Conversion of carbohydrates to levulinic acid esters

Abstract

The present invention relates to the field of converting carbohydrates into levulinic acid, a platform chemical for many chemical end products. More specifically the invention relates to a method for converting carbohydrates such as mono-, di- or polysaccharides, obtained from for example biomass production into a suitable levulinic acid ester in the presence of a zeolite or zeotype catalyst and a suitable alcohol, and the ester may be further converted into levulinic acid if desired.

Claims

1. A method of converting a saccharide into a levulinic acid ester, comprising heating the saccharide to a temperature between 120 C. and 200 C. in the presence of a zeolite or zeotype catalyst with a pore size diameter of at least 6 , wherein the effective number of total acid sites (measured by the TPD-NH3 method) in the catalyst is above 250 mol/g, and the ratio between the number of medium acid sites (type 1) and strong acid sites (type 2) (measured by the TPD-NH3 method) is between 1:0.3 and 1:1; and a C.sub.1-C.sub.4 alcohol solvent in a pressurized batch reactor; wherein the saccharide is selected from the group consisting of fructose, a fructose-containing disaccharide, a fructose-containing oligosaccharide, and a fructose-containing polysaccharide or a mixture thereof; and wherein the pressure in the reactor is at least 2 bar.

2. The method according to claim 1, wherein the catalyst has a pore size diameter between 6 and 8 .

3. The method according to claim 1, wherein the ratio between the number of medium acid sites (type 1) and strong acid sites (type 2) (measured by the TPD-NH3 method) is between 1:0.3 and 1:0.95.

4. The method according to claim 1, wherein the ratio between the number of medium acid sites (type 1) and strong acid sites (type 2) (measured by the TPD-NH3 method) is between 1:0.5 and 1:0.95.

5. The method according to claim 1, wherein the saccharide is a monosaccharide, a disaccharide, or a polysaccharide, or a combination thereof.

6. The method according claim 1, wherein the catalyst is selected from the group consisting of HY zeolites and Hbeta zeolites.

7. The method according to claim 1, wherein the catalyst is selected from the group consisting of HY 6, HY 30, Hbeta 12.5 and Hbeta 19.

8. The method according to claim 1, wherein the catalyst is selected from the group consisting of HY 6, Hbeta 12.5 and Hbeta 19.

9. The method according to claim 1, wherein the catalyst is a mesozeolite.

10. A method according to claim 1, wherein the C.sub.1-C.sub.4 alcohol solvent is selected from the group consisting of methanol, ethanol, n-propanol, isopropanol, 1-butanol, 2-butanol and tert-butanol.

11. The method according to claim 1, wherein the saccharide is heated in the solvent in the presence of the catalyst in a pressurized reactor to a suitable temperature at or above 140 C.

12. The method according to claim 11, wherein the saccharide and solvent is heated in the pressurized reactor to or above 160 C.

13. The method according to claim 1, wherein the pressure in the reactor is about 20 bar.

14. The method according to claim 1, wherein the levulinic acid ester is methyl levulinate (MLevu) or ethyl levulinate (ELevu).

15. The method according to claim 1, wherein the levulinic acid ester is methyl levulinate (MLevu).

16. The method according to claim 15, wherein the yield of methyl levulinate (MLevu) is at least 40%.

17. A method for the manufacture of levulinic acid, wherein the levulinic acid ester obtained according to claim 1 is hydrolyzed to obtain levulinic acid, wherein the hydrolysis is acidic, basic or enzymatic.

18. The method according to claim 1, wherein the saccharide is obtained from a biomass or biowaste.

19. The method according to claim 1, wherein the pressure in the reactor is at least 4 bar.

20. The method according to claim 1, wherein the pressure in the reactor is at least 10 bar.

21. The method according to claim 1, wherein the saccharide is selected from the group consisting of fructose, sucrose and inulin, or a mixture thereof.

Description

DRAWINGS

(1) FIG. 1 shows a time-course study on the conversion of glucose to methyl levulinate (each set of data's are from individual experiment).

(2) FIG. 2 shows recyclability of HY6 for glucose to ethyl levulinate.

(3) FIG. 3 shows a NH.sub.3-TPD profile of Y, Beta and ZSM-5 zeolites.

DETAILED DISCLOSURE OF THE INVENTION

(4) The present invention concerns a method for converting a saccharide, such as a mono-, di- or polysaccharide or any combination thereof into a levulinic acid ester (alkyl levulinate).

(5) The present method comprises the steps of providing a saccharide or combination of saccharides in a suitable reactor together with a suitable solvent and a suitable solid acid catalyst selected from zeolites and zeotypes, and heating said saccharide(s) in the presence of the catalyst and the solvent.

(6) The reactor is preferably a pressure batch reactor wherein the reaction mixture is pressurized during the conversion in order to maintain the solvent below the boiling point at the applied temperature.

