Research Article
Mechanism investigation for remarkable decreases in sensitivities from micron to nano nitroamine
Nanomaterials and Nanotechnology Volume 6: 1–10 ª The Author(s) 2016 DOI: 10.1177/1847980416663678 nax.sagepub.com
Yi Wang1, Xiaolan Song2, Dan Song3, Chongwei An2, Jingyu Wang2, and Fengsheng Li4
Abstract Raw hexahydro-1,3,5-trinitro-1,3,5-triazine (RDX) was pulverized to nano RDX by mechanical milling, and their micron morphology and surface elements were probed by transmission electron microscope and X-ray photoelectron spectroscopy analyses. Thermal analysis was employed to take a kinetic evaluation on thermal decomposition of raw and nano RDX. The result indicated that activation energy for thermal decomposition of nano RDX is closed to the value of raw RDX, which means nano RDX had similar thermal reactivity as raw RDX. However, the sensitivity tests showed that when raw RDX was pulverized to nanoparticles, its mechanical and shock sensitivities decreased by more than 45%. Since it was impossible to use kinetic evaluation to explain the reason why the difference on sensitivities between raw and nano RDX was so distinct, we recruited classic detonation models to solve the problem. By combining the models of Khasainov’s and Merzhanov’s, we related the detonation parameters such as temperature of hot spots, critical temperature of hot spots (TC), critical size of hot spots (δ C), and mean size of explosive particles, and concluded that: (a) under the same condition, mean size of hot spot in nano RDX charge was much smaller than that of raw RDX charge; (b) at the same δ C, TC of nano RDX (776 K) was higher than that of raw RDX (459 K); and (c) particle size was not an important factor to affect sensitivities of explosives unless size of explosive particles was less than 400 nm. These results must base on a steady thermal reactivity from micron to nano RDX. Keywords Nano explosives, sensitivities, kinetics, detonation models, mechanism Date received: 26 February 2016; accepted: 29 June 2016 Topic: Nanophase, Nanostructured Materials and Nanoscale Character Topic Editor: Leander Tapfer
Introduction In the past decade, scholars paid much attention to researches about nanometer energetic materials. Nano explosives, such as nano hexahydro-1,3,5-trinitro-1,3,5triazine (RDX), 1–5 nano octahydro-1,3,5,7-tetranitro1,3,5,7-tetrazocine (HMX),6–11 nano 2,4,6,8,10,12-hexanitro-2,4,6,8,10,12-hexaazaisowurtzitane (HNIW),12,13 nano 2,2’, 4,4’, 6,6’-hexanitro-stilbene (HNS),14,15 nano 3-nitro1,2,4-triazol-5-one (NTO),16 nano 1,3,5-triamino-2,4,6-trinitrobenzene (TATB),17 nano nitrocellulose (NC),18 and so on, were fabricated and their performance was investigated. However, there is no one proclaimed that nano energetics was successfully used in any practical type of ammunition.
1
School of Materials Science and Engineering, North University of China, Taiyuan, People’s Republic of China 2 School of Chemical Engineering and Environment, North University of China, Taiyuan, People’s Republic of China 3 China Ordnance Institute of Science and Technology, Beijing, People’s Republic of China 4 School of Chemical Engineering, Nanjing University of Science and Technology, Nanjing, People’s Republic of China Corresponding authors: Yi Wang, School of Materials Science and Engineering, North University of China, Taiyuan 030051, People’s Republic of China. Email:
[email protected] Xiaolan Song, School of Chemical Engineering and Environment, North University of China, Taiyuan 030051, People’s Republic of China. Email:
[email protected]
Creative Commons CC-BY: This article is distributed under the terms of the Creative Commons Attribution 3.0 License (http://www.creativecommons.org/licenses/by/3.0/) which permits any use, reproduction and distribution of the work without further permission provided the original work is attributed as specified on the SAGE and Open Access pages (https://us.sagepub.com/en-us/nam/ open-access-at-sage).
2 Why does this work? The most important reason is the inherent difficulty in scale-up fabrication of nano energetics. Meanwhile, another question arises. Since the considerable hindrance in fabrication, why did scientists pursue them relentlessly? It is generally thought that nanometer energetic materials should be easily ignited, decompose rapidly, give a violent reaction, and so on. For example, nanometer thermites increase in thermal reactivity as their particle size falls into the nanometer scale range.19–21 For this, someone may say that thermite is a kind of composite energetics, whose ingredients consist of separate fuel and oxidizer. Using nanometer materials as fuel and oxidizer would result in a remarkable increase of interface area between fuel and oxidizer particles, which greatly promoted the diffusion process in the redox reaction. However, for monomolecular energetics, the situation is changed. Just like nitroamine explosives, their thermolysis begins with the activation and rupture of the weakest bond (N–NO2) in their molecular structure. This activation and rupture are independent of particle size, despite the nanometer scale. Thus, in theory, nano explosives will possess the similar thermal reactivity as micron explosives, such as closed values of activation energy and onset temperature of thermal decomposition. Now, the same question remains: why did we take more effort on preparation of nano explosives? The answer lies in much higher safety of nano explosives than micron ones. At present, development of insensitive ammunition was becoming the ‘‘protagonist’’ in fields of military science and technology. Many studies had confirmed that nano explosives are more insensitive than their micron counterparts.22–25 In the aspect of safety, it becomes an advantage that nano explosives show similar thermal reactivity as micron ones. However, even so, it is necessary to disclose which factor leads to a remarkable decrease in sensitivities when particle size of explosives falls into nanometer scale. Herein, we would employ classic detonation models incorporated our experimental data to clarify this mechanism.
