Differential thermal analysis under quasi-isothermal, quasi-isobaric conditions (Q-DTA): Part III. Mechanism of congruent and incongruent phase transformations of salt hydrates

Article abstract of DOI:10.1016/j.tca.2004.12.011

By using the "simultaneous Q-DTA, Q-TG measuring technique" elaborated recently, conditions near to the requirements of thermodynamics can be created, thus the "normalized" course of curves taken by this method, their characteristic temperatures or the va

Full text of DOI:10.1016/j.tca.2004.12.011

Thermochimica Acta 430 (2005) 59–65
Differential thermal analysis under quasi-isothermal,
quasi-isobaric conditions (Q-DTA)
Part III. Mechanism of congruent and incongruent
phase transformations of salt hydrates
F. Paulik∗, E. Bessenyey-Paulik, K. Walther-Paulik
Institute for General and Analytical Chemistry, Technical University Budapest, Budapest 1521, Hungary
Received 30 August 2004; received in revised form 19 December 2004; accepted 22 December 2004
Available online 25 January 2005
Abstract
Byusingthe“simultaneousQ-DTA, Q-TGmeasuringtechnique”elaboratedrecently, conditionsneartotherequirementsofthermodynamics
can be created, thus the “normalized” course of curves taken by this method, their characteristic temperatures or the values measured for the
enthalpy change can be considered to be near to the theoretical values.
For testing this new method, CuSO4·5H2O, MnSO4·5H2O, Mg(NO3)2·6H2O and Na2SO4·10H2O have been investigated. These model
substances can be prepared in stoichiometric composition free of impurities, their decomposition mechanism is well-known, thus the effect
of the altered experimental conditions by the simultaneous Q-DTA, Q-TG method on the course of their dehydration can be calculated.
© 2005 Elsevier B.V. All rights reserved.
Keywords: Q-DTA and Q-TG measuring technique; Transformation-governed heating control (TGHC); Self-generated atmosphere (SGA)
1. Introduction
and “dynamic” heating control. This kind of heating con-
trol is also called “transformation-governed heating control”
(TGHC) referring to its operation principle. This name is
correct, as e.g. the heating control of the Q-TG instrument
is governed by the strictly constant and very low tempera-
ture difference between the sample and the furnace by the
feed-back principle, which difference is proportional to the
heat flux, which in turn, is proportional to the heat absorbed,
that is proportional to the rate of the transformation, being
proportional to the change in the potential of the DTG signal.
The “Q-TG method” applying “quasi-isothermal heating
control” [1–3] has already a past of four decades. After the ap-
plication for the invention [4], more papers were published
introducing measuring techniques (constant decomposition
rate thermal analysis (CRTA) [7,8], stepwise isothermal anal-
ysis [9]) which, similarly to the Q-TG method, intended also
to eliminate the disadvantages of the non-isothermal heat-
ing control, but they differed from it as well in the mode
The unusual name of “quasi-isothermal, quasi-isobaric
measuring technique” (Q-TG, Q-EGA, Q-TD) [1–3] requires
some explanation. This name should emphasize the fact that
in these investigations a heating technique and a sample
holder with a specific form are applied which are totally dif-
ferent from the traditional ones.
The authors of the Q-TG method [1–3] applied first
the heating control named “quasi-isothermal” [1–4]. This
method unifies the advantages of the traditional “isothermal”
and “non-isothermal” heating control modes without having
their disadvantages. The heating control with the name of
“quasi-static” [2,6] represents a transition between “static”
∗
Corresponding author at: H-1022 Boga´r Street 18/b, Budapest, Hungary.
Tel.: +36 1 3268167; fax: +36 1 3268167.
E-mail address: bessenyey.paulik@axelero.hu (F. Paulik).
