Selective Oxidation of Glyoxal to Glyoxalic Acid by Air over Mesoporous Silica Supported Pd Catalysts

Article abstract of DOI:10.1007/s10562-019-02799-3

Abstract: A series of mesoporous silica (KIT-6, MCM-41 and SBA-15) supported Pd catalysts were successfully synthesized and applied for selective oxidation of glyoxal. All of these catalysts exhibited significantly higher activity than commercial Pd/C. Among them, Pd/KIT-6 exhibited the best activity and selectivity with 41.3% glyoxal conversion and 57.0% selectivity to glyoxalic acid. The better performance of Pd/KIT-6 was attributed to its three-dimensional mesoporous structure. The three-dimensional mesoporous structure of KIT-6 could enhance Pd dispersion, providing sufficient accessible active sites which improved the conversion of glyoxal. Meanwhile, the better mass transfer capability of Pd/KIT-6 allowed glyoxalic acid to leave the catalyst easily, reducing the probability of over-oxidation. The ratio of k_I (rate constant of initial oxidation reaction) to k_II (rate constant of over-oxidation) was compared among three catalysts. The k_I/k_II of Pd/KIT-6 (0.50) was higher than that of Pd/MCM-41 (0.39) and Pd/SBA-15 (0.34), which reflected its best selectivity from kinetic aspect. Graphical Abstract: [Figure not available: see fulltext.]

Full text of DOI:10.1007/s10562-019-02799-3

Catalysis Letters
https://doi.org/10.1007/s10562-019-02799-3
Selective Oxidation of Glyoxal to Glyoxalic Acid by Air
over Mesoporous Silica Supported Pd Catalysts
1
1
1
1
1
1
1
Junchi Liu · Feng Qin · Zhen Huang · Liang Huang · Zhenan Liao · Hualong Xu · Wei Shen
Received: 3 February 2019 / Accepted: 15 April 2019
©
Springer Science+Business Media, LLC, part of Springer Nature 2019
Abstract
A series of mesoporous silica (KIT-6, MCM-41 and SBA-15) supported Pd catalysts were successfully synthesized and
applied for selective oxidation of glyoxal. All of these catalysts exhibited signiꢀcantly higher activity than commercial Pd/C.
Among them, Pd/KIT-6 exhibited the best activity and selectivity with 41.3% glyoxal conversion and 57.0% selectivity to
glyoxalic acid. The better performance of Pd/KIT-6 was attributed to its three-dimensional mesoporous structure. The three-
dimensional mesoporous structure of KIT-6 could enhance Pd dispersion, providing suꢁcient accessible active sites which
improved the conversion of glyoxal. Meanwhile, the better mass transfer capability of Pd/KIT-6 allowed glyoxalic acid to
leave the catalyst easily, reducing the probability of over-oxidation. The ratio of k (rate constant of initial oxidation reaction)
I
to k (rate constant of over-oxidation) was compared among three catalysts. The k /k of Pd/KIT-6 (0.50) was higher than
II
I
II
that of Pd/MCM-41 (0.39) and Pd/SBA-15 (0.34), which reꢂected its best selectivity from kinetic aspect.
Graphical Abstract
Keywords Selective catalytic oxidation · Glyoxal · Glyoxalic acid · Mesoporous silica supported Pd
*
Wei Shen
1
Introduction
wshen@fudan.edu.cn
1
Glyoxalic acid is a valuable ꢀne chemical and an important
intermediate of numerous products, such as ꢂavor, perfume,
cosmetics and drug modiꢀers, especially in the preparation
of penicillin and vanillin [1–3]. Nowadays, glyoxalic acid is
Department of Chemistry, Shanghai Key Laboratory
of Molecular Catalysis and Innovative Materials, Laboratory
of Advanced Materials, Collaborative Innovation Center
of Chemistry for Energy Materials, Fudan University,
Shanghai 200433, People’s Republic of China
Vol.:(0
1
123456789)
3

J. Liu et al.
industrially manufactured by glyoxal oxidation with nitric
acid. However, this process produces large amounts of nitro-
mesopore favored the mass transfer of reactants, which
helped to reduce the chance of over-oxidation. Consequently,
the dimensionality eꢃect of supports on the catalytic activity
and selectivity was investigated.
gen oxides (NO ), which causes severe potential environ-
x
mental problems. It is therefore urgent to be substituted by
green synthesis route. Direct selective oxidation of glyoxal
to glyoxalic acid, using air as oxidant, is an attractive and
environmentally-friendly route to solve these problems.