Experiment In this study, raw RDX was pulverized by a mechanical milling method.23,24 Fabrication process was very easy. Some RDX was added into the mill, and then the ball and the solvent were also introduced into the mill. The comminuting course lasted 20 h. Then the ball and the material were brought out. After filtrating and drying, nano RDX powder was obtained. The solvent mixture consists of H2O (50 wt%) and ethanol (50 wt%). The particle morphology was examined by a PhilipsTecnai-12 (Philips in Oregon, U.S.) transmission electron microscope. The purity of comminuted samples was characterized by X-ray photoelectron spectroscopy (XPS), PHI5000 Versa-Probe (ULVAC-PHI, Japan). Thermal analysis was performed on a differential scanning calorimeter
Nanomaterials and Nanotechnology (TA Model Q600, North University of China) at heating rates of 5, 10, 15, and 20 C min1 (nitrogen atmosphere, sample mass of approximately 5 mg, and aluminum oxide crucible), respectively. An HGZ-1 (North University of China) impact instrument was used to test the impact sensitivity of explosives. The special height (H50) value represents the height from which 2 kg (or 5 kg) drop hammer will result in an explosive event in 50% of the trials. A WM-1 (North University of China) friction instrument was employed to test the friction sensitivity of the explosives and test standards (66 , 2.45 MPa and 80 , 2.45 MPa) were utilized. In each test, an explosion probability (P, %) was obtained. The shock sensitivity of the explosives was measured by the small scale gap test. The density of the donor columns (RDX) was 1.48 + 0.01 g cm3. The acceptor columns were loaded with a density of 90% theoretical maximum density. The valid gap thickness (GT) was calculated. No binder was used in all small scale gap tests.
Results Sample characterization Micron morphology of raw and nano RDX was presented in Figure 1(a) and (b). The particle size of raw RDX is very large (30–80 mm). After pulverized, its particle size fell into nanoscale (approximately 50 nm). We carefully measured the sizes of approximately 150 particles in Figure 1(a) and (b) and calculated its particle size distribution and the mean particle size (X). The result indicated that the mean size RDX was very small (see Figure 1(d)). The surface element of the milled samples was detected by XPS analysis (see Figure 1(e)). There are peaks associated with O1s, C1s, and N1s. O1s peak relates to electron excitation in 1s orbit of O atom of –NO2 radical. C1s peak is corresponding to electron excitation in 1s orbit of C atom of CH2 group. The excitation of the N1s electron results in two peaks that represent ammonia nitrogen (N–NO2) and nitrate nitrogen (N–NO2) in this nitroamine, respectively. Figure 1(e) indicated that RDX powder was not contaminated by mechanical milling. In Figure 1(f), it was affirmed that X-ray diffractometry (XRD) pattern of nano RDX is almost as same as that of micro RDX. Two XRD patterns implied that the milled sample was of the same crystal phase as raw RDX.
Kinetic evaluation Thermal analysis is very important to estimate whether thermal reactivity of nano RDX is much higher than that of raw RDX. If thermolysis of nano RDX possesses of very lower activation energy, it is much easier to be ignited by hot spots under the same condition. This will be adverse to ensuring insensitive property of nano RDX. Herein, we performed the kinetic evaluation by a differential isoconversion method (Friedman–Reich–Levi Equation25).
Wang et al.
3
(c) 22 18
X = 32.8 nm
21
16 14 12 10 8
18 15 12
6
9 6
4
3
2 0
nano RDX
24
Frequency (%)
Frequency (%)
(d) 27
raw RDX X = 94.5 µm
20
0
25
0
75 100 125 150 175 200 225 250 275
50
5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85
Particle diameter (µm)
(e) 24,000
Particle diameter (nm)
(f)
O1s
16,000
21,000
Intensity (a.u.)
CPS
15,000 O KLL
9000
N1s C1s
12,000 10,000 8000 6000
6000
4000
3000
2000 0
0 1080
raw RDX nano RDX
14,000
18,000
12,000
18,000
960
840
720
600
480
360
240
120
Binding energy (eV)
0
5
10
15
20
25
30
35
40
45
50
55
2θ (°)
Figure 1. Characterization of samples: (a) microscopic photograph of raw RDX; (b) TEM image of nano RDX; (c) and (d) particle size distribution of raw and nano RDX calculated with Figure 1(a) and (b), respectively; (e) XPS spectrum of nano RDX; (f) XRD patterns of raw and nano RDX. RDX: hexahydro-1,3,5-trinitro-1,3,5-triazine; TEM: transmission electron microscope; XPS: X-ray photoelectron spectroscopy.
