Thermal Decomposition of Ammonium Perchlorate and Ammonium Nitrate

Sci. Tech. Energetic Materials, Vol．７７,No．３,２０１６ 65

Research paper

Thermal decomposition of ammonium perchlorate and ammonium nitrate doped with nanometer 2- SO4 /Fe2O3 solid strong acid as catalyst ４ ０ ９ Xiaolan Song＊†,YiWang＊＊,DanSong＊＊＊,Chongwei An＊,andJingyu Wang＊

＊School of Chemical Engineering and Environment, North University of China, Taiyuan 030051, CHINA Phone : +8615034015691 †Corresponding author : [email protected]

＊＊School of Materials Science and Engineering, North University of China, Taiyuan 030051, CHINA

＊＊＊China Ordnance Institute of Science and Technology, Beijing 100089, CHINA

Received : December 18, 2015 Accepted : February 9, 2016

Abstract

Amorphous Fe２O３ nanoparticles were prepared by precipitation method. These nanoparticles were dipped in dilute 2- sulfuric acid and then werecalcinedatdifferent temperature. After that,theSO4 /Fe２O３ solid strong acids with particle size of 30-40 nm were fabricated. Thermal analyses were recruited to probe the catalysis of nanometer Fe２O３ and 2- nanometer SO4 /Fe２O３ on decomposition of ammonium perchlorate (AP) and ammonium nitrate (AN). The results ο ο 2- indicated that in case of AP, the peak temperature decreased by 35.1 Cand98.7 CinuseofFe２O３ and SO4 /Fe２O３ as ο ο 2- catalysts respectively. In the case of AN, the peak temperature lowered by 2.2 Cand18.0 CinuseofFe２O３ and SO4 / 2- Fe２O３ as catalysts respectively. DSC-IR analysis were performed to investigatedecomposition products of [AP+3%SO4 / 2- 2- Fe２O３]and[AN+3%SO4 /Fe２O３]. The results demonstrated that [AP+3%SO4 /Fe２O３]decomposed to NO２, NOCl, N２O, 2- HClO４, NO, HCl, and H２O. For [AN+3%SO4 /Fe２O３], the decomposition products were massive N２OandfewH２O. Note that there was no NH３ presented in products, which meant thatthedecomposition reaction proceeded very completely. According to the detected products, the possible decomposition mechanismofAPandANwere derived. Meanwhile, the 2- catalysis mechanisms of SO4 /Fe２O３ were discussed in detail.

Keywords :nanoparticles, solid strong acid, oxidizer, thermal decomposition, mechanism

1. Introduction decomposition. Therefore, adding some solid catalysts to Ammonium perchlorate (AP) is the most common decrease decomposition temperature of AP became a hot oxidizer used in composite solid propellants. It possesses of research topic since 20031）-５）.Inparticular, nanometer traits such as high oxygen balance, high energy, and catalysts exhibited higher catalysis than the micron excellent ignition and combustion properties. Due to the catalysts６）-10）.For example, the AP/Al/hydroxyl superior comprehensive performance, it isverydifficult to terminated polybutadiene (HTPB) propellant was used in replace AP by another oxidizer. Despite these advantages, the first stage booster for launch of space shuttle, in which there are still two problems in use. One is the agreat deal of Fe２O３ severed as combustion catalyst. Of hygroscopicity ; another is the much high decomposition course, there were many biting criticisms successively ο temperature (>400 C). For hygroscopicity, it is not so came from environmental experts because the combustion serious that to block its application. However, too high of those propellants discharged tons of HCl that caused decomposition temperature would deteriorate its serious air pollution. Different from AP based propellants, reactions in condense and gas phase, i.e. pure AP the propellants using ammonium nitrate (AN) as oxidizer consumes too many heats for sustaining its thermal are of the advantages such as non-toxic discharge, low 66 Xiaolan Song et al. signature, and low sensitivities etc. HCl and oxychloride burning rate was very small ("!mm s-１). The catalysis of are not contained in combustion products of AN based NaCl, BaCl２,andNaFoncombustion of AN/GAP propellants. Moreover, sensitivity of AN is much lower propellants were also probed by Sinditskii18）.Hedisclosed than that of AP. For example, pure AN can not be ignited that the burning of propellants added with 7%NaCl could by flame or heating evenifthepressure accesses to 100 sustain at pressure of 0.5MPa, but the burning rate was MPa. Its impact, friction, and shock sensitivities are closed still less than 1 mm s-１.Miyataalsoconfirmed that the to zero. Meanwhile, it explosion heat accesses 2640 J g-１ mixture of AN and aminoguanidinium 5,5’-azobis-1H- and detonation velocity reaches 5270 m s-１ (#!!!#"g tetrazolate (AGAT) (50/50) can be ignited at pressure of cm-３). For AP, its explosion heat is 1112.9 J g-１ and its 0.5MPa, but the burning rate was only closed to detonation velocity is only 3800 m s-１.Thus,ANisatypical 1mm s-1 19）.Although abovementioned results were insensitive energetic material. Despite the super unsatisfactory, theresearch about AN based propellants insensitivity, now it is very early to say we can use AN to was rejuvenated. replace AP. In practice, AN had not been applied as main Now we found that since 2011, the studies about AP oxidizer in the propellants of certain type of missile (or decreased but the researches about AN increased. In rocket, or booster), and it were just slightly used in some particular the researchers from Russia, Japan, and India formulation of gas generation agents11）,12）.Whatdoes paid more attention on ANbased propellants20）-23）.Inthis results in this? Three inherent defects account for the paper, after studied the decomposition mechanism of AP uselessness of AN13）.Firstly, AN presents much more and AN in detail, we proposed that solid strong acid may hygroscopicity than AP ; secondly, phase transformation exhibit good catalysis ability because their super high ο will occur at low temperature (30-80 C) ; the third, i.e. the acidity may strongly promote the decomposition. So we most fatal factor, the ignition and combustion performance enlisted TG-DTG-DSC and DSC-IR technology to affirm of AN based propellants are very poor. They burned so the suggestion. slowly that their combustion could not provide with sufficient thrust to boost the missile to fly at high speed. 2. Experimental Therefore, the investigation about AN based propellants 2.1 Sample preparation were in silence for many years. With violent stirring, Fe(NO３)３ solution was dropped into

