Non-Isothermal Decomposition of 2-( 2-Hydroxybenzylideneamino )-3-phenylpropanoic Acid in Nitrogen Atmosphere

The non-isothermal decomposition properties of 2-(2-hydroxybenzylideneamino)-3-phenylpropanoic acid [HBAPPA] have been studied using microanalysis, FT-IR, UV, DTA, DTG and TG techniques. The TG studies were carried out at different heating rates of 10, 15 and 20 K/min. The Schiff base decomposed in three stages. The kinetic parameters were deduced for each stage. A probable mechanism has been proposed for the decomposition process.


Introduction
The well established thermogravimetric analysis (TGA), differential thermal analysis (DTA) and differential thermogravimetric analysis (DTG) techniques have been widely used for studying the thermal behaviour of various types of materials and the thermodynamic parameters for the decomposition processes could be evaluated.A method to determine the activation energy by using integral dynamic curves from TG at multiple heating rates has been proposed by Flynn-Wall 1 and Ozawa 2 .Several studies have been carried out for the determination of kinetic parameters viz., activation energy, pre-exponential factor and life time of several materials such as polymers, organic and inorganic compounds [3][4][5][6] .
The Schiff base derived from salicylaldehyde and L-phenylalanine and its complexes are used as antibacterial 7 , antifungal 7,8 , antimalarial 9 agents and also for DNA cleavage 10 , amino oxidase 11 , deamination antiradical activity 12 , bio-distribution and blood clearance 13 .We are reporting here, the non-isothermal decomposition of 2-(2-hydroxybenzylideneamino)-3phenylpropanoic acid at different heating rates in nitrogen atmosphere.

Experimental
Phenylalanine, salicylaldehyde and other chemicals used were of AnalaR grade from BDH.

Preparation of 2-(2-hydroxybenzylideneamino)-3-phenylpropanoic acid
The compound was synthesized by mixing an aqueous alcoholic solution of sodium salt of L-phenyalanine (0.01 M) and salicylaldehyde (0.01 M) (Scheme 1).The reaction mixture was heated for about 4 h.The resultant solution was cooled to room temperature and treated with 1:1 HCl.The pale yellow Schiff base was filtered, washed thoroughly with deionised water and dried in vacuum.The solid obtained was recrystallized using DMSO-water (50%).The melting point of the compound is 189 °C (lit: 189 14 ).Micro anal calcd % for C 16

Measurements
Microanalysis of the compound was carried out in a HERAEUS Carlo Erba 1108 model at Central Drug Research Institute, Lucknow, India.The UV-Visible spectrum of the sample was recorded on a HITACHIU 2001 UV-Visible spectrophotometer in DMF medium.The FT-IR spectrum was recorded on a AVATAR model 360 using KBr pellets.Thermogravimetric analysis was carried out using a NETZSCH -Gerate bare GMBH thermal analysis, STA 409 PC.The weight of the sample was maintained constant (10 mg) for all the heating rates.

Kinetic theory
The degree of conversion of the decomposition process is expressed as Where, m 0 is the initial mass of the sample, m t the mass of the sample at time t and m ∞ , the final mass of the sample in that reaction.The general kinetic equation is written as Where α is the degree of conversion of the sample and t is time, T in the absolute temperature, f(α) is a function, the type of which depends on the reaction mechanism, k(T), the temperature dependent rate constant, usually described by Arrhenius equation as, k = A exp Where, A is pre-exponential or frequency factor.E a is the activation energy and R is the universal gas constant.Inserting Eqn (2) into Eqn (1) gives An integration function is shown as Where g(α) is the integral kinetic function or integral reaction model when its form is mathematically defined, β = dT/dt, the heating rate of decomposition.The kinetic parameters were calculated using model-fitting Coats and Redfern method 15,16 , model-free Flynn-Wall-Ozawa 1,2 (FWO), Kissinger-Akahira and Sunose 17,18 (KAS) and Tang 19,20 (T) methods.

Model-free method
The following procedures are used for the determination of kinetic parameters by isoconversional methods.
Flynn-Wall-Ozawa method 1,2 The isoconversional integral method suggested independently by Flynn-Wall 1 and Ozawa 2 uses Doyle's approximation of p(x).This method is based on the Eqn (6).
Thus, for a constant α value, the plot log β versus (1/T), obtained from thermograms recorded at several heating rates should be a straight line whose slope can be used to evaluate the apparent activation energy.
Thus, for a fixed α value, the plot of ln versus (1/T) should be a straight line whose slope can be used to evaluate the apparent activation energy.
Tang method 19,20 This is based on the eqn.(8)   ln versus 1/T, should be a straight line whose slope can be used to calculate E a .
Model-fitting method (Coats and Refern) 15,16 Modified Coats-Redfern equation is used for the determination of Arrhenius parameters.ln Different models are used for the calculation of kinetic parameters (Table 1).
Kissinger method 17,21 Kissinger proposed a kinetic analysis method for thermal reactions.According to Kissinger, the maximum reaction rate occurs when the second derivative is zero from which the following equation can be obtained: Where, T m is the temperature of the maximum rate and α m is the conversion value at that rate.The maximum reaction rate represents the peak (i.e.inflection point) of a DSC or DTG curve.Taking the natural logarithm of Eqn.(10) and rearranging gives, The activation energy is obtained by plotting the left-hand side of the above equation versus m T 1 for a series of runs at different heating rates.Eqn.(11) has been generalized to any reaction model (f(α)).The peak temperatures are noted from DTG curves at different heating rates 22,23 .The thermodynamic parameters can be calculated using the following equations: A exp(-E/RT) = υ exp(-∆G ≠ /RT) ( 12) Where ∆G ≠ is the Gibbs free energy of activation, ∆H ≠ the enthalpy of activation, ∆S ≠ the entropy of activation and υ the Einstein vibrational frequency, υ = k B T/h (where k B and h are Boltzmann and Planck's constants, respectively).The values of entropy, enthalpy and free energy of activation at the peak temperatures are obtained on the basis of Eqns.( 12) - (14).

