Fullerene bisadduct stabilizers: The effect of different addition positions on inhibiting the autocatalytic decomposition of nitrocellulose absorbed nitroglycerin

Ling Liao, Bo Jin, Zhi-cheng Guo, Fei Xian, Chen-jie Hou, Ru-fang Peng

State Key Laboratory of Environment-friendly Energy Materials, School of Materials Science and Engineering, Southwest University of Science and Technology, Sichuan, Mianyang, 621010, China

Keywords:Fulleropyrrolidine bisadducts Stabilization effect Nitrocellulose/nitroglycerine

ABSTRACT To explore the effect of different positions and number of pyrrolidine bound to the carbon cage on the stabilization effect of fulleropyrrolidine derivatives to nitrocellulose (NC)/nitroglycerine (NG), we synthesized N-(4-methoxy) phenylpyrrolidine-C60 and four different of bis(N-(4-methoxy) phenylpyrrolidine)-C60 compounds through Prato reaction. Their structures were characterized by UV-vis, 1H NMR, 13C NMR, high-resolution mass spectroscopy, and single-crystal X-ray diffraction. Their stabilization effect to NC/NG were investigated using differential scanning calorimetry, methyl violet, vacuum stabilization effect, weight loss, and accelerating rate calorimeter tests. The results indicated these compounds had excellent stabilization effect to NC/NG. The stabilization effect of the fulleropyrrolidine bisadducts to NC/NG is significantly better than that of fulleropyrrolidine monoadduct and C60.Moreover,the position where pyrrolidine binds to fullerene in fulleropyrrolidine bisadducts is different, and its stabilization effect to NC is also different. The stabilization effect order of different bisadduct isomers to nitrocellulose is as follows:e-edge>trans-2>cis-2>trans-3.Electron paramagnetic resonance(EPR)and FT-IR were used to study the stabilization mechanism of fulleropyrrolidine derivatives to NC/NG.The EPR results also show that fulleropyrrolidine bisadducts with different addition sites have different abilities to absorb nitroxide, and their ability is better than that of the monoadduct and C60, which is consistent with the results of stabilization effect performance test.

1. Introduction

Double base propellant plays an increasingly important role in various military and civil fields.Its structure is mainly composed of NC and NG[1-6].Because of the low bond energy of nitrate groups,NC and NG are prone to autocatalytic decomposition. The decomposition of NC and NG are further promoted by the nitroxyl radicals and nitrogen oxides in the decomposition products.As a result,the stabilization effect and performance of propellants gradually become worse, and even lead to explosion [7-9]. This situation is mitigated by adding traditional stabilizers, such as diphenylamine(DPA), p-methyl-N-nitroaniline(MNA), and N-methyl-N′, N′′-diphenylurea(AKII), into the propellants [9,10]. Considering the poor thermal stabilization effect of traditional stabilizers, the stabilization effect of traditional stabilizers to NC/NG decreases when the ambient temperature increases. Moreover, these traditional stabilizers are consumed immediately under specific situations,such as high-temperature or-pressure environments [11-15]. It is very meaningful to design a new stabilizer which can be satisfied the harsh application environment of propellant.

Considering the unique structure and properties of fullerenes,they exhibit high reactivity, excellent electronic properties, and strong ability to scavenge free radicals.They have been widely used in biomedicine, material chemistry, solar photovoltaic materials,and other fields[16-22].Therefore,our team proposed a new idea for the development of novel stabilizers by combining fullerene with stabilizing groups.After proposing the application of fullerene derivatives as NC stabilizer, various fullerene derivatives were synthesized, including 4,11,15,30-tetramethylhophenyl fullereno[1,2:2′,3′] dihydrobenzofuran [23], 4,11,15,30-tetraalylamino fullerenoarylaziridines [24], cis-1 and cis-2 bis (benzofuro) [60]fullerene derivatives [25], fullerene pyrrolidine monoaddition derivatives [6], and fullerene malonamide derivatives [26]. From the results,the effect of fullerene derivatives on the stabilization effect of NC is better than that of the typical stabilizers (C2,DPA,and AK II).