At present, development of Insensitive Ammunition 6mol/L NH３·H２Osolutionthat originallycontained 10% was becoming the “protagonist” in fields of military ethanol as dispersant agent. pH value was adjusted to 9-10, science and technology. Therefore, the studies about how and then red brown precipitation (Fe(OH)３)wasobtained. to improve the performance of AN based propellants After aging 24h, the precipitation was washed with H２O proliferated over the past five years. Especially, for better three times ; then it was washed with ethanol two times ; performance on the aspect of decomposition and finally, it was washed with acetone one time. After ο combustion of AN, many studies were performed but the lavation, the precipitation was dried at 50 C. The dried results were not satisfactory. In studies about AP, the powder was carefully grinded in agate mortar, and then researchers found that a good catalyst could always the amorphous Fe２O３ tiny particles were gained. decrease the decomposition temperature by more than With stirring and ultrasonic, 2 g amorphous Fe２O３ was ο 100 C. However, this kind of results could not present in put into sulfuric acid solution and dipped in 30 min. After the studies about AN. If a catalyst could lower the dipping, the amorphous Fe２O３ was filtrated out and dried ο ο decomposition temperature of AN by more than 10 C, we at 80 C. The dried powders were grinded and calcined at ο ο ο ο think it exhibited a good catalysis ability. For example, 300 C, 400 C, 500 C, or 800 C, and then solid strong acids 2- Naya investigated the catalysis of MnO２ on thermal were obtained. This solid acid was marked as “SO4 / 14） decomposition of AN .Hefound that the catalyst did not Fe２O３”. ο -１ 2- work at higher heating rate (20 Cmin ); astheheating In order to investigate the catalysis of SO4 /Fe２O３ on ο rate was lowered to 2 Cmin-１,itscatalysis action was decomposition of AP and AN, four samples were prepared. 2- perceived, i.e. the DSC peak temperature of AN doped 4% (1) 0.97g AP was blended with 0.03 g SO4 /Fe２O３ by ο MnO２ decreased by 16 Ccomparing with that of pure AN ; carefully manual grinding, and the obtained sample was 2- nevertheless, the propellant added MnO２ as catalyst could tagged as [AP+3%SO4 /Fe２O３]. (2) For comparing, not be ignited at low pressure ("!MPa). In the study of another 0.97 g AP was alsoblended with0.03gFe２O３ by

Popok, he used nano Al and "-Al２O３ as catalyst on thermal carefully manual grinding, and the gained sample was 15） decomposition of AN .NanoAland"-Al２O３ decreased tagged as [AP+3%Fe２O３]. (3) 0.97 g AN was blended with ο ο 2- the peak temperature of AN by 20 Cand7Crespectively. 0.03 g SO4 /Fe２O３ by carefully manual grinding, and the 2- Vargeese investigated the catalysis of CuO, TiO２,andLiF. obtained sample was tagged as [AN+3%SO4 /Fe２O３]. (4)

