ARTICLE pubs.acs.org/IECR
Green Thermal Analysis Technology for Evaluating the Thermal Hazard of Di-tert-butyl Peroxide Jo-Ming Tseng Institute of Safety and Disaster Prevention Technology, Central Taiwan University of Science and Technology, 666 Buzih Road, Taichung, Taiwan 40601, ROC
Chun-Ping Lin* Department of Health and Nutrition Biotechnology, College of Health Science, Asia University, 500 Lioufeng Road, Wufeng, Taichung, Taiwan 41354, ROC ABSTRACT: Di-tert-butyl peroxide (DTBP) has been widely employed in chemical industries. Unfortunately, many serious explosions and fires have occurred during the manufacturing process, storage, and transportation of organic peroxides. This study investigated the thermokinetic parameters by isothermal-kinetic and nonisothermal-kinetic simulation, using differential scanning calorimetry (DSC) tests. Green thermal analysis technology was applied to assess the kinetic parameters, such as the kinetic model, frequency factor (ln k0), activation energy (Ea), reaction order, heat of reaction (ΔHd), etc. Comparisons of nonisothermal and isothermal kinetic model simulation led to a beneficial kinetic model of thermal decomposition to predict the thermal hazard of DTBP. Simulations in 0.5 L Dewar vessel and 25 kg barrel commercial package in liquid thermal explosion models were performed and compared to the results in the literature. From the results, we determined the DTBP of the optimal conditions to avoid violent runaway reactions during the storage and transportation. This study established the features of green thermal analysis technology that could be executed as a reduction of energy potential and storage conditions in view of loss prevention.
1. INTRODUCTION Organic peroxide, which has been widely employed in chemical industries, is used to manufacture polymer materials.13 Ditert-butyl peroxide (DTBP) is a commercial liquid organic peroxide that is to be transported stored under a limited temperature range.46 In terms of manufacturing and international management, many serious explosions and high ambient temperature occur because of the thermal decomposition.13 The goal of this study was to develop a green thermal analysis technology that can be applied to industrial manufacturing processes to avoid a reaction disaster. Green thermal analysis technology is a special method. It can via simple differential scanning calorimetry (DSC) tests and kinetic model simulation be applied in the evaluation of the thermal hazard of organic peroxides. DSC is a popular thermal analysis instrument worldwide that can, through a swift and effective procedure, obtain the thermal decomposition properties of materials. Traditional thermal hazard assessment of a chemical involves a complex method for determining the energy consumption and warming gas production process. Through green thermal analysis technology, the accurate kinetic estimate can acquire the precise thermal hazard parameters in preventing pollution, reducing energy, and protecting the environment. Comparisons of nonisothermal and isothermal-kinetic model simulation led to a beneficial kinetic model of thermal decomposition to predict the thermal hazard of DTBP. The chosen approach was to establish a green thermal analysis technology for the thermal decomposition that included the r 2011 American Chemical Society
kinetic parameters and thermal hazard properties,710 such as the frequency factor (ln k0), reaction order (n), activation energy (Ea), heat of decomposition (ΔHd), isothermal time to maximum rate (TMRiso), time to conversion limit (TCL), selfaccelerating decomposition temperature (SADT), control temperature (CT), emergency temperature (ET) and the critical temperature (TCR), and total energy release (TER) for a package containing DTBP. SADT is an important parameter for the safety management of reactive substances during storage and transportation.46 It is generally determined by one of four testing methods recommended in the UN orange book: the United States (US) SADT test, the adiabatic storage test, the isothermal storage test, and the heat accumulation storage test.1122 The most commonly used tests for organic peroxides are the UN and US SADT tests. The study was applied to simulate 0.5 L and 25 kg containers with the aim of developing a green thermal analysis technology to replace the traditional method for evaluating the thermokinetic parameters and predicting the thermal hazard of DTBP. The model may be applied to the optimal conditions to avoid violent runaway reactions during manufacturing, storage, and transportation. Received: April 5, 2011 Accepted: July 6, 2011 Revised: June 20, 2011 Published: July 06, 2011 9487
dx.doi.org/10.1021/ie200712n | Ind. Eng. Chem. Res. 2011, 50, 9487–9494
Industrial & Engineering Chemistry Research
ARTICLE
Table 1. Boundary Conditions for 0.5 L Dewar Vessel and 25 kg Barrel Packages size shell
boundary
package shape
radius (m)
height (m)
thickness (m)
conditions
0.5 L vessel
0.0286
0.18
0.00286
25 kg barrel
a
Yang et al.
