Biomacromolecules 2004, 5, 1799-1803
1799
Facilitated Synthesis of Inulin Esters by Transesterification T. M. Rogge and C. V. Stevens* Department of Organic Chemistry, Faculty of Agricultural and Applied Biological Sciences, Ghent University, Coupure links 653, B-9000 Ghent, Belgium Received March 3, 2004; Revised Manuscript Received May 24, 2004
Inulin, the polydisperse polyfructose, extracted from chicory, has been modified via transesterification, using fatty acid methyl esters (FAME). The grafting of an alkyl chain onto the inulin backbone under different conditions for the development of potential tensio-active derivatives is described. The modification of the biopolymer was performed in polar organic solvents, such as dimethyl sulfoxide (DMSO) and Nmethylpyrrolidinone (NMP). Depending on the type of solvent, different catalytic systems, such as DMSO-Na+, NaH, and NaOMe, were used and compared in reaction efficiency and reproducibility. Therefore the synthesized derivatives were characterized by 1H- and 13C NMR. The methods using NaH had a mean reaction efficiency of 80%, whereas the one using NaOMe showed a slight decrease in reaction efficiency to 75%. However, the method using NaOMe in NMP proved to be the preferred way to graft the inulin backbone with FAME on a bigger scale. The methods using DMSO as a solvent were not attractive since the end products had a specific bad smell. Introduction The synthesis and the use of low as well as high molecular weight carbohydrate esters have been studied extensively.1 Low molecular weight carbohydrate esters, for example, sucrose esters, are used as nonionic surfactants, as binders in paints,2 or as plasticizers or softeners, whereas the high molecular weight carbohydrate esters, such as starch esters, are used as sizing agents in the textile industry, as thickeners in the film and fiber industry,3 and as substitutes for purely synthetic grafted polymeric surfactants. Polymeric surfactants also attract considerable attention as dispersants for solids in liquid media and as emulsifiers.4 Most of the copolymeric surfactants consist of blocks (A-B and A-B-A) or graft structures (BAn) whereby the B chain represents the anchor of the molecule on which the A chains are randomly grafted. In recent years, considerable attention was given to surfactants that are based on polysaccharides via the grafting of alkyl chains.5 In this respect the alkyl groups represent the A chains that are randomly distributed on the polysaccharide backbone and are strongly adsorbed on hydrophobic surfaces, such as carbon black or an oil droplet. The polysaccharide chain acts as the stabilizing part as it is water-soluble. The type of polysaccharide has a considerable influence on the performance of the end-product and the difference in properties and applicability between modified starch, cellulose, inulin, or other polysaccharides is therefore enormous. Inulin, the reserve polysaccharide of chicory (Cichorium intybus), is gaining more and more interest for chemical modification due to its interesting properties.6 It consists mainly of β(2-1) fructosyl fructose units (Fm) with normally, * Corresponding author. E-mail:
[email protected]. Telephone: +32 (0)9 264 59 57. Fax: +32 (0)9 264 62 43.
Figure 1. Structure of native inulin.
