High-Efficiency Preparation of 2,5-Diformylfuran with a Keto-ABNO Catalyst Under Mild Conditions

In this paper, we report a new catalytic system for realizing the rapid and efficient oxidation of 5-hydroxymethylfurfural (HMF). First, we used 9-azabicyclo [3.3.1]nonan-3-one-N-oxyl (keto-ABNO) as a catalyst for the aerobic oxidation of 5-hydroxymethylfurfural (HMF) to 2,5-diformylfuran (DFF) in acetic acid. Then, we systematically studied the important reaction parameters, including the solvent, co-catalyst, and temperature. The results demonstrate that the acidic solvent used is crucial for the efficient oxidation of HMF to DFF. Under optimal conditions, we achieved a 93.4% yield of DFF within half an hour at room temperature. We also proposed the possible mechanism for this system.

Fructose is a representative sugar molecule from biomass, which can be converted into DFF via a one-pot reaction. As early as 2003, Gary et al. reported a catalytic system of Amberlyst-15 and polyaniline-VO (acac) for the one-pot conversion of fructose into DFF with a yield of 42.1% [15]. Recently, Yang et al. [16] developed bifunctional catalyst Ce-Mo composite oxides to catalyze the transformation of fructose into DFF, with a yield of 80%. However, the transformation of fructose to DFF involves a complex reaction system comprising both dehydration and oxidation reactions and the yield of DFF is always poor. The aerobic oxidation of HMF to DFF is more simple and efficient. In general, there are two strategies for the aerobic oxidation of HMF to DFF: heterogeneous catalysis and homogeneous catalysis. For heterogeneous catalysis, VO [17,18], Co [19,20], Mn [21], and Ru [22,23] show promising catalytic activity. However, heterogeneous catalysis has some obvious shortcomings, i.e., high temperature and a long reaction period, which are not conducive to industrial-scale production.

3
Homogeneous catalysis, in contrast, reacts under milder conditions. A well-known example is the TEMPO catalytic system, which is well reported in the studies. For example, Fang et al. [24] developed a TEMPO/Fe catalytic system for the selective oxidation of HMF to DFF in air at room temperature. Mittal et al. [25] developed a TEMPO/BAIB/ CH 3 COOH catalytic system for the selective oxidation of HMF to DFF with a yield of 66% at 30 °C within 45 min.
Recently, Lauber and Stahl [26] described the catalytic oxidation of a wide range of secondary alcohols using readily available ABNO-based catalysts in combination with HNO 3 and NaNO 2 . The authors reported that both 9-azabicyclo[3.3.1]nonane-N-oxyl (ABNO) and 9-azabicyclo[3.3.1] nonan-3-one-N-oxyl (keto-ABNO) exhibited remarkable catalytic efficiencies. Considering that HMF contains an alcohol hydroxyl, which can be considered to be a primary alcohol, we speculated that keto-ABNO might be able to catalyze the oxidation of HMF to DFF with a good yield.
In this study, we investigated the catalytic ability of keto-ABNO for the oxidation reaction of HMF to DFF and systematically studied the effects of solvents, co-catalysts, amounts of catalyst and co-catalyst, atmosphere, temperature, and reaction time. We found that the acidic solvent used is crucial for the efficient oxidation of HMF. With keto-ABNO serving as the catalyst and NaNO 2 as the co-catalyst, we achieved a DFF yield of 93.4% within half an hour at room temperature.

Materials
We purchased Zn(NO 3

General Procedure for the Aerobic Oxidation of HMF
We performed the oxidation of HMF in a glass tube equipped with a total reflux condenser and magnetic stirrer and introduced molecular oxygen through an airway. First, we charged the glass tube with HMF (63 mg, 0.5 mmol), and then we added the keto-ABNO (2 mg), NaNO 2 (3.4 mg, 0.025 mmol), and 4 mL of acetic acid. We heated the reactor at room temperature with an O 2 flow rate of 10 mL/ min. Lastly, we analyzed the furan compounds by gas chromatography.

