Supporting Information: Renewable Isoprene by Sequential Hydrogenation of Itaconic Acid and Dehydra-Decyclization of 3-Methyl-Tetrahydrofuran

Transcription

1 Supporting Information: Renewable Isoprene by Sequential Hydrogenation of Itaconic Acid and Dehydra-Decyclization of 3-Methyl-Tetrahydrofuran 1, Omar A. Abdelrahman, 1,2, Dae Sung Park, 1 Katherine P. Vinter, 1,2 Charles S. Spanjers, 1 Limin Ren, 3 Hong Je Cho, 1,2 Kechun Zhang, 3 Wei Fan, 1 Michael Tsapatsis, 1,2,* Paul J. Dauenhauer 1. University of Minnesota, Department of Chemical Engineering and Materials Science, 421 Washington Ave. SE, Minneapolis, MN, U.S.A. 2. Center for Sustainable Polymers, a National Science Foundation Center for Chemical Innovation. 209 Smith Hall, 207 Pleasant Street SE, Minneapolis, MN, U.S.A. 3. University of Massachusetts Amherst, Department of Chemical Engineering, 686 North Pleasant Street, Amherst, MA U.S.A. *Corresponding Author: hauer@umn.edu Authors Contributed Equally. TABLE OF CONTENTS 1.0 Materials Catalyst Synthesis Catalyst Characterization Reaction Methods: Flow Reactor Reaction Methods: Microcatalytic reactor Reaction Methods: Itaconic Acid Hydrogenation Reaction Methods: MBDO dehydration to 3-MTHF Thermodynamic Calculations Catalytic Experimental Results: Itaconic Acid Hydrogenation Catalytic Experimental Results: 3-MTHF dehydra-decyclization Catalytic Experimental Results: Isopropanol dehydration Product Identification: Mass spectrum fragmentation patterns Experimental three-dimensional selectivity map Time-on-stream study of P-SPP catalyst 48 Page 1 / 49

2 1.0 Materials 1.1 Reagents and standard chemicals: Itaconic acid (>99 %) and 2-methyl-1,4-butanediol (MBDO, >97 %) used for reactants in aqueous-phase reaction were obtained from Sigma-Aldrich. 3-methyltetrahydrofuran (3- MTHF,>95 %, TCI America) was used in gas-phase dehydra-decylization (Microcatalytic and flow reactor) to make isoprene. Standards including isoprene (>99 %), 2-methyltetrahydrofuran (>99 %), 3-methyl-3-buten-1-ol (>97 %), 2-methyl-3-buten-1-ol (>98 %) and α-methyl-ɤbutyrolactone (>98 %) were purchased from Sigma-Aldrich. 1.2 Catalysts A variety of commercial catalysts were utilized in the vapor phase conversion of 3- MTHF including ZSM-5 zeolite (Zeolyst CBV28014, Si/Al ratio = 140), H-Y zeolite (Zeolyst CBV760, Si/Al = 30), tricalcium phosphate (TCP, Sigma Aldrich), silica alumina (SiO 2 Al 2 O 3, Sigma Aldrich), niobium oxide (Nb 2 O 5, Sigma Aldrich). For the aqueous production of 3-MTHF from IA, 10 wt% Pd/C (Sigma Aldrich), 5wt% Ru/C (Sigma Aldrich), 5 wt% Pd/SiO 2 (Strem Chemicals), 5 wt% Ru/Al 2 O 3 (Sigma Aldrich) and Amberlyst-15 (Sigma Aldrich) were employed. In addition to commercially available catalysts, Sn-BEA, Pd-Re/C and various phosphorous containing zeolites were prepared in house. Details of the preparation methods are provided in subsequent sections. Page 2 / 49

3 2.0 Catalyst Synthesis 2.1 Pd-Re/C The carbon supported Pd-Re catalysts were prepared using incipient-wetness impregnation, where an incipient volume of 2 ml/g of catalyst was employed. Briefly, an aqueous solution of ammonium perrhenate (NH 4 ReO 4, Sigma Aldrich) was used to impregnate a commercial Pd/C catalyst. Once impregnated, the catalysts were dried at 100 C for 12 hr and reduced at 400 C in 10 % H 2 /Ar for 3 hr. 2.2 P-BEA P-BEA was synthesized according to the literature [31]. Commercial zeolite Al-BEA (Zeolyst CP814E, Si/Al = 12.5) was dealuminated by treatment with 70 wt % nitric acid (HNO 3, Fisher Scientific). Typically, 0.5 g of the Al-BEA was mixed with 25 ml of 70 wt % HNO 3 in a Teflonlined stainless steel autoclave. The autoclave was then put into an oven at 80 C for 24 h under a static condition. The dealuminated zeolite BEA (DeAl-BEA) was washed extensively with deionized water and dried overnight at 100 C. In order to prepare P-BEA with Si/P = 27 (confirmed by ICP analysis), wet impregnation was performed by stirring 0.4 g of DeAl-BEA and 18.2 µl of 85 wt% phosphoric acid (H 3 PO 4, Sigma-Aldrich) in 3.33 ml of deionized water. The impregnated sample was dried at 90 C overnight, followed by calcination in a tube furnace with dry air at 600 C for 25 min Page 3 / 49

