Specificity of decomposition of solids in non-isothermal conditions

Article abstract of DOI:10.1023/A:1024537902405

The thermal stability of the food additives Na metabisulphite, Na and K acetates, glutamic and citric acids, respective of the pharmaceuticals nifedipine and acetyl salicylic acid was studied by means of the non-isothermal kinetic (Friedman differential m

Full text of DOI:10.1023/A:1024537902405

Journal of Thermal Analysis and Calorimetry, Vol. 72 (2003) 597–604
SPECIFICITY OF DECOMPOSITION OF SOLIDS IN
NON-ISOTHERMAL CONDITIONS
T. Vlase*, G. Vlase, M. Doca and N. Doca
West University of Timiºoara, Department of Chemistry, Str. Pestalozzi 16, RO-1900 Timiºoara,
Romania
Abstract
The thermal stability of the food additives Na metabisulphite, Na and K acetates, glutamic and citric
acids, respective of the pharmaceuticals nifedipine and acetyl salicylic acid was studied by means of
the non-isothermal kinetic (Friedman differential method).
The specificity of the thermal decomposition was characterized by identification of the bond to
be selective activated due to energy absorption at vibrational level. These bonds were assigned by
comparison of calculated wave numbers with the wave number of the IR spectra.
Keywords: compensation effect, IR wave number, isokinetic temperature, non-isothermal kinetics,
pharmaceutical and food additives
Introduction
Many years ago the thermal analysis earned its place as a current instrumentation
technique in assisting the analytical problems of pharmaceuticals [1]. A relative new
trend in this field is the study of the thermal stability of pharmaceuticals and food ad-
ditives in connection with the molecular architecture.
In an earlier paper [2] we suggested a correlation between the isokinetic temper-
ature and the wave number assigned to the bond responsible for the beginning of the
thermal decomposition of a certain compound. In this paper we extend this study to
some usual food additives and pharmaceuticals.
Experimental
The thermal curves were recorded with a MOM Derivatograph devices in static air at-
mosphere. For the thermal analysis, 100 mg samples were used, with heating rates of
20, 10, 5 and 2.5 K min–1.
The IR spectra were determined in KBr tablet for the initial, respectively the
thermal treated samples, using a NICOLET-510 FTIR Spectrometer.
The TG data were interpolated with logistical function and finally numerical
derivate. With the values of α, β(dα/dT) and T, the Friedman`s method for estimating
the non-isothermal kinetic parameters was used.
*
Author for correspondence: E-mail: tvlase@cbg.uvt.ro
1388–6150/2003/ $ 20.00
Akadémiai Kiadó, Budapest
© 2003 Akadémiai Kiadó, Budapest
Kluwer Academic Publishers, Dordrecht

598
VLASE et al.: SPECIFICITY OF DECOMPOSITION
Results and discussion
In Table 1 the main characteristics of the thermal decomposition steps of the studied
compounds are systematized.
Table 1 Decomposition characteristics of the studied compounds, determined at different heating rates
Mass loss/% by heating rate of °C min–1
Thermal
effect
Calculated
mass loss/%
No. Compound
Na2S2O5
20
10
5
2.5
1
Exo
Exo
25.7
24.5
23.4
24.1
25.0 (Eq. 1)
2a Na(CH3COO)
2b K(CH3COO)
37.0
28.5
33.3
29.0
33.3
31.1
33.9
30.8
35.4
29.5 (Eq. 2)
12.2
(Scheme 1)
3
Glutamic acid
Endo
11.8
11.5
11.5
12.9
4
5
Citric acid
Nifedipine
Endo
Exo
83.5
58.0
83.0
57.2
84.5
55.4
81
78.2
–
54.3
Acetylsalicylic
acid
6
Endo
36.5
35.7
36.4
37
33.0
The sodium metabisulphite presents a single important step, an exothermic pro-
cess with mass loss and without an ‘a priori’ melting. This step is assigned to a de-
composition/oxidation process:
Na2S2O5 → Na2SO3+SO2
(1)
The concordance between the experimental and the calculated mass loss is very good.
The alkaline acetates presents (after dehydration) an exothermic process with
mass loss, corresponding to a decomposition according to the equation
2M(CH3COO) → 2MCO3+(CH3)2CO
The experimental determined mass loss is in very good agreement with the cal-
culated values.
(2)
The glutamic acid presents a decomposition step with a significant change in the
molecular structure, i.e. the dimerization to dicetopiperazine (Scheme 1).
Scheme 1
J. Therm. Anal. Cal., 72, 2003