Thermal Gravity (TG) curves (in Figure 2(a) and (b)), collected at different heating rates, were transformed to plot of conversion (α) to temperature (T) (see Figure 2(c) and (d)). Based on Arrhenius equation (equation (1)), it can be derived that there is a linear relationship between ln[β(dα/dT)] and 1/T at certain conversion (equation
(2)).26 Thus, we took a derivative with α-T curves and then calculated the value of ln[β(dα/dT)] at different α value. Every α value was corresponding to certain temperature. Consequently, the linear relationship between ln[β(dα/ dT)] and 1/T was fitted and depicted in Figure 2(e) and (f). Then the values of activation energy (E) at each α value
4
Nanomaterials and Nanotechnology
(b)
100
5 °C·min–1
80 15 °C·min–1
Weight loss (%)
100 90 80 70 60 50 40 30 20 10 0
Weight loss (%)
(a)
10 °C·min–1 20 °C·min–1
5 °C·min–1
60
15 °C·min–1 10 °C·min–1
40 20
nano RDX
raw RDX
0 60 90 120 150 180 210 240 270 300 330
60 90 120 150 180 210 240 270 300 330
Temperature (°C)
(c) 1.0
1.0 0.9 0.8
10 °C·min–1
5 °C·min–1
0.7 15 °C·min–1
0.6
α
α
Temperature (°C)
(d)
5 °C·min–1
0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0
10 °C·min–1
0.5 0.4
20 °C·min–1
15 °C·min–1
0.3 0.2
20 °C·min–1
0.1
raw RDX
nano RDX
0.0
480 485 490 495 500 505 510 515 520 525 530 535
550 552 554 556 558 560 562 564 566 568 570 572
Temperature (K)
Temperature (K)
(e) 0.0
(f) 0.0 raw RDX
–0.3
–0.9 –1.2 –1.5
nano RDX
–0.3
ln[β·(dα/dT)]
ln[β·(dα/dT)]
–0.6
–1.8
–0.6 –0.9 –1.2 –1.5 α = 0.95
–1.8
–2.1
α = 0.90
–2.1
α = 0.90
α = 0.95
–2.4
–2.4 1.89
1.92
1.95
1.98 –1
2.01
2.04
2.07
1.89
1.92
1.95
–1
1000·T (K )
2.04
2.07
–1
raw RDX nano RDX
–1
–1
E (kJ·mol )
140
E (kJ·mol )
130
110
90
2.01
1000·T (K )
(h)150
120
1.98 –1
(g) 140
100
20 °C·min–1
raw RDX nano RDX
80 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
α
130 120 110 100 0.0
0.1
0.2
0.3
α
0.4
0.5
0.6
0.7
Figure 2. Kinetic analyzis: (a) and (b) TG curves of raw and nano RDX; (c) and (d) plots of conversion (α) to temperature (T) derived from TG traces; (e) and (f) plots of ln[(βdα/dT)] to 1000/T for calculation of kinetic parameters with iso-conversional method; (g) and (h) activation energy calculated from Figure 2(e) and (f) as a function of α. RDX: hexahydro-1,3,5-trinitro-1,3,5-triazine.
Wang et al.
5
Table 1. Impact, friction, and shock sensitivities of raw and nano RDX. Impact sensitivity (H50, cm) Samples Raw RDX Nano RDX
Friction sensitivity (P, %)
Shock sensitivity
2 kg hammar
5 kg hammar
66 , 2.45 MPa
80 , 2.45 MPa
GTa (mm)
46.08 68.52
15.26 23.86
68.67 38.67
74.00 58.67
13.82–14.42 7.40–8.20
RDX: hexahydro-1,3,5-trinitro-1,3,5-triazine. a GT is the range of adequate gap thickness.
were obtained as the slope of lines, which were shown in Figure 2(g). dα A E ¼ exp f ðαÞ; (1) dT β RT dα E ; (2) ln β ¼ ln½A f ðαÞ dT RT where α is the conversion, T is the temperature (K), A is the pre-exponential factor (s1), R is the gas constant, β is the heating rate (Kmin 1 ), E is the activation energy (kJmol1), and f(α) is the mechanism function. From Figure 2(g), we can find that the E value for nano RDX changes seldom as α value increases from 0.05 to 0.65. When conversion is beyond 0.65, activation energy decreases distinctly. The situation of raw RDX is similar as that of nano RDX. In hotspot theory, it is recognized that a self-sustained decomposition with conversion less than 0.1 would cause an explosion of the charge.26 So we calculated the mean value of E at α ¼ 0.05–0.65, and the results were E nano ¼ 123 kJmol 1 and E raw ¼ 128 kJmol1 (see the insert figure in Figure 2(g)). The difference between E nano and E raw is only 5 kJmol1. This is within the measurement uncertainties. Hence, there is not enough reason to say nano RDX is easier to decompose than raw RDX, that is, at the aspect of thermal reactivity, nano RDX is as safe as raw RDX.
