Isomerization of aldo-pentoses into keto-pentoses catalyzed by phosphates

Transcription

1 Electronic Supplementary Material (ESI) for Green Chemistry. This journal is The Royal Society of Chemistry 2017 Isomerization of aldo-pentoses into keto-pentoses catalyzed by phosphates Irina Delidovich,* a Maria S. Gyngazova, a Nuria Sánchez-Bastardo, b Julia P. Wohland, a Corinna Hoppe a and Peter Drabo a a Chair of Heterogeneous Catalysis and Chemical Technology, RWTH Aachen University, Worringerweg 2, Aachen, Germany. Delidovich@itmc.rwth-aachen.de b High Pressure Processes Group, Chemical Engineering and Environmental Technology Department, Escuela de Ingenierías Industriales, University of Valladolid, Valladolid, Spain Fig. 1S Equilibrium constants of the isomerization reactions (left) and equilibrium yields (right) of keto-pentoses as functions of temperature according to the thermodynamic data reported by Tewari et al. 1, 2 S1

2 13 CH 13 CH H C H C H C CH 2 Ribose Retro aldol CH 2 Glycolaldehyde + CH Isomerization H C CH 2 Glyceraldehyde CH 2 C CH 2 Dihydroxyacetone Aldol 13 CH + CH 2 CH 2 C H C 13 H C CH 2 Ribulose 13 CH H C H C H C CH 2 Ribose Isomerization 13 CH 2 C H C H C CH 2 Ribulose Retro aldol 13 CH 2 C CH 2 Dihydroxyacetone + CH CH 2 Glycolaldehyde Aldol 13 CH 2 C H C H C CH 2 Ribulose + CH 2 C 13 H C H C CH 2 Ribulose Scheme 1S. Isotope label scrambling with D-(1-13 C)-ribose as a substrate via consecutive retro aldol and aldol reactions. According to the previously published data, isomerization of glyceraldehyde into dihydroxyacetone (shown in the upper sequence of reactions) is highly favorable in the presence of NaH 2 P 4 +Na 2 HP 4. 3 S2

3 Table 1S. Assignments of the resonance signals in the NMR spectra recorded after the isomerization of D-(1-13 C)-ribose in the presence of NaH 2 P 4 +Na 2 HP 4 (Fig. 2). F and P denote a furanose and a pyranose cyclic form, respectively. Peak number Chemical shift, ppm Assignment of the resonance signal C2-β-Xylulose C1-βF-Ribose C1-αP-Arabinose C1-αF-Ribose C1-βP-Ribose C1-αP-Ribose C5-α-Xylulose C5-β-Ribulose C5-β-Xylulose C1-keto-Ribulose C1-keto-Xylulose C1-β-Xylulose C1-α-Ribulose C1-β-Ribulose C1-α-Xylulose M 49.5 Methanol (internal standard) S3

4 Fig. 2S. Linearized dependence of R 0, the initial reaction rate of xylose isomerization, on C 0, the initial concentration of xylose to determine the reaction order α: lnr 0 =αlnc 0 +lnk. Reaction conditions: 2, 5 or 7 wt.% aqueous solution of a substrate, 78 C, 750 rpm, 0.5M NaH 2 P 4 +Na 2 HP 4. S4

5 By-product 3 By-product 4 CH CH k 10 H H k 8 H H H H k 1 k 5 H k CH 2-1 k -5 CH 2 Ribose CH 2 CH 2 Xylose k 4 k -2 k 2 H H H k -7 k 7 H k -4 H CH 2 CH 2 CH Ribulose Xylulose CH k 11 H H H H k 3 k 6 H H H k 9 H k -3 k -6 H CH 2 CH 2 Arabinose Lyxose By-product 1 By-product 2 Scheme 2S. Extended reaction network used for kinetic modeling Based on this reaction network, we derive the following system of ordinary differential equations (DEs): d[ribose] (k 1 k 2 k 10 ) [Ribose] k 1 [Ribulose] k 2 [Arabinose] dt d[arabinose] (k 2 k 3 k 11 ) [Arabinose] k 3 [Ribulose] k 2 [Ribose] dt d[ribulose] (k 1 k 3 k 4 ) [Ribulose] k 1 [Ribose] k 4 [Xylulose] k 3 [Arabinose] dt d[xylulose] (k 4 k 5 k 6 ) [Xylulose] k 4 [Ribulose] k 5 [Xylose] k 6 [Lyxose] dt d[xylose] (k 5 k 7 k 8 ) [Xylose] k 5 [Xylulose] k 7 [Lyxose] dt d[lyxose] (k 6 k 7 k 9 ) [Lyxose] k 6 [Xylulose] k 7 [Xylose] dt d[by product 1] k 8 [Xylose] dt d[by product 2] k 9 [Lyxose] dt d[by product 3] k 10 [Ribose] dt d[by product 4] k11 [Arabinose] dt (1S) (2S) (3S) (4S) (5S) (6S) (7S) (8S) (9S) (10S) S5

