Part B: Unraveling the mechanism of catalytic reactions through kinetics and thermodynamics

Transcription

1 Part B: Unraveling the mechanism of catalytic reactions through kinetics and thermodynamics F.C. Meunier, J. Scalbert and F. Thibault-Starzyk Appl. Catal. A: Gen. (2015), in press, doi: /j.apcata Fred Meunier Institut de Recherche sur la Catalyse et l Environnement de Lyon Villeurbanne, France EUROKIN meeting, Milan 18/2/

2 Outline: 1. Why studying ethanol condensation to butanol? 2. Suggested reaction mechanisms in the literature 3. Methods: - Thermodynamic calculations - Reactor + analytics 4. Ethanol condensation over a transition metal-free hydroxyapatite Ca 10 (PO 4 ) 6 (OH) 2 5. Other examples: - n-alkane hydroisomerisation - Methanol steam reforming - NO oxidation to NO 2 during SCR 2

3 Oxo process: propene, CO, H 2 (Co, Rh, Ni) Solvent, chemicals Butanol Aldol condensation: Base + Metal M-free basic solids! ABE Fermentation Ethanol 3

4 Outline: 1. Why studying ethanol condensation to butanol? 2. Suggested reaction mechanisms in the literature 3. Methods: - Thermodynamic calculations - Reactor + analytics 4. Ethanol condensation over a transition metal-free hydroxyapatite Ca 10 (PO 4 ) 6 (OH) 2 Scalbert et al., J. Catal. 311 (2014) 28 4

5 Metal-free ethanol condensation to butanol: mechanism? «Guerbet» mechanism (Acetaldehyde self-aldolisation) ethanol (éthanal) (3-hydroxybutenal) (butenal) Direct dimerisation (butanal) Ogo et al., Appl. Catal. A 402 (2011) 188 ethanol ethanol Semi-direct dimerisation Yang and Meng, J. Catal. 142 (1993) 37 Yang and Meng, J. Catal. 142 (1993) 37 ethanol acetaldehyde 5

6 Ethanol condensation mechanism over Mg x AlO y butanol Direct Acetaldehyde Self-aldolisation Di Cosimo, Apesteguia, Gines, Iglesia J. Catal. 190 (2000) 261 6

7 Outline: 1. Why studying ethanol condensation to butanol? 2. Suggested reaction mechanisms in the literature 3. Methods: - Thermodynamic calculations - Reactor + analytics 4. Ethanol condensation over a transition metal-free hydroxyapatite Ca 10 (PO 4 ) 6 (OH) 2 Scalbert et al., J. Catal. 311 (2014) 28

8 Molar Gibbs energy of reaction: r G = r G ø + RT ln Q = RT ln Q/K where Q = reaction quotient and K = equilibrium constant (éthanal) (3-hydroxybutenal) (butenal) (butanal) 2 acetaldehyde + 2 H 2 butanol + water Calculating Q requires quantifying: butanol, acetaldehyde, H 2 O and H 2 8

9 During a catalytic test, Q goes from a value of 0 to K or less, but not higher r G(T) = RT ln Q/K Q/K = η = progress to equilibrium Mechanism validity criterion: Q/K = η 1 9

10 During a catalytic test, Q goes from a value of 0 to K or less, but not higher r G(T) = RT ln Q/K Q/K = η Mechanism validity criterion: Q/K = η 1 10

11 ethanol Reactor: quartz plug flow reactor Catalyst: Hydroxyapatite Ca 10 (PO 4 ) 6 (OH) 2 from Accros Operating conditions : 350 C < T < 450 C 3.8 < P ethanol < 19.1 kpa 1.4 < WHSV < 56 g ethanol g -1 cata h -1 Gas Chromatograph + Mass Spectrometer + FT-IR gas cell ethanol acetaldehyde H 2 butanol H 2 O acetaldehyde ethene butenol butadiene 11

12 Outline: 1. Why studying ethanol condensation to butanol? 2. Suggested reaction mechanisms in the literature 3. Methods: - Thermodynamic calculations - Reactor + analytics 4. Ethanol condensation over a transition metal-free hydroxyapatite Ca 10 (PO 4 ) 6 (OH) 2 Scalbert et al., J. Catal. 311 (2014) 28

13 Hydroxyapatite: selectivity vs. temperature %Ethanol = 15.2 % ; WHSV = 14 h -1 Sélectivités (%) 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% Autres butenol éthylene butadiene acétaldéhyde butanol 0% Température de réaction ( C) most abundant product is butanol 13

14 Effect of contact time at 400 C Pression partielle (kpa) Butanol Acétaldéhyde butanol acetaldehyde secondary butanol primary butanol W/F (g cata.h.mol -1 Ethanol) Two routes to butanol? one direct, one involving acetaldehyde? 14

15 HSC equilibrium calculation based on self-aldolisation intermediates % (molaire) 10 H 2 (g) H 2 O(g) Butanol(g) Acetaldehyde(g) Butenal(g) 1 Butenol(g) Température ( C) At 400 C: acetaldehyde and butenal should be more abundant than butanol 15

