M.S. Ide, R.J. Davis / Journal of Catalysis 308 (2013) 50–59
51
showed that supported Au catalysts can be active for 1,4-butane-
diol oxidation in non-aqueous solvent as long as strongly basic
sites are present on the support [7].
tion over Au/TiO2 to deactivation by strongly adsorbed byproducts
[22]. Interestingly, oxidation of HMF over Pt/C did not require ex-
cess base to achieve high selectivity to the diacid, while under sim-
ilar conditions diacid selectivity was still quite low during glycerol
oxidation [8]. Likewise, the oxidation of 1,2-ethanediol and 1,3-
The conversion of glycerol at low temperature over supported
Au, Pt, and Pd at basic conditions does not proceed without an oxi-
dant, such as dioxygen. Interestingly, multiple studies have re-
ported the order of reaction with respect to dioxygen to be
essentially zero for glycerol and 5-hydroxymethyl furfural (HMF)
oxidation in semi-batch and flow reactor studies [5,8–10]. The role
of dioxygen in the mechanism has been discussed previously in the
literature [11–14]. Isotopic-labeling experiments involving the oxi-
dation of glycerol, ethanol, and HMF in basic aqueous solution over
supported Pt and Au showed that when 18O2 was used as the oxi-
dant no labeled oxygen was incorporated into the acid product
[5,15]. However, when H218O was used as the solvent one or more
labeled oxygen atoms were incorporated into the acid product.
Analogous labeling experiments involving the aqueous phase oxi-
dation of glycerol and HMF over supported Pt in the absence of
added base confirmed that water was the source of the oxygen
atoms incorporated into the acid product [5,15]. Thus, the mecha-
nism for aldehyde oxidation to the acid was proposed to proceed
through reversible hydration to form a geminal diol followed by
dehydrogenation of the diol to form the acid.
propanediol to their corresponding
highly selective over Au/C, but no production of the diacid was ob-
served, even when -hydroxy acid was used as the substrate
x-hydroxy acid products was
x
[23,24]. A low selectivity was also observed during 1,2-ethanediol
oxidation over Pt/C and Pd/C at basic conditions [25].
Clearly, there are many unresolved issues in the area of metal-
catalyzed oxidation that need to be examined. In this work, the
rates of alcohol oxidation reactions in aqueous solutions using
dioxygen over supported Pt catalysts were investigated at low
pH to avoid the need for eventual product neutralization. The influ-
ence of acid type (mineral vs. organic) on the rate of reaction was
also determined. In addition, the effects of carbon support and Pt
particle size on the rate were explored. The oxidation of various
a,x-diols was evaluated over carbon-supported Pt catalysts and
compared to the oxidation of analogous mono-alcohols. Finally,
the influence of dioxygen pressure and substrate concentration
on rate was determined at two different start-up conditions so that
a mechanistic model could be proposed for alcohol oxidation at
acidic conditions over supported Pt.
Although dioxygen did not provide oxygen atoms that incorpo-
rate into the acid product during alcohol oxidation, it was still re-
quired for the reaction to proceed. During glycerol oxidation over
supported Pt, Pd, and Au, the presence of hydrogen peroxide was
detected in the product solution [4,16]. The presence of peroxide
indicates dioxygen is partially reduced during the reaction. In this
way, dioxygen serves as a scavenger of electrons deposited into the
metal catalyst during the alcohol oxidation cycle. Supporting evi-
dence from DFT calculations showed that reduction of dioxygen
to hydroxide is energetically feasible in these systems [5]. Careful
measurement of the solution pH during alcohol oxidation suggests
that some of the hydroxide consumed in the reaction is regener-
ated by dioxygen reduction [8]. The combination of experimental
and theoretical results indicates the primary role of dioxygen in
the oxidation reaction is to scavenge electrons from the metal cat-
alysts, whereas the water and hydroxide are critical to the transfor-
mation of the alcohol to the acid. In fact, electrochemical
experiments involving glycerol oxidation in high pH electrolyte
over a gold electrode revealed the same distribution of products
as those reported over Au catalysts with dioxygen [17]. In the case
of the electrochemical experiments, electrons generated at the gold
electrode are removed via an external circuit so that dioxygen was
not required for oxidation to proceed.