VLASE et al.: SPECIFICITY OF DECOMPOSITION
599
This step is endothermic and the determined mass loss agrees very well with the
calculated ones.
The citric acid presents a melting step with the endothermic maximum at 140ºC.
After this, a unique degradation step occurs, with the characteristics presented in Ta-
ble 1. The theoretical mass loss was calculated for a combined dehydration/decompo-
Scheme 2
sition process, i.e. for the elimination of one water molecule and three carbon dioxide
molecules per citric acid molecule.
The nifedipine presents a melting step with the endothermic maximum between
Scheme 3
110–150ºC. The decomposition occurs due to the oxidation process with maximum
in the range 200–240ºC.
The acetyl salicylic acid presents a first decomposition step with rate maximum
between 110 and 140ºC (dependent on heating rate). A relatively good accordance
Scheme 4
exists between the experimental mass loss and that one determined for the elimina-
tion of acetic acid.
Kinetic analysis
From the generally accepted equation of the non-isothermal kinetics [3]
dα
dT
E
β
= f (α )Aexp −
(3)
RT
(where β is the heating rate and T is the temperature in K), the equation corresponding
to the Friedman’s differential isoconversional method, was obtained
J. Therm. Anal. Cal., 72, 2003

Table 2 Activation energy (E10–5 J mol–1) obtained by Friedman method for heating rates of 2.5, 5, 10 and 20 K min–1
α
Compound
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
Na2S2O5
0.64
0.68
0.70
0.71
0.71
0.70
0.67
0.62
0.51
Na(CH3COO)
K(CH3COO)
0.21
3.08
0.18
4.16
0.16
5.27
0.15
6.16
0.13
8.13
0.12
10.10
0.10
11.90
0.08
8.72
–
–
Glutamic acid
Citric acid
0.66
1.87
1.08
0.56
0.62
1.67
1.18
0.59
0.61
1.56
1.22
0.62
0.60
1.49
1.25
0.64
0.61
1.43
1.24
0.66
0.63
1.40
1.21
0.67
0.66
1.40
1.12
0.68
0.75
–
1.15
–
Nifedipine
0.96
0.68
0.58
0.49
Acetylsalicylic acid

VLASE et al.: SPECIFICITY OF DECOMPOSITION
601
dα
dT
E
ln β
=ln[A f (α )]−
(4)
RT
dα
dT
For α=ct and using various heating rates, the plot ln β
vs. (1/T) should be
linear. From the slope and the intercept of the straight line the value of activation en-
ergy and product [Af(α)] were obtained. This enable E(α) and A(α) to be estimated
using various values of α.
The procedure is based on the following assumptions [4–6]:
1) E and A depend on the degree of conversion and do not depend on the heating
rate;
2) E and A are correlated through the compensation effect (CE)
lnA=aE+b
(5)
(6)
3) The dependence of E on the degree of conversion is given by
E=E0+E1ln(1–α)
where E0 and E1 are constants.
4) The differential conversion function, f(α) has the form
f(α)=(1–α)n
(7)
From the differential conversion function (7) it follows that
ln[Af(α)]=lnA+nln(1–α)
(8)
and from the values of [Af(α)] and α, the values of lnA corresponding to various val-
ues of n can be obtained. The plot of lnA vs. E should be linear as required by the exis-
tence of compensation effect (Eq. (5)). The correct value of n will be the one which
gives a correlation coefficient closest to 1.00 for the straight line lnA vs. E.
The presence of the kinetic compensation effect presumes the existence of an
isokinetic temperature which can be estimated from the slope of the straight line lnA vs. E.
1
Ti =
(9)
Ra
From Eq. (3)–(7) we obtain:
E0
E1
dα
ln β
= b+aE0 −
+ aE1−
+n ln(1−α )
(10)
dT
RT
RT
The values of the activation energy obtained according to Eq. 4 by four heating
rates at different reaction degrees are presented in Table 2. A significant variation of
E vs. α is observed. By means of the kinetic analysis discussed before (Eq. (5)–(10)),
the data in Table 3 were obtained. The reaction order was tested for values between
0.1–2.0, with the step of 0.1.
J. Therm. Anal. Cal., 72, 2003