6 Fig. 3S. Model (solid line) and experimental data (markers) of xylose isomerization in the presence of NaH 2 P 4 +Na 2 HP 4 at different temperatures. Reaction conditions: 10wt. % of the substrate, 0.5M NaH 2 P 4 +Na 2 HP 4 at ph 7.5, 750 rpm. S6

7 Fig. 4S. Model (solid line) and experimental data (markers) of lyxose isomerization in the presence of NaH 2 P 4 +Na 2 HP 4 at different temperatures. Reaction conditions: 10wt. % of the substrate, 0.5M NaH 2 P 4 +Na 2 HP 4 at ph 7.5, 750 rpm. S7

8 Fig. 5S. Model (solid line) and experimental data (markers) of arabinose isomerization in the presence of NaH 2 P 4 +Na 2 HP 4 at different temperatures. Reaction conditions: 10wt. % of the substrate, 0.5M NaH 2 P 4 +Na 2 HP 4 at ph 7.5, 750 rpm. S8

9 Fig. 6S. Model (solid line) and experimental data (markers) of ribose isomerization in the presence of NaH 2 P 4 +Na 2 HP 4 at different temperatures. Reaction conditions: 10wt. % of the substrate, 0.5M NaH 2 P 4 +Na 2 HP 4 at ph 7.5, 750 rpm. S9

10 Table 2S. Effective pre-exponential factors (k 0 ) and activation energies (E a ) obtained for the degradation of pentoses in the presence of NaH 2 P 4 +Na 2 HP 4. Reaction ln k 0 E a, kj/mol Xylose k8 By products Lyxose k9 By products Ribose k10 By products Arabinose k11 By products S10

11 Table 3S. Comparison of the obtained values for the equilibrium constants (K) with the literature data Reaction Expression for K value a K at 78 C obtained in this work K at 78 C, literature data Ribose Ribulose K 1 =k 1 /k Ribose Arabinose K 2 =k 2 /k Arabinose Ribulose K 3 =k 3 /k Ribulose Xylulose K 4 =k 4 /k No data Xylose Xylulose K 5 =k -5 /k Lyxose Xylulose K 6 =k -6 /k No data Lyxose Xylose K 7 =k -7 /k b 4 a Expressions were derived in accordance with the reaction network presented in Scheme 2S. b Temperature was not specified in the reference. S11

12 Estimating accuracy of rate constants Interconversions of ribose-ribulose-arabinose and their equilibrium constants can be presented as follows: Ribose k 1 Ribulose K 1 =[Ribulose]/[Ribose]=k 1 /k -1 k -1 Ribose k 2 k -2 Arabinose K 2 =[Arabinose]/[Ribose]=k 2 /k -2 Arabinose k 3 k -3 Ribulose K 3 =[Ribulose]/[Arabinose]=k 3 /k -3 A certain combination of the equilibrium constants would be equal to 1 because a microscopic reversibility determines the Gibbs energy of the cycles equal to zero: Y = 1 = [Ribulose] [Ribose] [Arabinose] [Ribulose] [Ribose] [Arabinose] = K 1 = k 1 k 3 k 2 K 3 K 2 k 1 k 3 k 2 (11S) An absolute error of a function Y (x 1, x 2,, x n ) can be calculated as follows: n Y = i = 1 Y x x i i (12S) In our case Y(k 1, k -1, k 2, k -2, k 3, k -3 )=(k 1 k -3 k -2 )/(k -1 k 3 k 2 ), as indicated by the equation (11S). Assuming the same relative error for all the obtained rate constants, i.e. Δk i /k i is independent of i, the following formula could be derived after differentiation of the Y function: Y Y = n i = 1 k i = 6 k i k i k i (13S) As a result, a relative rate of determining the rate constants can be estimated as follows: k i k i = ( Y Y ) /6 (14S) Thus obtained values are provided in Table 4S. The same approach was used to calculate the relative errors of the rate constants of lyxose-xylulose-xylose interconversions. S12

