Carbonaceous Material Modified MoO2 Nanospheres with Oxygen Vacancies for Enhanced Visible-Light Photocatalytic Oxidative Coupling of Benzylamine

Surface oxygen vacancy (OV) plays a pivotal role in the activation of molecular oxygen and separation of electrons and holes in photocatalysis. Herein, carbonaceous materials-modified MoO2 nanospheres with abundant surface OVs (MoO2/C-OV) were successfully synthesized via glucose hydrothermal processes. In situ introduction of carbonaceous materials triggered a reconstruction of the MoO2 surface, which introduced abundant surface OVs on the MoO2/C composites. The surface oxygen vacancies on the obtained MoO2/C-OV were confirmed via electron spin resonance spectroscopy (ESR) and X-ray photoelectron spectroscopy (XPS). The surface OVs and carbonaceous materials boosted the activation of molecular oxygen to singlet oxygen (1O2) and superoxide anion radical (•O2−) in selectively photocatalytic oxidation of benzylamine to imine. The conversion of benzylamine was 10 times that of pristine MoO2 nanospheres with a high selectivity under visible light irradiation at 1 atm air pressure. These results open an avenue to modify Mo-based materials for visible light-driven photocatalysis.


Introduction
Photocatalytic oxidative coupling of amines to imines with a green, efficient and economical process has attracted great attention owing to the essential intermediates of imines for the production of chemical and pharmaceutical agents [1][2][3]. Researchers have made enormous efforts to develop high-performance photocatalysts for an innovative imines synthetic strategy [4][5][6]. Many kinds of photocatalysts have been developed for oxidation coupling of amines, such as plasmonic metals (Au, Ag, Cu, etc.) [7,8], semiconductor oxides (TiO 2 , WO 3 , BiOCl, etc.) [9,10], carbon materials [11,12], polymers [13] and so on. Semiconductor oxides possess environmental friendliness, cost-effectiveness, and excellent capability to harvest sunlight [14,15]. Among them, molybdenum dioxide is a promising candidate for semiconductor photocatalyst due to its cheapness, chemical stability and green synthesis [16]. MoO 2 is the rutile-type transition metallic oxide and has been widely studied due to its visible light absorption capacity and specific metallic properties, which differ from those of other oxides [17]. The existence of a large amount of free electrons makes MoO 2 trap electrons and enhance charge separation, which is profitable for improving the catalytic performance [18]. Additionally, MoO 2 is widely utilized in the various fields owing to acid and alkali resistance and thermal stability [19,20]. In order to enhance the photocatalytic performance of MoO 2 , the surficial state of semiconductor oxides could be adjusted by post-treatment to construct a surface defect structure, improving charge separation and the interaction between photocatalyst and reactant molecules [21].
Oxygen vacancies (OVs) have been confirmed to act as electron donors and adsorption sites and photocatalysis active sites for O 2 adsorption, which is highly desirable for enhancing the photocatalytic performance [22][23][24]. It has been accepted that the bulk defects are the recombination centers of photo-induced electrons and holes, which is unfavorable for photocatalytic reaction [25,26]. In comparison, surface OVs play an important role not only in the extension of the photo response, but also in increasing the charge separation efficiency [27][28][29]. Semiconductor photocatalysts with rich surface OVs, such as TiO 2 , BiO 2 , MoO 3 , WO 3 , ZnO and so on, have been extensively reported [16,[30][31][32][33]. BiOCl nanosheets with surface OVs resulted in abundant surface low-coordinated Bi atoms, contributing to the assembly of O 2 molecules on the surface of photocatalyst [34]. The appropriate concentration of surface OVs on TiO 2 nanosheets promotes the activation of O 2 in photocatalytic oxidative of benzylamine to imine [35]. The visible-light response ability of these photocatalysts still needs to be further improved. The common strategies for introducing surface OVs into the semiconductors include thermal treatment, chemical reduction, ultraviolet irradiation and so on [25,36,37]. Generally, these methods always involved high temperature or production of hydrogen [36]. Therefore, it makes sense to develop a simple and effective method to incorporate surface OVs into photocatalysts [27,38]. In situ introduction of carbonaceous precursors can trigger a reconstruction of the oxide surface and the formation of nanocomposites with interfacial disordered regions. [39] In this paper, carbonaceous materials modified MoO 2 nanospheres with abundant surface OVs (MoO 2 /C-OV) were synthesized by glucose hydrothermal processes aimed at the oxidative coupling of benzylamine. The surface OVs were constructed by the reconstruction of the MoO 2 surface caused by the in situ introduction of carbonaceous precursors after glucose hydrothermal treatment. The presence of surface OVs could effectively improve the photocatalytic transformation of benzylamine to its corresponding imines under visible light irradiation at 1 atm pressure. The morphology, surface chemical states and electronic structures of the MoO 2 /C-OV were further elucidated by a series of characterization techniques. The unsaturated sites caused by surface OVs sites strongly interacted with molecular oxygen, which could establish a photoexcited electron transport channel. The adsorption and activation of O 2 molecules produced active 1 O 2 and •O 2 − species with the assistance of the surface OVs and carbonaceous materials.