0.20
0.3
0.015
initial χ (W m2 K1)
top/third kind
1.4567a
side/third kind
1.4567a
bottom/third kind
1.4567a
top/third kind
2.8386a
sides/third kind
2.8386a
bottom/third kind
2.8386a
temperature (°C) 20
20
21
Figure 1. DSC thermal curves of heat flow versus temperature for DTBP decomposition with scanning rates of 1, 2, 4, and 8 °C min1.
2. EXPERIMENTAL DETAILS AND METHOD
Figure 2. DSC thermal curves for DTBP decomposition with isothermal temperature at 140, 145, 150, and 155 °C.
Table 2. Results of DSC Tests of DTBP at Various Scanning Rates of 1, 2, 4, and 8 °C min1 a scanning rate (°C min1)
2.1. Samples. DTBP 99 mass %, which was purchased
directly from ACE Chemical Corp in Taiwan, was stored in a refrigerator at 4 °C. Experiments were carried out at various scanning rates of 1, 2, 4, and 8 °C min1 and holding at four isothermal conditions of 140, 145, 150, and 155 °C. 2.2. Differential Scanning Calorimetry (DSC). Temperatureprogrammed screening experiments were performed with DSC (TA Q20). The test cell was used to carry out the experiment for withstanding relatively high pressure to approximately 10 MPa. ASTM E698 was used to obtain thermal curves for calculating kinetic parameters. For better thermal equilibrium, the heating rate chosen for the temperatureprogrammed ramp was not to exceed 10 °C min1. Approximately 2 mg of the sample was used to acquire the experimental data. Nonisothermal tests of the scanning rate selected for the programmed temperature ramp were 1, 2, 4, and 8 °C min1. The range of temperature rise chosen was from 30300 °C for each experiment. Isothermal tests of the holding isothermal condition were several at 140, 145, 150, and 155 °C. 2.3. Liquid Thermal Hazard Simulation. The liquid thermal explosion model and the algorithms that were used have been previously described. 7,9 We used a 0.5 L Dewar vessel
1
2
4
8
sample mass (mg)
2.0
2.1
2.4
2.2
onset temperature, To (°C) peak of temperature, Tp (°C)
135.05 169.87
144.96 177.03
152.36 185.54
164.73 194.67
reaction heat, ΔHd (kJ kg1)
747
1,018
920
1,106
Standard deviation: temperature accuracy (0.1; temperature precision (0.05; calorimetric reproducibility (1%; sensitivity 1.0 uW.
a
Table 3. Results of DSC Tests of DTBP under Isothermal Conditions of 140, 145, 150, and 155 °Ca isothermal conditions (°C)
140
145
150
155
sample mass (mg)
2.7
2.5
2.3
2.4
ΔHd (J/g)
237.1
522.1
491.1
727.4
Standard deviation: temperature accuracy (0.1; temperature precision (0.05; calorimetric reproducibility (1%; sensitivity 1.0 uW.