602
VLASE et al.: SPECIFICITY OF DECOMPOSITION
Table 3 Kinetic parameters according to Eq. 10
a 104/mol J–1
(Eq. 5)
b
E0/
E1/
Correlation
coefficient
Compound
Na2S2O5
n
kJ mol–1 kJ mol–1
72.5
6.9
3.12
–3.03
0.3
0.9992
Na(CH3COO)
K(CH3COO)
19.3
660.0
7.2
22.3
1.41
1.79
1.83
1.07
1.6
1.9
0.9992
0.9999
Glutamic acid
Citric acid
51.9
171.0
131.0
–20.4
25.8
25.0
2.16
2.96
2.37
0.18
–4.72
–0.96
0.1
1.2
0.9
0.998
0.99996
0.9998
Nifedipine
Acetylsalicylic
acid
63.7
1.6
3.09
–2.64
1.5
0.998
Specificity by non-isothermal decomposition
The specificity by decomposition under non-isothermal conditions is due to a selec-
tive vibrational energy accumulation on a certain bond. This breaking bond is assimi-
lated with a Morse oscillators [7] coupled non-linear [8] with the harmonic oscillators
of the thermic field. Following a theoretical treatment developed in our papers [2, 9],
a correspondence will be established between the kinetic parameter Ti and the wave
number of the activated bond:
k
ωcalc
=
B Ti =0.695Ti
(11)
hc
with kB, h – the Boltzmann, respective Planck constants, and c – the light velocity.
Because the breaking bond has an unharmonic behavior, the specific activation is
possible also due to more than one quanta, or by a higher harmonic: ωsp =qωcalc, q N,
where ωsp is the assigned spectroscopic number for the bond supposed to break.
In order to corroborate the calculated data with the spectroscopic ones, we drew
up the FT-IR spectra of the studied compound.
In Table 4 the ωcalc values with the ωsp determined from the spectra, together
with the assignments of the corresponding oscillations are compared. The studied
compounds exhibit a very good agreement between the wave number calculated from
the isokinetic temperature and the corresponding wave number accepted in IR spec-
troscopy for the bonds suggested to be broken.
Conclusions
The kinetic analysis of the decomposition of some food additives and pharma-
ceuticals, under non-isothermal conditions, allows the evidence of a compensation
effect, for the kinetic data at different heating rates. By means of the thermodynamic
functions of the activated complex, a correlation between the isokinetic temperature
and the wave number assigned to the breaking bond is possible. It is possible to ana-
J. Therm. Anal. Cal., 72, 2003

Table 4 Comparison between kinetic and spectroscopic data
Ti/K
Eq. (9)
ω
calc/cm–1
Eq. (11)
ωsp/cm–1
present in FT-IR spectra and the assignments [10, 11]
qωcalc
Samples
Na2S2O5
q
4
6
1072
1608
900–1000
1600–1620 bisulphite anion
385
268
2
3
1182
1773
1330–1440
C=O νs by carboxyl anions
Na(CH3COO)
K(CH3COO)
Glutamic acid
851
591
467
387
672
3
1401
–
6
7
2322
2709
556.5
O–H in carboxyl group
1700 ν (C=O) by sat, carboxylic acid
3000 OH of carboxylic group in H bond
3500–3550 OH in carboxylic group
Citric acid
405
282
353
271
6
1692
3
4
5
1056
1408
1760
1250–1300 esters of aromatic acids
1580–1650 C=C and C–H stretching
1370–1385 (C–CH3) νs; 1435–1470 (C–CH3) νas
Nifedipine
507.6
389.7
4
5
6
1084
1355
1626
1230–1250 acetates
1680–1700 C=O stretching in acyl acid
1770–1800 esters with phenyl group
Acetylsalicylic acid

604
VLASE et al.: SPECIFICITY OF DECOMPOSITION
lyze the thermal ‘sensitive’ part of a molecule, by means of an adequate processing of
the thermogravimetric data, in connection with the IR spectra.
References
1 H. Ferrari, ‘Thermal Analysis’ ed. R. F. Schwenker and P. D. Garn, Academic Press,
New York 1969, Vol.1, p. 41.
2 T. Vlase, G. Jurca and N. Doca, Thermochim. Acta, 379 (2001) 65.
3 P. Budrugeac and E. Segal, Int. J. Chem. Kinetic, 33 (2001) 564.
4 P. Budrugeac and E. Segal, Thermochim. Acta, 260 (1995) 75.
5 P. Budrugeac, J. Therm. Anal. Cal., 64 (2001) 821.
6 P. Budrugeac, J. Therm. Anal. Cal., 66 (2001) 557.
7 G. Herzberg, ‘Molekülspektren und Molekülstruktur. I. Zweiatomige Moleküle’, Steinkopff,
Dresden 1939.
8 N. B. Brandt and S. M. Chudinov, ‘Electronic Structure of Metals’ MIR Publishers,
Moskow 1975.
9 T. Vlase, G. Vlase and N. Doca, J. Therm. Anal. Cal., in press.
10 R. M. Silverstein, G. Clayton Bassler and T. C. Morrell, ‘Spectrometric Identification of the
Compound’, 5th Ed., Wiley 1995.
11 N. B. Colthup, L. H. Daly and S. E. Wiberley, ‘Introduction to Infrared and Raman Spectros-
copy’, Academic Press, New York 1964.
J. Therm. Anal. Cal., 72, 2003