Supporting Information for. Discovery of Multicomponent Heterogeneous Catalysts via Admixture Screening:

Size: px

Start display at page:

Download "Supporting Information for. Discovery of Multicomponent Heterogeneous Catalysts via Admixture Screening:"

Percival Pope
5 years ago
Views:

1 Supporting Information for Discovery of Multicomponent Heterogeneous Catalysts via Admixture Screening: PdBiTe Catalysts for Aerobic xidative Esterification of Primary Alcohols David S. Mannel a, Maaz S. Ahmed b, Thatcher W. Root a, *, and Shannon S. Stahl b, * a Department of Chemical and Biological Engineering b Department of Chemistry University of Wisconsin-Madison, Madison Wisconsin, 5706 * thatcher@engr.wisc.edu, stahl@chem.wisc.edu Table of Contents 1) Response-Surface Methodology Data for PdBiTe Catalyst ptimization ) rigin of Catalyst Poisoning by 1-ctanol in the Flow Reactor ) Reaction Dependence on K 2 C ) Catalyst Stability in Batch and Flow ) SEM-EDX of PBT-1 and PBT-2 Catalysts ) ptimization of oxygen flow-rate and pressure in the continuous flow reactor ) Leaching of Pd, Bi, and Te from 5 day continuous flow run ) xidation of 1-ctanol with Single Promoter Catalysts ) Base Dependence with 1-ctanol in Flow ) Experimental Data for Arrhenius Plot S1

2 1) Response-Surface Methodology Data for PdBiTe Catalyst ptimization. Scheme S1. General reaction conditions for the oxidative methyl esterification of 1-octanol by PdBiTe catalysts. 1 M H 1 mol % Pd (PdBi x Te y /C, 5 wt. % Pd) 25 mol % K 2 C, 60 C 1 bar 2, MeH, 12 h Me In the development of catalysts with defined catalyst compositions, paths of steepest ascent were followed from a starting point of Pd 1 Bi 1 Te 1 /C towards an optimum catalyst. Data from two independent efforts (Paths A and B) are presented below. All experiments were performed under the conditions in Scheme S1. Path A (Figure S1) was carried out by taking constant step sizes of 0.1 unit, and path B (Figure S2) was carried out by taking larger step sizes of approximately half the distance to an axis. Similar paths were taken for both with the path of steepest ascent first decreasing x Bi of approx. 0., then decreasing x Te to approx Both paths ended at a region of high activity (Figure S4) from which PdBi 0.5 Te 0.2 /C (PBT-2) and PdBi 0.47 Te 0.09 /C (PBT-1) were chosen. S2

Experimental Yields from Each Step (Path A): 1) 2) ) Calculated Path of Steepest Ascent (Path A) 1) 2) ) Experimental Yields from Each Step (Path A) (cont'd): 4) 5) Calculated Path of Steepest Ascent

3 Experimental Yields from Each Step (Path A): 1) 2) ) Calculated Path of Steepest Ascent (Path A) 1) 2) ) Experimental Yields from Each Step (Path A) (cont'd): 4) 5) Calculated Path of Steepest Ascent (Path A) (cont'd) 4) 5) Figure S1. Experimental yields (left) for catalysts along path A and the calculated path of steepest ascent (right). Reaction conditions given in Scheme S1. Color axis is yield. S

4 Experimental Yields from Each Step (Path B): 1) 2) ) Calculated Path of Steepest Ascent (Path B) 1) Experimental Yield and Calculated Path of Steepest Ascent (Path B) (cont'd): 4) Figure S2. Experimental yields (left) for catalysts along path B and the calculated path of steepest ascent (right). Reaction conditions given in Scheme S1. Color axis is yield. Figure S. Paths of steepest ascent followed for Path A and Path B. S4