but not always, a glucopyranose at the reducing end (GFn). See the structure in Figure 1. The fructose units in the polysaccharide are all present in the furanose form, except when the reducing end consists of fructose (as in Fm). Then, this terminal fructose is present in the pyranose form. Materials and Methods Materials. Inulin (INUTECN25) was supplied by Orafti (Tienen, Belgium) and was used without purification. The mean degree of polymerization was approximately 25 and was determined by HPLC analysis after enzymatic hydrolysis.7 NMP, acetone, and triethylamine were obtained from ACROS or Sigma-Aldrich Belgium and were used as received. DMSO and dichloromethane were also obtained from ACROS Belgium and were dried by distillation over calcium hydride. Sodium hydride was used after removal of
10.1021/bm049869q CCC: $27.50 © 2004 American Chemical Society Published on Web 08/17/2004
1800
Biomacromolecules, Vol. 5, No. 5, 2004
the mineral oil with petroleum ether; sodium methoxide (NaOMe) was used in its pure form, immediately after preparation by putting Na in dry MeOH. Synthesis. General Synthesis of Inulin Esters in DMSO (Dimethyl Sulfoxide) with NaH as Catalyst. In a roundbottom flask of 100 mL under nitrogen atmosphere at 60 °C, 0.2 g of NaH (0.22 equiv based on fructose equivalents) was added to 9 g of dried DMSO. After 2-3 h, when the reaction mixture became clear, 6.0 g of inulin (37.0 mmol of fructose equivalents) was added and the temperature was raised to 85 °C. After stirring until the inulin was completely dissolved, 1.59 g (0.2 equiv) of methyl laurate was added, to obtain a DS of 0.16. After heating the mixture at 85 °C for 24 h, the reaction mixture was poured into 200 mL dry dichloromethane or acetone under vigorous stirring. The inulin derivative crystallized immediately and was filtered over a sintered glass filter. The modified inulin was washed with dichloromethane and acetone to remove the DMSO and some remaining methyl laurate. The resulting powder was dried under high vacuum (crystallization yield: 84%). General Synthesis of Inulin Esters in N-Methylpyrrolidinone (NMP) with NaH as Catalyst. 6.0 g of inulin (37.0 mmol of fructose equivalents) was dissolved in 20 g of N-methylpyrrolidinone (NMP) at 50 °C. The solution of inulin in NMP was evaporated under high vacuum up to a 40% (w/w) solution to remove moisture and crystal water that is present in the inulin-NMP system. After stripping of the water, the inulin solution was transferred to a roundbottom flask of 100 mL under a nitrogen atmosphere and heated at 60 °C. To this solution was then added 0.2 g of NaH (0.22 equiv based on fructose equivalents). The temperature was raised to 85 °C. After stirring until the reaction mixture became homogeneous, 1.59 g (0.2 equiv) of methyl laurate was added, to obtain a DS of 0.16. After heating for 24 h, the reaction mixture was poured into 200 mL dry dichloromethane under vigorous stirring. The inulin derivative crystallized immediately and was filtered over a sintered glass filter. The modified inulin was washed with dichloromethane and acetone to remove the NMP and some remaining methyl laurate. The resulting powder was dried under high vacuum (crystallization yield: 89%). General Synthesis of Inulin Esters in N-Methylpyrrolidinone (NMP) with NaOMe as Catalyst. 6.0 g of inulin (37.0 mmol of fructose equivalents) was dissolved in 20 g of N-methylpyrrolidinone (NMP) at 50 °C. The solution of inulin in NMP was evaporated under high vacuum up to a 40% (w/w) solution in order to remove moisture and crystal water that is present in the inulin-NMP system. After stripping of the water, the inulin solution was transferred to a round-bottom flask of 100 mL under a nitrogen atmosphere and heated to 60 °C. To this solution was added 0.44 g of NaOMe (0.22 equiv based on fructose equivalents) and the temperature was raised to 85 °C. After stirring until the reaction mixture became homogeneous, 1.59 (0.2 equiv) of methyl laurate was added, to obtain a DS of 0.15. After a 24-h reaction time, the reaction mixture was poured into 200 mL of dry dichloromethane or acetone under vigorous stirring. The inulin derivative crystallized im-
Rogge and Stevens
