Selective Base-Catalyzed Isomerization of Glucose to Fructose
Supporting Information
Selective Base-Catalyzed Isomerization of Glucose to Fructose Chi Liu, Jack M. Carraher, Jordan L. Swedberg, Caitlyn R. Herndon, Chelsea N. Fleitman, Jean-Philippe Tessonnier* Department of Chemical & Biological Engineering, Iowa State University, Ames, Iowa 50011, United States. NSF Engineering Research Center for Biorenewable Chemicals (CBiRC), Ames, Iowa 50011, United States.
* Author to whom correspondence should be addressed: [email protected] Experimental methods Isomerization experiments were carried out in 5 ml thick-walled glass reactors (Chemglass Life Sciences) heated with a digital stirring hotplate (IKA RCT) equipped with an aluminum pie-block system (Chemglass Life Sciences). The PT1000 thermocouple that controls the heating plate was placed in a reaction vial containing 5 ml silicon oil, near the other reaction vials. In a typical experiment, 500 mg +/2 mg D-Glucose (≥99.5% Sigma-Aldrich MO, USA), a V-shaped stir bar, and 5 ml of an aqueous solution containing the amine catalyst (NR3:glucose molar ratio of 2, 5, 10, 20, 30%) were added to the glass reactor and sealed. The glucose concentration in the solution was kept constant at 10% (w/w) for all the experiments. The reactor was then placed in the pre-heated pie-block and heated for a specific time. The reaction was quenched by placing the reactor in an ice bath and the reactor was later stored in a freezer until analysis. All the catalysts were purchased from Sigma-Aldrich: Morpholine (≥99%), Piperazine (99%), Ethylenediamine (≥99.5%), Triethylamine (≥99%), Piperidine (99%), and Pyrrolidine (99.5%). Sample analysis was performed using a Waters Acquity H-Class ultra-performance liquid chromatograph (UPLC) equipped with photodiode array (PDA) and evaporative light scattering (ELS) detectors. Glucose, fructose, and mannose were analyzed with a Waters UPLC BEH Amide 1.7 μm (2.1 x 100 mm) column. Solutions were allowed to warm up to room temperature and were then diluted with appropriate amounts (typically 20 fold) of a 50:50 acetonitrile:water mixture before injection. A mobile phase gradient based on mixtures of 0.2 v/v% of TEA in Acetonitrile and 0.2 v/v% of TEA in Water was used to achieve a good separation of all the reactants and products. UV spectra were taken with a Varian Cary 50 Bio UV-Vis Spectrophotometer using Far UV quartz cells purchased from Spectrocell (Oreland, PA). The spectra are background corrected using DI water as a reference. 1
H NMR spectra were obtained with a Bruker AVIII-600 (600 MHz, rd = 1 s, NS = 32) and analyzed with MestReNova version 8.1 software.
Page 1 of 7
Figure S1. Detailled catalytic results for the amine-catalyzed isomerization of glucose to fructose
Figure S1 summarizes the results presented in Table 1. The catalytic tests were performed with a 10% (w/w) glucose/water solution, 10 mol.% N relative to glucose, 100 °C, 30 min. The amines were ranked by increasing basicity, from morpholine (pKa 8.4) to pyrrolidine (pKa 11.3). The pKa values are reported on the graph for easier comparison. Each experiment was repeated at least 3 times. Figure S1 shows the average conversion, selectivity, and yield for each amine. The error bars represent the standard error of the mean (the detailed method can be found at http://www.ncsu.edu/labwrite/res/gt/gt-stat-home.html ). Some of the error bars are very small, thus difficult to see on the graph.
Page 2 of 7
Figure S2. 1H NMR spectrum of the reacted solution after purification with activated carbon
Catalytic test performed with a 10% (w/w) glucose/D2O solution using 10 mol.% triethylamine as a catalyst. The reaction was carried out for 20 min at 100 °C. The solution was mixed after reaction with 5 wt.% of Darco® KB-G activated carbon, stirred for 1 h, filtered, and analyzed by 1H NMR. The chemical shifts in Figure S2 are consistent with glucose, fructose, and triethylamine. The relative integration of -CH2- and -CH3 signals (2:3) from triethylamine indicates that the catalyst has not undergone chemical transformation in a browning reaction. Other byproducts were not detected, thus confirming that their relative concentration was below the detection limit of this technique.
