(700d) Metal-Modified Carbides for Conversion of Alcohols to Syngas and Aldehydes

Metal-Modified Carbides
for Conversion of Alcohols to Syngas and Aldehydes

Thomas G. Kelly and Jingguang G. Chen¹*

¹Center for Catalytic Science and Technology,
Department of Chemical and Biomolecular Engineering,
University of Delaware, Newark, Delaware 19716 (USA)

tgkelly@udel.edu, *jgchen@udel.edu

In the search for energy sources to replace fossil
fuels, biorenewable fuels have become a distinct
possibility.  One class of biorenewable molecules are small-chain alcohols.  These alcohols can be reformed to produce syngas or thermochemically
converted to useful products.  Recently,
3d/Pt(111) bimetallic surfaces have been shown to be
active for methanol and ethanol reforming. 
The Ni/Pt(111) surface displayed higher
reforming yield than both of the parent metals [1,2].  However, Pt-based bimetallics
face the problems of Pt scarcity and poor overlayer
stability.  Metal carbides such as
tungsten carbide (WC) and molybdenum carbide (Mo₂C) have been shown
to exhibit Pt-like properties [3].  Also,
WC can be used as a less expensive and more stable support in place of Pt, as
was shown in using Ni/WC to replace Ni/Pt for ethanol decomposition [4].  In the current study, we have extended our
previous surface science studies for methanol on Pt-modified WC [5] to C1-C3
alcohols on carbides modified with Ni, Rh, Cu, and Au.  Furthermore, an understanding of bond scission
sequence should also provide insights into the utilization of carbides for
catalytic conversion of alcohols.

Density Functional Theory (DFT) was used to calculate
the binding energies of alcohols and relevant intermediates on model surfaces
of WC and Mo₂C, which showed the possibility to correlate reactivity
with binding energy. Temperature-programmed desorption (TPD) was used on model
WC and Mo₂C surfaces  modified
with different metals to quantitatively determine the activity of the surfaces
and their selectivity towards C1-C3 alcohol decomposition.  High resolution electron energy loss
spectroscopy (HREELS) identified different reaction intermediates on carbide
and metal-modified carbide surfaces.  The
first step in alcohol decomposition on these surfaces was O-H bond cleavage to
form an adsorbed alkoxy species.  In the case of methanol, the predominant
reaction product on clean WC was methane; adding small amounts of Ni or Rh promoted
the C-H bond scission and shifted the selectivity towards CO and H₂.  Au simply acted as a site blocker since
activity decreased with increasing Au coverage [6].  Similar experiments have been conducted for
ethanol and propanol decomposition on Rh/WC to investigate the pathway for C-C
bond scission.  The selectivity on clean
WC was for C-O bond scission to produce ethylene.  Adding Rh led to C-H and C-C bond cleavage;
HREELS showed that the C-C bond in ethanol is broken by 200 K.  Lastly, the clean Mo₂C surface was
shown to completely decompose ethanol to its constituent elements with a small
amount of ethylene and methane production. 
Ni-modified Mo₂C produced CO as the major product, indicating
C-C bond scission.  Similarly to methanol
on Au-modified WC, ethanol desorbed intact on Au-modified Mo₂C. Interestingly,
a pathway divergent from reforming was observed on Cu-modified Mo₂C,
which produced acetaldehyde.  Apparently
Cu was efficient at breaking the O-H and α C-H bonds but further
dehydrogenation was not favored over desorption. 

Of further interest is the bond scission sequence of additional
C2 and C3 oxygenates on metal-modified WC. 
HREELS experiments on WC and Rh/WC were conducted for ethanol,
acetaldehyde, propanol, and propanal
to investigate the bond scission sequence in these oxygenates.  On WC, ethanol followed a decomposition
pathway similar to acetaldehyde; the spectra suggested that both molecules
formed a di-σ species that facilitates C-O bond cleavage.  This behavior was mirrored by propanol and propanal on WC.  On Rh/WC, ethanol and acetaldehyde again
followed similar pathways, except that the pathway in this case is through C-C
bond scission instead of C-O scission. 

The results of decomposition for C1-C3 alcohols on carbides
hold promise for applications in heterogeneous reforming.  On native carbides, C-O scission occurred to
produce hydrocarbons.  Ni and Rh
supported on carbides have been shown to be efficient at cleaving C-H and C-C
bonds of alcohols while leaving the C-O bond intact.  Alternatively, modifying Mo₂C with
Cu resulted in acetaldehyde production from ethanol.  The ability to selectively decompose alcohols
by using various admetals demonstrates the
versatility of metal-modified carbides. 
Future experiments will involve synthesis of supported
metal-modified carbide catalysts and reactor testing for alcohol decomposition.

References

1.     
Skoplyak, O., Barteau, M.A., Chen,
J.G. J. Phys. Chem. B. 110 (2006) 1686.

2.     
Skoplyak, O., Menning, C.A., Barteau, M.A., Chen, J.G. J. Chem. Phys. 127 (2007)
114707.

3.      Chen, J.G. Chem. Rev. 96 (1996) 1477.

4.      Ren, H., Hansgen, D.A., Stottlemyer A.L., Kelly, T.G., Chen, J.G. ACS Catal. 1 (2011) 390.

5.      Stottlemyer, A.L., Liu, P., Chen, J.G. J. Chem. Phys. 133
(2010) 104702.

6.      Kelly, T.G., Stottlemyer,
A.L., Ren, H., Chen J.G. J. Phys. Chem. C. 115 (2011)
6644.