Synthesis and Characterization of the Samples
The phase, morphology and crystal structure of the samples were characterized using an X-ray powder diffractometer (XRD), a scanning electron microscope (SEM), a transmission electron microscope (TEM), a high-resolution transmission electron microscope (HRTEM), selected area electron diffraction (SAED) and an X-ray energy spectrometer (EDS). The morphology of the prepared samples was analyzed via SEM and TEM. As shown in Figure S1a, MoO 2 nanospheres with a diameter of about 500 nm consisted of a large number of tightly connected irregular nanoparticles. As shown in Figure 1a, there was no obvious change in the morphology of MoO 2 after hydrothermal treatment with glucose. High-magnification SEM images (Figure 1b) further exhibited rough surface and sharp projections of MoO 2 nanospheres, which would provide more surface catalytic active sites. As shown in Figure 1c,d, the self-assembled nanospheres with a rough surface are composed of many small nanoparticles approximately several tens of nanometers in size. These nanospheres are not loose aggregations of the small nanoparticles. As shown in Figure S2, the structure and morphology of MoO 2 nanoparticles remained stable after hydrothermal and high-temperature calcination. Figure 1e displayed that the MoO 2 nanoparticle was tightly connected with carbonaceous materials. A clear disordered area was constructed at the interface between the MoO 2 nanosphere and carbonaceous material, revealing more active Mo atoms and providing active sites. The inset in Figure 1e is the corresponding High resolution-TEM (HRTEM), it could be confirmed that the MoO 2 nanospheres owned clear lattice fringes. The spacing with 0.28 nm of the adjacent lattice planes was consistent with Molecules 2023, 28, 4739 3 of 13 the (−102) plane of MoO 2 . As shown in Figure 1f, the characteristic polycrystalline SAED pattern showed the (011), (020), (012), (220) plane, in accordance with the XRD diffraction pattern. The diffraction ring of carbon is not clearly observed in the pattern, indicating that the carbonaceous material was in an amorphous state. As shown in the elemental mapping images (Figure 1g), a certain amount of C was uniformly distributed on the whole MoO 2 nanosphere, expressing the successful fabrication of the MoO 2 /C-OV catalyst. The elemental mapping images of MoO 2 and MoO 2 /C were shown in Figures S3 and S4. The results revealed that a unique interface microstructure was formed after glucose hydrothermal treatment.
Molecules 2023, 28, x FOR PEER REVIEW 3 of 13 constructed at the interface between the MoO2 nanosphere and carbonaceous material, revealing more active Mo atoms and providing active sites. The inset in Figure 1e is the corresponding High resolution-TEM (HRTEM), it could be confirmed that the MoO2 nanospheres owned clear lattice fringes. The spacing with 0.28 nm of the adjacent lattice planes was consistent with the (−102) plane of MoO2. As shown in Figure 1f, the characteristic polycrystalline SAED pattern showed the (011), (020), (012), (220) plane, in accordance with the XRD diffraction pattern. The diffraction ring of carbon is not clearly observed in the pattern, indicating that the carbonaceous material was in an amorphous state. As shown in the elemental mapping images (Figure 1g), a certain amount of C was uniformly distributed on the whole MoO2 nanosphere, expressing the successful fabrication of the MoO2/C-OV catalyst. The elemental mapping images of MoO2 and MoO2/C were shown in Figures S3 and S4. The results revealed that a unique interface microstructure was formed after glucose hydrothermal treatment.  The crystal structures of the prepared catalysts were characterized using XRD patterns. As shown in Figure 2a (220), (013) and (131) crystal planes. There was no change in the crystal structure of MoO 2 after glucose hydrothermal and high-temperature calcination treatment, which shows that the oxygen vacancies did not disrupt the crystal structure. All the prepared products owned strong diffraction peaks, and no other peaks existed in the XRD pattern, which confirmed the formation of samples with high crystallinity and phase purity. The MoO 2 /C-OV and MoO 2 /C composites were synthesized successfully via a hydrothermal and calcination method using glucose as the carbon source and MoO 2 (acac) 2 as the Mo source.
