Article pubs.acs.org/JPCC
Radical Cation Intermediates in Propane Dehydrogenation and Propene Hydrogenation over H‑[Fe] Zeolites Jang Ho Yun and Raul F. Lobo* Center for Catalytic Science and Technology, Department of Chemical and Biomolecular Engineering, University of Delaware, 150 Academy Street, Newark, Delaware 19716, United States S Supporting Information *
ABSTRACT: This report investigates the mechanistic relationship between the monomolecular propane dehydrogenation reaction and the reverse reaction, the propene hydrogenation, over H-[Fe]ZSM-5 catalysts. It is shown that the difference in the apparent activation energies of the forward and reverse reactions is equal to the reaction enthalpy (∼130 kJ mol−1) and that the rate constants of the reactions have an isokinetic relationship. The ratios of the rate constants of the forward to the reverse reactions are equal to the equilibrium constant (e.g., KP ≈ 0.033 bar at 773 K) even if the reactions occur separately, away from equilibrium. The results are consistent with the principle of microscopic reversibility and indicate that both forward and reverse reactions are structurally related and proceed through the same elementary steps and reaction intermediates. The pattern of selectivity, the activation energy, and the estimated enthalpy and entropy of formation of the transition states in H-[Fe]ZSM-5 are very different from the observed values for the isostructural H-[Al]ZSM-5, indicating that despite their structural similarities the reactions proceed through different mechanisms in each catalyst. Analysis of the energy change along the reaction coordinate, including the reaction enthalpy and the apparent activation energies, suggests that in H-[Fe]ZSM-5 the reaction proceeds through radical cation-like intermediates. Analysis of a putative reaction mechanism and the energetics of electron transfer in the zeolite channels shows that dehydrogenation of propane is kinetically favored (as observed) over cracking of propane because ethene radical cations are less stable than propene radical cations.
■
were found in chromium oxide catalysts after calcination,16 but reduced Cr3+ ions are the active components of the catalyst.17,18 Because the reduced Cr3+ species are reoxidized to Cr6+, a redox mechanism for dehydrogenation of alkanes has been proposed on chromium oxide catalysts.10,18−20 In a previous report we investigated the catalytic behavior of iron−silicate zeolites for monomolecular propane conversion.21 The reaction proceeds through only two pathways: cracking and dehydrogenation. The ratio of dehydrogenation to cracking reaction rates at a reaction temperature of 773 K was about 22 over H-[Fe]ZSM-5, while H-[Al]ZSM-5 exhibited higher selectivity for cracking; the ratio was about 0.5. At higher conversion (20%) than what is typically used for the monomolecular reaction ( 2500) to prevent formation of undesired products, such as propene oligomers. The selectivity of propane conversion over H-[Fe]ZSM-5 was analyzed from a thermochemical perspective. Energy changes along the reaction coordinates were estimated based on the apparent activation energies and the ionization energies of propane and propene. The results indicate that in H-[Fe]ZSM-5 the main reaction pathway proceeds through radical cation intermediates and not through alkanium-like species.
and determined that the deprotonation enthalpy of the iron sample is higher by 23 kJ mol−1. As a consequence, acidcatalyzed reactions on the iron zeolite should be slower and have higher activation energies than on the aluminum forms.22 Thus, activation energy differences between Fe and Al zeolites indicate that the reaction proceeds through a nonprotolytic mechanism in the iron samples. Enhanced dehydrogenation selectivity (dehydrogenation rate/cracking rate 0.3:1) and lower apparent activation energies (187 versus 127 kJ mol−1) were also reported for dehydroxylated aluminosilicate zeolites. It has been suggested23 that dehydroxylation of Brønsted acid sites can occur by homolytic decomposition, producing an oxygen atom with a single-occupied molecular orbital that might lead to redox reactions at high temperatures. These results suggest that propane conversion mechanism over H[Fe]ZSM-5 is not protolytic, the mechanism that is now accepted to occur over the acid form of aluminosilicate zeolites.24−26 The protolytic mechanism over H-[Al]ZSM-5 involves formation of alkanium-like transition states, formed after the transfer of a proton from the zeolite acid site to the adsorbed propane molecule (Scheme 1a). A redox mechanism Scheme 1. Reaction Mechanisms for Dehydrogenation of Propane (a) via an Alkanium Ion over H-[Al]-ZSM-5 and (b) via Radical Cation over H-[Fe]-ZSM-5