4 2.3 P-Al-BEA P-Al-BEA was prepared by impregnation of H 3 PO 4 on zeolite Al-BEA (Zeolyst, CP814E, Si/Al = 12.5), according to the same procedure described in the preparation of P-BEA, without the dealumination step. 2.4 P-SPP P-SPP with Si/P = 27 (confirmed by ICP analysis) was synthesized according to the literature [31]. Typically, tetra(n-butyl) phosphonium hydroxide (TBPOH, 40 wt %, Sigma- Aldrich) as a structure-directing agent (SDA) was added dropwise into tetraethylorthosilicate (TEOS, 98%, Sigma-Aldrich) under stirring. Deionized water was then added to this mixture, and stirred for 24 h. The mixture became a clear sol with a composition of 1 SiO 2 : 0.3 TBPOH : 10 H 2 O : 4 EtOH. The sol was sealed in a Teflon lined stainless steel autoclave and heated for 3 days in an oven at 115 C. After crystallization, the solid product was extensively washed with deionized water by centrifugation and decanting of the supernatant. This process was repeated until the ph of the final supernatant was lower than 9. Subsequently, the collected sample was dried at 90 C overnight and calcined in a tube furnace at 550 C for 12 h under dry air. 2.5 P-MFI First, pure silica zeolite Si-MFI with a composition of 1 SiO 2 : 0.26 TPAOH : 15 H 2 O : 4 EtOH was synthesized by mixing structure directing agent tetrapropylammonium hydroxide solution (TPAOH, 40 wt%, SACHEM), water and tetraethylorthosilicate (TEOS, 98%, Sigma- Aldrich). The mixed gel was sealed in a Teflon lined stainless steel autoclave and heated for 3 days in an oven at 180 C. The solid products were centrifuged, washed with distilled water Page 4 / 49

5 (until ph<8) and then dried at 70 C overnight and calcined at 550 C for 6 h in air under static conditions. In order to prepare P-MFI with similar P content as P-BEA, the same wet impregnation procedure was performed for P-MFI as in preparation of P-BEA. First, 0.4 g of Si- MFI, 18.2 µl of 85wt % phosphoric acid (H 3 PO 4, Sigma-Aldrich) and 3.33 ml of deionized water were mixed. Then, the impregnated sample was dried at 90 C overnight, followed by calcination in a tube furnace with dry air at 600 C for 25 min. 2.6 Si-SPP Si-SPP without P was synthesized by modifying P-SPP synthesis method. First, P-SPP was synthesized according to the above procedures. Then, the calcined P-SPP solid products were dispersed in water in a centrifuge tube. After the centrifuge tube was heated in a 70 C oven for 1 h, the solid products were separated by centrifugation. The dispersion and centrifugation procedures were repeated until the suspension ph is around 7. Then, the solid products were dried at 70 C oven overnight and calcined at 500 C for 4 h in air under static conditions. For a better removal of P, the calcined products were dispersed in water and washed by centrifugation for another 3 times. The final Si-SPP solid products were dried at 70 C oven overnight and calcined at 500 C for 4 h in air under static conditions. 2.7 Pelletization procedure In order to maintain a relatively uniform particle distribution and avoid excessive pressure drops, catalyst powders utilized in the microcatalytic and packed bed reactors were pelletized to achieve a particle diameter of µm. Briefly, mg of catalyst powder were placed into a 13 mm pellet press (Pike technologies) and pressed to 2 tons of pressure, holding for mins. The catalyst pellet was then extracted and broken up into smaller Page 5 / 49