Because fullerene based stabilizer is more expensive than traditional stabilizer,the application of fullerene based stabilizer is still in theory. If fullerene based stabilizers want to replace traditional stabilizers, they must have the same or even better stabilizing effect than traditional stabilizers at the same cost.In order to improve the stabilization efficiency of fullerene based stabilizer and obtain the same stabilization effect,less fullerene based stabilizer is needed to reduce the use cost of fullerene based stabilizer. It is of great significance to explore what structure can improve the stabilization efficiency of fullerene based stabilizer.

Among these compounds, fullerene pyrrolidine monoadduct and the bis(benzofuro)[60]fullerene derivatives (cis-1 and cis-2)show excellent compatibility and stabilization effect in NC. However, the above research did not explore the effect of the different numbers and positions of molecular group bound to carbon cage on its stabilization effect to NC. Hence, the effect of the position and quantity of the molecular group bound in the fullerene carbon cage on the stabilization effect of NC/NG should be determined. This research gap has provided significance for the synthesis of fullerene derivatives with excellent stabilization effect.

In the cases of bisaddition of a symmetrical addend,eight regioisomers are possible (Fig.1) [27]. The maximum number of bisadduct fullerene pyrrolidine isomers can be obtained when the tether-directed remote multi-functionalization methodology is not used. In this work, we have successfully synthesized N-(4-methoxy) phenylpyrrolidine-C60compound and four kinds of bis(N-(4-methoxy) phenylpyrrolidine)-C60compounds (trans-2,trans-3, e-edge, and cis-2) by Prato reaction to explore the influence of different positions and number of pyrrolidines bounded to carbon cage on the stabilization effect of fullerene pyrrolidine derivatives to NC/NG. Results show that among these fullerene pyrrolidine derivatives,the stabilization effect of bisadducts to NC/NG was better than that of monoadditions,and the stabilization effect of bisadducts with different structures to NC/NG was distinguishing. Moreover, the possible mechanisms involved in this phenomenon were further clearly explored. Our research could provide a new insight into the design and synthesis of novel fullerene-based stabilizers.

Fig.1. Nomenclature for the eight possible region-isomers of C60 (first addition site is designated with a black bond).

2. Experimental

2.1. Synthetic procedures

2.1.1. Materials

NC with nitrogen content of 12.76% and NC/NG consists of NC(72.7%) and NG (27.3%) were provided by China Academy of Engineering Physics. p-Methoxyaniline was obtained from Aladdin Industrial Corporation (purity ≥ 99%). Ethyl chloroacetate was provided by J&K Scientific Corporation (purity ≥99%). All the solvents used in the test were purchased from Kelong Chemical reagents Corporation. Raw material C60(purity > 99.5%) were produced by Puyang Yongxin Fullerene Technology Corporation.

2.1.2. Synthetic procedures

N-substituted (4-methoxy) phenyl glycin [6] was synthesized from p-methoxyaniline and ethyl chloroacetate. C60(155 mg,0.2 mmol), N-substituted (4-methoxy) phenyl glycine (358 mg,2 mmol), and chlorobenzene (50 mL) were added into pressureresistant bottle (150 mL) and stirred at 130°C for 24 h. After the reaction was completed, the reaction system was cooled to room temperature, and then chlorobenzene was removed by vacuum distillation. Pure isolated products were obtained by column chromatography and further purified by HPLC(Buckyprep column,toluene/cyclohexane = 2/1). The structure and yield of each product were recorded in Fig. 2.

2.2. Equipment and conditions

2.2.1. Characterization

1H nuclear magnetic resonance (NMR) and13C NMR spectra were obtained using an AVANCE III600 MHz NMR spectrometer.CDCl3was used as solvents, and TMS was used as the internal standard.Ultraviolet-visible(UV-vis) spectra were obtained with CS2as solvent and UV-1600 spectrophotometer as testing instrument.High-resolution mass spectrometry(HRMS)was obtained by using 4-hydroxy-α-cyanocarcinic acid as the matrix,MALDI-TOF as testing instrument in positive-ion mode. The data of single crystal were obtained by using the intelligent Apex CCD diffractometer(Bruker) equipped with graphite monochromatic Cu Kα radiation(λ=1.54184)as the measuring instrument in the ω and φ scanning modes.