0040-6031/$ – see front matter © 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.tca.2004.12.011

60
F. Paulik et al. / Thermochimica Acta 430 (2005) 59–65
of heating control (heating mode governed by the change in
the decomposition pressure under atmospheric depression, or
alternately changing isothermal and non-isothermal heating
mode, etc. [7–9]) as well as in experimental conditions, con-
sequentlyalsointheexperimentalresults. Heatingtechniques
notfollowingthetraditionalwayobtainedthecollectivename
of “quasi-static” methods [2,13].
pounds, namely, seem to be suitable for testing the “simul-
taneous Q-DTA, Q-TG method”. The results of these studies
are provided in the second part (Part II) of this series [13].
In this paper, for the same purpose, the mechanisms of
the congruent and incongruent transformations of several salt
hydrates are studied. The results of these experiments are
summarized in the present paper.
The name “quasi-isobaric” used together with “quasi-
isothermal” indicates that—instead of traditional open sam-
ple holders—a specific, so-called “labyrinth” sample holder
[1,5] is used, from which, immediately at the start of decom-
position, the decomposition products sweep along the air and
prevent its flowing back. Thus the decomposition reactions
can be studied at a constant pressure of 100 kPa of the so-
called “self-generated atmosphere” (SGA).
The authors of the present paper applied a new, recently
elaborated measuring technique in their investigations, the
so-called “simultaneous Q-DTA, Q-TG method” [10–13].
This method differs from the Q-TG method used success-
fully earlier in two points. The heating control system of the
simultaneous Q-DTA, Q-TG instrument is governed by the
Q-DTA signal, i.e. by the rate of enthalpy change, and not by
the Q-DTG signal (Fig. 1). The other point is that with this
instrument, after its simple calibration, the enthalpy changes
of transformations can be measured simply and accurately
[6,12,13] (Fig. 2b). The principal basis for the method and
the implementation of measurements are described in the first
part (Part I) of this series [6].
2. Experimental conditions
Dehydration of salt hydrates CuSO4·5H2O (Figs. 1 and 2),
MnSO4·5H2O (Fig. 3), Mg(NO3)2·6H2O (Fig. 4) and
Na2SO4·10H2O (Fig. 5) were studied, once by using the tra-
ditional simultaneous TG, DTG, DTA [1] method, and then
by the simultaneous Q-DTA, Q-TG [12] method. Results of
the former studies are shown at the left side of the figures
(Figs. 2a–5a), whereas those of the latter are found at the
right side of the same (Figs. 2b–5b).
Measurements were carried out by the instrument
Derivatograph C (MOM, Hungarian Optical Works, Bu-
dapest), system Paulik, Paulik and Erdey. This instruments
was fabricated originally for performing the “simultane-
ous TG, DTG, DTA, EGA” and in an alternative way, for
the “quasi-isthermal, quasi-isobaric thermogravimetric” (Q-
TG) measurements [1], but later it was made suitable for
carrying out “simultaneous Q-DTA, Q-TG” [10–13] investi-
gations, as well.
If we want to prove that a task can be solved authentically
by using a new measuring technique elaborated with the in-
tention of improvement, then we have to study compounds
preparable with a stoichiometric composition, and the crystal
structure of which, as well as every detail of its decomposition
mechanism are well known.
The “simultaneous Q-DTA, Q-TG instrument” is an
adapter by which not only Derivatograph C is complemented,
but it can be connected to any “simultaneous TG, DTA in-
strument” in an alternative way. This instrument is based
on two inventions [10,11]. Fig. 1 was taken by the measur-
ing technique according to the first invention [10], whereas
Figs. 2b–5b according to the second one [11]. Fig. 1 is a not
transformed, original curve, only some data are drawn on it
This is why the authors have chosen for this purpose salt
hydrates containing structural and crystal water. These com-
Fig. 1. Original record. Study of the dehydration of CuSO4·5H2O.

F. Paulik et al. / Thermochimica Acta 430 (2005) 59–65
61
Fig. 2. Dehydration of CuSO4·5H2O studied by the traditional DTA, TG instrument (a) and by the simultaneous Q-DTA, Q-TG instrument (b).
