(129d) Chemical Transformations Via Electrochemical Reduction of Organics to Products

In recent years
there has been growing interest in using electrochemical methods for
hydrogenation of biomass-derived molecules for the formation of biofuels.
Conventional hydrogenation processes require moderate temperatures between 433
and 678 K, pressures of 14,000 kPa, and external sources of dihydrogen for the
stabilization and formation of biofuels. However, the same reaction can be
performed using electrochemical reactors and much lower temperatures and
pressures (293 K and 101 kPa) and with no supplied molecular hydrogen.

We
have been exploring electrocatalytic upgrading of bio-oils. Our initial tests
successfully decreased the carbonyl and aromatic species; however, the
fundamental chemistry was not well understood. In this work, we
use a combined experimental–theoretical approach to identify atomistic features and parameters
that affect the electrochemical hydrogenation (ECH) rates
of aldehydes, ketones, and carboxylic acids which are typically found in
biocrude. In particular, we evaluated Pd, Rh, Ru, Cu, Co, Ni, and Zn metals
supported on a carbon foam for the organic ECH and correlated the activity with
computationally-derived parameters. We then used these parameters to explain
the dependence of molecule functionality on the ECH rates. Our results show that whereas the
reduction of aldehydes and ketones into alcohols can be achieved
electrocatalytically at normal temperature and pressure, carboxylic acids
cannot be reduced. However, carboxylic acids can be oxidized to alcohols,
olefins, and ketones via Kolbe and non-Kolbe mechanism. Further, we observed
there the turn over frequency (TOF) and faradaic efficiency (FE) depend on
metal, fractional exposure, substrate concentration and half-cell, working vs
reference (WvR), potentials. For example, as shown in Figure 1, the FE and TOF ECH
are the highest when using Pd electrodes with low Pd weight loading (and small
metal particle size). The FE and TOF ECH decreases as the Pd weight loading
(and particle size) increases. Eley-Rideal-type kinetics were used to
understand the structure sensitivity of the Pd electrocatalysts and revealed
that it is related to changes in rate constants and substrate surface coverage.
That is, small Pd nanoparticles have higher ECH rate constant and higher
benzaldehyde surface coverage compared to large Pd particles. Structure
sensitivity was also observed on Pt and Rh-based electrocatalysts, but with
opposite trend. That is, large Pt and Rh nanoparticles were more active than
small nanoparticles. Cu-based catalysts were structure insensitive. 

Figure
1. TOFECH (A) and Faradaic Efficiency (B) as a function of WvR potentials under
normal conditions for the benzaldehyde ECH to benzyl alcohol. Catalysts with
different Pd loadings were used where  
represents 0.05wt% Pd/CF,  
represents 0.10wt% Pd/CF,  
represents 0.50wt% Pd/CF, represents 1.0wt% Pd/CF, and    represents 4.0wt% Pd/CF.

To
better understand the reactions, we used computational theory. Classical
molecular dynamics simulations were performed on the metal’s most stable
surfaces. All atoms of the electrode were fixed in position. The GROMACS
package(83) was used for all classical molecular dynamics (MD) simulations. The
force field’s nonbonded interaction (van der Waals and electrostatics)
parameters for classical MD simulations were fit to match the DFT (with
Grimme’s D3 correction(72)) binding energies to appropriately incorporate
short-range bonded and long-range nonbonded interactions among the organic, solvent,
and electrode molecules. Our computational model showed that the differences in
catalytic activity were most likely due to differences in binding energy. The
results followed the Sabatier principle which indicated that for the highest
activity, the substrate must bond strong enough to be activated, but not so
strongly that it cannot be released after reaction (Figure 2). Therefore computational
modeling using our approach can be used to guide metal selection for the
desired reactant reduction.

Figure 2.  Turn over frequency vs binding energy of
different metal catalysts with furfural and heptanal