a
and a 25 kg commercial barrel package, as the reactor sizes, to simulate the thermal hazard. The radius, width, height, and shell thickness and the reactors were as listed in Table 1. 9488
dx.doi.org/10.1021/ie200712n |Ind. Eng. Chem. Res. 2011, 50, 9487–9494
Industrial & Engineering Chemistry Research
ARTICLE
Table 4. Comparisons of the Thermokinetic Parameters for the Evaluation of nth Order and Autocatalytic Models under Nonisothermal Conditions sample kinetic model ln(k0)/ln(sec1) Ea (kJ mol1) reaction order (n)/nth reaction order (n1)/auto
1 nth order
2 autocatalytic
nth order
4 autocatalytic
nth order
8 autocatalytic
nth order
23.4788
42.8503
23.0873
16.5944
23.3209
24.1370
23.6450
10.9662
111.2818
173.7119
109.1452
84.6957
109.5693
112.3753
110.7794
65.8989
0.7398
1.9877
0.7054
0.7766
0.7241
1.0510
0.7005
0.4013
reaction order (n2)
N/A
5.6504
N/A
0.4851
N/A
1.5799
N/A
autocatalytic constant (z)
N/A
0.2280
N/A
0.2274
N/A
0.8706
N/A
ΔHd (kJ kg1)
autocatalytic
754.0774
734.2785
1046.8737
1040.2696
941.2732
924.8643
1117.9358
1.3407 1.5462 1154.1706
Table 5. Comparisons of the Thermokinetic Parameters for the Evaluation of nth Order and Autocatalytic Models under Isothermal Conditions sample kinetic model ln(k0)/ln(sec1) Ea (kJ mol1) reaction order (n)/nth
140 nth order
145
autocatalytic
nth order
150
autocatalytic
nth order
155
autocatalytic
nth order
autocatalytic
23.3938
16.2855
22.7686
21.8515
23.5018
21.6512
23.4352
35.3061
106.1112 1.1903
83.9496 1.7351
105.2945 1.1465
104.1035 2.0425
109.1066 1.0566
99.0756 1.2683
109.8128 0.8693
150.7834 1.8039
reaction order (n1)/auto reaction order (n2)
N/A
0.4335
N/A
0.5980
N/A
5.8567
N/A
autocatalytic constant (z)
N/A
1.4429
N/A
1.3642
N/A
0.4146
N/A
ΔHd (kJ kg1)
305.1110
334.1026
512.8460
611.7439
783.8132
759.0846
957.8832
0.2944 0.1642 1065.4857
Figure 3. DTBP heat production versus time curves of the nth order reaction with scanning rates of 1, 2, 4, and 8 °C min1 by experiment and simulation.
Figure 4. DTBP heat production rate versus time curves of the nth order reaction with scanning rates of 1, 2, 4, and 8 °C min1 by experiment and simulation.
3. RESULTS AND DISCUSSION
Single-stage for nth order reaction dR ¼ k0 eEa =RT ð1 RÞn dt
3.1. Determination of Thermokinetic Parameters by DSC.
Simulations of kinetic models can be complex multistage reactions that may consist of several independent, parallel, and consecutive stages:710 Simple single-stage reaction dR ¼ k0 eEa =RT f ðRÞ dt
ð1Þ
ð2Þ
Multistage for autocatalytic reaction f ðRÞ ¼ ð1 RÞn1 ðRn2 þ zÞ
ð3Þ
where Ea is the activation energy, k0 is the pre-exponential factor, z is the autocatalytic constant, and n1 and n2 are the reaction orders of a specific stage.710Reactions that include two 9489
dx.doi.org/10.1021/ie200712n |Ind. Eng. Chem. Res. 2011, 50, 9487–9494
Industrial & Engineering Chemistry Research
ARTICLE
Figure 5. DTBP heat production versus time curves of the autocatalytic reaction simulation with scanning rates of 1, 2, 4, and 8 °C min1 by experiment and simulation.
Figure 6. DTBP heat production rate versus time curves of the autocatalytic reaction simulation with scanning rates of 1, 2, 4, and 8 °C min1 by experiment and simulation.
consecutive stages are the following: dR dγ ¼ k1 eE1 =RT ð1 RÞn1 ; ¼ k2 eE2 =RT ðR γÞn2 dt dt
ð4Þ
where R and γ are the conversions of the reactant A and product C, respectively. E1 and E2 are the activation energies of the stages.710 Two parallel reactions for full autocatalysis are the following: dR ¼ r1 ðRÞr2 ðRÞ; dt
r1 ðRÞ ¼ k1 ðTÞð1 RÞn1 r2 ðRÞ ¼ k2 ðTÞRn2 ð1 RÞn3
ð5Þ
where r1 and r2 are the rates of each stage and n3 is the reaction order of stage three.710 The kinetic parameters were determined from the DSC experimental data at various scanning rates of 1, 2, 4, and 8 °C min1 and isothermal tests holding four isothermal conditions of 140, 145, 150, and 155 °C as displayed in Figures 1 and 2, respectively. The experimental results of nonisothermal and isothermal of DSC
Figure 7. DTBP heat production versus time curves of the experimental data and nth order reaction simulation with isothermal temperature at 140, 145, 150, and 155 °C.