Fig. 3. Dehydration of MnSO4·5H2O studied by the traditional DTA, TG instrument (a) and by the simultaneous Q-DTA, Q-TG instrument (b).
later. In Fig. 1, taken according to invention in [10], i.e. by
using “transformation-governed heating control” [1a,2,10]
and in “self-generated atmosphere” [1b,2,10], the enthalpy
change (curve Q-DTAt), weight change (curve Q-TGt) and
temperature change (curve Tt) of the sample (s) is registered
in function of time (t). The above curves are let to be regis-
tered by the adapter only for special request, otherwise only
the end result of the measurement is registered, i.e. curves
the samples (ꢀms) expressed in the weight change (g) of 1 g
the sample, also as a function of sample temperature.
3. Dehydration of CuSO4·5H2O
n
n
The Q-TAT and Q-TGT curves taken by the simulta-
s
s
neous Q-DTA, Q-TG method shown in Fig. 2b illustrate
the tearing down and removal of the first four crystal
water molecules of CuSO4·5H2O. According to curve 3,
CuSO4·5H2O melted incongruently at 97 ◦C, and solid
CuSO4·3H2O and a saturated solution (Ls1at) were formed
(Fig. 2b, curve 3, sections a–b). While the temperature
increased from 97 to 103 ◦C, the saturated solution Ls1at dis-
solved a further portion of CuSO4·3H2O (sections b–c). The
n
n
Q-TA and Q-TG are drawn which are more easy to in-
Ts
Ts
terpret and can be more accurately evaluated.
n
The Q-TA curves in Figs. 2b–5b provide the enthalpy
Ts
change of the transformations in kJ mol−1, whereas their
course shows the course of the transformations under nor-
malized (n) conditions in function of the sample (s) temper-
ature (Tn).Q-TG curves show the change in the weight of
n
Ts
s

62
F. Paulik et al. / Thermochimica Acta 430 (2005) 59–65
Fig. 4. Dehydration of Mg(NO3)2·6H2O studied by the traditional DTA, TG instrument (a) and by the simultaneous Q-DTA, Q-TG instrument (b).
saturated solution Ls2at started to boil at 103 ◦C, then, without
a change in temperature, it evaporated, the intermediate
CuSO4·3H2O crystallized out in its whole mass (sections
c–d). The isothermal course of sections c–d in curve 3 proves
that the solution Ls2at remains saturated. During boiling,
namely, simultaneously with the amount of water vapour
leaving the system, an amount of CuSO4·3H2O proportional
to the amount of the evaporated water was precipitated
(sections c–d), thus till the end of the process the concen-
tration, and consequently the boiling point of Ls2at did not
change. The intermediate solid CuSO4·3H2O decomposed
to CuSO4·H2O and water vapour at 116 ◦C (sections e–f).
The proper separation of the partial processes made it pos-
sible to determine the enthalpy changes of all the three partial
sections e–f CuSO4 · 3H2Os
→ CuSO4 · H2Os + 2H2Og ↑ · · · (+Q4)
Between 97and 103 ◦C, as well as between 103 and 116 ◦C,
the system absorbed an amount of heat corresponding to its
heat capacity (Q5, Q6). This equals to the product of the aver-
age value of the molar heats (C1, C2) and the corresponding
temperature range (T2–T1, T4–T3).
¯
sections b–c Q5 = C1(T2 − T1)
¯
sections d–e Q6 = C2(T4 − T3)
The amounts of heat proportional to the heat capacity of
the samples, however, are usually so small that—as it is the
case here—they cannot be observed on the Q-TAT curves,
n
processes from the course of the Q-TAT curve.
n
s
s
The fact that the solution Ls2at contained only a relatively
therefore—similarly to dissolution heats (Qs)—they cannot
small amount of CuSO4·3H2O is made probable by its boil-
ing point not being significantly different from that of wa-
ter. Since the amount of CuSO4·3H2O held dissolved in L2sat
was small, the dissolution heat might also be small, there-
fore no heat uptake could be recorded in sections b–c of
curve 3.
be indicated. Therefore, for the sake of simplicity, further on
they will be neglected.