(7) Boiling Point of Alcohols at Different Pressure with Argon

(8) TABLE-US-00002 Pressure Methanol Ethanol (bar) bp ( C.) bp ( C.) 1 64.7 78.37 2 83 98 4 103 120 10 133 153 20 160 182

(9) Selecting solid acid catalysts suitable for converting saccharides into levulinic acid ester is based on many properties of the catalyst in addition to properties of the selected solvent and the saccharide selected for conversion. Other factors to consider are reaction conditions such as heat, reaction time and reaction vessel. In the examples of the present invention, the reaction mixture in the pressure batch reactor is pressurized to 20 bar with argon. Other pressures may be applied in accordance with the selected temperature.

DEFINITIONS

(10) A carbohydrate is an organic compound that consists only of carbon, hydrogen, and oxygen, usually with a hydrogen:oxygen atom ratio of 2:1 and the empirical formula C.sub.m(H.sub.2O).sub.n. The term is most common in biochemistry, where it is a synonym of saccharide. The carbohydrates (saccharides) are divided into four chemical groupings: monosaccharides, disaccharides, oligosaccharides, and polysaccharides. In general, the monosaccharides and disaccharides, which are smaller (lower molecular weight) carbohydrates, are commonly referred to as sugars. The terms carbohydrate, saccharide and sugar may be used interchangeably throughout the text.

(11) Monosaccharides are the simplest carbohydrates in that they cannot be hydrolyzed to smaller carbohydrates. They are aldehydes (aldoses) or ketones (ketoses) with two or more hydroxyl groups. The general chemical formula of an unmodified monosaccharide is (C.H2O) n, literally a carbon hydrate. Monosaccharides with five carbon atoms (C5) are called pentoses and with six carbon atoms (C6) are hexoses. For example, glucose is an aldohexose (a six-carbon aldehyde), ribose is an aldopentose (a five-carbon aldehyde), and fructose is a ketohexose (a six-carbon ketone).

(12) Two joined monosaccharides are called a disaccharide and these are the simplest polysaccharides. Examples include sucrose and lactose. They are composed of two monosaccharide units bound together by a covalent bond known as a glycosidic linkage formed via a dehydration reaction, resulting in the loss of a hydrogen atom from one monosaccharide and a hydroxyl group from the other. The formula of unmodified disaccharides is C.sub.12H.sub.22O.sub.11. There are numerous kinds of disaccharides. In the present invention, the most important are: sucrose (glucose linked to fructose via a (1.fwdarw.2) bond), maltose (two glucose units linked via a (1.fwdarw.1) bond, a hydrolysis product of the polysaccharide starch) and cellobiose (two glucose units linked via a (1.fwdarw.4) bond, a hydrolysis product of the polysaccharide cellulose).

(13) An oligosaccharide is a saccharide polymer (a polysaccharide) containing a small number (typically 2 to 10) of component monosaccharides. In the present invention oligosaccharides are considered a polysaccharide.

(14) Polysaccharides are composed of long chains of monosaccharide units bound together by glycosidic bonds. Polysaccharides contain more than two monosaccharide units. They range in structure from linear to highly branched. Polysaccharides are often quite heterogeneous, containing slight modifications of the repeating unit. Depending on the structure, these macromolecules can have distinct properties from their monosaccharide building blocks.

(15) Catalyst

(16) As discussed above, many properties of a catalyst define the functionality of the catalyst. In the present invention, focus has been put on two important measurable properties when selecting catalysts for use in the present method. The first property is the pore size. By testing different catalysts, it has become clear, that a large pore size is needed in order to allow access of the substrate and solvent through the channels into the interior part of the catalyst and to allow the resulting levulinate to leave the catalyst after conversion. The second important property is the acidity of the catalyst. Many other properties are however involved in a successful conversion, such as adsorption properties of the pores and cavities inside the catalyst and polarity of the substrate. As also discussed above and in the background literature, it is not possible to predict optimal selection of catalysts or catalyst properties for conversion of carbohydrates. The skilled person would need to try out many different catalysts in the hope of finding suitable solid catalyst(s). Only if suggestions have been made in the art, such as predefined properties of useful catalysts in relation to certain substrates, the skilled person may select catalysts with expected properties suitable for his need. The present invention addresses a way of pre-selecting catalysts with the right properties such as pore size and acidity, measurable features that define useful catalysts for the conversion of mono-, di- or polysaccharides to levulinic acid ester.

(17) The group of materials with a molecule sieves structure possessing catalytic properties useful in the method according to the present invention is the so-called zeolites. In 2011, over 40 naturally occurring zeolite frameworks were known. In addition over 180 synthetic zeolites and zeotypes with different pore sizes and structures have been produced over the time. According to the present invention, only zeolites and zeotypes with large pores (by definition between 6 and 8 ) are efficient for use in the present method of converting carbohydrates (saccharides) into levulinic acid esters. This means that in practice, the pore size diameter should be above 6 . Table 1 summarizes a number of zeolite and zeotype catalysts.