13 Table 4S. Estimating relative errors of determining the rate constants (Δk i /k i ) Entry Temperature, C Y a calc ΔY/Y, % b Δk i /k i, % Ribose-ribulose-arabinose, Y=(k 1 k -3 k -2 )/(k -1 k 3 k 2 ) Lyxose-xylulose-xylose, Y=(k -5 k 6 k -7 )/ (k 5 k -6 k 7 ) a Y calc denotes Y value calculated according to the equation (11S) based on the determined rate constants; b Relative error of Y was determined as follows: ΔY/Y=( Y calc -Y true /Y true ) 100%, where Y true =1, since Gibbs energy of the cycles is equal to zero. S13

14 Fig. 7S. Parity plot for the substrates and the products. The following data were used for the parity plot: concentrations of all the substrates (xylose, lyxose, arabinose, and ribose) at all the temperatures (56, 64, 70, and 78 C) as well as concentrations of all the isomeric products upon isomerization of xylose at all the temperatures. S14

15 Xylose H H H H CH 2 H CH 2 H -D-Xylopyranoside 63% -D-Xylopyranoside 35.5% -D-Xylofuranoside <1% -D-Xylofuranoside <1% Lyxose H H H H H H CH 2 CH 2 H H -D-Lyxopyranoside 28% -D-Lyxopyranoside 70% -D-Lyxofuranoside 0.5% -D-Lyxofuranoside 1.5% Ribose H H CH 2 H CH 2 H -D-Ribopyranoside 58.5% -D-Ribopyranoside 21.5% -D-Ribofuranoside 13.6% -D-Ribofuranoside 6.5% Arabinose H -D-Arabinopyranoside 35.5% -D-Arabinopyranoside 60% Fig. 8S. Cyclic forms of pentose present in aqueous solution 5 H H H CH 2 H -D-Arabinofuranoside 2% CH 2 H -D-Arabinofuranoside 2.5% S15

16 Table 5S. Conformational free energies of the pyranosides in aqueous equilibrium solutions of D- aldopyranoses. Aldose Free energy, kcal mol -1 α form β form average a Xylose Lyxose Ribose Arabinose Glucose a Average free energy = (Average free energy of α form) (Percentage of α form)+ (Average free energy of β form) (Percentage of β form) S16

17 Fig. 9S. A correlation between activation energy of isomerization of the aldoses into the corresponding ketoses with relative free energy of the substrates. S17