Results and Discussion
TG, DTA and DTG curves of salicylidene-phenylalanine are shown in Figure 1 at various heating rates, i.e. 10, 15 and 20 K/min.As can be seen from Figure 1, the compound is stable up to 200 °C in all heating rates.After the endothermic process of melting at 104.8 °C at the heating rate of 10 K/min, it shows three decomposition stages of mass loss.The first stage starts at 480 K and ends at about 550 K with the corresponding mass loss of 68.71, 62.27 and 65.90% at different heating rates.It is probably due to the partial release of phenylalanine (found 62.27, calculated 60.57%) characterized by a strong endothermic peak.The second stage starts at 550 K and ends at 693 K, with the mass loss of 16.48, 16.78 and 22.48% respectively for different heating rates (weak endothermic) attributed to the partial decomposition of salicylaldehyde (cyclobutadione) (found 16.46; calculated 19.33%) with weak endothermic peak observed at 387.3 °C.In the third stage, which starts at 694 K ends at 1073 K.This is due to the mass loss of cyclopropanic oxide.Similar trends are observed for other heating rates also.

Kinetic analysis
TG, DTG, DTA curves of the title compound are shown in Figure 1.The data obtained for the non-isothermal decomposition process at 10, 15 and 20 K/min were processed according to equation (1).The variation of E a with α for all the three decomposition stages are given in Figures.2-4.It is seen that E a value depends upon the extent of conversion in stages I and III (Figures 2 & 4), whereas in stage II (Figure 3) it is independent of the extent of conversion (0.2 ≤ α ≤ 0.8).The data indicate that the kinetics are likely to be governed by a single step reaction i.e., F 1 mechanism for state II, whereas as multi-step reactions are involved in stages I and III.It is seen from the Figures 2 and 4, that there is a decrease in activation energy for stage I and for the stage III it increases up to α ≤ 0.5 and then decreases.This suggests that it follows a multi-step kinetics.Conversely, in stage II, the E a values do not significantly vary with α (0.1 ≤ α ≤ 0.8). Figure 2 shows a decrease in activation energy from 200 to 102 kJ/mol as α increases.In second stage, the E a value increases from 93 to 190 kJ/mol (Figure 3).In the third stage, the E a value increases from 26 to 96 kJ/mol (0.1 ≤ α ≤ 0.5) (Figure 4).

Invariant kinetic parameters
The invariant kinetic parameters, E inv and A inv are calculated using Coats-Redfern method 15,16,24 (Table 1).The values of E a and lnA depend on the kinetic model as well as on the heating rate, as shown in Tables 2-4.The evaluation of the invariant parameters is performed using supper correlation equation and the values are listed in Table 5. 1. Algebraic expressions of f(α) and g(α) for the reaction models considered in the present work 15,16

Symbol
Reaction Model a In some references f(α) and g(α) have opposite designations    The plot of a β versus b β , obtained for three constant heating rates, is a straight line whose parameters allow the computation of lnA inv and E inv (Table 6).For several groups of apparent activation parameters listed in Table 5, obtained by different kinetic models, we tried to establish the best correlation and the closest value to the mean isoconversional activation energies [FWO; E a = 138.97±4.31kJ/mol].For AKM-{D 1 -D 4 ), the plot of lnA versus E a has the highest correlation coefficient E a = 146.23 kJ/mol; lnA = 33.18;(r = -0.990)and is a straight line for stage I.For AKM-{F 3 , D 1 -D 4 ), the plot of lnA versus E a has the highest correlation coefficient (E a = 105.94kJ/mol; lnA = 19.63;r =0.950) and is true straight line for the second stage (FWO; E a = 131.42± 3.02 kJ/mol).Whereas for AKM-{P 2 -P 4 , D 1 -D 4 , A 2 -A 3 , R 2 , R 3 }, the plot of lnA versus E a has the highest correlation coefficient E a = 54.4 kJ/mol; lnA = 5.02 (r = 0.998) and is a straight line for the third stage (FWO; E a = 70.75±3.26kJ/mol).

Thermodynamic parameters
As can see from the Table 7, the value of ∆S ≠ is positive for the first stage and negative for the second and third stages.The negative values suggest that the activated complexes have a higher degree of arrangement than the initial stage.The positive values of enthalpy and free energy of activation for this compound show that the process is connected with absorption of heat is non-spontaneous in nature 25,26 .

Conclusion
The title compound is decomposed in three stages.The energy of activation of first stage is more than that of the other two stages, which indicates that the first stage decompose in rather slow.∆G ≠ is more positive indicating that the decomposition process is nonspontaneous in nature.

Figure 2 .Figure 3 .
Figure 2. Variation of the E a with α for the first stage decomposition of HBAPPA by FWO method Figure 3. Variation of E a with α for the second stage decomposition of HBAPPA by FWO method

Figure 4 .
Figure 4. Variation of E a with α for the third stage decomposition of HBAPPA by FWO method

Table 2 .
Kinetic parameters for the first stage decomposition of HBAPPA

Table 3 .
Kinetic parameters for the second stage decomposition of HBAPPA

Table 4 .
Kinetic parameters for the third stage decomposition of HBAPPA

Table 5 .
Compensation effect parameters for several combinations of kinetic models for the decomposition of HBAPPA

Table 6 .
IKP for several combinations of kinetic models for HBAPPA

Table 7 .
Kinetic and thermodynamic parameters for the thermal behaviour of HBAPPA byKissinger method