6 particles, which were then passed through a set of sieves from which the µm sieve cut was collected and used for catalytic testing throughout the study. 3.0 Catalyst Characterization Textural information of the various catalysts employed in this study was characterized through Ar physisorption in an Autosorb iq2 porosimetry instrument (Quantachrom). Prior to analysis, catalysts were outgassed at 573 K for 6 hours and subsequently cooled down to room temperature under vacuum. BET specific surface area measurements were used to represent the total surface area of the catalyst materials; total pore volume was determined using a single point measurement at P/P 0 = Brønsted acid site concentration (H + ) was determined by isopropylamine temperature programmed desorption (IPA-TPD). Catalyst samples were dehydrated at 400 C under a stream of He for one hour then allowed to cool down to 120 C where the catalyst was contacted with IPA for 15 minutes. The catalyst was then held under a stream of He for one hour to remove any physisorbed IPA, after which the temperature was ramped at a rate of 10 C/min to 700 C. IPA-TPD was not conducted for the phosphorous containing materials, since earlier studies with P-containing materials did not result in any detectable acidity [30]. Table S1. Catalyst Characteristics. Catalyst Surface Area (m 2 g -1 ) Pore Volume (cm 3 g -1 ) Si:Al a H + (µmol g -1 ) ZSM H-Y SiAl Sn-BEA b Nb 2 O a- Provided by the manufacturer b- Cannot be determined by existing methods Page 6 / 49

7 Table S2. Phosphorous-containing Catalyst Characteristics. Catalyst Surface Area (m 2 g -1 ) Pore Volume (cm 3 g -1 ) Si:P a P-SPP P-BEA P-MFI P-Al-BEA Si-SPP a- Determined by ICP-MS 3.1 ICP-MS Elemental analysis was performed on inductively coupled plasma optical emission spectroscopy (ICP-OES, icap 6500 Dual view, Thermo Scientific) in Analytical Geochemistry Lab, Department of Earth Sciences in University of Minnesota. 4.0 Reaction Methods: Flow reactor The steady state measurements of the dehydra-decyclization of 3-MTHF to isoprene over supported acid catalysts was performed in a downflow packed bed reactor. Liquid 3-MTHF was fed into a vaporization section through a 1/16 PEEK capillary line (0.01 ID), where a syringe pump (KDS-100, KD Scientific) equipped with a gas tight syringe (Hamilton Company) controlled the liquid flow rate (50 µl/hr). The vaporization section consisted of an insulated aluminum tube wrapped with a heating tape, inside where the liquid feed was vaporized and swept with a helium carrier gas regulated by a mass flow controller (Brooks, 5850S). Helium flow rates were adjusted according to desired reaction conditions anywhere between sccm. The vaporized stream of 3-MTHF in helium was then passed through a switching section, consisting of a heated 6-port valve (Vici Valco, DL6UWE), from which the stream could be sent either directly to the analysis section or through the reactor first. The carbon balance across the packed catalyst bed was calculated as per Eq. (S1) Page 7 / 49

8 % = 100 (S1) Analysis of the vapor stream as it by-passed the reactor and sent directly to the analysis section, allowed the verification of the molar flow rate carbon in to the reactor. A direct measure of the molar flowrate of carbon was possible given the quantitative carbon detector employed in the analysis. Details of the possible flow patterns are illustrated in Figure S1. The reactor consisted of a ½ 316 SS tube packed with quartz chips and catalyst operated in a down flow mode at 1.1 bar, where the mass of catalyst loaded varied between mg. The pressure drop across the catalyst bed was maintained below 5% of the total pressure, achieved by maintaining the catalyst particle diameter between µm. The catalyst bed was placed between two plugs of inert quartz wool, above which the void volume was reduced using inert quartz packing in the range of µm (Sigma Aldrich). Temperature measurements were made using a 1/16 type-k thermocouple (Omega) placed directly below the catalyst bed, while a second thermocouple was placed in the furnace used to control the temperature using a PID temperature controller (OMEGA,CN-7800). Prior to introducing any reactant, the catalyst bed was heated at 3 K min -1 to 673 K and held for 1 hour under a 50 sccm stream of air. The catalyst bed was then cooled down to the desired reaction temperature and flushed with a stream of He prior to introducing the vapor phase reactant stream. Analysis of the vapor phase products was performed using an online gas sampling gas chromatograph (Agilent, 7890A) equipped with a quantitative carbon detector (QCD, POLYARC) in conjunction with a flame ionization detector (FID). Separation of the various products and reactant in the GC were performed using an HP-PLOT Q column (Agilent, 19091P-QO4). Page 8 / 49

9 Figure S1. Schematic of the Packed Bed Catalyst Flow Reactor. (a) Reaction Configuration. (b) Reactor bypass configuration. Page 9 / 49

10 5.0 Reaction Methods: Microcatalytic reactor Reactions to screen for catalysts and operating conditions were performed in a microcatalytic reactor. The reactor consists of a glass reactor (i.e., inlet liner) placed in the inlet of a gas chromatograph (Aglient 7890A). The reactor was packed with quartz wool below (7.5 mg) and above (15 mg) the catalyst (~45 mg). A diagram of the reactor is shown in Figure S2 below: Figure S2. Diagram of high throughput pulsed flow reactor (HTPFR) implementing the microcatalytic method. Experiments were performed at reaction temperatures of 200 to 400 C. The space velocity was controlled by adjusting the carrier gas (He, %) flow rate to the split vent. Space velocities of s -1 were tested with the reactor, where space velocity is defined as follows: = = = (S2) where F He is the flow rate of He through in sccm min -1 and V C is the volume of the catalyst bed. Helium pressure was kept constant at 30 psi. Each experiment was performed by injecting 1 ul Page 10 / 49