Recently, some carbon supported noble metal catalysts, such
as Pd/C, Pt/C and Au/C have been applied in the title reac-
tion as active components [4–7]. Among them, Pd/C exhib-
ited superior activity for selective oxidation of glyoxal under
mild condition. However, there are still some challenges for
Pd/C in this reaction. First, the most active monometallic
Pd/C catalyst exhibits relative low activity with about 15%
yield of glyoxalic acid after 20 h. By bimetallic modiꢀca-
tion, the Ru–Pd/C catalyst can reach around 20% yield of
glyoxalic acid [4, 8, 9]. Besides, the product, glyoxalic acid,
is easily over-oxidized to oxalic acid, which leads to low
selectivity [10]. Therefore, there is a signiꢀcant demand for
development of new catalyst systems to achieve higher activ-
ity and selectivity.
2 Experimental Sections
2.1 Synthesis of MS Supports
Mesoporous KIT-6, MCM-41 and SBA-15 were synthesized
according to the literature [20, 22]. For KIT-6, 4.0 g Pluronic
P123 was dissolved in 144 g water, 4.0 g Butan-1-ol and
7.9 g (35 wt%) hydrochloric acid with stirring at 40 °C. Then
8.6 g tetraethyl orthosilicate was added dropwise and stirred
for 24 h. Subsequently, the mixture was aged at 100 °C for
24 h without stirring. The obtained suspension was ꢀltered
and washed with water, dried overnight before calcination at
550 °C for 6 h. For MCM-41, 4.0 g hexadecyltrimethylam-
monium bromide was dissolved in 200 g water and 15.9 g
(25 wt%) aqueous ammonia with stirring at 25 °C. Then
16.7 g of tetraethyl orthosilicate was added dropwise and
stirred for 1 h. The obtained suspension was ꢀltered, washed
with water and dried overnight before calcination at 550 °C
for 6 h. For SBA-15, 4.0 g Pluronic P123 was dissolved in
127 g water and 85 g of 4 M hydrochloric acid with stirring
at 35 °C. Then 11.3 g of tetraethyl orthosilicate was added
dropwise and stirred for 20 h. The obtained suspension was
ꢀltered and washed with water, dried overnight before cal-
cination at 550 °C for 6 h samples.
Mesoporous silica (MS), as a kind of excellent catalytic
supports with high surface area, ordered pore structure and
ꢂexible pore architecture [11, 12], is widely used in the het-
erogeneous catalysis [13–15]. In the past decades, various
mesoporous silicas with tunable pore size have been synthe-
sized, such as M41S family, SBA series and MSU-X. Their
superior catalytic performance in photocatalysis reaction,
selective oxidation reaction and organic synthesis was veri-
ꢀ
ed [16–18]. Karimi et al. [19] reported Pd immobilized on
SBA-15 for the aerobic oxidation of alcohols. The Pd nano-
particles were well conꢀned inside the channel of SBA-15
and they acted as active species for the aerobic oxidation of
a wide range of alcohols. Afterwards, Parlett et al. [20, 21]
found that Pd dispersed on SBA-16 and KIT-6 exhibited
enhanced aerobic alcohol oxidation performance owing to
their three-dimensional (3D) interconnected architecture.
For the aerobic alcohol oxidation, the 3D architecture of
MS supports led to the obvious enhancement of incorporated
metal dispersion and mass transfer in liquid phase. Since
these investigated oxidation reactions did not face the prob-
lem of severe over-oxidation, their work mainly focused on
the promotion of conversion by the 3D mesoporous structure
of supports.
2.2 Preparation of Pd/MS Catalysts
The Pd (1 wt%)/MS catalysts were prepared by impregnation
method in acetic solution. 0.011 g Pd(OAc) was dissolved
2
in 50 mL acetic solution. The synthesized MS support
(0.50 g) was then added to the above solution and stirred
for 4 h at 50 °C. After that, the solvent in suspension solu-
tion was vaporized under a vacuum in a rotary evaporator.
The obtained precursor was then dried at 90 °C overnight.
Finally, the dried powder was calcined in air at 350 °C for
3 h, followed by reduction at 350 °C for 3 h in hydrogen.
These catalysts were denoted as Pd/SBA-15, Pd/MCM-41
and Pd/KIT-6.
Despite these excellent features of MS, to our knowledge,
there is no report on MS supported Pd catalyst for selective
oxidation of glyoxal to glyoxalic acid by air. Herein, we pre-
pared a series of Pd/MS catalysts (Pd/KIT-6, Pd/MCM-41
and Pd/SBA-15) to investigate their catalytic performance.
