J Therm Anal Calorim (2014) 117:473–480 DOI 10.1007/s10973-014-3695-5
Thermal studies and decomposition kinetics of alkaline earth metal trichloroacetates Sukhjinder Singh • Deepika Saini • S. K. Mehta Ravneet Kaur • Valeria Ferretti
Received: 31 July 2013 / Accepted: 9 February 2014 / Published online: 5 March 2014 Ó Akade´miai Kiado´, Budapest, Hungary 2014
Abstract Alkaline earth metal trichloroacetates M(O2 CCCl3)2nH2O, where M = Be (1), n = 4; M = Mg (2), n = 6; M = Ca (3) or Sr (4) or Ba (5), n = 4, were synthesized and their thermal behavior analyzed using thermogravimetric analysis (TG/DTG/DSC). A critical examination was made for the apparent activation energy by means of non-isothermal kinetic methods employing multiple heating rates. A systematic and comparative study of thermal decomposition was carried out at different heating rates i.e., 5, 10, 15, and 20 °C min-1 for various trichloroacetates synthesized. It was observed that the Ca, Sr, and Ba trichloroacetates decompose preferentially to respective metal halides while Be and Mg compounds decompose to metal and metal oxide, respectively. The composition of the final residues was also confirmed using FT-IR spectroscopy. The activation energy follows the order: Mg [ Ca [ Sr [ Ba, Be being the exception. Results reveal that each metal trichloroacetate decomposes through its unique thermolysis mechanism.
Electronic supplementary material The online version of this article (doi:10.1007/s10973-014-3695-5) contains supplementary material, which is available to authorized users. S. Singh (&) D. Saini S. K. Mehta (&) R. Kaur Department of Chemistry and Centre of Advanced Studies in Chemistry, Panjab University, Chandigarh 160014, India e-mail: [email protected] S. K. Mehta e-mail: [email protected] V. Ferretti Centro di Strutturistica Diffrattometrica and Dipartimento de Chimica, University of Ferrara, Via L. Borsari 46, 44100 Ferrara, Italy
Keywords Alkaline earth metal trichloroacetates Activation energy Thermal behavior TG/DTG DSC Thermal decomposition kinetics Introduction The research in the area of alkaline earth metal compounds is significantly increasing these days because of the wide range of applications shown by them. They act as precursors for metal organic chemical vapor deposition (MOCVD), metal organic deposition (MOD), and/or metal organic chemical solution deposition (MOCSD) techniques useful for semiconductor, and functional ceramics such as biomaterials, fertilizers, sensors, ferroelectrics, and catalysis [1–5]. Moreover, the biochemistry of virtually all living organisms is fairly dependent on the alkaline earth metals (i.e., Mg, Ca, Sr, etc.) . Therefore, an improved understanding of the functioning and properties of these biological cations is required. It has also become desirable to investigate the thermal properties of these complexes. Generally, the metal carboxylates decompose generating insoluble metal carbonates which is undesirable and adversely affects their use in advanced material applications (MOD, MOCSD). However, trihaloacetates (highly halogen substituted carboxylates), i.e., the alkaline earth metal trifluoroacetates decompose into corresponding fluorides at relatively low temperatures. Hence, this avoids the formation of insoluble metal carbonates which affects their use as precursors for advanced material applications. Therefore, the emerging popularity of alkaline earth metal trifluoroacetates, M(O2CCF3)2 (M = Ca, Sr or Ba) as precursors for the deposition of thin films and generation of nanopowders [7, 8] may be attributed to their easy and convenient syntheses [9, 10] and further transformation into fluorides via thermolysis .
We undertook the synthesis of alkaline earth metal trichloroacetates, i.e., Be(O2CCCl3)24H2O, [Mg(H2O)6] [O2CCCl3]2, M(O2CCCl3)2(H2O)4, where M = Ca, Sr, Ba and their thermal characterization by TG/DTG/DSC. TG can help to understand the structural variety in these compounds, determine the coordination of the metals with different ligands, and explore the thermal (kinetic as well as thermodynamic) properties of these compounds. It enables to determine apparent kinetic parameters of decomposition reactions the reaction order n and the apparent activation energy E. The activation energy of the thermal decomposition reaction is an indication of the relative bond strengths within the molecules studied and can give us a fair indication about the bonding properties. The present work deals with the thermal decomposition studies (TG, DTG, and DSC) of different metal trichloroacetates. The apparent activation energy E for thermal decomposition has been determined by means of Flynn– Wall–Ozawa method at various heating rates, i.e., 5, 10, 15, and 20 °C min-1. The advantage of using Flynn–Wall– Ozawa method over the popularly used Coats Redfern or Horowitz–Metzger method is that it eliminates the need for assumption of any model, i.e., it is a model-free calculation method while rest of the methods are based on models. However, Flynn–Wall–Ozawa method is not valid in the case of E varying with degree of conversion, a. Therefore, the thermal calculations have been carried out using a modified Friedman’s method proposed by Budrugeac and Segal  that takes into account the a-dependence of activation energy. The different decomposition mechanisms for the compounds synthesized have also been proposed based on the TG/DTG curves.