mediately and was filtered over a sintered glass filter. The modified inulin was washed with dichloromethane and acetone to remove the NMP and some remaining ester. The resulting powder was dried under high vacuum (crystallization yield: 71%). NMR Spectroscopy. All spectra were recorded on a JEOL NMR spectrometer under a static magnetic field of 300 MHz for 1H NMR measurements and of 75.57 MHz for 13C NMR. The spectra were all recorded in DMSO-d6 at room temperature for 1H NMR and at 50°C for 13C NMR. Results and Discussion In a previous study on the chemical modification of inulin for the development of inulin-based surfactants, very interesting properties were found for some inulin carbamates.8 To obtain comparable tensio-active properties, the esterification of this polysaccharide was studied in detail. The most classical way to esterify an alcohol is through condensation with an acid chloride. This method has been described in dimethyl formamide and pyridine for dextrin9 and in water for inulin.10 When performing the esterification on inulin in aqueous medium, with NaOH as catalyst, the reaction is very hard to reproduce and the purification of the crude reaction mixture is very difficult, since a lot of side products have to be removed. The esterification of polysaccharides by alkylsuccinic anhydrides with sodium hydroxide was patented for the first time in 1953 and was initially described on starch.11,12 Since then, several improvements with slight variations in reaction conditions and purification methods have been extensively described and some new applications have been patented.13-17 Recently, a study was performed on the modification of several other polysaccharides, including inulin, using this esterification method in water.18 The disadvantage of water as a solvent, however, is the formation of several sideproducts and the very difficult purification procedure. Similarly, Su¨dzucker described the esterification of inulin with the C4-C7 anhydrides in pyridine (Figure 2) and reported the properties of the derivatives. The speed of the reaction could be increased using alkaline catalysts such as 4-(dimethylamino)pyridine, potassium carbonate, or an ionexchange resin in its acid or in its alkaline form.10 Again, the major disadvantage of this reaction is the formation of an equivalent amount of free fatty acid that has to be removed. Further, the use of pyridine is undesirable for an industrial process because of its toxicity profile. Apart from the chemical esterification, it is also possible to perform the esterification enzymatically, using lipases, with a fatty acid ester such as vinyl or ethyl laurate as alkylating agents in a mixture of organic solvents such as dimethyl sulfoxide (DMSO) and 2-methyl-2-butanol.19 A major drawback for enzymatic esterification is the fact that it needs to be performed under very diluted conditions, which is undesirable for industrial application. Furthermore, some organic solvents that are suitable to dissolve polysaccharides (e.g., DMF) will denaturate the expensive enzyme; therefore, most reactions are still performed in diluted aqueous media.20 The esterification of inulin using fatty acid methyl esters (FAME) in dimethylacetamide (DMA) (Figure 3) was
Facilitated Synthesis of Inulin Esters
Biomacromolecules, Vol. 5, No. 5, 2004 1801
Figure 2. Modification of inulin using the acid anhydrides.
Figure 3. Esterification of inulin using fatty acid methyl esters (FAME).
patented recently,21 but the described reaction conditions were very stringent, with reactions at 160 °C for 6 h. The described reactions were repeated in our lab, but could not be reproduced. Therefore, we studied the transesterification in depth to develop a more elegant and less stringent method, which ensures a very good reproducibility in a wide range of organic solvents. However, due to the limited solubility of inulin in organic solvents, the choice is reduced to pyridine, dimethylformamide (DMF), dimethyl sulfoxide (DMSO), N-methylpyrrolidinone (NMP), and dimethylacetamide (DMA). In combination with the solvents, several catalysts can be evaluated, such as Et3N, NaH, NaOMe, etc. When performing the reaction in NMP, Et3N, NaOMe, and NaH are compatible as catalyst. However, after only a few experiments it became obvious that Et3N, which has been used to catalyze the etherification of inulin with ethylene oxide and propylene oxide,22 is too weak to obtain an efficient transesterification. The two stronger bases, NaH and NaOMe, were both suitable for the catalysis of the reaction in NMP. However, special attention had to be paid to exclude traces of water in the reaction mixture, to prevent decomposition of the catalyst. Therefore, these reactions were all performed under a nitrogen atmosphere in combination with extensively dried solvents.
