Deqin
Liang
a,
Yu
Wang
a,
Xiaojing
Yin
a,
Ziyan
Liu
a,
Jizhou
Du
b,
Junfeng
Qian
a,
Mingyang
He
*a and
Weiyou
Zhou
*a
aJiangsu Key Laboratory of Advanced Catalytic Materials and Technology, Changzhou University, Changzhou, 213164, China. E-mail: hemy_cczu@126.com; zhouwy426@126.com
bIntellectual Property Management Office, PetroChina Co., Ltd, Beijing, 100101, China
First published on 29th November 2025
The efficient and selective oxidation of toluene derivatives to produce the corresponding aldehydes over a heterogeneous catalyst remains a challenge in industrial production. In this study, an effective catalytic system based on a Mg-doped Mn-based layered double oxide (MnMgAl-LDO) catalyst and NHPI (N-hydroxyphthalimide) has been developed. This system enables the liquid-phase oxidation of toluene and its derivatives into aromatic aldehydes, with hexafluoroisopropanol (HFIP) serving as the solvent. The incorporation of Mg increases the amount of oxygen vacancies and the surface basicity of the catalyst, thereby enhancing the reaction efficiency and aldehyde selectivity. In the model reaction over MgMnAl-LDO, the conversion of toluene reached 77%, with an 86% selectivity of benzaldehyde under the selected conditions. A possible reaction pathway for the selective oxidation of toluene to benzaldehyde was proposed based on the obtained results and a series of control experiments, and the active intermediates PINO (phthalimide-N-oxyl) were generated from NHPI simultaneously through the HAT (hydrogen atom transfer) and PCET (proton-coupled electron transfer) processes in the present reaction system. This catalytic system also exhibits good substrate applicability, as well as excellent stability and reusability, providing an efficient strategy for the green synthesis of aromatic aldehydes.
The key issue in the liquid-phase oxidation synthesis of benzaldehyde compounds from toluene derivatives is the inertness of the C–H bond, which makes its activation challenging. In addition, the fact that the product aldehyde is more susceptible to oxidation to carboxylic acids than the raw materials leads to low selectivity. There is an urgent necessity to develop an efficient and highly selective catalytic reaction system to achieve the green synthesis of aromatic aldehydes.6,7 NHPI (N-hydroxyphthalimide) can efficiently catalyze the oxidation of C–H bonds on the side chains of aromatic hydrocarbons via the generation of active PINO (phthalimide-N-oxyl) radicals.8,9 Early studies (such as that by the Ishii team) utilized NHPI and Co(OAc)2 to oxidize toluene at room temperature using oxygen as the oxidant. The main product obtained was benzoic acid with a yield of 81%, while the yield of aldehyde was merely 3%.10 Although numerous catalytic materials have been developed for NHPI/O2 systems, the selectivity of aldehydes remains at a relatively low level. In addition to Co-based catalysts, there are also several catalytic systems based on Cu, and Fe is developed for the molecular oxygen oxidation system of toluene in the presence of NHPI.11–13 However, none of these reaction systems can achieve satisfactory selectivity for benzaldehyde.
The Pappo group was the first to employ hexafluoroisopropanol (HFIP) solvent to inhibit the over-oxidation of aldehydes by constructing hydrogen bond complexes. Under the catalysis of cobalt acetate, the conversion rate of toluene to benzaldehyde reached 91% within 4 hours, with a selectivity of 90%, far exceeding the efficiency of the reported catalytic systems. Nevertheless, this system has limitations such as the difficulty in recovering homogeneous catalysts and the susceptibility of products to secondary oxidation during distillation, which restrict its industrial application.14,15 The development of heterogeneous catalytic materials to achieve highly selective molecular oxygen oxidation of toluene holds significant practical importance.
