(689f) Reactivity and Selectivity Descriptors for the Activation of C-H Bonds in Hydrocarbons and Oxygenates on Metal Oxides

Reactivity and Selectivity Descriptors
for the Activation of C-H Bonds in Hydrocarbons and Oxygenates on Metal Oxides

Prashant Deshlahra and
Enrique Iglesia*

 Department
of Chemical Engineering, University of California, Berkeley, CA 94720

*iglesia@berkeley.edu

C-H bond activations at
lattice O-atoms on oxides mediate some of the most important chemical
transformations of small organic molecules [1-2]. The relations between
molecular and catalyst properties and C-H activation energies are
discerned in this study for the diverse C-H bonds prevalent in C₁-C₄
hydrocarbons and oxygenates using lattice O-atoms of polyoxometalate (POM)
clusters with a broad range of H-atom abstraction properties. These activation
energies determine, in turn, attainable selectivities
and yields of desired oxidation products, which differ from reactants in their
C-H bond strength. Brønsted-Evans-Polanyi (BEP) linear scaling relations [3-4]
predict that C-H activation energies depend solely and linearly on the C-H bond
dissociation energies (BDE) in molecules and on the H-atom addition energies
(HAE) of the lattice oxygen abstractors. These relations omit critical
interactions between organic radicals and surface OH groups that form at
transition states that mediate the H-atom transfer, which depend on both
molecular and catalyst properties (Fig. 1a); they also neglect deviations from
linear relations caused by the lateness of transition states. Thus, HAE and BDE
values, properties that are specific to a catalyst and a molecule in isolation,
represent incomplete descriptors of reactivity and selectivity in oxidation
catalysis (Fig. 1b). These effects are included here through crossing potential
formalisms that account for the lateness of transition states in estimates of
activation energies from HAE and BDE and by estimates of molecule-dependent,
but catalyst-independent, parameters that account for diradical interactions
that differ markedly for allylic and non-allylic C-H bonds (Fig. 1c). The
systematic ensemble-averaging of activation energies for all C-H bonds in a
given molecule show how strong abstractors and high temperatures decrease an
otherwise ubiquitous preference for activating the weakest C-H bonds in
molecules, thus allowing higher yields of products with C-H bonds weaker than
in reactants than predicted from linear scaling relations based on molecule and
abstractor properties [5].  Such
conclusions contradict the prevailing guidance to improve such yields by softer
oxidants and lower temperatures, a self-contradictory strategy, given the lower
reactivity of such weaker H-abstractors. The diradical-type interactions, not
previously considered as essential reactivity descriptors in catalytic
oxidations, may expand the narrow yield limits imposed by linear free energy
relations by guiding the design of solids with surfaces that preferentially
destabilize allylic radicals relative to those formed from saturated reactants
at C-H activation transition states.

Figure
1.
(a) A thermochemical cycle
description of the C-H activation transition-state energy relative to a gaseous
organic reactant and a surface metal oxide site (MO*) as a sum of, (i) the energy required to separate H atom from the C atom
in the reactant to form a radical species (C-H BDE), (ii) the energy of adding
the H atom to the O atom of MO* (HAE), (iii) and interaction energy between the
radical (R•) and hydroxylated metal oxide (•MOH*) at the transition state (
). (b) DFT-derived
C-H bond activation energies
in alkanes (closed symbols), alkanols (open symbols), alkenes (half-filled
symbols) and alkanals (crossed symbols) at a specific O-atom location in H₃PMo₁₂O₄₀
POM cluster as a function of their C-H
bond dissociation energy (BDE). (c) DFT-derived
C-H bond activation energies as a function of the sum of the C-H BDE, of
reactants, the HAE of different abstractors and the product state interaction
energy (
). Dashed curves represent regressed values of the activation
energies to the functional form described by C-H O-H crossing-potential models
for C-H activation.

Acknowledgments

This
work was supported by the U.S. Department of Energy, Office of Science, Office
of Basic Energy Sciences, under contract number DE-AC05-76RL0-1830.
Computational facilities were provided by the Environmental Molecular Science
Laboratory (EMSL) at Pacific Northwest National Laboratory (PNNL), a DOE Office
of Science User Facility, under proposal number 48772.

References

[1]
Mamedov, E. A.; Corberán, V. C. Appl. Catal. A: General 1995, 127, 1-40.

[2]
Labinger, J. A.; Bercaw J. E. Nature 2002, 417, 507-514.

[3]
Bronsted, J. N. Chem. Rev. 1928, 5,
231-338.

[4]
Evans, M. G.; Polanyi, M. Trans. Faraday Soc. 1938, 34, 11-24.

[5]
Zboray, M.; Bell, A. T.; Iglesia,
E. J. Phys. Chem. C 2009, 113 12380-12386.