18 Xylose to xylulose Aldo-pentose to epimeric aldopentose Xylose to lyxose Xylose to byproduct 1 Conversion rate of a keto-pentose, mol L -1 s -1 a Formation rate of an epimeric aldopentose, Table 6S. Reaction rates contributing to the overall kinetics of the isomerization of the pentoses in the presence of NaH 2 P 4 +Na 2 HP 4. The reaction rates were calculated based on the results of kinetic modeling at 10% substrate conversion in accordance with the reaction network shown in Scheme S2. R denotes reaction rate. Entry T, Conversion rate of a substrate aldo-pentose, C mol L -1 s -1 a mol L -1 s -1 b Aldo-pentose to keto-pentose Aldo-pentose to by-product Keto-pentose to epimeric aldo- From substrate aldo-pentose From ketopentose pentose Substrate: Xylose Xylulose to lyxose Keto-pentose to C-3 epimeric keto-pentose Xylulose to ribulose Keto-pentose to substrate aldopentose Xylulose to xylose Lyxose formation based on xylose Lyxose formation based on xylulose R = k -5 [Xylose] R = k 7 [Xylose] R = k 8 [Xylose] R = k 6 [Xylulose] R = k -4 [Xylulose] R = k 5 [Xylulose] R = k 7 [Xylose] R = k 6 [Xylulose] (58%) (2%) (40%) (52%) (37%) (11%) (5%) (95%) (65%) (2%) (33%) (55%) (34%) (11%) (4%) (96%) (69%) (2%) (29%) (56%) (33%) (11%) (5%) (95%) (74%) (2%) (24%) (58%) (30%) (12%) (5%) (95%) Substrate: Lyxose Lyxose to xylulose Lyxose to xylose Lyxose to byproduct 2 Xylulose to xylose Xylulose to ribulose Xylulose to lyxose Xylose formation based on lyxose Xylose formation based on xylulose R = k -6 [Lyxose] R = k -7 [Lyxose] R = k 9 [Lyxose] R = k 5 [Xylulose] R = k -4 [Xylulose] R = k 6 [Xylulose] R = k -7 [Lyxose] R = k 5 [Xylulose] (84%) (1%) (15%) (11%) (37%) (52%) (13%) (87%) (84%) (1%) (15%) (11%) (34%) (55%) (13%) (87%) (84%) (1%) (15%) (11%) (33%) (56%) (14%) (86%) (84%) (1%) (15%) (11%) (30%) (58%) (14%) (86%) a differential selectivities are given in parentheses; b percentages of epimeric aldo-pentoses formed via a corresponding reaction pathway are given in parentheses. S18

19 Table 6S (Continuation). Reaction rates contributing to the overall kinetics of the isomerization of the pentoses in the presence of NaH 2 P 4 +Na 2 HP 4. The reaction rates were calculated based on the results of kinetic modeling at 10% substrate conversion in accordance with the reaction network shown in Scheme S2. R denotes reaction rate. Entry T, Conversion rate of a substrate aldo-pentose, Conversion rate of a keto-pentose, mol L -1 s -1 a Formation rate of an epimeric aldopentose, mol L -1 s -1 C mol L -1 s -1 a b Aldo-pentose to keto-pentose Ribose to ribulose Aldo-pentose to epimeric aldopentose Ribose to arabinose Aldo-pentose to by-product Ribose to byproduct 3 Keto-pentose to epimeric aldopentose Substrate: Ribose Ribulose to arabinose Keto-pentose to C-3 epimeric keto-pentose Ribulose to xylulose Keto-pentose to substrate aldopentose Ribulose to ribose From substrate aldo-pentose Arabinose formation based on ribose From ketopentose Arabinose formation based on ribulose R = k 1 [Ribose] R = k 2 [Ribose] R = k 10 [Ribose] R = k -3 [Ribulose] R = k 4 [Ribulose] R = k -1 [Ribulose] R = k 2 [Ribose] R = k -3 [Ribulose] (78%) (1%) (21%) (9%) (52%) (39%) (27%) (73%) (82%) (1%) (17%) (9%) (50%) (41%) (30%) (70%) (84%) (1%) (14%) (9%) (49%) (42%) (38%) (62%) (87%) (2%) (11%) (9%) (47%) (44%) (29%) (71%) Substrate: Arabinose Arabinose to ribulose Arabinose to ribose Arabinose to by-product 4 Ribulose to ribose Ribulose to xylulose Ribulose to arabinose Ribose formation based on arabinose Ribose formation based on ribulose R = k 3 [Arabin.] R = k -2 [Arabin.] R = k 11 [Arabin.] R = k -1 [Ribulose] R = k 4 [Ribulose] R = k -3 [Ribulose] R = k -2 [Arabin.] R = k -1 [Ribulose] (76%) (14%) (10%) (39%) (52%) (9%) (42%) (58%) (71%) (11%) (17%) (41%) (50%) (9%) (20%) (80%) (66%) (9%) (25%) (42%) (49%) (9%) (18%) (82%) (57%) (7%) (36%) (44%) (47%) (9%) (16%) (84%) a differential selectivities are given in parentheses; b percentages of epimeric aldo-pentoses formed via a corresponding reaction pathway are given in parentheses. S19