The results indicated that TiO２ almost showed no For comparing, another 0.97 g AN was also blended with catalysis ; CuO worked a little ; LiF presented a negative 0.03 g Fe２O３ by carefully manual grinding, and the gained 16） effect on thermal decomposition of AN .Hasuestudied sample was tagged as [AN+3%Fe２O３]. the combustion performance of bis (1H-tetrazolyl) amine ammonium salt (BTA · NH３)andphase-stabilized 2.2 Methods and measurements ammonium nitrate (PSAN) mixture17）.Hefound that the The morphology was observed with a field-emission mixture could be ignited at pressure of 1MPa, but the scanning electron microscope (SEM, JEOL JSM-7500). The Sci. Tech. Energetic Materials, Vol．７７,No．３,２０１６ 67

2- phases of the samples were investigated with an X-ray drying process in fabrication of SO4 /Fe２O３.Freeze- diffractometer (XRD, Bruker Advance D8), using Cu K_! drying or supercritical drying may be more feasible. If we radiation at 40 kV and 30 mA. XPS analysis was could use freeze-drying or supercritical technology to dry performed byX-rayphotoelectron spectroscopy (XPS), the power, we might obtain the particles with better PHI5000 Versa-Probe (ULVAC-PHI). TG-DSC analysis and dispersion. The particle size distribution was obtained by DSC-IR analysis were carried out by using a thermal carefully measure the specific size of each particle in SEM analyzer system (TG/DSC, Mettler Toledo) coupled with a images (Figure 3 (a) and Figure 3 (b)), in which the Fourier transform infrared spectrometer. number of the particles that were gauged was more than 120 for each sample. The measurement was carried out by 3. Results and discussion software based on Equation 1. Then the data were 3.1 Morphology and structure displayed as a statistic figure (Figure (3c) and Figure (3d)). ４ 2- The XRD analyses were performed and the results It indicated that Fe２O３ and SO4 /Fe２O３ have their mean ０ were illustrated in Figure 1. It was indicated that the particle size of 27 nm and 25 nm respectively. This was ９ amorphous powder did not transform into crystal phase accordance with result in XRD analysis, i.e. surface ο ο after calcined at 300 Cor400 C. When the temperature acidification slightly retarded the growth of particles (or ο climbed to 500 C, !-Fe２O３ formed and its average grain grains). size was 24.9 nm. When the temperature increased to ! $ # ο "!#"# "&*)#$! "'(%!" 800 C, the peak intensity became very strong and its % (1) ###$%" %## ## average grain size was more than 100 nm. In Figure 1 (b), XRD patterns of two samples were compared. One was where "!#"is particle size distribution ; ## is the the powder underwent surface acidification in sulfuric acid standard deviation of the diameters, ## is themean 2- (SO4 /Fe２O３), and another was the powder without diameter, !is a constant. surface acidification (Fe２O３). The result showed that the 2- peak intensity of Fe２O３ was stronger than that of SO4 / 3.2 Thermal analysis 2- Fe２O３,whichmeant that surface acidification could retard Catalysis of SO4 /Fe２O３ was probed with thermal the crystallization of Fe２O３.Average grain size of Fe２O３ analyses and the results were illustrated in Figure 4. was 29.8 nm. Figure 4 (a) showed that DSC trace of raw AP comprised 2- Figure 2 (a) is the XPS spectrum of SO4 /Fe２O３ that one endothermic peak and two exothermic peaks, which was the powder subjected to acidification in 0.3 mol/L refer to the phase transformation and thermal ο sulfuric acid andcalcinedin500C(Figure 2 (b) is the decomposition of AP. For [AP+3%Fe２O３], the peaks for expansion of Figure 2 (a)). In XPS spectrum, three phase transformation and low temperature decomposition elements of O, Fe, and S were detected. Fe and O peaks changed seldom compared with those of pure AP. The located at binding energy of 706.7eV and 529.9eV peak temperature of high temperature decomposition ο respectively, which should reflect to the elements in Fe２O３ decreased by 35.1 Candthedecomposition heats 2- because their peak intensity was very strong. The peak at increased. For [AP+3%SO4 /Fe２O３], compared with pure binding energy of 168.5eV should 1000 400 relate to 2p electron transition of (a) 2- (b) SO4 /Fe2O3 Selement.The S element should 800 2- 300 origin from the SO4 radical on o Fe2O3 calcined at 500 C 2- 600 the surface of SO4 /Fe２O３.This 800 oC 200 2- implied that SO4 had been fixed 400 500 oC on the surface of Fe２O３ after 100 SO 2-/Fe O calcined at 500 oC acidification. 200 400 oC 4 2 3