16 Computing Q/K for: 2 Acetaldehyde + 2 H 2 Butanol + H 2 O Temperature %EtOH WHSV K Q η = Q/K / C ,2 % 28 h ,2 % 28 h ,2 % 28 h ,6 % 1,4 h ,6 % 2,1 h ,6 % 4,2 h ,6 % 7,0 h Q/K >> 1 16

17 Suggested reaction mechanisms: Main: direct reaction mechanism Q/K < 1 - H 2 O 2 OH OH ethanol butanol Minor: semi-direct reaction mechanism Q/K < 1 ethanol acetaldehyde O O OH H OH - H 2 + H OH - H 2 O Butenol OH + 2 H butanol OH Final hydrogenation: H-tranfer from sacrifical ethanol (not via H 2 dissociation) Minor route is intrinsically less selective: one ethanol is sacrificed 17

18 Conclusion: ethanol condensation at high T over hydroxyapatite without metals O - 2 H OH H Indirect + ethanol -acetaldehyde -H 2 O -H 2 Direct - H 2 O - H 2 O Aldol selfcondensation O OH + 2 H 2 H Target? try to hinder the less selective mechanism that involves acetaldehyde, contrary to the case of low T metal-promoted reactions! Scalbert et al., J. Catal. 311 (2014) 28 18

19 5. Other cases with useful insight from thermodynamics n-alkane hydroisomerisation over MoOx: bifunctional mechanism, RDS is different for C 4 and C 5 Meunier et al., Chem. Commun.1999, 259 CH 3 OH + H 2 O = 3 H 2 + CO 2 over Cu/Zn/Zr/Al: CO not reaction intermediate, secondary reaction product Meunier, Chem. Commun. 2003, 1954 C 3 H 6 -SCR of NO over Ag/ γ-al 2 O 3 : NO 2 not formed via NO + O 2, but via CxHyNzO +O 2 Breen et al., Chem. Commun. 1999,

20 Case 1: n-alkane hydroisomerisation Alkane hydroisomerisation over reduced MoO 3 Reaction mechanism? - bifunctional (metallic site + acidic site) - metallacyclobutane Matsuda et al., Catal. Lett. 47 (1997) 99 Blekkan et al., Ind. Eng. Chem. Res. 33 (1994) 1657 Experimental details for n-butane to isobutane : MoO 3 (Fluka), BET < 2 m 2 g % n-butane / H 2 Total flowrate: ml min -1 Mass of catalyst: mg T = 350 C (623 K) 20

21 Case 1: n-alkane hydroisomerisation n-butane hydroisomerisation on MoO 3 Yield /% isobutane ( 100) isobutene trans-butene cis-butene butene Time on stream /h 21

22 Case 1: n-alkane hydroisomerisation De/hydrogenation steps ± H 2 ± H 2 1h 2h 6h Steady-state Thermodynamic ratio n-butane 1-butene isobutane isobutene ± ± De/hydrogenation rates increases with TOS towards equilibrium 22

23 Case 1: n-alkane hydroisomerisation Skeletal isomerisation steps 1h 2h 6h Steady-state Thermodynamic ratio isobutane n-butane isobutene 1-butene ± ± Skeletal isomerisation rate gradually increases, not at equilibrium: it is the rate-determining step (RDS) 23

24 Case 1: n-alkane hydroisomerisation Conclusion 1: n-butane hydroisomerisation: bifunctional mechanism operates Acid site? ± H 2 + H 2 RDS MoO Reduced MoO 3 Meunier, Chem. Commun, 2003,

25 Case 1: n-alkane hydroisomerisation Can the activity be promoted by a n-alkene isomerisation catalyst? CoAlPO-11 (AEL structure) monodimensional 10-member ring framework [001] 4.5 nm x 6.5 nm Houzvicka and Ponec, Catal. Rev. 39 (1997)

26 Case 1: n-alkane hydroisomerisation n-butane hydroisomerisation: promotion by CoAlPO-11? Isobutane Yield/ % mg of MoO mg of CoAlPO mg of MoO mg of CoAlPO mg of MoO 3 50 mg of CoAlPO Time on stream /min CoAlPO-11 promotes MoO3 for n-butane hydroisomerisation Meunier et al., Catal. Today, 2006, 112, 64 26

27 Case 1: n-alkane hydroisomerisation nc5 and nc6 hydroisomerisation/moo 3 : promotion by CoAlPO-11? Methyl-2-butane Yield /% Methyl-2-pentane Yield /% mg of MoO mg of MoO mg of MoO mg of CoAlPO mg of MoO mg of CoAlPO Time on stream /min Time on stream /min CoAlPO-11: - Initially promotes nc 5+ to ic 5 - At steady-state: small inhibition 27

28 Case 1: n-alkane hydroisomerisation Short TOS RDS ± H 2 + H 2 Reduced MoO 3 Steady-state RDS More Reduced MoO 3 Conclusion 2: nc5 and nc6 hydroisomerisation: RDS limited by de/hydrogenation steps at steady-state Meunier et al., Catal. Today, 2006, 112, 64 28