tests are listed in Tables 2 and 3, respectively. We hypothesized that the thermal decomposition of DTBP represents an unknown reaction, such as an nth order or autocatalytic reaction. We used the nth order and autocatalytic simulations to calculate the thermokinetic parameters and, then, to compare the results of nonisothermal and isothermal of kinetic model simulation. The simulation results are presented in Tables 4 and 5. From comparisons of nonisothermal vs isothermal of kinetic model simulation, nth order reaction vs autocatalytic reaction, and literature data, we obtained better results for the thermal hazard parameters with the nth order of thermal decomposition. Meanwhile, comparisons of experimental data and data derived from simulated n-th order and autocatalytic reaction for heat production and heat production rate versus time are shown in Figures 310, respectively. In contrast to the fact that the use of simulated nth order kinetic models to match original DSC experimental data (from Figures 310) was proven to give superior results, not all of the data are compatible with the model. In addition, from comparisons of nth order thermokinetic parameters of nonisothermal and isothermal kinetic model simulation, the parameters can be matched very well to each other. This data set was excluded from further analysis. The analysis of thermokinetic parameters of the thermal decomposition of DTBP depended on the reliability of the kinetic model. This study applied the isothermal and nonisothermal kinetic model for the evaluation of thermokinetic parameters and compared the results to simulated thermal analysis. This approach led to the development of a precise and effective procedure for the evaluation of thermal decomposition properties of DTBP. In contrast to Tables 4 and 5, we could observe the results of an nth order reaction kinetic simulation and autocatalytic reaction kinetic simulation for DTBP, in which thermokinetic parameters were providing the disorderly and confusing results by autocatalytic reaction simulation. The result is explicit: The nth order reaction simulation is appropriately applied on DTBP’s thermokinetic parameter evaluation. Moreover, comparison of Tables 4 and 5 shows the samples were tested under the high isothermal conditions; the overheating effect was greater than nonisothermal 9490
dx.doi.org/10.1021/ie200712n |Ind. Eng. Chem. Res. 2011, 50, 9487–9494
Industrial & Engineering Chemistry Research
ARTICLE
Figure 8. DTBP heat production rate versus time curves of the experimental data and nth order reaction simulation with isothermal temperature at 140, 145, 150, and 155 °C.
DSC tests. Thus, the result of isothermal-kinetic-model simulation was concerned with overheating effect of the kinetic parameters they were excluded from further analysis. Here, the TMRiso, TCL, and TER of DTBP were acquired by simulating nth order of nonisothermal and isothermal-kinetic-model simulation, as displayed in Figures 1114. Figure 11 shows the TMRiso of DTBP that was obtained, which values were ca. less than 30 °C and exceeded the upper limit of 100 days. The TCL of DTBP was less than 30 °C, which is beyond the upper limit of one year. Especially, Figure 12 shows the TER of DTBP is stored in less than 20 °C, which is safe. All of the safety parameters were acquired by nth order nonisothermal kinetic model simulation. In contrast to Figures 12 and 14, we could observe the DTBP’s TER of nth order kinetic model simulation of nonisothermal and isothermal, in which TER was providing disorderly and confusing results of energy release by isothermal kinetic simulation. The result is explicit: the nth order reaction of isothermal kinetic model simulation cannot be appropriately applied on DTBP’s thermal hazard parameter evaluation. Figures 3 and 12 show that not all of the data are compatible with the model. The heat effect at the 1 °C min1 scanning rate is smaller than that observed at the other scanning rates. While analyzing the thermokinetic parameters by kinetic model simulation, we acquired three numbers for the nth order thermokinetic parameters by using scanning rates of 2, 4, and 8 °C min1 in the thermal hazard simulation. This data set was excluded from further analysis. While analyzing the thermokinetic parameters with comparing the kinetic model method, we acquired three numbers for the nth order thermokinetic parameters by nonisothermal DSC tests. In addition, we also added the literature data to simulate the thermal hazard in this study. 3.2. 0.5 L Dewar Vessel and 25 kg Barrel Package Thermal Hazard Simulations. To simulate the thermal hazard of liquid organic peroxide, the critical parameters for the thermal hazard were determined numerically from the chemical kinetics for several types of reactor geometries and various boundary conditions, including the possibility setoff setting boundary shells. For liquid thermal hazard simulations, the following
Figure 9. DTBP heat production versus time curves of the experimental data and autocatalytic simulation with isothermal temperature at 140, 145, 150, and 155 °C.