(18) Two commercially available zeolite catalysts with large pores, H-Y (a fajusite (FAU) catalyst) and H-beta (a BEA catalyst) were selected to illustrate the present invention as shown below. This is by no means to be seen as a limitation to these two types of catalysts. Many other zeolite and zeotype catalysts with large pores (see Table 1) may be used as catalyst in method of the present invention.

(19) Fajusite

(20) Fajusite (FAU) zeolites are divided into two types X and Y. Y zeolites can be dehydrated and dealuminated to produce ultra-stable Y (USY) zeolites which are commercially available. Zeolite Y is classified under large pore zeolites since it has pore diameter of 7.4 . Zeolite Y has a 3-dimensional pore structure with pores running perpendicular to each other in the x, y, and z planes, and is made of secondary building units 4, 6, and 6-6. The pore diameter is as large as 7.4 since the aperture is defined by a 12 member oxygen ring, and leads into larger cavities of a diameter of 12 . The cavity is surrounded by ten sodalite cages (truncated octahedra) connected on their hexagonal faces. Commercially available Y (H-Y or HY) catalysts are often named H-VUSY (e.g. H-VUSY-6), H-SDUSY (e.g. H-SDUSY-30) or H-USY, indicating specific preparation before use, for example, steam treatment.

(21) Beta Zeolite

(22) Zeolite Beta (BEA; H-beta) is three-dimensional large-pore zeolites with a 12-membered ring system. Zeolite BEA is the only large-pore zeolite having chiral pore intersections. The pore structure of BEA consists of 12-membered rings interconnected by cages formed by the intersections of the channels. The channel system of zeolite Beta has pore diameters of about 7 , which are similar to other large-pore molecular catalysts such as FAU.

(23) Si/Al Ratio

(24) The adsorption characteristics of zeolites will allow discrimination between competing reactants and products by modifying their relative adsorption interaction, which can be done for example by changing the Si/AI ratios in the catalytic structures. This can be achieved by changing the framework Si/AI ratio by either synthesis or post synthesis treatments. For example, alkaline treatment of zeolites may lead to extensive silicon extraction at mild treatment conditions. This leads to a lower Si/AI ratio, but also impacts on the microporous and acidic properties of the resulting catalyst. As will become clear from the present invention, optimization of the Si/AI ratio is of very high importance for successful conversion of saccharides to levulinic acid ester. It will also become clear, that an optimal window of Si/AI ratios identified for one zeolite family does not necessary apply to other zeolite families. The optimal Si/AI window needs to be established for each zeolite family individually. Extraction of silicon may lead to substantial mesoporosity of the treated catalyst and presumably an improved transport of the reactants.

(25) Acidity

(26) In addition to pore size, the catalysts for use in the present method should have the right acidic properties. One way of securing, controlling and/or changing acidic properties is by selecting a catalyst with the right Si/AI ratio, i.e. Si/AI ratio window, or changing the ratio of other catalysts with less efficient structures. The complexity of the catalyst structures also lead to different acids sites with different strengths in a particular zeolite framework. It has turned out, that the two illustrative zeolites possess medium (type 1) and a strong (type 2) acid sites as measured by the NH.sub.3-TPD method. A measure of the acidic properties of a certain catalyst for use in the present invention is thus the number of acidic sites and the ratio between type 1 and type 2 acid sites.

(27) Adsorption of volatile amines such as NH.sub.3 can be used to determine the number of acidic sites on a solid catalyst. NH.sub.3 is adsorbed on both Brnsted and Lewis sites if both types of acid sites are catalytically active. NH.sub.3 chemically adsorbed on the catalyst is evacuated by thermal desorption (TPD) and the acid strength calculated according to the proportion evacuated at various temperatures. The two acid sited provide for a strong acid site (type 2) and a medium acid site (type 1), the adsorption being stronger at the strongest acid site, thus calling for a higher temperature for desorption. The NH.sub.3-TPD method is a conventional method that is extensively used to measure the acidity of solid catalysts (Chem. Rev., 1995, 95, 559-614).

(28) The number of acidic sites in some of the catalysts tested in a method according to the present invention has been measured according to the NH.sub.3-TPD method (example 11) and the results are shown in Table 2.