Intensity [arbitrary unit] [arbitrary Intensity o 300 C unit] [arbitrary Intensity To disclose the micron 0 0 morphology and particle size of 10 20 30 40 50 60 70 80 10 20 30 40 50 60 70 80 θ 2θ [degree] 2- 2 [degree] SO4 /Fe２O３,SEManalysis were 2- Figure１ XRD patterns of samples : (a) SO4 /Fe２O３ calcined at different temperature ; performed and SEM images of ο 2- (b) Fe２O３ and SO4 /Fe２O３ calcined at 500 C. samples with and without surface (b) 6500 acidification were showed in (a)300000 S2p Figure 3. Figure 3 (a) imaged the 6000 250000 O1s sample without acidification (Fe２ Fe2p 5500 200000 O３). Fe２O３ with particles size of 30 CPS ~40nm were agglomerate. 150000 CPS 5000

Figure 3 (b) indicated that 100000 4500 S2p 2- particles size of SO4 /Fe２O３ were 50000 4000 2- at nano scale. SO4 /Fe２O３ 0 particles were also agglomerate 3500 1400 1200 1000 800 600 400 200 0 174 172 170 168 166 164 162 160 158 each other. Therefore, we had Binding Energy [eV] Binding Energy [eV] 2- enough reason to reconsider the Figure２ XPS spectra of SO4 /Fe２O３ nanoparticles. Figure 2 (b) is the expansion of Figure 2 (a). 68 Xiaolan Song et al.

Figure 4 (c) indicated that there were four endothermic peaks and no exothermic peaks in DSC trace of pure AN. The peak ο at 52.4 Creflected to a phase transformation (phase III to phase II, #!!"#!' Jg-１). The ο peak at 125.8 Calsorefer to a phase transformation (phase II to phase I, #!!$!!" Jg-１). The ο peak at 167.2 Crelated to melting 18 2- (c) nano-Fe O (d)30 nano SO /Fe O 2 3 4 2 3 #!!%%!$ -１ 16 X = 27nm X = 25nm course of AN ( Jg ). 14 25 The strongest endothermic peak 12 20 reflected to thermal 10 cy [%] decomposition of AN. The

en 15 8 ο decomposition began at 239.2 C

equency [%] 6 10 Frequ

Fr and showed its peak point at 4 ο 5 #!!&&&!' -１ 2 277.4 C( Jg ). As the 0 0 first step of condense phase 15 20 25 30 35 40 10 15 20 25 30 35 40 45 50 Particle diameter [nm] Particle diameter [nm] reaction, NH３ (g) and HNO３ (g) ο 2- would form bydissociationofNH４ Figure３ SEM images of Fe２O３ (a) and SO4 /Fe２O３ (b) nanoparticles calcined at 500 C. NO３.Thisdissociation course 100 0.0 -1 24） (a) for AP 342.5 (b) would absorb heat of 2.18kJ·g .

10 dm DTG,

EXO. -0.5 80 Although subsequent reactions 5 -1.0 were exothermic, the released -1.5

ht [%] 60 0 -1 442.6 /dT [%· /dT heat was less than 2.18kJ·g .So, 2- -2.0 AP+3%SO4 /Fe2O3 -5 406.1 40 402.7 the whole decomposition of AN 332.8 -2.5 pure AP

244.9 AP+3% Fe O -3.0 o was endothermic. Therefore, -10 20 C AP+3%Fe O TG, Weig 2 3 2 3 2-

AP+3% SO /Fe O -1 Heat flow [mW] Heat 4 2 3 330.5 -3.5 ] 441.2 341.2 pure AN can not be ignited in air -15 243.1 0 pure AP -4.0 or in N２ because its 242.3 150 200 250300350400 450 500 -20 o 150 200 250 300 350 400 450 500 Temperature [ C] decomposition is not self- o Temperature [ C] sustained. This is why AN

