- Email: [email protected]
Non-isothermal decomposition kinetics of hydrogarnet in sodium carbonate solution
Chinese Journal of Chemical Engineering 23 (2015) 1634–1639
Contents lists available at ScienceDirect
Chinese Journal of Chemical Engineering journal homepage: www.elsevier.com/locate/CJChE
Catalysis, Kinetics and Reaction Engineering
Non-isothermal decomposition kinetics of hydrogarnet in sodium carbonate solution☆ Xiaofeng Zhu 1, Tingan Zhang 1,⁎, Yanxiu Wang 1, Guozhi Lu 1, Weiguang Zhang 1, Cong Wang 1, Aichun Zhao 2 1 2
Key Laboratory of Ecological Utilization of Multi-metal Intergrown Ores of Ministry of Education, School of Materials and Metallurgy, Northeastern University, Shenyang 110819, China School of Material Science and Engineering, Taiyuan University of Science and Technology, Taiyuan 030024, China
a r t i c l e
i n f o
Article history: Received 9 January 2015 Received in revised form 16 July 2015 Accepted 17 July 2015 Available online 5 August 2015 Keywords: Hydrogarnet Differential scanning calorimetry (DSC) Activation energy Mechanism function Carbonation Alumina
a b s t r a c t Carbonation decomposition of hydrogarnet is a significant reaction of the calcification–carbonation new method for alumina production by using low-grade bauxite. In this work, non-isothermal decomposition kinetics of hydrogarnet in sodium carbonate solution was studied by high-pressure differential scanning calorimetry (HPDSC) at different heating rates of 2, 5, 8, 10, 15 and 20 K·min−1, respectively. The activation energy (Ea) was calculated with the help of isoconversional method (model-free), and the reaction mechanism was determined by the differential equation method. The calculated activation energy of this reaction was 115.66 kJ·mol−1. Furthermore, the mechanism for decomposition reaction is Avrami–Erofeev (n = 1.5), and the decomposition process is diffusion-controlled. © 2015 The Chemical Industry and Engineering Society of China, and Chemical Industry Press. All rights reserved.
1. Introduction As the leader of alumina production in the world, China produced more than 40 million tons of alumina, which accounts for approximately 40% of the global output in 2012 [1]. However, the lack of bauxite resource, especially high-grade bauxite is in stark contrast to the huge amount of the alumina output in China. More than 65% of bauxite in China is low-grade ore with the mass ratio of Al2O3 to SiO2 (A/S) below 7, and only 18.5% of bauxite with the A/S higher than 9. It is estimated that the high-grade bauxite will be used up within 10 years in China [2]. Bayer process is the most common method for alumina production both in China and abroad. However, it is not suitable for processing the low-grade bauxite because of the high consumption of caustic alkali and the low recovery of alumina [3]. Although the sintering process can be employed for the low-grade bauxite, it is being phased out gradually due to its complex process and high energy consumption. Furthermore, alumina industry in China results in severe environmental problems, due to the huge amount of the red mud with a high alkaline after alumina refinery [4]. Therefore,
☆ Supported by the Joint Funds of the National Natural Science Foundation of China (U1202274), the National Natural Science Foundation of China (51204040), the Research Fund for the Doctoral Program of Higher Education of China (201200421100 11), and the Doctor Start-up Foundation in Taiyuan University of Science and Technology (20142001). ⁎ Corresponding author. E-mail address: [email protected] (T. Zhang).
an environment-friendly method to efficiently utilize the abundance of the low-grade bauxite resources is urgently needed to be developed in China. In order to produce alumina with low-grade bauxite, a novel calcification–carbonation method was proposed by our team [5,6]. In the calcification process, silica-containing phase in bauxite is transformed into hydrogarnet, which is a kind of hydroxy aluminosilicate with the general chemical formula of 3CaO · Al2O3·xSiO2 · (6 − 2x)H2O (0 b x b 3). Thereafter, the generated hydrogarnet is decomposed as CaCO3, 2CaO·SiO2·nH2O and Al(OH)3 through the carbonation process. After that, the Al(OH)3 is extracted by NaOH solution at temperature below 373 K. The final red mud is composed of alkali-lean and alumina-lean CaCO3 and 2CaO·SiO2·nH2O. This process provides an eco-friendly design for the effective use of low-grade bauxite. The main reactions involved in this process can be presented as:
Al2 O3 2SiO2 2H2 O þ CaO þ AlOOH þ NaOHðaqÞ→3CaO Al2 O3 xSiO2 ð6−2xÞH2 O þ H2 O þ NaAlðOHÞ4 ðaqÞ
ð1Þ
3CaO Al2 O3 xSiO2 ð6−2xÞH2 O þ CO2− 3 þ H2 O→CaCO3
ð2Þ
− þ AlðOHÞ− 4 þ 2CaO SiO2 nH2 O þ OH
In previous works, the effect of reaction conditions on calcification process and transformation of equilibrium phase during the
http://dx.doi.org/10.1016/j.cjche.2015.07.029 1004-9541/© 2015 The Chemical Industry and Engineering Society of China, and Chemical Industry Press. All rights reserved.
X. Zhu et al. / Chinese Journal of Chemical Engineering 23 (2015) 1634–1639
process has been investigated in detail [7–9]. However, as a significant intermediate process, the kinetics of carbonation decomposition reaction (Eq. (2)) of hydrogarnet is still unclear. The differential scanning calorimetry (DSC) technique is widely used to determine the kinetics of reactions [10,11], curing [12], crystallization [13] and thermo-decomposition [14–16]. As reported by BAO et al. [10,11], a high-pressure DSC was used to determine the kinetics of gibbsite, boehmite and diaspore dissolving in caustic solution. Mechanism functions and kinetic parameters were featured by model-fitting method. SULTANIA et al. [12] studied the cure kinetics of vinyl ester–styrene system by non-isothermal DSC at four different heating rates (2.5, 5, 7.5, 10 K·min− 1), and obtained the apparent activation energy (E a ) of curing process by isoconversional method. Their results indicated that there is a good agreement between experiment and model. In this paper, the decomposition kinetics of hydrogarnet in sodium carbonate solution was studied by means of a high-pressure differential scanning calorimetry (HP-DSC). The isoconversional and differential equation methods were used to analyze the DSC curve data. Furthermore, the apparent activation energy was calculated and the most probable mechanism of this reaction was proposed.
