Introduction

Aromatic compound nitration is a significant industrial process for the production of chemical intermediates [1, 2]. In particular, nitration of 1-nitronaphthalene (1-NN) or naphthalene yields dinitronaphthalenes (1,5-DNN, 1,8-DNN, 1,3-DNN, and 1,4-DNN), which can be used to make polyurethanes, plastics, and dyes [3]. 1,5-DNN, for example, is an important raw material for the preparation of intermediate polyurethanes and 1,5-naphthalene diisocyanate [4]. Because 1,5-DNN is the most commercially valuable and in-demand dinitronaphthalene product, considerable studies are devoted to boosting its output. The typical approach for preparing 1,5-DNN involves nitration of 1-NN or naphthalene, which is usually catalyzed by significant volumes of concentrated H2SO4 and HNO3. However, this method has unavoidable drawbacks, such as low selectivity (30–35%), extensive energy requirement, high corrosion, and negative environmental effect [5, 6]. As a result, it is necessary to develop a benign and economically feasible nitration approach to increase the selectivity of 1,5-DNN and improve the notorious production environment.

Solid acid catalysts for the nitration of 1-NN or naphthalene have recently attracted growing attention as a way to make dinitronaphthalene more cleanly and with higher selectivity. Various solid acid catalysts have been reported for this purpose, such as acid-treated clay [7], ionic liquid-based catalysts [8,9,10], modified zeolites [11, 12], and sulfated metal oxides (SO42−/MxOy) [13,14,15]. Nevertheless, the process still produces large quantities of undesirable by-products, consumes too much concentrated HNO3 as a nitrating agent, or obtains low target selectivity. This has motivated the study of eco-friendly nitrating agents to obtain a high selectivity of 1,5-DNN. Mascal et al. found that when [Ag-K-Na]NO3 is used as a nitrating agent, the selectivity of 1,5-DNN for the nitration of naphthalene under the catalysis of K3Fe(CN)6 was only 2.1% [16]. According to Wang et al., the selectivity of 1,5-DNN was 56% for the nitration of 1-NN with the NO2-O2 system using a zeolite catalyst [17]. Our previous work demonstrated that when 1-NN was nitrated with NO2 employing a HY catalyst, the selectivity of 1,5-DNN was 35% [12]. Consequently, it is of great significance to develop high-efficiency catalysts for clean and green nitration processes under facile conditions.

Among solid acid catalysts, the SO42−/MxOy catalyst has been widely evaluated for many acid catalytic reactions, such as isomerization, alkylation, acetalization, and nitration reactions, due to its easy synthesis, good thermal stability, and super acidity [18,19,20,21]. Jiao et al. presented a solvent-free liquid phase toluene nitration process, utilizing NO2 as the nitrifying agent and SO42−/WO3 as the catalyst, which exhibited good catalytic activity [13]. Yan et al. observed that catalyzing the nitration of 1-NN with NO2 and SO42−/ZrO2 resulted in 93.8% 1-NN conversion and 52.8% 1,5-DNN selectivity [14]. Based on the literature, transition metal-modified SO42−/MxOy catalysts can significantly enhance catalytic performance. CuO/SO42−/ZrO2, for example, was shown by Occelli et al. to have better catalytic performance and stability than SO42−/ZrO2 [22]. After the inclusion of Pd and Co metals, Córdoba et al. discovered that the lifetime of SO42−/ZrO2 in NOx reduction by methane was extended [23]. In the nitration of aromatic compounds, Wang et al. revealed that TiO2-Fe3O4-SO42−/ZrO2 presented higher catalytic activity than SO42−/ZrO2, possibly due to the synergistic interaction between metal oxides [24]. Using Fe- and Mn-modified SO42−/ZrO2 catalyzed nitration of 1-NN with NO2 driven by O2-Ac2O, our previous work indicated that 92.6% conversion to 1-NN and 57.2% selectivity to 1,5-DNN were achieved [25].