Mechanical and shock sensitivities To investigate the safety of nano RDX, sensitivities of raw and nano RDX were tested and the results were listed in Table 1. Impact sensitivity of nano RDX is lower than that of raw RDX by 50%. For friction sensitivity, the result is similar as that of impact sensitivity test, that is, explosion probability of nano RDX is distinctly lower than that of raw RDX. The highlight is the result of shock sensitivity test, which is indispensable to affirm whether an explosive is insensitive. It is obvious that, comparing with raw RDX, nano RDX is more insensitive to the action of shock waves because the valid GT decreased by 47%.
Discussions Concerning results of kinetic evaluation and sensitivity tests, we found that when particle size of RDX decreased from micron to nano scale, its sensitivities also decreased
Figure 3. RDX samples: (a) crystal defects, voids, and solvent pores in RDX crystal27 and (b) RDX particles with high surface complexity.30 RDX: hexahydro-1,3,5-trinitro-1,3,5-triazine.
clearly, however, its thermal reactivity did not change. Herein, we will discuss the effect of nanometer particle size on sensitivity of this nitroamine explosive. According to reported articles, we found that some factors could affect sensitivities of explosives (such as particles density and density distribution, crystal defects, micron morphology, as well as particle size and size distribution). In terms of particle density and density distribution, the density of an explosive crystal is certain. However, because of the existence of crystal defects, voids, and solvent pores, the density would change and show a normal distribution (see Figure 3(a)27). Low particle density and broaden density distribution would lead to high sensitivities because hot spot easily formed therein.27–29 When raw explosive particles were pulverized to nano particles, these crystal defects, voids, and solvent pores were eliminated, and the particle density was closed to its theory density of the crystal. The second factor is the micron morphology of explosive particles. Czerski affirmed that the particles with high surface complexity related to high sensitivities.30 Figure 3(b) shows the micron morphology of RDX with high surface complexity in Czerski’s study. There are many ‘‘horns’’ fixed on the particle surfaces. We knew that when explosive charge was undergone a shock wave, plastic deformation occurred. Czerski considered energy generated from plastic deformation would gather around the horns to promote the formation of hot spots. Figure 1(b) shows that this disadvantage does not exist on the surface of nano RDX. The third factor is particle size. In previous studies, we found that the mechanical sensitivity of nitroamine explosives was changed as their particle size decreased, however, these changes were very weak. 31,32 For shock sensitivity, Czerski’s study also indicated that sometimes
6
Nanomaterials and Nanotechnology
the critical gap of HMX sample with mean size of explosive particles (d50) ¼ 237 mm is smaller than that of sample with d50 ¼ 12 mm, but sometimes the critical gap of HMX sample with d50 ¼ 165 mm is also larger than that of sample with d50 ¼ 27 mm.31 There are no rules to follow. Thus, it seems that particle size is not a crucial factor on sensitivities. However, this statement is correct only under the condition that the explosive particles are of micron size. If the particle size fell into nanoscale, the situation changed and nanometer particle size became the most important factor on sensitivities. Before clarifying the mechanism of nano size, we must understand the basic hypothesis of detonation process for explosives that are subjecting to a stimulation of impact, frication, or shock wave. Now, for investigation of sensitivity mechanism, the ‘‘hotspot’’ theory is generally accepted. The mechanism for formation of hotspot was established by the model of voids collapse. In this model, it is considered that when explosive charge underwent an action of shock wave, the voids embedding in the charge would collapse. This would result in generation of many viscoplastic power and adiabatic compressions at the points where the holes collapse. The heat generated from viscoplastic power is much larger than the heat generated from adiabatic compressions. Hence, peer scholars recognized that the action of viscoplastic power dominates the main mechanism for generation of hotspots. Based on this, the model, which depicted the detonation process of explosive charges undergoing an action of shock wave, was established by Khasainov.33 This model had been accepted by peer scholars and was based on four root hypothesizes. 1. Each explosive particle is incompressible, that is, every particle is rigid body. 2. The voids embedding in charge are of spherical morphology. Meanwhile, deformation of the voids is spherical symmetric. 3. Ignore the melting on surfaces of the voids, that is, no phase transformation heat. 4. There are no interactions among voids in the course of adiabatic compression. Upon these, when viscoplastic deformation occurred (i.e. Ps >> Py þ Pg,0), the model of voids collapse was depicted as: drþ ¼ vþ ; dt
(3)
in which,
∂vþ vþ 3 1 2 Ps Pg Py þ 4μ s þ ρ s vþ : (4) ¼ ρ s rþ ∂t rþ 2
Using equation (4), we can solve rþ and vþ. According to the principles of fluid mechanics, under the condition of