(29) TABLE-US-00003 TABLE 2 Influence of number of acid sites and pore type of the zeolites on the yield of methyl levulinate. Acid sites Acid sites Total Acid sites type 1 type 2 acid type 1: Methyl Pore Si/Al (100-270 C.) (270-500 C.) sites type 2 levulinate Catalyst type ratio (mol/g) (mol/g) (mol/g) ratio (%) H-Y Large 2.6 699 252 951 1:0.36 32 pores 6 461 374 835 1:0.81 49 30 182 165 347 1:0.91 37 H-Beta 12.5 563 292 855 1:0.52 44 19 440 366 806 1:0.83 47 150 71 76 147 1:1.07 10 H-ZSM-5 Medium 40 211 240 451 1:1.14 6 pores

(30) In accordance with the results obtained by the NH.sub.3-TPD method, the effective number of acid sites needs to be above 150 mol/g, more particular or above 200 mol/g and even more particular above 300 mol/g. It can also be seen that there appears to be an optimum for the amounts of acid sites around 800 mol/g, above which the efficiency of the catalyst seems to decrease. An amount of 500 or 700 mol/g acid sites may be sufficient to obtain acceptable yields.

(31) From table 2, it can be inferred that HY 6 and Hbeta 19 have almost the same ratio of acid sites type 1 and 2 (1:0.8), which catalysts gave the highest yields of methyl levulinate. For HY 30, the acid sites ratio (1:0.91) is higher than HY6 and Hbeta 12.5 giving lower yield of methyl levulinate (37%). Hbeta 12.5 with ratio of 1:0.52 gave slightly lower yield of methyl levulinate compared to HY6 and Hbeta 19. From these results, it can be understood that acid sites type ratio above 1:0.3, more particular between 1:0.3 and 1:1.2 and more particular between 1:0.3 and 1:1.1 or between 1:0.3 and 1:1 along with total acid sites above 200 could give significant yield of methyl levulinate. The optimum ratio window seems to be 1:0.5 to 1:0.9. Another interesting medium pore zeolite, HZSM-5(40) has been considered for comparison on the yield of methyl levulinate. Even though it has relatively large number of total acid sites and the acid sites ratio is within the range to have a good yield of methyl levulinate, it yielded only 6% of methyl levulinate because of bulky molecular transport limitations in the medium pore channels of ZSM-5.

(32) In an initial experiment, the conversion of glucose to ethyl levulinate (ELevu) and methyl levulinate (MLevu) respectively over different zeolites was carried out in presence of the solvents ethanol or methanol. The results are presented in Table 3, and show that H-Y zeolite with a silicon to aluminum ratio of 2.6 yielded a relatively low amount of ELevu (15%), but interestingly yielded a high amount of MLevu (32%). H-Y zeolite with silicon to aluminum ratio of 6 gave 41% ELevu, and 49% MLevu respectively. H-Y zeolite with silicon to aluminum ratio of 30 gave 26% ELevu and 37% MLevu respectively.

(33) TABLE-US-00004 TABLE 3 Catalytic conversion of glucose to ethyl levulinate or methyl levulinate over zeolites (see Examples 1 and 11 for experimental details; 160 C.; 20 hours) Yield of ELevu Yield of MLevu Catalyst.sup.[a] (Conversion) (%) (Conversion) (%) HY (2.6) 15 (>99) 32 (>99) HY (6) 41 (>99) 49 (>99) HY (30) 26 (>96) 37 (93) HBeta (12.5) 27 (>99) 44 (>99) HBeta (19) 28 (>99) 47 (>99) Hbeta (150) 3 (74) 10 (80) HZSM-5 (40) 6 (81) No Catalyst 0 (45) <0.5 (77) .sup.[a]Numbers in the parenthesis are silicon to aluminum ratio

(34) Two other zeolites, H-Beta, and HZSM-5, were also used for the conversion of glucose and the results are also shown in Table 3. The H-Beta zeolites with ratio of 12.5 and 19 gave 27 and 28% of ELevu and 44 and 47% of MLevu, respectively. Practically all glucose was converted. The H-Beta zeolite with ratio of 150 gave only 3% of ELevu and 10% MLevu, the conversion rate being 74-80% of the glucose after 20 hours at 160 C. The HZSM-5 zeolite with a Si/AI ratio of 40 gave only a yield of 8% of MLevu. HZSM-5 zeolite has a medium pore size (4.5-6 ) which would appear to be the reason for the low yield.

(35) The formation of ethyl levulinate can possibly be formed from glucose via three pathways as shown in scheme 1 (illustrated with ethanol as solvent): The first one, glucose can isomerize to fructose and subsequently reacted with ethanol to form ethyl-D-fructofuranoside (EDFF) (intermediate not confirmed yet) followed by dehydration to form HMF-ether and then rehydration to form ELevu. The second one, glucose can directly react with ethanol to form ethyl-D-glucopyranoside (EDGP) and then isomerise to EDFF and follows the same pathway as mentioned in the first one. Thirdly, after the isomerisation to fructose, it can directly dehydrate to form HMF, HMF-ether and then to ELevu. Methyl levulinate is most likely formed via the same pathways.

(36) The importance of selecting the right catalyst (in combination with the right solvent) for a given carbohydrate substrate is illustrated by the conversion of glucose (Table 3) and fructose (Table 4) over different catalysts (and in the presence of different solvents). One of the reasons for a higher sensitivity in conversion of fructose compared to glucose seems to be that the predominant pathway for fructose to ELevu/MLevu is through the intermediates either HMF or EDFF (and not through EDGP/EDFF as for glucose, mannose and cellobiose), which could be responsible for more degradation or formation of more humins.