On the other hand, in comparison with SO42−, S2O82− is easier to coordinate with metal oxides, which can provide more superacid sites and sulfur active components to enhance catalytic performance [26,27,28]. Moreover, compared with the H2SO4, (NH4)2S2O8 is a safe acidifying agent during the preparation of the catalyst (impregnation process). The previous work considered that S2O82−/MxOy presented stronger acidity, higher catalytic effect, and stability than SO42−/MxOy [29,30,31]. It is therefore of great interest to adopt the metal oxide promoted S2O82−/ZrO2 catalyst for the liquid phase nitration of 1-NN with NO2 assisted by O2-Ac2O to make this process more environmentally and economically feasible. To the authors’ knowledge, the use of Fe-modified S2O82−/ZrO2 in the nitration of the 1-NN process has rarely been reported heretofore.

Experimental section

Materials

The materials employed in this study are presented in supplementary information (SI).

Catalyst synthesis

The coprecipitation–impregnation method was used to prepare the Fe-modified S2O82−/ZrO2 catalyst (S2O82−/Fe-ZrO2), and the regenerated S2O82−/Fe-ZrO2 catalyst means that the catalyst was washed, dried, and then calcined at 600 °C after being recycled for five times. Details are displayed in SI.

Catalyst characterization

The catalysts were characterized by X-ray diffraction (XRD), Fourier transform infrared spectroscopy(FT-IR), N2 adsorption/desorption experiment, inductively coupled plasma-optical emission spectrometer (ICP-OES), scanning electron microscope (SEM), high-resolution transmission electron microscope (HRTEM), X-ray photoelectron spectroscopy (XPS), temperature programmed desorption of ammonia (NH3-TPD), and pyridine adsorption infrared spectroscopy (Py-FT-IR). Detailed information about these methods is available in SI.

Typical experimental process

The catalytic nitration of 1-NN with the NO2 test was performed based on the method reported in our previous work [14, 25], and detailed information about this test is listed in SI.

Results and discussion

Catalytic performance of different catalysts

Several typical solid superacid catalysts were evaluated under the same reaction conditions to find an efficient catalyst for the nitration of 1-NN with NO2 aided by O2-Ac2O, and the results are tabulated in Table 1. For the non-catalytic reaction, the nitration of 1-NN with NO2 assisted by acetic anhydride (Ac2O) could result in the formation of 1,5-DNN, and 16.1% of 1-NN conversion with 39.9% selectivity to 1,5-DNN was obtained. When O2 was introduced into the nitration reaction, the conversion of 1-NN increased to 83.9%, and the selectivity of 1,5-DNN was 40.7%. Stabilization of NO2+ and formation of an active acetyl nitrate (AcONO2) nitrating reagent are thought to be responsible for the beneficial effect of Ac2O on enhancing reaction activity.

Table 1 Comparison of the catalytic nitrification performance of different systems
Full size table

Regarding the catalytic reaction process, all tested solid superacid catalysts S2O82−/MxOy improved catalytic activity in the metal oxides and non-catalytic process, with S2O82−/ZrO2 exhibiting the highest catalytic performance with a selectivity of 47.6% to 1,5-DNN. On the other hand, the use of metal oxide-modified S2O82−/ZrO2 catalysts greatly improved reaction activity, and the Fe-modified solid superacid catalyst, S2O82−/Fe-ZrO2 improved catalytic performance even more with 53.4% selectivity of 1,5-DNN. S2O82−/Fe-ZrO2 was found to have the highest catalytic activity, resulting in 96.8% conversion of 1-NN and 62.6% selectivity of 1,5-DNN under optimal reaction conditions (Sect. 3.2).