Fig. 1 proves that the hydrogarnet is synthesized successfully. The data in Table 1 show that the mass percentages of oxides of calcium, aluminum and silicon are 40.09%, 25.80% and 4.69%, respectively. The molar ratio of CaO:Al2O3:SiO2 is 2.83:1:0.31, which corresponds to the stoichiometry of 2.83CaO·Al2O3·0.31SiO2·5.38H2O. 2.2. DSC measurements The DSC measurements were performed with a high-pressure DSC (204HP, NETZSCH, Germany). Dry argon was used as the purge gas at a rate of 20 ml·min−1, and a baseline was obtained with DSC crucible first. The mixture of prepared hydrogarnet and saturated sodium carbonate solution was sealed in a gold crucible with a stainless steel cap at the heating rate of 2, 5 and 10 K·min−1 when the DSC was performed. The Proteus Software was used to collect and analyze the DSC data. 2.3. Calculation of kinetic parameters According to non-isothermal kinetics theory, the general decomposition reaction rate can be expressed by [17]: dα Ea f ðα Þ ¼ kðT Þf ðα Þ ¼ A exp − RT dt
2. Materials and Methods 2.1. Preparation and characterization of materials The hydrogarnet used in this experiment was synthesized by hydrothermal synthesis in a 2 L scale autoclave with a magnetic stirring. The materials for preparing hydrogarnet were analytical reagent CaO, NaOH, Al(OH)3, and Na2SiO3·9H2O (Sinopharm Chemical Reagent Co., Ltd, China). 1 L sodium aluminate solution with concentration of 240 g·L−1 Na2O and 197 g·L−1 Al2O3 was prepared as the hydrothermal medium, which was subsequently mixed with 50 g CaO and 25 g Na2SiO3·9H2O (SiO2 5 g · L−1) and reacted in the autoclave at 513 K for 4 h. Then, the product was filtered and washed with distilled water to weak alkaline, and dried in an oven at 353 K for 8 h. The percentages of oxides of sample, as determined by X-ray fluorescence (XRF), are shown in Table 1. The mineralogy of sample was characterized by X-ray diffraction (XRD, D8 ADVANCE of Bruker company, 40 kV, 40 mA, CuKα, 2θ 10°–90°, increment 0.0095°) as shown in Fig. 1. Table 1 Chemical composition of the synthesized hydrogarnet Composition
Al2O3
SiO2
CaO
Na2O
Loss on ignition
Content (by mass)/%
25.80
4.69
40.09
1.85
25.14
1635
ð3Þ
where α is the reaction fraction, k(T) is the rate constant, f(α) is the differential mechanism function, T is the absolute temperature, Ea is the apparent activation energy, A is the pre-exponential factor and R is the universal gas constant (R = 8.314 J·mol−1·K−1). When a sample is heated at a constant rate under non-isothermal conditions, β = dT/dt, and Eq. (3) is modified as follows: β
dα Ea f ðα Þ: ¼ A exp − RT dT
ð4Þ
Taking the logarithm of Eq. (4) and the Friedman–Reich–Levi equation can be described as follows [18]: dα Ea : ¼ ln ½A f ðα Þ− ln β RT dT
ð5Þ
The Friedman–Reich–Levi method is considered as one of the most reliable isoconversional methods to calculate activation energy (Ea) of reactions without preselecting a reaction model [19]. By this method, activation energy can be evaluated from the slope of linear fitting ln(βdα/dT) against T−1 under a given value of reaction fraction (α). In order to obtain a reliable value, the activation energy Ea was determined by the isoconversional method (Friedman–Reich–Levi), which avoids the choosing of mechanism function. Herein, the differential equation method of non-isothermal kinetics was applied to study the reaction mechanism of the decomposition reaction, and it can be presented as follows [20]: 9 > > = dα=dT A Ea ¼ ln − ln > > Ea ðT−T 0 Þ β RT > > : f ðα Þ þ1 ; RT 2 8 > > <
ð6Þ
where f(α) is a differential expression for mechanism functions listed in Table 2. 3. Results and Discussion 3.1. XRD analysis of decomposition products
Fig. 1. XRD patterns of sample prepared by hydrothermal synthesis.
In order to characterize the product after decomposition of hydrogarnet by sodium carbonate solution, 6 g synthesized
1636
X. Zhu et al. / Chinese Journal of Chemical Engineering 23 (2015) 1634–1639
Table 2 Mechanism functions of differential equation No.
Differential equations
1 2 3 4–5 6 7 8 9 10–16 17–22 23–27 28 29
α−1/2 −[ln(1 − α)]−1 (3/2)[(1 − α)−1/3 − 1]−1 (3/n)(1 − α)2/3[1 − (1 − α)1/3]−(n − 1) (n = 2, 1/2) 4(1 − α)1/2[1 − (1 − α)1/2]1/2 (3/2)(1 + α)2/3[(1 + α)1/3 − 1]−1 (3/2)(1 − α)4/3[(1 − α)−1/3 − 1]−1 1−α (1/n)(1 − α)[−ln(1 − α)]−(n − 1) (n = 2/3, 1/2, 1/3, 4, 1/4, 2, 3) (1/n)(1 − α)−(n − 1) (n = 1/2, 3, 2, 4, 1/3, 1/4) (1/n)α−(n − 1) (n = 1, 3/2, 1/2, 1/3, 1/4) (1 − α)2 2(1 − α)3/2
hydrogarnet was added to a 200 ml saturated sodium carbonate solution and the reaction mixture was stirred at 350 rpm at 343, 353 and 363 K in a water bath for 1.5 h. Then the reactants were filtered under vacuum, washed by distilled water, and dried at 353 K for 8 h. Fig. 2 shows the XRD patterns of decomposition products at different temperatures. It indicates that the mineralogy of decomposition at different temperature is essentially the same, mainly composed of calcium carbonate, and hydrogarnet can also be found in XRD patterns due to incomplete decomposition. It is noted that the hydrated calcium silicate (2CaO · SiO 2 · nH 2 O) is not found in the products, which may be formed as an amorphous phase with low crystallinity [21].
Fig. 2. XRD patterns of decomposition products at different temperatures.
3.2. Calculation of activation energy (Ea) The DSC curves of at six different heating rates (2, 5, 8, 10, 15 and 20 K·min−1) are shown in Fig. 3. It can be seen from the curves that the thermal effects of reaction are more obvious under a higher heating rate. Simultaneously, the peaks of thermal effects broaden with an increased heating rate, which might be resulted from the reduced resolution [22]. According to Fig. 3, the reaction temperature range is significantly dependent on the heating rate, and the peak endothermic temperature (Tmax) shifts to a higher temperature region as the heating rate rises.
Fig. 3. DSC curves at different heating rates of 2, 5, 8, 10, 15 and 20 K·min−1, respectively.
By integrating the DSC curve [23], the fractional conversion as a function of temperature is obtained and shown in Fig. 4. All the T-α curves at different heating rates show a S-shape, and it shifts to a higher temperature region with an increased heating rate, which might be due to the K(T) and f(α) varying simultaneously under non-isothermal conditions [12]. The isoconversional temperatures at six different heating rates were obtained from the crossover point of fractional conversion curves.
Fig. 4. Fractional conversion as a function of temperature at different heating rates of 2, 5, 8, 10, 15 and 20 K·min−1, respectively.
By meaning of Friedman–Reich–Levi method, the diagrams of ln(βdα/dT) ~ T− 1 for the decomposition reaction were showed in Fig. 5. The slopes, activation energy, residual standard deviation (S) and correlation coefficients (R2) of each line with different degrees of conversion α were listed in Table 3. It can be seen that the activation energy changed in the range of 133.10 kJ·mol−1 to 93.19 kJ·mol−1 at different conversion α between 0.2 and 0.8. In solid-state reactions, the variation of activation energy with the degree of conversion α may be caused by the heterogeneous nature of sample under nonisothermal conditions [24]. The average activation energy of the decomposition reaction was 115.66 kJ·mol−1, and it will be subsequently used as a criterion to determine the reaction mechanism.
X. Zhu et al. / Chinese Journal of Chemical Engineering 23 (2015) 1634–1639
Fig. 5. ln(βdα/dTα,i) against T−1 α,i of Friedman–Reich–Levi method in the interval 0.2 b α b 0.8.