Since inulin contains several percent crystal water (35%), which can be partially removed by drying the polyfructose in a vacuum oven at 70 °C, a pretreatment of the solution of inulin in NMP needs to be performed. This consisted of a high vacuum distillation of NMP, with removal of water. The starting concentration of the inulin/NMP solution was approximately 20%, and was raised to 40% by evaporation under high vacuum (pressure less than 0.5 mbar) at 70 °C. Karl Fisher analysis proved that the water content of the starting solution was reduced to less than 0.5% after the stripping operation. The advantage of the use of NaOMe in comparison to NaH is that this catalyst is less reactive toward the humidity in the atmosphere and is therefore easier to handle on a bigger scale. The series of synthesized products was built up by varying the degree of substitution on one hand (from 0.1 to 0.3) and the length of the alkyl chain (from C6 to C18) on the other hand. A disadvantage concerning the application of these inulin esters for certain use (e.g., in cosmetic applications) was the brown color of the reaction mixture and the light brown caramel color of the crystallized powder. In both cases, the presence of traces of water in the reaction mixture considerably diminished the reaction efficiency, which is defined as the percentage of the fatty acid ester that reacted onto the inulin backbone and which could
1802
Biomacromolecules, Vol. 5, No. 5, 2004
Rogge and Stevens
Table 1. Yieldsa of the Inulin Esters by Transesterification via the NaOMe/NMP Method R/aimed DS C6 C8 C12 C14 C18
0.1
0.2
0.3
65% (DS ) 0.07)b 72% (DS ) 0.08) 65% (DS ) 0.08) 67% (DS ) 0.07) 92% (DS ) 0.07)
67% (DS ) 0.16) 66% (DS ) 0.16) 71% (DS ) 0.15) 72% (DS ) 0.18) 81% (DS ) 0.13)
68% (DS ) 0.24) 71% (DS ) 0.21) 66% (DS ) 0.23) 83% (DS ) 0.26) 78% (DS ) 0.24)
a The reported yields are based on the amount of product recovered after crystallization in a nonsolvent. b The reported DS is the DS calculated by NMR.
Table 2. Yieldsa of the Inulin Esters by Transesterification via the NaH/NMP Method R/aimed DS C12 C14 C18
0.1
0.2
0.3
94% (DS ) 0.1)b 93% (DS ) 0.07) 92% (DS ) 0.08)
84% (DS ) 0.15) 97% (DS ) 0.15) 90% (DS ) 0.20)
89% (DS ) 0.25) 91% (DS ) 0.20) 88% (DS ) 0.28)
a The reported yields are based on the amount of product recovered after crystallization in a nonsolvent. b The reported DS is the DS calculated by NMR.
range from 90% to 50%. Therefore, good reproducibility of the reaction is not evident, and fluctuations can form a tremendous problem during the scale-up of the reaction. When performing the distillation of the NMP very accurately, immediately followed by the addition of the catalyst under a nitrogen atmosphere, the reproducibility of the reaction is good and a mean reaction efficiency of 75% using NaOMe can be reached without any problem (Table 1). Using NaH as base in NMP, a slight improvement of the reaction efficiency (up to 80%) was obtained (Table 2). When DMSO was used as solvent in combination with NaH as base for the reaction, the actual catalyst is the sodium salt of DMSO, obtained after heating for 2-3 h at 60 °C. After the formation of the dimsylsodium, the inulin was dissolved in the DMSO/DMSO-Na+ system. Consequently, the methyl ester of the fatty acid was added under a nitrogen atmosphere, and the reaction mixture was stirred for 24 h at 85 °C. The reaction efficiency was low (50%) when the temperature was less then 70 °C. With the increase of the reaction temperature and the extra drying of DMSO over CaH2, the mean efficiency was mostly about 70% due to the presence of the crystal water in inulin. Using vacuumdried inulin the reaction efficiency could be increased up to 80% (Table 3), which was very reproducible, but the drying of the inulin is too expensive on an industrial scale in comparison to the minor losses of the cheap reagents. A variety of products was synthesized, but all the synthesized derivatives had a very specific bad smell, even after purification via crystallization in a nonsolvent. Therefore, this method seems to be less useful for the synthesis of inulin-based surfactants for cosmetic formulations. The isolation of the inulin esters was performed by pouring the
Table 3. Yieldsa of the Inulin Esters by Transesterification via the NaH/DMSO Method R/aimed DS C8 C12 C14 C18
0.1
0.2
0.3
96% (DS ) 0.1)b 94% (DS ) 0.08) 93% (DS ) 0.07) 92% (DS ) 0.06)
95% (DS ) 0.17) 84% (DS ) 0.16) 97% (DS ) 0.12) 90% (DS ) 0.11)
82% (DS ) 0.24) 89% (DS ) 0.18) 91% (DS ) 0.26) 88% (DS ) 0.18)
a The reported yields are based on the amount of product recovered after crystallization in a nonsolvent. b The reported DS is the DS calculated by NMR.