Molecules 2023, 28, x FOR PEER REVIEW 5 of 13 vacancies for MoO2/C-OV was about 55.3%, which was higher than that in MoO2 (18.7%) and MoO2/C (15.9%). It was further testified that many more oxygen vacancies were formed on the surface of MoO2/C-OV. As shown in Figure 2e, the deconvoluted C 1s spectrum showed three characteristic peaks, located at 284.5, 285.8 and 288.4 eV, belong to C−C, C−O−Mo, C=O−O [38,44,45]. It is inferred that a carbonaceous complex layer may be constructed around MoO2 nanospheres and combined with the MoO2 surface via C−O−Mo bonds after glucose hydrothermal treatment. The deconvoluted C 1s spectra of MoO2 and MoO2/C-OV were shown in Figures S4 and S5. The above results revealed that the introduction of carbonaceous precursors in situ could trigger the reconstruction and fabricate surface oxygen vacancies on the surface of MoO2 via a hydrothermal process. The presence of oxygen vacancies was further confirmed via room-temperature EPR. As shown in Figure 2f, no signal was observed in the EPR spectrum of the pristine MoO2 nanospheres. In comparison, the obvious oxygen vacancies signal with a g value of 2.003 was exhibited for the MoO2/C-OV, which could be identified as unpaired electrons trapped in the oxygen vacancy [34,46]. More oxygen vacancies were generated due to the introduction of carbon atoms. The results were consistent with the XPS spectra. This demonstrated that the MoO2/C-OV possesses more oxygen vacancies in the surface after the glucose hydrothermal process. The oxygen vacancy and chemical states of the surface atoms on MoO 2 , MoO 2 /C-OV and MoO 2 /C nanospheres were further characterized via X-ray photoelectron spectroscopy (XPS). As shown in Figure 2b, the full-survey XPS spectra of the three samples displayed the copresence of Mo, O and C elements. The contents of the Mo, O, and C elements according to the XPS spectra were shown in Table S1. The contents ratios of C and Mo elements for MoO 2 /C-OV and MoO 2 /C were higher than those of MoO 2 nanospheres, indicating that carbonaceous matter was introduced into the surface of MoO 2 nanospheres after the treatment via a glucose hydrothermal process. As shown in Figure 2c and Mo 4+ 3d 3/2 in MoO 2 . In addition, the peaks at 231,1.0 and 234.4 eV are ascribed to the high-valence Mo 6+ 3d 5/2 and Mo 6+ 3d 3/2 . After glucose hydrothermal treatment, the Mo 3d core levels presented significant shifting to lower binding energy, indicating that more electrons were transferred from carbonaceous material and a more reduced state of Mo compared to the pristine MoO 2 nanosphere. Compared to the MoO 2 and MoO 2 /C-OV samples, the chemical state of MoO 2 /C composite differed significantly. Two peaks of the Mo 6+ species were observed at 235.7 and 232.7 eV; this was ascribed to the Mo 3d 3/2 and Mo 3d 5/2 states. Two peaks of the Mo 4+ species were observed at 231.0 and 229.4 eV, which was ascribed to the Mo 3d 3/2 and Mo 3d 5/2 states [40][41][42]. This revealed the remarkable surface oxidation of MoO 2 after high-temperature calcination treatment. The stability of oxygen vacancies is due to the adsorption of oxygen species, which is a typical feature of defect-rich oxides. As shown in Figure 2d, the deconvoluted O 1s spectrum of MoO 2 /C-OV exhibited four peaks at 529.8, 531.6, 532.8, 533.4, which could be assigned to the lattice oxygen, the surface-adsorbed oxygen or oxygen vacancies, the adsorption of H 2 O on the surface and the oxygen of C=O-C. At the same time, the characteristic peak area of the lattice oxygen is relatively small, further demonstrating that the lattice oxygen atoms have reduced and oxygen vacancies have been introduced in the MoO 2 /C-OV catalyst [43]. As listed in Table S2, the atomic percentages of each oxygen species (at%) are calculated from the XPS data of the MoO 2 , MoO 2 /C-OV and MoO 2 /C nanospheres, respectively. The results showed that the proportion of the surface-adsorbed oxygen or oxygen vacancies on the surface of MoO 2 /C-OV was higher than that of MoO 2 and MoO 2 /C. In addition, the ratio of the lattice oxygen on the surface of MoO 2 /C-OV was about 5.4%, which was less than that of MoO 2 and MoO 2 /C (63.4% and 52.7%). In addition, the O 1s spectrum showed the