Compared with 2D architecture of MCM-41 and SBA-15,
we speculated that the 3D architecture of KIT-6 could pro-
vide highly dispersed active sites, which were more acces-
sible to reactant molecules. Meanwhile, its interconnected
2.3 Characterization
Nitrogen adsorption and desorption isotherms were per-
formed by Micromeritics ASAP2020. Samples were
degassed at 200 °C before measurements. The speciꢀc sur-
face areas were calculated by the BET equation. The pore
diameters and volumes were determined by applying the
1
3

Selective Oxidation of Glyoxal to Glyoxalic Acid by Air over Mesoporous Silica Supported Pd…
Barrett–Joyner–Halenda (BJH) method to the desorption
branch of the isotherms.
O
O
O
O
1
/2 O2
kII
H
OH
HO
OH
Small-angle X-ray scattering (SAXS) measurements
were taken on a Nanostar U small-angle X-ray scattering
glyoxalic acid
(AGLY)
oxalic acid
(OX)
O
O
system (Bruker, Germany) using Cu radiation (40 kV,
Kα
H
H
3
5 mA). Wide-angle X-ray diꢃraction (XRD) patterns were
glyoxal
(GLY)
H
H
O
recorded on a Bruker AXS D8 diꢃractometer (Cu radia-
Kα
OH
tion, λ=0.154 nm), operated at 40 kV and 40 mA.
High resolution TEM images were recorded on an FEI
Tecnai F20 ꢀeld emission gun (FEG-)TEM operating at
HO
glycolic acid
(GLYC)
2
00 kV equipped with a Gatan Orius SC600A camera.
Images were analyzed by ImageJ 1.41 and the mean Pd
particle diameters were calculated by analyzing 100–150
particles for each sample using Nano Measurer program.
CO pulse chemisorption (Autochem II 2920, Micromerit-
ics Co., USA) measurements were used to measure the Pd
dispersion. Samples were outgassed at 150 °C under ꢂow-
ing hydrogen (20 mL/min) for 1 h, followed by ꢂowing He
Scheme 1 Reaction pathways and corresponding products of glyoxal
oxidation by air
C
product
Y =
C
GLY at t=0
(
20 mL/min) for 1 h before analysis at room temperature.
This reduction temperature was much lower than the process
during catalysts synthesis thus did not induce additional sin-
tering of Pd particles.
∗
GLY
X
= YAGLY + YOX
ꢀ
ꢁ × 100
∗
AGLY
S
= YAGLY∕XGLY
2.4 Catalytic Performance
3
Results and Discussion
The catalytic reaction was performed in a glass flask.
00 mL (100 mmol/L) glyoxal solution and 100 mg cata-
3.1 Catalyst Characterization
1
lyst were added in the reactor and then heated at 40 °C with
constant stirring rate of 1000 rpm. The pH of the reaction
mixture was constantly measured by a pH electrode. In
order to keep the pH value at 7.2, a solution of 0.3 mol/L
Na CO was added into the reaction mixture to neutralize
SAXS patterns of these MS supports are shown in Fig. 1a.
The SAXS results indicated the successful synthesis of KIT-
6
, MCM-41 and SBA-15. For KIT-6, it displayed a strong
peak and a hump arising from (211) and (220) reꢂections.
This pattern demonstrated that the prepared support had a
well-ordered mesoporous structure, which belonged to the
2
3
the generated acid by an automatic titration device. Air,
as oxidant, was bubbled through the solution with a con-
stant ꢂow rate of 400 mL/min. In recycle studies, the same
reaction conditions were performed except using recovered
catalysts.
3
D cubic Ia3d space group. For MCM-41 and SBA-15,
both of them exhibited one well-resolved peak and other
two peaks, indexed as (100), (110) and (200) reꢂections,
respectively. These distinct peaks of MCM-41 and SBA-15
corresponded to an ordered, 2D hexagonal mesophase of
P6mm. Wide-angle XRD patterns of these Pd/MS catalysts
are shown in Fig. 1b. All of them showed a broad peak at
The reaction products were sampled after removal of
catalysts, and the components of ꢀltrate were quantiꢀed by
Shimadzu HPLC, on a Phenomenex Synergi 4u Hydro-RP
8
2
0A column, heated at 25 °C. The injection volume was
0 μL, and a sulfuric acid solution (pH 2.7) was used as the
2
2.2° attributed to the amorphous silica in pore wall. The
other two sharp peaks at 40.2° and 46.6° were indexed as
mobile phase with a ꢂux of 1 mL/min. The products were
detected at 212 nm on a UV–Vis Spectra System 6000 LP
photodiode array detector.