Experimental All starting materials were obtained commercially and used without further purification. Sample preparation The alkaline earth metal trichloroacetates, i.e., Be(O2CCCl3)2 4H2O, [Mg(H2O)6][O2CCCl3]2, M(O2CCCl3)2(H2O)4, where M = Ca, Sr, Ba were synthesized in the laboratory. The synthesis, full characterization, and single crystal X-ray studies (for Ca, Sr, Ba) of the compounds have already been described elsewhere . Instrumentation Simultaneous TG–DTG–DSC curves were obtained with thermal analysis system; model SDT Q-600, from TA
S. Singh et al.
Instruments. The purge gas was nitrogen flow of 100.0 mL min-1. Heating rates of 5, 10, 15 and 20 °C min-1 were used, with samples weighing about 10.0 mg. This technique consists of heating the sample to a given temperature at a fixed heating rate (b) and then simultaneously recording the mass loss. Alumina crucibles were used for recording the curves. The FT-IR spectra were recorded on a Nicolet iS 50 Thermo Scientific instrument using the samples in powdered form.
Results and discussion TG was employed to determine and compare the (a) thermal decomposition mechanism and (b) the apparent activation energy (E) of various metal trichloroacetates synthesized. Thermal behavior The thermal behavior of different metal trichloroacetates was studied with varying heating rates. The TG/DSC curves (Fig. 1) show that all the metal trichloroacetates except Mg compound decompose in two steps. However, Mg follows a three-step decomposition process. This goes well with the literature findings by Chiaretto et al.  as well. They observed the thermal decomposition of Magnesium cinnamate in three stages whereas the rest of alkaline earth metals follow two-step decomposition. Therefore, it is evident that different thermolysis mechanisms are followed just owing to the presence of different metals even if the groups coordinated remain same. The plausible decomposition mechanisms are as summarized below (Scheme 1). The results of thermogravimetric analysis performed on compounds 1–5 under nitrogen atmosphere are presented in Table 1. Transition temperatures are indicated by the DTG curves (not shown for the sake of clarity). Compound 1 is thermally decomposed in two successive decomposition steps within the temperature range 25–800 °C. The first step (obs = 22.19 %, calc = 21.62 %) within the range 120–180 °C is accounted to the partial loss of some H2O molecules from the hydrated compound. The second decomposition step found within the temperature range 180–360 °C (obs = 75.48 %, calc = 74.30 %) may be attributed to the complete expulsion of coordinated trichloroacetate anions as well as remaining H2O molecules leaving behind beryllium metal as the residual product. Total observed mass loss (98.15 %) is in excellent agreement with the calculated mass loss (97.90 %). TG curve of 2 represents the thermal decomposition of the complex within temperature range 40–800 °C, in three successive steps. From the TG curve, the first step of
Alkaline earth metal trichloroacetates
Strontium Magnesium Calcium Beryllium Berium
80 70 60
1.0 0.5 0.0
50 –0.5 40
Fig. 1 (a) TG traces, (b) DSC curves for the metal trichloroacetates
120-180oC Be4O(O2CCCl3)6(H2O)12 -(CCl3CO)2O -4H2O (endo) -CCl3COCl + CO2 + Cl2 (exo) 180-360oC -12H O (endo) 2
Table 1 TG for different decomposition steps of all the complexes at b = 10 °C min-1 Metal trichloroacetates
Mass loss/% Calculated
56-150oC Mg(O2CCCl3)2(H2O) -5H2O (endo)
-C2Cl6 + CO (exo) 150-310oC -H O (endo) 2 MgCO3
52-220oC -4H2O (endo)
220-270oC -CCl3COCl + CO + CO2 (exo)
50-140oC -CO (exo) -4H2O (endo)
Sr(OCCl3)(O2CCCl3) 140-230oC -CCl3COCl + CO2 (exo)
50-154oC -2H2O (endo)
