ACS Catalysis
Research Article
equipped with a thermal conductivity detector. Approximately,
the Pd/Al O catalyst and an α-Al O diluent were finely
2
3
2
3
50 mg of catalyst was loaded into a quartz U-tube reactor and
ground to a similar consistency prior to mixing, which was
found to yield consistent results in repeated measurements
with different catalyst beds. (Error bars for different catalyst
beds are reported in the results). Both the feed streams and
reactor effluents were analyzed using an SRI Instruments
8610C gas chromatograph equipped with a Restek MXT-5
capillary column and a flame ionization detector. Each reaction
experiment was run to steady state, which was defined here as
running for at least 150 min and having a conversion within
held as a fixed bed on a plug of quartz wool. The sample was
−
1
pretreated in 4% H /Ar flowing at 50 mL min by heating at
1
2
−
1
0 °C min to 180 °C and holding for 2 h. After the
pretreatment, the sample was cooled to 30 °C and flushed with
He flowing at 50 mL min− for 10 min to remove any weakly
adsorbed hydrogen. The sample was then exposed to
sequential 500 μL pulses of a 10% CO/He gas mixture until
saturation. A 500 μL sample loop was used to calibrate the
TCD response. The surface area of the 1% Pd/Al O catalyst
1
±
0.5% over four consecutive measurements, taken over a
2
3
period of 1 h. Final conversions of FA and 2MF were 5−15
and 15−30%, respectively, except for C18SH-modified
catalysts which had conversions below 5% for both reactants.
Average conversions and selectivities determined from the final
steady state measurements were used to calculate the reported
reaction rates.
was found to be approximately 49 ± 4 μmol/g, corresponding
to a dispersion estimated at 53% ± 5%.
Diffuse reflectance infrared Fourier transform spectroscopy
(
DRIFTS) was performed on powder catalyst samples using a
Thermo Scientific Nicolet 6700 FT-IR with a Harrick
Scientific closed cell attachment. Spectra were collected in
the hydrocarbon stretching region, using 100 scans and a
2.4. DFT Calculations. Density functional theory calcu-
−
1
lations were performed using a plane-wave basis set as
resolution of 4 cm , in order to confirm the identity of alkyl
tails in the surface modifiers. Spectra of the unmodified Pd/
45
implemented in the Vienna ab initio simulation package.
The generalized gradient corrected PBE exchange−correlation
Al O catalyst were used as the background for all experiments
2
3
46
functional was used, and the ion-electron interactions were
so that the reported spectra of the modified catalysts represent
only changes due to surface modification.
47
described using projector augmented-wave potentials with an
energy cutoff of 400 eV. Calculations were also performed
using the optPBE-vdW functional to account for van der Waals
interactions, and the corresponding energies are reported in
Hydrogen−deuterium exchange experiments were per-
formed to measure the effect of each surface modifier on the
catalytically active surface area for hydrogenation. Approx-
imately 0.5 mg of catalyst, along with α-Al O as a diluent, was
48
enclosed square brackets following the PBE values. The
Pd(111) surface was modeled by a (3 × 3) unit cell,
periodically repeated in a supercell geometry with successive
four-layer slabs separated by 20 Å vacuum. The surface
Brillouin zone was sampled using a 5 × 5 × 1 Monkhorst−
Pack mesh. Surface adsorbates and atoms in the top two layers
of the metal slab were allowed to relax whereas atoms in the
bottom two layers were fixed in the bulk truncated positions.
Optimizations were considered complete when forces fell
below 0.05 eV/Å. The periodic Pd slab was modeled using an
optimized lattice constant of a = 3.96 Å which closely matches
2
3
loaded into a Pyrex tube, packed bed, continuous-flow reactor
at atmospheric temperature and pressure. The effluent of the
reactor was analyzed with a Pfeiffer Vacuum Prisma 80
quadrupole mass spectrometer. Samples were first purged in a
−
1
1
5 mL min Ar stream for 10 min, followed by a mixture of 15
−
1
−1
mL min Ar and 5 mL min H for at least 10 min. Next,
2
while continuing to flow the Ar/H mixture, three separate
2
injections of 0.2 mL D were slowly introduced to the feed
2
stream and the resulting D signals were measured. As a
2
control, the same feed stream was then allowed to bypass the
reactor, and three injections of 0.2 mL D2 were again
49
the experimental value of 3.89 Å. Binding energies (BE)
reported on Pd(111) were calculated relative to the energy of
the clean Pd(111) slab (EPd(111)) and the energy of the
adsorbed species in the gas phase (Egas)
performed. The amount of D detected in the reactor effluent
2
with and without the presence of the catalyst was used to
calculate the D conversion, used as a measure of the H /D
2
2
2
2
exchange activity. Due to the large excess of H relative to D
BE = Etotal − EPd(111) − Egas
2
(
∼10:1 M ratio), the conversions measured in these pulse
where Etotal is the total energy of the slab with the adsorbed
species. The differential BEs of FA and 2MF on the PA-
modified Pd surfaces were calculated relative to the energy of
experiments are far from equilibrium. Data reported here were
measured at conversions below 70%, while the use of larger
catalyst masses allowed for conversions >95%, suggesting that
the reverse reaction rate of H /D scrambling was negligible.
the Pd(111) slab with the specified PA adsorbate (EPA/Pd(111)
and the energy of the furanic species in the gas phase
)
2
2
Experiments using only α-Al O and no Pd catalyst did not
2
3
yield any observable conversion of D2.
differential BE = Etotal − EPA/Pd(111) − Egas
2
.3. Reactor Studies. Hydrogenation reactions were
performed on FA and 2MF using a Pyrex tube, packed bed,
continuous-flow reactor under atmospheric pressure. Helium
was bubbled through the desired liquid reactant and held in a
temperature-controlled bath (FA at 60 °C, 2MF at −10 °C),
where Etotal is the total energy of the slab with FA or 2MF
coadsorbed with the PA species.
3. RESULTS AND DISCUSSION
and the resulting stream was mixed with H and make-up He
2
3
.1. Catalyst Characterization. The uncoated Pd/Al O
2 3
upstream of the reactor. The reactor feed streams had a gas-
phase mole fraction of YH2 = 0.10 and YH2 = 0.15 for reactions
(
1 wt %) catalyst was modified with either one of two thiols or
one of two PAs to probe the effect of the modifier head group
on reactivity. The thiols, 1-octadecanethiol (C18SH) and 1-
adamantanethiol (AT), were selected because they have been
shown to form well-ordered SAMs with considerably different
of FA and 2MF, respectively. All reactions were run using a
3
0:1 M ratio of H /reactant at a temperature of 180 °C. The
2
mass of the catalyst used for each reaction was varied as
needed (typically between 0.3 and 1.2 mg) to achieve a desired
conversion. Due to the low mass of the catalyst required, both
33,34,50,51
coverage on Pd catalysts.
More specifically, the bulkier
tail group present in AT results in a more sparsely packed
3
732
ACS Catal. 2021, 11, 3730−3739