(37) TABLE-US-00005 TABLE 4 Catalytic conversion of fructose to methyl levulinate over different zeolites. Conversion Yield of Catalyst.sup.[a] (%) MLevu (%) HY (2.6) 99 32 HY (6) >99 51 HBeta (12.5) 99 44 HBeta (19) >99 48 No Catalyst 70 0.5 [a]Numbers in the parenthesis are silicon to aluminum ratio. (see example 2a-d and 11 for experimental details).
Durability of the Catalyst

(38) It is industrially important to know the stability of a catalyst after repeated cycles. Based on this view point, HY6 was been tested for the conversion of glucose and sucrose to ELevu for five cycles (FIG. 2). After the first run, the yields of ethyl levulinate were 48 and 43% from glucose and sucrose, respectively. First run-through fifth, no significant changes in yields of ethyl levulinate from glucose and sucrose were observed. After the fifth use of HY6 catalyst, negligible changes in pore volumes and surface areas were observed. Likewise, a small decrease in surface area and a small increase in pore volume was observed after the fifth use of HY6 for sucrose, thus confirming the structural integrity of HY6 after five cycles.

(39) Substrate

(40) Carbohydrates such as mono-, di- and polysaccharides are abundant and inexpensive carbonaceous resources available in nature which, however, may be converted into useful chemicals, such as for example levulinic acid or esters thereof in order to be economically interesting. In the present invention conversion of simple sugars such as pentoses and hexoses, in particular glucose and its isomers to levulinate have been exploited. The product of the present method, levulinic acid ester or levulinate, may easily be converted into levulinic acid in a number of commonly and commercially attractive ways, including hydrolysis in the presence of a base, acid or suitable enzyme, e.g. a lipase, an esterase, etc. The substrate, e.g. glucose, fructose, sucrose, maltose, cellobiose, xylose, etc., may be provided after pretreatment of suitable biomass, such as straw, grass, wood, biowaste. Polysaccharides such as cellulose, hemicellulose, starch and inulin may be treated directly in the present method or after a pre-treatment to liberate mono and disaccharide (sugar) units. Such pretreatment is commonly known in the art.

(41) ##STR00001##

(42) Under identical reaction conditions, other mono-, disaccharides related to glucose (Scheme 1) as well as some polysaccharides were also studied, and the yield of ELevu is shown in Tables 5 and 6.

(43) TABLE-US-00006 TABLE 5 Catalytic conversion of mono-, di-, and polysaccharides to ethyl and methyl levulinate over H-Y6 (See Examples 1-8 for experimental details; 160 C.; 20 hours) Yield of ELevu Yield of MLevu Substrate (Conversion) (%) (Conversion) (%) Glucose 41 (>99) 49 (>99) Fructose 40 (>99) 51 (>99) Mannose 44 (>99) 53 (>99) Sorbose 47 (>99) 51 (>99) Sucrose 35 (>99) 49 (>99) Cellobiose 44 (>99) 53 (>99) Maltose 33 (>99) 41 (>99) Xylose.sup.[b] 11 (>99) Inulin 39 50 (>99) Starch 4 31 Cellulose 2 3 (NA) .sup.[b]A pentose (C5 sugar).

(44) TABLE-US-00007 TABLE 6 Catalytic conversion of mono-, di-, and polysaccharides to ethyl and methyl levulinate over H-beta 19 (See Examples 1-6 for experimental details; 160 C.; 20 hours) Yield of ELevu Yield of MLevu Substrate (Conversion) (%) (Conversion) (%) Glucose 28 (>99) 47 (>99) Fructose 48 (>99) 48 (>99) Mannose 25 (>99) 35 (>99) Sucrose 34 (>99) 44 (>99) Cellobiose 28 (>99) 44 (>99) Xylose.sup.[a] 11 (>99) Inulin 34 47 (>99) .sup.[a]A pentose (C5 sugar).

(45) As shown in scheme 1, the predominant pathway for fructose to ELevu is through the intermediates either HMF or EDFF which could be responsible for more degradation or formation of more humins. For glucose, mannose and cellobiose, the predominant pathway to form ELevu is through the formation of EDGP/EDFF. This is apparently supported by the time-course experiment that the formation of significant amount of EDGP from glucose after 10 minutes (FIG. 1).

(46) The polysaccharide inulin, yielded more than 39% (H-Y) and 34% (H-beta) of ethyl levulinate and more than 50% (H-Y) and 47% (H-beta) of methyl levulinate thus implying that under this experimental conditions, inulin was able to dissolve and hydrolyze in ethanol and methanol and eventually, be converted into ethyl levulinate or methyl levulinate respectively. Low yields of ELevu were observed for starch and cellulose even at increased temperatures (Tables 5 and 8), however formation of 31% MLevu could be obtained for conversion of starch in the presence of methanol.