(
111) and (200) planes of metallic Pd crystallites, and there
was no PdO phase detected over all catalysts. The Pd dif-
fraction peak of Pd/SBA-15 was the highest due to its large
particle size among these catalysts. Compared with Pd/SBA-
The main catalytic oxidation pathways are presented in
Scheme 1. Because glycolic acid was formed from Canniz-
zaro dismutation of glyoxal, which depended only on the
pH of reaction mixture and not on the catalyst. According to
literature [9, 10], the corrected conversion (X*=conversion
without considering glycolic acid, %) and corrected selectiv-
ity (S*, %) can be used to compare the natural properties of
catalysts, deꢀned as follows:
1
5 and Pd/MCM-41, Pd/KIT-6 displayed a broad diꢃrac-
tion peak of metallic Pd, implying that the 3D mesoporous
structure of KIT-6 eꢃectively promoted the dispersion of Pd
with smaller particle size.
The nitrogen sorption isotherms of three MS supports at
diꢃerent stages (prior to depositing Pd, after deposition and
1
3

J. Liu et al.
Fig. 1 a SAXS patterns of
KIT-6, MCM-41 and SBA-15
supports; b wide-angle XRD
patterns of Pd/KIT-6, Pd/MCM-
4
1 and Pd/SBA-15 catalysts
post-reaction) are shown in Fig. 2. All samples demonstrated
type IV isotherms according to IUPAC classiꢀcation, and
the isotherms of each sample were in good agreement with
previous reports [21, 23–25]. For MCM-41 support without
loading, the adsorption and desorption branches coincided
and no hysteresis loops were observed. These isotherms
exhibited sharp increase in the nitrogen uptake over a nar-
row range of p/p = 0.2–0.35 due to the capillary conden-
0
sation inside the mesopores. KIT-6 and SBA-15 showed
typical H -type hysteresis arising from ordered mesoporous
1
architecture. The BJH pore size distributions of all MS sup-
ports were in the mesoporous region with pore diameters
7
.1 nm, 3.3 nm and 5.0 nm for KIT-6, MCM-41 and SBA-
5, respectively. Their speciꢀc surface areas were consistent
1
with literature values for parent MS [25, 26]. Textual proper-
ties of these MS supports before and after Pd impregnation
are listed in Table 1. The types of isotherms, hysteresis and
mean mesoporous diameters were almost unchanged after
loading of Pd. Impregnation of Pd systematically depressed
the surface area and pore volume, which reꢂected the partial
blockage of the mesopores. Besides, the isotherms of all
catalysts did not change obviously after reaction, indicating
that the mesoporous structure of catalyst was well main-
tained during the oxidation of glyoxal.
Fig. 2 Nitrogen adsorption–desorption isotherms of MS supports: a
prior to depositing Pd, b after deposition, c post-reaction
Table 1 Textural properties of
as-prepared MS supports and
corresponding Pd/MS catalysts
2
Sample
SBET (m /g)
Pore volume Pore diam-
Dispersion of dPd (nm)
3
a
(
cm /g)
eter (nm)
Pd (%)
CO chem.a
STEMb
KIT-6
847
715
1679
1453
579
0.89
0.76
0.69
0.61
0.47
0.44
7.1
7.0
3.3
3.1
5.0
4.8
Pd/KIT-6
MCM-41
Pd/MCM-41
SBA-15
49
40
30
2.5
2.8
3.8
2.5
2.9
4.0
Pd/SBA-15
505
a
Dispersion of Pd and dPd were calculated from CO chemisorption
dPd was calculated by STEM analysis
b
1
3

Selective Oxidation of Glyoxal to Glyoxalic Acid by Air over Mesoporous Silica Supported Pd…
Fig. 3 Bright-ꢀeld HRTEM images: a Pd/KIT-6, b Pd/MCM-41, c Pd/SBA-15; HAADF-STEM images: d Pd/KIT-6, e Pd/MCM-41, f Pd/SBA-
1
5; Pd particle size distributions: g Pd/KIT-6, h Pd/MCM-41, i Pd/SBA-15. Analyses based upon 100–150 particles
The representative bright-field high-resolution TEM
HRTEM) and high-angle annular dark-ꢀeld (HAADF)-
Table 2 Catalytic results of glyoxal oxidation over Pd/KIT-6, Pd/
MCM-41 and Pd/SBA-15 catalysts
(
STEM of these prepared Pd/MS catalysts are shown in
Fig. 3. The HRTEM images clearly illustrated the reten-
tion of ordered and long parallel channels of Pd/MCM-41
and Pd/SBA-15 as well as cubic packed pore structure of
Pd/KIT-6. This suggested that impregnation of Pd did not
change the ordered mesoporous structure of MS supports.
