Applied Catalysis B: Environmental 284 (2021) 119803

Contents lists available at ScienceDirect

Applied Catalysis B: Environmental

journal homepage: www.elsevier.com/locate/apcatb

Engineering Pt-Mn2O3 interface to boost selective oxidation of ethylene

glycol to glycolic acid
Hao Yan a, Shuang Yao a, Jinyao Wang a, Siming Zhao a, Yinghao Sun a, Mengyuan Liu a,
Xin Zhou a, Guangyu Zhang a, Xin Jin a, Xiang Feng a, *, Yibin Liu a, *, Xiaobo Chen a, De Chen b,
Chaohe Yang a
a
State Key Laboratory of Heavy Oil Processing, China University of Petroleum, Qingdao, 266580, China
b
Department of Chemical Engineering, Norwegian University of Science and Technology, Trondheim, 7491, Norway

A R T I C L E I N F O A B S T R A C T

Keywords: Rational design of desirable active sites is still a grand challenge for the efficient conversion of polyols to value-
Interfacial active site added products. Herein, we successfully constructed the Pt-Mn2O3 interfacial sites rather than Pt-MnOx solid
Oxidation of ethylene glycol solution to boost selective oxidation of ethylene glycol to glycolic acid under mild conditions. X-ray absorption
DFT calculations
spectroscopy and high-resolution transmission electron microscope revealed that the pre-distribution of Mn2O3
inside support unexpectedly induced the formation of Pt-Mn2O3 interfacial active sites with strong electron
coupling effect, leading to an unprecedented catalytic activity (turnover frequency: 3196.9 h−1) and glycolic acid
yield (86.4 %). In addition, quantitative analysis of the intrinsic active sites was performed, and a “volcano-
shape” relationship was established between initial reaction rate and Pt/Mn ratio. Moreover, the structure-
dependent reaction kinetics and density functional theory calculation revealed that the synergistic effect be-
tween the Mn2O3 redox cycle and Pt promotes the adsorption of ethylene glycol and the activation of C–H bond,
resulting in the lower activation energy of ethylene glycol oxidation. The outcome of this work offers a promising
avenue for the direct construction of efficient Pt-based catalysts with desired active sites.

1. Introduction suffer from complex product separation, high investment cost and toxic
compounds in raw materials or intermediate products [17–20]. Sus-
Ethylene glycol (EG), as the simplest diol, is widely used in various tainable development of EG oxidation to GA over heterogeneous cata-
fields, such as a coolant and antifreeze for cars, a desiccant in the natural lysts offers an alternative green and efficient process. However, up to
gas production process and a deicing fluid for windshields [1–5]. The now, there are very limited reports on the oxidation of EG to GA over
global EG industry will face the problem of overcapacity in the future heterogeneous catalysts. Berndt et al. [13] prepared the Au/Al2O3
due to the continuous development of biomass conversion and coal catalyst via deposition precipitation method for the EG oxidation to GA
chemical industry [6–9]. In this scenario, it is of great significance to in the presence of NaOH. Lately, Shi et al. [4] realized the EG oxidation
explore new industrial application fields of EG. Up to date, intensive over the PtFe/CeO2 catalyst in the NaOH solution. Unfortunately, the
efforts have been made in the synthesis of fine chemicals [1,2,10–12], introduction of NaOH produces glycolate instead of GA, and the sub-
and the selective oxidation of EG is an attractive way because its sequent acid neutralization process will inevitably bring problems such
oxidation products are essential platform chemicals (e.g., glyoxal, as secondary pollution and increased production cost. Therefore, it is
methyl glycolate and glycolic acid) [13–15]. In particular, glycolic acid urgent to develop more efficient heterogeneous catalysts for the selec-
(GA) has versatile applications in leather processing, texture washing, tive oxidation of EG to GA in base-free medium.
packaging material, metal cleaning and adhesives [16]. Generally, the oxidation of polyols to carboxylic acids involves
Commercially, the majority of GA is prepared by the carbonylation of multiple-step reactions in a combinatory fashion on the catalyst surface
formaldehyde on the acid catalysts, the alkaline hydrolysis of chloro- [21–29]. This requires the multi-functional active sites to achieve the
acetic acid and the acid hydrolysis of hydroxyacetonitrile, which all tandem activation of C–H, C–O, C– – O bond. Recently, metal-metal (or

* Corresponding authors.
E-mail addresses: xiangfeng@upc.edu.cn (X. Feng), liuyibin@upc.edu.cn (Y. Liu).

https://doi.org/10.1016/j.apcatb.2020.119803
Received 11 October 2020; Received in revised form 22 November 2020; Accepted 28 November 2020
Available online 9 December 2020
0926-3373/© 2020 Elsevier B.V. All rights reserved.
H. Yan et al. Applied Catalysis B: Environmental 284 (2021) 119803

metal oxide) interfacial active sites have been demonstrated to be effi- was dried (at 90 ◦ C for 2 h) and calcined (at 450 ◦ C for 3 h). All the
cient in the activation of oxygen-containing functional groups [30–32]. catalysts were reduced by 10 vol.% H2 and Ar mixture gas at 350 ◦ C.
Chen et al. [2] observed that the Cu-CeO2 interface is highly active for
the low-temperature water-gas shift reaction. Reaction kinetics and in 2.3. Catalyst characterization
situ characterization revealed that CO could be chemically adsorbed on
the Cu+ site whereas the oxygen vacancies (Ov) of ceria could activate X-ray diffraction (XRD, X’pert PRO MPD diffractometer, CuKα) was
H2O separately. Pan et al. [25] identified the interfacial between Au and employed to determine the crystal phase of the catalysts. N2 phys-
ZnO as the specific active site for the selective oxidation of glycerol to 1, isorption (BET, Micromeritics ASAP 2020) was used to characterize the
3-dihydroxyacetone. The interfacial active sites formed by heat treat- pore volume and diameters. Raman spectra was employed on the Lab-
ment promote the generation of positively charged Au sites resulting in RAM HR Evolution (HORIBA JobinYvon), which was equipped with
the formation of oxygen vacancy defects [33]. Essentially, aqueous with the 532 nm laser lines. 29Si NMR of the samples was measured on
phase oxidation of polyols involves the transformation of the JNM-ECZ600R. High-angle annular dark-field scanning transmission
oxygen-containing groups (e.g., the activation of oxygen and water, and electron microscopy (HAADF-STEM) images, energy dispersive X-ray
the oxidation of hydroxyl groups). Clearly, the rational designed catalyst (EDX) mapping and element line scan were recorded on the Tecnai G2
with metal-metal (or metal oxide) interfacial active sites based on the F20 S-Twin. X-ray photoelectron spectroscopy (XPS, Thermo ESCALAB
corresponding reaction mechanism is highly desirable to facilitate the 250Xi) was performed to measure the composition and valence states of
oxidation of EG to GA. surface metals. In the process of the XPS deconvolution, we have fol-
Herein, interfacial Pt-Mn2O3 active sites in the PtMn/MCM-41 lowed the corresponding principles: (1) The peak intensity ratio of Pt
catalyst were constructed to boost the catalytic performance for selec- 4ƒ7/2 to Pt 4ƒ5/2 is 4:3; (2) the fixed distance between Pt 4ƒ7/2 and Pt 4ƒ5/
tive oxidation of ethylene glycol to glycolic acid under mild conditions. 2 is 3.35 eV; (3) Full width at half maxima (FWHM) is no more than
Pre-distribution of Mn2O3 inside MCM-41 support induced the forma- 2.5 eV. (4) Lorentzian-Gaussian ration is about 20 %. Inductively
tion of Pt-Mn2O3 interfacial structure. Compared with Pt-MnOx solid coupled plasma optical emission spectrometry (ICP-OES, VARIAN 720-
solution, multi-characterizations (such as XANES, EXAFS, HRTEM and ES) was used to obtain the Pt and Mn metal content. XAS measure-
XPS) demonstrated that there is a strong electronic coupling effect at the ments for the Mn K-edge were conducted in transmission mode (fluo-
Pt-Mn2O3 interface of the PtMn/MCM-41 (In-70) catalyst, resulting in rescence mode on beamline 12-BM in the Advanced Photon Source at
the high TOF (3196.9 h−1) and glycolic acid yield (86.4 %). Through Argonne National Laboratory). Athena software was employed to obtain
accurate quantitative analysis of the intrinsic active sites, a “volcanic the X-ray absorption near edge structure (XANES) and extended X-ray
curve” relationship was observed between initial reaction rate and Mn absorption fine structure spectroscopy (EXAFS) data. The chemical
content, and excessive or insufficient Mn could affect the exposure or adsorption/reduction properties of catalyst surface are tested on the
formation of the Pt-Mn2O3 interfacial active sites. Furthermore, kinetic Micromeritics AutoChem II 2920 instrument. For H2 temperature pro-
analysis and DFT calculation were employed to reveal the synergistic grammed reduction (H2-TPR), a certain amount of unreduced samples
effect between the Mn2O3 redox cycle and Pt in the reaction mechanism. was heated to 900 ◦ C in the 10 vol.% H2/Ar mixed gas (30 mL/min) at a
The methodology employed here may serve as a unique tool to rationally rate of 15 ◦ C mL/min [21,35]. In a typical H2 temperature programmed
develop efficient catalysts in the conversion of polyols. desorption (H2-TPD), the samples were firstly pretreated by 10 vol.%
H2/Ar mixed gas (30 mL/min) for 1 h. As the quartz tube was cooled to
2. Experimental room temperature (25 ◦ C), the sample was purged by inert gas (Ar,
30 mL/min). Then, the sample was ramped to 500 ◦ C and held for 0.5 h
2.1. Chemicals [21]. For the ethanol pulse adsorption, a small amount of catalyst was
pretreated in a flow of He (30 mL/min) for 1 h at 300 ◦ C. When the
Manganese nitrate solution (Mn(NO3)2, 50 wt.%), tetraethyl ortho- temperature was cooled to 50 ◦ C, ethanol molecule was pulsed into the
silicate (TEOS (98 %)), chloroplatinic acid (H2PtCl6⋅6H2O (99.9 % ACS, quartz tube. The flask and reflux temperatures of ethanol steam gener-
Pt min 37.5 %)) and ammonia solution tetrahydrate (NH3⋅H2O, 25 %–28 ator were 50 and 40 ◦ C. Three consecutive peaks show the same peak
%) were purchased from Sinopharm Chemical Reagent Co., Ltd. Hex- area, indicating that the sample test is completed.23
adecyl trimethyl ammonium bromide (CTAB, 99 %) was provided by
Sigma-Aldrichk.
2.4. Ethylene glycol oxidation