Table 3 The values of slopes, activation energy, residual standard and correlation coefficient at different degrees of conversion α α
Slope −16,010.33 −14,875.63 −15,514.60 −14,537.81 −13,152.25 −12,082.14 −11,209.30
0.2 0.3 0.4 0.5 0.6 0.7 0.8 Average
Ea/kJ·mol−1 133.10 123.68 128.99 120.87 109.35 100.45 93.19 115.66
−2
2.42 × 10 1.05 × 10−2 3.28 × 10−2 2.08 × 10−2 2.94 × 10−2 1.27 × 10−2 1.25 × 10−2
Fig. 6. dα/dTi as a function of temperature at a heating rate of 2 K·min−1.
Eq. (6) can be presented as: zi ¼ ayi þ bði ¼ 1; 2; …; LÞ
R2
S
0.992 0.996 0.988 0.992 0.986 0.993 0.993
L a¼
L X
ð7Þ
let zi ¼ ln
yi ¼
1 : Ti
9 > > > > =
ðdα=dT Þi " # > > > > Ea ðT i −T 0 Þ > > > > þ 1 : f ðα i Þ ; 2 RT i
i¼1
L X
y2i −
i¼1
Ti, α and dHt/dt can be obtained from the DSC curve, thus the corresponding dα/dTi can be calculated as shown in Fig. 6. Due to that there is a linear relationship between the left part of Eq. (6) and 1/T for each mechanism function listed in Table 2, Eq. (6) could be solved by the iterative method [10]. With an initial value (N 0) for Ea, the left side of Eq. (6) can be calculated for each dα/dTi. Then, a new Ea can be obtained by linear least square method from the slope and A from the intercept of Eq. (6). Take the new Ea as the initial value and do iteration. The calculating process is shown as follows [20]: 8 > > > > <
L L X X yi zi − yi zi
i¼1
L
For the thermal analysis kinetics, the extent of a chemical reaction can be measured by its thermal effects. This can be presented as a = Ht/H0, where H0 is the total heat effect of a reaction, Ht is the heat effects at a transient time t, in the non-thermal process, Ti = T0 + βt. Therefore, dα/dT is easy to be expressed as:
ð8Þ
ð9Þ
ð10Þ
where a = −Ea/R, b = ln(A/β) and L is the number of data point. Eq. (10) can be solved by the least squares method through the following equation:
3.3. Determination of the reaction mechanism function
dα 1 dH t : ¼ dT i H 0 β dt
1637
L X
i¼1 !2
ð11Þ
yi
i¼1
ð12Þ
b ¼ z−ay L
L
i¼1
i¼1
where z ¼ 1L ∑ zi and y ¼ 1L ∑ yi . Then, Ea and A can be obtained via the equation Ea = −aR, A = βeb. The linear correlation coefficient r and the remaining variance are calculated using the following equations: X L L L X X yi zi − yi zi L i¼1 i¼1 i¼1 r ¼ vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 3 2 u2 !2 !2 3 u L L L L X X X u4 X 2 2 5 4 t L yi − yi zi − zi 5 L i¼1
Q¼
L X
i¼1
2
½zi −ðayi þ bÞ :
i¼1
ð11Þ
i¼1
ð12Þ
i¼1
29 dα/dTi points shown in Fig. 6 were used to analyze the 29 mechanism functions listed in Table 2, and the results have been summarized in Table 4. It was found that six mechanism functions (No. 18, 19, 20, 25, 26, 27) yield invalid values through the iteration process. The reaction mechanism function was determined by matching the activation energy calculated via isoconversional method and correlation coefficient of linear fit is more than 0.98. It can be found that only function No. 10 with the value of Ea 111.97 kJ·mol− 1 and correlation coefficient 0.9816 can satisfy both requirements, which indicated that it should be the most probable mechanism function for the decomposition reaction. Function No. 10 is a
1638
X. Zhu et al. / Chinese Journal of Chemical Engineering 23 (2015) 1634–1639
Table 4 Kinetics parameters of hydrogarnet decomposed by sodium carbonate solution calculated by 29 differential mechanism functions No.
Apparent activation energy, Ea/J·mol−1
Pre-exponential constant, A/s−1
Correlation coefficient, r
Variance, Q
Iterative times
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29
2.0124 × 105 2.7811 × 105 3.3035 × 105 4.2029 × 105 2.444 × 105 4.3338 × 103 1.6831 × 105 6.2231 × 105 2.1423 × 105 1.1197 × 105 6.7187 × 105 2.7890 × 104 1.1848 × 106 1.2412 × 104 5.2674 × 105 8.5425 × 105 1.1992 × 105 0 0 0 1.5057 × 105 1.6622 × 105 2.1268 × 104 9.3969 × 104 0 0 0 3.9070 × 105 3.0125 × 105
6.8948 × 1028 7.7317 × 1039 8.6602 × 1046 1.2584 × 1058 1.7901 × 104 1.2594 × 101 9.6947 × 1022 1.5030 × 1090 1.5633 × 1031 1.5933 × 1017 4.1575 × 1011 1.1849 × 105 5.2946 × 10171 6.6844 × 102 8.3627 × 1076 1.2506 × 10124 6.1977 × 1015 0 0 0 3.4033 × 1020 8.1750 × 1022 2.4073 × 104 2.7564 × 1018 0 0 0 3.1102 × 1063 5.9697 × 1044
0.3374 0.7782 0.8715 0.9486 0.4470 0.0626 0.5287 0.9991 0.9899 0.9816 0.9533 0.7644 0.9863 0.4352 0.9889 0.9873 0.7345 0 0 0 0.8837 0.9296 0.1498 0.5062 0 0 0 0.9582 0.9913
267.6310 80.4761 55.2191 28.3394 3.4458 6.8771 116.6821 2.2379 1.2953 0.6831 0.6521 0.7970 56.6127 0.9497 9.0268 27.1893 14.0198 0 0 0 7.9328 5.5872 44.4280 66.5998 0 0 0 23.4862 2.1055
5 5 7 5 8 6 5 6 6 6 6 7 4 5 5 5 6 2 2 2 5 5 5 5 2 2 2 7 7
Function No. 10 (bold data) was the most probable mechanism function for the decomposition reaction.
nucleation–growth mechanism derived by Avrami–Erofeev equation (n = 1.5), f(α) = (3/2)(1 − α)[− ln(1 − α)](1/3), Pre-exponential constant is 1.5933 × 1017 s− 1. For this decomposition process, the calculated reaction order n is 1.5, which means the decomposition of hydrogarnet in sodium carbonate solution should be occurred in the presence of amorphous phase, and the process is controlled by diffusion [25]. For this reaction, the amorphous phase can be attributed to the generation of hydrated calcium silicate (2CaO·SiO2·nH2O) as a product, which was not detected in the XRD patterns (Fig. 2). On the other hand, the diffusion mechanism can be interpreted that on the decomposition of hydrogarnet, a solid layer of CaCO3 was formed, and covered on ions the surface of hydrogarnet particles. For further reaction, the CO2− 3 have to diffuse in through the solid layer of CaCO3, therefore, the hydrogarnet cannot be decomposed completely, and can be detected in the XRD patterns (Fig. 2). Based on the values of activation energy and reaction mechanism, the kinetic equation of this reaction can be described as: ! dα 1:157 105 ð1=3Þ 17 ð1−α Þ½− ln ð1−α Þ ¼ 2:39 10 exp − RT dt which is a diffusion-controlled mechanism. 4. Conclusions (1) The hydrogarnet was synthesized in the CaO–SiO 2 –NaAl (OH) 4 –H 2 O system by hydrothermal method in laboratory conditions, and the synthesized products were characterized by chemical analysis and X-ray diffraction. The results indicated that synthesized hydrogarnet had the stoichiometry of 2.83CaO·Al2O3·0.31SiO2·5.38H2O. After decomposition by
sodium carbonate solution at different temperatures, the products were mainly composed of calcium carbonate and unreacted hydrogarnet. The hydrated calcium silicate was not detected by XRD maybe due to the formation of the amorphous phase. (2) The high-pressure DSC technic was used to investigate the kinetics of hydrogarnet decomposed by sodium carbonate solution. Measurements at six different heating rates provided a determination of E a by the Friedman–Reich–Levi isoconversional method, and the average activation energy of the decomposition reaction was 115.66 kJ·mol− 1. The reaction mechanism function was selected from 29 types of differential equations, which can be presented as f(α) = (3/2)(1 − α)[− ln(1 − α)] (1/3) derived by Avrami–Erofeev equation. The reaction mechanism indicated that the decomposition of hydrogarnet in sodium carbonate solution should occur in the presence of amorphous phase, and the reaction is diffusion-controlled mechanism.