2.2.2. Preparation of the NC/NG/stabilizers samples

The simple solvent evaporation method was used to obtain the mixtures of NC/NG/stabilizers (3 wt %). The specific steps were as follows:970 mg NC/NG and stabilizer(30 mg)were added into the beaker,then 20 ml CS2was added to stir evenly,and the stabilizer and NC/NG were mixed evenly during the solvent evaporation process.Finally,the mixture was dried in a vacuum oven(45°C)for 48 h to obtain a spare sample.The details of the formula are shown in Table 1.

2.2.3. DSC test

Fig. 2. Product yields and reaction conditions for the reaction of C60 with N-substituted (4-methoxy) phenyl glycine.

Table 1 Basic formulations of NC/NG/stabilizers samples.

The differential scanning calorimetry(DSC)test was carried out with the Q200 model of TA company. Using platinum plates with aluminum inserts, the sample mass was about 0.75 mg (NC/NG)and 1.5 mg (NC/NG/stabilizer). The DSC data of the samples were obtained under the condition of nitrogen flow atmosphere(50 mL/min) and temperature from 40°C to 250°C.

2.2.4. Methyl violet test

The experiment of methyl violet was carried out with VDY00-01 device under the condition of 120°C and 300 mg sample mass.At the same time, the time of methyl violet test paper changing from purple to orange red was recorded.

2.2.5. Vacuum stabilization effect test (VST)

Under the experimental conditions of 90°C, 100 mg sample mass and 48 h retention time, the vacuum stabilization effect test was carried out with YC-1 device of Xi'an Modern Chemistry Research Institute. Record the amount of gas released from the sample after the experiment.

2.2.6. TG test

Under the experimental conditions of nitrogen flowing atmosphere (40 mL/min), temperature 120°C, constant temperature time 6 h and sample mass 0.75 mg,the weight loss experiment was carried out by 404 F1 differential scanning calorimeter(Netzsch).At the same time,the weight loss rate of the sample was recorded.

2.2.7. ARC test

This test device(ARC 244)is produced by NETZSCH Co.,Ltd.The sample(35 mg,specific heat capacity of 1.24 J/g?°C)was placed in a titanium bomb with a mass of 10 g,a volume of 7.3 ml and specific heat capacity of 0.59 J/g?°C, sealed in air atmosphere. Initially heated to 120°C (the sample does not decompose under the condition below 120°C),and then equilibrated for 20 min,followed by a 10 min seek for an exothermic signal,which can be detected if the self-heat rate > 0.02 K/min. In short, the exothermic threshold is 0.02 k/min. If the exothermic signal was not detected, then the temperature was increased by 1°C with the subsequent repetition of the heat-wait-seek (H-W-S) periods. This H-W-S mode continued until to the decomposition of sample completed. At the same time, the curves of temperature, temperature change rate,pressure and pressure change rate with time under the adiabatic condition were recorded.

2.2.8. EPR test

The instrument for EPR test is Bruker-EMXnano. The test was conducted under the conditions of sweep width 150.0 G, center magnetic field 3373.05 G and sample g-factor 2.0400. These auxiliary reagents were sodium nitroprusside (SNP) (2 mM), FeSO4?(H2O)7(20 mM), sodium diethyldithiocarbamate trihydrate(DETC) (40 mM). Fullerene derivatives were dissolved in CS2to prepare solutions with different concentrations (0.2, 0.4, 0.8, 1.6 and 3.2 mM).

3. Result and discussion

3.1. Structural characterization of fullerene derivatives

The monoadduct N-methylpyrrolidine-C60was synthesized as previously described[6].Further,flash chromatographic separation on silica gel gave (in order of elution following) yielded four products (compounds 2-5). The high-performance liquid chromatography (HPLC) spectra of compounds 2-5 were obtained using HPLC LC-9104(Japan Analytical Industry Co.,Ltd.)with toluene as the mobile phase at a flow rate of 2 mL/min.The results showed that the purity of compounds 2-5 exceeded 99% (Fig. S1). The HRMS spectra (Figs. S11-14) results of compounds 2-5 were 1018.1681, 1018.1684, 1018.1680, and 1018.1684 respectively, indicating that the four products were N-(4-methoxy) phenylpyrrolidine-C60bisadducts. The results of UV-vis spectra (Fig. S2)indicated that the characteristic absorption peaks of compounds 2-5 were as follows: compound 2: 485, 658, 690, and 726 nm;compound 3: 419, 469, and 701 nm; compound 4: 422 nm; and compound 5: 455 nm; Based on the previously reported UV-vis spectra and chromatographic elution sequence of different bis(ethoxycarbonyl)methylene-C60isomers in the 400-700 nm region[27-31], some tentative structural assignments could be made for compounds 2-5, that is, compounds 2-5 are trans-2, trans-3, eedge,and cis-2 isomers, respectively.