2.2. Catalyst preparation

The prepared catalyst was evaluated in a 50 mL high-pressure
reactor. Briefly, 0.2 g catalyst was added into 25 mL ethylene glycol
The Mn-MCM-41 s were prepared by the in situ doping method [29,
solution (0.3 mol/L). Then, the reactor with final 1 MPa O2 was purged
34]. The molar ratio of the reagents used is as follows: TEOS: CTAB:
four times with O2. The liquid products were analyzed by high perfor-
NH3: H2O: 50 wt.% Mn(NO3)2 = 1 : 0.12 : 0.6 : 100 : x. Firstly, CTAB was
mance liquid chromatography (HPLC, Shimadzu LC-20AT). The refrac-
dissolved in the mixed solution of ammonia and deionized water. TEOS
tive index (RID-10A) and UV detectors together with the phenomenex
together with Mn(NO3)2 was slowly dropped into the above solution
chromatographic column (Rezex ROA-Organic Acid H+ (8%)) were
resulting in the formation of a white gel. Subsequently, the gel was
equipped on the HPLC. 0.005 M H2SO4 solution was employed as the
stirred for 30 min and crystallized for 52 h. The as-formed product was
mobile phase. The CO2 is analyzed by a Scion 456-GC Chromatography
filtered, dried (at 90 ◦ C for 2 h) and calcined (at 550 ◦ C for 6 h). The
equipped with a TCD detector. The definitions of conversion, selectivity
samples obtained by in situ doping method were denoted as
and turnover frequency (TOF) have been described in previous literature
Mn/MCM-41 (In-38), Mn/MCM-41 (In-70), Mn/MCM-41 (In-200) and
[21,23]. The turnover frequency (TOF), glycerol conversion (X) and
Mn/MCM-41 (In-360), and the numbers here represent the different
product selectivity (S) were calculated by the following formula:
molar ratio of TEOS/50 wt.% Mn(NO3)2.
1.5 wt.% Pt metal was loaded on the above supports via equivalent- N
TOF =
volumetic impregnation method. PtMn/MCM-41 (IM) was prepared by M×T
the co-impregnation of Mn (1.0 wt.%) and Pt (1.5 wt.%) metal pre-
K0 − Kt
cursors. A certain amount of Pt metal precursor and Mn metal precursor X(%) = × 100%
K0
was added into deionized water (8 mL) to impregnate the MCM-41
support (1 g) and stood at room temperature for 6 h. Then, the sample

2
H. Yan et al. Applied Catalysis B: Environmental 284 (2021) 119803

Ct 3. Results and discussion

S(%) = × 100%
K0 − Kt
3.1. Engineering Pt-Mn2O3 interface of the PtMn/MCM-41 catalysts
TOF was calculated as mole of EG converted per mole of surface Pt
atoms per hour. M, T and N represent the number of moles of Pt on the
Engineering the desired interfacial active sites is the key to enhance
surface, reaction time and the number of EG moles converted during the
catalytic performance. Herein, the Pt/MCM-41, PtMn/MCM-41 (IM) and
reaction time, respectively. It should be noted that TOF is estimated at
PtMn/MCM-41 (In-70) catalysts prepared by impregnation and/or in situ
conversions less than 10 %, which could ensure that the average reac-
doping methods were evaluated for the selective oxidation of EG to GA
tion rate equaled the instantaneous reaction rate [21,22]. In addition, K0
(Fig. 1a). The TOF values are calculated based on the Pt metal dispersion
is the initial concentration of EG (mol/L), Kt is the concentration after
measured by TEM, and other details of the catalysts are summarized in
the EG oxidation (mol/L), and Ct is the concentration of product formed
Table S1. Blank experiments show that MCM-41 or Mn-MCM-41 (x)
in the reaction (mol/L).
supports have almost no EG oxidation activity (Table S2). Fig. 1b shows
that the PtMn/MCM-41 (IM) and PtMn/MCM-41 (In-70) catalysts
2.5. Density functional theory (DFT) calculations
display higher catalytic activity and GA yield than the monometallic Pt/
MCM-41 catalyst, suggesting that the incorporation of Mn could pro-
Pt (111) surface was constructed using a p(4 × 4) slab containing
mote EG oxidation. Moreover, the PtMn/MCM-41 (In-70) catalyst with
four layers with 16 atoms per layer. The thickness of vacuum is 15 Å.
uniform pre-distribution of Mn species in MCM-41 support (Fig. S1)
Mn2O3 cluster was loaded on Pt (111) surface to represent the Pt-Mn2O3
exhibits superior catalytic performance (TOF: 3196.9 h−1, 86.4 % GA
interfacial active sites according to the results of EXAFS. The two bottom
yield) than that of the PtMn/MCM-41 (IM) catalyst (TOF: 2403.8 h−1,
Pt (111) layers were fixed, while the other layers and the Mn2O3 cluster
68.9 % GA yield). The main difference between the two catalysts is the
were fully relaxed. The density functional theory (DFT) calculations
incorporation method of Pt and Mn. The PtMn/MCM-41 (IM) catalyst is
were performed on the Materials Studio package. The exchange-
prepared by co-impregnating of Pt and Mn onto the pure MCM-41
correlation effects were described by generalized gradient approxima-
support, while the PtMn/MCM-41 (In-70) catalyst is obtained through
tion (GGA) using the Perdew–Burke–Ernzerhof (PBE) functional [36].
loading Pt onto the Mn-doped MCM-41 support (Mn-MCM-41) (Fig. 1a).
The double numerical plus polarization (DNP) was employed to consider
Meanwhile, 29Si NMR shows that the coordination environment of Si in
the electronic polarization effect. A 3 × 3×1 k-point grid was selected
the PtMn/MCM-41 (In-70) catalyst is almost the same as that of MCM-41
by Monkhorst-Pack method. All the calculation results were corrected
(Fig. S1a), indicating that the doped Mn does not enter into the frame-
by zero-point energy correction [37]. In the process of exploring
work of MCM-41. On this basis, we infer that the distinct structures of Pt
elementary reaction, the transition state (TS) was searched by LST/QST
and Mn have been generated in the two catalysts resulting in the
method. Detailed calculation parameters have been described in previ-
different catalytic performance.
ous literatures [24,29]. The definition formula of adsorption energy (or
EXAFS was first performed to investigate the atomic arrangements
binding energy) (Eads or EBE) is: Eads=Ead/sub-Ead-Esub, where Ead/sub, Ead
around photoabsorbers of the PtMn/MCM-41 (In-70) and PtMn/MCM-
and Esub are the total energy of the calculation system, the energy of
41 (IM) catalysts. Fig. 1c shows that the peaks observed at ~1.5 (Mn-
adsorbate and model, respectively [38].
O1) and 2.7 Å (Mn-Mn) spectra corresponding to the presence of Mn2O3
[39–41]. Obviously, the Mn metal in the PtMn/MCM-41 (In-70) catalyst
mainly exists in the form of independent Mn2O3. This independent metal

Fig. 1. (a) Schematic diagram of different catalysts. Surface Mn/Pt molar ratios were measured from XPS. (b) Catalytic performance of ethylene glycol oxidation on
different catalysts (experimental conditions: 25 mL 0.3 M EG solution, 0.2 g solid catalyst, 60 ◦ C, 1 MPa O2, 8 h). (c) Fourier transform of Mn K-edge extended EXAFS
oscillations, (d) the first-order derivatives of Mn K-edge XANES in the PtMn/MCM-41 (IM) and PtMn/MCM-41 (In-70) catalysts.