statements were used:7,9 FCP
∂T ¼ divðλΔTÞ þ W ∂t
Thermal conductivity equation
ð6Þ ∂Ri ¼ ri i ¼ 1, :::NC ∂t
Kinetic equations ðformal modelsÞ ð7Þ
W ¼
∑ðiÞ Qi∞ ri
Heat power equation
ð8Þ
where T is the temperature, t is the time, F is the density, CP is the specific heat, λ is the heat conductivity, Q∞ i is the reaction calorific effect, W is the heat power, r is the reaction rate constant, R is the degree of conversion for a component, NC is the number of components, and i is the component number.7,9 9491
dx.doi.org/10.1021/ie200712n |Ind. Eng. Chem. Res. 2011, 50, 9487–9494
Industrial & Engineering Chemistry Research
ARTICLE
Figure 10. DTBP heat production rate versus time curves of the experimental data and autocatalytic simulation with isothermal temperature at 140, 145, 150, and 155 °C.
Figure 11. Simulated time until 10% conversion of the thermal decomposition of DTBP and the time until the maximum rates were achieved with DSC nonisothermal tests at scanning rates of 1, 2, 4, and 8 °C min1.
Figure 13. Simulated time until 10% conversion of the thermal decomposition of DTBP and the time until the maximum rate were achieved under isothermal conditions with DSC isothermal tests at 140, 145, 150, and 155 °C.
Figure 12. Total energy release with DSC nonisothermal tests at scanning rates of 1, 2, 4, and 8 °C min1.
Figure 14. Total energy release under isothermal conditions with DSC isothermal tests at 140, 145, 150, and 155 °C. 9492
dx.doi.org/10.1021/ie200712n |Ind. Eng. Chem. Res. 2011, 50, 9487–9494
Industrial & Engineering Chemistry Research
ARTICLE
Table 6. Comparison of the Values from the Literature and the Thermal Hazard Simulation for SADT, CT, ET, and TCR in the 0.5 L Dewar Vessel and 25 kg Barrel Package size 0.5 L
25 kg
a i
sample
SADT (°C)
CT (°C)
ET (°C)
TCR (°C)
2
77
67
72
76.76
4 8
77 76
67 66
72 71
76.83 75.78
literature data simulationj
39
29
34
37.71
2
68
58
63
67.76
4
68
58
63
67.86
8
67
57
62
66.89
literature data simulationj
30
15
20
21
SADT (°C) in literature 74d, 79b, 80a, 81.7h
80c,e,f,g,i
Akzo Nobel1 b Nippon Oil and Fats Co. Ltd.2 c Fisher and Goetz12,13 d Li and Koseki15 e Li and Hasegawa16 f Sun et al.17 g Wilberforce19 h Yang et al.21 Yu and Hasegawa22 j Literature data: Ea =137.93 kJ mol1, ln A = 38.74, reaction order = 0.95 (Yu and Hasegawa22); ΔHd = 1340 kJ kg1 (Yang et al.21).