References [1] G.Z. Lv, T.A. Zhang, X.X. Wang, X.H. Zhang, Y. Liu, Q.Y. Zhao, Z.H. Dou, Effects of intense magnetic field on digestion and settling performances of bauxite, J. Cent. South Univ. 21 (6) (2014) 2168–2175. [2] W.L. Liu, Y.H. Hu, W. Sun, Separation of diaspore from bauxite by selective flocculation using hydrolyzed polyacrylamide, J. Cent. South Univ. 21 (4) (2014) 1470–1476. [3] P. Smith, The processing of high silica bauxites — Review of existing and potential processes, Hydrometallurgy 98 (1–2) (2009) 162–176. [4] S.H. Ma, Z.G. Wen, J.N. Chen, S.L. Zheng, An environmentally friendly design for low-grade diasporic-bauxite processing, Miner. Eng. 22 (9–10) (2009) 793–798. [5] Zhang, T.A., Lv, G.Z., Liu, Y., Dou, Z.H., Zhao, Q.Y., Niu, L.P., He, J.C., “A process for alumina production based on calcification–carbonation”, China Pat., CN201110275013, (2012). (in Chinese). [6] Zhang, T.A., Lv, G.Z., Liu, Y., Dou, Z.H., Zhao, Q.Y., Niu, L.P., He, J.C., “Method for dissolving Bayer process red mud”, China Pat., CN201110275030, (2012). (in Chinese). [7] T.A. Zhang, X.F. Zhu, G.Z. Lv, L. Pan, Y. Liu, Q.Y. Zhao, Y. Li, J.C. He, Calcification– carbonation method for alumina production by using low-grade bauxite, TMS Light Metals, San Antonio 2013, pp. 233–238. [8] X.F. Zhu, T.A. Zhang, G.Z. Lv, Y. Liu, Q.Y. Zhao, Y. Li, Z.H. Dou, Basic research on calcification transformation process of low grade bauxite, TMS Light Metals, San Antonio 2013, pp. 239–244. [9] G.Z. Lv, T.A. Zhang, X.F. Zhu, Y. Liu, Y.X. Wang, F.F. Guo, Q.Y. Zhao, C.Z. Zheng, Calcification–carbonation method for cleaner alumina production and CO2 utilization, JOM 66 (9) (2014) 1616–1621. [10] L. Bao, T.A. Zhang, Z.H. Dou, G.Z. Lv, Y.N. Guo, P.Y. Ni, X.J. Wu, J. Ma, Kinetics of AlOOH dissolving in caustic solution studied by high-pressure DSC, Trans. Nonferrous Metals Soc. 21 (1) (2011) 173–178. [11] L. Bao, T.A. Zhang, Y. Liu, Z.H. Dou, G.Z. Lv, X.M. Wang, J. Ma, X.L. Jiang, The most probable mechanism function and kinetic parameters of gibbsite dissolution in NaOH, Chin. J. Chem. Eng. 18 (4) (2010) 630–634. [12] M. Sultania, J.S.P. Rai, D. Srivastava, Modeling and simulation of curing kinetics for the cardanol-based vinyl esterresin by means of non-isothermal DSC measurements, Mater. Chem. Phys. 132 (1) (2012) 180–186. [13] R. Svoboda, J. Malek, Non-isothermal crystallization kinetics of As2Se3 glass studied by DSC, Thermochim. Acta 579 (10) (2014) 56–63. [14] Q.W. Long, Y. Xia, S. Liao, Y. Li, W.W. Wu, Y.H. Huang, Facile synthesis of hydrotalcite and its thermal decomposition kinetics mechanism study with masterplots method, Thermochim. Acta 579 (10) (2014) 50–55. [15] W.X. Zhang, S.Y. Liu, C.Q. Ma, D.H. Jiang, Thermo-decomposition kinetics and representative structure of homologous dinuclear complexes, Thermochim. Acta 376 (2) (2001) 163–168. [16] W.C. Xie, J. Tang, X.H. Gu, C.R. Luo, G.Y. Wang, Thermal decomposition study of menthyl-glycoside by TGA/SDTA, DSC and simultaneous Py-GC–MS analysis, J. Anal. Appl. Pyrolysis 78 (1) (2007) 180–184. [17] Y.G. Liang, B.J. Cheng, Y.B. Si, D.J. Cao, H.Y. Jiang, G.M. Han, X.H. Liu, Thermal decomposition kinetics and characteristics of Spartina alterniflora via thermogravimetric analysis, Renew. Energy 78 (2014) 111–117. [18] K.M. Rajkotia, M.M. Kamani, P.H. Parsania, Synthesis of and physico-chemical studies on poly(4,4′-cyclopentylidene diphenylene toluene-2,4-disulfonate), Polymer 38 (3) (1997) 715–719. [19] A.A. Abu-Sehly, Variation of the activation energy of crystallization in Se81.5Te16Sb2.5 chalcogenide glass: Isoconversional analysis, Thermochim. Acta 485 (1–2) (2009) 14–19. [20] R.Z. Hu, S.L. Gao, F. Zhao, Thermal Analysis Kinetics, Science Press, Beijing, 2008 80–82 (in Chinese).
X. Zhu et al. / Chinese Journal of Chemical Engineering 23 (2015) 1634–1639 [21] P. Pena, J.M. Rivas Mercury, A.H. De Aza, X. Turrillas, I. Sobrados, J. Sanz, Solid-state 27 Al and 29Si NMR characterization of hydrates formed in calcium aluminate–silica fume mixtures, J. Solid State Chem. 181 (8) (2008) 1744–1752. [22] B. Ning, R.Z. Hu, H. Zhang, Z.M. Xia, P.J. Guo, R. Liu, G.E. Lu, J.Y. Jiang, Estimation of the critical rate of temperature rise for thermal explosion of autocatalytic decomposing reaction of nitrocellulose using non-isothermal DSC, Thermochim. Acta 416 (1–2) (2004) 47–50. [23] M. Abu Ei-oypun, DSC studies on the transformation kinetics of two separated crystallization peaks of Si12.5Te87.5 chalcogenide glass: an application of the
1639
theoretical method developed and isoconversional method, Mater. Chem. Phys. 131 (1–2) (2011) 495–506. [24] B. Janković, Kinetic analysis of the nonisothermal decomposition of potassium metabisulfite using the model-fitting and isoconversional (model-free) methods, Chem. Eng. J. 139 (1) (2008) 128–135. [25] L.H. He, X.H. Liu, Z.W. Zhao, Non-isothermal kinetics study on synthesis of LiFePO4 via carbothermal reduction method, Thermochim. Acta 566 (8) (2013) 298–304.
