[Show abstract] [Hide abstract]
ABSTRACT: Fructose dehydration to 5-hydroxymethylfurfural (HMF) is a well-studied reaction and it is considered a characteristic example of converting carbohydrates to useful chemical intermediates . The choice of solvent (DMSO, water or DMSO – water mixture) significantly affects its selectivity and yield [1,2]. For example, the acid catalyzed dehydration in an aqueous environment is accompanied by side reactions which lead to the formation of acyclic dehydration and fragmentation products, insoluble humins and HMF rehydration products. On the other hand, HMF yields higher than 90% can be achieved in DMSO . Unfortunately, issues related to the separation of HMF from DMSO render this solvent unattractive for commercial applications. The goal of this work is to understand the solvation of fructose and HMF in DMSO and to provide insights about the effect of solvation on the selectivity of fructose dehydration. To achieve this goal we used a combination of vibrational spectroscopy (ATR/FTIR, Raman) and molecular simulations (Molecular Dynamics).
The analysis of the fructose hydroxyl hydrogen – DMSO oxygen radial distribution function showed that the coordination number of DMSO around fructofuranose is ~3.5. This number is smaller than the number of hydroxyl groups of fructose because one DMSO is shared between two hydroxyl groups and because of the formation of intramolecular hydrogen bonds. In the case of fructose - DMSO mixtures, a red shift of the Raman S=O asymmetric stretching is observed which indicates that fructose breaks the DMSO clusters owing to strong hydrogen bonding between the hydrogen of its hydroxyl groups and the oxygen of DMSO. The Raman scattering cross-sections of the DMSO S=O stretching when a DMSO molecule interacts with another DMSO, fructose, water or HMF were estimated from the spectra of the binary mixtures using the coordination numbers from the MD simulations. It was also possible to use these values together with the MD predicted coordination numbers to satisfactorily predict the effect of water fraction on the Raman scattering intensity of the S=O stretching band in ternary mixtures when the H2O/DMSO molar ratio was less than ~0.3. MD simulations also showed that with increasing water content the DMSO orientation around fructose changes with the sulfur atom moving away from the carbohydrate. The analysis of the ATR/FTIR spectra reveal that with increasing water content the average hydrogen bond enthalpy of the fructose hydroxyls decreases by ~2.5 kJ/mol. The MD simulations indicate that this decrease is due to the reduction of the strength of DMSO – fructose hydrogen bond, and due to the replacement of a fraction of the stronger DMSO – fructose hydrogen bonds by the weaker fructose – water ones. Finally, MD simulations also show the preferential coordination of DMSO around HMF carbonyl group could shield HMF from further rehydration to levulinic acid and formic acid and from humins formation.
1. Kuster, B. F. M., Starch 1990, 42 (8), 314-321.
2. Chheda, J. N.; Roman-Leshkov, Y.; Dumesic, J. A. Green Chemistry 2007, 9 (4), 342-350
3. Musau, R. M.; Munavu, R. M., J of Chem Tech. and Biotechnology 1987, 32, 920-924
4. Hess, B.; Kutzner, C.; van der Spoel, D.; Lindahl, E., J. Chem Theory and Comp. 2008, 4 (3), 435-447.
12 AIChE Annual Meeting; 10/2012
[Show abstract] [Hide abstract]
ABSTRACT: The solvation of fructose in dimethyl sulfoxide (DMSO) and DMSO-H(2)O (or DMSO-D(2)O) mixtures was investigated using vibrational spectroscopy (Raman, ATR/FTIR) and molecular dynamics (MD) simulations. The analysis of the fructose hydroxyl hydrogen-DMSO oxygen radial distribution function showed that the coordination number of DMSO around the furanose form of fructose is ∼3.5. This number is smaller than the number of hydroxyl groups of fructose because one DMSO molecule is shared between two hydroxyl groups and because intramolecular hydrogen bonds are formed. In the case of fructose-DMSO mixtures, a red shift of the Raman S═O asymmetric stretch is observed, which indicates that fructose breaks the DMSO clusters through strong hydrogen bonding between the hydrogen atoms of its hydroxyl groups and the oxygen atom of DMSO. The Raman scattering cross sections of the DMSO S═O stretch when a DMSO molecule interacts with another DMSO molecule, a fructose molecule, or a water molecule were estimated from the spectra of the binary mixtures using the coordination numbers from MD simulations. It was also possible to use these values together with the MD-estimated coordination numbers to satisfactorily predict the effect of the water fraction on the Raman scattering intensity of the S═O stretching band in ternary mixtures. MD simulations also showed that, with increasing water content, the DMSO orientation around fructose changed, with the sulfur atom moving away from the carbohydrate. The deconvolution of the fructose IR OH stretching region revealed that the hydroxyls of fructose can be separated into two groups that participate in hydrogen bonds of different strengths. MD simulations showed that the three hydroxyls of the fructose ring form stronger hydrogen bonds with the solvent than the remaining hydroxyls, providing an explanation for the experimental observations. Finally, analysis of ATR/FTIR spectra revealed that, with increasing water content, the average hydrogen-bond enthalpy of the fructose hydroxyls decreases by ∼2.5 kJ/mol.
The Journal of Physical Chemistry B 08/2012; 116(36):11274-83. · 3.38 Impact Factor
[Show abstract] [Hide abstract]
ABSTRACT: Surface plasmon resonance spectroscopy is sensitive to near-surface (<300 nm) chemical and physical events that result in refractive index changes. The non-specific nature of the stimulus implies that chemical selectivity in SPR sensing configurations entirely relies upon the chemical recognition scheme employed. Biosensing applications commonly use surface layers composed of antibodies or enzymes for biomolecular recognition. Monitoring of volatile compounds with SPR spectroscopy, however, has not been widely discussed due to the difficulty in selectively responding to small molecules (<100 Da) in addition to the limited refractive index changes resulting from the interaction between the plasmon wave and volatile compounds. Different strategies explored thus far for sensing of small molecules have relied on optical and electrical changes of the recognition layer upon exposure to the analyte, yielding an indirect measurement. Examples of coatings used for gas-phase sensing with SPR include conducting metal oxides, polymers and organometallic dyes. Electrically conducting polymers, like polyaniline and polypyrrole, display dramatic conductivity changes in the presence of certain compounds. This property has resulted in their routine incorporation into different sensing schemes. However, application of electrically conducting polymers to SPR gas-phase sensing has been limited to a few examples, despite encouraging results. The emeraldine salt form of polyaniline (PAni) demonstrates a decreased electrical conductivity correlated to NH(3) concentration. In this contribution, PAni doped with camphorsulfonic acid (PAni-CSA) was applied to gas-phase sensing of NH(3) by way of SPR spectroscopy. Spectroscopic ellipsometry was used to determine the optical constants (n and k) for emeraldine salt and emeraldine base forms of PAni, confirming the wavelength-dependent response observed via SPR. The analytical performance of the coatings show that a limit of detection of 32 ppm NH(3) based on precision of the mass-flow controllers used and an estimated method limit of detection of ∼0.2 ppm based on three standard deviations of the blank. This is directly comparable to other, more established sensing architectures.
Talanta 09/2011; 85(3):1369-75. · 3.50 Impact Factor