Y.B. Yin, et al.
CatalysisCommunications135(2020)105895
(8 × 40 mm, VWR), and analyzed by ion-exclusion high-performance
liquid chromatography (HPLC).
A
Shimadzu Prominence SIL 20 system (Shimadzu Scientific
Instruments, Inc., Columbia, MD, USA) equipped with an HPX-87H
organic acid column (Bio-Rad, Hercules, CA, USA), a refractive index
detector (RID), and a UV detector (210 nm) was used to separate and
detect the reaction products. The operation conditions for the Shimadzu
HPLC were at 42 °C with 30 mM H2SO4 as mobile phase flowing at
0.3 cm3 min−1
.
Pure reactant MG and reaction products (MGA, glyceric acid, oxalic
acid, glycolic acid, formic acid, acetic acid, and lactic acid) standards
were used to determine retention times, and concentration-peak area
calibration curves were prepared in the range of 0 to 0.2 M. A slight loss
in liquid volume due to water evaporation was observed during the
reaction (~0.65 mL volume loss after 1 h at 50 °C), for which the
measured concentrations were corrected during calculation. pH was
checked before and after reaction, and no significant pH change was
observed (< 0.4).
Scheme 1. Direct synthesis of glucuronic acid via biocatalyzed glucose oxida-
tion (red path), and indirect synthesis of glucuronic acid via metal-catalyzed
MG oxidation (blue path).
Bimetallic catalysts (e.g. PdPt, PdAu, and PtAu) have enhanced
activity, selectivity and/or stability compared with monometallic cat-
alysts for aqueous alcohol oxidation, which is commonly attributed to a
combination of geometric and electronic effects [17,18]. PdAu is one of
the most widely studied bimetallic catalysts for primary alcohol oxi-
dation. Silva et al. reported > 5 times improvement in activity using
Au@Pd core-shell NP catalysts compared to monometallic Au and Pd
for the oxidation of benzyl alcohol, for example [19]. Hutchings's group
prepared carbon-supported PdAu alloy NPs catalysts and found they
gave ~80% selectivity to glyceric acid from glycerol oxidation and
were 50% more active than supported monometallic Pd and Au NPs
[20]. Our group has investigated structure–property relationships of
Pd-on-Au NPs for the oxidation of glycerol; the catalytic NPs exhibited
volcano-shaped activity dependence on Pd surface coverage, and had
~10 times higher activity than monometallic Pd or Au catalysts
[21,22]. Also, these Pd-on-Au NPs showed improved resistance to de-
activation from Pd over-oxidation [21,23].
The conversion of MG (XMG) was defined as
CMG,0 − CMG
XMG
=
CMG,0
(1)
where CMG,0 is the initial MG concentration and CMG is the MG con-
centration at each sampling time. The MG oxidation reaction over the
whole reaction time (8 h) cannot be modeled as pseudo first-order re-
action since the generated products can also react on the catalyst sur-
face competing with the MG oxidation. Due to the low amount of re-
action products at beginning of reaction, MG oxidation was the
dominant reaction. Thus, initial first-order reaction rate constants were
calculated using data collected in the first 2 h:
dCMG
dt
= kmeas × CMG
(2)
The initial apparent first-order reaction rate constant kmeas (with
units of h−1) was obtained from the following equation:
In this work, we explored the oxidation of MG using Pd-on-Au NP
model catalysts to generate MGA as a precursor to glucuronic acid
(Scheme 1). We studied if these materials also exhibited volcano-shape
oxidation activity dependence on Pd surface coverage. We quantified
MGA selectivity and yield in relation to Pd surface coverage.
CMG
CMG,0
⎛
⎜
⎞
⎟
= k
× t
meas
(3)
⎝
⎠
Initial turnover frequency, normalized to the estimated moles of
exposed catalytic surface atoms, (TOF, with units of mol-MG mol-sur-
2. Experimental
face-atom−1 −1) was given by:
h
2.1. Catalyst preparation and characterization
kmeas × CMG
TOF =
Cmetal
(4)
The carbon-supported NP catalysts were synthesized as reported
previously [21,24]. The detailed material usage, preparation proce-
dures and catalyst characterization were provided in Supplementary
information.
where Cmetal is the surface metal content of the catalysts. The amount of
surface metal was estimated using the magic cluster model [25–28],
where the 4 nm carbon-supported monometallic Au and Pd NPs were
modeled as clusters with 7 shells of atoms, and the surface atoms were
estimated to be the atoms in the 7th shell of the cluster. For carbon-
supported Pd-on-Au NPs with Pd surface coverages < 100 sc%, all Pd
atoms were assumed to be surface atoms in the 8th layer and uncovered
Au atoms in the 7th layer were also counted as surface atoms. For Pd
surface coverages > 100 sc%, the surface atoms were the Pd atoms in
the 9th layer and the exposed Pd atoms in 8th layer. The charge
amounts of catalysts with different surface coverage to reactor are
shown in Table S4.
2.2. Catalytic testing
MG oxidation was conducted using a screw-cap bottle (100 mL,
Alltech) as a semi-batch reactor. The bottle was capped by a te-
flon‑silicone septum with one stainless steel needle inserted into the
liquid as oxygen inlet, and a shorter one above the liquid as gas outlet.
MG was added to 50 mL DI water at a concentration of 0.1 M. The
solution pH was adjusted to 13 by adding 80 mg NaOH to the solution.
The reactor was placed in an oil bath on a heating plate to control the
temperature at 50 °C and magnetically stirred. After the temperature of
liquid stabilized (~30 min), O2 flow (100 mL min−1) was started, the
stirring rate was increased to 1200 rpm, and 10 mg catalyst was added
to start the reaction.
To quantify the percentage of reactant MG converted to the desired
product MGA for different surface coverage catalysts, yield (YMGA) of
MGA using each catalyst was defined as:
CMGA
CMG
YMGA
=
(5)
Aliquots of the reaction fluid (1 mL) were periodically withdrawn
by a 5 mL plastic syringe via a stainless steel needle, filtered by a 0.2 μm
syringe filter (25 mm, VWR), stored in the 1 mL clear shell vials
where CMGA and CMGA is the MGA and MG concentration, respectively.
MGA selectivity (SMGA) was defined as:
2
Yin, Y. Ben