spherical symmetry, the energy equation of viscous flow was obtained as: " 2 # 2 2μ s ∂T ∂T a ∂ ∂v v 2 ∂T þv ¼ r þ2 : þ ∂t ∂r r 2 ∂r ∂r ∂r r ρ s Cs (5) Because the explosive particles adjacent on the surface of voids is incompressible, the equation of v ¼ v þ r þ 2/r2 is obtained. Giving the integral of equation (5), equation (6) is obtained as: Z d 1 2 ρs Cs r ðT T 0 Þ dr ¼ 4μ s vþ 2 rþ þ rþ 2αðT g T0 Þ: dt rþ (6) In fact, the amount of air in voids is very small. Thus, the increase of surface temperature of voids is determined by energy diffusion rate caused by viscosity. Meanwhile, this increase is also determined by heat loss rate caused by heat conduction. This can be depicted by: 12μ s vþ 2 ∂T þ ∂2 T : ¼ a 2 jþ þ ∂r ∂t ρ s Cs rþ 2
(7)
Consequently, the relationship between Tþ and mean diameter of voids is established and illustrated as: 2
9 ðPs Py Pg;0 Þ Tþ ¼ Ta¼0 20 ρ s Cs μ s ( " 3 #) 2 2 2 I0 r0 8at t 1 1 2 2 : I0 r0 4a
(8)
Now, the model is obtained, but another question arises. In equation (8), it is clear that r0 is the crucial factor that determines the Tþ. As r0 increases, Tþ also increased. When Tþ arrives at its critical value (TC), explosive charge is of the maximum probability to explode. This Tþ value corresponds to a critical value of r0 (δ C). How to ascertain TC and δ C? And what is the relationship between TC and δ C? For this, Merzhanov considered that in the system containing chemical heat source, the rate of heat increase/decrease is equal to a sum of heat conduction rate and heat release rate of thermal decomposition.34 This can be depicted in equation (9). Upon equation (9), we can relate TC and δ C as equation (10) under the boundary condition: Ea ∂T 2 CE ρE ¼ λ E ∇ T þ QA exp ; (9) ∂t RT δ c ¼ 3:84Tc
λER ρE QAEa
1 2
exp
Ea Ea ðTc T 0 Þ 0:6 ; : ln 2RTc RTc2
(10) where δ C is the critical size of hot spots (m), TC is the critical temperature of hot spots (K), λ E is the thermal conductivity of explosives (Wm1K1), R is the gas constant, ρE is the density of explosive charge (gcm3), Q is
Wang et al.
7
Figure 4. Illustration of shock induced, hotspot formation and growth, in a porous, condensed-phase, explosive-based energetic formulation.
the heat release from the decomposition of explosive particles (J), A is the pre-exponential factor (s1); Ea is the activation energy (kJmol1), and T0 is the initial temperature of explosive charge (K). Other parameters and the derivation of above equations were stated in the studies by Khasainov and Merzhanov et al.33,34 Mean diameter of voids is immeasurable, but this parameter relates to the mean diameter (d50) of explosive particles. In Patom’s study, he related r0 and d50 in equation (11) under the premises that micron morphology of explosive particles is spherical and the compactness of explosive charge is very high.35 In equation (11), parameter Φ is charge porosity that can be calculated by charge density (Φ ¼ 1ρ/ρT, ρ is the charge density and ρT is the theoretical density of the charge): 1 6Φ 3 r 0 ¼ 0:5d50 : (11) π Combining equations (10) and (11), it affirms that larger d50 corresponds larger r0. Figure 4 describes the detonation process of explosive charge undergoing action of shock waves. First, it is supposed that there are many voids existing in a compacted charge. When it is subjecting to an action of shock, wave generated from the detonation of donor charge. Some voids are compacted and collapse. Due to the extreme short time of action, this compression is adiabatic. Another more important case lies in the generation of viscoplastic power at points where voids collapse.
Consequently, hotspots form, then they heat the adjacent explosive particles. When the temperature achieve and exceed the decomposition temperature of this explosive, the particles decompose and the excess heat is released. If the heat is sufficient to heat other particles to decompose, the valid heated layers form and the decomposition becomes self-sustained. In this case, the explosive charge may explode. In this course, one factor, which determines whether the charge will explode, is the thermal reactivity of the explosive. For example, there are two kinds of explosives: A and B. Explosive A has almost same particle size as explosive B. According to above-mentioned model, the size and Tþ in charge A should be closed to that in charge B under the same condition. Here if the activation energy for thermal decomposition of explosive A is much lower than that of explosive B, explosive A decomposes easily, and its decomposition will transform to detonation with more possibility. In our study, thermal decomposition of raw and nano RDX is of almost same activation energy, that is, nano RDX is not easier to decompose than raw RDX. But why did nano RDX show much lower sensitivities than raw RDX? The focus consists of the size ant Tþ. According to equation (11), it is very clear that mean size of void in charge composed of nano RDX is much smaller than that of raw RDX. Subjected to the same action, there are much more hotspots formed in raw RDX charge of which the sizes exceed the critical value (δ C). Of course, the straighter
8
Nanomaterials and Nanotechnology
1.5
×10
(b)
–8
Critical temperature (K)
–1.5
dTC /d C
1200
2250
400nm
0.0
–3.0 –4.5 –6.0 –7.5
2000 1750 1500 1250 1000
raw RDX nano-RDX
1100 1000 900 800
750
700
–8
0.0
–8
2.0 × 10
500
–8
4.0 × 10
6.0 × 10
–8
8.0 × 10
Critical diameter (m)
250
–9.0 0.0
2500
Critical temperature (K)
(a)
–6
2.0 × 10
–6
4.0 × 10
–6
6.0 × 10
–6
8.0 × 10
0
–5
0.000
1.0 × 10
0.005
δ C × 10 (m)
0.015
0.010
0.020
Critical diameter (m)
Figure 5. Results derived from Merzhanov’s model: (a) relationship between dTC/dδ C and δ C and (b) relationship between TC and δ C. TC: critical hotspot temperature; δ C: critical hotspot diameter.