The initial fields for the temperature and the conversions were constant throughout the reactor volume: Tjt ¼ 0 ¼ T 0 Ri jt ¼ 0 ¼ Ri0
ð9Þ
Here, the index 0 indicates the initial values of the temperature and conversion. The boundary conditions of the first, second, and third kind were specified as follows:7,9 1st kind : Tjwall ¼ T e ðtÞ Temperature
ð10Þ
4. CONCLUSIONS The thermokinetic parameters and thermal hazard of DTBP were studied using nonisothermal and isothermal-kinetic models of thermal decomposition. Modeling the thermokinetic and the safety parameters provided precise hazard information concerning the avoidance of thermal accidents during process manufacturing, storage, and transportation. We developed a beneficial analysis model for the thermokinetic and thermal hazard parameters of DTBP with the green thermal analysis technology. ’ AUTHOR INFORMATION Corresponding Author
2nd kind:
Heat flow
ð11Þ
∂T j ¼ xðTwall Te Þ ∂n s
Newton0 s cooling law
qjwall ¼ qðtÞ
3rd kind : λ
ð12Þ Here, the indices “wall” and “e” relate to the parameters on the boundary and the environment, respectively. q is the heat flow, and n is the unit outer normal on the boundary.7,9 The results of the thermal hazard simulation for the SADT, CT, ET, and TCR are presented in Table 6. The thermal decomposition stability of 0.5 L Dewar vessels was greater than that of the 25 kg barrel package. The stability and applicability worsened as the reactor size increased. The results of thermal hazard simulation also proved the smaller size container has a better beneficial exothermal effect for containment of organic peroxide than a huge package. Therefore, Table 6 shows the 25 kg reactor SADT of literature data is questionable; it should smaller even than the 0.5 L reactor. We developed a green thermal analysis technology to determine the thermokinetic parameters and the thermal hazard of DTBP. These results could be applied toward energy reduction and safer designs for use and in storage. In addition to analyzing the thermal decomposition kinetic parameters through comparing nonisothermal and isothermal-kinetic model simulation, we found that the results presented a reasonable model to calculate the kinetic parameters of thermal decomposition. The validity of the results significantly depends on the reliability of the applied kinetic model, which can be validated by the proper selection of a kinetic model for a reaction, and the correctness of the methods used for the kinetics evaluation. The model can be applied to evaluating other organic peroxides or chemicals.
*Fax: +886-4-2332-1126. E-mail:
[email protected].
’ ACKNOWLEDGMENT We are indebted to the donors of the National Science Council (NSC) in Taiwan under the contract number NSC 100-2218-E468-001 for financial support. In addition, the authors are grateful to ACE Chemical Corp. in Taiwan, ROC. ’ NOMENCLATURE CP = specific heat capacity (J g1 K1) CT = control temperature (°C) Ea = activation energy (kJ mol1) E1 = activation energy of the first stage (kJ mol1) E2 = activation energy of the second stage (kJ mol1) ET = emergency temperature (°C) fi = kinetic functions of the ith stage; i = 1, 2, 3 f(R) = kinetic functions k0 = pre-exponential factor (m3 mol1 s1) ki = reaction rate constant (mol L1 s1); i = 1, 2 n = reaction order or unit outer normal on the boundary, dimensionless NC = number of components, dimensionless ni = reaction order of the ith stage, dimensionless; i = 1, 2, 3 1 Q∞ i = specific heat effect of a reaction (J kg ) 1 q = heat flow (J g ) R = gas constant (8.31415 J K1 mol1) ri = reaction rate of the ith stage, (g s1); i = 1, 2, 3, 4 S = heat-exchange surface (m2) SADT = self-accelerating decomposition temperature (°C) T = absolute temperature (K) T0 = exothermic onset temperature (°C) TCL = time to conversion limit (year) 9493
dx.doi.org/10.1021/ie200712n |Ind. Eng. Chem. Res. 2011, 50, 9487–9494
Industrial & Engineering Chemistry Research TCR = critical temperature (°C) TER = total energy release (kJ kg1) Te = ambient temperature (°C) TMRiso = time to maximum rate under isothermal conditions (day) Twall = temperature on the wall (°C) t = time, s W = heat power (W g1) z = autocatalytic constant, dimensionless R = degree of conversion, dimensionless γ = degree of conversion, dimensionless F = density (kg m3) λ = heat conductivity (W m1 K1) χ = heat transfer coefficient (W m2 K1) ΔHd = heat of decomposition (kJ kg1)
ARTICLE
(19) Whitmore, M. W.; Wilberforce, J. K. Use of the accelerating rate calorimeter and the thermal activity monitor to estimate stability temperatures. J. Loss Prev. Process Ind. 1993, 6 (2), 95–101. (20) Wilberforce, J. K. The use of the accelerating rate calorimeter to determine the SADT of organic peroxides, internal report; Columbia Scientific Corp.: Texas, 1981. (21) Yang, D.; Koseki, H.; Hasegawa, K. Predicting the self-accelerating decomposition temperature (SADT) of organic peroxides based on non-isothermal decomposition behavior. J. Loss Prev. Process Ind. 2003, 16 (5), 411–416. (22) Yu, Y. H.; Hasegawa, K. Derivation of the self-accelerating decomposition temperature for self-reactive substances using isothermal calorimetry. J. Hazard. Mater. 1996, 45 (23), 193–205.