3
H. Yan et al. Applied Catalysis B: Environmental 284 (2021) 119803

particle distribution is highly possible to induce the formation of transfer of Mn in the PtMn/MCM-41 (IM) catalyst is less [47,48]. In
Pt-Mn2O3 interfacial active sites. However, the PtMn/MCM-41 (IM) other words, the Pt-Mn2O3 interfacial active sites exhibit stronger
catalyst prepared by the co-impregnation method has many small peaks electron coupling effect than the Pt-MnOx solid solution. Combined with
in addition to the high Mn-O1 peaks (~1.5 Å). EXAFS curve-fitting pa- the results of catalyst evaluation, we can infer that the Pt-Mn2O3
rameters in Table S3 demonstrated that the fitted Mn-M coordination interfacial active sites (in the PtMn/MCM-41 (In-70) catalyst) display
number for the PtMn/MCM-41 (IM) catalyst (3.2 and 1.9) is higher than superior catalytic performance than the active sites of the Pt-MnOx solid
that of the PtMn/MCM-41 (In-70) catalyst (2.5 and 1.7). The change of solution (in the PtMn/MCM-41 (IM) catalyst). Therefore, it is necessary
coordination environment is probably due to the formation of Pt-MnOx to further explore the underlying relationship between catalyst structure
solid solution generated in the co-impregnation process. Similar phe- and catalytic performance of EG oxidation.
nomenon of solid solution formation between Pt and other metal oxides The physico-chemical structures of Pt-Mn active sites for different
has also been confirmed in other literatures [42–44]. Consequently, the catalysts were further studied. HAADF-STEM was employed to directly
Pt-MnOx solid solution nanoparticles may exist in the PtMn/MCM-41 observe the distribution of Pt and Mn nanoparticles. Fig. 2a shows that
(IM) catalyst. Moreover, almost same surface (2.4) and bulk Mn/Pt the nanoparticles of all the catalysts are well-dispersed (concentrated at
molar ratio (2.3) also confirmed the formation of Pt-MnOx solid solution. 2.0–3.0 nm). Such a small difference in particle size could not be the
The electronic state of Mn species in the two catalysts was further decisive factor for the difference in catalyst performance. Actually, the
investigated by XANES (Fig. 1d) [45,46]. The Mn K-edge XANES spec- MCM-41 support plays an important role in the dispersion of Pt and Mn
trum of the PtMn/MCM-41 (IM) and PtMn/MCM-41 (In-70) catalysts nanoparticles due to its high surface area and regular arrangement of
exhibits the main features at 6558 eV, indicating the presence of Mn3+ hexagonal channels. The results of N2 physisorption (Fig. S2 and
in these two catalysts. The white-line intensity of the PtMn/MCM-41 Table S4) also confirm all the synthesized Mn-MCM-41 supports display
(IM) catalyst (1.24 eV) is slightly higher than that of the the typical type-IV isotherm with a high pore volume (> 0.5 cm3/g) and
PtMn/MCM-41 (In-70) catalyst (1.20 eV), suggesting that the electron surface area (> 570 m2/g) [29]. In addition, the XRD patterns also

Fig. 2. (a) HAADF-STEM images and particle size distribution of (a-1) Pt/MCM-41, (a-2) PtMn/MCM-41 (IM), (a-3) PtMn/MCM-41 (In-70) and (a-4) PtMn/MCM-41
(In-26) catalysts. (b) HRTEM images and (c) EDX-mapping of the PtMn/MCM-41 (In-70) catalyst. (d) XRD patterns of the Pt/MCM-41 and PtMn/MCM-41 catalysts.

4
H. Yan et al. Applied Catalysis B: Environmental 284 (2021) 119803

proved the uniform distribution of Pt and Mn nanoparticles on the binding energy (71.0 eV) than the PtMn/MCM-41 (IM) catalyst. This
MCM-41 support (Fig. 2d), and the only small peak located at approxi- indicates that the surface electrons in the Pt-Mn2O3 interfacial active
mately 45◦ corresponds to the existence of Mn2O3 [49]. To further sites are more abundant. In addition, doping excessive or insufficient Mn
observe the configuration of Pt and Mn nanoparticles, Raman spectra, could not further reduce the binding energy of Pt◦ (Fig. S4a), proving
element line scan, HRTEM images and EDX-mapping were further per- that only the appropriate amount of Mn could strength the electronic
formed. Raman spectra in Fig. S1b shows that the PtMn/MCM-41 (In-70) coupling effect of the interfacial active sites. The Mn 2p spectra were
catalyst exhibits a characteristic peak of Mn2O3 at ~650 cm−1 [50,51]. further analyzed to confirm that the origin of Pt electron transfer is
However, the intensity of this Mn2O3 peak in the PtMn/MCM-41 (IM) caused by the addition of Mn (Fig. 3b and Fig. S4b). Since Mn metal on
catalyst is greatly reduced and there is no other Raman characteristic the catalyst surface has many valence states, we select three most
peak of Mn species. It is highly possible that the formation of Pt-MnOx common valence states of Mn (Mn2+: ~641.0 eV, Mn3+: ~642.0 eV and
solid solution results in the decrease or disappearance of Raman char- Mn4+: ~644.0 eV) [53,54]. Apparently, the surface Mn on all the cat-
acteristic peaks of Mn2O3 in the PtMn/MCM-41 (IM) catalyst. Mean- alysts is still mainly in the form of Mn3+ (Mn3+ content > 40 %). This
while, the element line scan clearly shows that both Pt and Mn are fully indicates that Mn exists mainly in trivalent form (i.e. Mn2O3),
distributed throughout the nanoparticles in the PtMn/MCM-41 (IM) which is consistent with the results of XRD and EXAFS. Furthermore,
catalyst, further confirming the formation of Pt-MnOx solid solution compared with the reduced Mn-MCM-41 (In) sample, all the PtMn cat-
(Fig. S3a). In contrast, for the PtMn/MCM-41 (In-70) catalyst, PtMn alysts exhibit higher binding energy of the Mn2p3/2 peak, verifying the
nanoparticles tend to exist mainly in the form of adjacent isolated Pt and electron transfer from Mn to Pt. In addition, the PtMn/MCM-41 (In-70)
Mn nanoparticles (Fig. S3b). HRTEM images of the PtMn/MCM-41 catalyst with the optimal ratio (Si/Mn = 70) exhibits the higher binding
(In-70) catalyst shows that Pt (111) and Pt (100) are the most energy of Mn (642.2 eV) than that of the PtMn/MCM-41 (IM) catalyst
frequently exposed crystal faces of Pt particles [43,52], and Mn mainly (642.0 eV). Consequently, XPS spectra of Pt4f and Mn2p confirm the
exposes different crystal faces in the form of Mn2O3 oxide (Fig. 2b). electronic coupling effect of the Pt-Mn2O3 interfacial active sites is
Moreover, it is found that Pt and Mn nanoparticles are uniformly stronger than that of the active sites of the Pt-MnOx solid solution.
distributed in the form of mutual contact, and Pt nanoparticles are DFT calculation was also used to study the charge transfer between
located on the surface of Mn particles. EDX-mapping also confirms that Pt and Mn2O3 more intuitively on the basis of the interfacial structure
Mn and Pt in the PtMn/MCM-41 (In-70) catalyst are not completely revealed by EXAFS, XPS and STEM. Fig. 4a shows that the surface Pt
mixed (Fig. 2c). Therefore, Raman spectra and HRTEM together with atoms on the Pt (111) only have few negative charge (-0.058 |e|).
EXAFS provides strong evidence for the identification of the Pt-Mn2O3 Comparatively, the introduction of Mn2O3 cluster leads to electron
interfacial active sites in the PtMn/MCM-41 (In-70) catalyst. enrichment on the nearby Pt atom (increased to -0.138 and -0.176 |e|).
This electron transfer from Mn to Pt is consistent with the XANES and
3.2. Electronic coupling effect of Pt-Mn2O3 interfacial active site XPS results (Figs. 1d and 3). Moreover, the d-band of the Pt-Mn2O3
interfacial model (Pt-Mn2O3 (111)) tends to approach the Fermi level
The formation of Pt-Mn2O3 interfacial active sites in the PtMn/MCM- (0 eV) (Fig. 4b). Under the combined effect of charge transfer and d-
41 (In-70) catalyst inevitably changes the electronic coupling effect and band shift, the adsorption energy of EG increases from 45.13 kcal/mol
metal-support (or metal) interaction. XPS spectra were first measured to [Pt (111)] to 49.41 kcal/mol [Pt-Mn2O3 (111) model]. It has been
study the plausible electronic coupling significance in the interfacial demonstrated that the adsorption ability of catalyst for reactants is
active site. As shown in Fig. 3, the Pt/MCM-41 exhibits the highest closely related to its catalytic activity [55]. Therefore, the enhanced
binding energy of Pt 4f7/2 compared with other PtMn catalysts, sug- electron coupling effect of the Pt-Mn2O3 interfacial active site is critical
gesting the increased electron density on Pt surface promoted by Mn. for the excellent catalytic performance of the PtMn/MCM-41 (In-70)
Moreover, the PtMn/MCM-41 (In-70) catalyst shows much lower Pt◦ catalyst.