Ba(O2CCCl3)2(H2O)2 -CCl3COCl + CO + CO2 (exo) 154-210oC -2H O (endo) 2 BaCl2
Scheme 1 Schematic presentation of thermolysis of compounds 1–5
decomposition is found to occur between 56 and 150 °C (obs = 19.14 %, calc = 19.68 %), corresponding to the removal of 5 molecules of H2O out of total 6H2O
molecules associated with the compound. The mass loss (obs = 60.38 %, calc = 61.89 %) from 150 to 310 °C is accounted to the elimination of one unit of H2O and decomposition of both trichloroacetate anions to expel CO ? C2Cl6 as volatiles. MgCO3 is formed at the end of the step. The conclusive step of thermolysis between 310 and 570 °C suggests MgO as the final residue (obs = 11.54 %, calc = 9.62 %) of the process with the loss of CO2. Total mass loss (obs = 91.06 %, calc = 91.19 %) confirms the composition of Mg(H2 O)6(CCl3COO)2 for 2 and MgO as the final residue. Thermal decomposition of 3 was studied within the temperature range 40–800 °C and evaluated as two major steps of the TG curve. The first step corresponds to the removal of 4H2O molecules from the complex in the
temperature range 52–220 °C (obs = 17.02 %, calc = 16.50 %). Further mass loss (obs = 59.22 %, calc = 58.12 %) between 220 and 270 °C, is accounted to the decomposition of trichloroacetate groups as CCl3COCl ? CO ? CO2 and halide transfer to the metal atom resulting in the formation of CaCl2. Calcium chloride is identified as the final residue from excellent agreement between the observed mass loss (74.62 %) and calculated mass loss (76.24 %). The step between 220 and 270 °C is accompanied by an exothermic peak revealing breakdown of (CCl3COO)2 to CCl3COCl, CO, and CO2. An endotherm/peak at *760 °C in DSC curve of 3 may be coinciding with the reported  m.pt. of CaCl2 (772 °C). Thermal decomposition of 4 proceeds in two major steps. The first step between 50 and 140 °C corresponds to the elimination of 4H2O molecules and simultaneous decarboxylation of CCl3COO group sets in and one molecule of CO escapes during the process (obs = 21.18 %, calc = 20.63 %). This step involves one endotherm in initial stage and then exotherm. The final step within temperature range 140–230 °C accounts to the removal of volatiles as CCl3COCl, CO2 (obs = 48.98 %, calc = 46.64 %). SrCl2 remains as the final product of decomposition (obs = 70.16 %, calc = 67.27 %). The slight variation in calculated and observed values may be partially due to some volatility of final product. The compound 5 again decomposes mainly in two steps. The first step occurs between 50 and 154 °C and is accounted to the liberation of 2H2O molecules (obs = 3.61 %, calc = 6.74 %). The last step concludes between 154 and 220 °C with a mass loss (obs = 49.74 %, calc = 54.27 %) corresponding to the elimination of 2 H2O ? CCl3COCl ? CO ? CO2 from the compound to give BaCl2 as final residue. The formation of BaCl2 as the final residue is supported by the observed endothermic peaks in the DSC curve in the region 945–960 °C (Fig. 1). These peaks in DSC curve are related to the melting point of BaCl2 (reported m.pt. is 961 °C) . The proposed final residue based on % mass loss during thermal analysis has been confirmed separately by heating 1 at 380 °C, 2 at 600 °C, 3 at 290 °C, 4 at 240 °C, and 5 at 230 °C, respectively, under N2 atmosphere. Thermal analysis of 1–5 reveals that polymeric compounds 3, 4, or 5 produce respective metal halides as final residual materials while 1 and 2 with different structures produce Be and MgO, respectively, as residues. FT-IR spectroscopy The composition of final residues has been confirmed using FT-IR Spectroscopy ex situ. FT-IR spectroscopy has been carried out for the metal trichloroacetates and the residues obtained after decomposition. The samples 1, 3, 4, and 5