(47) Solvent

(48) Levulinic acid is an important source for producing useful chemicals and fuels. Levulinic acid may be produced as a bio-derived chemical by converting saccharides in the presence of water as previously described. However, as shown in the present application, under identical reaction conditions, the yield of levulinic acid is significantly reduced when water was used as solvent compared to use of an alcohol. The yield of levulinic acid was 11% along with 99% conversion of glucose (Table 7). A lot of other side products were also observed in HPLC including lactic acid, acetic acid and formic acid. Once these acids are formed, they may take part in the degradation of glucose and lead the reaction to form undesired products, thus decreasing the yield of levulinic acid.

(49) The yield of levulinate ester increased substantially to 41% when ethanol was used as solvent and even to 49% when methanol was used as solvent. However, when propanol and butanol were used as solvent the yield of propyl and butyl levulinate was as low as 17 and 12% respectively, which could be due to predominant side reactions, changed polarity or bulkiness. It is therefore preferred to use methanol or ethanol as solvent in the method according to the present invention. A suitable pressure is applied to the reaction mixture in order to keep the boiling point below the applied temperature. An inert gas, such as argon may be used to pressurize the reaction chamber. A preferred pressure is about 20 bar, as shown in the examples.

(50) TABLE-US-00008 TABLE 7 Catalytic conversion of glucose to levulinic acid and its esters over HY6 (See Example 1 for experimental details; 160 C.; 20 hours) Conversion Yield of Levulinic Solvent (%) acid ester (%) Water >99 11 Methanol >99 49 Ethanol >99 41 Propanol >99 17 Butanol >99 12

(51) After conversion of glucose with the use of water as solvent, the HY6 catalyst was subjected to N.sub.2 sorption analysis. It revealed that the surface area was decreased from 657 to 595 m.sup.2/g and pore volume decreased from 0.2295 to 0.2164 cm.sup.3/g, thus implying that a relatively significant loss of structural integrity in the zeolite lattice occurred during the process. These results show the importance of using alcohol, preferably methanol or ethanol, as solvent to obtain levulinic acid esters rather than to produce levulinic acid in water.

(52) Method Conditions

(53) The method in the present invention can be run in any kind of pressure batch reactor that can withstand high pressure that may develop during heating of reactants, including the selected alcohol. A pressure of 20 bar is applied in the examples. A thermocouple and stirrer should be connected in the reactor to monitor reaction temperature and rotation per minute (rpm).

(54) The catalytic experiments could possibly also be performed in suitable fixed bed reactors, loop reactors or in pressure and temperature controlled microwave reactors to decrease the reaction time.

(55) For production of levulinic acid esters in commercial scale, large reaction vessels suited for the disclosed method can be applied by the skilled person. The up-scaling and optimization of the process is very important and may include adjustment of many process parameters known by a process engineer.

(56) Process Temperature

(57) A very important aspect of the present invention is the choice of the optimum temperature suited for the chosen catalyst, and reactants, i.e. saccharide(s) and solvent. In an experiment (Example 9), the conversion of the monosaccharide glucose, was tested in the presence of H-Y 6 and methanol at different temperatures, in order to optimize the process conditions. The results are shown in Table 8. The optimum temperature for the tested combination would appear to be above 120 C., with good result at 120 C., an even better result at 140 C. At 160 C. the best results are seen with the tested catalyst, monosaccharide and solvent. Raising the temperature to 180 C. seems to decrease the output of MLevu, thus indicating an optimum temperature around 160 C. The conversion was run for 20 hours and the pressure being 20 bar (290 psi). Dependent on the complexity of the saccharide (di- and polysaccharides), it may be beneficial to the conversion rate and yield to increase the temperature to above 160 C., for example to 180 C. or 200 C. Alternatively, di- and polysaccharides may be hydrolyzed to obtain suitable mono- and/or disaccharides for use in the present invention.

(58) TABLE-US-00009 TABLE 8 Catalytic conversion of substrates to methyl levulinate over HY6 (see Example 10 for experimental details; 20 hours, 20 bar) Temperature Conversion Yield of ( C.) Substrate (%) MLevu (%) 100 Glucose 66 13 120 Glucose 90 33 140 Glucose 98 47 160 Glucose >99 49 180 Glucose >99 39 160 Cellulose 3 180 Cellulose 9 200 Cellulose 13

(59) TABLE-US-00010 TABLE 10 Results from pressure batch reactor (autoclave) experiments at different pressure of argon. Pressure Conversion of Yield of (bar) Fructose (%) Elevulinate (%) 20 99 15 No pressure* >99 7 Comparison with S. Saravanamurugan, A. Riisager, in Catal. Commun. 2012, 17, 71-75. Experimental conditions: Fructose = 0.5400 g; Catalyst weight = 0.160 g (H-beta 12.5); Ethanol: 8.0 g; Temperature = 140 C.; Time = 20 h; Pressurized with Ar; *not pressurized with Ar
Process Time

