Praseodymium oxychloride

In this work, we will show that the addition of TCM to the feedstream in the methane conversion process results in the enhancement of the conversion of methane and the selectivity to C2 hydrocarbons on praseodymium oxide primarily as a result of the formation of praseodymium oxychloride, in contrast with the production of carbon oxides on praseodymium oxide in the absence of TCM (8-10). The surface properties of these catalysts are characterized by application of adsorption experiments and X-ray photoelectron spectroscopy (XPS). 

The adsorption of carbon dioxide or oxygen on praseodymium samples was measured by a constant-volume method using a calibrated Pirani vacuum gauge. Praseodymium oxide was heated in oxygen (4 kPa) at 775°C for 1 h, then evacuated at 750°C for 0.5 h just before the measurement. The sample of praseodymium oxychloride was prepared from praseodymium chloride by heating under oxygen flow... 

Stable catalytic activity was observed with praseodymium chloride pretreated under an oxygen stream at 750°C for 15 h (Figure 4). The longer pretreatment produced high selectivity to C2+ compounds, that is, 58.8% in the presence of TCM and 58.5% in the absence of TCM after 0.5 h on-stream while methane conversions were 19.7 and 18.6%, respectively. Without TCM the conversion and the selectivity decreased gradually to 17.0 and 54.0% after 6 h on-stream, respectively. The catalytic activity was partially restored by addition of TCM to the reaction stream for 1 h. The conversion was 17.3% and the selectivity was 55.0% after 0.5 h on-stream following the period of the TCM feeding. The BET surface areas of the catalysts measured after the reactions in the presence and absence of TCM were both 1.5 m g The XRD patterns of the catalysts used for the reactions were identical with that of praseodymium oxychloride. A small amount of methyl chloride was detected when praseodymium chloride was used as a precursor of the catalyst both in the presence and in the absence of TCM. 

Adsorption of carbon dioxide or oxygen on the praseodymium samples was carried out in the pressure range of 1-40 Pa to evaluate the number of chemisorption sites on the samples. Praseodymium oxide irreversibly adsorbed 9.5 x 10" mol g of carbon dioxide. The amount of oxygen irreversibly adsorbed on the sample was 15.2 x 10" mol g Carbon dioxide or oxygen was not adsorbed on the samples containing chlorine, i.e., praseodymium chloride and praseodymium oxychloride prepared from the chloride by heating under oxygen flow at 750°C for 1 h. 

In the case of the praseodymium oxychloride produced from praseodymium oxide by the reaction in the presence of TCM (white-green portion), an O Is peak at 529.3 eV with a shoulder at 531 eV was observed. The peak shifted to 528.9 eV after xenon ion-sputtering for 0.5 min. The peak of Pr 3d was similar to that for praseodymium chloride, that is, the main peak was observed at 932.9 eV with a shoulder at 928 eV which was more clearly defined than that in the spectrum for praseodymium chloride. The peak shifted to 932.5 eV after the sputtering and the shoulder at 928 eV intensified as seen in the spectra for praseodymium chloride. The peak of Cl 2p was present at 198.8 eV and the position of the peak did not change after the sputtering. 

There were two peaks at 528.9 and 531.1 eV in the spectrum for O Is of the praseodymium oxychloride taken from the inlet portion of the reactor after the reaction in the absence of TCM (the sample originated from praseodymium chloride heated at 750°C for 1 h). The peak at 531.1 eV disappeared after xenon-ion sputtering while the main peak was present at 529.1 eV. The spectra for Pr 3d before and after xenon-ion sputtering were similar to those for praseodymium oxychloride which originated from praseodymium oxide. Although the Cl 2p spectra for the... 

Praseodymium oxychloride produced from praseodymium chloride with a pretreatment in oxygen followed by methane conversion in the presence of feedstream TCM (as shown in Figure 3) displayed an O Is peak at 529.0 eV but no shoulder. The peak position did not change after xenon-ion sputtering. The spectra for Pr 3d were similar to those for other praseodymium oxychloride samples. The binding energy of Cl 2p was 198.9 eV while the peak had a shoulder at 200 eV. The shoulder disappeared after the sputtering. 

No adsorption of carbon dioxide or oxygen was observed on either praseodymium chloride or oxychloride. This finding is consistent with the XPS results. The main peaks at 529 eV in the spectra for praseodymium oxychloride samples are also attributed to the lattice oxygen of the oxychloride while the peaks at 531 eV are assignable to O Is for praseodymium oxide, suggesting that the surfaces of the oxychloride samples are partially oxidized to praseodymium oxide. The 3d binding energy of 933 eV for praseodymium in the chloride and oxychloride implies that the valence of praseodymium is 3+, while the shoulder at 928 eV could be attributed to metallic praseodymium (77). 

Figure 6 Schemes for the interaction of methane with praseodymium oxychloride.

Although generalization of the present results to other catalyst systems is tempting, the evidence currently available is undoubtedly insufficient to justify such an extrapolation. It is clear, however, from both the present and previous work in this laboratory, that the chlorine atoms interact with the surface of the catalyst. There is also strong evidence to support the contention that the effect produced by the introduction of TCM into the feedstream can be primarily attributed to a modification of the catalyst surface and not to a gas phase process. The present work demonstrates that, at least with praseodymium oxide, the oxychloride is produced on addition of TCM and further, the oxychloride is largely responsible for the beneficial effects. Since the enhancements observed with TCM have been shown (2, 4-7) to be related to the nature of the catalyst, it is conceivable that these effects, while dependent on the formation of the oxychloride, are also a function of the thermodynamic stability of the oxychloride. Further work is in progress. 

---