–5
–1.50 × 10
(a)
Critical diameter (δC, m)
–5
–5
–5
–5
–5
–5
0.00 1.50 × 10 3.00 × 10 4.50 × 10 6.00 × 10 7.50 × 10 9.00 × 10
4000 (b) 3200 2400
Hot-spots temperature (T +, K)
4000 3200 2400 P s =1.0 GPa P s =2.0 GPa P s =3.0 GPa P s =4.0 GPa P s =3.16 GPa P s =3.33 GPa P s =3.4 GPa
1000 900 800 700
1000 900 800 700
600
600 r
r
r
1.0
2.0
4.0
+
++3.0 +
500
r
500
400
400
raw RDX
300
300 –8
2.0 × 10
0.0
0.0
1.0 × 10
4.0 × 10
–8
6.0 × 10
–8
Time ( t, s)
8.0 × 10
–8
1.0 × 10
Critical diameter (δ C, m)
–7
2.0 × 10
(c) 4000
–7
3.0 × 10
–7
4.0 × 10
–7
5.0 × 10
–7
–7
6.0 × 10
800
4.0
n
n
3.0
+ +
900
P s =1.0 GPa P s =2.0 GPa P s =3.0 GPa P s =4.0 GPa P s =6.62 GPa P s =7.11 GPa P s =7.28 GPa
2.0
+
700 600 500 400
800 700 600 500 400
nano RDX
300
Critical temperature ( T C, K)
900 n
–7
4000 (d) 3200 2400
Time ( t, s)
3200 2400
Hot-spots temperature (T+, K)
Critical temperature (T C, K)
factor is the critical temperature (TC). The relationship between δ C and TC has been established in equation (10). After substituting correlative parameters of raw and nano RDX into equation (10), we got the plot of δ C to TC. Then we take a derivative of the plot with the aspect of δ C and the plot of dTC/dδ C to δ C was presented in Figure 5(a). It indicates that when d50 is larger than 400 nm, there is no obvious change for value of dTC/dδ C. Once the mean size decreases to less than 400 nm, then the value of dTC/dδ C also decreases linearly. This means that 400 nm is a critical point. The change rate of TC is independent on δ C until δ C is less than 400 nm. This accounted for the experimental results of Czerski’s. In addition, the plot of δ C to TC was illustrated in Figure 5(b). It disclosed that at certain δ C, TC of nano RDX is higher than that of raw RDX. In voids collapse model, it is acquiescent that the δ C is somewhat smaller than mean size of voids (r0). Thus, in critical condition, we supposed that r0 was equal to δ C. Then the relationship among Tþ, t, and Ps is depicted in Figure 6, in which t is the duration time of the action and Ps is the pressure of the shock wave. The coordinates at the cross points where plot (TCδ C) intersects with plot (Tþt) are the critical values (i.e. TC, tC, and δ C). Under different Ps, TC values of raw and nano RDX are different because of the remarkable difference between particle size of raw and nano RDX. The coordinate values were tabulated in Table 2. Table 2 indicates that TC of raw and nano RDX is 459 K and 776 K, respectively. This means that for raw RDX, the charge may explode when temperature is 459 K; for nano RDX, explosion will not happen unless the temperature increases to 776 K. Meanwhile, tC of nano RDX is somewhat smaller than tC of raw RDX, which means under the same pressure, heating nano RDX to explode will cost more time. In addition, according to Patom’s study, mean void size in raw RDX charge is much larger than the size in nano RDX charge. That is to say the sizes of hot spots in raw
300
0.0
–8
2.0 × 10
4.0 × 10
–8
6.0 × 10
Time ( t, s)
–8
8.0 × 10
–8
1.0 × 10
–7
Figure 6. Relationships among Ps, Tþ, duration time (t), δ C, and TC. (a) and (b) for raw RDX; (c) and (d) for nano RDX. RDX: hexahydro-1,3,5-trinitro-1,3,5-triazine; Ps: shock pressure; Tþ: hotspots temperature; δ C: critical hotspot diameter; TC: critical hotspot temperature.
Wang et al.
9
Table 2. Calculation values of Ps, TC, and t for raw and nano RDX. Raw RDX
Nano RDX
Points Ps (GPa) TC (K) tC (ns) Points Ps (GPa) TC (K) tC (ns) r
1.0 2.0 r 3.0 r 4.0 r 3.16 r 3.33 r 3.40 r
1.0 2.0 3.0 4.0 3.16 3.33 3.40
459.31 60.1 459.31 16.7 459.31 9.73 459.31 7.37 459.31 9.14 459.31 8.7 459.31 8.51
– 2.0 n 3.0 n 4.0 n 6.62 n 7.11 n 7.28 n
1.0 2.0 3.0 4.0 6.62 7.11 7.28
– – 776.01 50.3 776.01 21.6 776.01 11.9 776.01 8.09 776.01 7.54 776.01 7.37
RDX: hexahydro-1,3,5-trinitro-1,3,5-triazine; Ps: shock pressure; Tþ: hotspots temperature; TC: critical temperature; t: critical time.