11 of pure 3-MTHF into the reactor followed by immediate separation and quantification of the products. Reaction products were separated using a chromatographic column (Agilent, HP-Plot Q, 30 m, 0.32 mm ID, 20 µm film thickness; temperature program: 40 ºC for 2 min, 10 ºC/min to 270 ºC, hold 10 min) followed by quantification using a quantitative carbon detector (Polyarc/FID). 6.0 Reaction Methods: Itaconic Acid hydrogenation Sequential hydrogenation and dehydration of itaconic acid (IA) for the production of 3- MTHF was performed in a 100 ml high pressure Parr Reactor (model 4598HPHT, Parr Instrument Co.). Typically, 1.5 g (11.5 ml) of IA was added to 28.5 ml of water, and g of catalyst was introduced to the mixture. The reactor was purged with N 2 twice to remove any residual air in the reactor. The reactor was then pressurized to psig with H 2 (at room temperature) and heated to reaction temperature ( C) under vigorous stirring (1,000 rpm). Products were identified via GC-MS (Agilent 7890A connected with Triple-Axis MS detector, Agilent 5975C) and quantified using a liquid sampling GC (Agilent 7890A) equipped with a HP-Plot Q column and a QCD/FID combination. 7.0 Reaction Methods: MBDO dehydration to 3-MTHF Dehydration of MBDO to 3-MTHF was performed in a 100 ml high pressure Parr reactor. 1 ml (28.6 mmol) of MBDO was added to 29 ml of water, and 0.2 g of catalyst was added to the mixture. The reactor was purged with N 2 twice and pressurized to 200 psi to keep reactants in the condensed phase. Reactions were performed at temperatures ranging from o C at a stirring rate of 1000 rpm. Products were identified via GC-MS (Agilent 7890A connected with a Triple-Axis MS detector, Agilent 5975C) and quantified using a liquid sampling GC (Agilent 7890A) equipped with a HP-Plot Q column and a QCD/FID combination. Page 11 / 49

12 8.0 Thermodynamic Calculations Aqueous phase thermodynamic calculations were performed with the Gaussian 09 (Rev C.1) program [32] using the M062X/ G(3df,3pd) level of theory and the SMD model [33] with water as the solvent for liquid phase calculations and the CBS-QB3 Complete Basis Set method for calculations in the vapor phase. Vibrational frequencies, rotational temperature, electronic energy, among other values calculated by Gaussian for each molecule are used to calculate first, partition functions and finally thermochemical data resulting in Gibbs Free energies. Each calculation considered contributions from translational, rotational, electric, and vibrational energies to determine the overall partition function. All molecules are assumed to be non-interacting, ideal gases. Theory and guidelines of this process can be found in the cited resource by Ochterski [34]. The CBS-QB3 method was chosen due to its agreement (+ 1 kcal/mole), in most cases, to reported experimental data from NIST. The exception was found with the reactions forming formaldehyde and propene. In this case, data reported by NIST was included instead. The Gibbs free energies of reaction were calculated by the following equation: = (S3) in which G i, G j correspond to Gibbs free energies calculated using Gaussian data and ν i, ν j represent stoichiometric values for the given compounds in the chemical reaction of interest. Another ability inherent in these calculations was to be able to examine sensitivity of G to varying experimental temperatures and pressures. However, trends of G with respect to Page 12 / 49

13 pressure have been largely omitted from this work. This has been justified due to the negligible change in G in the ranges of pressures feasible to the reaction. Instead, focus has been directed at temperature dependence. Table S3. Calculated Free Energies of the Hydrogenation Products of Itaconic Acid. Temperature (K) MGBL (Kcal mol -1 ) MBDO (Kcal mol -1 ) 3-MTHF (Kcal mol -1 ) Isoprene a (Kcal mol -1 ) a- Vapor phase free energy relative to that of 3-MTHF Page 13 / 49