Whereas on curve 3 in Fig. 2b the incongruent phase trans-
formation can be shown unambiguously, there is no trace of
such a process on the DTA curve in Fig. 2a. According to
experience [1a], this process can only be recorded by a DTA
investigation by using special experimental conditions.
sections a–b CuSO4 · 5H2Os
→ (1 − x)CuSO4 · 3H2Os
+Ls1at(= xCuSO4 · 3H2O + 2H2Ol) · · · (+Q1)
4. Dehydration of MnSO4·5H2O
Curves 3 and 4 in Fig. 3b illustrate the course of the
dehydration of MnSO4·5H2O under “normal” conditions:
as is seen the salt hydrate melted at 45 ◦C producing solid
MnSO4·4H2Os and a saturated solution (Ls1at) (Fig. 3b, curve
3, sections a–b). At 75 ◦C, the MnSO4·4H2O decomposed
further incongruently, and MnSO4·H2O was formed, and the
composition of the solution phase was changed (Ls2at) (sec-
tions c–d). Then, the solution containing further only small
amounts of the intermediate salt hydrate Ls2at started to boil
sections b–c (1 − x)CuSO4 · 3H2O + L1sat
→ (1 − y)CuSO4 · 3H2Os
+Ls2at(= yCuSO4 · 3H2O + 2H2Ol) · · · (+Q2)
sections c–d (1 − y)CuSO4 · 3H2Os + L2sat
→ CuSO4 · 3H2Os + 2H2Og ↑ · · · (+Q3)

F. Paulik et al. / Thermochimica Acta 430 (2005) 59–65
63
sections e–f (1 − y)MnSO4 · H2Os + L2sat
→ MnSO4 · H2Os + 4H2Og ↑ · · · (+Q3)
and evaporate at 103 ◦C. As the MnSO4·H2Os intermediate
precipitated from the Ls2at solution in an amount proportional
to the evaporating water vapour also in this case, the boil-
ing point of the solution did not change until the temperature
reached point fꢀ.
sections f–g MnSO4 · H2Os
As an end of evaporation, the remaining water left Ls2at be-
tween points fꢀ–f, whilst the temperature of the MnSO4·H2Os
and Ls2at mixture being in a mass ratio of 20:1 increased from
103 to 270 ◦C. It can only be guessed why the temperature of
the mixture should be elevated by nearly 170 ◦C for the water
to leave the system (sections fꢀ–f). At 270 ◦C, MnSO4·H2Os
decomposed to MnSO4s and water vapour. This process was
also anomalous. The process was introduced by nucleus for-
mation. At the surface of the crystals, the number of nuclei
is usually small. The temperature of the sample should be
increased in order to reach the prescribed rate for the decom-
position. At the point, however, when the number of nuclei
is large enough, the temperature falls back spontaneously
to the value required for the transformation to proceed under
equilibrium conditions, with the rate prescribed. This process
s
→ MnSO4 + H2O ↑ · · · (+Q4)
5. Dehydration of Mg(NO3)2·6H2O
Curves 3 and 4 in Fig. 4b illustrate the course of water
loss of Mg(NO3)2·6H2O when this process is studied by
the simultaneous Q-DTA, Q-TG method. According to cu◦rve
n
Q − TAT , the salt hydrate melted incongruently at 95 C,
s
and an unsaturated solution (Lu1nsat) appeared, and the simul-
taneously formed Mg(NO3)2·2H2O immediately dissolved
in it (sections a–b). The solution phase did not loose water
till 150 ◦C (sections c–d), which fact could be proven from
n
the Q-TGT curve.