22.8 raw RDX nano RDX
22.5 22.2
lnPs
21.9 21.6 21.3 21.0 20.7 20.4 –19.0
–18.5
–18.0
–17.5
–17.0
–16.5
lntC Figure 7. Plots of lnPs to lntC. Ps: shock pressure; TC: critical hotspot temperature.
RDX charge are considerably larger than that in nano RDX charge under the same condition. For an exaggerated example, the difference of power between a 500 C spark and a 500 C hot iron ball is enormous. Therefore, nano RDX presented very lower sensitivities in the tests (especial in shock sensitivity test). On the whole, Khasainov’s model is suitable to analyzing detonation process of RDX charge. The plot of lnPs to lntC is shown in Figure 7, we found that for the case of raw RDX, there is a good linear relationship between lnPs and lntC; for the case of nano RDX, there is a reluctant linear relationship between lnPs and lntC. To a certain extent, Figure 7 confirmed the accuracy of Khasainov’s model that was used to explain the difference of sensitivity between raw and nano explosives. But the lower linear constant also proposed, it needs some supplement to improve the model that was used to analyze detonation process of nano RDX charge.
Conclusions Based on the successful fabrication of nano RDX, this study paid more efforts to clarify the mechanism why nano
RDX possess of much lower sensitivities than micron RDX. Before this, we imaged the micron morphology of nano RDX and carefully performed the kinetic evaluation for thermal decomposition of raw and nano RDX; meanwhile, the mechanical and shock sensitivities of raw and nano RDX were also tested. As was expected, mechanical sensitivities of nano RDX were lower by 40% than that of raw RDX, and, in particular, shock sensitivity was decreased by about 50% when raw RDX was pulverized to nano RDX. However, upon the results of kinetic evaluation, it was unexpected that thermal reactivity of nano RDX was as similar as that of raw RDX. Since thermal analysis could not disclose the difference in sensitivities between raw and nano RDX, we employed classic theory of detonation physics to depict the process that explosive charge was undergoing an action of shock wave. By combining Khasainov’s and Merzhanov’s models, we got the relationship between δ C and TC. Upon the relationship, it was concluded that (a) under the same condition, mean size of hot spot in nano RDX charge was much smaller than that in raw RDX charge; (b) at the same δ C, TC of nano RDX (776 K) was higher than that of raw RDX (459 K); and (c) particle size was not an important factor to affect sensitivities of explosives unless size of explosive particles was less than 400 nm. These conclusions partially explained the results of sensitivity tests. Of course, this explanation is rational only if the thermal reactivity is steady to the change of particle size. For a kind of energetic material, if its thermal reactivity increases considerably when its particle size decreases to nano scale, its sensitivities must increase and any detonation model becomes unsuitable. Author contribution Authors Yi Wang and Xiaolan Song contributed equally to this work.
Declaration of conflicting interests The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding The author(s) disclosed receipt of the following financial support for the research, authorship and/or publication of this article: This research was supported by the National Natural Science Foundation of China (grant no.: 51206081). Second author (Xiaolan Song) paid the same effort and contribution to this article as the first author (Yi Wang).
Reference 1. Fathollahi M, Mohammadi B, and Mohammadi J. Kinetic investigation on thermal decomposition of hexahydro-1,3,5trinitro-1,3,5-triazine (rdx) nanoparticles. Fuel 2013; 104: 95–100. 2. Kumar R, Siril PF, and Soni P. Preparation of nano-RDX by evaporation assisted solvent antisolvent interaction. propel. Explos Pyrotech 2014; 39: 383–389.