Fig. 3. (a) Pt 4ƒ spectra of the (a-1) Pt/MCM-41, (a-2) PtMn/MCM-41 (IM) and (a-3) PtMn/MCM-41 (In-70) catalysts. (b) Mn 2p spectra of the (b-1) reduced Mn-
MCM-41 (In-70), (b-2) PtMn/MCM-41 (IM) and (b-3) PtMn/MCM-41 (In-70) catalysts.

5
H. Yan et al. Applied Catalysis B: Environmental 284 (2021) 119803

Fig. 4. (a) Mulliken charge (|e|) distribution, (b) d-partial density and EG adsorption energy on Pt (111) and Pt-Mn2O3 (111) (dark blue ball represents the Pt atom,
red ball represents the O atom, and purple ball represents the Mn atom). (c) TPR profiles of the Mn-MCM-41 samples and PtMn/MCM-41 catalysts (For interpretation
of the references to colour in this figure legend, the reader is referred to the web version of this article.).

Essentially, the modification of electronic structure could also the existence of the Pt-Mn2O3 interfacial active sites in the PtMn/MCM-
change the reduction ability of metals. H2-TPR was employed to study 41 (In-70) catalysts is of significant importance to enhance catalytic
the reduction properties of Pt and Mn metal on the MCM-41 support. As performance. On this basis, we further prepared and evaluated the
shown in Fig. 4c, there are two reduction peaks in the Mn/MCM-41 (IM) PtMn/MCM-41 (In-x) catalysts with various in-situ doping amount of Mn
and Mn/MCM-41 (In-70) samples corresponding to the reduction of for the EG oxidation. Fig. 5 clearly shows that TOF value increases
MnO2 to Mn2O3 (~220 ◦ C) and Mn2O3 to MnO (~520 ◦ C) respectively gradually and then decreases with the increase of Mn. The PtMn/MCM-
[56,57]. Obviously, the two samples tend to be reduced to Mn2O3 under 41 (In-70) catalyst (Pt: 1.4 wt.%; Mn: 0.9 wt.%) exhibits the maximum
350 ◦ C reduction conditions. Subsequently, the reduction properties of TOF of 3196.9 h−1. XRD and N2 physisorption confirmed that intro-
Pt were investigated. For the monometallic Pt/MCM-41 catalyst, there duction of Mn does not significantly affect the physical structure of the
are two reduction peaks corresponding to hydrogen overflow (< 100 ◦ C) PtMn/MCM-41 (In-x) catalysts (Fig. S2, Fig. S5 and Table S4). Although
and the reduction of Pt species to Pt◦ (~ 300 ◦ C) [35,58]. Different the particle sizes of the Pt/MCM-41, PtMn/MCM-41 (In-360), PtMn/
incorporation methods of Mn have a noticeable effect on the reduction MCM-41 (In-200) and PtMn/MCM-41 (In-70) catalysts are almost the
temperature of Pt. The reduction temperature of Pt in the same, their catalytic activity is quite different. The decrease of the
PtMn/MCM-41 (IM) catalyst is increased due to the formation of
PtMnOx solid solution in the reduction process [56,21]. A relatively low
Pt reduction temperature (~ 300 ◦ C) could be obtained in the
PtMn/MCM-41 (In-70) catalyst, which is similar to that of the
Pt/MCM-41 catalyst, confirming the independent Pt metal at the
Pt-Mn2O3 interface. Meanwhile, the introduction of Pt can significantly
reduce the reduction temperature of MnO2 to Mn2O3 in the PtMn cata-
lysts. Notably, the PtMn/MCM-41 (In-70) catalyst with more uniform
pre-distribution of Mn displays the lowest reduction temperature of Mn
species (~ 100 ◦ C). It is highly possible that the PtMn interfacial active
sites could transfer more H generated by the hydrogen overflow acti-
vation of Pt metal to the adjacent Mn species resulting in the lower
reduction temperature. In other words, the decrease of reduction tem-
perature suggests a strong interaction between Pt and Mn2O3 in the
PtMn/MCM-41 (In-70) catalyst.

3.3. Quantitative analysis of the Pt-Mn2O3 interfacial active sites

Fig. 5. Effect of the amount of the doped Mn on the TOF, particle size, and
Previous characterization and evaluation results have proved that binding energy of Pt0.

6
H. Yan et al. Applied Catalysis B: Environmental 284 (2021) 119803

binding energy of Pt (from 71.5 eV in the Pt/MCM-41 catalyst to 71.0 eV 0.42, 0.28 and 0.22, respectively. We tried to correlate the initial reac-
in the PtMn/MCM-41 (In-70) catalyst) indicates that the more Mn is tion rate with the active sites detected by H2-TPD, while no direct
added, the more electrons are enriched on the surface of Pt metal. At this relationship was found. The initial reaction rate first increased and then
time, the enhanced electron coupling effect promotes the increase of decreased with the increase of Mn content. In contrast, the number of
TOF. However, further increase of Mn content could reduce the TOF active sites determined by H2-TPD increased all the time in the presence
values. Although the particle size increases slightly due to the increase of of Pt metal. It can be preliminarily inferred that the PtMn/MCM-41 (In-
metal loading, it is found that the catalytic activity of the PtMn/MCM-41 x) catalysts contain three active sites: I Pt metal active sites, II Mn metal
(In-6 days) catalyst (2787.5 h−1), which was prepared by loading Pt on sites and III Pt-Mn2O3 interfacial active sites. It should be noted that
the MCM-41 (In-70) crystallized for 6 days, is still higher than that of the both Pt (1692.3 μmolH2/gPt) and Mn metals (3330.7 μmolH2/gMn) have
PtMn/MCM-41 (In-26) (1619.7 h−1) and PtMn/MCM-41 (In-38) cata- strong H2 adsorption capacity. This means that although isolated Mn
lysts (1923.1 h−1) with similar particle size of 2.8 nm (Fig. S6). This metal site does not have the ability for EG oxidation, H2-TPD still regards
suggests that the particle size is not the dominant factor affecting the it as the active site and detects it. In other words, H2-TPD can detect
catalytic activity. Meanwhile, the binding energies of the PtMn/MCM- these three active sites with no distinction. Besides, both Pt and Mn
41 (In-70), PtMn/MCM-41 (In-38) and PtMn/MCM-41 (In-26) catalysts metals tend to overestimate the exposed metal sites due to hydrogen
are similar (~71.0 eV), revealing that the electron coupling effect is also spillover. Moreover, it is found that the PtMn/MCM-41 catalysts hardly
not the main reason. To investigate the underlying reason for this phe- adsorbed any CO probe molecules. Therefore, the Pt-Mn2O3 interfacial
nomenon, we further carried out the quantitative analysis of the active sites with high catalytic activity could not be identified effectively
intrinsic Pt-Mn2O3 interfacial active sites in the PtMn/MCM-41 (In-x) through the traditional H2 or CO adsorption test.
catalysts. The oxidation of EG to GA is precisely a process of hydroxyl
H2 was first used as the probe molecule to determine the active sites adsorption followed by oxidation reaction. Consequently, it is highly
of the catalysts (Fig. 6 (a-1)). The Pt/Mn molar ratios of the PtMn/MCM- possible to determine more intrinsic surface active sites with oxidation
41 (In-360), PtMn/MCM-41 (In-200), PtMn/MCM-41 (In-70), PtMn/ activity by selecting ethanol as the probe molecule. Fig. 6 (a-2) clearly
MCM-41 (In-38) and PtMn/MCM-41 (In-26) catalysts are 1.97, 0.92, shows that both the initial reaction rate and the active sites detected by

Fig. 6. (a) Active sites determined by chemisorption with different probe molecules ((a-1) H2 and (a-2) ethanol). (b) Schematic diagram of the Pt-Mn2O3 interfacial
active sites.