Optimization of catalyst preparation conditions

In order to further improve its catalytic performance, the preparation conditions of the S2O82−/Fe-ZrO2 catalyst that included the mass fraction of Fe2O3, the concentration of the acidifying reagent (NH4)2S2O8, and the calcination temperature were optimized. The results for the nitration of 1-NN are depicted in Fig. S1. Figure S1(a) presents the catalytic performance of S2O82−/Fe-ZrO2 with various contents of Fe2O3. When the iron content is 2 wt%, the selectivity of 1,5-DNN is highest at 53.4%, and the conversion of 1-NN is 94.9%.

The influence of varied (NH4)2S2O8 impregnation concentrations on 1-NN nitration is displayed in Fig. S1(b). The selectivity of 1,5-DNN decreases significantly as the impregnation concentration lowers, which is due to a reduction in acid content, a reduction in the quantity of acid supported by the catalyst, and a decrease in catalytic efficiency. However, the catalytic performance is lower than that of 0.5 M when the concentration is 1 M. This might be related to excessive (NH4)2S2O8 that cannot be properly supported on the carrier surface, the reduction of the number of metal oxide particles on the catalyst surface, and the formation of persulfate salts that cover the surface activity sites of the catalyst [32]. As a result, a concentration of (NH4)2S2O8 of 0.5 M is preferred.

As shown in Fig. S1(c), the catalyst exhibits the best catalytic activity when the calcination temperature is 600 °C, with a selectivity of 1,5-DNN of 53.4%. With a change in the calcination temperature, the catalytic performance has an obvious trend of change. This is because the lower calcination temperature cannot allow ZrO2 to form a relatively complete tetragonal phase, which is not conducive to the formation of stable acid sites. When the calcination temperature is too high, sulfur combined with the catalyst will be lost and the tetragonal phase will be weakened, resulting in the loss of the surface-active centers [33]. Accordingly, the optimized calcination temperature is 600 °C.

Optimization of reaction conditions and stability of the catalyst

The S2O82−/Fe-ZrO2 catalyst obtained under optimized preparation conditions, that is, the iron content of 2 wt%, an acid treatment concentration of 0.5 mol/L, and a calcination temperature of 600 °C, was used for further studies. Reaction factors, including the molar ratio of 1-NN to NO2, reaction temperature, oxygen pressure, reaction time, and catalyst dosage were optimized for the nitration of 1-NN with the prepared S2O82−/Fe-ZrO2 catalyst, and the results are displayed in Fig. 1a–e. The impact of the molar ratio of 1-NN to NO2 is given in Fig. 1a. Note that the conversion of 1-NN is increased with increased content of NO2, and when 1-NN/NO2 was 1:3, the selectivity of 1,5-DNN reaches the highest value. Similarly, as shown in Fig. 1b–d, the reaction temperature, oxygen pressure, and reaction time present similar change trends and results. The effect of the amount of catalyst on the nitration reaction results is depicted in Fig. 1e. It can be seen that a suitable amount of S2O82−/Fe-ZrO2 catalyst is favorable for the formation of 1,5-DNN. Consequently, the optimization factors are obtained: the molar ratio of 1-NN to NO2: 1:3, reaction temperature: 40 °C, partial pressure of O2: 0.7 MPa, reaction time: 5 h, and catalyst dosage: 0.5 g. Under these reaction conditions, 96.8% 1-NN conversion with 62.6% 1,5-DNN selectivity is achieved over the S2O82−/Fe-ZrO2 catalyst.

Fig. 1
figure 1

Effects of parameters on the nitration reaction. Reaction condition: a 35 °C, 0.5 MPa O2, 4 h, catalyst dosage 0.5 g, solvent 10.0 g; b 1-NN to NO2 is 1:3, 0.5 MPa O2, 4 h, catalyst dosage 0.5 g, solvent 10.0 g; c 1-NN to NO2 is 1:3, 40 °C, 4 h, catalyst dosage 0.5 g, solvent 10.0 g; d 1-NN to NO2 is 1:3, 40 °C, 0.7 MPa O2, catalyst dosage 0.5 g, solvent 10.0 g; e 1-NN to NO2 is 1:3, 40 °C, 0.7 MPa O2, 5 h, solvent 10.0 g. f Cyclic tests of the S2O82−/Fe-ZrO2 catalyst. Reaction conditions: 1-NN to NO2 is 1:3, 40 °C, 0.7 MPa O2, 5 h, catalyst dosage 0.5 g, solvent 10.0 g