warm reaction mixture in a 10-fold excess of nonsolvent (mostly dichloromethane, diethyl ether, or acetone) under vigorous stirring. The modified inulin was crystallized slowly and the crystals were isolated by filtration. In a second step, the modified inulin was added again to an excess of dichloromethane in order to remove trace amounts of fatty acid ester. In the final step, the modified inulin was filtered again and dried. This method was used for the purification of all the described derivatives of inulin. The advantage of performing the reaction in organic solvents is the convenient purification of the derivatives in a nonsolvent. The straightforward crystallization in a nonsolvent is sufficient to purify the compounds to less than 1% of impurity (solvent, fatty acid ester). More expensive and elaborate methods such as membrane filtration or dialysis can therefore be avoided. After isolation of the products, the degree of substitution could be measured by NMR-spectroscopy at 300 MHz, comparing the integration of the signal for the three protons of the CH3 group of the alkyl chain (at approximately 0.8 ppm) to a specific signal of the fructose unit integrating for 1 proton (at approximately 4.0 ppm) or for the 3 OH groups at 4.5-5.5 ppm. Using these integrations, the reaction efficiency was calculated. The accuracy of the integration signals is not quantitative due to the broadness of the peaks of the inulin. But, in a previous study, where a quantitative reaction was obtained using alkyl isocyanates, it was proven that the fault was less than 10% when the DS was determined via the NMR-measurements. This was confirmed with an HPLC analysis after acid hydrolysis of the inulin backbone of the carbamate derivatives.8 For a representative lauryl inulin ester, δ (ppm, 300 MHz, DMSO-d6): 0.8 (3H, br.t., CH3); 1.0-1.2 (16H, br.s., (CH2)8); 1.3-1.5 (2H, br.s., CH2CH2CO); 2.2-2.4 (2H, br. t., CH2CO) 3.3-3.6 (5H, br.m., CH2 at pos. 6 and 1 and CH at pos. 5), 3.7-3.8 (1H, br.s., CH at pos. 4); 3.9-4.0 (1H, br.s., CH at pos. 3); 4.5-5.5 (3H, m, 3OH). Only the integrations of the peaks change with a changing DS of the sample and an increase or a decrease in the length of the alkyl chain. The esterification of the inulin backbone was also confirmed using 13C NMR spectroscopy. For a representative myristyl inulin ester, δ (ppm, 75 MHz, DMSO-d6): 14.4 (CH3); 22.7 (CH2); 25.0 (CH2); 29-31 ((CH2)8); 31.9 (CH2); 33.9 (CH2); 62.1-62.3 (CH2 at pos. 1 and at pos. 6); 74.474.8 (CH at pos. 4); 77.3-77.7 (CH at pos. 3); 82.2 (CH at pos. 5); 103.7 (Cquat at pos. 2); 173.3 (CdO). (Also see Figure 4.) From the spectroscopic data, however, no conclusions
Facilitated Synthesis of Inulin Esters
Biomacromolecules, Vol. 5, No. 5, 2004 1803
The careful study of the transesterification of inulin here described has resulted in the development of an industrially useful synthetic method for the preparation of inulin esters. References and Notes
Figure 4. 1H NMR (bold) and derivatives.
13C
NMR data for esterified inulin
could be drawn regarding the regioselectivity toward a certain hydroxyl group of the inulin. Conclusion The esterification of inulin, which has been described before in the literature, poses a lot of problems regarding the isolation and stability of the pure derivatives. All previously described reactions had very poor reproducibility and none of them could prevent the formation of considerable amounts of side products, which were difficult to remove afterward. In this manuscript the transesterification of inulin, using fatty acid methyl esters, has been studied in different organic solvents, using different bases. For the evaluation of the different types of solvent, it can be concluded that DMSO is less favorable, since the reaction mixture with NaH as catalyst had a very bad smell, as did the obtained powder. The smell makes the use of the obtained compounds, without further purification methods, impossible in, for example, cosmetic applications. For other industrial use, however, this method can be considered very useful because of the good reproducibility and the easy purification. The advantageous purification is valid for all methods described and therefore, the two methods in NMP can both be considered to be very useful. NaOMe is the preferred catalyst on an industrial scale, because of the safety advantage in comparison to NaH and its water tolerance, although NaH shows slightly better reaction efficiencies. Although all reactions showed a colored end product, this coloration will be negligible upon dissolution and use in formulations.