Contents lists available at ScienceDirect
Chinese Journal of Chemical Engineering journal homepage: www.elsevier.com/locate/CJChE
Catalysis, Kinetics and Reaction Engineering
Non-isothermal decomposition kinetics of hydrogarnet in sodium carbonate solution☆ Xiaofeng Zhu 1, Tingan Zhang 1,⁎, Yanxiu Wang 1, Guozhi Lu 1, Weiguang Zhang 1, Cong Wang 1, Aichun Zhao 2 1 2
Key Laboratory of Ecological Utilization of Multi-metal Intergrown Ores of Ministry of Education, School of Materials and Metallurgy, Northeastern University, Shenyang 110819, China School of Material Science and Engineering, Taiyuan University of Science and Technology, Taiyuan 030024, China
a r t i c l e
i n f o
Article history: Received 9 January 2015 Received in revised form 16 July 2015 Accepted 17 July 2015 Available online 5 August 2015 Keywords: Hydrogarnet Differential scanning calorimetry (DSC) Activation energy Mechanism function Carbonation Alumina
a b s t r a c t Carbonation decomposition of hydrogarnet is a significant reaction of the calcification–carbonation new method for alumina production by using low-grade bauxite. In this work, non-isothermal decomposition kinetics of hydrogarnet in sodium carbonate solution was studied by high-pressure differential scanning calorimetry (HPDSC) at different heating rates of 2, 5, 8, 10, 15 and 20 K·min−1, respectively. The activation energy (Ea) was calculated with the help of isoconversional method (model-free), and the reaction mechanism was determined by the differential equation method. The calculated activation energy of this reaction was 115.66 kJ·mol−1. Furthermore, the mechanism for decomposition reaction is Avrami–Erofeev (n = 1.5), and the decomposition process is diffusion-controlled. © 2015 The Chemical Industry and Engineering Society of China, and Chemical Industry Press. All rights reserved.
1. Introduction As the leader of alumina production in the world, China produced more than 40 million tons of alumina, which accounts for approximately 40% of the global output in 2012 [1]. However, the lack of bauxite resource, especially high-grade bauxite is in stark contrast to the huge amount of the alumina output in China. More than 65% of bauxite in China is low-grade ore with the mass ratio of Al2O3 to SiO2 (A/S) below 7, and only 18.5% of bauxite with the A/S higher than 9. It is estimated that the high-grade bauxite will be used up within 10 years in China [2]. Bayer process is the most common method for alumina production both in China and abroad. However, it is not suitable for processing the low-grade bauxite because of the high consumption of caustic alkali and the low recovery of alumina [3]. Although the sintering process can be employed for the low-grade bauxite, it is being phased out gradually due to its complex process and high energy consumption. Furthermore, alumina industry in China results in severe environmental problems, due to the huge amount of the red mud with a high alkaline after alumina refinery [4]. Therefore,
☆ Supported by the Joint Funds of the National Natural Science Foundation of China (U1202274), the National Natural Science Foundation of China (51204040), the Research Fund for the Doctoral Program of Higher Education of China (201200421100 11), and the Doctor Start-up Foundation in Taiyuan University of Science and Technology (20142001). ⁎ Corresponding author. E-mail address: [email protected] (T. Zhang).
an environment-friendly method to efficiently utilize the abundance of the low-grade bauxite resources is urgently needed to be developed in China. In order to produce alumina with low-grade bauxite, a novel calcification–carbonation method was proposed by our team [5,6]. In the calcification process, silica-containing phase in bauxite is transformed into hydrogarnet, which is a kind of hydroxy aluminosilicate with the general chemical formula of 3CaO · Al2O3·xSiO2 · (6 − 2x)H2O (0 b x b 3). Thereafter, the generated hydrogarnet is decomposed as CaCO3, 2CaO·SiO2·nH2O and Al(OH)3 through the carbonation process. After that, the Al(OH)3 is extracted by NaOH solution at temperature below 373 K. The final red mud is composed of alkali-lean and alumina-lean CaCO3 and 2CaO·SiO2·nH2O. This process provides an eco-friendly design for the effective use of low-grade bauxite. The main reactions involved in this process can be presented as:
Al2 O3 2SiO2 2H2 O þ CaO þ AlOOH þ NaOHðaqÞ→3CaO Al2 O3 xSiO2 ð6−2xÞH2 O þ H2 O þ NaAlðOHÞ4 ðaqÞ
ð1Þ
3CaO Al2 O3 xSiO2 ð6−2xÞH2 O þ CO2− 3 þ H2 O→CaCO3
ð2Þ
− þ AlðOHÞ− 4 þ 2CaO SiO2 nH2 O þ OH
In previous works, the effect of reaction conditions on calcification process and transformation of equilibrium phase during the
http://dx.doi.org/10.1016/j.cjche.2015.07.029 1004-9541/© 2015 The Chemical Industry and Engineering Society of China, and Chemical Industry Press. All rights reserved.
X. Zhu et al. / Chinese Journal of Chemical Engineering 23 (2015) 1634–1639
process has been investigated in detail [7–9]. However, as a significant intermediate process, the kinetics of carbonation decomposition reaction (Eq. (2)) of hydrogarnet is still unclear. The differential scanning calorimetry (DSC) technique is widely used to determine the kinetics of reactions [10,11], curing [12], crystallization [13] and thermo-decomposition [14–16]. As reported by BAO et al. [10,11], a high-pressure DSC was used to determine the kinetics of gibbsite, boehmite and diaspore dissolving in caustic solution. Mechanism functions and kinetic parameters were featured by model-fitting method. SULTANIA et al. [12] studied the cure kinetics of vinyl ester–styrene system by non-isothermal DSC at four different heating rates (2.5, 5, 7.5, 10 K·min− 1), and obtained the apparent activation energy (E a ) of curing process by isoconversional method. Their results indicated that there is a good agreement between experiment and model. In this paper, the decomposition kinetics of hydrogarnet in sodium carbonate solution was studied by means of a high-pressure differential scanning calorimetry (HP-DSC). The isoconversional and differential equation methods were used to analyze the DSC curve data. Furthermore, the apparent activation energy was calculated and the most probable mechanism of this reaction was proposed.