5 2) rigin of Catalyst Poisoning by 1-ctanol in the Flow Reactor. Catalyst poisoning under flow reaction conditions was manifested as a limit to the maximum attainable conversion of 1-octanol. Recycling of reactor effluent back through the reactor showed no increase in concentration of methyl octanoate (Scheme S2). In addition, spiking the feed solution with methyl octanoate or octanoic acid yielded 80% conversion of 1-octanol to methyl octanoate, showing that neither of these species poisons the catalyst (Scheme S). Scheme S2. Recycling of Reactor Effluent. C 7 H 15 H 0.25M PBT-2 (40 mg) MeH, 60 C, 0.1 ml/min (WHSV = 9.6 h -1 ) 0.56 bar 2 (9 % 2 in N 2 ) 8:1 2 :sub 20 % mesitylene as I.S. 25 mol % K 2 C C 7 H 15 Me 0.2 M + C 7 H MH PBT-2 (40 mg) MeH, 60 C, 0.1 ml/min (WHSV = 9.6 h -1 ) 0.56 bar 2 (9 % 2 in N 2 ) 8:1 2 :sub 20 % mesitylene as I.S. 25 mol % K 2 C C 7 H 15 Me 0.2 M + C 7 H MH Scheme S. Reaction Conversion with Added Methyl ctanoate or ctanoic Acid. C 7 H 15 H M PBT-2 (40 mg) + C 7 H 15 H + C 7 H 15 MeH, 60 C, 0.1 ml/min liquid M M 0.56 bar 2 (9 % 2 in N 2 ) 8:1 2 :sub 20 % mesitylene as I.S., M K 2 C (85 % conv.) C 7 H 15 Me 0.25 M C 7 H 15 H M PBT-2 (40 mg) + C 7 H 15 H + C 7 H 15 H MeH, 60 C, 0.1 ml/min liquid 0.04 M M 0.56 bar 2 (9 % 2 in N 2 ) 8:1 2 :sub 20 % mesitylene as I.S., M K 2 C (8 % conv.) + C 7 H 15 H M C 7 H 15 Me 0.17 M In contrast, use of octyl aldehyde as the starting material afforded only 50% yield of methyl octanoate at 100% conversion of octyl aldehyde. A fresh sample of 1-octanol was added to the reactor effluent and the mixture was re-fed through the reactor under standard reaction conditions. This run afforded only 7.5% conversion of the alcohol, indicating that a catalyst S5

6 poison is generated during the oxidation of octyl aldehyde (Scheme S4). The missing mass from the initial reaction of octyl aldehyde was identified as 2-hexyl-2-decenal and methyl 2-hexyl-2- decenoate by GC-MS, 1 H NMR, and 1 C NMR, reflecting aldol self-condensation of octyl aldehyde. Scheme S4. Catalyst Poisoning After Reaction of ctyl Aldehyde. C 7 H M PBT-2 (40 mg) MeH, 60 C, 0.1 ml/min liquid 0.56 bar 2 (9 % 2 in N 2 ) 8:1 2 :sub 20 % mesitylene as I.S., M K 2 C C 7 H 15 Me 0.12 M (100 % Conv) M 1-octanol PBT-2 (40 mg) MeH, 60 C, 0.1 ml/min liquid 0.56 bar 2 (9 % 2 in N 2 ) 8:1 2 :sub 20 % mesitylene as I.S., M K 2 C C 7 H 15 H 0.2 M (7.5 % conv) + C 7 H 15 Me 0.1 M 2-Hexyl-2-decenal (aldol condensation product) was independently synthesized (Scheme S5) to confirm the identity of the catalyst poison. Inclusion of the aldol condensation species in a batch oxidation reaction of 1-octanol resulted in significant reduction in the yield and conversion of the reaction (Figure S4). Scheme S5. Synthesis of 2-hexyl-decenal and ligomers from ctyl Aldehyde. 200 µl 1 M NaH in H ml 20 mg PBT-2 80 C, 22 hours + oligomers >71 % conv. S6