HAADF-STEM was used to observe the size distribution of
Pd by analyzing 100–150 particles for each sample, and the
Pd particles (brighter spots labeled by red circles) were well
dispersed throughout all MS supports with mean particle
size of 2.5 nm, 2.9 nm and 4.0 nm for Pd/KIT-6, Pd/MCM-
∗
GLY
∗
AGLY
Catalysts
Reaction
time (h)
X
(%)
S
(%)
YAGLY (%)
Pd/KIT-6
5
16.5
30.4
41.3
48.9
14.8
28.0
39.2
47.1
10.3
19.7
32.8
37.0
5.5
81.2
68.1
57.0
48.7
81.8
65.7
50.5
45.2
86.4
71.6
56.4
49.7
76.4
81.2
57.0
54.3
54.5
13.4
20.7
23.5
23.8
12.1
18.4
19.8
21.3
8.9
14.1
18.5
18.4
4.2
1
1
2
0
5
0
5
0
5
0
5
0
5
0
20
20
20
20
20
Pd/MCM-41
Pd/SBA-15
1
1
2
4
1 and Pd/SBA-15, respectively. For Pd/MCM-41 and Pd/
1
1
2
SBA-15, it could be observed that Pd species existed as
nanorods (appeared as light, necklace-like objects along the
channels) and nanoparticles. However, Pd species of KIT-6
appeared as spherical shape in the 3D mesopore.
Pd/C
Pd/C [5]
Pd/C [8]
Bi–Pd/C [9]
Ru–Pd/C [4]
3.8
7.4
26.3
37.6
3.1
4.3
14.3
20.5
The diꢃerence of Pd dispersion on these MS supports was
also tested by CO chemisorption, from which, assuming a
CO:Pd
stoichiometry of 2:1 [27–29]. The dispersion of
surface
Pd decreased in the order of Pd/KIT-6>Pd/MCM-41>Pd/
SBA-15. It indicated that KIT-6 support possessed the small-
est Pd particle size, consistent with wide-angle XRD and
HRTEM measurements. These results revealed that increas-
ing mesopore interconnectivity and surface area of MS both
Reaction condition: 40 °C, 100 mg catalyst, 100 mL glyoxal
100 mmol/L), 1000 rpm of stirring rate
(
1
3

J. Liu et al.
Fig. 5 Eꢃect of stirring rate on catalytic performance toward glyoxal
oxidation of Pd/KIT-6, Pd/MCM-41 and Pd/SBA-15 after 15 h. Reac-
tion conditions: 40 °C, 100 mg catalyst and 10 mmol glyoxal
Fig. 4 Eꢃect of stirring rate on turnover frequency (TOF) toward gly-
oxal oxidation of Pd/KIT-6, Pd/MCM-41 and Pd/SBA-15 after 1 h.
Reaction conditions: 40 °C, 100 mg catalyst and 10 mmol glyoxal
helped to enhance the dispersion of Pd with smaller particle
size.
3.2 Catalytic Performance
The catalytic performances of the synthesized Pd/MS cata-
lysts and commercial Pd/C (J&K Scientiꢀc Ltd) catalyst
were tested for selective oxidation of glyoxal by air, and
the results are listed in Table 2. The results indicated that
catalytic activity was very sensitive to supports. Compared
with 5.5% glyoxal conversion after 20 h for Pd/C catalyst,
three Pd/MS catalysts had at least 6 times higher activity
than Pd/C in this reaction. To further investigate the cata-
lytic behaviors of Pd/MS catalysts, their catalytic activity
under diꢃerent reaction time was also measured. For all Pd/
MS catalysts, with extension of reaction time, the conver-
sion of glyoxal increased and the selectivity of glyoxalic
acid (AGLY) decreased. The yield of AGLY approximately
reached the maximum at 15 h. It implied that glyoxalic acid
was easily over-oxidized to oxalic acid. Among these cata-
lysts, Pd/KIT-6 exhibited the best catalytic performance with
∗
AGLY
∗
GLY
Fig. 6 S
against X
under 500 rpm and 1000 rpm: the solid
lines represent 1000 rpm of stirring rate and the dotted lines represent
5
00 rpm of stirring rate. Reaction conditions: 40 °C, 100 mg catalyst
and 10 mmol glyoxal
also observed in the oxidation of glyoxal. To further inves-
tigate the mass transfer of reactants in these Pd/MS cata-
lysts, the values of turnover frequency (TOF) after 1 h were
calculated and plotted against stirring rate in Fig. 4. For Pd/
KIT-6, the plateau of TOF reached at 700 rpm, which was
lower than 800 rpm of Pd/MCM-41 and Pd/SBA-15. This
indicated that the diꢃusion of reactants over Pd/KIT-6 was
better than those over Pd/MCM-41 and Pd/SBA-15. The 3D
mesoporous structure of Pd/KIT-6 could favor the reactant
molecules contacting active sites, leading to improved con-
version of glyoxal. However, the 1D channels of Pd/MCM-
41 and Pd/SBA-15, to some extent, hindered the reactant to
access the active sites due to the longer diꢃusion pathway.