(60) In an experiment (Example 8), the conversion of the monosaccharide glucose, was tested in the presence of HY-6 and methanol at different process times, in order to further optimize the process conditions. The results are shown in Table 9. The optimum process time for the tested combination would appear to be above 10 minutes, with good results after 10 minutes, even better results after 30 minutes, 1 hour, 3 hours and the best result after 20 hours. The results also show that for the tested combination, in order to have a full conversion of the glucose, a process time of above 3 hours is needed. Only the results obtained after 20 hours show more than 99% conversion. Process time should be selected in combination with the chosen process temperature in order to optimize the process.

(61) TABLE-US-00011 TABLE 9 Catalytic conversion of glucose to methyl levulinate over HY-6 (see Example 9 for experimental details; 160 C.) Conversion Yield of Time (%) MLevu (%) 10 min 86 23 30 min 88 27 1 hour 90 31 3 hours 95 40 20 hours >99 49

(62) Levulinic acid esters may be converted into levulinic acid by any known way of hydrolyzing a ester, including the use of a base, an acid or any suitable enzyme, such as an esterase a lipase, etc.

EXAMPLES

Example 1a-l

(63) A 50 ml autoclave (Microclave reactor from Autoclave Engineers) was charged with 250 mg glucose, 150 mg of one of the catalysts: HY 2.6, HY 6, HY 30, Hbeta 12.5, Hbeta 19, Hbeta 150, and HZSM-5 (purchased from Zeolyst International, USA) and 10 mL solvent (methanol, ethanol, propanol or butanol) and then pressurized with argon (20 bar). The autoclave was heated to 160 C. and the stirring was started at 300 rpm once the temperature reached 150 C. After 20 hours of stirring, the autoclave was quenched with cold water and 50 mg naphthalene (internal standard) added to the reaction mixture and the content thereafter analyzed. Aliquots of the reaction mixtures were subjected to GC-FID analysis (Agilent 6890N instrument, HP-5 capillary column 30.0 m320 m0.25 m). The reaction mixtures were also analyzed by HPLC with RI detection (Agilent 1200 series, 30 cm Aminex HPX-87H column, 0.005 M aqueous sulfuric acid solution as eluent at a flow rate of 0.6 ml/min). An Agilent 6850 GC system coupled with an Agilent 5975C mass detector was used for qualitative analysis. Sugar conversions to levulinic acid esters were determined by HPLC. The amounts of unreacted monosaccharides were calculated from their individual HPLC standards. Sucrose inverts to fructose and glucose on the acidic HPLC-column used, and the conversions were calculated from the monosaccharides fructose, glucose and mannose on a carbon-basis. The yield of levulinate esters were calculated from GC results on series of individual levulinate ester standards with naphthalene as internal standard. The results are shown in Tables 3 and 7.

Example 2a-d

(64) A 50 ml autoclave (Microclave reactor from Autoclave Engineers) was charged with 250 mg of fructose, 150 mg of a catalyst (HY 6, Hbeta 19), and 10 ml solvent (methanol, ethanol) and then pressurized with argon (20 bar). The autoclave was heated to 160 C. and the stirring was started once the temperature reached 150 C. (300 rpm). After 20 hours of stirring, the autoclave was quenched with cold water, 50 mg naphthalene (internal standard) added and the reaction mixture analyzed as described in examples 1. The results are shown in Tables 4, 5 and 6.

Example 3a-d

(65) A 50 ml autoclave (Microclave reactor from Autoclave Engineers) was charged with 250 mg of mannose, 150 mg of a catalyst (HY 6, Hbeta 19), and 10 ml solvent (methanol, ethanol) and then pressurized with argon (20 bar). The autoclave was heated to 160 C. and the stirring was started once the temperature reached 150 C. (300 rpm). After 20 hours of stirring, the autoclave was quenched with cold water, 50 mg naphthalene (internal standard) added and the reaction mixture analyzed as described in examples 1. The results are shown in the Tables 5 and 6.

Example 4a-d

(66) A 50 ml autoclave (Microclave reactor from Autoclave Engineers) was charged with 237.5 mg of sucrose, 150 mg of a catalyst (HY 6, Hbeta 19), and 10 ml solvent (methanol, ethanol) and then pressurized with argon (20 bar). The autoclave was heated to 160 C. and the stirring was started once the temperature reached 150 C. (300 rpm). After 20 hours of stirring, the autoclave was quenched with cold water, 50 mg naphthalene (internal standard) added and the reaction mixture analyzed as described in examples 1. The results are shown in Tables 5 and 6.