5 100 0.0 (c) for AN (d) presents much low sensitivities. 239.2 -0.5 [%· dm/dT DTG, 80 In addition, we could find that pure AN -1.0 0 -1.5 nanometer Fe２O３ and nanometer ht [%] 60 2- AN+3%Fe2O3 -2.0 SO4 /Fe２O３ did not change the -5 277.4 40 -2.5 261.5 decomposition process of AN, pure AN 273.6 -3.0 II and III 275.2 AN+3% Fe O 2 3 277.3 2- TG, Weig 2- because the DSC curve of [AN+ -10 I AN+3%SO /FeO 20 AN+3% SO /FeO -3.5 o 4 2 3 4 2 3 C 2- -1 Heat flow [mW] Heat 52.4 -4.0 3%Fe２O３]or[AN+3%SO4 /Fe２ 125.8 0 257.5 ] 167.2 276.1 -4.5 -15 O３ ]alsoexhibited four 259.4 150 175 200 225 250275300325 350 Temperature [ oC] endothermic peaks (two to phase 50 100 150 200 250 300 350 400 Temperature [ oC] transformation, one to melting, and one to thermal Figure４ DSC traces of samples : (a, b) for AP ; (c, d) for AN. decomposition). This meant that AP, the phase transformation peak did not move but the adding nano catalyst could not change the reaction heats ο exothermic peak temperature decreased by 98.7 C; in thermodynamics. However, comparing with pure AN, 2- meanwhile, the peaks for low and high temperature the decomposition peak of AN doped 3%SO4 /Fe２O３ ο decomposition incorporated in one exothermic peak. decreased by 18.0 C; andthenano Fe２O３ almost showed These results meant that the catalysis action of nanometer no catalysis because the peak temperature only decreased ο 2- 2- SO4 /Fe２O３ was higher than that of nano Fe２O３.Figure 4 by 2.2 C. This meant that the existence of SO4 (instead of

(b) illustrated the TG-DTG curves of samples, in which the pure Fe２O３)favored to thermal decomposition of AN. DTG curves were obtained by derivation calculus to TG Figure 4 (d) showed the TG-DTG curves of pure AN, [AN 2- curves. In terms of the curves, the difference among the +3%Fe２O３], and [AN+3%SO4 /Fe２O３]. There was not 2- samples was obvious. The DTG peak of [AP+3%SO4 / distinct difference in TG curves. After derivation calculus,

Fe２O３]wasstronger than others, and it was of the highest the DTG curves were gained and the curve depicted the value of "")#")" *(+.Thisimplied that the decomposition decomposition rateofsamples. Their DTG peak profile 2- rate of [AP+3%SO4 /Fe２O３]wasfaster. The peak points was as similar as that of DSC peaks. in DTG curves, where the "")#")" *(+ located, were closed to the peak points in DSC curves respectively. Sci. Tech. Energetic Materials, Vol．７７,No．３,２０１６ 69

3.3 Catalysis mechanism because their rate constants were very small. So if we By means of DSC-IR analysis, we investigated introduced some strong acid (HA, #%"!$#"!#%"(HNO３)) 2- - ＋ decomposition products for [AP+3%SO4 /Fe２O３]and[AN into the decomposing AN, the reaction of NO３ +H ⇌ 2- +3%SO4 /Fe２O３]andtheresults were showed in Figure 5. HNO３ would happen due to the principle of chemical 2- For [AP+3%SO4 /Fe２O３], we interceptedtheIRspectra replacement of weak acid by strong acid. This would ο ο ο ο ο at 221.8 C, 296.1 C, 341.5 C, 346.8 C, and 351.9 C, considerably increase the concentration of HNO３ in the respectively. From Figure 5 (b), it was found that the decomposition system, which resulted in a remarkable decomposition products were NO２, NOCl, N２O, HClO４, NO, promotion on the keyreactions (i.e. Equations 5~7).

HCl, and H２O. In particular, the peak intensity of NO２, Therefore, the decomposition of AN was accelerated.

NOCl, and N２Oweremuchstronger than that of others. In Especially, when #%"!$#"!#%" ( HNO３ ), the 25） Reference ,Cl２ and O２ usually formed in decomposition of decomposition of HNO３ would be changed (see Equation ４ AP. However, these non-polar molecules can not be 11). ０ detected by IR technology. Thus, we also considered there ９ NH４NO３→NH３+HNO３+2.18kJ/g (4) were some Cl２ and O２ generated. According to the ＋ - 2- 2HNO３⇌[NO２ NO３ ]+H２O(5) products, the decomposition reactions of [AP+3%SO4 / - ＋ - [NO２+NO3 ]NO２ +NO３ (6) Fe２O３]werededuced (i.e. Equations 2 and 3). Obviously, ＋ ＋ ＋ NO２ +NH３⇌[NH３·NO２ ]→N２O+H３O (7) despite existence of some HClO４,NH３ were not detected in HNO３→·OH+NO２ (8) the products. It meant that the oxidation of NH３ proceeded NH３+·OH→·NH２+H２O(9) very completely. We inferred that the generated NH３ gas 2- NH２+NO２→[NH２NO２]→N２O+H２O(10) adsorbed on the surface of AP and nanometer SO4 / ＋ ＋ ＋ 2- HNO３+H ⇌[H２ONO２] →NO２ +H２O(11) Fe２O３.Especially, when #%"(SO4 /Fe２O３)wasmuchlower ＋ ＋ than #%"(NH３), the reaction of NH３+H ⇌NH４ occurred The “strong acids” are several kinds of solid acids whose