2. Materials and methods
2.1. Catalyst preparation
The 2.69 wt% Pt/C was obtained from Aldrich Chemical Co. The
2.69% Pt/C catalyst was reduced in 100 cm3 minꢀ1 of flowing H2 for
4 h at 473 K, cooled, exposed to air, and stored at ambient temper-
ature. A portion of the 2.69% Pt/C catalyst was reduced similarly for
4 h at 673 K. The activated charcoal (Norit, SX ultra) supported Pt
(3 wt%) catalyst was prepared by incipient wetness impregnation
of chloroplatinic acid (Sigma–Aldrich). The 3% Pt/NoritC was dried
in air for 12 h at 393 K, reduced in 100 cm3 minꢀ1 of flowing H2 for
4 h at 673 K, cooled, exposed to air, and stored at ambient temper-
ature. A Vulcan (Cabot, VXC72R) supported Pt (1 wt%) catalyst was
prepared in a similar manner, except the 1% Pt/VulcanC was re-
duced for 4 h at 573 K.
2.2. Catalyst characterization
The metal dispersion of the Pt catalysts was determined by H2
One motivation to study alcohol and aldehyde oxidation reac-
tions is to produce diacids, which are important monomer com-
pounds. For example, adipic acid is a 6-carbon diacid used in the
production of nylon-6,6 that can be derived from renewable re-
sources, such as fructose. Dehydration of fructose to HMF can be
accomplished with 72% yield and selective hydrogenolysis of
HMF can produce 1,6-hexanediol in 86% yield [18,19]. The subse-
quent oxidation of 1,6-hexanediol forms the desired adipic acid.
Currently, adipic acid is produced from a cyclohexanol mixture
(KA oil) via nitric acid oxidation, which is a process that contrib-
utes considerably to nitrous oxide emissions [20,21]. Recent work
indicates the selectivity to diacid products during both glycerol
and HMF oxidation is very low (<20%) over supported Au at 2:1
or 4:1 M ratios of sodium hydroxide to substrate concentration.
However, the use of excessive amounts of base (20:1 M ratios) al-
lowed high selectivities to diacid products (>70%) from HMF and
glycerol [8,9]. The use of high concentrations of base is likely to
be unrealistic for an industrial process because of the cost of neu-
tralizing the product stream to recover the free acid. Also, a recent
study attributed the lack of diacid selectivity during glycerol oxida-
chemisorption using a Micromeritics ASAP 2020 automated
adsorption analyzer. The supported Pt catalysts were heated to
473 K at 4 K minꢀ1 under flowing H2 (GT&S 99.999%) and reduced
for 2 h. The samples were then evacuated and held for 2 h at 473 K
before being cooled to 308 K for analysis in the pressure range of
10–450 Torr. The amount of metal on the surface was evaluated
by the total amount of H2 adsorbed, extrapolated to zero pressure,
assuming a stoichiometry (H/Ptsurf) equal to unity.
Elemental analysis (using ICP–AES analysis performed by Gal-
braith Laboratories, 2323 Sycamore Drive, Knoxville, TN 37921)
determined a Pt loading of 2.69 wt% for Pt/C (Aldrich).
The X-ray diffraction (XRD) patterns were recorded on a Scintag
automated diffractometer using Cu K
a radiation (40 kV, 30 mA)
and continuous scanning of 2h from 20° to 80° with a step
size of 0.05° at a rate of 0.3° minꢀ1. The crystalline size of
the metal particles was calculated by the Scherrer equation
[d = Kk(b1/2cos(2h))ꢀ1] using the Pt(111) peak at 2h = 39.7°.
To prepare the 2.69% Pt/C sample for transmission electron
microscopy (TEM), ꢁ1 mg of sample was suspended in 2 cm3 of
ethanol by agitating the mixture for 30 min in a sonication bath.