7
H. Yan et al. Applied Catalysis B: Environmental 284 (2021) 119803

ethanol-pulse exhibit “volcano-shape” relationship with the increase of MCM-41 (In-26) catalysts are as high as 4.9 and 10.3 (Table S1), which
Mn content, suggesting that the ethanol-pulse could obtain more further confirms our inference. Therefore, the PtMn/MCM-41 (In-70)
intrinsic active sites with high catalytic activity. It should be noted that catalyst with the optimal Pt/Mn molar ratio of 0.42 has the maximum
Pt metal exhibits much stronger adsorption capacity (308.5 μmolEt/gPt) Pt-Mn2O3 interfacial active sites and the highest initial reaction rate.
for ethanol compared with Mn metal (122.2 μmolEt/gMn). This differ- Detailed discussions on the active site and TOF were provided in the
ence in adsorption capacity provides a prerequisite for the accurate Supporting Information.
determination of Pt based active sites. Initially, with the introduction of
a small amount of Mn, the measured active sites increased, and the 3.4. Structure-dependent kinetics of selective oxidation of EG to GA
PtMn/MCM-41 (In-70) catalyst (molar ratios: 0.42) possesses the largest
number of effective active sites. This suggests that the as-formed Pt- The optimum PtMn/MCM-41 (In-70) catalyst was selected to reveal
Mn2O3 interfacial active sites strengthen the adsorption of ethanol the effect of parameters (including temperature, pressure and stability)
because of the enhanced electron transfer between Pt and Mn2O3, which on the EG oxidation performance. Fig. 7a clearly shows that the con-
is consistent with the DFT calculation results (Fig. 4b). However, further version of EG increases from 55.7 % to 100 % with reaction temperature
increase of Mn content results in the rapid decrease of the initial reaction increasing from 50 ◦ C to 90 ◦ C. Due to unfavorable deep oxidation
rate and the active sites. It is because the isolated Mn site formed by the (oxalic acid) and C–C bond cleavage (formic acid and CO2), the selec-
introduction of excess Mn does not have strong EG adsorption capacity tivity of GA and GAD is lower at relatively higher reaction temperature.
compared with Pt and Pt-Mn2O3 interfacial active sites. On this basis, it Similarly, the conversion of EG increases with the prolongation of re-
implies that the optimal Pt/Mn molar ratio is 0.42. It should be noted action time (Fig. 7b). When the reaction time is extended to more than
that the number of moles of Mn is much higher than that of Pt at this 8 h, the conversion of EG can be higher than 90 %. Meanwhile, the
time. Similar phenomenon has also been observed in the oxidation of n- selectivity of GA is also dependent on reaction time. It increases up to
hexane [59]. It is highly possible that the first introduction of more Mn 93.8 % in 8 h and then decreases slightly. Under the optimal reaction
metals can provide sufficient conditions for the formation of Pt-Mn2O3 temperature (60 ◦ C) and reaction time (8 h), the PtMn/MCM-41 (In-70)
interfacial in the actual catalyst synthesis process. catalyst exhibits the optimal EG conversion (92.1 %) and GA selectivity
Fig. 6b intuitively concludes the effect of Pt/Mn molar ratio on the (93.8 %). In contrast, under the same reaction conditions, the Pt/MCM-
formation of the interfacial active sites. When the Pt/Mn molar ratio is 41 catalyst shows only 61.2 % EG conversion and 82.4 % GA selectivity
higher than 0.42 (such as the PtMn/MCM-41 (In-360) and PtMn/MCM- (Table S5). The GAD selectivity and FA selectivity over the Pt/MCM-41
41 (In-200) catalysts), insufficient Mn could not form saturated Pt- catalyst (10.3 % and 2.3 %) is higher than that of the PtMn/MCM-41 (In-
Mn2O3 interfacial active sites with Pt. When the Pt/Mn molar ratio is 70) catalyst (2.5 % and 0.1 %). This indicates that the poor catalytic
lower than 0.42 (such as the PtMn/MCM-41 (In-38) and PtMn/MCM-41 activity of the Pt/MCM-41 catalyst could not convert GAD to GA rapidly,
(In-26) catalysts), excessive Mn could block the generated Pt-Mn2O3 and even lead to the occurrence of side reactions such as C–C bond
interfacial active sites, preventing the adsorption and activation of EG. cleavage. In addition, the catalytic stability of the Pt/MCM-41 and
Moreover, surface Mn/Pt ratios of the PtMn/MCM-41 (In-38) and PtMn/ PtMn/MCM-41 (In-70) catalyst was also investigated under mild

Fig. 7. Catalytic performance of EG oxidation on the PtMn/MCM-41 (In-70) catalyst as a function of (a) reaction temperature (0.2 g catalyst, 25 mL 0.3 M EG
soloution, 8 h, 1 MPa O2) and (b) reaction time (0.2 g catalyst, 25 mL 0.3 M EG soloution, 1 MPa O2, 60 ◦ C). GAD and FA represent glycolaldehyde and formic acid
respectively. (c) Catalytic stability of the PtMn/MCM-41 (In-70) catalyst. (Reaction condition: 0.2 g catalyst, 25 mL 0.3 M EG soloution, 1 MPa O2, 60 ◦ C, 8 h). (d)
Apparent activation energies of the Pt/MCM-41, PtMn/MCM-41 (IM) and PtMn/MCM-41 (In-70) catalysts.

8
H. Yan et al. Applied Catalysis B: Environmental 284 (2021) 119803

reaction conditions (Fig. 7c, Fig. S7, Table S5 and Table S6). After four (Fig. S9), meaning that the PtMn/MCM-41 (In-70) catalyst strongly
recycles, the PtMn/MCM-41 (In-70) catalyst still exhibits approximately adsorbs oxygen, probably due to the relatively faster replenishment rate
80 % EA yield, while the EA yield on the Pt/MCM-41 catalyst rapidly of the consumed surface molecular oxygen [60]. Moreover, the activa-
decreases to 27.0 %. Table S6 shows that Pt metal leaching occurs in the tion energies for EG oxidation on the Pt/MCM-41, PtMn/MCM-41 (IM)
used Pt/MCM-41 catalyst with the increasing of particle size (from and PtMn/MCM-41 (In-70) catalysts are 48.8, 36.6 and 28.3 kJ/mol,
2.3 nm to 4.1 nm) during the recycle experiment, and the used Pt/MCM- respectively (Fig. 7d). Obviously, the PtMn/MCM-41 (In-70) catalyst
41 catalyst exhibits almost no significant change in the binding energy of exhibits the lowest activation energy of EG oxidation. Previous XPS and
Pt◦ . Therefore, the rapid decline of catalytic performance of the Pt/ Mulliken charge distribution have proved that the electrons on the Pt
MCM-41 catalyst is mainly due to the leaching of Pt and agglomera- surface are enriched due to the existence of the Pt-Mn interfacial
tion of Pt nanoparticles. In contrast, the content of Pt metal in the used structure of the PtMn/MCM-41 (In-70) catalyst. It has been demon-
PtMn/MCM-41 (In-70) catalyst is almost unchanged, and the particle strated that the enrichment of electrons on Pt surface is beneficial to
size is still 2.2 nm. However, the Mn content in the PtMn/MCM-41 (In- enhance the catalytic activity in polyol oxidation system [6,21,29].
70) catalyst decreased from 0.9 wt.% to 0.3 wt.%, which was probably Therefore, we infer that the electron transfer from Mn2O3 to Pt could
due to the corrosion of formed GA product. Furthermore, the loss of Mn strengthen the adsorption of EG on the Pt-Mn interfacial active sites,
weakens the electron transfer ability between Pt and Mn, which in- resulting in the lower the activation energy of EG oxidation over the
creases the binding energy of Pt◦ from 71.0 eV to 71.3 eV (Fig. S7). It PtMn/MCM-41 (In-70) catalyst.
should be noted that the catalytic performance of the PtMn/MCM-41 (In-
70) catalyst decreased slowly, since the remaining Mn metal can still
3.5. Oxidation mechanism of EG to GA
form part of the Pt-Mn2O3 interfacial active sites.
The structure-dependent kinetics on the Pt/MCM-41, PtMn/MCM-41
DFT theoretical calculation was further employed to explore the
(IM) and PtMn/MCM-41 (In-70) catalysts were then systematically
detailed reaction elementary steps of the EG oxidation to GA at the Pt-
studied for EG oxidation. All the catalysts were evaluated with negligible
Mn2O3 interfacial active sites as well as Pt (111). There are four steps in
mass transfer limitations. Detailed calculations were provided in the
the formation of GA: (Step 1) dehydrogenation of hydroxyl group, (Step
Supporting Information (Fig. S8-S9). The purpose of kinetic analysis was
2) β-hydride elimination, (Step 3) oxidation of GAD and (Step 4)
to quantitatively analyze the role of Mn metal and Pt-Mn2O3 interfacial
dehydrogenation of C–H bond (Fig. 8). Commonly, before starting the
active sites in EG oxidation. Commonly, the EG concentration (CG) and
alcohol oxidation reaction, H2O could initially assist O2 to dissociate
O2 pressure (PO) could affect the reaction rate. The reaction order for CG
into hydroxyl groups [38,61]. The as-formed hydroxyl groups could
(a) of the Pt/MCM-41, PtMn/MCM-41 (IM) and PtMn/MCM-41 (In-70)
further participate subsequent oxidation reaction, and the dissociated
catalysts are 0.50, 0.44 and 0.38, respectively (Fig. S8). This suggests
O* will continue to complete the cycle reaction of O2 and H2O with the
that the PtMn catalysts have stronger adsorption ability for EG than
formation of OH* by combining with H protons (formed in the dehy-
monometallic Pt/MCM-41 catalyst, and the Pt-Mn2O3 interfacial active
drogenation reaction: Step-2 and Step-4). As shown in Fig. 8 and
site in the PtMn/MCM-41 (In-70) catalyst further promotes the adsorp-
Fig. S10a, the Pt-Mn2O3 (111) model shows lower activation energy of
tion of EG, which is well consistent with the DFT calculation results.
H2O-assisted O2 dissociation than the Pt (111) model. Subsequently, the
Meanwhile, the PtMn/MCM-41 (In-70) catalyst displays the minimum
activation energy barriers of detailed oxidation steps were calculated on
reaction order for PO (b) (0.26) compared with the other two catalysts
the Pt (111) and Pt-Mn2O3 (111) models (Fig. 8: Step 1 - Step 5). It is