7 Figure S4. Reaction poisoning by aldol condensation products in the batch reaction aerobic methyl esterification of 1-octanol. 1 M 1-octanol, 1 mol % Pd (PdBi 0.47 Te 0.09 /C, 5 wt. % Pd), 60 C, 1 bar 2, MeH, 1 ml total volume, 20 mol % Mesitylene. 0 or 10 µl aldol condensation products added. ) Reaction Dependence on K 2 C. Batch reactions with different concentrations of K 2 C revealed that the yield maximized at 25 mol % K 2 C (Figure S5). 1 M H 1 mol % PBT-2 25 mol % K 2 C, 60 C 1 bar 2, MeH, 24 h Me Figure S5. Yield of methyl octanoate obtained at various [K 2 C ] showing maximum yield being reached at 0.25 M. S7

8 4) Catalyst Stability in Batch and Flow. To assess the activity and long-term stability of the two PBT catalysts, a continuous flow run was performed at different WHSV s and over an extended reaction time at a constant WHSV (Figure S6A and B). The PBT-2 catalyst has higher activity for the oxidation of benzyl alcohol in the flow reactor with quantitative yields obtained at a WHSV of 500 h -1 (1/WHSV = 7.2 s). The catalyst stability was measured by holding the reactor at a constant elevated WHSV of 700 h -1 (1/WHSV ~ 5 s), and the product yield was monitored over an extended time period. After h and 22,000 turnovers, the yield of methyl benzoate produced from the reaction with PBT-1 decreased from 7 to 50%. In contrast, the reaction with PBT-2 showed no loss in activity after 4 h and 27,000 turnovers. Therefore, PBT-2 was chosen for the demonstration of a 5 day continuous process. 0.5 M H PdBiTe 25 mol % K 2 C, MeH, 60 C 0.54 bar 2 (9% 2 in N 2 ) 8:1 mol 2 :mol substrate Me A. B. Figure S6. Continuous flow oxidation of benzyl alcohol to methyl benzoate by PdBiTe catalysts. (A) WHSV scan of PBT-1 and PBT-2 catalysts showing higher activity of PBT-2. (B) Stability of PBT-1 and PBT-2 catalysts over extended times in the flow reactor measured from yield of methyl benzoate at the highlighted WHSV (1/WHSV = 5 s). S8

9 5) SEM-EDX of PBT-1 and PBT-2 Catalysts. SEM-EDX mapping of fresh and used PdBiTe catalysts showed higher concentrations of Bi and Te on Pd (Figures S7-S9). No strong changes were observed between the fresh and used catalyst suggesting that leaching of Bi and Te was relatively uniform from both the Pd surface and the carbon support. The molar ratios of Bi and Te to Pd are in agreement between the catalyst synthesis, the SEM-EDX integrated signal, and the total metal content analyzed from ICP-AES (Table S1). Both PBT-1 and PBT-2 showed significant losses of both Bi and Te after use in the continuous flow reactor. Figure S7. SEM-EDX mapping of Pd Bi and Te on freshly prepared PBT-1. Figure S8. SEM-EDX mapping of Pd Bi and Te on freshly prepared PBT-2. Figure S9. SEM-EDX mapping of Pd, Bi and Te on PBT-2 after a 5 day continuous run. S9

10 Table S1. Elemental Composition of PBT Catalysts from Different Analysis Techniques. Catalyst Detection Method Bi:Pd Te:Pd PBT-1 (Fresh) From Catalyst Synthesis SEM-EDX PBT-1 (Used) SEM-EDX From Catalyst Synthesis PBT-2 (Fresh) SEM-EDX ICP-AES PBT-2 (Used) SEM-EDX ICP-AES Fresh Catalysts are as-prepared prior to any reactions. PBT-1 (Used) was recovered from flow reactor after undergoing 22,000 turnovers (see Figure S6B). PBT-2 (Used) was recovered from flow reactor after undergoing 58,000 turnovers (see Figure 6). 6) ptimization of oxygen flow-rate and pressure in the continuous flow reactor. With higher pressures obtainable in the flow reactor, additional screening with benzyl alcohol was done to optimize the catalysts performance. Varying the 2 pressure and the 2 to substrate ratio landed on optimum values of 0.5 bar 2 and 8 mol 2 /mol substrate (Table S2). Table S2. ptimization of p 2 and 2 Flow Rate in the xidation of Benzyl Alcohol. 0.5 M H PBT-2 (4.6 mg) WHSV =,500 h -1, MeH, 60 C x bar 2 (9% 2 in N 2 ) y:1 2 :sub 25 mol % K 2 C Me Entry mol 2 :mol substrate p 2 (bar) Yield (%) :1 4:1 4:1 8:1 8:1 8:1 16:1 16:1 16: ) Leaching of Pd, Bi, and Te from 5 day continuous flow run. Effluent concentrations of Pd, Bi, and Te were measured from different fractions of the fiveday continuous flow run (Table S). The recovered mass of Pd, Bi, and Te in combination with the quantities leached from the reactor were within 5% of the original mass of Pd, Bi, and Te. S10