existence of C=O-C on the surface of MoO 2 /C-OV and MoO 2 /C, which further demonstrated the introduction of carbonaceous matter onto the surface of MoO 2 /C-OV and MoO 2 /C. The proportion of the oxygen species adsorbed at oxygen vacancies for MoO 2 /C-OV was about 55.3%, which was higher than that in MoO 2 (18.7%) and MoO 2 /C (15.9%). It was further testified that many more oxygen vacancies were formed on the surface of MoO 2 /C-OV. As shown in Figure 2e, the deconvoluted C 1s spectrum showed three characteristic peaks, located at 284. The presence of oxygen vacancies was further confirmed via room-temperature EPR. As shown in Figure 2f, no signal was observed in the EPR spectrum of the pristine MoO 2 nanospheres. In comparison, the obvious oxygen vacancies signal with a g value of 2.003 was exhibited for the MoO 2 /C-OV, which could be identified as unpaired electrons trapped in the oxygen vacancy [34,46]. More oxygen vacancies were generated due to the introduction of carbon atoms. The results were consistent with the XPS spectra. This demonstrated that the MoO 2 /C-OV possesses more oxygen vacancies in the surface after the glucose hydrothermal process.
The photo-response of catalysts is an important indicator of photocatalytic efficiency. The optical properties of the samples with different structures were investigated via UV-vis diffuse reflectance spectroscopy (DRS) measurement. As shown in Figure 3a, the visiblelight absorption intensity of MoO 2 , especially in the near-infrared region, was significantly enhanced after the glucose hydrothermal process. MoO 2 /C-OV exhibited a powerful absorption intensity from the UV to the NIR region. The band gap energy (E g ) of the photocatalysts plays an important role in evaluating the physical and optical properties, which could be calculated according to the following Formula [40].
(1) The optical properties of the samples with different structures were investigated via UVvis diffuse reflectance spectroscopy (DRS) measurement. As shown in Figure 3a, the visible-light absorption intensity of MoO2, especially in the near-infrared region, was significantly enhanced after the glucose hydrothermal process. MoO2/C-OV exhibited a powerful absorption intensity from the UV to the NIR region. The band gap energy (Eg) of the photocatalysts plays an important role in evaluating the physical and optical properties, which could be calculated according to the following Formula [40].
where α, hν, A and Eg represent the absorption coefficient, Planck constant, light frequency, proportionality and band gap energy, respectively. As shown in Figure 3b  Transient photocurrent measurements were further carried out to verify the charge generation and separation efficiency, as shown in Figure 3d. Compared with the MoO2 and MoO2/C, MoO2/C-OV exhibited a higher photocurrent response, which suggested that the MoO2/C-OV would produce more photo-electrons and holes than MoO2 and MoO2/C in the same timeframe. Moreover, the separation and transfer of these generated photo-electrons and holes occurred more efficiently in the MoO2/C-OV than MoO2 and MoO2/C. The results demonstrated that the surface OVs would hasten the photo-electrons and holes separation to further induce the photo-electron transfer from the inside to the surface of the MoO2/C-OV. Therefore, it revealed that surface OVs promote the generation Transient photocurrent measurements were further carried out to verify the charge generation and separation efficiency, as shown in Figure 3d. Compared with the MoO 2 and MoO 2 /C, MoO 2 /C-OV exhibited a higher photocurrent response, which suggested that the MoO 2 /C-OV would produce more photo-electrons and holes than MoO 2 and MoO 2 /C in the same timeframe. Moreover, the separation and transfer of these generated photo-electrons and holes occurred more efficiently in the MoO 2 /C-OV than MoO 2 and MoO 2 /C. The results demonstrated that the surface OVs would hasten the photo-electrons and holes separation to further induce the photo-electron transfer from the inside to the surface of the MoO 2 /C-OV. Therefore, it revealed that surface OVs promote the generation of the photo-electrons and accelerate the charges transfer on the interface to improve the transformation of the benzylamine molecules on the MoO 2 /C-OV.