Fig. 1 proves that the hydrogarnet is synthesized successfully. The data in Table 1 show that the mass percentages of oxides of calcium, aluminum and silicon are 40.09%, 25.80% and 4.69%, respectively. The molar ratio of CaO:Al2O3:SiO2 is 2.83:1:0.31, which corresponds to the stoichiometry of 2.83CaO·Al2O3·0.31SiO2·5.38H2O. 2.2. DSC measurements The DSC measurements were performed with a high-pressure DSC (204HP, NETZSCH, Germany). Dry argon was used as the purge gas at a rate of 20 ml·min−1, and a baseline was obtained with DSC crucible first. The mixture of prepared hydrogarnet and saturated sodium carbonate solution was sealed in a gold crucible with a stainless steel cap at the heating rate of 2, 5 and 10 K·min−1 when the DSC was performed. The Proteus Software was used to collect and analyze the DSC data. 2.3. Calculation of kinetic parameters According to non-isothermal kinetics theory, the general decomposition reaction rate can be expressed by [17]: dα Ea f ðα Þ ¼ kðT Þf ðα Þ ¼ A exp − RT dt
2. Materials and Methods 2.1. Preparation and characterization of materials The hydrogarnet used in this experiment was synthesized by hydrothermal synthesis in a 2 L scale autoclave with a magnetic stirring. The materials for preparing hydrogarnet were analytical reagent CaO, NaOH, Al(OH)3, and Na2SiO3·9H2O (Sinopharm Chemical Reagent Co., Ltd, China). 1 L sodium aluminate solution with concentration of 240 g·L−1 Na2O and 197 g·L−1 Al2O3 was prepared as the hydrothermal medium, which was subsequently mixed with 50 g CaO and 25 g Na2SiO3·9H2O (SiO2 5 g · L−1) and reacted in the autoclave at 513 K for 4 h. Then, the product was filtered and washed with distilled water to weak alkaline, and dried in an oven at 353 K for 8 h. The percentages of oxides of sample, as determined by X-ray fluorescence (XRF), are shown in Table 1. The mineralogy of sample was characterized by X-ray diffraction (XRD, D8 ADVANCE of Bruker company, 40 kV, 40 mA, CuKα, 2θ 10°–90°, increment 0.0095°) as shown in Fig. 1. Table 1 Chemical composition of the synthesized hydrogarnet Composition
Al2O3
SiO2
CaO
Na2O
Loss on ignition
Content (by mass)/%
25.80
4.69
40.09
1.85
25.14
1635
ð3Þ
where α is the reaction fraction, k(T) is the rate constant, f(α) is the differential mechanism function, T is the absolute temperature, Ea is the apparent activation energy, A is the pre-exponential factor and R is the universal gas constant (R = 8.314 J·mol−1·K−1). When a sample is heated at a constant rate under non-isothermal conditions, β = dT/dt, and Eq. (3) is modified as follows: β
dα Ea f ðα Þ: ¼ A exp − RT dT
ð4Þ
Taking the logarithm of Eq. (4) and the Friedman–Reich–Levi equation can be described as follows [18]: dα Ea : ¼ ln ½A f ðα Þ− ln β RT dT
ð5Þ
The Friedman–Reich–Levi method is considered as one of the most reliable isoconversional methods to calculate activation energy (Ea) of reactions without preselecting a reaction model [19]. By this method, activation energy can be evaluated from the slope of linear fitting ln(βdα/dT) against T−1 under a given value of reaction fraction (α). In order to obtain a reliable value, the activation energy Ea was determined by the isoconversional method (Friedman–Reich–Levi), which avoids the choosing of mechanism function. Herein, the differential equation method of non-isothermal kinetics was applied to study the reaction mechanism of the decomposition reaction, and it can be presented as follows [20]: 9 > > = dα=dT A Ea ¼ ln − ln > > Ea ðT−T 0 Þ β RT > > : f ðα Þ þ1 ; RT 2 8 > > <
ð6Þ
where f(α) is a differential expression for mechanism functions listed in Table 2. 3. Results and Discussion 3.1. XRD analysis of decomposition products
Fig. 1. XRD patterns of sample prepared by hydrothermal synthesis.
In order to characterize the product after decomposition of hydrogarnet by sodium carbonate solution, 6 g synthesized
1636
X. Zhu et al. / Chinese Journal of Chemical Engineering 23 (2015) 1634–1639
Table 2 Mechanism functions of differential equation No.
Differential equations
1 2 3 4–5 6 7 8 9 10–16 17–22 23–27 28 29
α−1/2 −[ln(1 − α)]−1 (3/2)[(1 − α)−1/3 − 1]−1 (3/n)(1 − α)2/3[1 − (1 − α)1/3]−(n − 1) (n = 2, 1/2) 4(1 − α)1/2[1 − (1 − α)1/2]1/2 (3/2)(1 + α)2/3[(1 + α)1/3 − 1]−1 (3/2)(1 − α)4/3[(1 − α)−1/3 − 1]−1 1−α (1/n)(1 − α)[−ln(1 − α)]−(n − 1) (n = 2/3, 1/2, 1/3, 4, 1/4, 2, 3) (1/n)(1 − α)−(n − 1) (n = 1/2, 3, 2, 4, 1/3, 1/4) (1/n)α−(n − 1) (n = 1, 3/2, 1/2, 1/3, 1/4) (1 − α)2 2(1 − α)3/2
hydrogarnet was added to a 200 ml saturated sodium carbonate solution and the reaction mixture was stirred at 350 rpm at 343, 353 and 363 K in a water bath for 1.5 h. Then the reactants were filtered under vacuum, washed by distilled water, and dried at 353 K for 8 h. Fig. 2 shows the XRD patterns of decomposition products at different temperatures. It indicates that the mineralogy of decomposition at different temperature is essentially the same, mainly composed of calcium carbonate, and hydrogarnet can also be found in XRD patterns due to incomplete decomposition. It is noted that the hydrated calcium silicate (2CaO · SiO 2 · nH 2 O) is not found in the products, which may be formed as an amorphous phase with low crystallinity [21].
Fig. 2. XRD patterns of decomposition products at different temperatures.
3.2. Calculation of activation energy (Ea) The DSC curves of at six different heating rates (2, 5, 8, 10, 15 and 20 K·min−1) are shown in Fig. 3. It can be seen from the curves that the thermal effects of reaction are more obvious under a higher heating rate. Simultaneously, the peaks of thermal effects broaden with an increased heating rate, which might be resulted from the reduced resolution [22]. According to Fig. 3, the reaction temperature range is significantly dependent on the heating rate, and the peak endothermic temperature (Tmax) shifts to a higher temperature region as the heating rate rises.
Fig. 3. DSC curves at different heating rates of 2, 5, 8, 10, 15 and 20 K·min−1, respectively.
By integrating the DSC curve [23], the fractional conversion as a function of temperature is obtained and shown in Fig. 4. All the T-α curves at different heating rates show a S-shape, and it shifts to a higher temperature region with an increased heating rate, which might be due to the K(T) and f(α) varying simultaneously under non-isothermal conditions [12]. The isoconversional temperatures at six different heating rates were obtained from the crossover point of fractional conversion curves.
Fig. 4. Fractional conversion as a function of temperature at different heating rates of 2, 5, 8, 10, 15 and 20 K·min−1, respectively.
By meaning of Friedman–Reich–Levi method, the diagrams of ln(βdα/dT) ~ T− 1 for the decomposition reaction were showed in Fig. 5. The slopes, activation energy, residual standard deviation (S) and correlation coefficients (R2) of each line with different degrees of conversion α were listed in Table 3. It can be seen that the activation energy changed in the range of 133.10 kJ·mol−1 to 93.19 kJ·mol−1 at different conversion α between 0.2 and 0.8. In solid-state reactions, the variation of activation energy with the degree of conversion α may be caused by the heterogeneous nature of sample under nonisothermal conditions [24]. The average activation energy of the decomposition reaction was 115.66 kJ·mol−1, and it will be subsequently used as a criterion to determine the reaction mechanism.
X. Zhu et al. / Chinese Journal of Chemical Engineering 23 (2015) 1634–1639
Fig. 5. ln(βdα/dTα,i) against T−1 α,i of Friedman–Reich–Levi method in the interval 0.2 b α b 0.8.
Table 3 The values of slopes, activation energy, residual standard and correlation coefficient at different degrees of conversion α α
Slope −16,010.33 −14,875.63 −15,514.60 −14,537.81 −13,152.25 −12,082.14 −11,209.30
0.2 0.3 0.4 0.5 0.6 0.7 0.8 Average
Ea/kJ·mol−1 133.10 123.68 128.99 120.87 109.35 100.45 93.19 115.66
−2
2.42 × 10 1.05 × 10−2 3.28 × 10−2 2.08 × 10−2 2.94 × 10−2 1.27 × 10−2 1.25 × 10−2
Fig. 6. dα/dTi as a function of temperature at a heating rate of 2 K·min−1.