The1H and13C NMR spectra of compounds 2-5 were analyzed based on symmetry considerations (Figs. S3-10). The number and intensity of the13C signals of the residual sp2fullerene carbons in the eight potential bisadduct isomers were reported in Table S1[31]. Four groups of symmetry were defined, and each group resulted in a well-defined number of resonances in the1H and13C NMR spectra.Notably,the13C NMR spectra signals of carbon atoms connected with nitrogen and oxygen atoms in benzene ring were near 141 and 152.The number and intensity of the sp2carbons NMR signals of the four different fulleropyrrolidine bisadducts 2-5 were listed in Table 2. The13C NMR spectrum of 2 displayed 32 signals between δ 125.53-160.94 ppm,with 28 of them integrated for two sp2carbon atoms,and the four remaining were integrated for one.Therefore, compound 2 could be assigned as having a trans-2 structure, consistent with the conclusion of UV-vis spectrum.Moreover,its1H NMR spectrum showed one singlet at δ 3.86 ppm for the six methoxy protons，four doublets at δ 4.94-5.29 ppm for the eight methylene protons of two pyrrolidine rings，and two double peaks at δ 6.99-7.28 ppm for the eight protons of two different p-substituted benzene rings, consistent with the C2symmetry of the trans-2 isomer[29-31].The13C NMR spectrum of 3 displayed 32 signals between δ 126.42-155.25 ppm, with 28 of them integrating for two sp2carbon atoms,and the four remaining were integrated for one, consistent with the conclusion of UV-vis spectrum [31]. Therefore, compound 3 could be assigned as having a trans-3 structure. Moreover, the1H NMR spectrum of compound 3 showed one singlet at δ 3.79 ppm for the six methoxy protons,four doublets at δ 4.66-5.03 ppm for the eight methylene protons of the two pyrrolidine rings, and two double peaks at δ 6.89-7.13 ppm for the eight protons of two different p-substitutedbenzene rings, consistent with the C2symmetry of the trans-3 isomer [29-31]. The13C NMR spectrum of compound 4 displayed 33 signals between δ 135.8-158.6 ppm,where 27 signals integrate for two sp2carbon atoms,and the six remaining integrate for one,consistent with the conclusion of UV-vis spectrum. Therefore,compound 4 could be assigned with an e-edge structure. The1H NMR spectrum of compound 4 showed two singlets at δ 3.83-3.84 ppm for the six methoxy protons, two doublets and two singlets at δ 4.37-5.30 ppm for the eight methylene protons of the two pyrrolidine rings, and two multiple peaks at δ 6.96-7.14 ppm for the eight protons of two different p-substituted benzene rings, consistent with the CSsymmetry of the e-edge isomer [29-31]. The13C NMR spectrum of compound 5 displayed 34 signals between δ 128.37-158.02 ppm,where 26 carbon atoms integrate for two sp2carbon atoms, and the eight remaining integrate for one, consistent with the conclusion of UV-vis spectrum.Therefore, compound 5 could be assigned as having a cis-2 structure.The1H NMR spectrum of compound 5 showed one singlet at δ 3.79 ppm for the six methoxy protons, four doublets at δ 4.37-4.63 ppm for the eight methylene protons of the two pyrrolidine rings, and two double peaks at δ 6.88-7.05 ppm for the eight protons of two different p-substituted benzene rings,consistent with the Cssymmetry of the cis-2 isomer [29-31].

Table 2 Symmetry groups with number and intensities (in parentheses)of the13C fullerene sp2 signals of the four different fulleropyrrolidine bisadducts 2-5.

The structures of compounds 2-5 were clearly assigned as having trans-2, trans-3, e-edge and cis-2 structures based on1H NMR,13C NMR, UV-vis, and HRMS results. Accordingly, we used CS2and n-hexane as benign and non-benign solvents,respectively,to obtain the single crystal of compound 3 by liquid-phase diffusion. Its structure was confirmed as trans-3 based on previous characterization results. As shown in the single crystal X-ray diffraction analysis,this single crystal had an orthorhombic crystal structure that pertains to Fdd2 space group, and the two pyrrolidine rings were on the correct position of the fullerene cage(Fig.3 and Table S2). The results of single crystal X-ray diffraction were highly consistent with the structure of trans-3,thus supporting our previous conclusion on distinguishing compounds 2-5 by1H NMR,13C NMR, UV vis, and HRMS.

