Improvement of the Structural Model for the M1 Phase Mo–V–Nb–Te–O Propane (Amm)oxidation Catalyst

Abstract

An improved structural model for the M1 phase in the Mo-V-Nb-Te-O propane (amm)oxidation catalyst has been refined after accounting for a molybdenumsubstituted-V 2 O 5 impurity and by making adjustments based on aberration-corrected imaging results. The newly refined unit cell has Pba2 symmetry with a = 21.134(1) Å , b = 26.647(1) Å , c = 4.0140(2) Å , and Z = 4, in good agreement with our earlier findings (DeSanto et al. Top Catal 23:23 [20], DeSanto et al. Z Kristallogr 219:152[22]). From the newly refined occupancies, the formula unit is {TeO} 0.86(1) ÁMo 7.48(6) V 1.52(6) NbO 28 . As in the earlier models, V is concentrated in sites that link the pentagonal rings of M1. Careful analysis of bond valences, in combination with the electroneutrality constraint, suggest that the linking sites S3, S4, and S7 all have mixed Mo/V occupancies and valences (d 1 /d 0 ). Furthermore, these sites may contain a mix of Mo 5? and V 5? , which is consistent with the proposed catalytic mechanism in which V 5? plays an important role in propane activation.

The Mechanism of Alkane Selective Oxidation by the M1 Phase of Mo–V–Nb–Te Mixed Metal Oxides: Suggestions for Improved Catalysts

“…[10,11] Assuming that the S13 site is empty and that the S12 sites are fully occupied, one single unit cell of M1 still has 160 atoms. [4] Therefore it is computationally impractical to use a surface model containing more than one layer. Fortunately, the experiments show that M1 has a distinct layered structure along the c axis, with two adjacent layers connected only through coordinate bonds, M=O---M=O motifs (M = Mo, V, Nb and Te).…”

“…Fortunately, the experiments show that M1 has a distinct layered structure along the c axis, with two adjacent layers connected only through coordinate bonds, M=O---M=O motifs (M = Mo, V, Nb and Te). [4] This makes the preparation of the one-layer surface model from the crystal structure unambiguous by breaking only the longer M---O coordinate bond, while keeping shorter M=O intact. The surface is arranged in such a way that oxygen of the M=O bonds in sites S2, S5, S6, S8, S10, and S11 point toward the same direction as that of Te=O at S12, while those of M=O in sites S1, S3, S4, S7, and S9 point toward the opposite direction.…”

“…Note that we assume that Te is at both S12 sites, although this site is only 71.1 % occupied by Te (colored by blown), and we assume S13 is empty although experimentally it is 14.6 % occupied. [4] Besides those sites, S9 is occupied by Nb (colored by green), while all other sites are occupied by Mo (colored by pink). Polyhedrons with violet labels mean that their oxo's point down, whereas those with white labels mean that they oxo's point up.…”

“…[5] Secondly, eight of the 13 distinct sites exhibit partial occupation by either molybdenum or vanadium, while two other sites are only partially occupied, making it difficult to design an unambiguous surface model. [4] To do this we should enlarge the unit cell sufficiently that every site can be occupied by whole atoms. Indeed we did this for the M2 phase using the ReaxFF reactive force field to find the optimum configuration, which required a 2×3×4 supercell.…”

“…[1][2][3] In section 4, we use quantum mechanics (QM) to examine the activation of propane by the M1 phase, where we use the improved experimental analysis of the partial occupations of the various sites. [4] Here we examine how the activation energy for propane depends on the specific distributions over the sites. [5] We conclude that critical to activation of propane is to have Te=O at both S12 sites of the S12-S2-S12 triad site coupled with two or more V at S2, S4, and S7 sites.…”