Influence of Particle Size on Reaction Selectivity in Cyclohexene Hydrogenation and Dehydrogenation over Silica-Supported Monodisperse Pt Particles

Abstract

The role of particle size during the hydrogenation/dehydrogenation of cyclohexene (10 Torr C₆H₁₀, 200–600 Torr H₂, and 273–650 K) was studied over a series of monodisperse Pt/SBA-15 catalysts. The conversion of cyclohexene in the presence of excess H₂ (H₂: C₆H₁₀ ratio = 20:60) is characterized by three regimes: hydrogenation of cyclohexene to cyclohexane at low temperature (<423 K), an intermediate temperature range in which both hydrogenation and dehydrogenation occur; and a high temperature regime in which the dehydrogenation of cyclohexene dominates (>573 K). The rate of both reactions demonstrated maxima with temperature, regardless of Pt particle size. For the hydrogenation of cyclohexene, a non-Arrhenius temperature dependence (apparent negative activation energy) was observed. Hydrogenation is structure insensitive at low temperatures, and apparently structure sensitive in the non-Arrhenius regime; the origin of the particle-size dependent reactivity with temperature is attributed to a change in the coverage of reactive hydrogen. Small particles were more active for dehydrogenation and had lower apparent activation energies than large particles. The selectivity can be controlled by changing the particle size, which is attributed to the structure sensitivity of both reactions in the temperature regime where hydrogenation and dehydrogenation are catalyzed simultaneously.

Appendix I. Deviations from Thermodynamic Selectivity at High Temperatures: Influence of Hydrogen Coverage and Surface Reaction Segregation

The thermodynamics of cyclohexene hydrogenation are favored at low temperatures (ΔG _h = −18 kcal mol⁻¹ at 298 K), while the dehydrogenation is favored at high temperatures (ΔG _d = −11 kcal mol⁻¹ at 600 K) [37]. The thermodynamic selectivity, defined as $\frac{{K_{\text{h}} }}{{\left( {K_{\text{h}} + K_{\text{d}} } \right)}}$, where K _h and K _d are the respective equilibrium constants (and related to the free-energy of reaction, $K = exp\left( {\frac{{ - \Updelta G_{{{\text{h or}} {\text{d}}}} }}{\text{RT}}} \right)$ for hydrogenation and dehydrogenation is plotted as the solid line in Fig. 6. The free-energy reaction data was obtained from [38]. At 400 ≤ T ≤ 450 K, the experimentally-measured selectivity for hydrogenation is lower than the thermodynamic selectivity for most samples, although a few of the low-temperature experimentally-measured selectivity data was greater than the thermodynamic value. This is attributed to error in the measurement of the individual reaction rates in this regime where the dehydrogenation activity is low. Figure 6a demonstrates the selectivity for cyclohexane predicted by thermodynamics is much greater than the experimentally-measured values at temperatures >450 K for all catalysts (supported and single crystal). The deviation in this temperature range is related to a much larger decrease in the rate of hydrogenation over this temperature range (see Fig. 1a, the rate of hydrogenation actually decreases with increasing temperature) than the rate of dehydrogenation which has begun to plateau (Fig. 1b). The overall trend of increased hydrogenation selectivity on catalysts containing larger particles is apparent over the entire temperature range as discussed in Sect. 3.2 and demonstrated in Fig. 3. Due to the fact that cyclohexane selectivity is lower than the thermodynamic selectivity, the benzene selectivity will be correspondingly greater than the thermodynamic selectivity (Fig. 6b). For all but a few data points, the selectivity to benzene is greater than that predicted by thermodynamics.

Fig. 6

The selectivity to (a) cyclohexane and (b) benzene over a number of catalysts: (●) 3.2% Pt(1 nm)/SiO₂; (□) 0.6% Pt(1.7 nm)/SBA-15; (▲) 0.77% Pt/(2.9 nm)/SBA-15; (○) 0.6% Pt(3.6 nm)/SBA-15; and ( ) 0.62% Pt(7.1 nm)/SBA-15, (★) Pt(111) and (✩) Pt(100) single crystal surfaces [8]. The solid line represents the thermodynamic selectivity, based on the temperature dependent Gibbs free-energy (ΔG _rxn(T)) of both hydrogenation and dehydrogenation reactions [38]. The points in (a) that lie above the thermodynamic selectivity represent typical uncertainties in the rate measurements, while the points in (b) lie above the thermodynamic selectivity line because the decreased hydrogenation rate leads to a selectivity in benzene greater than that predicted by thermodynamics. The reaction conditions were 10 Torr C₆H₁₀, 200 H₂ Torr and the temperature stated on the x-axis

Significant deviations of kinetic selectivity from thermodynamic selectivity has been documented previously for this reaction [9]. The authors suggested that both reactions surface segregate to specific sites on the nanoparticle surface [9]. Cyclohexene dehydrogenates preferentially at low coordination sites (which are more prevalent on smaller particles and consistent with particle-size dependent data (see Table 1)), and the adsorbed hydrogen diffuses away from low coordination sites to higher-coordination sites where it desorbs ($2 {\text{H}}^{*} \rightleftharpoons {\text{H}}_{ 2} \left( {\text{g}} \right) + 2^{*}$) immediately or reacts with cyclohexene adsorbed at high coordination sites to form cyclohexane. The hydrogen at the high coordination sites in the presence of adsorbed cyclohexane may represent the high temperature population observed in H₂-TPD studies [35]. The equilibrium occurs in isolation, effectively separating adsorbed hydrogen and hydrocarbon, allowing non-equilibrium yields of benzene. Sermon and co-workers propose that under conditions where both cyclohexene hydrogenation and dehydrogenation occur on the same surface, the two reaction pathways segregate [9]. This picture is consistent with the cyclohexene hydrogenation rate and the observed changes in selectivity with particle size (see Figs. 2 and 3, respectively).