(490e) Kinetic Modeling of Ethane and Propane Pyrolysis at SOFC Operating Conditions

            Fossil fuels have long been the world's dominant source of energy. This dependence on oil and other fossil fuel resources presents many difficulties, including resource depletion, accelerating global warming, escalating cost of oil, and national security issues. One way to address this problem is to employ more efficient energy conversion devices such as fuel cells. Solid-oxide fuel cells SOFCs in particular are promising because they have the potential to operate directly on hydrocarbons. Typical operating temperatures for SOFCs range from 600°C to 800°C. Because of these high operating temperatures, thermal decomposition of the hydrocarbon fuel within the anode channel is significant and is likely to influence SOFC operation. A particular concern is the potential for deposit formation within the anode channel. Such deposits can be formed via gas phase reactions as well as catalytic surface chemistry mechanisms. In the gas phase, molecular weight growth reactions can lead to PAH formation that might condense on the catalyst within the anode. Characterization of such reactions can provide guidance as to the combinations of temperature, time and fuel composition that might promote such molecular weight growth, thereby suggesting operating conditions to avoid. Alternatively, this knowledge might suggest approaches to mitigate such deposit formation by interrupting the molecular weight growth reaction sequence. In this talk we will analyze the pyrolysis of ethane and propane, with particular emphasis upon characterization of the reactions that lead to molecular weight growth.

            Ethane and propane pyrolysis experiments were performed in a 6 mm i.d. quartz flow reactor at ambient pressure (~ 0.8 atm) over a 550 ~ 800^oC temperature range with a residence time of ~5 s. The fuels were diluted with N₂ in a mole ratio of ~50/50. Fuel conversion and the product distribution as a function of temperature were monitored with a GC (HP 5890 Series II) and a mass spectrometer (MKS Cirrus LM99). A wide range of products was observed, with the molecular weight growth species (up to C₆) appearing at the higher temperatures. Quantification of the exit gas composition led to H and C atom masses that closed within 2-3%.

            The experimental results were modeled with a plug flow model in CHEMKIN PRO using the measured temperature–distance profiles. The mechanism used was based on our earlier work [1] but updated with improved hydrogen abstraction rate rules based on CBS-QB3 calculations. [2] Although this model properly described the temperature dependence for both ethane conversion and the major product distributions (Fig. 1), it was not so successful for some minor products (Fig. 2). Methane and propylene were underpredicted while 1,3-butadiene was overpredicted. A reaction path analysis revealed that 1,3-butadiene was mainly produced from C₄H₇, formed by vinyl radical addition to ethylene. The main production pathway of propylene was beta-scission of 2-C₄H₉, while it was mainly consumed through H addition to form i-C₃H₇. This analysis thus suggests that the predictions can be improved by incorporating updated potential energy surfaces for C₃H₇, C₄H₇ and C₄H₉. For these updates we performed electronic structure calculations at the CBS-QB3 level, calculated the high pressure rate constants with transition state theory and used QRRK/MSC theory [3] to derive pressure-dependent rate expressions. When these updated kinetic data were incorporated into the mechanism, we observed significantly improved predictions for methane and propylene, as shown in Fig. 2.

           Although the predictions for 1,3-butadiene also improved, there is still room for further improvement (Fig. 3). This remaining overprediction in 1,3-butadiene may be connected to the underprediction of C₅ and C₆ species (Fig. 3) in that this may be due to reaction pathways from 1,3-butadiene to these products being too slow. We have already examined some of these pathways in detail by performing CBS-QB3 calculations for addition reactions to 1,3-butadiene as well as the Diels-Alder reaction between 1,3-butadiene and ethylene and incorporating these rate constants into the mechanism. Studies of other reactions such as methyl addition to 1,3-butadiene are in progress.

            One unexpected result that emerged from the model calculations is that several radical addition reactions become partially equilibrated very early in the reaction. The consequence of this observation is the possibility that some of the potentially important molecular weight growth reactions are controlled by thermodynamic properties rather than by kinetics.

            The modifications to the kinetic mechanism that improved the description of ethane pyrolysis also improved the predictions of the major products (CH₄, C₂H₄, and H₂) in propane pyrolysis. However, propylene, the other major product, is overpredicted at the higher temperatures while the minor products C₄, C₅ and C₆ species are underpredicted. It would appear that one way to improve the description of the molecular weight pathways is to identify more rapid radical addition reactions to propylene. Work is underway on such reactions, including, vinyl addition to propylene and addition reactions of the resonantly stabilized allyl radical.

            The results from this study suggest that the difficulty in predicting the observed molecular weight growth in both ethane and propane is related to the same problem: an overprediction of intermediate olefin species. A likely common cause of this overprediction is that the subsequent radical addition reactions of these species are either too slow or missing. It also appears that some of the reactions are limited by the rapid onset of partial equilibria. For such cases, it is not sufficient to increase the rate coefficient; instead one needs to identify a higher molecular weight species that can serve as a thermodynamic sink to allow these equilibria to shift toward the heavier species.

1.   Randolph, K.L. and A.M. Dean, Hydrocarbon Fuel Effects in SOFC Operation: An Experimental and Modeling Study of n-hexane Pyrolysis. PCCP, 2007. 9: p. 4245 - 4258.

2.   Carstensen, H.-H. and A.M. Dean, Rate Constant Rules for the Automated Generation of Gas-Phase Reaction Mechanisms. J. Phys. Chem. A, 2009. 113(2): p. 367-380.

3.   Chang, A.Y., J.W. Bozzelli, and A.M. Dean, Kinetic analysis of complex chemical activation and unimolecular dissociation reactions using QRRK theory and the modified strong collision approximation. Z. Physik. Chem., 2000. 214: p. 1533-1568.