Page 3 of 7
Figure S3. UV-vis difference spectra of reacted solutions utilizing different organic bases
Catalytic test performed with a 10% (w/w) glucose/H2O solution using 10 mol.% N relative to glucose with morpholine, piperazine, piperidine, pyrrolidine, triethylamine, and ethylenediamine as catalysts. Aliquots of reaction solutions were collected before reaction and diluted 10-fold for acquisition of UV-vis spectra. After 30 min at 100 °C, reaction solutions were diluted 100-fold with H2O, and UV-vis spectra acquired. The path length for acquisition of UV-vis spectra was 1 cm. The red-shift in absorbance maxima observed with secondary amines is inversely related to pKa, and is consistent with the browning reaction requiring deprotonated amine. However, even the most extreme shift observed with secondary amines (morpholine) appears more closely related to the thermal degradation observed with triethylamine than browning with ethylenediamine.
Page 4 of 7
Figure S4. 1H NMR of 10 w/w % glucose with 0.12 mol % TEA in D2O before and after 25 min at 80 oC
1
H NMR spectra acquired at room temperature, 600 MHz, D2O. 0.56 M glucose + 67 mM TEA (top) and after 25 minutes at 80 oC (bottom). Shift of -CH2- signal of triethylamine from 2.78 ppm (top) to 2.83 ppm (bottom) indicates a change in pD from 11.0 to 10.9 (see Figures S5 and S6). Identical experiments carried out in H2O had pH measured by pH electrode. Initial and final pH were measured at 11.0 and 10.8, respectively. No change in triethylamine concentration or structure was observed, thus demonstrating that triethylamine is not consumed during the reaction.
Page 5 of 7
Figure S5. 1H NMR of Triethylamine with HCl
1
H NMR spectra acquired at room temperature, 600 MHz, D2O. Solutions contain 0.10 M triethylamine and 0 (bottom), 0.50 (middle), and 0.99 (top) equivalents of HCl. Ionic strength was held constant at 0.10 M with NaCl. Shifts in peak position are explained in Figure S6.
Page 6 of 7
Figure S6. Plot of 1H NMR chemical shift relative to HCl equivalents
3.5
1H
NMR Chemical Shift
3 2.5 2 1.5 1 0.5 0 0
0.2
0.4
0.6
0.8
1
Equivalents of HCl
Chemical shift of triethylamine -CH2- • and -CH3 • as a function of HCl equivalents.
Page 7 of 7
Selective Base-Catalyzed Isomerization of Glucose to Fructose Chi Liu, Jack M. Carraher, Jordan L. Swedberg, Caitlyn R. Herndon, Chelsea N. Fleitman, Jean-Philippe Tessonnier* Department of Chemical & Biological Engineering, Iowa State University, Ames, Iowa 50011, United States. NSF Engineering Research Center for Biorenewable Chemicals (CBiRC), Ames, Iowa 50011, United States.
* Author to whom correspondence should be addressed: [email protected] Experimental methods Isomerization experiments were carried out in 5 ml thick-walled glass reactors (Chemglass Life Sciences) heated with a digital stirring hotplate (IKA RCT) equipped with an aluminum pie-block system (Chemglass Life Sciences). The PT1000 thermocouple that controls the heating plate was placed in a reaction vial containing 5 ml silicon oil, near the other reaction vials. In a typical experiment, 500 mg +/2 mg D-Glucose (≥99.5% Sigma-Aldrich MO, USA), a V-shaped stir bar, and 5 ml of an aqueous solution containing the amine catalyst (NR3:glucose molar ratio of 2, 5, 10, 20, 30%) were added to the glass reactor and sealed. The glucose concentration in the solution was kept constant at 10% (w/w) for all the experiments. The reactor was then placed in the pre-heated pie-block and heated for a specific time. The reaction was quenched by placing the reactor in an ice bath and the reactor was later stored in a freezer until analysis. All the catalysts were purchased from Sigma-Aldrich: Morpholine (≥99%), Piperazine (99%), Ethylenediamine (≥99.5%), Triethylamine (≥99%), Piperidine (99%), and Pyrrolidine (99.5%). Sample analysis was performed using a Waters Acquity H-Class ultra-performance liquid chromatograph (UPLC) equipped with photodiode array (PDA) and evaporative light scattering (ELS) detectors. Glucose, fructose, and mannose were analyzed with a Waters UPLC BEH Amide 1.7 μm (2.1 x 100 mm) column. Solutions were allowed to warm up to room temperature and were then diluted with appropriate amounts (typically 20 fold) of a 50:50 acetonitrile:water mixture before injection. A mobile phase gradient based on mixtures of 0.2 v/v% of TEA in Acetonitrile and 0.2 v/v% of TEA in Water was used to achieve a good separation of all the reactants and products. UV spectra were taken with a Varian Cary 50 Bio UV-Vis Spectrophotometer using Far UV quartz cells purchased from Spectrocell (Oreland, PA). The spectra are background corrected using DI water as a reference. 1
H NMR spectra were obtained with a Bruker AVIII-600 (600 MHz, rd = 1 s, NS = 32) and analyzed with MestReNova version 8.1 software.