(The parameter “#!"”, which can be used to quantify the acidity are much higher than that of 100% H２SO４. 26）,27） strength of an acid, is called acidity coefficient. Each acid Generally, their #!" are less than -11.93 .Inindustry, has its own “#!"”value.).This reaction meant that a part many organic synthesis reactions must use acid as of NH３ could be fixed as condense phase on the surface of catalyst. Comparing with liquid acid, solid acids are of 2- nanometer SO4 /Fe２O３.Thisbenefited to the oxidation of advantages such as high catalysis, high selectivity,

NH３ because it avoided the cessation of AP decomposition pollution-free, as well as easy to separate from reaction by NH３ poisoning. Hence, the decomposition of [AP+3% system. Thus, they have a bright application prospect in 2- SO4 /Fe２O３]onlyshowed one exothermic peak in DSC organic industry. Mainly, there are four kinds of solid acid trace. that are classified as strong acid. Kind 1 is the supported

heteropoly acids, such as HF-SbF５-AlF３/Al２O３, SbP３-Pt/ NH４ClO４⇌NH３(s)+HClO４(s)⇌NH３(g)+HClO４(g) (2) Graphite, SbP３-HF/F-A12O３, SbF５-FSO３H/Graphite, and so 5NH３+5HClO４→NO２+N２O+NO+2HCl+NOCl+Cl２+3O２ on. Kind 2 is the mixture of inorganic salts, such as AlCl３- +9H２O(3) CuCl２,MCl３-Ti２(SO４)３,A1C13-Fe２(SO４)３,andsoon.Kind3is 2- For [AN+3%SO4 /Fe２O３], we NO 2- (a)0.07 AP+3%SO 2-/Fe O 346.8 (b)0.5 2 AP+3%SO /FeO intercepted the IR spectra at 4 2 3 4 2 3 ο ο ο ο 0.06 161.5 C, 215.7 C, 238.8 C, 264.9 C, 0.4 N O ο ο 0.05 2 286.3 C, and 345.7 C, respectively (in NOCl [-] 0.04 [-] 0.3 Figure 5 (b)). Figure 5 (d) revealed o HClO 0.03 351.9 C 4 Abs Abs

341.5 Abs HCl H O that the decomposition products of 0.2 NO 2 0.02 351.9 346.8oC 2- [AN+3%SO4 /Fe２O３]wereN２O 296.1 341.5oC 0.01 0.1 o and seldom H２O. This meant the 221.8 296.1 C 0.00 221.8oC decomposition proceeded very 0.0 200 250 300 350 400 450 500 1000 1500 2000 2500 3000 3500 4000 o completely. In theory, Temperature [ C] Wavenumber [cm-1] decomposition of AN complies with 2- 0.35 (c) 0.025 AN+3%SO /Fe O 264.9 4 2 3 (d) 2- N O ionic reactions (Equations 4~7) or AN+3%SO /Fe O 2 0.30 4 2 3 radical reactions (Equations 8~10). 0.020 238.8 0.25

The gas product of both the [-] 286.3 o 0.015 345.7 C [-] 0.20 N2O H O o channels was N２O, which consisted 2 286.3 C Abs Abs 215.7 0.15 0.010 o with our experimental results. Abs 264.9 C

0.10 o 238.8 C Distinctly, at low temperature, 0.005 345.7 161.5 thermal decomposition of AN would 0.05 215.7 oC 0.000 o 0.00 161.5 C comply with ionic reactions since 100 150 200250300350 400 Temperature [ oC] 500 1000 1500 2000 2500 3000 3500 4000 the rupture of O-N bond (i.e. Wavenumber [cm-1] Equation 8 would happen only at 2- ο Figure５ IR spectra of decomposition products : (a, b) [AP+3%SO4 /Fe２O３]; (c,d)[AN+ temperature more than 1300 C. In 2- 3%SO4 /Fe２O３]; (a,c)are total absorbance of gas products ; (b, d) are IR ionic mechanism,thereaction of spectra of gas products intercepted at different temperature (the temperature Equation 5 was the limiting step nodes of interception were illustrated in Figure 5 (a) and (c), respectively.). 70 Xiaolan Song et al.

lose from the surface. Thus, the H２Oincondense phase 2- could obviously weaken the catalysis of nanometer SO4 /