20 Table 6S presents reaction rates and values of differential selectivities towards the products. By definition, differential selectivity is the ratio of the rate of desired product formation to the rate of reaction consumption. 6 The differential selectivities were calculated as follows: 1. Differential selectivities for conversion of a substrate into a ketose, an epimeric aldopentose and by products The calculations for xylose as a substrate were performed according to the following formula: S diff (Xylose for xylulose) R(Xylose intoxylulose) R(Xylose intoxylulose) R(Xylose intolyxose) R(Xylose intoby product 1) k k 5 [Xylose] 5 5 [Xylose] k k 7 [Xylose] k 8 [Xylose] k 5 k 7 k 8 (15S) Analogously differential selectivities can be expressed for all the products and all the substrates k 7 Sdiff (Xylose for lyxose) k 5 k 7 k 8 S k8 diff (Xylose for by product1) k 5 k 7 k 8 k S 6 diff (Lyxose for xylulose) k 6 k 7 k 9 k S 7 diff (Lyxose for xylose) k 6 k 7 k 9 S k9 diff (Lyxose for by product 2) k 6 k 7 k9 S k1 diff (Ribose for ribulose) k1 k2 k10 S k2 diff (Ribose for arabinose) k1 k2 k10 S k10 diff (Ribose for by product 3) k 1 k 2 k 10 S k3 diff (Arabinose for ribulose) k3 k 2 k11 k S 2 diff (Arabinose for ribose) k3 k 2 k11 k11 Sdiff (Arabinose for by product 4) k3 k 2 k11 (16S) (17S) (18S) (19S) (20S) (21S) (22S) (23S) (24S) (25S) (26S) 2. Differential selectivities for conversion of a keto-pentose into a C-3 epimeric ketose and aldoses S diff (Xylulose for ribulose) k 4 k 4 k5 k6 (27S) S20

21 k 5 Sdiff (Xylulose for xylose) k 4 k 5 k 6 k 6 Sdiff (Xylulose for lyxose) k 4 k 5 k 6 k 4 Sdiff (Ribulose for xylulose) k 4 k 1 k 3 k 1 Sdiff (Ribulose for ribose) k 4 k 1 k 3 k 3 Sdiff (Ribulose for arabinose) k 4 k 1 k 3 (28S) (29S) (30S) (31S) (32S) 3. Percentage of an epimeric aldo-pentose formed via direct epimerization of an aldo-pentose and isomerization of a ketose Xylose as a substrate R(Xylose into lyxose) %(Lyxose from xylose) R(Xylose into lyxose) R(Xylulose into lyxose) k 7 [Xylose] k7[xylose] k6[xylulose] k 6 [Xylulose] %(Lyxose from xylulose) k 7 [Xylose] k 6 [Xylulose] (33S) (34S) Lyxose as a substrate k %(Xylose from lyxose) 7 [Lyxose] k 7 [Lyxose] k 5 [Xylulose] k 5 [Xylulose] %(Xylose from xylulose) k 7 [Lyxose] k 5 [Xylulose] (35S) (36S) Arabinose as a substrate k %(Ribose from arabinose) 2 [Arabinose] k 2 [Arabinose] k 1 [Ribulose] k %(Ribose from ribulose) 1 [Ribulose] k 2[Arabinose] k 1[Ribulose] (37S) (38S) Ribose as a substrate k2[ribose] %(Arabinose from ribose) k2[ribose] k 3[Ribulose] k %(Arabinose from ribulose) 3[Ribulose] k2[ribose] k 3[Ribulose] (39S) (40S) S21

22 Fig. 10S. Concentration profiles of xylulose formation based on lyxose (left) and ribulose formation based on ribose (right) at various temperatures. S22

23 Calculation of amount of extracted monosaccharides and selectivity of extraction In order to optimize the conditions of the anionic extraction, the solutions after reactions in the presence of NaH 2 P 4 +Na 2 HP 4 were utilized. In addition to the substrate and the ketoses, the solutions contained minor amounts of other saccharides (Table 7S): Table 7S. Concentrations of the saccharides in the solutions used for optimization of the extraction conditions. Separation of aldoseketose [Lyxose], M [Xylose], M [Ribose], M [Arabinose], M [Ribulose], M [Xylulose], M Lyxosexylulose Riboseribulose Percentages of the extracted components and selectivity of extraction shown in Table 3 were calculated as follows: Percentage of the extracted component% c 0 i ci Selectivity, % 100% c 0 i ci i c 0 c i i after extraction, respectively. c 0 i ci 100% c 0 i Where and correspond to concentrations of the saccharides in aqueous phase before and (41S) (42S) S23