3.2. Compatibility of fulleropyrrolidine derivatives with NC/NG

The evaluation of compatibility between NC/NG and different stabilizers was based on the standardization agreement STANAG 4147 of the NATO.According to this agreement,if the temperature difference between the exothermic peaks in the DSC analysis of NC/NG and mixed samples is less than 4°C, they are considered to be compatible[32].The compatibility of NC/NG with C60and 1-5 were studied by DSC.Fig.S15 showed that the DSC curves of all samples were single decomposition peaks. Table 3 showed that the exothermic peak temperature of NC/NG/C60and NC/NG/compound 1-5 stabilizers were 0.04-1.71°C higher than that of NC/NG.These results indicate that C60and 1-5 had good compatibility with NC/NG.

Table 3 Compatibility data of mixture sample(NC/NG/stabilizers)for DSC peak temperature.

3.3. Stabilization effect of fulleropyrrolidine derivatives

3.3.1. Methyl violet test

Based on the methyl violet test, the stabilization effect of stabilizer was evaluated based on the time consumed for the methyl violet test paper to change from purple to orange. The nitrogen oxides produced by thermal decomposition of NC/NG can make methyl violet test paper turn orange.Because the nitroxide radical produced by thermal decomposition of NC/NG is absorbed by stabilizer, the autocatalytic decomposition rate of NC/NG is reduced,the discoloration time of methyl violet test paper is prolonged.Therefore, the discoloration time of the methyl violet test paper could directly reflect the stabilization effect of stabilizer. The discoloration process of methyl violet test paper and the time taken for each sample to change color were recorded in Fig. 4. The discoloration times of S-1～S-7 were 60, 75, 84, 99, 96, 96, and 102 min,respectively,which indicated that the addition of fullerene derivatives delayed the discoloration time of methyl violet test paper by 15-42 min.The order of the discoloration time was S-6>S-4 > S-7 = S-5 > S-3 > S-2 > S-1. Therefore, the order of stabilization effect was 4 > 2 > 5 = 3 > 1 > C60, indicating that the stabilization effect of the bisadducts with different structures differed(e-edge > trans-2 > cis-2, trans-3) and was better than the monoaddition and C60. Notably, the discoloration time of S-7 and S-5 is the same, and this condition is the result of a vacuum period between two adjacent records of methyl violet test paper discoloration.

3.3.2. Vacuum stabilization effect test

VST was used to determine the stabilization effect of the stabilizer in the sample by measuring the volume of gas released from a unit mass of sample under experimental conditions. Under the same conditions, the less gas released from the sample decomposition, the better the stabilization effect of the stabilizer. Fig. 5a shows the pressure curves and the volume of gas released from decomposition for S-1～S-7 under test conditions. The pressure followed the order of S-6< S-4

Fig. 3. (a) X-ray diffraction patterns of 3. (b) Stacking diagram of 3 single crystal (solvent molecules have been removed).

Fig.4. (a)Color change of samples in methyl violet test at 120 °C.(b)The discoloration time of methyl violet test paper in the sample under the experimental conditions of methyl violet.

Fig. 5. (a) Pressure curves and (b) outgassing volumes of vacuum stabilization effect test samples.

Fig. 6. (a) Weight loss curves and (b) final weight remaining rates of the loss test samples.

Where, the volume of gas released by the sample under standard conditions is expressed by VH,the sample pressure is expressed by P (Pa), the reactor volume is expressed by V0(mL), the sample volume is expressed by VG(mL),and the experimental temperature is expressed by T(K).The detailed results are shown in Fig.5b.The gas released by per unit mass of S-1～S-7 were 3.88,3.57,3.25,2.73,3.01,2.52,and 2.86 mL/g,respectively.Consequently,the volume of gas released per unit mass of all samples was in the following order:S-6 < S-4 < S-7 < S-5 < S-3 < S-2 < S-1. Hence, the stabilization effect followed the order 4 > 2 > 5 > 3 > 1 > C60. These results revealed that the stabilization effect of the bisadducts with different structures were different (e-edge > trans-2 > cis-2>trans-3)and better than the monoadditions and C60.Meanwhile,these results more accurately distinguished the stabilities of compounds 3 and 5.