Fig. 8. Energy profiles and configuration diagrams of reactants (R), transition states (TS) and products (P) for the oxidation of ethylene glycol to glycolic acid on the
Pt (111) and Pt-Mn2O3 (111) models. (O atoms colored red. C atoms colored gray. H atoms colored white. Pt atom colored dark blue. Mn atom colored violet.) (For
interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.).

9
H. Yan et al. Applied Catalysis B: Environmental 284 (2021) 119803

found that the Pt-Mn2O3 (111) model has a lower energy barrier for Step original draft. Shuang Yao: Methodology, Formal analysis, Investiga-
2 (0.23 eV) and Step 4 (0.21 eV). These two steps are the rupture of the tion. Jinyao Wang: Conceptualization. Siming Zhao: Investigation.
C–H bond, suggesting that C–H bond activation is more prone to occur Yinghao Sun: Investigation. Mengyuan Liu: Supervision, Funding
on the Pt-Mn2O3 interfacial active sites. Interestingly, there is an obvious acquisition. Xin Zhou: Project administration, Validation. Guangyu
synergistic catalytic effect between Mn2O3 and Pt. Fig. S10b and Zhang: Writing - review & editing. Xin Jin: Investigation. Xiang Feng:
Fig. S10c shows that the Mn2O3 could assist the adsorption of interme- Writing - review & editing. Yibin Liu: Writing - review & editing.
diate products on the Pt-Mn2O3 interfacial active sites, leading to the Xiaobo Chen: Formal analysis. De Chen: Project administration, Vali-
increase of the electronegativity of Pt atom and adsorbed C atom in the dation, Writing - review & editing. Chaohe Yang: Supervision, Funding
transition state. The adsorption energies of intermediates CH2OHCH2O* acquisition, Validation.
(Step 2) and CH2OHCHOOH* (Step 4) on the Pt-Mn2O3 (3.47 eV and
3.41 eV) are significantly enhanced compared with those on the Pt
Declaration of Competing Interest
(2.45 eV and 2.26 eV). This synergistic effect between the Mn2O3 redox
cycle and Pt makes the dehydrogenation of C–H bond more favorable.
The authors report no declarations of interest.
Consequently, it can be concluded that the electron transfer and syn-
ergistic adsorption effect between Pt and Mn2O3 together facilitate the
Acknowledgments
activation of C–H bond in the intermediate products. Combined with
the results of reaction kinetics, the Pt-Mn2O3 interfacial active sites
Our work was supported by the National Science Foundation of
could reduce the apparent activation energy by promoting the activation
China (Grant No. 21776312, 21978325 and 21606254) and the Fun-
of the C–H bond, thus enhancing the catalytic activity of EG oxidation.
damentalResearch Funds for the Central Universities (20CX02210A,
Actually, the synergistic effect of similar interfacial active sites has also
20CX06072A, 20CX06095A and 20CX06096A).
been observed on the Pt-Co/MCM-41 catalyst [29]. It has been
demonstrated that the introduction of Co to Pt could lead to the for-
mation of Pt-Co interfacial active sites, thus promoting the selective Appendix A. Supplementary data
oxidation of glycerol to glyceric acid. Similarly, the cooperative effect
between Pt and SbOx was observed in the oxidation of glycerol to Supplementary material related to this article can be found, in the
dihydroxyacetone [43]. These further indicate that the interfacial active online version, at doi:https://doi.org/10.1016/j.apcatb.2020.119803.
sites could be beneficial to the oxidation of polyols. In addition, the
activation energy of undesired C–C bond cleavage side reaction (Step 5) References
on the Pt-Mn2O3 (111) model (1.56 eV) is significantly higher than that
[1] G. Cui, X. Meng, X. Zhang, W. Wang, S. Xu, Y. Ye, K. Tang, W. Wang, J. Zhu,
on the Pt (111) model. Consequently, the Pt-Mn2O3 interfacial active site M. Wei, D.G. Evans, X. Duan, Low-temperature hydrogenation of dimethyl oxalate
facilitates the activation of C–H bond and prevents the cleavage of C–C to ethylene glycol via ternary synergistic catalysis of Cu and acid−base sites, Appl.
bond resulting in the improvement of catalytic performance. Catal. B 248 (2019) 394–404.
[2] A. Chen, X. Yu, Y. Zhou, S. Miao, Y. Li, S. Kuld, J. Sehested, J. Liu, T. Aoki, S. Hong,
On the basis of the above results, the Pt-Mn2O3 interfacial rather than M.F. Camellone, S. Fabris, J. Ning, C. Jin, C. Yang, A. Nefedov, C. Wöll, Y. Wang,
Pt-MnOx solid solution is the desirable active site for EG oxidation to GA. W. Shen, Structure of the catalytically active copper-ceria interfacial perimeter,
The pre-distribution of Mn2O3 inside support could induce the formation Nat. Catal. 2 (2019) 334–341.
[3] H. Yue, Y. Zhao, X. Ma, J. Gong, Ethylene glycol: properties, synthesis, and
of Pt-Mn2O3 interfacial active sites with strong electron coupling effect. applications, Chem. Soc. Rev. 41 (2012) 4218–4244.
By performing the ethanol-pulse test reaction kinetics and DFT calcu- [4] H. Shi, X. Yin, B. Subramaniam, R.V. Chaudhari, Liquid-phase oxidation of
lation, the number of intrinsic active sites (mainly the Pt-Mn2O3 inter- ethylene glycol on Pt and Pt−Fe catalysts for the production of glycolic acid:
remarkable bimetallic effect and reaction mechanism, Ind. Eng. Chem. Res. 58
face active sites) could be quantitatively determined, and the Pt-Mn2O3
(2019) 18561–18568.
interfacial active sites could promote the adsorption of ethylene glycol [5] H. Du, K. Wang, P. Tsiakaras, P.K. Shen, Excavated and dendritic Pt-Co nanocubes
and the activation of C–H bond. Moreover, the occurrence of C–C as efficient ethylene glycol and glycerol oxidation electrocatalysts, Appl. Catal. B
258 (2019), 117951.
cleavage side reaction could be effectively inhibited. It is expected that
[6] Y. Wang, M. Zheng, H. Sun, X. Zhang, C. Luan, Y. Li, L. Zhao, H. Zhao, X. Dai, J. Ye,
the methodology of engineering desirable active sites could be extended H. Wang, S. Sun, Catalytic Ru containing Pt3Mn nanocrystals enclosed with high-
to the synthesis of other high-performance Pt based catalysts for the indexed facets: Surface alloyed Ru makes Pt more active than Ru particles for
efficient conversion of polyols. ethylene glycol oxidation, Appl. Catal. B 253 (2019) 11–20.
[7] Q. Yang, S. Zhu, P. Yu, Q. Yang, D. Zhang, Thermodynamic and techno-economic
analysis of coal to ethylene glycol process (CtEG) with different coal gasifiers,
4. Conclusion Energ. Convers. Manage. 191 (2019) 80–92.
[8] L.S. Ribeiro, N. Rey-Raap, J.L. Figueiredo, J.J. Melo Órfão, M.F.R. Pereira, Glucose-
based carbon materials as supports for the efficient catalytic transformation of
In summary, electronically coupled PtMn/MCM-41 catalyst with Pt- cellulose directly to ethylene glycol, Cellulose 26 (2019) 7337–7353.
Mn2O3 interface was prepared by the combination of impregnation and [9] S. Kandasamy, S.P. Samudrala, S. Bhattacharya, The route towards sustainable
in situ doping method. Multi-characterizations (mainly XANES, EXAFS, production of ethylene glycol from a renewable resource, biodiesel waste: a review,
Catal. Sci. Technol. 9 (2019) 567–577.
HRTEM and XPS) confirmed the existence of Pt-Mn2O3 interfacial active [10] F. Gao, Y. Zhang, P. Song, J. Wang, T. Song, C. Wang, L. Song, Y. Shiraishi, Y. Du,
sites and observed enhanced electron transfer from Mn2O3 to Pt. Precursor-mediated size tuning of monodisperse PtRh nanocubes as efficient
Furthermore, the ethanol-pulse test accurately detected the intrinsic electrocatalysts for ethylene glycol oxidation, J. Mater. Chem. A 7 (2019)
7891–7896.
active sites with high catalytic activity (mainly the Pt-Mn2O3 interface [11] Y. Zou, N. von Wolff, A. Anaby, Y. Xie, D. Milstein, Ethylene glycol as an efficient
active sites). In addition, structure-dependent kinetics proved the pro- and reversible liquid-organic hydrogen carrier, Nat. Catal. 2 (2019) 415–422.
motion effect of the interfacial active sites on the adsorption and acti- [12] Y. Zhang, F. Gao, P. Song, J. Wang, J. Guo, Y. Shiraishi, Y. Du, Glycine-assisted
fabrication of N-doped graphene-supported uniform multipetal PtAg nanoflowers
vation of EG. Meanwhile, the detailed reaction pathway clarified by DFT
for enhanced ethanol and ethylene glycol oxidation, ACS Sustain. Chem. Eng. 7
calculation demonstrated that the Pt-Mn2O3 interfacial active site fa- (2019) 3176–3184.
cilitates the activation of C–H bond and prevents the cleavage of C–C [13] H. Berndt, I. Pitsch, S. Evert, K. Struve, M.M. Pohl, J. Radnik, A. Martin, Oxygen
bond, resulting in the improvement of catalytic performance of EG adsorption on Au/Al2O3 catalysts and relation to the catalytic oxidation of
ethylene glycol to glycolic acid, Appl. Catal. A 244 (2003) 169–179.
oxidation to GA. [14] G. Mattioda, B. Metivier, J.P. Guette, What you can do with glyoxal, Chemtech 13
(1983) 478–481.
CRediT authorship contribution statement [15] M.J. Climent, A. Corma, S. Iborra, Heterogeneous catalysts for the one-pot
synthesis of chemicals and fine chemicals, Chem. Rev. 111 (2011) 1072–1133.
[16] T. Mallat, A. Baiker, Oxidation of alcohols with molecular oxygen on solid
Hao Yan: Conceptualization, Methodology, Investigation, Writing - catalysts, Chem. Rev. 104 (2004) 3037–3058.