’ REFERENCES (1) Organic Peroxide Initiators; Product Brochure, Akzo Nobel, 2001. (2) Organic peroxides, 4th ed.; Nippon Oil and Fats Co. Ltd.: Tokyo, Japan, 1996. (3) NFPA 432. Code for the storage of organic peroxide formulations; National Fire Protection Association: Quincy, MA, 2008. (4) European agreement concerning the international carriage of dangerous goods by road (ADR); United Nations: New York, 2009. (5) Recommendations on the transport of dangerous goods, manual of tests and criteria, 4th rev. ed.; United Nations: New York, 2003. (6) Recommendations on the transport of dangerous goods, model regulations, 16th rev. ed.; United Nations: New York, 2009. (7) Lin, C.-P.; Chang, C.-P.; Chou, Y.-C.; Chu, Y.-C.; Shu, C.-M. Modeling solid thermal explosion containment on reactor HNIW and HMX. J. Hazard. Mater. 2010, 176, 549–558. (8) Lin, C.-P.; Chang, Y.-M.; Tseng, J.-M.; Shu, C.-M. Comparisons of nth order kinetic algorithms and kinetic model simulation on HMX by DSC tests. J. Therm. Anal. Calorim. 2010, 100 (2), 607–614. (9) Lin, C.-P.; Tseng, J.-M.; Chang, Y.-M.; Liu, S.-H.; Cheng, Y.-C.; Shu, C.-M. Modeling liquid thermal explosion reactor containing tertbutyl peroxybenzoate. J. Therm. Anal. Calorim. 2010, 102, 587–595. (10) Tsai, L.-C.; Wei, J.-M.; Chu, Y.-C.; Chen, W.-T.; Tsai, F.-C.; Shu, C.-M.; Lin, C.-P. RDX kinetic model evaluation by nth order kinetic algorithms and model simulations. Adv. Mater. Res. 2011, 189193, 1413–1416. (11) Fierz, H. Influence of heat transport mechanisms on transport classification by SADT-measurement as measured by the Dewar-method. J. Hazard. Mater. 2003, 96 (23), 121–126. (12) Fisher, H. G.; Goetz, D. D. Determination of self-accelerating decomposition temperatures using the accelerating rate calorimeter. J. Loss Prev. Process Ind. 1991, 4, 305–316. (13) Fisher, H. G.; Goetz, D. D. Determination of self-accelerating decomposition temperatures for self-reactive substances. J. Loss Prev. Process Ind. 1993, 6 (3), 183–194. (14) Heitkamp, D.; Eberz, A. Use of kinetic methods for SADT determination. Loss Prevention and Safety Promotion in the Process Industries. 11th International Symposium on Loss Prevention, Prague, Czech, 2004. (15) Li, X. R.; Koseki, H. Thermal decomposition kinetic of liquid organic peroxides. J. Loss Prev. Process Ind. 2005, 18, 460–464. (16) Li, Y. F.; Hasegawa, K. On the thermal decomposition mechanism self-accelerating materials and evaluating method for their SADTs. 9th International Symposium Loss Prevention in the Process Industries, Spain, 1998; pp 555569. (17) Kotoyori, T. Critical temperatures for the thermal explosion of solid organic peroxides. Process Saf. Prog. 1995, 14 (1), 37–44. (18) Sun, J. H.; Li, Y. F.; Hasegawa, K. A study of self-accelerating decomposition temperature (SADT) using reaction calorimetry. J. Loss Prev. Process Ind. 2001, 14 (5), 331–336. 9494
dx.doi.org/10.1021/ie200712n |Ind. Eng. Chem. Res. 2011, 50, 9487–9494