Photocatalytic Activity for Oxidation of Amines to Imines
The photocatalytic performance of the MoO 2 samples was evaluated via selective oxidative coupling of amines to imines under white LED light irradiation at air pressure. The oxidative coupling of BA to N-benzylidenebenzylamine was chosen as the type of reaction to explore the activation of molecular oxygen. The control experiment was run in the absence of catalyst, and there was no product under visible light irradiation for 3.5 h. As shown in Figure 4a, the conversion was only 25% with >99% selectivity when air was replaced with N 2 , demonstrating the indispensable role of molecular oxygen for the photocatalytic oxidative coupling of BA. To confirm the effect of illuminant on the oxidation reaction, the dark reaction was carried out with other things being equal. The control reaction was testified to be very slow, and the conversion was situated at 9% in the presence of MoO 2 /C-OV under the dark condition. In the presence of visible light under the same conditions, the MoO 2 /C-OV exhibited much higher conversion (96%) and selectivity (>99%) for the formation of N-benzylidenebenzylamine. The concentration of oxygen vacancy sites of MoO 2 /C sample was less than that of MoO 2 /C-OV. As the comparison, the conversion of MoO 2 /C was only 50% with >99% selectivity of N-benzylidenebenzylamine under identical conditions. The results confirmed that the presence of surface OVs was the critical requirement for the activation of molecular oxygen. The conversion of MoO 2 (7%) was much lower than that of MoO 2 /C-OV, which indicated that the synergistic effect of carbonaceous material and oxygen vacancies played an important role in the photocatalytic reaction. lytic reaction.
The main photo-generated active species during the photocatalytic BA oxidation were verified through radical trapping experiments with a series of scavengers over the MoO2/C-OV sample. As shown in Figure 4b, the conversion of BA dramatically decreased with the addition of KI (h + scavenger), AgNO3 (e − scavenger), p-BQ (•O2 − scavenger) or NaN3 ( 1 O2 scavenger), while IPA (•OH scavenger) exhibited a negligible quenching effect [47][48][49]. The BA conversion significantly reduced with the addition of trapping agents: 34.5% (with KI), 32.4% (NaN3), 18.7% (p-BQ) and 51.9% (AgNO3). The results indicated that h + , e − , •O2 − and 1 O2 played a key role in the reaction process for the MoO2/C-OV sample, where •O2 − and 1 O2 are the main active species in the photocatalytic reaction. The results proved that both photo-generated holes and electrons were the initial driving forces for the photocatalytic coupling reaction. The conversions of MoO2/C with the addition of trapping agents were lower than that of MoO2/C-OV. As shown in Figure 4c, the MoO2/C sample has the same changing tendency as the results of the MoO2/C-OV sample, which implied that a similar photocatalytic process was involved. The results further revealed that the oxygen vacancy played an important role in visible light photocatalytic selective oxidation of BA. The formation of H2O2 was detected via the iodometry method in the photocatalytic oxidation of BA. The result illustrated that the H2O2, produced by the reaction of •O2 − and protons, was an intermediate in the photocatalytic aerobic oxidative coupling reaction of BA. Hence, the radical experiments indicated that h + , e − , •O2 − and 1 O2 species played a key role in the photocatalytic oxidation of BA. In order to prove the versatility of the as-prepared photocatalyst, the selective oxidation of amine derivatives was considered. The MoO 2 /C-OV sample showed a high efficiency for the visible light-driven photocatalytic oxidation of amines at 1 atm air pressure. As listed in Table 1, the arylamines with electron-withdrawing groups would be more challenging to transform into the corresponding imines than that of arylamines with electron-donating groups. It is reported that the electron-induced effect of an electronwithdrawing group could enhance the oxidative coupling of the C-N bond via the α-carbon activation. The electron inductive effect of the -Cl group is relatively weaker than that of the -F group. For Entry 2 and 3, the reactant substrate with the -F group showed high conversion under the same conditions. Entries 6 and 7, the reactant substrates with an electron-donating group, exhibited a high yield to produce their corresponding imines with high selectivity under the same reaction conditions, owing to the conjugative effect of the substituent groups. In addition, for Entries 3-5 and 7-9, the para and meta substitution of the reactant substrates was beneficial, offering greater activation of α-carbon than that of ortho substitution due to the steric effect. Table 1. Aerobic photocatalytic oxidation of benzylamine substitutes over MoO 2 /C-OV. tion of amine derivatives was considered. The MoO2/C-OV sample showed a high efficiency for the visible light-driven photocatalytic oxidation of amines at 1 atm air pressure. As listed in Table 1, the arylamines with electron-withdrawing groups would be more challenging to transform into the corresponding imines than that of arylamines with electron-donating groups. It is reported that the electron-induced effect of an electron-withdrawing group could enhance the oxidative coupling of the C-N bond via the α-carbon activation. The electron inductive effect of the -Cl group is relatively weaker than that of the -F group. For Entry 2 and 3, the reactant substrate with the -F group showed high conversion under the same conditions. Entries 6 and 7, the reactant substrates with an electron-donating group, exhibited a high yield to produce their corresponding imines with high selectivity under the same reaction conditions, owing to the conjugative effect of the substituent groups. In addition, for Entries 3-5 and 7-9, the para and meta substitution of the reactant substrates was