10
H. Yan et al. Applied Catalysis B: Environmental 284 (2021) 119803

[17] S. Suzuki, Process for the production of glycolic acid and oxydiacetic acid, Google [38] H. Yan, S. Yao, W. Liang, X. Feng, X. Jin, Y. Liu, X. Chen, C. Yang, Selective
Patents, 1975. oxidation of glycerol to carboxylic acids on Pt(111) in base-free medium: a periodic
[18] O.M. Koivistoinen, J. Kuivanen, D. Barth, H. Turkia, J.P. Pitkanen, M. Penttila, density functional theory investigation, Appl. Surf. Sci. 497 (2019), 143661.
P. Richard, Microb. Glycolic acid production in the engineered yeasts [39] G. Zhang, X. Xu, Q. Ji, R. Liu, H. Liu, J. Qu, J. Li, Porous nanobimetallic Fe-Mn
Saccharomyces cerevisiae and Kluyveromyces lactis, Cell Fact. 12 (2013) 82. cubes with high valent Mn and highly efficient removal of arsenic(III), ACS Appl.
[19] A. Panova, L.J. Mersinger, Q. Liu, T. Foo, D.C. Roe, W.L. Spillan, A.E. Sigmund, Mater. Inter. 9 (2017) 14868–14877.
A. Ben-Bassat, L.W. Wagner, D.P. O’Keefe, S. Wu, K.L. Petrillo, M.S. Payne, S. [40] Z. Luo, Y. Ouyang, H. Zhang, M. Xiao, J. Ge, Z. Jiang, J. Wang, D. Tang, X. Cao,
T. Breske, F.G. Gallagher, R. DiCosimo, Chemoenzymatic synthesis of glycolic acid, C. Liu, W. Xing, Chemically activating MoS2 via spontaneous atomic palladium
Adv. Synth. Catal. 349 (2007) 1462–1474. interfacial doping towards efficient hydrogen evolution, Nat. Commun. 9 (2018).
[20] M. Kataoka, M. Sasaki, A.G.D. Hidalgo, M. Nakano, S. Shimizu, Glycolic acid [41] X. Wang, W. Zhao, T. Zhang, Y. Zhang, L. Jiang, S. Yin, Facile fabrication of shape-
production using ethylene glycol-oxidizing microorganisms, Biosci Biotech Bioch. controlled CoxMnyOβ nanocatalysts for benzene oxidation at low temperatures,
65 (2001) 2265–2270. Chem. Commun. 54 (2018) 2154–2157.
[21] H. Yan, S. Yao, B. Yin, W. Liang, X. Jin, X. Feng, Y. Liu, X. Chen, C. Yang, [42] X. Jin, C. Zeng, W. Yan, M. Zhao, P. Bobba, H. Shi, P.S. Thapa, B. Subramaniam, R.
Synergistic effects of bimetallic PtRu/MCM-41 nanocatalysts for glycerol oxidation V. Chaudhari, Lattice distortion induced electronic coupling results in exceptional
in base-free medium: structure and electronic coupling dependent activity, Appl. enhancement in the activity of bimetallic PtMn nanocatalysts, Appl. Catal. A Gen.
Catal. B 259 (2019), 118070. 534 (2017) 46–57.
[22] X. Zhang, D. Zhou, X. Wang, J. Zhou, J. Li, M. Zhang, Y. Shen, H. Chu, Y. Qu, [43] X. Duan, Y. Zhang, M. Pan, H. Dong, B. Chen, Y. Ma, G. Qian, X. Zhou, J. Yang,
Overcoming the deactivation of Pt/CNT by introducing CeO2 for selective base-free D. Chen, SbOx-promoted Pt nanoparticles supported on CNTs as catalysts for base-
glycerol-to-Glyceric acid oxidation, ACS Catal. (2020) 3832–3837. free oxidation of glycerol to dihydroxyacetone, AIChE J. 64 (2018) 3979–3987.
[23] H. Yan, S. Yao, W. Liang, S. Zhao, X. Jin, X. Feng, Y. Liu, X. Chen, C. Yang, Ni-Co [44] Y. Xiao, J. Greeley, A. Varma, Z. Zhao, G. Xiao, An experimental and theoretical
oxide catalysts with lattice distortions for enhanced oxidation of glycerol to study of glycerol oxidation to 1,3-dihydroxyacetone over bimetallic Pt-Bi catalysts,
glyceric acid, J. Catal. 381 (2020) 248–260. AIChE J. 63 (2017) 705–715.
[24] H. Yan, H. Qin, X. Feng, X. Jin, W. Liang, N. Sheng, C. Zhu, H. Wang, B. Yin, Y. Liu, [45] A. Campos, N. Lohitharn, A. Roy, E. Lotero, J.G. Goodwin, J.J. Spivey, An activity
X. Chen, C. Yang, Synergistic Pt/MgO/SBA-15 nanocatalysts for glycerol oxidation and XANES study of Mn-promoted, Fe-based Fischer-Tropsch catalysts, Appl. Catal.
in base-free medium: catalyst design and mechanistic study, J. Catal. 370 (2019) A Gen. 375 (2010) 12–16.
434–446. [46] D. Fandeur, F. Juillot, G. Morin, L. Olivi, A. Cognigni, S.M. Webb, J. Ambrosi,
[25] Y. Pan, G. Wu, Y. He, J. Feng, D. Li, Identification of the Au/ZnO interface as the E. Fritsch, F. Guyot, J.G.E. Brown, XANES evidence for oxidation of Cr(III) to Cr
specific active site for the selective oxidation of the secondary alcohol group in (VI) by Mn-Oxides in a lateritic regolith developed on serpentinized ultramafic
glycerol, J. Catal. 369 (2019) 222–232. rocks of New Caledonia, Environ. Sci. Technol. 43 (2009) 7384–7390.
[26] Y. Shen, Y. Li, H. Liu, Base-free aerobic oxidation of glycerol on TiO2-supported [47] Z. Zhang, Y. Zhu, H. Asakura, B. Zhang, J. Zhang, M. Zhou, Y. Han, T. Tanaka,
bimetallic Au-Pt catalysts, J. Energy Chem. 24 (2015) 669–673. A. Wang, T. Zhang, N. Yan, Thermally stable single atom Pt/m-Al2O3 for selective
[27] T. Gao, J. Chen, W. Fang, Q. Cao, W. Su, F. Dumeignil, Ru/MnCe1O catalysts with hydrogenation and CO oxidation, Nat. Commun. 8 (2017).
enhanced oxygen mobility and strong metal-support interaction: exceptional [48] A.L. Ankudinov, J.J. Rehr, J.J. Low, S.R. Bare, Theoretical interpretation of XAFS
performances in 5-hydroxymethylfurfural base-free aerobic oxidation, J. Catal. 368 and XANES in Pt clusters, Top. Catal. 18 (2002) 3–7.