14 9.0 Catalytic Experimental Results: Itaconic Acid Hydrogenation Table S4. Summary of the sequential hydrogenation-dehydration of itaconic acid (IA) to 3- methyltetrahydrofuran (3-MTHF) over Ru-based catalysts. Catalyst Temperature ( C) P Hydrogen (psig) Conversion (%) Selectivity (%) Carbon Balance IA 3-MTHF MBDO MGBL Others 1 (%) Ru/C (0.3g) 100 1, Ru/C (0.3g) 120 1, Ru/C (0.1g) Ru/Al 2 O 3 (0.3g) 100 1, Ru/Al 2 O 3 (0.1g) Ru/Al 2 O 3 (0.05g) Ru/C (0.3g) + SiO 2 -Al 2 O 3 (0.2g) Ru/C (0.3g) + Amberlyst-15 (0.2g) Ru/C (0.1g) + SiO 2 -Al 2 O 3 (0.2g) Ru/Al 2 O 3 (0.05g) + SiO 2 -Al 2 O 3 (0.2g) 100 1, , Reaction conditions: 1.5 g (11.5 mmol) of IA in D.I. water (28.5 ml), H 2 pressurized at room temperature) 1 Others: 2-methyltetrahydrofuran, 2(or 3)-methyl-3-buten-1-ol and some unknown products. Page 14 / 49

15 Table S5. Summary of the sequential hydrogenation-dehydration of itaconic acid (IA) to 3- methyltetrahydrofuran (3-MTHF) over Pd-based catalysts. Catalysts Temperature ( C) P Hydrogen (psig) Conversion (%) Selectivity (%) IA 3-MTHF MBO MBDO MGBL Others 1 Carbon Balance (%) Pd/C (0.3g) + Ru/C (0.2g) Pd/C (0.3g) + SiO 2 -Al 2 O 3 (0.2g) Pd/C (0.3g) + SiO 2 -Al 2 O 3 (0.2g) Pd/ SiO 2 (0.3g) + SiO 2 -Al 2 O 3 (0.2g) 10Pd-10Re/C (0.2g) 10Pd-5Re/C (0.2g) 10Pd-10Re/C (0.2g) 10Pd-10Re/C (0.1g) 10Pd-10Re/C (0.2g) 120 1, , , , , , , , Reaction conditions: 1.5 g (11.5 mmol) of IA in D.I. water (28.5 ml), H 2 pressurized at room temperature) 1 Others: 2-methyltetrahydrofuran, 2(or 3)-methyl-3-buten-1-ol and some unknown products. Page 15 / 49

16 Figure S3. Yield to 3-MTHF from MBDO over various solid acid catalysts in the aqueous phase at o C under 200 psig of N 2 after 24 hrs. Page 16 / 49

17 Figure S4. Results for the Sn-BEA catalyzed dehydration of MBDO to 3-MTHF. A) yield to 3- MTHF at C and B) Conversion of MBDO and selectivity to 3-MTHF at 200 C. Page 17 / 49

18 Figure S5. Comparison of various Ru based strategies for the sequential hydrogenation and dehydration of IA to 3-MTHF at 100 C under 1000 psi of H 2 (Others: 2-methyl-tetrahydrofuran, 2-methyl-3-buten-1-ol, 3-methyl-3-buten-1-ol and unknown products). Reaction conditions: 1.5 g (11.5 mmol) of IA in D.I. water (28.5 ml), Ru/C and Ru/Al 2 O g, SiO 2 -Al 2 O 3 and Amberlyst g. Page 18 / 49

19 Figure S6. Comparison of various Ru based strategies for the sequential hydrogenation and dehydration of IA to 3-MTHF at 150 and 200 o C under 500 psi of H 2 (Others: 2-methyltetrahydrofuran, 2-methyl-3-buten-1-ol, 3-methyl-3-buten-1-ol and unknown products). Reaction conditions: 1.5 g (11.5 mmol) of IA in D.I. water (28.5 ml), Ru/C and Ru/Al 2 O g, SiO 2 - Al 2 O g. Page 19 / 49

20 Figure S7. Comparison of various Pd based strategies for the sequential hydrogenation and dehydration of IA to 3-MTHF at 160, 180 and 200 o C under 1000 psi of H 2 (Others: 2-methyltetrahydrofuran, 2-methyl-3-buten-1-ol, 3-methyl-3-buten-1-ol and unknown products). Reaction conditions: 1.5 g (11.5 mmol) of IA in D.I. water (28.5 ml), Pd/C, Pd/SiO 2, Pd-Re/C 0.2 g, SiO 2 -Al 2 O g. Page 20 / 49

21 10.0 Catalytic Experimental Results: 3-MTHF dehydra-decylization Table S6. Microcatalytic summary of 3-MTHF dehydra-decyclization to isoprene over P-SPP presented in Figure 3. T( o C) SV (s -1 ) Conversion (%) Selectivity (%) Isoprene C 3 C 4 1,4 pentadiene 1,3 pentadiene (E,Z) Page 21 / 49