beneficial, offering greater activation of α-carbon than that of ortho substitution due to the steric effect. challenging to transform into the corresponding imines than that of arylamines with electron-donating groups. It is reported that the electron-induced effect of an electron-withdrawing group could enhance the oxidative coupling of the C-N bond via the α-carbon activation. The electron inductive effect of the -Cl group is relatively weaker than that of the -F group. For Entry 2 and 3, the reactant substrate with the -F group showed high conversion under the same conditions. Entries 6 and 7, the reactant substrates with an electron-donating group, exhibited a high yield to produce their corresponding imines with high selectivity under the same reaction conditions, owing to the conjugative effect of the substituent groups. In addition, for Entries 3-5 and 7-9, the para and meta substitution of the reactant substrates was beneficial, offering greater activation of α-carbon than that of ortho substitution due to the steric effect. challenging to transform into the corresponding imines than that of arylamines with electron-donating groups. It is reported that the electron-induced effect of an electron-withdrawing group could enhance the oxidative coupling of the C-N bond via the α-carbon activation. The electron inductive effect of the -Cl group is relatively weaker than that of the -F group. For Entry 2 and 3, the reactant substrate with the -F group showed high conversion under the same conditions. Entries 6 and 7, the reactant substrates with an electron-donating group, exhibited a high yield to produce their corresponding imines with high selectivity under the same reaction conditions, owing to the conjugative effect of the substituent groups. In addition, for Entries 3-5 and 7-9, the para and meta substitution of the reactant substrates was beneficial, offering greater activation of α-carbon than that of ortho substitution due to the steric effect. drawing group could enhance the oxidative coupling of the C-N bond via the α-carbon activation. The electron inductive effect of the -Cl group is relatively weaker than that of the -F group. For Entry 2 and 3, the reactant substrate with the -F group showed high conversion under the same conditions. Entries 6 and 7, the reactant substrates with an electron-donating group, exhibited a high yield to produce their corresponding imines with high selectivity under the same reaction conditions, owing to the conjugative effect of the substituent groups. In addition, for Entries 3-5 and 7-9, the para and meta substitution of the reactant substrates was beneficial, offering greater activation of α-carbon than that of ortho substitution due to the steric effect. drawing group could enhance the oxidative coupling of the C-N bond via the α-carbon activation. The electron inductive effect of the -Cl group is relatively weaker than that of the -F group. For Entry 2 and 3, the reactant substrate with the -F group showed high conversion under the same conditions. Entries 6 and 7, the reactant substrates with an electron-donating group, exhibited a high yield to produce their corresponding imines with high selectivity under the same reaction conditions, owing to the conjugative effect of the substituent groups. In addition, for Entries 3-5 and 7-9, the para and meta substitution of the reactant substrates was beneficial, offering greater activation of α-carbon than that of ortho substitution due to the steric effect. the -F group. For Entry 2 and 3, the reactant substrate with the -F group showed high conversion under the same conditions. Entries 6 and 7, the reactant substrates with an electron-donating group, exhibited a high yield to produce their corresponding imines with high selectivity under the same reaction conditions, owing to the conjugative effect of the substituent groups. In addition, for Entries 3-5 and 7-9, the para and meta substitution of the reactant substrates was beneficial, offering greater activation of α-carbon than that of ortho substitution due to the steric effect. the -F group. For Entry 2 and 3, the reactant substrate with the -F group showed high conversion under the same conditions. Entries 6 and 7, the reactant substrates with an electron-donating group, exhibited a high yield to produce their corresponding imines with high selectivity under the same reaction conditions, owing to the conjugative effect of the substituent groups. In addition, for Entries 3-5 and 7-9, the para and meta substitution of the reactant substrates was beneficial, offering greater activation of α-carbon than that of ortho substitution due to the steric effect. electron-donating group, exhibited a high yield to produce their corresponding imines with high selectivity under the same reaction conditions, owing to the conjugative effect of the substituent groups. In addition, for Entries 3-5 and 7-9, the para and meta substitution of the reactant substrates was beneficial, offering greater activation of α-carbon than that of ortho substitution due to the steric effect. electron-donating group, exhibited a high yield to produce their corresponding imines with high selectivity under the same reaction conditions, owing to the conjugative effect of the substituent groups. In addition, for Entries 3-5 and 7-9, the para and meta substitution of the reactant substrates was beneficial, offering greater activation of α-carbon than that of ortho substitution due to the steric effect. of the substituent groups. In addition, for Entries 3-5 and 7-9, the para and meta substitution of the reactant substrates was beneficial, offering greater activation of α-carbon than that of ortho substitution due to the steric effect. of the substituent groups. In addition, for Entries 3-5 and 7-9, the para and meta substitution of the reactant substrates was beneficial, offering greater activation of α-carbon than that of ortho substitution due to the steric effect.