Page 1 of 7
Figure S1. Detailled catalytic results for the amine-catalyzed isomerization of glucose to fructose
Figure S1 summarizes the results presented in Table 1. The catalytic tests were performed with a 10% (w/w) glucose/water solution, 10 mol.% N relative to glucose, 100 °C, 30 min. The amines were ranked by increasing basicity, from morpholine (pKa 8.4) to pyrrolidine (pKa 11.3). The pKa values are reported on the graph for easier comparison. Each experiment was repeated at least 3 times. Figure S1 shows the average conversion, selectivity, and yield for each amine. The error bars represent the standard error of the mean (the detailed method can be found at http://www.ncsu.edu/labwrite/res/gt/gt-stat-home.html ). Some of the error bars are very small, thus difficult to see on the graph.
Page 2 of 7
Figure S2. 1H NMR spectrum of the reacted solution after purification with activated carbon
Catalytic test performed with a 10% (w/w) glucose/D2O solution using 10 mol.% triethylamine as a catalyst. The reaction was carried out for 20 min at 100 °C. The solution was mixed after reaction with 5 wt.% of Darco® KB-G activated carbon, stirred for 1 h, filtered, and analyzed by 1H NMR. The chemical shifts in Figure S2 are consistent with glucose, fructose, and triethylamine. The relative integration of -CH2- and -CH3 signals (2:3) from triethylamine indicates that the catalyst has not undergone chemical transformation in a browning reaction. Other byproducts were not detected, thus confirming that their relative concentration was below the detection limit of this technique.
Page 3 of 7
Figure S3. UV-vis difference spectra of reacted solutions utilizing different organic bases
Catalytic test performed with a 10% (w/w) glucose/H2O solution using 10 mol.% N relative to glucose with morpholine, piperazine, piperidine, pyrrolidine, triethylamine, and ethylenediamine as catalysts. Aliquots of reaction solutions were collected before reaction and diluted 10-fold for acquisition of UV-vis spectra. After 30 min at 100 °C, reaction solutions were diluted 100-fold with H2O, and UV-vis spectra acquired. The path length for acquisition of UV-vis spectra was 1 cm. The red-shift in absorbance maxima observed with secondary amines is inversely related to pKa, and is consistent with the browning reaction requiring deprotonated amine. However, even the most extreme shift observed with secondary amines (morpholine) appears more closely related to the thermal degradation observed with triethylamine than browning with ethylenediamine.
Page 4 of 7
Figure S4. 1H NMR of 10 w/w % glucose with 0.12 mol % TEA in D2O before and after 25 min at 80 oC
1
H NMR spectra acquired at room temperature, 600 MHz, D2O. 0.56 M glucose + 67 mM TEA (top) and after 25 minutes at 80 oC (bottom). Shift of -CH2- signal of triethylamine from 2.78 ppm (top) to 2.83 ppm (bottom) indicates a change in pD from 11.0 to 10.9 (see Figures S5 and S6). Identical experiments carried out in H2O had pH measured by pH electrode. Initial and final pH were measured at 11.0 and 10.8, respectively. No change in triethylamine concentration or structure was observed, thus demonstrating that triethylamine is not consumed during the reaction.
Page 5 of 7
Figure S5. 1H NMR of Triethylamine with HCl
1
H NMR spectra acquired at room temperature, 600 MHz, D2O. Solutions contain 0.10 M triethylamine and 0 (bottom), 0.50 (middle), and 0.99 (top) equivalents of HCl. Ionic strength was held constant at 0.10 M with NaCl. Shifts in peak position are explained in Figure S6.
Page 6 of 7
Figure S6. Plot of 1H NMR chemical shift relative to HCl equivalents
3.5
1H
NMR Chemical Shift
3 2.5 2 1.5 1 0.5 0 0
0.2
0.4
0.6
0.8
1
Equivalents of HCl
Chemical shift of triethylamine -CH2- • and -CH3 • as a function of HCl equivalents.
Page 7 of 7