Among the various heterogeneous catalytic materials, layered bimetallic hydroxides (LDHs) have emerged as crucial catalytic materials or precursors due to their adjustable surface alkalinity and metal composition. And, highly dispersed composite metal oxide catalysts can be prepared just through calcination using LDHs as precursors. It has been known that the surface basicity of catalysts, to a certain extent, can stabilize the product aldehyde and mitigate its over-oxidation reaction.16 There have also been some research reports on the utilization of hydrotalcite-based catalytic materials in the molecular oxygen oxidation reaction of toluene. Our group prepared a Co–Al LDH catalyst for the liquid phase oxidation of toluene using acetonitrile as the solvent, but the conversion rate was only 8.2% and the selectivity was 77.6%.17 The Wu research team developed Ti3+ modified Mg–Al LDHs. Under solvent-free conditions for the aerobic oxidation of toluene, the selectivity of benzaldehyde reached 97.5%; nevertheless, the conversion rate was only 8.7%. Moreover, it required stringent reaction conditions at 150 °C and 1 MPa, posing challenges for industrialization.18 Zhang et al. achieved a 58% selectivity for benzaldehyde by using a Co–Mn–Al flake metal oxide as the catalyst and HFIP as the solvent. Evidently, the current catalytic systems for toluene oxidation generally face problems such as low catalytic efficiency, harsh reaction conditions, and low selectivity of aldehydes.19
During our continuous research on catalytic oxidation catalysts based on hydrotalcite materials, it was discovered that, in the NHPI/O2 oxidation system, the Mg-doped manganese-based layered double oxide (MnMgAl-LDO) catalyst prepared by calcining the MnMgAl-LDH precursor can selectively catalyze the molecular oxygen oxidation of toluene and its derivatives using HFIP as the solvent. This thesis systematically examined the performance, structure–activity relationship, and substrate tolerance of the catalyst. Additionally, based on a series of control experiments and in situ EPR analysis, possible reaction mechanisms were proposed. This research is conducive to promoting the green synthesis of aromatic aldehydes.
A range of characterization techniques were employed to perform a systematic analysis of the materials. The crystal structure was characterized using an X-ray diffractometer (XRD, Rigaku D/max 2500 PC), and the microscopic morphology was observed via a field emission scanning electron microscope (SEM, JSM-6010plus/LV). The composition of metal elements was determined using an inductively coupled plasma spectrometer (ICP, Varian Vista-AX). Fourier transform infrared (FT-IR) spectra were recorded in the wavenumber range of 4000–400 cm−1, and KBr was used as a reference. The specific surface area and pore structure parameters were measured using an ASAP 2010C physical adsorption instrument, and the specific surface area and pore size distribution were calculated, respectively, based on the BET theory and the BJH model. The chemical state of the surface Mn element was analyzed by X-ray photoelectron spectroscopy (XPS, Thermo Scientific K-alpha, Al Kα excitation source hv = 1486.6 eV), and the C 1s peak at 284.80 eV was used as calibration. Electron paramagnetic resonance (EPR) measurements were carried out using a Bruker E500–10/12 at low temperature (T = −140 °C) with a field modulation of 100 kHz. The test conditions for all spectra were as follows: frequency, 9.840264 GHz; attenuation, 20.0 dB; and modulation amplitude, 1.000 G.
Temperature programmed desorption of CO2 (CO2-TPD) was conducted using an AutoChem II 2920 TPR/TPD instrument. Approximately 50 mg samples were pretreated at 150 °C in He airflow (50 mL min−1) for 2 hours to clean the surface. After cooling to adsorption temperature (50 °C), 10% CO2/He (50 mL min−1) was passed through for 1 hour. Subsequently, the gas flow was switched to pure He (50 mL min−1), and the sample was purged at 50 °C for 1 hour to remove physically adsorbed CO2. Finally, the temperature was increased at a linear heating rate of 10 °C min−1 from 50 °C to a final temperature of 700 °C. The desorbed CO2 was monitored online using a thermal conductivity detector (TCD).
![]() | ||
| Fig. 1 SEM images (a), XRD patterns (b), and EDS spectra (c) of the MnMgAl-LDH, MnMgAl-LDO, and Mn2Al-LDO samples. | ||
Fig. 2a depicts the nitrogen adsorption and desorption curves of three samples. According to the international IUPAC classification standard, the three catalysts present type II isotherms. Among them, the adsorption of MnMgAl-LDH and Mn2Al-LDO under low-pressure conditions showed a gradual upward trend, which indicates that it belongs to the multi-layer adsorption of adsorbates on the material surface. MnMgAl-LDO falls into the second case because the pores in the sample are too large to be filled when approaching saturation. The adsorption and desorption curves of Mn2Al-LDO materials nearly overlap, from which it can be roughly inferred that the pores of this catalyst are cylindrical or V-shaped closed-end pores. The desorption of MnMgAl-LDH shows an H4-type hysteresis isotherm. The reason for this is that the evaporation and desorption of the pore size or pore body involve cavitation, and it may also be caused by the growth of bubbles.
