.3 80.69 6.37 69.11 6.34 100.00 Sulfuric 58.20 3.36 57.81 6.37 97.53

COMPARATIVE EXAMPLE 12 AND EXAMPLES 31-36

(17) The acids were evaluated at the lower catalyst load of 0.05 mol percent used in Examples 25-30, but at the higher temperature of 160 degrees Celsius and with a reaction time of one hour rather than two after introduction of the catalyst being evaluated, with the results shown in Table 7 as follows:

(18) TABLE-US-00007 TABLE 7 Isosorbide 1,4-sorbitan 2,5-sorbitan Account- % sorbitol yield (mol yield (mol yield (mol ability Acid conversion %) %) %) (wt %) Bi(OTf).sub.3 99.43 25.25 60.18 9.13 97.87 In(OTf).sub.3 100.00 31.12 55.71 9.83 92.75 Sc(OTf).sub.3 96.69 15.38 68.14 8.43 97.99 Ga(OTf).sub.3 100.00 71.31 7.64 8.30 86.95 Sn(OTf).sub.3 85.51 10.65 69.05 7.17 100.00 Al(OTf).sub.3 100.00 26.08 58.80 8.85 95.65 Sulfuric 69.22 4.05 53.67 5.52 95.53

COMPARATIVE EXAMPLE 13 AND EXAMPLES 37-42

(19) The acids were evaluated at a still lower catalyst load of 0.01 mol percent, at a temperature of 160 degrees Celsius and with a reaction time of one hour after introduction of the catalyst being evaluated, with the results shown in Table 8 as follows:

(20) TABLE-US-00008 TABLE 8 Isosorbide 1,4-sorbitan 2,5-sorbitan Account- % sorbitol yield (mol yield (mol yield (mol ability Acid conversion %) %) %) (wt %) Bi(OTf).sub.3 58.18 2.95 49.02 4.92 100.00 In(OTf).sub.3 71.67 5.19 58.41 6.23 98.70 Sc(OTf).sub.3 40.40 1.74 34.48 3.43 99.46 Ga(OTf).sub.3 67.72 4.42 58.30 6.08 99.65 Sn(OTf).sub.3 71.00 4.86 58.58 6.10 99.21 Al(OTf).sub.3 64.41 3.56 54.54 5.65 100.00 Sulfuric 26.90 0.00 26.17 0.61 99.98

COMPARATIVE EXAMPLE 14 AND EXAMPLES 43-44

(21) The acids were evaluated at a still lower catalyst load of 0.005 mol percent, at a temperature of 160 degrees Celsius and with a reaction time of one hour after introduction of the catalyst being evaluated, with the results shown in Table 9 as follows:

(22) TABLE-US-00009 TABLE 9 Isosorbide 1,4-sorbitan 2,5-sorbitan Account- % sorbitol yield (mol yield (mol yield (mol ability Acid conversion %) %) %) (wt %) Bi(OTf).sub.3 71.06 4.68 58.47 6.38 100.00 In(OTf).sub.3 88.69 9.99 68.93 8.26 100.00 Sulfuric 21.10 0.00 19.22 0.87 100.00

COMPARATIVE EXAMPLES 15 and 16, WITH EXAMPLE 45

(23) For these examples, two runs were conducted using differing amounts of sulfuric acid (0.1 mol percent for Comparative Example 15 and 1 mol percent for Comparative Example 16) to catalyze the dehydration of mannitol to isomannide and anhydromannitols, and the results were compared to a run using 0.1 mol percent of bismuth (III) triflate under the same conditions of 160 degrees Celsius, one hour run time and a reduced pressure of 20 torr.

(24) For the two sulfuric acid experiments, a three neck 250 mL round bottomed flask equipped with a magnetic stir bar was charged with 100 grams of mannitol (0.549 mol), then immersed in an oil bath maintained at 160 degrees Celsius. Once the mannitol liquefied and attained an internal temperature of 160 degrees as measured by an internal temperature probe, a condenser was outfitted onto one of the flask necks and vacuum was initiated. The sulfuric acid was then introduced via syringe through a rubber septum capped neck. After an hour, the vacuum was broken, and the crude product mixture was cooled, weighed and quantitatively analyzed by gas chromatography.

(25) For the run with the inventive bismuth triflate catalyst, a three neck 250 ml round bottomed flask was charged with the mannitol and with 360 milligrams of the bismuth triflate catalyst, than immersed in the 160 degree Celsius oil bath. Once the mannitol liquefied and the bismuth triflate dissolved in the mannitol, and as the mixture achieved an internal temperature of 160 degrees Celsius, then a condenser was outfitted onto one of the flask necks and vacuum was initiated down to a pressure of 20 torr. After one hour, the vacuum was broken, and the crude product mixture was cooled, weighed and quantitatively analyzed by gas chromatography.

(26) The results were that 100% conversion of the mannitol was realized in all three runs, the yields of isomannide (expressed in mol percents) were much greater using the bismuth triflate catalyst: sulfuric acid at 0.1 mol percent gave only 2 percent of isomannide, whereas at 1 mol percent addition the isomannide yield was 25 percent. However, by comparison, the inventive bismuth triflate gave 61 percent of isomannide.