Full size image

The stability of S2O82−/Fe-ZrO2 catalyst for the catalytic 1-NN nitration reaction was evaluated under the optimized reaction conditions, and the results are displayed in Fig. 1f. For the cyclic tests, after the catalytic reaction was finished, the spent catalyst was filtered, washed, dried, and then used for the next cycle test. After the catalyst was recycled five times, the conversion of 1-NN did not change and the selectivity of 1,5-DNN decreased only by 3.8% (62.6% to 60.2%). Furthermore, as demonstrated in Sect. 3.4, the characterization results (XRD, FT-IR, N2 adsorption–desorption experiment, and ICP-OES) revealed that the structure, composition, and acidity of the regenerated catalyst hardly changed after five cycles. These results demonstrated that the S2O82−/Fe-ZrO2 catalyst had outstanding recyclability for the nitration of 1-NN.

Meanwhile, comparison of catalytic activities in the nitration of 1-NN with various reported catalysts and nitration systems is summarized in Table 2. Compared with treated clay, metal salt, and ionic liquids-based catalysts, S2O82−/Fe-ZrO2 has excellent recyclability. The NO2-O2-Ac2O nitration system is more environmentally friendly than the concentrated HNO3 as nitrating agent. Furthermore, the catalytic activity of the S2O82−/Fe-ZrO2 catalyst is better than most of the reported solid acid catalysts for the nitration of 1-NN (Table 2). Accordingly, the combination of the S2O82−/Fe-ZrO2 catalyst and the NO2-O2-Ac2O nitration system provides a practical pathway for the low-cost, efficient and green synthesis of 1,5-DNN with great potential for industrial application.

Table 2 Comparison of the results of 1-NN or naphthalene nitration in different nitration systems
Full size table

Results of catalyst characterization

Figure 2a displays the XRD patterns of fresh, used, and regenerated S2O82−/Fe-ZrO2 catalysts. Note that all patterns show the same diffraction peaks at 2 thetas of 30.1, 35.2, 50.2, and 60.2°, which can be assigned to the pure tetragonal phase ZrO2 (JCPDS No. 88–1007) [39]. The non-appearance of diffraction reflection of iron oxides suggests that these oxides are well dispersed on the surface or may migrate evenly into the interior of the ZrO2 tetragonal phase lattice [40]. Moreover, after the catalyst was used five times, the crystal phase structure of the catalyst remains unchanged.

Fig. 2
figure 2

Catalyst characterization results. a XRD patterns. b FT-IR spectra. c N2 adsorption–desorption isotherms. d Pore diameter distributions

Full size image

FT-IR spectra of catalysts are presented in Fig. 2b. Bands around 545 cm−1 are attributed to the characteristic peak of crystalline ZrO2 [41]. Peaks at around 3440 and 1632 cm−1 refer to the H–O-H bending frequency of adsorbed water [42]. Two peaks at about 1270 and 1074 cm−1 can be assigned to chelating bidentate persulfate groups coordinated with the surface of the metal oxide supports [43]. These results indicated that the persulfate group has been successfully bonded to the supports. Furthermore, the regenerated catalyst presents peaks similar to those of the fresh S2O82−/Fe-ZrO2 catalyst, demonstrating that the framework of the catalyst remained intact after five catalytic reaction cycles.