Catalytic activity of these Pd/MS catalysts under diꢃer-
ent stirring rate after 15 h is shown in Fig. 5. These results
4
1.3% glyoxal conversion and 57.0% selectivity to AGLY
after 15 h. Pd/MCM-41 exhibited 39.2% glyoxal conversion
and 50.5% selectivity to AGLY after 15 h, which were both
lower than those of Pd/KIT-6. Although Pd/SBA-15 attained
similar selectivity with Pd/KIT-6, 32.8% glyoxal conversion
was lower than 41.3% of Pd/KIT-6. Therefore, Pd/KIT-6
catalyst achieved higher conversion and selectivity at the
same time. In addition, the catalytic activity of Pd/KIT-6
also exceeded those reported Pd-based catalysts.
In previous reports, KIT-6 supported Pd catalyst can
improve the aerobic reaction rate of alcohols by its unique
3
D mesoporous structure [20, 21]. This phenomenon was
1
3

Selective Oxidation of Glyoxal to Glyoxalic Acid by Air over Mesoporous Silica Supported Pd…
Fig. 7 Recycling tests of Pd/KIT-6, Pd/MCM-41 and Pd/SBA-15.
Reaction condition: 40 °C, 10 h, and 10 mmol glyoxal
indicated that the catalytic activity and selectivity were obvi-
ously aꢃected by mixing state. When stirring rate was higher
than 800 rpm, the external diꢃusion was eliminated for all
catalysts, and further increase of stirring rate had no eꢃect
on activity and selectivity. To investigate the eꢃect of mixing
state on selectivity, the selectivity against conversion under
5
00 rpm and 1000 rpm is shown in Fig. 6. With the increase
of conversion, selectivity decreased. It was observed that
all catalysts achieved higher selectivity under higher stir-
ring rate. Meanwhile, Pd/KIT-6 exhibited higher selectivity
between 500 and 1000 rpm. It suggested that the selectivity
of this consecutive reaction was obviously inꢂuenced by dif-
fusion condition. Due to the poor diꢃusion under 500 rpm
of stirring rate, the oxidized product (glyoxalic acid) might
be diꢁcult to depart from reaction region to solution, thus
it would be over-oxidized to oxalic acid on the active sites.
Pd/KIT-6 with 3D mesoporous structure exhibited excellent
capability to facilitate the diꢃusion of product, resulting in
higher selectivity under the same conversion. Besides, it was
also obvious that Pd/KIT-6 had a higher conversion when
the three catalysts reached the same selectivity.
Fig. 8 General kinetics for Pd/KIT-6, Pd/MCM-41 and Pd/SBA-15
catalysts: a ln(a/(a–x)) as a function of time and b Y
as a function
GLYC
of X
GLY
is presented in Scheme 1. k represented the initial oxida-
I
tion rate of glyoxal to glyoxalic acid and k represented the
II
over-oxidation rate of glyoxalic acid to oxalic acid. k
CAN
represented the dismutation reaction rate of glyoxal. Thus,
k
(k = k + k
) represented the consumption rate of
CAN
exp
exp
I
glyoxal. To measure the reaction kinetics, external diꢃu-
sion limitation was eliminated under 1000 rpm according to
above discussion. Derivation process of the kinetics was the
same as literature based on ꢀrst-order reaction hypotheses
[4, 10].
To test the stability of these heterogenous catalysts, four
consecutive runs were carried out. After 10 h reaction, cata-
lysts were collected and reused for a new batch. Recycling
results of the three catalysts are displayed in Fig. 7. The
conversion and selectivity of these catalysts showed a slight
decrease after four consecutive runs, which indicated the
good stability of three Pd/MS catalysts.
k
× t = ln (a∕(a − x)), kCAN∕kexp = YGLYC∕XGLY
exp
with a=1, x=X /100.