Eq. (6) can be presented as: zi ¼ ayi þ bði ¼ 1; 2; …; LÞ
R2
S
0.992 0.996 0.988 0.992 0.986 0.993 0.993
L a¼
L X
ð7Þ
let zi ¼ ln
yi ¼
1 : Ti
9 > > > > =
ðdα=dT Þi " # > > > > Ea ðT i −T 0 Þ > > > > þ 1 : f ðα i Þ ; 2 RT i
i¼1
L X
y2i −
i¼1
Ti, α and dHt/dt can be obtained from the DSC curve, thus the corresponding dα/dTi can be calculated as shown in Fig. 6. Due to that there is a linear relationship between the left part of Eq. (6) and 1/T for each mechanism function listed in Table 2, Eq. (6) could be solved by the iterative method [10]. With an initial value (N 0) for Ea, the left side of Eq. (6) can be calculated for each dα/dTi. Then, a new Ea can be obtained by linear least square method from the slope and A from the intercept of Eq. (6). Take the new Ea as the initial value and do iteration. The calculating process is shown as follows [20]: 8 > > > > <
L L X X yi zi − yi zi
i¼1
L
For the thermal analysis kinetics, the extent of a chemical reaction can be measured by its thermal effects. This can be presented as a = Ht/H0, where H0 is the total heat effect of a reaction, Ht is the heat effects at a transient time t, in the non-thermal process, Ti = T0 + βt. Therefore, dα/dT is easy to be expressed as:
ð8Þ
ð9Þ
ð10Þ
where a = −Ea/R, b = ln(A/β) and L is the number of data point. Eq. (10) can be solved by the least squares method through the following equation:
3.3. Determination of the reaction mechanism function
dα 1 dH t : ¼ dT i H 0 β dt
1637
L X
i¼1 !2
ð11Þ
yi
i¼1
ð12Þ
b ¼ z−ay L
L
i¼1
i¼1
where z ¼ 1L ∑ zi and y ¼ 1L ∑ yi . Then, Ea and A can be obtained via the equation Ea = −aR, A = βeb. The linear correlation coefficient r and the remaining variance are calculated using the following equations: X L L L X X yi zi − yi zi L i¼1 i¼1 i¼1 r ¼ vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 3 2 u2 !2 !2 3 u L L L L X X X u4 X 2 2 5 4 t L yi − yi zi − zi 5 L i¼1
Q¼
L X
i¼1
2
½zi −ðayi þ bÞ :
i¼1
ð11Þ
i¼1
ð12Þ
i¼1
29 dα/dTi points shown in Fig. 6 were used to analyze the 29 mechanism functions listed in Table 2, and the results have been summarized in Table 4. It was found that six mechanism functions (No. 18, 19, 20, 25, 26, 27) yield invalid values through the iteration process. The reaction mechanism function was determined by matching the activation energy calculated via isoconversional method and correlation coefficient of linear fit is more than 0.98. It can be found that only function No. 10 with the value of Ea 111.97 kJ·mol− 1 and correlation coefficient 0.9816 can satisfy both requirements, which indicated that it should be the most probable mechanism function for the decomposition reaction. Function No. 10 is a
1638
X. Zhu et al. / Chinese Journal of Chemical Engineering 23 (2015) 1634–1639
Table 4 Kinetics parameters of hydrogarnet decomposed by sodium carbonate solution calculated by 29 differential mechanism functions No.
Apparent activation energy, Ea/J·mol−1
Pre-exponential constant, A/s−1
Correlation coefficient, r
Variance, Q
Iterative times
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29
2.0124 × 105 2.7811 × 105 3.3035 × 105 4.2029 × 105 2.444 × 105 4.3338 × 103 1.6831 × 105 6.2231 × 105 2.1423 × 105 1.1197 × 105 6.7187 × 105 2.7890 × 104 1.1848 × 106 1.2412 × 104 5.2674 × 105 8.5425 × 105 1.1992 × 105 0 0 0 1.5057 × 105 1.6622 × 105 2.1268 × 104 9.3969 × 104 0 0 0 3.9070 × 105 3.0125 × 105
6.8948 × 1028 7.7317 × 1039 8.6602 × 1046 1.2584 × 1058 1.7901 × 104 1.2594 × 101 9.6947 × 1022 1.5030 × 1090 1.5633 × 1031 1.5933 × 1017 4.1575 × 1011 1.1849 × 105 5.2946 × 10171 6.6844 × 102 8.3627 × 1076 1.2506 × 10124 6.1977 × 1015 0 0 0 3.4033 × 1020 8.1750 × 1022 2.4073 × 104 2.7564 × 1018 0 0 0 3.1102 × 1063 5.9697 × 1044
0.3374 0.7782 0.8715 0.9486 0.4470 0.0626 0.5287 0.9991 0.9899 0.9816 0.9533 0.7644 0.9863 0.4352 0.9889 0.9873 0.7345 0 0 0 0.8837 0.9296 0.1498 0.5062 0 0 0 0.9582 0.9913
267.6310 80.4761 55.2191 28.3394 3.4458 6.8771 116.6821 2.2379 1.2953 0.6831 0.6521 0.7970 56.6127 0.9497 9.0268 27.1893 14.0198 0 0 0 7.9328 5.5872 44.4280 66.5998 0 0 0 23.4862 2.1055
5 5 7 5 8 6 5 6 6 6 6 7 4 5 5 5 6 2 2 2 5 5 5 5 2 2 2 7 7
Function No. 10 (bold data) was the most probable mechanism function for the decomposition reaction.
nucleation–growth mechanism derived by Avrami–Erofeev equation (n = 1.5), f(α) = (3/2)(1 − α)[− ln(1 − α)](1/3), Pre-exponential constant is 1.5933 × 1017 s− 1. For this decomposition process, the calculated reaction order n is 1.5, which means the decomposition of hydrogarnet in sodium carbonate solution should be occurred in the presence of amorphous phase, and the process is controlled by diffusion [25]. For this reaction, the amorphous phase can be attributed to the generation of hydrated calcium silicate (2CaO·SiO2·nH2O) as a product, which was not detected in the XRD patterns (Fig. 2). On the other hand, the diffusion mechanism can be interpreted that on the decomposition of hydrogarnet, a solid layer of CaCO3 was formed, and covered on ions the surface of hydrogarnet particles. For further reaction, the CO2− 3 have to diffuse in through the solid layer of CaCO3, therefore, the hydrogarnet cannot be decomposed completely, and can be detected in the XRD patterns (Fig. 2). Based on the values of activation energy and reaction mechanism, the kinetic equation of this reaction can be described as: ! dα 1:157 105 ð1=3Þ 17 ð1−α Þ½− ln ð1−α Þ ¼ 2:39 10 exp − RT dt which is a diffusion-controlled mechanism. 4. Conclusions (1) The hydrogarnet was synthesized in the CaO–SiO 2 –NaAl (OH) 4 –H 2 O system by hydrothermal method in laboratory conditions, and the synthesized products were characterized by chemical analysis and X-ray diffraction. The results indicated that synthesized hydrogarnet had the stoichiometry of 2.83CaO·Al2O3·0.31SiO2·5.38H2O. After decomposition by
sodium carbonate solution at different temperatures, the products were mainly composed of calcium carbonate and unreacted hydrogarnet. The hydrated calcium silicate was not detected by XRD maybe due to the formation of the amorphous phase. (2) The high-pressure DSC technic was used to investigate the kinetics of hydrogarnet decomposed by sodium carbonate solution. Measurements at six different heating rates provided a determination of E a by the Friedman–Reich–Levi isoconversional method, and the average activation energy of the decomposition reaction was 115.66 kJ·mol− 1. The reaction mechanism function was selected from 29 types of differential equations, which can be presented as f(α) = (3/2)(1 − α)[− ln(1 − α)] (1/3) derived by Avrami–Erofeev equation. The reaction mechanism indicated that the decomposition of hydrogarnet in sodium carbonate solution should occur in the presence of amorphous phase, and the reaction is diffusion-controlled mechanism.