Figure 2c, d presents the isotherms of the N2 adsorption–desorption and pore diameter distribution curves of the fresh, used, and regenerated S2O82−/Fe-ZrO2 catalysts. Table 3 tablets the textural properties of the samples. It can be seen that the three samples present the same type IV isotherms with a H2 type hysteresis loop (P/P0 = 0.4–1.0), belonging to the feature of the mesoporous structure [44, 45]. Compared with the pure ZrO2, the surface area of S2O82−/Fe-ZrO2 increases, while the pore diameter decreases after acidification with ammonium persulfate. S2O82−/Fe-ZrO2 exhibits better textural properties factors than the ZrO2 catalyst, including surface area, pore volume, and pore diameter. In addition, the structural properties of S2O82−/Fe-ZrO2 decreased slightly after five recycles, while the structural factors of spent S2O82−/Fe-ZrO2 can be regenerated by calcination.

Table 3 Textural properties of catalysts
Full size table

The morphology of the fresh S2O82−/Fe-ZrO2 catalyst was researched by SEM (Fig. 3a) and HRTEM (Fig. 3b, c). As shown in Fig. 3a, the S2O82−/Fe-ZrO2 surface is rough and uneven particles are deposited into large congeries. The TEM image (Fig. 3b) reveals the morphology of the crystal, which is agglomerated by uneven nano-crystalline particles (Fig. 3b, c). Moreover, the HRTEM image (Fig. 3c) of the S2O82−/Fe-ZrO2 catalyst shows two distinct lattices spacing about 0.295 nm and 0.252 nm, which can be attributed to the (101) and (110) lattice planes of the tetragonal phase ZrO2, respectively [46]. Unfortunately, the lattice spacing of iron oxide was not found in the HRTEM image, possibly due to the low additional amount.

Fig. 3
figure 3

The morphology of the S2O82−/Fe-ZrO2 catalyst. a SEM image. b and c TEM and HRTEM images

Full size image

XPS can be utilized to analyze the chemical state of elemental species in the S2O82−/Fe-ZrO2 catalyst [47]. The XPS measurement results are shown in Fig. 4. As shown in Fig. 4a, the banding energy gap between Zr 3d 5/2 and Zr 3d 3/2 is higher than 2.3 eV, suggesting that the Zr species are primary in the Zr4+ form in the catalyst [48]. The peaks at 184.2 (3d 3/2) and 181.8 eV (3d 5/2) correspond to the Zr-O-Zr bond of ZrO2 [49]. Bands at 184.8 (3d 3/2) and 182.3 eV (3d 5/2) are due to the Zr-O-S bond [50]. For the O 1 s spectrum of a catalyst (Fig. 4b), the peaks at about 529.7, 531.3, and 532.1 eV represent the binding energies of the O-Zr bond (ZrO2), the O-S bond (persulfate group), and the O–H bond (surface hydroxyl group), respectively [51, 52]. The O 1 s peak for Fe–O has not appeared from the XPS result, which may be attributed to the content of iron is low. For the S 2p spectrum, as shown in Fig. 4c, the peak at around 167–172 eV is attributed to the S6+ element, which can be separated into two bands at 169.9 and 168.7 eV [25]. The peak at 168.7 eV is due to the deprotonated persulfate groups, while the band at 169.9 eV is due to the protonated groups [53]. In the Fe 2p spectrum (Fig. 4d), the peak at 718.3 eV is the satellite peak of Fe. Peaks at around 711.3 and 724.8 eV are related to the binding energies of Fe 2p 3/2 and Fe 2p 1/2, respectively. The binding energy gap between Fe 2p 3/2 and Fe 2p 1/2 is greater than 13 eV, demonstrating that Fe species exist mainly in the form of Fe3+ in the S2O82−/Fe-ZrO2 catalyst [42, 54]. The XPS results demonstrate that the persulfate groups have bonded to Fe-modified ZrO2, and Fe3+ exists in the S2O82−/Fe-ZrO2 catalyst.