Fe２O３.Infact,Sun et al had reported that some inorganic acids such as concentrated sulfuric acid and concentrated hydrochloric acid could promote the decomposition of AN ; andthe catalysis action of hydrochloric acid were Figure６ Proposed surface structural model and strong-acid distinctly higher than that of sulfuric acid26）.Thecatalysis 2- species on the surface of SO4 /Fe２O３ acid. mechanism of inorganic acids had been elucidated above. But we can not added liquid acid into a propellant because 2- 2- acid metal oxide coated with SO4 ,suchasSO4 /ZrO２, the processing property of the propellant will be 2- 2- SO4 /TiO２,SO4 /Fe２O３,etc.Kind4is composition of deteriorated. Hence, using solid acid may be a better metal oxides, such as WO３/ZrO２,MoO３/ZrO２,etc.Allthe choice. Moreover, in terms of acidity, the #+" of solid solid strong acids show their#!" are less than -10 at lest. strong acid (~-14) was far less than #+" of sulfuric acid However, the reason why they process of so strong acidity (#+""!$!!)and hydrochloric acid ( #+""!)!!). Thus, is obscure. Meanwhile, the different kind of strong acid solid strong acid may show higher catalysis action than may have different reason. In my study, we pay more those liquid acids. Meanwhile, the processing property of attention to the strong acid of kind 3, because it can be propellants will not be affected by adding little solid strong easily prepared and their particle sizes are easily acid. 2- controlled in nanometer scale. According to the studies of Overall, the catalysis action of SO4 /Fe２O３ Guo28）and Wang29）,theyconsidered that due to the strong nanoparticles was not as excellent as expectation. inductive effects of S=O group, the adsorbed H２O According to the reported studies, its catalysis ability contributes as Brφnsted acid center. The surface acid sites should be lower than metal nanoparticles (especial nano are associated with the metal ions whose acidic strength Cu) but higher than traditional metal oxides. However, can be strongly enhanced by induction effect of S=O owing to the very low cost in fabrication and storage, groups (please see Figure 6). using solid strong acids as catalysts is worthy to be In addition, solid strong acids do not like traditional attempted. Moreover, there are so many kinds of solid acids that can form a mass of H＋ ions when they dissolve strong acid that could exhibit different catalysis ability. into water. For solid strong acid, Brφnsted acidcenters are Hence, this study is just a beginning of these. only existent on their surface. The centers work only if the molecular of reactant(s) absorbed on the surface of the 4. Conclusions 2- catalyst. Thus, this kind of catalysis is not homogeneous SO4 /Fe２O３ nanoparticles were prepared by catalysis but rather heterogeneous catalysis. Of course, precipitation-dipping method. XRD analysis indicated that the acid groups on the surface of solid strong acid can be the amorphous Fe２O３ transformed to "-Fe２O３ after ο “apperceived” by the reactions with some chemical agents. calcined at 500 C. By using JADE5.0 software, the average 2- For example, a suitable method for determining the acid grain size of SO4 /Fe２O３ was calculated. SEM images 2- strength of solid acid may be “Steam Method” reported by showed that the particle size of SO4 /Fe２O３ was 30-40 nm. 26） 2- Li .InLi’s study, the agents, such as m-nitrotoluene XPS analysis presented that after acidification, the SO4