GLY
Thus k , k
and k (k =k −k ) could be obtained
I I exp CAN
exp CAN
3.3 Kinetic Study
from these ꢀtted lines, shown in Fig. 8. These linear plots
implied that there was no obvious deactivation or metal
leaching for all catalysts, and the calculated values of rate
constants are listed in Table 3. In all cases, the similar values
In order to better reꢂect the diꢃerences among the three cat-
alysts, kinetic study was also performed. Catalytic process
1
3

J. Liu et al.
Table 3 Values of ꢀrst-order rate constants (h−1) for Pd/KIT-6, Pd/
MCM-41 and Pd/SBA-15 catalysts
Catalysts
Pd/KIT-6
Pd/MCM-41
Pd/SBA-15
kexp
0.041
0.122
0.005
0.036
0.072
0.50
0.037
0.098
0.004
0.033
0.083
0.39
0.028
0.143
0.004
0.024
0.070
0.34
kCAN/kexp
kCAN
kI
kII
kI/kII
of k
proved that the glycolic formation was independ-
CAN
ent of catalyst. As for k , it was obvious that Pd/KIT-6 pos-
I
−
1
sessed the highest value with 0.036 h , demonstrating its
Fig. 9 General kinetics for Pd/KIT-6, Pd/MCM-41 and Pd/SBA-15
superior activity to 2D architecture supports of Pd/MCM-41
catalysts: Y
as a function of time
AGLY
−
1
−1
and Pd/SBA-15 with k equal to 0.033 h and 0.024 h ,
I
respectively. Compared with Pd/SBA-15, the k of Pd/KIT-6
I
increased by 50%, which implied an obvious improvement of
the highest value of k and relative low value of k , which
I
II
reaction rate. The highest k of Pd/KIT-6 might be ascribed
I
were both beneꢀcial to this reaction. The value of k /k was
I
II
to its 3D mesoporous structure, which provided suꢁcient
active sites and favored the contact between active sites and
reactant molecules. However, for the same 2D architecture,
used to intuitively reꢂect the selectivity diꢃerence among
these catalysts. The calculated k /k of Pd/KIT-6 was 0.50,
I
II
which was obviously higher than 0.39 of Pd/MCM-41 and
Pd/MCM-41 exhibited a higher value of k than Pd/SBA-15,
I
0
.34 of Pd/SBA-15. Compared with Pd/MCM-41, the k /k
I II
probably due to its better Pd dispersion.
of Pd/KIT-6 increased by 28%, indicating an improvement of
selectivity. This reꢂected the higher selectivity of Pd/KIT-6
from another aspect. Due to the 3D mesoporous architecture
of KIT-6, the catalyst provided suꢁcient active sites, which
were more accessible to reactants. Meanwhile, the oxidation
product could also diꢃuse from catalyst to solution easily
thanks to its better mass transfer capability. Therefore, the
In order to reꢂect the degree of over-oxidation, the reac-
tion rate of k was calculated as follows:
II
dCAGLY∕dt = k × C
− k × CAGLY ,
II
I
GLY
⇒
dCAGLY∕dt = k × a × exp(−k × t) − k × CAGLY,
I exp II
The solution of this diꢃerential equation was given by the
software of Mathematica.
C
= (k × a × exp(−k × t))∕(k − k )−(k × a)∕(k − k ) × exp(−k × t)
I exp II exp I II exp II
AGLY
Therefore, k reꢂected the over-oxidation rate and k /k
II
I
II
3D mesoporous structure of Pd/KIT-6 provided an eꢁcient
way to improve the conversion and avoid over-oxidation at
the same time.
could be used to compare the selectivity of diꢃerent cata-
lysts. The lower value of k /k indicated that glyoxalic acid
I
II
was slowly produced or easily over-oxidized, leading to
lower selectivity. The experimental plots of Y [t] and
AGLY
ꢀ
tting curve of corresponding catalysts are displayed in
Fig. 9. All ꢀtting curves were well ꢀtted with the experi-
mental results, which implied that the obtained values of k
4
Conclusions
II
Three series of Pd/MS catalysts have been synthesized
and applied for selective oxidation of glyoxal by air under
mild condition. Among these catalysts, Pd/KIT-6 exhibited
higher catalytic activity and selectivity than those of Pd/
MCM-41 and Pd/SBA-15. It was found that KIT-6 with 3D
mesoporous structure enhanced Pd dispersion, making the
active sites on Pd particles more accessible to reactants.
Such an eꢃect was reꢂected by the higher oxidation rate of
were reliable. Both the values of k and k /k are also listed
II
I
II
in Table 3.