(1) Douglas, A. K.; Bernard, Y. T. J. Surfactants Deterg. 1999, 2 (3), 383-390. (2) Oostveen, E. A.; Weijnen, J.; Van Haveren, J.; Gillard, M. PCT WO 03064498, 2003; Chem. Abstr. 2003, 139, 165937. (3) Bognolo, G. In Lipid Technologies and Applications; Gunstone, F. D., Padley, F. B., Eds.; Marcel Dekker: New York, 1997, p 633. (4) Tadros, T. F. Polymeric Surfactants: Stabilization of Emulsions and Dispersions. In Principles of Polymer Science and Technology in Cosmetics and Personal Care; Goddard, E. D.; Gruber, J. V., Eds.; Marcel Dekker: New York, 1999. (5) Menger, F. M.; Mbadugha, B. N. A. J. Am. Chem. Soc. 2001, 123, 875-885. (6) Stevens, C. V.; Meriggi, A.; Booten K. Biomacromolecules 2001, 2, 1-16. (7) Hubrechts. AOACS methods nr 997.08. J. AOAC Int. 1997, 80, 1029. (8) Stevens, C. V.; Meriggi, A.; Peristeropoulou, M.; Christov, P. P.; Booten, K.; Levecke, B.; Vandamme, A.; Pittevils N.; Tadros, T. F. Biomacromolecules 2001, 2, 1256-1259. (9) Suzuki T.; Amano, I.; Chiba, K.; Tofukuji, R. Eur. Patent 0736545, 1996; Chem. Abstr. 1996, 125, 332125. (10) Erhardt, S.; Haji Begli, A.; Kunz, M.; Sheiwe, L. US Patent 5877144, 1999; Chem. Abstr. 1997, 127, 249640. (11) Wurzburg O. B.; Caldwell, C. G. US Patent 2661349, 1953; Chem. Abstr. 1954, 48, 9428. (12) Wurzburg O. B.; Caldwell, C. G. US Patent 2654836, 1953; Chem. Abstr. 1954, 48, 5626. (13) Richards, C. N.; Bauer, D. US Patent 4035235, 1977; Chem. Abstr. 1977, 87, 137597. (14) Billmers R. L.; Mackewicz, V. L. Eur Patent 0761691, 1997; Chem. Abstr. 1997, 126, 279257. (15) Maliczyszyn, W.; Atkinson, J. G.; Tolchinsky, M. US Patent 6037466; 2000; Chem. Abstr. 2000, 132, 209385. (16) Nakajima T. US Patent 5580553, 1996; Chem. Abstr. 1996, 126, 148212. (17) Jureller S. H.; Trzasko P. T.; Harris R.; Humphreys R. W. R.; Kerschner J. L. US Patent 5977348; 1999; Chem. Abstr. 2000, 132, 209385. (18) Ward, F. M. WO PCT 02069981, 2002; Chem. Abstr. 2002, 137, 237723. (19) Ferrer, M.; Cruces, M. A.; Bernabe, M.; Ballesteros, A.; Plou, F. J. Biotechnol. Bioeng. 1999, 65, 10-16. (20) Cruces, M. A.; Plou, F. J.; Ferrer, M.; Bernabe, M.; Ballesteros, A. J. Am. Oil Chem. Soc. 2001, 78, 541-546. (21) Oostveen, E. A.; Weijnen, J.; Van Haveren, J.; Gillard, M. PCT WO 03064498; Chem. Abstr. 2003, 139, 165937. (22) Rogge, T. M.; Stevens, C. V.; Booten, K.; Levecke, B.; Vandamme, A.; Vercauteren, C.; Haelterman, B.; Corthouts, J.; D′hooge, C. Top. Catal. 2004, 27, 39-47.
BM049869Q
