Solvent Effects in the Catalytic Conversion of Fructose to Lactic Acid Christian G. Rivera-Goyco, Nitza Padilla Fuentes, Oscar Oyola-Rivera, Yomaira J. Pagan- Torres and Nelson Cardona-Martínez* Chemical Engineering Dep., University of Puerto Rico-Mayagüez, Mayagüez, PR 00681 *nelson.cardona@upr.edu Introduction Lactic acid is an industrial bulk chemical with production higher than 300,000 metric tons per year in 2010 [1]. Taarning and coworkers [2] found that Sn-BEA zeolite displays high activity and selectivity for the conversion of mono- and disaccharides to lactic acid derivatives in methanol. Using Sn-BEA zeolite (Si/Sn ratio of 125) they discovered that fructose in methanol is transformed to methyl lactate in yields of up to 44% after 20h of reaction at 433 K. When water was used as solvent lactic acid was the main product but its yield decreased to 27%. Recently, Dumesic and coworkers [3] found that polar aprotic organic solvents such as γ- valerolactone (GVL) cause significant increases in reaction rates compared to water in addition to increased product selectivity for Brønsted acid-catalyzed reactions like conversion of xylose into furfural, dehydration of 1,2-propanediol to propanal and for the hydrolysis of cellobiose to glucose. Here we report a similar effect for the Lewis acid-catalyzed conversion of fructose to lactic acid. We demonstrate that the combination of a Sn-Beta zeolite with mainly Lewis acidity prepared using a post-synthetic procedure and a solution of GVL and water as solvent is an effective catalytic process for the production of lactic acid from fructose. Materials and Methods The Sn-Beta zeolite was synthesized using the procedure reported by Sels et al. [4]. Commercial Beta zeolite (Grace Davison Division) with a Si/Al ratio of 15 was dealuminated by stirring in a 7 M aqueous nitric acid solution at 353 K overnight. Afterwards the powder was filtered, rinsed with water and dried. Tin (IV) was grafted on the dealuminated sample by suspending it in isopropanol and adding SnCl4·5H2O. The solution was refluxed under N2 for 7 h. The sample was calcined using a ramp of 3 K min -1 to 473 K, kept at 423 K for 6 h, ramped to 823 K and kept at 823 K for 6 h. The catalysts were characterized using nitrogen adsorption, X-ray Diffraction, X-ray Photoelectron Spectroscopy, Fourier Transform Infrared Spectroscopy and Inductively Coupled Plasma Atomic Emission Spectroscopy. For reaction kinetics experiments, a solution containing 46 mg of fructose, 3 g of solvent (either water or 9:1 GVL:H2O), and 33 mg of catalyst were added in 10 mL thick-walled glass reactors with a small magnetic stir bar. The reactors were placed in an oil bath at 433 K and stirred at 700 rpm. Reactors were removed from the oil bath at specific reaction times and cooled by placing in cold water. Concentrations in liquid solution were quantified using a Waters 600 HPLC system with a Waters 2410 refractive index detector and an ion-exclusion column (Bio-Rad Aminex HPX-87H). The mobile phase used was a 0.005 M sulfuric acid solution at a flow rate of 0.6 mL min -1 and a column temperature of 353 K. Results and Discussion Highly crystalline Sn-Beta zeolite catalyst was synthesized with a surface area of 536 m 2 g -1 and a composition of 0.131% Al, 0.907% Sn, and 41.4% Si. This corresponding to a Si/Sn ratio of 193 and a loading of 0.076 mmol Sn g -1 . XPS analysis did not show evidence of surface Al. Experimental results for the conversion of fructose to lactic acid with Sn-Beta zeolite catalyst are shown in Table 1. The results indicate that when the reaction is conducted using GVL:H2O as solvent the initial turnover frequency (TOF) is 262 h -1 , whereas in the presence of only water the initial TOF is 97 h -1 . This demonstrating an almost three-fold increase in the initial rate when GVL:H2O is used as solvent. Furthermore, a four-fold increase in the production rate of lactic acid (mol lactic acid per mol of Sn per h) from 78 to 293 is observed when GVL:H2O is used as solvent compared to only water. The lactic acid yields at equal conversions show a significant improvement when the reaction is conducted in GVL:H2O as shown in entries 1 and 4 (37 vs. 24% yield at 60-65% conversion) and in entries 2 and 5 (41 vs. 31% at 86% conversion). At longer reactions times of 5 h the yield towards lactic acid slightly increases to 48% in GVL:H2O, whereas in water the yield decreases to 31%. When GVL:H2O is used as solvent in the absence of a catalyst (entries 6 and 7) there is significant fructose conversion, but with low lactic acid selectivity and yield. The increase in reaction rate may be explained by an acceleration of the rate-determining step, i.e., the conversion of dihydroxyacetone to pyruvic aldehyde as suggested by Sels an coworkers [5]. That step is catalyzed by the low Brønsted acidity present on the sample (from the remaining Al) that is enhanced by GVL by the mechanism proposed by Dumesic [3]. A similar behavior for the Lewis acid catalyzed retro- aldol condensation and isomerization steps may explain the enhancement in selectivity. Table 1. Conversion of fructose to lactic acid in different solvents in the presence of Sn- Beta zeolite. a Entry Solvent Catalyst Time (h) Conversion (%) Yield (%) 1 9:1 GVL:H2O Sn-Beta 0.25 65 37 2 9:1 GVL:H2O Sn-Beta 0.5 86 41 3 9:1 GVL:H2O Sn-Beta 5 96 48 4 H2O Sn-Beta 1 60 24 5 H2O Sn-Beta 5 86 31 6 9:1 GVL:H2O No cat 1 72 6 7 9:1 GVL:H2O No cat 5 94 6 a Reactions were carried out at 433 K in a batch reactor containing (46 mg) of fructose, (3 g) of solvent and (33 mg) of Sn-Beta where indicated. Significance The use of Sn-Beta zeolite prepared using a post-synthetic procedure amenable to scale up and GVL:water as solvent offers a viable pathway for the conversion of fructose to the commodity chemical lactic acid in high yields. References 1. Nattrass, L., and Higson, A. The National Non-Food Crops Centr. National Non-Food Crops Centre (NNFCC) Renewable Chemicals Factsheet: Lactic Acid. (February 4, 2014). 2. Holm, M. S., Saravanamurugan, S., and Taarning, E. Science 328 (5978), 602 (2010). 3. Mellmer, M. A., Sener, C., Gallo, J. M. R., Luterbacher, J. S., Alonso, D. M., and Dumesic, J. A. Angew. Chem. Int. Ed. 53, 11872 (2014). 4. Dijkmans, J., Gabriels, D., Dusselier, M., de Clippel, F., Vanelderen, P., Houthoofd, K., Malfliet, A., Pontikes, Y., and Sels, B. F. Green Chem. 15 (10), 2777 (2013). 5. de Clippel, F., Dusselier, M., Van Rompaey, R., Vanelderen, P., Dijkmans, J., Makshina, E., Giebeler, L., Oswald, S., Baron, G. V., Denayer, J. F. M., Pescarmona, P. P., Jacobs, P. A., and Sels, B. F. J. Am. Chem. Soc. 134 (24), 10089 (2012).