Example 5a-d

(67) A 50 ml autoclave (Microclave reactor from Autoclave Engineers) was charged with 237.5 mg of cellobiose, 150 mg of a catalyst (HY 6, Hbeta 19), and 10 ml solvent (methanol, ethanol) and then pressurized with argon (20 bar). The autoclave was heated to 160 C. and the stirring was started once the temperature reached 150 C. (300 rpm). After 20 hours of stirring, the autoclave was quenched with cold water, 50 mg naphthalene (internal standard) added and the reaction mixture analyzed as described in examples 1. The results are shown in Tables 5 and 6.

Example 6a-d

(68) A 50 ml autoclave (Microclave reactor from Autoclave Engineers) was charged with 225 mg inulin, 150 mg of a catalyst (HY 6, Hbeta 19), and 10 ml solvent (methanol, ethanol) and then pressurized with argon (20 bar). The autoclave was heated to 160 C. and the stirring was started once the temperature reached 150 C. (300 rpm). After 20 hours of stirring, the autoclave was quenched with cold water, 50 mg naphthalene (internal standard) added and the reaction mixture analyzed as described in examples 1. The results are shown in Table 5 and 6.

Example 7a-c

(69) A 50 ml autoclave (Microclave reactor from Autoclave Engineers) was charged with 225 mg cellulose, 150 mg of the catalyst HY 6, and 10 ml methanol and then pressurized with argon (20 bar). The autoclave was heated to 160, 180 and 200 C. respectively and the stirring was started once the temperature reached 10 to 20 C. below the set temperature (300 rpm). After 20 hours of stirring, the autoclave was quenched with cold water, 50 mg naphthalene (internal standard) added and the reaction mixture analyzed as described in examples 1. The results are shown in Table 5.

Example 8a-e

(70) A 50 ml autoclave (Microclave reactor from Autoclave Engineers) was charged with 250 mg glucose, 150 mg of the catalyst HY 6, and 10 mL methanol and then pressurized with argon (20 bar). The autoclave was heated to 160 C. and the stirring was started once the temperature reached 150 C. (300 rpm). After 10 minutes, 30 minutes, 1 hour, 3 hours and 20 hours of stirring, the autoclave was quenched with cold water, 50 mg naphthalene (internal standard) added and the reaction mixture analyzed as described in examples 1. The results are shown in Table 9.

Example 9a-e

(71) A 50 ml autoclave (Microclave reactor from Autoclave Engineers) was charged with 250 mg glucose, 150 mg of the catalyst HY 6, and 10 mL methanol and then pressurized with argon (20 bar). The autoclave was heated to 100, 120, 140, 160 and 180 C., respectively and the stirring was started once the temperature reached 10 to 20 C. below the set temperature (300 rpm). After 20 hours of stirring, the autoclave was quenched with cold water, 50 mg naphthalene (internal standard) added and the reaction mixture analyzed as described in examples 1. The results are shown in Table 8.

Example 10a-c

Controls without a Catalyst

(72) A 50 ml autoclave (Microclave reactor from Autoclave Engineers) was charged with 250 mg substrate (glucose, fructose, sucrose), and 10 mL solvent (methanol, ethanol) and then pressurized with argon (20 bar). The autoclave was heated to 160 C. and the stirring was started once the temperature reached 150 C. (300 rpm). After 20 hours of stirring, the autoclave was quenched with cold water, 50 mg naphthalene (internal standard) and the reaction mixture analyzed as described in examples 1. The results are shown in Tables 3 and 4.

Example 11

NH3-TPD Measurement

(73) The number of acid sites present in the zeolites was measured by using a AutoChem II 2920 apparatus from Micromeritics. 100 mg of the sample was placed in a quartz reactor and degassed at 500 C. for 1 hr in a flow of helium at the rate of 50 ml/min. The reactor was then cooled to 100 C. and ammonia (50 ml/min) was allowed to get adsorbed at the same temperature for 2 hr. Before the ammonia desorption measurement, the sample was flushed with helium at the rate of 50 ml/min to remove the physisorbed ammonia. Ammonia desorption was measured every one second from 100 to 500 C. at a ramp of 10 C./min. The number of acid sites is calculated as the area under the curve. The results are shown in Table 2 and in FIG. 3. From the Figure, it can be apparently understood that zeolites contain both medium (approximately between 100 and 270 C.) and strong acid sites (approximately between 270 and 500 C.). The intensity of the peak for HY 2.6 is high at low temperature range (medium acid sites) and low at high temperature range (strong acid sites) compared to HY 6 and Hbeta 19. The yield of methyl levulinate is relatively high for HY6 and Hbeta 19 compared to HY 2.6. The ratio of the number of medium acidic sites (type 1) to strong acidic sites (type 2) can be taken as a measure defining the relative efficiency of the catalysts for the conversion of sugars. A ratio window between 1:03 and 1:1 seem to be a preferred range for the tested catalysts.