Analytic Methods
We analyzed the furan compounds with an Agilent gas chromatograph equipped with an HP-5 column (30 m × 320 μm × 0.25 μm). We calculated the HMF and DFF contents in the samples using an external standard calibration curve that had been constructed based on the pure compounds. The retention times of HMF and DFF were 2.7 and 5.4 min, respectively.

Effect of Solvents on the Oxidation of HMF to DFF
We used keto-ABNO and NaNO 2 to catalyze the oxidation of HMF to DFF (Table 1). In this system, keto-ABNO serves as the catalyst and NaNO 2 as the co-catalyst. As the solvent polarity, dielectric constant, steric hindrance, and acid-base properties play decisive roles in reactions [27], we first investigated the influence of various solvents on the catalytic reaction. Table 1 shows the results, in which we can see that reaction in acetic acid proved to be the best, with a full conversion of HMF and 81.1% selectivity of DFF through the catalysis of keto-ABNO and NaNO 2 for 2 h at 40 °C. When using other solvents, i.e., acetonitrile, dimethylsulfoxide, dimethylformamide, tetrahydrofuran, and ethyl acetate, we detected almost no DFF product. In studies, acetonitrile has been frequently used when ABNO is the catalyst. However, in our catalytic system, the yield of DFF in acetonitrile was very low. This distinct difference might be attributed to the use of a different co-catalyst from that had been used in previous work. In the keto-ABNO and NaNO 2 catalytic system, the NO x sources serve as an intermediate of the electron transfer, which is crucial for the efficient catalysis of HMF.
In the study of Wang et al. [28], NO x was generated by Fe 3+ and NO 3 − . However, in our catalytic system, NO x sources are generated from NO 2 − and H + . In the absence of H + , no NO x would form and thus the yield of DFF would be negligible.

DFF selectivity (% ) = DFF yield HMF conversion × 100%
This result indicates that the acidic solvent used is of utmost importance in achieving high efficiency and high selectivity in the oxidation of HMF to DFF.

Effects of Co-catalysts on the Oxidation of HMF to DFF
Based on the above results, a good yield of DFF can be achieved via the co-catalysis of NaNO 2 . To gain further insights into this catalytic system, we investigated the influence of other commercial sodium and nitrate salts as cocatalysts on the oxidation of HMF to DFF, the results of which are summarized in Table 2. As we can see in the table, none of the sodium salts worked well (

Effect of Atmosphere on the Oxidation of HMF to DFF
Compared to other oxidants, oxygen is inexpensive, green, and can be reused in liquid-phase reactions, so it has advantages in industrial fields. As such, in our experiment, oxygen was used as the oxidant. To investigate the effect of oxygen on the catalytic reactions, experiments were conducted under different atmospheres at 40 °C, the results of which are summarized in Fig. 1. In oxygen atmosphere, we achieved the highest DFF yield of 81.1% with a full conversion of HMF.
In an open-air atmosphere, the DFF yield dropped to 78.1%. Exceeding our expectations, when the reaction was conducted in an inert nitrogen atmosphere, we realized an HMF conversion of 28.6% with a DFF selectivity of 41.0%. In this situation, NaNO 2 might act as the oxidant. In weak acid solvents, NaNO 2 decomposes into NO 2 and NO (Scheme 2). In an oxygen atmosphere, the reduced nitric oxide is reoxidized to nitrogen dioxide, leading to the continuous conversion of HMF. However, in a nitrogen atmosphere, the intermediate of NO 2 serves as the oxidant for HMF and the conversion of HMF mainly depends on the amount of sodium nitrite used.