![]() | ||
| Fig. 2 N2 adsorption/desorption isotherms (a) and CO2-TPD (b) of the MnMgAl-LDH, MnMgAl-LDO and Mn2Al-LDO. | ||
The structural properties of the three catalysts are presented in Table 1. The ICP results indicate that, after calcination at 500 °C, the metal ratio remained unchanged. This can be ascribed to the fact that the calcination process mainly removes interlayer water and anions (such as CO32−), while metal ions remain stable at high temperatures and do not undergo lose. From the BET data, it is evident that the specific surface area of the calcined hydrotalcite increases significantly, which should be related to the evaporation of the moisture in the LDH interlayer during the sintering process.25 The elevated surface area should benefit to expose the active site. Compared with MnMgAl-LDH and Mn2Al-LDO, MnMgAl-LDO has a larger pore volume and pore diameter, which is consistent with the isotherm analysis results in the N2 adsorption–desorption curve. CO2-TPD was employed to analyze the number of basic sites on the surface of the oxide catalyst (Fig. 2b), because the pretreatment process would damage the structure of MnMgAl-LDH. The curve of Mn2Al-LDO can be fitted by two peaks at 576 and 615 °C, whereas that of MnMgAl-LDO shows a peak at 591 °C. All of these peaks correspond to strongly basic sites. MnMgAl-LDO has more basic sites than Mn2Al-LDO, and the number of strongly basic sites in MnMgAl-LDO is approximately 1.64 times that of Mn2Al-LDO. This also indicates that the addition of Mg is conducive to enhancing the alkalinity of the catalyst. In composite metal oxides, the strong basic sites were probably related to the M–O2−(H) structure, and the doping Mg2+, which has high positive charge density, is expected to increase the electron density of the oxygen atom in M–O2−(H) and enhance the basicity of MnMgAl-LDO.
| Samples | Content (wt%) | Mn : Mg : Al |
BET surface area (m2 g−1) | Pore volume (cm3 g−1) | Pore size (nm) | ||
|---|---|---|---|---|---|---|---|
| Mn | Mg | Al | |||||
| MnMgAl-LDH | 18.60 | 9.60 | 8.60 | 1.06 : 1.25 : 1 |
54.52 | 0.12 | 8.89 |
| MnMgAl-LDO | 25.80 | 13.80 | 12.40 | 1.02 : 1.25 : 1 |
88.45 | 0.55 | 24.89 |
| Mn2Al-LDO | 56.63 | — | 9.67 | 1 : 0 : 0.35 |
187.03 | 0.34 | 7.35 |
FT-IR spectra of MnMgAl-LDO and Mn2Al-LDO (Fig. S1) demonstrate that a strong absorption band observed at 3422 cm−1 corresponds to the stretching vibration mode of O–H; the bending vibration of O–H can also be observed around 1365 cm−1. Lattice vibration peaks of M–O (M = Mn, Al, or Mg) are observed at 620–641 cm−1.22–24
The metal composition and oxidation state on the sample surface were analyzed by XPS. Fig. 3a shows the spin–orbit splitting of the Mn 2p XPS spectra of MnMgAl-LDH, MnMgAl-LDO, and Mn2Al-LDO. Previous studies have indicated that the peaks at 640.7–641.0 eV are related to Mn2+, those at 642.1–642.2 eV to Mn3+, and those at 644.7–644.8 eV to Mn4+.26 As shown in Fig. 3a, the content of metal ions in the high-valent state of hydrotalcite increases significantly after calcination. Compared with Mn2Al-LDO, MnMgAl-LDO contains more Mn3+ and Mn4+ ions. The O 1s spectra of MnMgAl-LDO and Mn2Al-LDO are compared in Fig. 3b. The peaks at 529.3–529.4 eV, 530.7–530.9 eV, and 532.0–532.1 eV can be ascribed to surface lattice oxygen (Olatt), surface adsorbed oxygen (Oads), and surface hydroxyl oxygen (OOH),27 respectively. Notably, the Oads of the MnMgAl-LDO catalyst (33.4%) is significantly higher than that of Mn2Al-LDO (22.2%), indicating that the addition of Mg is beneficial to increase the oxygen vacancy of MnMgAl-LDO. Previous studies have reported that an alkaline environment is conducive to elevating the selectivity of benzaldehyde.16,28,29 The peaks of the lattice oxygen atoms in MnMgAl-LDO shift towards lower binding energies to some extent. In combination with Fig. 3a, this may be attributed to the transfer of electrons from Mn3+ to the adjacent lattice oxygen.