In all cases, the value of k was higher than k , which con-
II
I
ꢀ
rmed that glyoxalic acid was easily over-oxidized to oxalic
acid. For Pd/MCM-41 and Pd/SBA-15, the higher value of
k corresponded to higher value of k , indicating that accel-
I
II
erating the reaction activity would also improve the reaction
rate of over-oxidation. Taken together, Pd/KIT-6 possessed
1
3

Selective Oxidation of Glyoxal to Glyoxalic Acid by Air over Mesoporous Silica Supported Pd…
k . On the other hand, the 3D mesoporous structure of KIT-6
9. Alardin F, Delmon B, Ruiz P, Devillers M (2000) Catal Today
I
6
1:255–262
with better mass transfer capability allowed the product
molecules to depart from catalysts to solution phase easily,
which reduced the probability of over-oxidation side reac-
tion. This advantage could also be conꢀrmed by the rela-
1
0. Alardin F, Wullens H, Hermans S, Devillers M (2005) J Mol Catal
A 225:79–89
11. Zhao DY, Feng JL, Huo QS, Melosh N, Fredrickson GH, Chmelka
BF, Stucky GD (1997) Science 279:548–552
1
1
1
2. Kleitz F, Choi SH, Ryoo R (2003) Chem Commun 17:2136–2137
3. Jiao F, Frei H (2009) Angew Chem Int Ed 48:1841–1844
4. Crudden CM, Sateesh M, Lewis R (2005) J Am Chem Soc
127:10045–10050
tive low over-oxidation rate of k . These conclusions might
II
also apply to other catalysts, and similar strategies could be
developed for other liquid phase oxidations to achieve higher
conversion and selectivity.
15. Kim SW, Son SU, Lee SI, Hyeon T, Chung YK (2000) J Am Chem
Soc 122:1550–1551
1
1
6. Zhu K, Hu J, Richards R (2005) Catal Lett 100:195–199
Acknowledgements This work was supported by the Shanghai Science
7. Belhekar AA, Awate SV, Anand R (2002) Catal Commun
and Technology Committee (Grant No. 14DZ2273900).
3
:453–458
1
1
8. Naik B, Hazra S, Prasad VS, Ghosh NN (2011) Catal Commun
12:1104–1108
Compliance with Ethical Standards
9. Karimi B, Abedi S, Clark JH, Budarin V (2006) Angew Chem Int
Ed 118:4894–4897
Conflict of interest All authors declare that they have no conꢂict of
interest.
20. Parlett CMA, Bruce DW, Hondow NS, Lee AF, Wilson K (2011)
ACS Catal 1:636–640
2
1. Parlett CMA, Bruce DW, Hondow NS, Newton MA, Lee AF,
Wilson K (2013) ChemCatChem 5:939–950
References
22. Cai Q, Lin WY, Xiao FS, Pang WQ, Chen XH, Zou BS (1999)
Microporous Mesoporous Mater 32:1–15
1
.
Alardin F, Ruiz P, Delmon B, Devillers M (2018) Appl Catal A
23. Mastalir Á, Rác B, Király Z, Molnár Á (2007) J Mol Catal A
264:170–178
2
15:125–136
2
3
.
.
Li K, Frost JW (1998) J Am Chem Soc 120:10545–10546
Meester WJN, Maarseveen JHV, Schoemaker HE, Hiemstra H,
Rutjes FPJT (2003) Eur J Org Chem 1:2519–2529. https://doi.
org/10.1002/ejoc.200200714
24. Sonwane CG, Bhatia SK (1999) Langmuir 15:2809–2816
25. Grün M, Unger KK, Matsumoto A, Tsutsumi K (1999) Micropo-
rous Mesoporous Mater 27:207–216
26. Li C, Zhang Q, Wang Y, Wan H (2008) Catal Lett 120:126–136
27. Marx S, Krumeich F, Baiker A (2011) J Phys Chem C
115:8195–8205
4
5
6
.
.
.
Deffernez A, Hermans S, Devillers M (2005) Appl Catal A
2
82:303–313
Hermans S, Thiltges F, Deꢃernez A, Devillers M (2012) Catal
Lett 142:521–530
28. Cabilla GC, Bonivardi AL, Baltanás MA (1998) Catal Lett
55:147–156
Jia ML, Liu CX, Wang J, Bao S, Bao Z (2014) Kinet Catal
29. Canton P, Menegazzo F, Polizzi S, Pinna F, Pernicone N, Riello
P, Fagherazzi G (2003) Catal Lett 88:141–146
5
5:671–675
7
8
.
.
Hermans S, Devillers M (2005) Catal Lett 99:55–64
Deꢃernez A, Hermans S, Devillers M (2007) J Phys Chem C
Publisher’s Note Springer Nature remains neutral with regard to
1
11:9448–9459
jurisdictional claims in published maps and institutional aꢁliations.
1
3