10 3. Liu J, Jiang W, and Zeng JB. Effect of drying on particle size and sensitivities of nano hexahydro-1,3,5-trinitro-1,3,5-triazine. Defence Technol 2014; 10: 9–16. 4. Kenneth KK, Christopher GL, and James HA. Formation and characterization of nano-sized RDX particles produced using the RESS-AS process. Propel Explos Pyrotech 2012; 37: 699–706. 5. He B and Stepanov V, and Qiu H W. Production and characterization of composite nano-RDX by RESS co-precipitation. Propel Explos Pyrotech 2015; 40(5): 659–664. 6. Bayat Y, Pourmortazavi SM, and Iravani H. Statistical optimization of supercritical carbon dioxide antisolvent process for preparation of HMX nanoparticles. J Supercrit Fluid 2012; 72: 248–254. 7. Qiu HW, Stepanov V, and Chou TM. Single-step production and formulation of HMX nanocrystals. Powder Technol 2012; 226: 235–238. 8. Zhang YX, Liu DB, and Lv CX. Preparation and characterization of reticular nano-HMX. Propel Explos Pyrotech 2005; 30: 438–441. 9. An CW, Li HQ, and Guo WJ. Nano Cyclotetramethylene tetranitramine particles prepared by a green recrystallization process. Propel Explos Pyrotech 2014; 39: 701–706. 10. Liu J, Jiang W, and Li FS. Effect of drying conditions on the particle size, dispersion state, and mechanical sensitivities of nano HMX. Propel Explos Pyrotech 2014; 39: 30–39. 11. Risse B, Schnell F, and Spitzer D. Synthesis and desensitization of nano-β-HMX. Propel Explos Pyrotech 2014; 39: 397–401. 12. Liu J, Jiang W, and Yang Q. Study of nano-nitramine explosives: preparation, sensitivity and application. Defence Technol 2014; 10: 184–189. 13. Guo XD, Ouyang G, and Liu J. Massive preparation of reduced-sensitivity nano cl-20 and its characterization. J Energ Mater 2015; 33: 24–33. 14. Wang JY, Huang H, and Xu WZ. Prefilming twin-fluid nozzle assisted precipitation method for preparing nanocrystalline HNS and its characterization. J Hazard Mater 2009; 162: 842–847. 15. Huang H, Wang JY, and Xu WZ. Effect of habit modifiers on morphology and properties of nano-HNS explosive in prefilming twin-fluid nozzle-assisted precipitation. Propel Explos Pyrotech 2009; 34: 78–83. 16. Yang GC, Nie FD, and Li JS. Preparation and characterization of nano-NTO explosive. J Energ Mater 2007; 25: 35–47. 17. Yang GC, Nie FD, and Huang H. Preparation and characterization of nano-TATB explosive. Propel Explos Pyrotech 2006; 31: 390–394. 18. Sovizi MR, Hajimirsadeghi SS, and Naderizadeh B. Effect of particle size on thermal decomposition of nitrocellulose. J Hazard Mater 2009; 168: 1134–1139. 19. Wang Y, Song XL, Jiang W, et al. Mechanism investigation for thermite reactions of aluminum/iron-oxide nanocomposites based on residues analysis. T Nonferr Metal Soc 2014; 24: 263–270.
Nanomaterials and Nanotechnology 20. Comet M, Siegert B, and Pichot V. Reactive characterization of nanothermites: correlation structure/reactivity. J Therm Anal Calorim 2013; 111: 431–436. 21. Gibot P, Comet M, and Eichhorn A. Highly insensitive/ reactive thermite prepared from Cr2O3 nanoparticles. Propel Explos Pyrotech 2011; 36: 80–87. 22. Wang Y, Song XL and Song D. A versatile methodology using sol-gel, supercritical extraction, and etching to fabricate a nitroamine explosive: nanometer HNIW. J Energ Mater 2013; 31: 49–59. 23. Wang Y, Jiang W, and Song D. A feature on ensuring safety of superfine explosives: the similar thermolysis characteristics between micro and nano nitroamines. J Therm Anal Calorim 2013; 111: 85–92. 24. Wang Y, Jiang W and Song XL. Insensitive HMX (octahydro-1,3,5,7-tetranitro-1,3,5,7- tetrazocine) nanocrystals fabricated by high-yield, low-cost mechanical milling. Cent Eur J Energ Mater 2013; 10: 3–15. 25. Hu RZ. Thermal analysis kinetics. Beijing: Science Press of China, 2008. 26. Vyazovkin S, Burnham AK, and Criado JM. ICTAC kinetics committee recommendations for performing kinetic computations on thermal analysis data. Thermochim Acta 2011; 520: 1–19. 27. Borne L, Mory J, and Schlesser F. Reduced sensitivity RDX (RS-RDX) in pressed formulations: respective effects of intra-granular pores, extra-granular pores and pore sizes. Propel Explos Pyrotech 2008; 33: 37–43. 28. Borne L and Patedoye JC. Quantitative characterization of internal defects in RDX crystals. Propel Explos Pyrotech 1999; 24: 255–259. 29. Doherty RM. Relationship between RDX properties and sensitivity. Propel Explos Pyrotech 2008; 33: 4–13. 30. Czerski H.relationship between the morphology of granular cyclotrimethyene- trinitramine and its shock sensitivity. J Appl Phys 2007; 102: 1–8. 31. Song XL, Wang Y and An CW. Dependence of particle morphology and size on the mechanical sensitivity and thermal stability of octahydro-1,3,5,7-tetranitro- 1,3,5,7-tetrazocine. J Hazard Mater 2008; 159: 222–229. 32. Song XL and Li FS. Dependence of particle size and size distribution on mechanical sensitivity and thermal stability of hexahydro-1, 3, 5-trinitro-1, 3, 5-triazine. Defence Sci J 2009; 59: 37–42. 33. Khasainov BA. Two-phase visco-plastic model of shock initiation of detonation in high density pressed explosives. In: Proceedings of 7th International Symposium on Detonation, 1981, pp. 435–447. 34. Merzhanov AG, Barzikin VV, and Gontkovskaya VT. The problem of hot-spot thermal explosion. Doklady AN SSSR 1963; 148: 380–391. 35. Partom Y. A viod collapse model for shock initiation. In: Proceedings of the 7th International Symposium on Detonation, 1981, pp. 506–516.