29 Case 1: n-alkane hydroisomerisation Why C 4 and C 5+ hydroisomerisation RDS are different? DFT calculation on energetics of branching of carbenium ions M. Boronat et al., Appl. Catal. A: Gen. 146 (1996) 207 +H + + -H + E a +H + + E a -H + 29

30 Other examples: 1. n-alkane hydroisomerisation 2. Methanol steam reforming 3. NO oxidation to NO 2 during SCR 30

31 Case 2: Methanol steam reforming methanol steam reforming: CH 3 OH + H 2 O = 3 H 2 + CO 2 Mechanism based on CO as reaction intermediates: WGS Other mechanisms not based on CO have been proposed, e.g.: 31

32 Case 2: Methanol steam reforming H 2 H 2 CO CO 2 CO CO: seems to be a secondary product If CO included in equilibrium calculation (lines n), H 2 and CO 2 concentrations go beyond equilibrium values! 32

33 Case 2: Methanol steam reforming Because the proportion of H 2 and CO 2 observed exceeds that predicted in the CH 3 OH /H 2 O /H 2 /CO 2 /CO system, this system is not relevant. (1) + (2) cannot be the primary reaction pathway, another reaction scheme (unknown) applies Only at higher contact time does the proportions of CO, H 2 and CO 2 in the CH 3 OH /H 2 O /H 2 /CO 2 /CO system becomes e consistent with that predicted by the thermodynamics: this is because the reverse WGS reaction is now significant. CO is formed consecutively from CO 2 Importance of selecting the set of species relevant to the reaction scheme Breen et al., Chem. Commun, 1999,

34 Other examples: 1. n-alkane hydroisomerisation 2. Methanol steam reforming 3. NO oxidation to NO 2 during SCR 34

35 Case 3: NO oxidation to NO 2 during SCR C x H y -SCR of NO over Al 2 O 3 -based catalysts Role of CoOx in CoO x / γ-al 2 O 3 NO + 1 / 2 O 2 NO 2 H. Hamada et al., Catal. Today 29 (1996) 53 N 2 yield during the C 3 H 6 -SCR of NOx 100 % Al 2 O 3 (NO 2 ) 50 Co/Al 2 O 3 (NO) Al 2 O 3 (NO) T / C Yet NO + O 2 to NO 2 activity low! J. Yan et al., J. Catal. 172 (1997) 178 Is NO + O 2 to NO 2 really important? 35

36 Case 3: NO oxidation to NO 2 during SCR C 3 H 6 -SCR of NO over Ag/ γ-al 2 O 3 C 3 H 6 conv. /% 100 N 2 yield /% N 2 O yield /% T / C 10% Ag /Al 2 O 3 1% Ag /Al 2 O 3 Al 2 O ppm NO 500 ppm C 3 H 6 2.5% O 2 W/F = 0.06 g s cm T / C NH 3 yield /% T / C T / C NO 2 yield /% NO + 1 / 2 O 2 NO ? T / C 36

37 Case 3: NO oxidation to NO 2 during SCR NO 2 formation over γ-al 2 O 3 -based catalysts NO 2 /NO ratio 500 ppm NO 500 ppmc 3 H 6 2.5% O 2 0.4% Co 1 (x 0.25) γ-al 2 O 3 NO + 1 / 2 O 2 NO SOx-1% Ag T / C 1% Ag The system to consider is not NO, O 2 and NO 2 37

38 Case 3: NO oxidation to NO 2 during SCR Origin of the NO 2 formation over γ-al 2 O 3 -based catalysts Homogeneous oxidation of alkanes: formation of R-NO 2, role of R-ONO K. Otsuka et al., Catal. Today 45 (1998) 23 NO + 1 / 2 O 2 => NO C 3 H / 2 O NO => 2 CH 3 NO 2 + CO -298 C H 3 NO / 4 O 2 => NO 2 + CO / 2 H 2 O -734 F.C. Meunier et al., Chem. Commun. (1999) 259 r G (540 C) /kj mol -1 The true R-NOx intermediate(s) is(are) likely more complex, but these calculi show that such routes afford high concentration of NO 2, contrary to the direct oxidation. 38

39 Case 3: NO oxidation to NO 2 during SCR Suggested mechanisms of the C 3 H 6 -SCR of NO over high and low Ag loading γ-al 2 O 3 NO 2 N 2 O + N 2 NO / O 2 C 3 H 6 OO O N N N CxHy NO 2 NO + O 2 C 3 H 6 O 2 NO 2 + CO x /H 2 O N 2 Ag NOx - Ag +? R-ONO R-NO 2 R-NCO R-NH 2 NH 3 γ-al 2 O 3 Meunier et al. J. Catal. 187 (1999)

40 Conclusion: Kinetics and Thermodynamics The comparison of the reaction quotient (Q) or similar ratios to the corresponding thermodynamic equilibrium ratios (e.g. K) can be very useful in supporting or rejecting a reaction mechanism. 40