Effect of the Amounts of Keto-ABNO and NaNO 2 on the Oxidation of HMF to DFF
It is known that the dosage of the catalyst will have a significant influence on a reaction. An insufficient amount of catalyst will lead to low catalytic efficiency, which is mainly manifested as a low conversion rate. However, an excessive amount of a catalyst leads to an excessive reaction and increases production costs. We investigated the influence of the amounts of catalyst and co-catalyst on the oxidation of HMF to DFF, and the results are shown in Table 3. We found both keto-ABNO and NaNO 2 to play crucial roles in the conversion of HMF to DFF. An experiment was also conducted without using keto-ABNO and NaNO 2 , for which no DFF could be detected (Table 3, entry 1). In the absence of keto-ABNO, only a HMF conversion of 69.4% was achieved, and no DFF could be detected (Table 3, entry 2). In the presence of keto-ABNO without NaNO 2 , a full conversion of HMF was achieved, but the selectivity to DFF was relatively low (Table 3, entry 6). These results indicate that the highly selective oxidation of HMF to DFF requires a synergic catalytic effect between keto-ABNO and the co-catalyst NaNO 2 . Moreover, the insufficient use of keto-ABNO results in the partial conversion of HMF to DFF. Increasing the amounts of keto-ABNO and NaNO 2 does not lead to a continuous  increase in the DFF yield. Considering the formation of other by-products (e.g., FFCA, FDCA), the optimal dosages of keto-ABNO and NaNO 2 are 2 mg and 3.4 mg, respectively. We note that the dosages of catalysts used in our catalytic system are much lower than those reported in Refs. [2,22], while the amount of catalyst used represents a major cost in the industrial production of DFF. The high catalytic efficiency of keto-ABNO and NaNO 2 makes them good candidates for the industrial-scale production of DFF.

Effects of Temperature and Reaction Time on the Oxidation of HMF to DFF
To gain a better understanding of the catalytic kinetics of the reaction, we investigated the oxidation of HMF at temperatures of 40, 50, and 60 °C. The results indicate that the DFF yield decreases slightly with increasing temperature (Fig. 2a). According to the data, it is the most suitable condition at room temperature, yielding a full conversion of HMF and a DFF yield of 88.1%. Moreover, we also investigated the effect of reaction time on the oxidation of HMF to DFF, the results of which are shown in Fig. 2b. Compared to results reported in the literature, our catalytic system is much faster and more efficient, with the HMF fully converted within 30 min. In the first 10 min, the conversion of HMF was very fast, which might be due to the high concentration of HMF. The highest yield of DFF was 93.4% at 30 min. Longer reaction times resulted in a lower selectivity of DFF, mainly because of the overreaction of HMF, which leads to the further transformation of DFF into FFCA or FDCA.

Possible Mechanism of Keto-ABNO Catalytic Oxidation of HMF to DFF
Based on our experiments and previous reports [29,30], we proposed a possible mechanism for the keto-ABNO catalytic oxidation of HMF to DFF, as shown in Scheme 2. In the catalytic reaction, we suggest that the oxidation of HMF to DFF is achieved through the coupling of multiple redox reactions. Cycle (I) involves an oxoammonium cation, which is initially oxidized from keto-ABNO, which has the ability to oxidize HMF to DFF. In this process, keto-ABNOH is continuously oxidized to oxoammonium cations with NO 2 , and the NO 2 is considered to play the role of oxidant, having been initially produced through the deposition of NaNO 2 (Scheme 2). Cycle (II) shows the transformation between NO 2 and NO. The NO, which is reduced by keto-ABNOH, is oxidized into NO 2 in the presence of oxygen. By the coupling of Cycles (I) and (II), a novel, coherent, and efficient aerobic oxidation system is achieved, leading to the highly selective production of DFF with a yield of 93.4% under optimal conditions.

Conclusions
In summary, we have successfully developed a highly efficient keto-ABNO catalytic system for the selective oxidation of HMF to DFF. This oxidation can occur at room temperature with 2.8% of keto-ABNO and a small amount of sodium nitrite. Under optimal reaction conditions, we achieved a full conversion of HMF and a 93.4% selectivity of DFF within 30 min. With the advantages of a high reaction rate, low catalyst dosage, and mild reaction conditions, the keto-ABNO catalytic system shows great promise for the industrial-scale production of DFF in achieving the cost-effective conversion of biomass into high-value chemicals. Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creat iveco mmons .org/licen ses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.