s
The solution started to boil at 150 ◦C (point c). After that,
n
used to produce a beak-like formation on both the Q-TAT
s
the boiling point of the solution was continuously raised, as
the system lost water, its concentration increased. This caused
the inclined course of sections c–d of curves 3 and 4, which is
an unambiguous sign for the solution being still unsaturated
(Lu2nsat). A turn in this process occurred at point d of curves
3 and 4. Namely, the unsaturated solution at 280 ◦C became
saturated (Lu2nsat → Ls1at). Following this, neither the boil-
ing point of the solution, nor its concentration had further
changed meaning that solid Mg(NO3)2·2H2O precipitated
from the solution in an amount proportional to the amount
of water vapour leaving the system (sections d–e). This in-
n
and Q-TGT curves [1b].
s
sections a–b MnSO4 · 5H2Os
→ (1 − x)MnSO4 · 4H2Os
+Ls1at(= xMnSO4 · 4H2O + H2Ol) · · · (+Q1)
sections c–d (1 − x)MnSO4 · 4H2Os + L1sat
→ (1 − y)MnSO4 · H2Os
+Ls2at(= yMnSO4 · H2O+ + 4H2Ol) · · · (+Q2)
termediate decomposed to Mg(NO3)2 and water vapour at
s
Fig. 5. Dehydration of Na2SO4·10H2O studied by the traditional DTA, TG instrument (a) and by the simultaneous Q-DTA, Q-TG instrument (b).

64
F. Paulik et al. / Thermochimica Acta 430 (2005) 59–65
390 ◦C (sections f–g).
Namely, the “simultaneous Q-DTA, Q-TG instrument”
provides experimental conditions being near to the require-
ments of thermodynamics [12], thus processes free of for-
eign effects and leading to equilibrium proceed always
strictly isothermally, and at a characteristic temperature be-
ing consequently the same and close to the theoretical value
(Figs. 2b–5b).
sections a–b Mg(NO3)2 · 6H2Os
→ Lu1nsat(= Mg(NO3)2 · 2H2O + 4H2Ol) · · · (+Q1)
sections c–d L1unsat
However, isothermal transformations can be influenced
by additionally proceeding foreign processes, but without
changing the character of the reaction of leading to equi-
librium [1c]. Such a process might be e.g. nucleus formation
[1b] (Fig. 3b, curve 3, sections f–g), the delaying effect of a
new solid phase on the leaving of the gaseous decomposition
product [1d,6], modification of curves indicating the forma-
tion of intermediates [12], the evaporation of an unsaturated
solution (Fig. 4, curve 3, sections c–d) [12], etc.
The irreversible transformations (decomposition of plas-
tics, solid phase reactions, etc.) or the reversible but non-
isothermal processes traceable to concentration changes in
multicomponent systems [13], etc.) proceed always conse-
quently non-isothermally, in the same temperature range with
an unchanged course. Such processes are normalized by us-
ing the TGHC technique, but they are not influenced by the
partial pressure of the gaseous decomposition products and
the form of the sample holder.
→ Lu2nsat(= Mg(NO3)2 · 2H2O + (4 − x)H2Ol)
+xH2Og ↑ · · · (+Q2)
point d Lu2nsat
→ Ls1at
sections d–e Ls1at
→ Mg(NO3)2 · 2H2Os + (4 − x)H2Og ↑ · · · (+Q3)
sections f–g Mg(NO3)2 · 2H2Os
s
→ Mg(NO3)2 + 2H2Og ↑ · · · (+Q4)
6. Dehydration of Na2SO4·10H2O
On the traditional DTA, DSC curves, the course of trans-
formations and their characteristic temperature change with
the experimental conditions, therefore the value of transfor-
mation heat determined also changes (Kirchhoff’s law). To
The dehydration of Na2SO4·10H2O is, according to
curves 3 and 4 in Fig. 5, a simple and unambiguous process.
The salt hydrate melted incongruently at 32 ◦C, i.e. it decom-
posed to Na2SO4s and a solution Ls1at (Fig. 5, curve 3, sections
a–b). The solution Ls1at of relatively low concentration started
to boil at 104 ◦C, and without a change in temperature it evap-
orated totally (sections c–d). A sure sign of being this process
isothermal was that the solution remained saturated.
n
the opposite, the course of Q-TAT curves and the character-
istic temperature determined undesr normalized conditions do
not change. The value of enthalpy changes read from these
curves remain consequently constant [12] and reproducible.