3.3.3. TG test

TG test was used to determine the stabilization effect of the corresponding stabilizer by recording the mass loss of the sample under the experimental conditions. The greater the weight remaining rate of the sample, the better the stabilization effect of the corresponding stabilizer. Fig. 6 shows the weight remaining curve and final weight residue of all samples.The weight remaining rates of samples S-1～S-7 were 51.84%, 57.15%, 61.32%, 67.36%,66.37%, 69.5%, and 66.51%, respectively. The thermal stabilization effect of all samples was in the following order:S-6>S-4>S-7>S-5 > S-3 > S-2 > S-1. Therefore, the stabilization effect order of the stabilizers was 4 > 2 > 5 > 3 > 1 > C60. This result show that the stabilization effect of the bisadducts with different structures differed (e-edge > trans-2 > cis-2 > trans-3) and was better than that of the monoadditions and C60.

3.3.4. ARC test

ARC is an effective method for evaluating the thermal hazard of reactive chemicals. The overall process of exothermic decomposition for a mass of sample was shown using ARC under adiabatic condition. Fig. S16 shows the curves of temperature, temperature change rate, pressure, and pressure change rate with time for S1-S7.The temperature recorded by ARC does not reflect the actual temperature of sample but that of the system of sample and closedbomb. Under an ideal adiabatic condition, the reaction heat is completely consumed by heating up the sample. Therefore, the experimental data should be modified by incorporating the thermal inertia factor φ before applying them into practice [33]. The expression for φ is as follows:

where y (°C) is the heating temperature, and x (min) is the time required. Table 4 shows the detailed parameters of the fitting formula. The TMR curves of S-1～S-7 showed that the thermal stabilization effect of all samples were in the following order S-6>S-4>S-7>S-5>S-3>S-2>S-1.Therefore,the stabilization effect order of the stabilizers was 4>2>5>3>1>C60.The results of ARC test revealed that the stabilization effect of the bisadducts with different structures differs(e-edge>trans-2>cis-2>trans-3)andis better than that of the mono-addition product and C60.

Table 4 Fitting curves parameters of TMR.

3.4. Stabilization effect mechanism

3.4.1. EPR test

In order to explore the stabilization mechanism of fullerenebased stabilizer for NC/NG, the absorption capacity of compounds 1-5 and C60for nitroxide radicals was investigated by EPR. The samples without compounds 1-5 and C60showed strong EPR signals of NO?,which were much higher than those with compounds 1-5 and C60(See Supporting Information Fig. S17 for details). In addition,the samples with high concentrations of compounds 1-5 and C60showed lower EPR signal intensity of NO?than those with low concentrations of compounds 1-5 and C60. Different compounds in the same concentration conditions, by comparing their EPR absorption peak strength to determine their ability to scavenge free radicals[36].High absorption intensity indicates poor ability of scavenging free radicals.The EPR curves of compounds 1-5 and C60at 3.2 mM and blank sample are depicted in Fig.8a.The absorption peak intensity of 1-5 and C60followed the order 4 < 2 < 5 < 3 <1 < C60< blank sample. The estimation formula of NO?clearance rate of 1-5 and C60were as follows:

where the NO?clearance rate was expressed as X(%),the EPR signal intensity of blank sample was expressed as I0, and the EPR signal intensity of NO?after the addition of compounds 1-5 and C60was expressed as IC. The NO?scavenging rates of compounds 1-5 and C60at different concentrations are shown in Fig. 8b. The NO?scavenging rates of 1-5 and C60at these concentrations (0.2, 0.4,0.8,1.6,and 3.2 mM/L)followed the order 4>2>5>3>1>C60.The IC50(half-maximal inhibitive concentration) of compounds 1-5 and C60were calculated to compare their ability to absorb NO?.Fig. 8c shows the fitting EPR curves of NO?clearance rates of compounds 1-5 and C60at different concentrations.The details of the fitting formula are as follows:

Fig. 7. (a) Detailed values of T0, T1, and θm of samples and (b) TMR fitting curves of samples decomposition temperature ranges.