Fig. 4
figure 4

XPS spectra of the S2O82−/Fe-ZrO2 catalyst. a–d Zr 3d, O 1S, S 2p, and Fe 2p spectra, respectively

Full size image

The acid types, Brønsted acid sites (B acid site) and Lewis acid sites (L acid site) of the ZrO2 and S2O82−/Fe-ZrO2 catalysts were measured using pyridine-FT-IR, and the results are given in Fig. 5a. Generally, for the pyridine-FT-IR spectra, the band at around 1449 and 1608 cm−1 can be attributed to the L acid sites, and the band at about 1544 cm−1 can be ascribed to B acid sites [48]. The peak at 1486 cm−1 is related to both B acid sites and L acid sites [45]. Compared to pure ZrO2, which only possesses a small amount of L acid sites, the S2O82−/Fe-ZrO2 catalyst has both B and L acid sites, while the L acid site is dominant. The abundant L acid sites of the S2O82−/Fe-ZrO2 catalyst may be one of the reasons for its excellent catalytic activity for the nitration of 1-NN.

Fig. 5
figure 5

Results of catalyst acidity characterization. a Pyridine adsorbed FT-IR spectra. b NH3-TPD curves

Full size image

NH3-TPD results of fresh S2O82−/Fe-ZrO2 and pure ZrO2 catalysts are presented in Fig. 5b. Regarding NH3-TPD measurement results, acid sites can be divided into weak (< 200 °C), medium (200–400 °C), and strong acid sites (> 400 °C) according to the desorption temperature of NH3 [44, 48]. According to Fig. 5b, pure ZrO2 only possesses a few weak and medium–strong acid sites. However, as shown in Fig. 5b, the broad NH3-TPD desorption curve of the S2O82−/Fe-ZrO2 catalyst demonstrates that the acid sites are widely distributed and that the weak, medium, and strong acid sites coexist. S2O82−/Fe-ZrO2 has a higher acid strength than pure ZrO2, especially for strong acid sites. Therefore, the strong acid sites may be one of the most important factors determining the catalytic nitration performance of the catalyst.

Catalytic mechanism

Figure 6 gives the possible mechanism of nitration 1-NN over the S2O82−/Fe-ZrO2 catalyst in the O2-Ac2O system. During the NO2 and O2 reaction, there is an oxidation process that might release the resultant NO2+·NO3− species [11, 55], and the liberated NO2+·NO3− can react with Ac2O to form AcONO2, which is then adsorbed on the acid site of the catalyst, resulting in the formation of an acetyl nitrate cation (AcONO2+) [25]. The electrophilic species AcONO2+ is likely to assault 1-NN to generate the Wheland intermediate and release acetic acid [11]. Finally, the S2O82−/Fe-ZrO2 catalyst is recovered by acquiring a proton from the intermediate and nitration products are also achieved.

Fig. 6
figure 6

Possible nitration mechanism of 1-NN with NO2 over S2O82−/Fe-ZrO2 in the O2-Ac2O system

Full size image

Conclusions

In summary, an effective and environmentally friendly method has been reported to prepare a high 1,5-DNN selectivity from the nitration of 1-NN with NO2 aided by O2-Ac2O using a S2O82−/Fe-ZrO2 catalyst. The findings demonstrate that the S2O82−/Fe-ZrO2 catalyst and O2-Ac2O play an important role in the synthesis of 1,5-DNN. The S2O82−/Fe-ZrO2 catalyst possesses high Lewis acid sites and strong acidity, thus exhibiting excellent catalytic activity and stability. 96.8% conversion of 1-NN and 62.6% selectivity of 1,5-DNN were achieved under optimal conditions. Compared to the previously reported catalyst for the nitration of 1-NN, the S2O82−/Fe-ZrO2 catalyst presents greater catalytic activity. The well-developed 1-NN nitration process for the creation of valuable 1,5-DNN, which uses NO2 as a benign nitrating agent instead of HNO3 and S2O82−/Fe-ZrO2 as a solid acid catalyst instead of H2SO4, has significant promise for real-world applications.