S. Singh et al.
were heated to 400 °C while 2 was heated to 600 °C. The FT-IR spectra were recorded after cooling the residues to room temperature. The final residue obtained in case of 1, i.e., Be was very less as only the metal remains, therefore, a clear FT-IR spectrum could not be obtained. The FT-IR spectra for 2, 3, 4, 5 have been compared with the final residues. Figure 2 indicates that although the metal trichloroacetates decompose to metal chlorides (except Mg which decomposes to MgO), but the residues obtained show a tendency to take up water of hydration. For the metal trichloroacetates, the FT-IR spectra show two sharp peaks at approx. 1,340 and 1,620 cm-1 corresponding to the C=O stretching, while the peaks between 600 and 800 cm-1 are suggestive of C–Cl stretching. The broad peak from 3,200 to 3,600 cm-1 corresponds to O–H stretching due to water of hydration. On the other hand for the final residues, the antisymmetric and symmetric C=O stretching and C–Cl stretching peaks are eliminated. The peaks observed at around 1,600 and 3,200–3,600 cm-1 could be attributed to the O–H stretching frequency due to water molecules which are in perfect agreement with the FT-IR spectrum observed by Miller and Wilkins  for BaCl22H2O. Actually, the peaks due to water arise because the cooling of the final residues and FT-IR measurement is carried out in air ex situ, therefore, the peaks corresponding to metal chloride or oxide with water of hydration are recorded in the FT-IR spectrum. The FT-IR spectra obtained agree very well with those of hydrated metal chloride/oxide spectra already available in literature [15–17]. Therefore, from the FT-IR spectra it can be fairly well implied that the final residue obtained after thermal decomposition of metal trichloroacetates is metal chloride for Ba, Ca, Sr, and MgO in case of Mg. These results fairly well corroborate the decomposition pathways predicted from TG studies, which predict metal halide as the decomposed product at the end of thermal decomposition process. Kinetics and calculation procedure The methods used for the calculation of kinetic parameters were chosen on the basis that any knowledge of the reaction mechanism or reaction model is not required. The first method employed was Flynn–Wall–Ozawa method [18, 19] which uses Doyle’s linear approximation of temperature integral. Several TG curves at different heating rates were recorded. It is assumed that reaction rate at constant a, where a represents the extent of reaction or conversion depends only on temperature. Hence, constant E values are expected for single-step decomposition processes while, E varies with a for multi-step processes. A master curve is derived from the data at different heating rates to calculate E at any particular value of a using the following equation:
Alkaline earth metal trichloroacetates
13471625 672 838
Fig. 2 FT-IR spectra for a Ba, b Ca, c Sr, and d Mg. Black line represents metal trichloroacetate; red line represents final residue. (Color figure online)
Table 2 Activation energy values for metal trichloroacetates at different heating rates using Flynn Wall Ozawa method E/kJ mol
α value 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
1.1 1.0 0.9
logb ¼ log
AE E 2:315 log gðaÞ 0:4567 : R RT
The log b values plotted versus reciprocal temperature give parallel lines for fixed a values and the slope gives the value of E, as slope = -0.4567(E/R). Therefore, the apparent values of reaction order (n), the activation energies (E), and pre exponential factor (A) of the thermal decomposition reactions have been calculated from the thermogravimetric curves taking into account the multiple heating rates. At different heating rates, the kinetic result of TG and DTG evaluations using FWO method for the synthesized compounds have been presented in Table 2. The linearization curves were plotted for the compounds and those
Fig. 3 Representative isoconversional plots for Mg trichloroacetate using Flynn–Wall–Ozawa method
corresponding to the maximum value of the correlation coefficient have been chosen for all the further calculations. Representative isoconversional plots (because they are carried out at fixed a value) for Mg trichloroacetate have been presented in Fig. 3. The curves for the rest of metal trichloroacetates have been given in Supplementary information. Figure 4 reveals the linear variation of log b with 1/T along with different a values. A variation of E with a value has been observed (Fig. 5; Table 2), i.e., the apparent activation energy does not
Fig. 4 Flynn–Wall–Ozawa plots at different a values 110 100
Ca Be Ca Sr Ba
E/kJ mol –1
90 80 70
Mg Sr Be Ba Ca
50 25 0.0
100 150 200 250 300 350 400 450 500 550
E/kJ mol –1
Fig. 7 The plot of ln A versus E for all the metal trichloroacetates
Fig. 5 a dependence of E for various metal trichloroacetates evaluated from slope of Flynn–Wall–Ozawa plots
According to Budrugeac and Segal, E and A values are found to be interrelated through compensation effect as: remain constant and keeps on changing with degree of conversion. However, Eq. 1 holds only with the assumption that the activation energy remains constant. The Ozawa method is only applicable if the process satisfies single-step approximation and remains invariant with temperature and degree of conversion, a . This type of a dependence of E is usually seen in case of complex multistep reactions. To overcome this problem, Budrugeac and Segal  proposed another method for the calculation of activation energy as well as preexponential factor, A, both of which depend upon the degree of conversion.