Proposed Mechanism of Photocatalytic BA Coupling to N-benzylidenebenzylamine
A reaction mechanism is proposed for the selective transformation of BA to N-benzylidenebenzylamine over MoO2/C-OV, as shown in Figure 5

Proposed Mechanism of Photocatalytic BA Coupling to N-benzylidenebenzylamine
A reaction mechanism is proposed for the selective transformation of BA to N-benzylidenebenzylamine over MoO2/C-OV, as shown in Figure 5

Proposed Mechanism of Photocatalytic BA Coupling to N-benzylidenebenzylamine
A reaction mechanism is proposed for the selective transformation of BA to N-benzylidenebenzylamine over MoO2/C-OV, as shown in Figure 5, based on the above results and discussions. The surface OVs sites of MoO2/C-OV would collect abundant O2 molecules from the air. Under the visible light irradiation, electrons and holes were generated in the MoO2/C-OV photocatalyst. The photogenerated electrons rapidly transferred to the positively charged surface OVs of the MoO2/C-OV, which played an important role in the further reduction of adsorbed O2 molecules to produce active •O2 − anion radicals. The

Proposed Mechanism of Photocatalytic BA Coupling to N-benzylidenebenzylamine
A reaction mechanism is proposed for the selective transformation of BA to N-benzylidenebenzylamine over MoO2/C-OV, as shown in Figure 5, based on the above results and discussions. The surface OVs sites of MoO2/C-OV would collect abundant O2 molecules from the air. Under the visible light irradiation, electrons and holes were generated in the MoO2/C-OV photocatalyst. The photogenerated electrons rapidly transferred to the positively charged surface OVs of the MoO2/C-OV, which played an important role in the further reduction of adsorbed O2 molecules to produce active •O2 − anion radicals. The holes generated transferred from bulk to surface and oxidized most of the•O2 − into 1 O2. Moreover, the adsorbed BA underwent oxidation via photogenerated holes to form amine radical cations, and subsequent activation by 1 O2 and •O2 − radicals promoted formation

Proposed Mechanism of Photocatalytic BA Coupling to N-benzylidenebenzylamine
A reaction mechanism is proposed for the selective transformation of BA to Nbenzylidenebenzylamine over MoO 2 /C-OV, as shown in Figure 5 Figure S9, the H 2 O 2 intermediate in the process of reaction was detected via the iodometry method.

Proposed Mechanism of Photocatalytic BA Coupling to N-benzylidenebenzylamine
A reaction mechanism is proposed for the selective transformation of BA to N-benzylidenebenzylamine over MoO2/C-OV, as shown in Figure 5, based on the above results and discussions. The surface OVs sites of MoO2/C-OV would collect abundant O2 molecules from the air. Under the visible light irradiation, electrons and holes were generated in the MoO2/C-OV photocatalyst. The photogenerated electrons rapidly transferred to the positively charged surface OVs of the MoO2/C-OV, which played an important role in the further reduction of adsorbed O2 molecules to produce active •O2 − anion radicals. The holes generated transferred from bulk to surface and oxidized most of the•O2 − into 1 O2. Moreover, the adsorbed BA underwent oxidation via photogenerated holes to form amine radical cations, and subsequent activation by 1 O2 and •O2 − radicals promoted formation of imine intermediate (RCH=NH) and H2O2. As shown in Figure S9, the H2O2 intermediate in the process of reaction was detected via the iodometry method.

Preparation of MoO2 Nanospheres
MoO2 nanospheres were synthesized via a facile one-pot solvothermal process. Typically, 600 mg of MoO2(acac)2 was dissolved into 12 mL of water, 12 mL of ethanol and 36 mL of isopropyl alcohol via magnetic stirring. After 2 h stirring, the mixed solution was transferred into a 100 mL Teflon-lined stainless-teel autoclave at 200 °C for 10 h. After cooling to room temperature, the obtained black and blue precipitates were washed with

Preparation of MoO 2 Nanospheres
MoO 2 nanospheres were synthesized via a facile one-pot solvothermal process. Typically, 600 mg of MoO 2 (acac) 2 was dissolved into 12 mL of water, 12 mL of ethanol and 36 mL of isopropyl alcohol via magnetic stirring. After 2 h stirring, the mixed solution was transferred into a 100 mL Teflon-lined stainless-teel autoclave at 200 • C for 10 h. After cooling to room temperature, the obtained black and blue precipitates were washed with absolute ethanol and distilled water, then dried at 60 • C overnight. Finally, the MoO 2 nanospheres were obtained.

Preparation of Defective MoO 2 /C Nanospheres
In a typical synthesis process, 80 mg of MoO 2 nanospheres and 1200 mg of glucose were added into 40 mL of ultrapure water, then stirred to obtain a homogeneous solution. The mixture was poured into a 100 mL Teflon-lined stainless-teel autoclave at 180 • C for 3 h. The precipitates were gathered via centrifugation and washed with absolute ethanol and water. Finally, the samples were marked MoO 2 /C-OV. The as-prepared MoO 2 /C-OV sample was calcined in an air flow at 300 • C to obtain MoO 2 /C.