To further elucidate the catalytic performance of MnMgAl-LDO, MnMgAl-LDH and Mn2Al-LDO were also examined under the standard conditions for comparison. As can be seen from Table 2 and Fig. 4B, the catalytic activity of MnMgAl-LDO is significantly higher than that of MnMgAl-LDH and Mn2Al-LDO. Moreover, the order of the performance is as follows: MnMgAl-LDO > Mn2Al-LDO > MnMgAl-LDH. Independent kinetic analyses were conducted on the reaction temperatures of the three catalysts while excluding diffusion effects, and the Arrhenius diagram is shown in Fig. 5d. The apparent activation energies (Ea) of MnMgAl-LDO, Mn2Al-LDO and MnMgAl-LDH were 32.7 kJ mol−1, 35.2 kJ mol−1 and 35.4 kJ mol−1, respectively, which is consistent with their catalytic performance. Combining the XPS analysis of the three catalysts, the higher catalytic performance of the two oxide samples might be related to the increased amount of high-valent Mn species, which always exhibits high oxidation capacity. Considering the reaction path discussed below, the high-valent Mn species can promote the activation of NHPI to PINO through a PCET (proton-coupled electron transfer) process during the reaction, thereby accelerating the aerobic oxidation of toluene. Furthermore, the higher performance of MnMgAl-LDO than Mn2Al-LDO is likely ascribed to the increased amount of oxygen vacancies, which can activate oxygen molecules. The TOF (turnover frequency) of the two catalysts, which can be used to further elucidate the activity of the active sites, was calculated. The TOF values were 17.8 and 10.5 h−1 for MnMgAl-LDO and Mn2Al-LDO, respectively, implying that the intrinsic activity of Mn species in MnMgAl-LDO is higher than that in Mn2Al-LDO. Regarding the selectivity of the target product aldehyde, MnMgAl-LDO gave obviously a higher value than Mn2Al-LDO, which is probably related to their surface basicity. It has been recognized that surface basicity is conducive to the stabilization of aldehyde and the prevention of its over-oxidation.30 CO2-TPD analysis has revealed more basic sites on MnMgAl-LDO, which can account for the improved catalytic performance.
| Entry | Catalyst | Additive | Conv./% | Sel. (1b)/% | TOF (h−1) |
|---|---|---|---|---|---|
| a Reaction conditions: toluene 1 mmol, catalyst 75 mg, NHPI 15% mmol, HFIP 2 mL, 40 °C, O2 5 mL min−1, and 12 h. b 1 Equiv. | |||||
| 1 | MnMgAl-LDH | — | 52.3 | 69.5 | — |
| 2 | MnMgAl-LDO | — | 77.0 | 86.0 | 17.8 |
| 3 | Mn2Al-LDO | — | 65.2 | 72.8 | 10.5 |
| 4 | Mg2Al-LDH | — | 8.2 | 36.3 | — |
| 5 | MnMgAl-LDO | Ar | 2.0 | — | — |
| 6 | MnMgAl-LDO | Air | 71.2 | 81.3 | — |
| 7 | — | — | 7.6 | — | — |
| 8 | MnMgAl-LDO | Without NHPI | 1.1 | — | — |
| 9 | MnMgAl-LDO | BHTb | 5.3 | — | — |
| 10 | MnMgAl-LDO | TEMPOb | — | — | — |
| 11 | MnMgAl-LDO | 1,4-Benzoquinoneb | 6.3 | 20.7 | — |
| 12 | MnMgAl-LDO | 9,10-Diphenylanthraceneb | 5.5 | 53.4 | — |
| 13 | MnMgAl-LDO | 1,4-Dinitrobenzeneb | Trace | — | — |
To gain a more in-depth understanding of the catalytic process, some control experiments were conducted. Controlled experiments under an argon atmosphere or in the absence of NHPI demonstrated the crucial roles of molecular oxygen and NHPI (Table 2, entries 5 and 8). The control experiment under an air atmosphere proved that the oxygen concentration in the air met the requirement for the oxidation of toluene to benzaldehyde (entry 6). The conversion rate is only 7.6% in the absence of a catalyst, which indicates the key role of MnMgAl-LDO in the system (entry 7). Mg2Al-LDH only achieved an 8.2% conversion of toluene, suggesting that Mn species should be responsible for the catalytic process (entry 4). It is well known that reduced Mn2+ can react with molecular oxygen to form Mn3+-superoxide, thereby promoting the HAT (hydrogen atom transfer) process. To verify this process, p-benzoquinone as an inhibitor was introduced into the reaction.