22 T( o C) SV (s -1 ) Conversion (%) Selectivity (%) Isoprene C 3 C 4 1,4 pentadiene 1,3 pentadiene (E,Z) Page 22 / 49

23 T( o C) SV (s -1 ) Conversion (%) Selectivity (%) Isoprene C 3 C 4 1,4 pentadiene 1,3 pentadiene (E,Z) Page 23 / 49

24 Figure S8. Distribution of produce C 5 dienes including isoprene (red, ), 1,4 Pentadiene (blue, ) and 1,3 Pentadiene (Cis and Trans, green, ) at conditions presented in Tables S6A-S6C. Page 24 / 49

25 Table S7. Summary of the dehydra-decylization of 3-methyltetrahydrofuran (3-MTHF) to isoprene over Nb 2 O 5 using the microcatalytic method. Temperature ( C) Space velocity (s -1 ) Conversion (%) Selectivity (%) C3 1 C4 2 Isoprene Pentadiene C3: Propenes 2 C4: Butenes and Butadienes 3 Pentadienes: 1,3-pentadiene and 1,4-pentadiene Page 25 / 49

26 Table S8. Summary of the dehydra-decylization of 3-methyltetrahydrofuran (3-MTHF) to isoprene over H-Y using the microcatalytic method. Temperature ( C) Space velocity (s -1 ) Conversion (%) Selectivity (%) C3 1 C4 2 Isoprene Pentadiene C3: Propenes 2 C4: Butenes and Butadienes 3 Pentadienes: 1,3-pentadiene and 1,4-pentadiene Page 26 / 49

27 Table S9. Summary of the dehydra-decylization of 3-methyltetrahydrofuran (3-MTHF) to isoprene over ZSM-5 using the microcatalytic method. Temperature ( C) Space velocity (s -1 ) Conversion (%) Selectivity (%) C3 1 C4 2 Isoprene Pentadiene C3: Propenes 2 C4: Butenes and Butadienes 3 Pentadienes: 1,3-pentadiene and 1,4-pentadiene Page 27 / 49

28 Table S10. Summary of the dehydra-decylization of 3-methyltetrahydrofuran (3-MTHF) to isoprene over Sn-BEA using the microcatalytic method. Temperature ( C) Space velocity (s -1 ) Conversion (%) Selectivity (%) C3 1 C4 2 Isoprene Pentadiene C3: Propenes 2 C4: Butenes and Butadienes 3 Pentadienes: 1,3-pentadiene and 1,4-pentadiene Page 28 / 49

29 Table S11. Summary of the dehydra-decylization of 3-methyltetrahydrofuran (3-MTHF) to isoprene over SiO 2 -Al 2 O 3 using the microcatalytic method. Temperature ( C) Space velocity (s -1 ) Conversion (%) Selectivity (%) C3 1 C4 2 Isoprene Pentadiene C3: Propenes 2 C4: Butenes and Butadienes 3 Pentadienes: 1,3-pentadiene and 1,4-pentadiene Page 29 / 49

30 Table S12. Summary of the dehydra-decylization of 3-methyltetrahydrofuran (3-MTHF) to isoprene over P-SPP using the microcatalytic method. Temperature ( C) Space velocity (S -1 ) Conversion (%) Selectivity (%) C3 1 C4 2 Isoprene Pentadiene C3: Propenes 2 C4: Butenes and Butadienes 3 Pentadienes: 1,3-pentadiene and 1,4-pentadiene Page 30 / 49

31 Table S13. Summary of the dehydra-decylization of 3-methyltetrahydrofuran (3-MTHF) to isoprene over P-MFI using the microcatalytic method. Temperature ( C) Space velocity (s -1 ) Conversion (%) Selectivity (%) C3 1 C4 2 Isoprene Pentadiene C3: Propenes 2 C4: Butenes and Butadienes 3 Pentadienes: 1,3-pentadiene and 1,4-pentadiene Page 31 / 49

32 Table S14. Summary of the dehydra-decylization of 3-methyltetrahydrofuran (3-MTHF) to isoprene over P-BEA using the microcatalytic method. Temperature ( C) Space velocity (s -1 ) Conversion (%) Selectivity (%) C3 1 C4 2 Isoprene Pentadiene C3: Propenes 2 C4: Butenes and Butadienes 3 Pentadienes: 1,3-pentadiene and 1,4-pentadiene Page 32 / 49

33 Table S15. Summary of the dehydra-decylization of 3-methyltetrahydrofuran (3-MTHF) to isoprene over P-Al-BEA using the microcatalytic method. Temperature ( C) Space velocity (s -1 ) Conversion (%) Selectivity (%) C3 1 C4 2 Isoprene Pentadiene C3: Propenes 2 C4: Butenes and Butadienes 3 Pentadienes: 1,3-pentadiene and 1,4-pentadiene Page 33 / 49