Electrochemistry Measurements
Electrochemical measurements were characterized on a CH660D electrochemical workstation in a three-electrode model, using the sample films as the working electrode, saturated calomel electrode (SCE) as the reference electrode and Pt wire as the counter electrode. The working electrode was prepared as follows: 5 mg of catalyst was dispersed in 0.5 mL of ethanol and 25 µL of Nafion solution, and ultrasonically treated until a homogeneous mixture was obtained. The Mott-Schottky, electrochemical impedance spectroscopy (EIS) and photocurrent response measurements were performed at the electrochemical workstation with the working electrodes immersed in 0.1 M Na 2 SO 4 aqueous solution.

Photocatalytic Oxidative Coupling of Benzylamine
The photocatalytic reactions were conducted under an air atmosphere at 45 • C for 3.5 h. The reaction system was composed of a catalyst (6 mg) and benzylamine (0.077 mmol) in acetonitrile (3 mL) using white LED light as the light source. The optical power density at the liquid surface was about 300 mW cm −2 . The reaction products were analyzed via gas chromatography (GC, Agilent 7890B, Agilent, Santa Clara, CA, USA). The reactive oxygen species (ROS) quenching experiments Radical trapping experiments were carried out under the above conditions, except for the addition of scavengers. Isopropanol (IPA), KI, p-benzoquinone (pBQ) and NaN 3 were used as the scavengers for hydroxyl radicals (•OH), holes (h + ), superoxide radicals (•O 2 − ) and singlet oxygen ( 1 O 2 ), respectively.

Detection of H 2 O 2
The iodometry method was selected to detect the H 2 O 2 produced in the photocatalytic reaction process. The reaction equation is as follows: H 2 O 2 + 3I − + 2H + → I 3− + 2H 2 O The H 2 O 2 reacted with I − under acidic conditions to form I 3− , which shows strong absorption at about 350 nm. The solution A: KI (249 mg) was dissolved in ultrapure water (15 mL). Solution B: the potassium biphthalate (1.225 g) was added to 15 mL of ultrapure water. The detection process is as follows: after mixing 2 mL of solution A and 2 mL of solution B, 0.1 mL reaction solution and 0.9 mL ultrapure water were added and sonicated for 1 min. The absorption spectrum was measured using a UV-vis spectrophotometer.

Conclusions
In summary, defective MoO 2 /C nanospheres were successfully prepared via the one-pot glucose hydrothermal method. The MoO 2 nanospheres combined closely with carbonaceous materials, forming a clear disordered region and a unique interface microstructure. In situ introduction of carbonaceous materials triggered a reconstruction of the MoO 2 surface, which introduced abundant surface OVs on the MoO 2 /C composites. The surface OVs and carbonaceous materials played a synergistic role in the activation of molecular oxygen into 1 O 2 and •O 2 − species. Furthermore, the surface OVs structure led to the effective separation of the photogenerated carriers. The MoO 2 /C-OV sample displayed superior photocatalytic oxidation of benzylamine under visible light irradiation at 1 atm air pressure, and the performance was 10 times that of pristine MoO 2 nanospheres with high selectivity (>99%). This work can provide an avenue for the design of photocatalysts used for photocatalytic transformation of amines to imines by introducing oxygen vacancy of Mo-based metal oxides.
Supplementary Materials: The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/molecules28124739/s1, characterizations and measurement details, Figures S1-S9, Tables S1 and S2. Figure S1. SEM images of (a) MoO 2 samples, (b) MoO 2 /C samples; Figure S2. TEM images of (a) MoO 2 and (b) MoO 2 /C; Figure S3. The corresponding TEM image and EDX elemental mapping of Mo and O for the MoO 2 ; Figure S4. The corresponding TEM image and EDX elemental mapping of C, Mo and O for the MoO 2 /C; Figure S5. XPS spectra of C 1s in the as-prepared MoO 2 ; Figure S6. XPS spectra of C 1s in the as-prepared MoO 2 /C; Figure S7. Mott-Schottky plots of MoO 2 ; Figure S8. EIS Nyquist plots of MoO 2 , MoO 2 /C-OV, MoO 2 /C; Figure S9. UV-vis absorption spectra of the solution after photocatalytic reaction for detection H 2 O 2 ; Table S1. The contents of the Mo, O, and C elements according to the XPS spectra; Table S2. The spectrum parameters of the fitted O 1s peaks.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.