Reliability of Small Cluster Models for Cu-Exchanged Zeolites

Abstract

The predictions of a previously developed small cluster model for bound Cu ions and their interactions with CO and NO in Cu-exchanged zeolites [Schneider, W. F.; Hass, K. C.; Ramprasad, R.; Adams, J. B. J. Phys. Chem. 1996, 100, 6032] are compared to those of a series of larger, more realistic cluster models. The hypothetical case of Cu n + (n = 0, 1, 2) coordinated to a single framework oxygen atom, described in the simplest model as CuOH2 n +, is first examined in detail. The larger clusters considered include explicit Si or Al tetrahedral sites and are terminated by either H or OH. For nominal Cu(0) and Cu(I) oxidation states, the qualitative predictions of the original “water−ligand” model are found to be extremely robust, as are quantitative aspects of CO and NO binding. More pronounced differences between models are observed for a nominal Cu(II) oxidation state, due primarily to the tendency of the zeolitic portions of the larger clusters to reduce the Cu. The coordination of Cu n + (n = 1, 2) to a more realistic 4-fold coordination site model is similarly examined and compared to the results for Cu(OH2)4 n +. Excellent agreement is obtained between the water−ligand and cluster models for both Cu(I) and Cu(II) in this case. The results indicate that proper treatment of the local coordination environment of a Cu ion exchanged into a zeolite is of greater importance for describing chemistry on the Cu site than is use of a “realistic” model for an individual bridge oxygen site. The implications of these results for future modeling studies of Cu zeolites are discussed.

Adsorption of Nitrogen Monoxide and Carbon Monoxide on Copper-Exchanged ZSM-5: A Cluster and Embedded Cluster Study

“…The majority of Cu + ions in ZSM-5 are expected to coordinate to two bridging oxygen atoms of the active site. 18,29,30,33,45,50,72 All structural parameters of our model, i.e., bond distances and the coordination number of the copper ion, are comparable to experimental results. Particularly, the calculated Cu-O bond distances of 2.013-2.032 Å (see Table 1) are in excellent agreement with the experimental data of 2.00 ( 0.02 Å, determined by Lamberti et al 8 In comparison with previous theoretical studies, our calculated Cu-O bond distances are slightly larger.…”

“…Note that previous calculated binding energies of NO adsorption are very scattered, ranging from 17 to 52 kcal/mol, as shown in Figure 6. 18,29,30,34,44,46,47,78,79 The NO binding energy estimated by Chakraborty et al 18 from TPD results of Li and Armor is 43.9-49.5 kcal/mol and that estimated from Schay and Guczi's results is 47.6-49.5 kcal/mol. These experimental estimations are much higher than our prediction for the NO binding energy and higher than the experimental CO binding energy as well.…”

“…In particular, Hass and Schneider used the simple 5T model at the DFT/LSDA level of theory and obtained the binding energy of 175 kcal/ mol. 29 Sauer and co-workers 45 determined the Cu + binding energy of 160 kcal/mol from a 3T model at the DFT/B3LYP level of theory. Sauer et al 39 also predicted a value of 148.6 kcal/mol from their mechanical embedded 3T cluster model (QM-pot) at the same level of theory.…”

“…42 This red shift corresponds to the lengthening of the NO bond, observed in our study and other theoretical works. 27,29,32,46,47 The stretching frequency of an adsorbed NO was calculated using our embedded 3T model to be 1921.7 cm -1 , red-shifted from 1990.4 cm -1 of the gas-phase NO. Our calculated binding energy of NO to the Cu-ZSM-5, including the BSSE and ZPE corrections, is about 22 kcal/mol.…”

“…24,43 Adsorption of CO and NO have also been the subject of numerous theoretical studies. 18,[27][28][29][30]34,35,[44][45][46][47][48][49][50] The calculated results for adsorption energy, frequency shift, geometry, and the coordination number are still scattered due to differences in the topological structure of the model, cluster size, constraints, basis sets, and the levels of theory employed. Almost all of the previous theoretical models, except the Cu-water cluster, were not able to predict the experimental blue shift of adsorbed CO.…”