34 Table S16. Summary of the dehydra-decylization of 3-methyltetrahydrofuran (3-MTHF) to isoprene over tricalcium phosphate using the microcatalytic method. Temperature ( C) Space velocity (s -1 ) Conversion (%) Selectivity (%) C3 1 C4 2 Isoprene Pentadiene C3: Propenes 2 C4: Butenes and Butadienes 3 Pentadienes: 1,3-pentadiene and 1,4-pentadiene Page 34 / 49

35 Table S17. Packed bed reactor summary of 3-MTHF dehydra-decyclization to isoprene over SiO 2.Al 2 O 3. T ( o C) WHSV (g 3-MTHF gcat -1 hr -1 ) P 3-MTHF (torr) Sel isoprene (%) Sel pentadiene (%) Conversion (%) Carbon Balance (%) Table S18. Packed bed reactor summary of 3-MTHF dehydra-decyclization to isoprene over P- SPP. T ( o C) WHSV (g 3-MTHF gcat -1 hr -1 ) P 3-MTHF (torr) Sel isoprene (%) Sel pentadiene (%) Conversion (%) Carbon Balance (%) Page 35 / 49

36 11.0 Catalytic Experimental Results: Isopropanol dehydration To assess the performance of the packed bed reactor, the dehydration reaction of isopropanol to propene was examined over the silica alumina catalyst. Typically, alcohol dehydration reactions exhibit a near zero order dependence in alcohol partial pressure and an apparent activation barrier of approximately 100 kj mol -1 under kinetic control [35,36]. At 398 K, the reaction order in isopropanol was measured to be 0.18 ± 0.03 with an apparent barrier of 92.3 ± 11.4 kj mol -1. Uncertainty in kinetic measurements are reported at a 95% confidence level. Page 36 / 49

37 12.0 Product identification: Mass Spectrum Fragmentation Patterns Verification of the identity of product chemical compounds was conducted through the injection of standards and comparison with retention times by gas chromatography. In addition, the microcatalytic method was coupled with a gas chromatography system with a mass spectrometer detector. In a manner identical to which the microcatalytic method was implemented (Section 5.0), a quartz tube packed with catalyst was placed in the liner of a GC equipped a mass spectrometer detector. Shown in Figures S9-S13 are the mass spectrum fragmentation patterns of the various products detected in the course of applying the microcatalytic method to a P-SPP catalyst with 3-MTHF as the injected reactant. Page 37 / 49

38 Figure S9. Mass spectrum fragmentation pattern of isoprene A produced over P-SPP from 3- MTHF using the microcatalytic method at 400 o C B NIST mass spectrum fragmentation pattern library Page 38 / 49

39 Figure S10. Mass spectrum fragmentation pattern of 1,3 Pentadiene A produced over P-SPP from 3-MTHF using the microcatalytic method at 400 o C B NIST mass spectrum fragmentation pattern library Page 39 / 49

40 Figure S11. Mass spectrum fragmentation pattern of 1,4 Pentadiene A produced over P-SPP from 3-MTHF using the microcatalytic method at 400 o C B NIST mass spectrum fragmentation pattern library Page 40 / 49

41 Figure S12. Mass spectrum fragmentation pattern of butene A produced over P-SPP from 3- MTHF using the microcatalytic method at 400 o C B NIST mass spectrum fragmentation pattern library Page 41 / 49

42 Figure S13. Mass spectrum fragmentation pattern of propene A produced over P-SPP from 3- MTHF using the microcatalytic method at 400 o C B NIST mass spectrum fragmentation pattern library Page 42 / 49

43 13.0 Experimental three-dimensional selectivity map An advantage of the microcatalytic method is the ability to rapidly screen multiple reaction conditions. Given that injections of the reactant (3-MTHF) were performed using a GC automated liquid sampler, a variety of reaction conditions could be sampled through an automated process (GC software). Automated GC software permits the organization of inlet condition sequences (i.e. long lists of temperature and flow conditions / space velocity). Therefore, in an effort to identify the reaction conditions that optimize the production of the desired product, isoprene, a grid of 112 reactions conditions was tested. The grid of points spanned twelve different temperatures ranging from 225 to 400 o C, in addition to fourteen different space velocities ranging from 0.9 to 180 s -1. To avoid systematic errors, the order of experimental trials with varying reaction temperature and space velocity depicted in Figure 3 was randomized. Also, given the nature of the reaction chemistry, oxygenated hydrocarbons reacting over solid acid catalysts, deactivation through coke deposition was present. Not accounting for parasitic phenomena such as deactivation can unfortunately lead to misleading representations of catalytic data. Therefore, to avoid the manifestation of artifacts resulting from catalyst deactivation, reference conditions were routinely utilized to ensure all catalytic data was compared equally. Once the catalyst was first placed on stream, a reference condition was selected and tested multiple times in a row to establish a catalytic baseline. Using this catalytic reference point, a bracketing technique was employed to explore a new condition followed by a return to the reference point. Essentially, every data point was bracketed on either end by the reference condition. This bracketing technique was continuously applied throughout the process of constructing the reaction Page 43 / 49