involving formation of propane radical cations as reaction intermediates27 over H-[Fe]ZSM-5 (Scheme 1b) has been suggested. This scheme shows a mechanism in which tetrahedral FeIII in the zeolite framework is reduced to FeII as an electron is transferred from a propane molecule occluded in the zeolite channel. The tetrahedral FeII is oxidized back to tetrahedral FeIII after formation of propene in the reaction cycle. Tetrahedral FeII in silicates and other oxides have been reported for synthetic glasses such as Na2FeSi3O8, K2FeSi3O8,28 MFe0.5Si2.5O6 (M = K, Na, Cs),29−31 Fe2SiO4,32,33 and Rb2FeSi3O834 as well as for minerals such as staurolite (Fe 1.5 Mg 0.5 Al 9 Si 3.9 Al 0.1 O 22 (OH) 2 13.1 wt % tetrahedral FeIIO4/2).34 These structures demonstrate that a tetrahedral FeIIO4/2 structure is chemically stable and could be formed during the reaction without destruction of the framework or migration of Fe atoms from the framework to an extra27293
dx.doi.org/10.1021/jp504453n | J. Phys. Chem. C 2014, 118, 27292−27300
The Journal of Physical Chemistry C
■
Article
EXPERIMENTAL SECTION Synthesis methods to prepare H-[Al]ZSM-5 and H-[Fe]ZSM-5 and the experimental procedure for propane conversion have been reported elsewhere.21,23,51 In general, about 20 mg of zeolite catalyst was transferred to a quartz tube reactor (i.d. = 5 mm). To support the sample in the center of the reactor chamber, quartz wool (∼0.05 g) was placed near the bottom of the reactor tube and quartz chips (∼1.6 mm diameter, Quartz Plus, Inc.) were placed between the quartz wool and the sample. The reactor was heated in a cylindrical furnace (C5232, Hoskins Mfg. Co.), controlled by a temperature controller (NC 74000, Omega Engineering) and a thermocouple (type K, Omega Engineering) inserted in the quartz tube directly over the sample. The samples were first treated in flowing N2 (100 sccm) by heating to 753 K with a rate of 2 K/min for 3 h to obtain the acid form of the zeolites. The reactants were flown into the reactor with dry N2 as the carrier gas to vary the partial pressure of H2 (0.2−1 bar) and C3H6 (from 5 × 10−5 to 4 × 10−4 bar). A high H2/C3H6 ratio (>2500) was maintained at all times to avoid oligomerization reactions. The reactor effluent is analyzed using gas chromatography (GC, Shimadzu 2014). A molsieve column is used for separation and eluted into a thermal conductivity detector (TCD). A RT-alumina column is connected to a flame ionization detector (FID) for separation and quantification of hydrocarbons. The temperature of the columns was increased from 308 to 393 K at a rate of 10 K/ min, and the products were collected during a 12 min program.
kDapp = k 2K1 =
app Ea,D
‡
app Ea,D
e(ΔS / R)e(− RT ) = ADappe(− RT ) app Ea,D = RT 2
(2)
∂(ln rD) = ΔH1 + E2 = ΔH2‡ − ΔHH+Z− ∂T
− ΔHC3H8(g)
(3)
⎡ ⎛ ek T ⎞⎤ ΔSDapp = R ⎢ln(ADapp) − ln⎜ B ⎟⎥ ⎝ h ⎠⎦ ⎣ = ΔS2‡ − ΔSH+Z− − ΔSC3H8(g)
(4)
where rD is the rate of propane dehydrogenation per site, PC3H8 is the partial pressure of propane, kDapp is the apparent dehydrogenation rate constant (first order), K1 is the equilibrium constant relating intrazeolite propane concentration to gas or extra-zeolite propane pressure, and k2 is the intrinsic rate constant for the rate-determining step. KD‡ is the equilibrium constant for formation of activated complexes from app gaseous propane, Ea,D is the apparent activation energy (experimentally determined; ; as can be seen in eq 3, it is the enthalpy difference between intrazeolite intermediates and extra-zeolite reactants), Aapp D is the pre-exponential factor, and ΔSapp D is the apparent activation entropy (the difference in entropy between the transition state in the zeolite channel and the reactants in the gas phase). The propene hydrogenation reaction was conducted from 733 to 803 K with excess H2 (H2/C3H6 > 2500); the temperature range was similar to the one used for propane dehydrogenation. Under these reaction conditions propane was produced as the major product (selectivity of over 80%, Figure 1a), while ethylene and methane were formed in equimolar ratios with low selectivity. The minor products can be formed through monomolecular cracking of propoxide species59,60 or through interconversion of dehydrogenation transition states (C−H−H) to cracking transition states (C−C−H).23,36 Propane production rates show a linear correlation with the partial pressures of H2 and C3H6 (Figure 2), and consequently, (i) intrazeolite H2 and C3H6 concentrations are equilibrated with their extrazeolite pressures, (ii) C3 intermediates are kinetically relevant, and (iii) zeolite surface coverage is low. The rate constants at a reaction temperature of 773 K are defined to be the slope of the propane formation rates vs the partial pressures of C3H6 and H2. The dehydrogenation average rate constant is 0.036 ± 0.006 (95% confidence interval) [mol (mol Al)−1 s−1 (bar H2)−1 (bar C3H6)−1] at 773 K (Supporting Information, Figure S2). Kinetics analysis of propene hydrogenation can be described as follows (based on the reverse reaction of the propane dehydrogenation mechanism shown in Scheme 2)
■
RESULTS AND DISCUSSION Catalytic Hydrogenation of Propene vs Dehydrogenation of Propane. Catalytic monomolecular propane conversion has been studied over a number of H-[Al] zeolites24,25,36,57,58 under conditions at which the zeolite surface is mostly empty, such as low reactant partial pressure, low conversion (