11 Table S. Average effluent concentrations of Pd, Bi, and Te between 0-1, 1-10, and h, and the corresponding mass of Pd, Bi, and Te lost from original catalyst. Element Pd Bi Te Time on Stream (h) Effluent Concentration (ppb) % Lost from Initial Catalyst ) xidation of 1-ctanol with Single Promoter Catalysts xidation of single promoter catalysts (PdBi x and PdTe y ) was done to compare the effect of additional promoter amounts (Table S4). With both Bi and Te containing catalysts increased amounts of Bi or Te did not increase the reaction yield, relative to those observed with the dual promoter PBT-1 and PBT-2 catalysts. Table S4. Yield of methyl octanoate with single promoter catalysts. a Catalyst Yield (%) PdBi PdBi PdBi 5 5 PdTe PdTe PdTe a 1 M 1-octanol, 1 eq. KMe, 1 bar 2, 24 h, 60 C, MeH solvent, 1 mol % Pd. 9) Base Dependence with 1-ctanol in Flow The oxidation of 1-octanol to methyl octanoate in a continuous process shows increased yields with decreasing [K 2 C ] (Figure S10). During the screening of the [K 2 C ] in the S11

12 oxidation of 1-octanol a single packed bed reactor was used showing sustained activity of the PBT-1 catalyst over a 180 h period with varying reaction conditions. This corresponds with the PBT-1 catalyst undergoing nearly 800 turnovers. Figure S10. Aerobic oxidative methyl esterification of 1-octanol with various concentrations of K 2 C. Catalyst activity is maintained over the 180 h run. 10) Experimental Data for Arrhenius Plot The activation energy and pre-exponential factor for the oxidation of benzyl alcohol were calculated from the conversion of the substrate at different temperatures and a constant flow rate (Figure S11). Rate constants were calculated assuming first order kinetics using equation S1 (Figure S12). S12

13 Figure S11. Yield of methyl benzoate at different reaction temperatures. Reaction Conditions: 0.5 M benzyl alcohol, 4.6 mg PBT-2, 25 mol % K 2 C, 0.54 bar 2 (9% 2 in N 2 ), 8:1 mol 2 :mol substrate, WHSV =,500 h -1. k!"" = ln 1 Conversion 1/WHSV (S1) Figure S12. Time course for the aerobic oxidative methyl esterification of benzyl alcohol showing conversion of benzyl alcohol (blue circles) to benzaldehyde intermediate (red diamonds) and subsequent reaction of benzaldehyde to methyl benzoate (green triangles) over PdBiTe catalysts. Conditions: 1 M benzyl alcohol and 0.25 M K 2 C in 1 ml MeH, 0.1 mol % PBT-2, 60 C, 1 bar 2. Fits are from treating both oxidation reactions as pseudo-first order reactions. S1

Nanocatalysis in Continuous Flow: Supported Iron. Oxidation of Benzyl Alcohol

This journal is The Royal Society of Chemistry 13 1 Supporting Information Nanocatalysis in Continuous Flow: Supported Iron xide Nanoparticles for the Heterogeneous Aerobic xidation of Benzyl Alcohol David