lnA ¼ aE þ b;
where a and b are constant coefficients. The equation used for estimating non-isothermal kinetic parameters is: E da ln b ð3Þ ¼ lnAf ðaÞ : dT RT The plot of ln bddTa versus 1/T is linear at constant a values, where the slope and intercept of the straight line yield the values of E and Af(a) [f(a) = (1-a)n, n, being the order of reaction]. Figure 6 shows the representative ln bddTa versus 1/T curves for Sr trichloroacetate with
Alkaline earth metal trichloroacetates
Table 3 Compensation parameters derived from ln A versus E plots Trichloroacetates
0.4 B a B 0.9 (linearization graphs for other metal trichloroacetates have been given in supplementary information). The a values are chosen corresponding to those which give maximum correlation coefficient. Therefore, various values of A can be obtained corresponding to different n values but the value of n is chosen such as to give correlation coefficient closest to 1 for the straight line between ln A and E due to the presence of compensation effect. The compensation effect gives rise to isokinetic temperature, Ti, which can also be determined from the slope of ln A versus E plot using the equation: Ti ¼
Although the order of E obtained remains the same as Flynn–Wall–Ozawa method; i.e., Mg (327.42 kJ mol-1) [ Ca (142.69 kJ mol-1) [ Sr (137.49 kJ mol-1) [ Ba (97.45 kJ mol-1), Be (73.54 kJ mol-1) However, the magnitude obtained is much lower. The linear plot of ln A versus E (Fig. 7) corresponding to the compensation effect gives Ti, with reaction order, n = 1. The compensation parameters obtained have been listed in Table 3. It is clear from the order of apparent activation energy values obtained by both the methods, that the highest apparent activation energy is required by the Mg compound. The first and foremost reason for this behavior may be that Mg compound is coordinated to six water molecules whereas the rest of the compounds have only four water molecules in coordination. These six coordinated H2O molecules in 2 are involved in extensive H-bonding (bifurcation/trifurcation) with CCl3COO ligands whereas fewer H2O molecules are involved in H-bonding with CCl3COO ligands in case of 3 and 4. The order followed in this case is Mg [ Ca [ Sr [ Ba. Hydrogen bonding between non-bonded CCl3COO and bonded H2O is maximum in case of Mg compound while the order of hydrogen bonding interaction between bonded H2O and non-bonded CCl3COO keeps on decreasing as we move from Mg to Ca to Sr to Ba trichloroacetates , i.e., it may be the order of their activation energy. However, the E of Be compound is different in the order of 1–5 and may be attributed to the smaller size and high charge distorting the electronic cloud of anion and changing hydrogen bonding between nonbonded CCl3COO and bonded H2O.
The present results may also get support from findings of Hall and Verhock  that activation energy required for thermal decomposition of metal trichloroacetates in solution increases as the water content is increased. This is because of hydration of cation, ultimately leading to the solvation of CCl3COO ions by coordinated H2O molecules via H-bonding, thereby, leading to an increase in activation energy. Studies on the series of metal trichloroacetates help to evaluate the effect of metal change on the stability of compounds, keeping all other factors constant. It is observed that the apparent activation energy decreases down the group. A possible explanation for this consistent decrease with increasing radii of ions (as move from Mg2? to Ba2?) could be that larger metal ion does not permit a closer approach of ligands coordinated. Therefore, it will lead to an increase in the metal–ligand distance and hence decrease in the bond energy. Hence, lesser apparent activation energy is required for Ba compound as compared to others. Therefore, only the metal change can highly affect the decomposition mechanism even if the coordinating ligands are same.
Conclusions In this paper, we have presented the thermal behavior of the hydrated alkaline earth metal trichloroacetates. The compounds 3–5 having polymeric structures preferentially decompose to respective metal halides as the final residue which has been confirmed using FT-IR spectroscopy. There are two important findings: (i) the apparent activation energy for Mg compound is high compared to others, which infers that the activation energy for thermal decomposition of metal trichloroacetates increases as the coordinated water content is increased, (ii) the apparent activation energy follows the order: Mg [ Ca [ Sr [ Ba. The possible explanation for consistent decrease of activation energy with increasing radii of ions (Mg to Ba) could be that the larger metal ion does not permit a closer approach of ligands or it can be the magnitude of hydrogen bonding which determines the activation energy. The solubility of 3–5 in a number of solvents and their preferential decomposition to respective metal halides at T \ 300 °C makes them promising potential precursors in advanced material applications.