The enthalpy changes of partial processes of complex re-
actions can also be read easily and with good accuracy from
sections a–b Na2SO4 · 10H2Os
n
the Q-TAT curves (Figs. 2b–5b).
s
Thoughstransformations proceed with a rate by orders of
magnitude smaller by applying TGHC than in traditional
DTA, DSC measurements using non-isothermal heating con-
trol, the total time for taking the curves by the new measur-
ing technique increases only by a factor of two. Only the
quasi-isothermal heating mode is time-consuming, as be-
fore and after the transformation the heating control may
increase the temperature in a non-isothermal way, even by
20–30 ◦C min−1 [6,12,13]. Theheatingcontrolswitchesfrom
isothermal to non-isothermal heating mode and vice verse
fully automatically.
→ (1 − x)Na2SO4
+Ls1at(= Na2SO4 + 10H2Ol) · · · (+Q1)
s
s
sections c–d (1 − x)Na2SO4 + L1sat
s
→ Na2SO4 + 10H2Og ↑ · · · (+Q2)
7. Conclusions
On the traditional DTA, DSC curves, the course of physi-
cal and chemical transformations leading to equilibrium ap-
pears always deteriorated. Due to the open sample holder
and the use of non-isothermal heating control, transforma-
tions proceed always in a non-isothermal way, in a more or
less broad temperature range producing the deceptive appear-
ance that they are not leading to equilibrium and they are
irreversible (Figs. 2a–5a).
References
[1] (a) F. Pasulik, Special Trends in Thermal Analysis, Wiley, Chich-
ester, 1995, p. 121 ;
(b) F. Pasulik, Special Trends in Thermal Analysis, Wiley, Chich-
ester, 1995, p. 194 ;
These problems are totally eliminated by using the “simul-
taneous Q-DTA, Q-TG measuring technique” (Figs. 2–5).
(c) F. Pasulik, Special Trends in Thermal Analysis, Wiley, Chich-
ester, 1995, p. 105 ;

F. Paulik et al. / Thermochimica Acta 430 (2005) 59–65
65
(d) F. Pasulik, Special Trends in Thermal Analysis, Wiley, Chich-
ester, 1995, p. 191.
[8] J. Rouquerol, J. Therm. Anal. 2 (1970) 123.
[9] O.T. Sorensen, J. Therm. Anal. 13 (1978) 429.
[2] F. Paulik, J. Paulik, Thermochim. Acta 100 (1986) 23.
[3] F. Paulik, J. Therm. Anal. Calorim. 58 (1999) 711.
[4] F. Paulik, J. Paulik, L. Erdey, Patents, Hungary152197 (1962),
France, Switzerland, England, Austria, Italy, USA, Germany.
[5] F. Paulik, J. Paulik, Patents, Hungary 163305 (1971), USA, Japan,
Germany, France, Italy, Switzerland, England, Austria.
[6] F. Paulik, E. Bessenyey-Paulik, K. Walther-Paulik, Thermochim.
Acta 424 (2004) 75.
[10] J. Paulik, F. Paulik, M. Arnold, Patents, Hungary 194405 (1985),
USA, England, France, Germany, Switzerland.
[11] F. Paulik, E. Bessenyey-Paulik, K. Walther-Paulik, Patents, Hungary:
HU 9700804/1997; European Patent: EP 0977983 B1/1998; USA
6,239,415 B1/1998.
[12] F. Paulik, E. Bessenyey-Paulik, K. Walther-Paulik, Thermochim.
Acta 402 (2003) 105.
[13] F. Paulik, E. Bessenyey-Paulik, K. Walther-Paulik, J. Therm. Anal.
Calorim. 58 (1999) 725.
[7] J. Rouquerol, Bull. Soc. Chim. Fr. (1964) 31.