(2018) 53–68. [49] X. Zhang, H. Li, F. Hou, Y. Yang, H. Dong, N. Liu, Y. Wang, L. Cui, Synthesis of
[28] X. Jin, M. Zhao, J. Shen, W. Yan, L. He, P.S. Thapa, S. Ren, B. Subramaniam, R. highly efficient Mn2O3 catalysts for CO oxidation derived from Mn-MIL-100, Appl.
V. Chaudhari, Exceptional performance of bimetallic Pt1Cu3/TiO2 nanocatalysts Surf. Sci. 411 (2017) 27–33.
for oxidation of gluconic acid and glucose with O2 to glucaric acid, J. Catal. 330 [50] S. Bernardini, F. Bellatreccia, A. Casanova Municchia, G. Della Ventura, A. Sodo,
(2015) 323–329. Raman spectra of natural manganese oxides, J. Raman Spectrosc. 50 (2019)
[29] H. Yan, H. Qin, W. Liang, X. Jin, Y. Zhang, X. Feng, Y. Liu, X. Chen, C. Yang, 873–888.
Enhanced performance of bimetallic PtCo/MCM-41 catalysts for glycerol oxidation [51] F. Buciuman, F. Patcas, R. Craciun, D.R.T. Zahn, Vibrational spectroscopy of bulk
in base-free medium, Catal. Sci. Technol. 9 (2019) 4909–4919. and supported manganese oxides, Phys. Chem. Chem. Phys. 1 (1999) 185–190.
[30] L. Lei, H. Liu, Z. Wu, Z. Qin, G. Wang, J. Ma, L. Luo, W. Fan, J. Wang, Aerobic [52] M. Pan, J. Wang, W. Fu, B. Chen, J. Lei, W. Chen, X. Duan, D. Chen, G. Qian,
oxidation of alcohols over isolated single Au atoms supported on CeO2 nanorods: X. Zhou, Active sites of Pt/CNTs nanocatalysts for aerobic base-free oxidation of
catalysis of interfacial [O-Ov-Ce-O-Au] sites, ACS Appl. Nano. Mat. 2 (2019) glycerol, Green Energy Environ. (2019).
5214–5223. [53] E.S. Ilton, J.E. Post, P.J. Heaney, F.T. Ling, S.N. Kerisit, XPS determination of Mn
[31] M. Tinoco, S. Fernandez-Garcia, A. Villa, J.M. Gonzalez, G. Blanco, A.B. Hungria, oxidation states in Mn (hydr)oxides, Appl. Surf. Sci. 366 (2016) 475–485.
L. Jiang, L. Prati, J.J. Calvino, X. Chen, Selective oxidation of glycerol on [54] M.C. Biesinger, B.P. Payne, A.P. Grosvenor, L.W.M. Lau, A.R. Gerson, R.S.C. Smart,
morphology controlled ceria nanomaterials, Catal. Sci. Technol. 9 (2019) Resolving surface chemical states in XPS analysis of first row transition metals,
2328–2334. oxides and hydroxides: Cr, Mn, Fe, Co and Ni, Appl. Surf. Sci. 257 (2011)
[32] J. Saavedra, H.A. Doan, C.J. Pursell, L.C. Grabow, B.D. Chandler, The critical role 2717–2730.
of water at the gold-titania interface in catalytic CO oxidation, Science 345 (2014) [55] B. Zhang, G. Sun, S. Ding, H. Asakura, J. Zhang, P. Sautet, N. Yan, Atomically
1599. dispersed Pt1 -Polyoxometalate catalysts: how does metal-support interaction affect
[33] G. Wu, G. Zhao, J. Sun, X. Cao, Y. He, J. Feng, D. Li, The effect of oxygen vacancies stability and hydrogenation activity? J. Am. Chem. Soc. (2019).
in ZnO at an Au/ZnO interface on its catalytic selective oxidation of glycerol, [56] E.R. Stobbe, B.A. de Boer, J.W. Geus, The reduction and oxidation behaviour of
J. Catal. 377 (2019) 271–282. manganese oxides, Catal. Today 47 (1999) 161–167.
[34] M. Selvaraj, P.K. Sinha, K. Lee, I. Ahn, A. Pandurangan, T.G. Lee, Synthesis and [57] P.R. Ettireddy, N. Ettireddy, S. Mamedov, P. Boolchand, P.G. Smirniotis, Surface
characterization of Mn-MCM-41and Zr-Mn-MCM-41, Micropor. Mesopor. Mat. 78 characterization studies of TiO2 supported manganese oxide catalysts for low
(2005) 139–149. temperature SCR of NO with NH3, Appl. Catal. B 76 (2007) 123–134.
[35] I. Gandarias, P.L. Arias, J. Requies, M.B. Güemez, J.L.G. Fierro, Hydrogenolysis of [58] P. Panagiotopoulou, J. Papavasiliou, G. Avgouropoulos, T. Ioannides, D.
glycerol to propanediols over a Pt/ASA catalyst: the role of acid and metal sites on I. Kondarides, Water-gas shift activity of doped Pt/CeO2 catalysts, Chem. Eng. J.
product selectivity and the reaction mechanism, Appl. Catal. B 97 (2010) 248–256. 134 (2007) 16–22.
[36] C. Hu, L. Zhang, Z. Zhao, J. Luo, J. Shi, Z. Huang, J. Gong, Edge sites with [59] B. Grbic, N. Radic, B. Markovic, P. Stefanov, D. Stoychev, T. Marinova, Influence of
unsaturated coordination on core-shell Mn3O4@MnxCo3-xO4Nanostructures for manganese oxide on the activity of Pt/Al2O3 catalyst for CO and n-hexane
electrocatalytic water oxidation, Adv. Mater. 29 (2017), 1701820. oxidation, Appl. Catal. B 64 (2006) 51–56.
[37] H. Su, Y. Gorlin, I.C. Man, F. Calle-Vallejo, J.K. Nørskov, T.F. Jaramillo, [60] Y. Feng, W. Li, M. Meng, H. Yin, J. Mi, Mesoporous Sn(IV) doping MCM-41
J. Rossmeisl, Identifying active surface phases for metal oxide electrocatalysts: a supported Pd nanoparticles for enhanced selective catalytic oxidation of 1,2-pro-
study of manganese oxide bi-functional catalysts for oxygen reduction and water panediol to pyruvic acid, Appl. Catal. B 253 (2019) 111–120.
oxidation catalysis, Phys. Chem. Chem. Phys. 14 (2012) 14010. [61] B.N. Zope, D.D. Hibbitts, M. Neurock, R.J. Davis, Reactivity of the gold/water
interface during selective oxidation catalysis, Science 330 (2010) 74–78.

11