480 Acknowledgements DS gratefully acknowledges the meritorious fellowship as a financial assistance from University Grants Commissions (UGC), New Delhi, India. SKM is thankful to Department of Science and Technology (DST), India for the financial assistance.
References 1. Gschwind F, Sereda O, Fromm KM. Multitopic ligand design: a concept for single-source precursors. Inorg Chem. 2009;48:10535–47. 2. Fromm KM. Coordination polymer networks with s-block metal ions. Coord Chem Rev. 2008;252:856–85. 3. Singh MK, Yang Y, Takoudisa CG. Synthesis of multifunctional multiferroic materials from metalorganics. Coord Chem Rev. 2009;253:2920–34. 4. Trnovcova V, Fedorov PP, Furar I. Fluoride solid electrolytes containing rare earth elements. J Rare Earth. 2008;26:225–32. 5. Deacon GB, Junk PC, Moley GJ, Ruhlandt-Senge K, Stprox C, Zungia MF. Charge-separated and molecular heterobimetallic rare earth–rare earth and alkaline earth–rare earth aryloxo complexes featuring intramolecular metal–p-arene interactions. Chem Eur J. 2009;15:5503–19. 6. Chen ZF, Xiong RG, Zhang J, Zuo JL, Guo Z, You XZ, Fun FK. X-ray crystal structures of Mg2? and Ca2? dimers of the antibacterial drug norfloxacin. J Chem Soc Dalton Trans. 2000;22:4013–4. 7. Lv Y, Wu X, Wu D, Huo D, Zhao S. Synthesis of barium fluoride nanoparticles by precipitation in ethanol–aqueous mixed solvents. Powder Technol. 2007;173:174–8. 8. Quan Z, Yang D, Li C, Yang P, Cheng Z, Yang J, Kong D, Lin J. SrF2 hierarchical flowerlike structures: solvothermal synthesis, formation mechanism, and optical properties. Mater Res Bull. 2009;44:1009–16. 9. Wojtczak WA, Atanassova P, Hampden-Smith MJ, Duesler E. Synthesis and characterization of polyether adducts of barium and strontium carboxylates and their use in the formation of MTiO(3) films. Inorg Chem. 1996;35:6995–7000.
S. Singh et al. 10. Boyle TJ, Pratt HD, Alam TM. Synthesis and characterization of polyether adducts of barium and strontium carboxylates and their use in the formation of MTiO3 films. Polyhedron. 2007;26:5095–6103. 11. Budrugeac P, Segal E. Non-isothermal kinetics of reactions whose activation energy depends on the degree of conversion. Thermochim Acta. 1995;260:75–85. 12. Singh S, Saini D, Mehta SK, Kaur R, Ferretti V. Synthesis and characterization of alkaline earth metal trichloroacetates. Inorg Chim Acta. 2014 (under revision). 13. Chiaretto Z, Glauce A, Filho C, da Silva Fernandes MA, Suely N, Massao I. Preparation and thermal decomposition of solid state cinnamates of alkali earth metals except beryllium and radium. Eclet Quı´m. 1998;23:91–3. 14. Lide DR. CRC handbook of chemistry and physics. 89th ed. New York: CRC; 2008. 15. Miller FA, Wilkins CH. Infrared spectra and characteristic frequencies of inorganic ions. Their use in qualitative analysis. Anal Chem. 1952;24:1253–94. 16. Buchanan RA, Caspers HH, Murphy J. Lattice vibration spectra of Mg(OH)2 and Ca(OH)2. Appl Opt. 1963;2:1147–50. 17. Benesi HA. Infrared spectrum of Mg(OH)2. J Chem Phys. 1959;30:852. 18. Ozawa T. A new method of analyzing thermogravimetric data. Bull Chem Soc Jpn. 1965;38:1881–6. 19. Flynn J, Wall L. A quick, direct method for the determination of activation energy from thermogravimetric data. J Polym Sci B. 1966;4:323–8. 20. Koga N. Ozawa’s kinetic method for analyzing thermoanalytical curves. J Therm Anal Calorim. 2013;. doi 10.1007/s10973-0122882-5. 21. Hall GA, Verhock FH. The kinetics of the decomposition of certain salts of trichloroacetic acid in ethanol–water mixtures. J Am Chem Soc. 1947;69:613–6.