44 conditions grid, allowing us to ensure that all catalytic data was compared relative to the same reference point established with a pristine catalyst. Once collected, the data at all conditions was arranged by temperature and space velocity with their corresponding selectivity to each of the products. This grid of catalytic information was then plotted in MATLAB, where a colored contour plot could be created using existing 3Dplotting tools. Given the discrete nature of the data collected, data interpolation was required to create a three-dimensional continuous contour plot. Presented in Figure 3 of the text is the contour plot created through interpolation via polynomial smoothing of the overall data set. The data was first empirically fit to a 5 th order polynomial in both temperature and space velocity while minimizing the residual error. Illustrated in Figure S14 is the parity plot between the polynomial fit (5 th order) and the experimental data for isoprene selectivity where reasonable agreement is obtained. It is worth noting that no form of extrapolation was applied to the data set. Page 44 / 49

45 Figure S14. Parity plot of the fifth order polynomial fit (Figure 3) of isoprene selectivity versus experimental data. Alternatively, a linear interpolation can be applied as shown in Figure S15, where the data between two experimental points is calculated to be a linear average; linear interpolation by this method is a MATLAB built-in function. The trends of selectivity to products appear to be the same when comparing the resulting plot (Figure S14) to that in Figure 3 of the text; isoprene formation is favored at low space velocities and an intermediate temperature of ~ 325 o C. The 3D plot resulting from linear interpolation, however, is more textured, which is likely the result of experimental variability. The linear interpolation, however, provides a tighter fit of the experimental data, shown in the parity plot for isoprene selectivity in Figure S16. Page 45 / 49

46 Figure S15. Linear Interpolation Depiction of Figure 3. A three-dimensional plot of 112 experimental measurements of selectivity to isoprene (large panel), pentadienes (upper right), and butane/propene (lower right). The color map is presented using a linear interpolation of the experimental data. Page 46 / 49

47 Figure S16. Parity plot of the linear interpolation fit (Figure S15) of isoprene selectivity versus experimental data. Page 47 / 49

48 14.0 Time-on-Stream (TOS) Study of P-SPP Catalyst. To evaluate the long-term activity of P-SPP catalyst and selectivity to dienes, we have also evaluated the stability of the P-SPP catalyst during the course of 3-MTHF dehydradecyclization at 325 C and a partial pressure of 5.5 torr. Shown in Figure S17 is the 3-MTHF conversion profile with time on stream (TOS), where the conversion drops from 43% initially to 16% after approximately 18 hours on stream. Interestingly, the selectivity to dienes appears to increase initially, where the selectivity to isoprene and dienes (isoprene plus 1,3 pentadiene) stabilizes at approximately 66 and 86 %, respectively. Throughout the course of evaluating the catalyst s stability, the carbon balance was maintained at an average value of 98%. Figure S17. Time-on-stream study of isoprene production. The vapor-phase conversion (blue, ) of 3-methyl-tetrahydrofuran was selective to isoprene (orange, ) and dienes (isoprene + 1,3-pentadiene, grey, ). The carbon balance (yellow, ) of each experiment equals 100% within experimental error. Catalyst: P-SPP, Temperature: 325 C, Partial pressure: 5.5 torr. Page 48 / 49

49 References [31] Cho, H. J.; Ren, L.; Vattipailli, V.; Yeh, Y. H.; Gould, N.; Xu, B.; Gorte, R. J.; Lobo, R.; Dauenhauer, P. J.; Tsapatsis, M.; Fan, W. ChemCatChem 2017, 9, 1-6. [32] Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenb, D. J. Gaussian 09 (Rev. A.2); Gaussian, Inc. Wallingford, CT, [33] Marenich, A.V.; Cramer, C.J.; Truhlar, D.G. J. Phys. Chem. B 2009, 113, [34] Ochterski, J.W. Gaussian Inc, Pittsburgh, PA 2000, [35] Bedia, J.; Ruiz-Rosas, R.; Rodríguez-Mirasol, J.; Cordero, T. J. Catal. 2010, 271, [36] Turek, W.; Haber, J.; Krowiak, A. Appl. Surf. Sci. 2005, 252, Page 49 / 49