In this case, the conversion rate decreased significantly (entry 11), which indicates that superoxide is involved in the reaction. When 1,4-dinitrobenzene was added to the reaction system as an electron transfer remover, the conversion rate also dropped substantially (entry 13), indicating that electron transfer occurred during the reaction process. Combining this with the valence state diagrams of Mn 2p in Fig. 3a indicates that HAT and PCET act in concert to promote the oxidation process of toluene in the present reaction system.
Fig. 6a presents the EPR spectrum obtained from the DMPO-trapping reaction solution, in which superoxide radicals (AN = 14.67 G and AH = 11.08 G) can be identified.28 Within this time frame, no clear signals related to the formation of PINO from NHPI were observed, indicating that the PINO radical reacts extremely rapidly. This means that its accumulation in the solution can be ignored and is beyond the detection range of EPR.31
By incorporating 9,10-dinitroanthracene, as a singlet oxygen inhibitor, into the reaction system, the conversion rate and yield were significantly decreased. This indicates the presence of 1O2 during the reaction process (entry 12), which may be related to oxygen vacancies.32 During the reaction process, the presence of 1O2 species was confirmed by TEMP capture, and the triple characteristic peak signals of 1O2 species could be observed in the EPR spectrum (Fig. 6b), with equal intensities.33
In addition, BHT (3,5-di-tert-butyl-4-hydroxytoluene) and TEMPO (2,2,6,6-tetramethylpiperidine oxide) were separately introduced into the reaction as free radical scavengers to clarify the possible free radical intermediates. The transformation of toluene (Table 2, entries 9 and 10) was significantly inhibited, which indicates that the intermediate products of the reaction contain free radical structures.
Based on the above results and analysis, a possible catalytic pathway for the oxidation of toluene by MnMgAl-LDO with the O2/NHPI system was proposed (Scheme 1). Mn2+ in MnMgAl-LDO can react with oxygen to form superoxide radicals (Mn3+OO˙), which then extracts a hydrogen atom from the NHPI molecule through the HAT process to generate the active PINO intermediate. Alternatively, PINO can also be formed from NHPI through a PCET process. Mn4+ and Mn3+ facilitate electron transfer, while the surface alkalinity promotes proton transfer, which is conducive to the PCET process. Then, the PINO radical extracts the α-H in toluene to produce a benzyl radical (I). Next, intermediate I reacts with singlet oxygen (1O2, which may be produced through oxygen activation by oxygen vacancies) to form the peroxyl radical II.33–36 Finally, the peroxyl radical extracts a hydrogen atom from the substrate, generating the unstable intermediate III, which decomposes to produce benzyl alcohol and benzaldehyde as products (Scheme 1).37 Part of benzyl alcohol can be further oxidized into an aldehyde in the present reaction system, leading to excellent selectivity for the aldehyde. When benzyl alcohol was tested as the substrate under the standard conditions, aldehyde was smoothly produced in high selectivity (Scheme S1), further confirming the above speculation.
Source data are provided with this paper.
| This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2026 |