Fig.8. (a)The EPR curves of compounds 1-5 and C60 at the 3.2 mM/L and blank sample;(b)NO?scavenging rates of compounds 1-5 and C60;(c)fitting curves of compounds 1-5 and C60 for the NO?scavenging rates; and (d) IC50 values of compounds 1-5 and C60 for the NO?scavenging rates.

where the NO?scavenging rate of compounds 1-5 and C60were expressed as y (%), and their concentration were expressed as x(mM/L).Table 5 recorded the detailed parameter information of the fitting formula. The smaller the IC50value of the compound, the stronger the absorption capacity of NO?. As shown in Fig. 8d, the IC50values of compounds 1-5 and C60for the NO?scavenging rates were 0.6234, 0.3861, 0.4312, 0.3381, 0.4042, and 1.0048 mM/L,respectively.The IC50values of compounds 1-5 and C60for the NO?scavenging rates had the following order 4 < 2 < 5 < 3 < 1 < C60.Hence, EPR results showed that the order of the ability of nitroxyl radicals absorption was 4>2>5>3>1>C60,thus supporting the stabilization effect test results.

According to the reports in Ref. [6], due to the different structures of fullerene groups, the IC50values of two fullerene pyrrolidine monoaddition derivatives with different structures differ by 0.729. According to the measured data, the difference of IC50between fullerene pyrrolidine monoaddition derivatives andfullerene pyrrolidine double addition derivatives is 0.285.It can be concluded that the scavenging efficiency of fullerene pyrrolidine double addition derivatives is higher than that of fullerene pyrrolidine single addition derivatives under the same structure of fullerene cage binding group.

Table 5 Fitting parameters of the fitting curve for the scavenging rate of all fullerene derivatives to nitroxide radicals.

3.4.2. FT-IR test

Fig. 9. FT-IR spectra of products 2-5 before and after interaction with NC/NG.

The FT-IR spectra of compounds 2-5 and the FT-IR spectra of compounds 2-5 after absorbing nitrogen oxides produced by NC/NG decomposition were recorded in Fig. 9. New strong [6] NO2vibrational peaks were observed at 1650 and 1274 cm-1.Moreover,the absorption peaks of C-N-O flexural vibration and C-N stretching vibration of aromatic nitro compounds were observed at 749 cm-1and 845 cm-1, respectively. The FT-IR spectra show that compounds 2-5 absorbed the nitrogen oxides in the mixture system effectively, while effectively reducing the autocatalytic decomposition rate of NC/NG and prolonging the safe storage time of NC/NG.

4. Conclusion

In the present work, four bis (N-(4-methoxy) phenylpyrrolidine)-C60compounds and one N-(4-methoxy) phenylpyrrolidine-C60compound were synthesized successfully by the Prato reaction. The structures of the four bisadducts were successfully defined by UV-vis,1H NMR,13C NMR, HRMS, and single crystal X-ray diffraction. The compatibility and stabilization effect of the compounds in NC/NG were determined by VST,methyl violet,TG, DSC, and ARC tests. The results of stabilization effect test indicate difference in the stabilization effect of the bisadducts with different structures (e-edge > trans-2 > cis-2 > trans-3), and their stabilization effect was better than that of the monoaddition products and C60. In order to further explored the stabilization mechanism of fullerene pyrrolidine derivatives on NC/NG, we characterized them in detail by EPR and FT-IR test.The results show that the order of the ability of absorb the nitroxyl radicals was eedge > trans-2 > cis-2 > trans-3 > monoaddition > C60, thus supporting the stabilization effect test results. Based the results, the more the pyrrolidine groups are incorporated into the carbon cage of the bisadducts, the stabilization effect on NC/NG is increased.Moreover, the different positions of pyrrolidine bound to the carbon cage of fullerene pyrrolidine made their stabilization effect on NC/NG quite different. Finally, the stabilization effect of the bisadducts with e-edge structure to NC/NG is stronger than that of trans-2, trans-3, and cis-2.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

This work was supported by National Natural Science Foundation of China (51972278), Outstanding Youth Science and Technology Talents Program of Sichuan (no. 19JCQN0085), and Open Project of State Key Laboratory of Environment-friendly Energy Materials, Southwest University of Science and Technology (No.19fksy04).

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.dt.2021.04.015.