(#!" "!""!**), p-nitrochlorobenzene (#!" "!"#!(!), m- was fixed well on the surface of Fe２O３. nitrochlorobenzene (#!" "!"$!'! ), and Thermal analyses were performed to probe thermal 2- dinitrofluorobenzene (#!" "!"%!&#), were used as decomposition of [AP+3%Fe２O３], [AP+3%SO4 /Fe２O３], 2- indicators. [AN+3%Fe２O３], and [AN+3%SO4 /Fe２O３]. The results 2- In fact, for solid strong acid, its preparation method and indicated that SO4 /Fe２O３ presented highercatalysis than the acid strength (very high) had become a common nano Fe２O３.DSC-IRanalyses were employed to detect the 2- recognition early. Thus, this paper took more efforts on gas products for thermal decomposition of [AP+3%SO4 / 2- thecatalysis ability and catalysis mechanism of solid Fe２O３]and[AN+3%SO4 /Fe２O３]. Via analyzing the IR strong acid on thermal decomposition of AP and AN. As spectra of gas products, we found that the gases such as for the property and the function of these acid groups (on NO２, NOCl, N２O, HClO４, NO, HCl, and H２Ogenerated in 2- 2- the surface of solid acid particles), I think these had been decomposition of [AP+3%SO4 /Fe２O３]. For [AN+3%SO4 clearly elucidated in many references (just like /Fe２O３], the main gas products were N２OandseldomH２O, Reference26）-29）). which meant that the decomposition reaction proceeded In fact, plenty of Lewis acid points and Bronsted acid very completely. According to the detected products, the 2- points were co-existent on the surface of nanometer SO4 possible decomposition mechanism of AP and AN were 2- /Fe２O３.However, what really matters was not Lewis acid derived. Meanwhile, the catalysis actions of SO4 /Fe２O３ but Bronsted acid because pure nano Fe２O３ did not were discussed in detail. present some catalysis action (there are so many Lewis acid points on the surface of nano Fe２O３). The liquid H２O, Acknowledgement which generated from decomposition of AN, could enable This research was supported by the National Natural the transformation (Lewis acid transforms to Bronsted Science Foundation of China (Grant No. : 51206081; 2- acid). However, the liquid H２Ocouldalso make the SO4 Recipient: Yi Wang). Secondauthor (Yi Wang) paid the Sci. Tech. Energetic Materials, Vol．７７,No．３,２０１６ 71 same effort andcontribution to this paper as the first 14) T. Naya and M. Kohga, Propellants Explos. Pyrotech., 38, 87 author (Xiaolan Song). -94 (2013). 15) V. N. Popok and N. V. Bychin, Nanotechnol. Russ., 9, 541- References 548 (2014). 1) Oide Sho, Takahashi Kenichi, and Kuwahara Takuo, Sci. 16) A. A. Vargeese, S. J. Mija, and K. Muralidharan, J. Energ. Tech. Energetic Materials, 73, 153-156 (2012). Mater., 32, 146-161 (2014). 2) Kohga Makoto, Sci. Tech. Energetic Materials, 71, 145-150 17) K. Hasue, J. Energ. Mater., 32, 199-206 (2014). (2010). 18) V. P.Sinditskii and V. Y. Egorshev, Propellants Explos. 3) A. P. Sanoop, R. Rajeev, and B. K. George, Thermochim. Pyrotech., 30, 269-280 (2005). Acta, 606, 34-40 (2015). 19) Y. Miyata and K. Hasue, J. Energ. Mater., 29, 344-359 4) Y. H. Wang, T. Xu, and L. L. Liu, Sci. Tech. Energetic (2011). ４ Materials, 75, 21-27 (2014). 20) V. P.Sinditskii and V. Y. Egorshev, Combust. Explo. Shock ０ 5) Kakami Akira, Terashita Shota, and Tachibana Takeshi, Waves, 48, 81-99 (2012). ９ Sci. Tech. Energetic Materials, 70, 145-151 (2009). 21) T. Naya and M. Kohga, J. Energ. Mater., 33, 73-90 (2015). 6) S. Singh, G. Singh, and N. Kulkarni, J. Therm. Anal. Calorim., 22) M. Kohga and K. Okamoto, Combust. Flame, 158, 573-582 119, 309-317 (2015). (2011). 7) Fujimura Kaori and Miyake, Sci. Tech. Energetic Materials, 23) S. Chaturvedi and P. N. Dave, J. Energ. Mater., 31, 1-26 69, 149-154 (2008). (2013). 8) K. Gao, G.P. Li, and Y.J. Luo, J. Therm. Anal. Calorim., 118, 24) R. Gunawan and D. K. Zhang, J. Hazard. Mater., 165, 751- 43-49 (2014). 758 (2009). 9) Fujimura Kaori and Miyake Atsumi, Sci. Tech. Energetic 25) P. W. M. Jacobs and G. S. Pearson, Combust. Flame., 13, 419 Materials, 71, 65-69 (2010). -430 (1969). 10) G. Z. Hao and J. Liu, Propellants Explos. Pyrotech., 40, 848- 26) Q. Y. Li, Petrochemical Technol., 34, 75-77 (2005). 853 (2015). 27) R. R. Zhan, Chin. Rare Earths, 3, 84-88 (2009). 11) Ohtake Toshiakim, Date Shingo, and Ikeda Ken, Sci. Tech. 28) H. F. Guo, Mater. Chemist. Phys., 112, 1065-1068 (2008). Energetic Materials, 76, 1-2 (2015). 29) Y. Wang, J. Mater. Sci., 44, 6736-6740 (2009). 12) Hasue Kazuo, Sci. Tech. Energetic Materials, 75, 5-6 (2014). 30) J. H. Sun, Z. H. Sun, and Q. S. Wang, J. Hazard. Mater., 127, 13) C. Oommen and S. R. Jain, J. Hazard. Mater., A67, 253-281 204-210 (2005). (1999).