References [1] G.Z. Lv, T.A. Zhang, X.X. Wang, X.H. Zhang, Y. Liu, Q.Y. Zhao, Z.H. Dou, Effects of intense magnetic field on digestion and settling performances of bauxite, J. Cent. South Univ. 21 (6) (2014) 2168–2175. [2] W.L. Liu, Y.H. Hu, W. Sun, Separation of diaspore from bauxite by selective flocculation using hydrolyzed polyacrylamide, J. Cent. South Univ. 21 (4) (2014) 1470–1476. [3] P. Smith, The processing of high silica bauxites — Review of existing and potential processes, Hydrometallurgy 98 (1–2) (2009) 162–176. [4] S.H. Ma, Z.G. Wen, J.N. Chen, S.L. Zheng, An environmentally friendly design for low-grade diasporic-bauxite processing, Miner. Eng. 22 (9–10) (2009) 793–798. [5] Zhang, T.A., Lv, G.Z., Liu, Y., Dou, Z.H., Zhao, Q.Y., Niu, L.P., He, J.C., “A process for alumina production based on calcification–carbonation”, China Pat., CN201110275013, (2012). (in Chinese). [6] Zhang, T.A., Lv, G.Z., Liu, Y., Dou, Z.H., Zhao, Q.Y., Niu, L.P., He, J.C., “Method for dissolving Bayer process red mud”, China Pat., CN201110275030, (2012). (in Chinese). [7] T.A. Zhang, X.F. Zhu, G.Z. Lv, L. Pan, Y. Liu, Q.Y. Zhao, Y. Li, J.C. He, Calcification– carbonation method for alumina production by using low-grade bauxite, TMS Light Metals, San Antonio 2013, pp. 233–238. [8] X.F. Zhu, T.A. Zhang, G.Z. Lv, Y. Liu, Q.Y. Zhao, Y. Li, Z.H. Dou, Basic research on calcification transformation process of low grade bauxite, TMS Light Metals, San Antonio 2013, pp. 239–244. [9] G.Z. Lv, T.A. Zhang, X.F. Zhu, Y. Liu, Y.X. Wang, F.F. Guo, Q.Y. Zhao, C.Z. Zheng, Calcification–carbonation method for cleaner alumina production and CO2 utilization, JOM 66 (9) (2014) 1616–1621. [10] L. Bao, T.A. Zhang, Z.H. Dou, G.Z. Lv, Y.N. Guo, P.Y. Ni, X.J. Wu, J. Ma, Kinetics of AlOOH dissolving in caustic solution studied by high-pressure DSC, Trans. Nonferrous Metals Soc. 21 (1) (2011) 173–178. [11] L. Bao, T.A. Zhang, Y. Liu, Z.H. Dou, G.Z. Lv, X.M. Wang, J. Ma, X.L. Jiang, The most probable mechanism function and kinetic parameters of gibbsite dissolution in NaOH, Chin. J. Chem. Eng. 18 (4) (2010) 630–634. [12] M. Sultania, J.S.P. Rai, D. Srivastava, Modeling and simulation of curing kinetics for the cardanol-based vinyl esterresin by means of non-isothermal DSC measurements, Mater. Chem. Phys. 132 (1) (2012) 180–186. [13] R. Svoboda, J. Malek, Non-isothermal crystallization kinetics of As2Se3 glass studied by DSC, Thermochim. Acta 579 (10) (2014) 56–63. [14] Q.W. Long, Y. Xia, S. Liao, Y. Li, W.W. Wu, Y.H. Huang, Facile synthesis of hydrotalcite and its thermal decomposition kinetics mechanism study with masterplots method, Thermochim. Acta 579 (10) (2014) 50–55. [15] W.X. Zhang, S.Y. Liu, C.Q. Ma, D.H. Jiang, Thermo-decomposition kinetics and representative structure of homologous dinuclear complexes, Thermochim. Acta 376 (2) (2001) 163–168. [16] W.C. Xie, J. Tang, X.H. Gu, C.R. Luo, G.Y. Wang, Thermal decomposition study of menthyl-glycoside by TGA/SDTA, DSC and simultaneous Py-GC–MS analysis, J. Anal. Appl. Pyrolysis 78 (1) (2007) 180–184. [17] Y.G. Liang, B.J. Cheng, Y.B. Si, D.J. Cao, H.Y. Jiang, G.M. Han, X.H. Liu, Thermal decomposition kinetics and characteristics of Spartina alterniflora via thermogravimetric analysis, Renew. Energy 78 (2014) 111–117. [18] K.M. Rajkotia, M.M. Kamani, P.H. Parsania, Synthesis of and physico-chemical studies on poly(4,4′-cyclopentylidene diphenylene toluene-2,4-disulfonate), Polymer 38 (3) (1997) 715–719. [19] A.A. Abu-Sehly, Variation of the activation energy of crystallization in Se81.5Te16Sb2.5 chalcogenide glass: Isoconversional analysis, Thermochim. Acta 485 (1–2) (2009) 14–19. [20] R.Z. Hu, S.L. Gao, F. Zhao, Thermal Analysis Kinetics, Science Press, Beijing, 2008 80–82 (in Chinese).
X. Zhu et al. / Chinese Journal of Chemical Engineering 23 (2015) 1634–1639 [21] P. Pena, J.M. Rivas Mercury, A.H. De Aza, X. Turrillas, I. Sobrados, J. Sanz, Solid-state 27 Al and 29Si NMR characterization of hydrates formed in calcium aluminate–silica fume mixtures, J. Solid State Chem. 181 (8) (2008) 1744–1752. [22] B. Ning, R.Z. Hu, H. Zhang, Z.M. Xia, P.J. Guo, R. Liu, G.E. Lu, J.Y. Jiang, Estimation of the critical rate of temperature rise for thermal explosion of autocatalytic decomposing reaction of nitrocellulose using non-isothermal DSC, Thermochim. Acta 416 (1–2) (2004) 47–50. [23] M. Abu Ei-oypun, DSC studies on the transformation kinetics of two separated crystallization peaks of Si12.5Te87.5 chalcogenide glass: an application of the
1639
theoretical method developed and isoconversional method, Mater. Chem. Phys. 131 (1–2) (2011) 495–506. [24] B. Janković, Kinetic analysis of the nonisothermal decomposition of potassium metabisulfite using the model-fitting and isoconversional (model-free) methods, Chem. Eng. J. 139 (1) (2008) 128–135. [25] L.H. He, X.H. Liu, Z.W. Zhao, Non-isothermal kinetics study on synthesis of LiFePO4 via carbothermal reduction method, Thermochim. Acta 566 (8) (2013) 298–304.