24 Fig. 11S Results of ESI-MS analysis in negative mode of the organic phase after anionic extraction of ribulose (up) and xylulose (down) with HMPBA. S24

25 Fig. 12S. HSQC NMR spectra of the extracted complexes of keto-pentoses with HMBPA S25

26 Table 8S. Chemical shifts (in ppm) observed in 13 C NMR spectra for complexes between ribulose and HMPBA as well as xylulose and HMPBA (CDCl 3 ) and their comparison with α-ribulose and β-xylulose (D 2 ). C1 C2 a C3 C4 C5 CH 2 -HMPBA b β-xylulose HMPBA complex β-xylulose Δ δ 13 C c α-ribulose HMPBA complex α-ribulose Δ δ 13 C c a 13 C NMR signals of the quaternary C2 atoms were not detected; b chemical shifts of the hydroxymethyl rest of HMPBA ligand; c 13 C chemical shift differences for a complex and a corresponding saccharide. S26

27 Table 9S. Results of extraction-assisted isomerization of lyxose into xylulose. Reaction conditions: isomerization over 0.5M NaH 2 P 4 +Na 2 HP 4, 89 C, 20 min, 750 rpm; extraction was conducted at room temperature for 1 h, 750 rpm using 0.1 M HMPBA and 0.1 M Aliquat 336 in 1-octanol as an organic phase, reaction mixture after the isomerization was used as an aqueous phase. Run Isomerization Extraction Concentration of saccharides Extracted amount, % Selectivity of extraction, % [Lyx] 0, a M [Xylu] 0, a M [Ribu] 0, a M [Lyx] fin, b M [Xylu] fin, b M [Ribu] fin, b M Lyx conversion, % Lyx Xylu Ribu Lyx Xylu Ribu < < < <1 37 a [Lyx] 0, [Xylu] 0, and [Ribu] 0 are concentrations of lyxose, xylulose, and ribulose in the reaction mixture prior to the isomerization b [Lyx] fin, [Xylu] fin, and [Ribu] fin, are concentrations of lyxose, xylulose, and ribulose, in the reaction mixture after the isomerization Yield of xylulose, % Calculation of conversion, selectivity, and yield for the reaction-assisted isomerization Lyxose conversion was calculated as follows ([Lyx] 0, i [Lyx]fin,i) Lyxose conversion i ([Lyx] 0, i [Lyx] extr, i 1) i Xylulose yield was calculated for the total amount of formed xylulose according to the formula: ([Xylu] fin,i [Xylu] 0,i ) Xylulose yield i ([Lyx] 0, i [Lyx] extr, i 1 ) i (43S) (44S) Used for Equations 43S and 44S parameter i denotes a number of run, whereas [Lyx] 0, [Xylu] 0, [Lyx] fin, [Xylu] fin, denote concentrations of lyxose and xylulose provided in Table 9S. [Lyx] extr, and [Xylu] extr denote concentrations of lyxose and xylulose in aqueous phase after extraction. S27

28 References 1. Y.B. Tewari, D.K. Steckler and R.N. Goldberg, Biophys. Chem., 1985, 22, Y.B. Tewari and R.N. Goldberg, Biophys. Chem., 1985, 22, I.V. Delidovich, A.N. Simonov,.P. Pestunova and V.N. Parmon, Kinet. Catal., 2009, 50, L. Petruš, M. Petrušová and Z. Hricovíniová, The Bílik Reaction. In Glycoscience, Springer Berlin / Heidelberg: 2001; Vol. 215, pp J.B. Lambert, G. Lu, S.R. Singer and V.M. Kolb, J. Am. Chem. Soc., 2004, 126, D. Murzin and T. Salmi, Chapter 4 - Complex reactions. In Catalytic Kinetics, Elsevier Science: Amsterdam, 2005; pp 111. S28