Journal of Thermal Analysis and Calorimetry

https://doi.org/10.1007/s10973-018-7888-1(0123456789().,-volV)(0123456789().,-volV)

Thermal decomposition of calcium oxalate: beyond appearances

Djamila Hourlier1

Received: 27 May 2018 / Accepted: 28 October 2018

 Akadémiai Kiadó, Budapest, Hungary 2018

Abstract
The goal of this study is twofold: to take a fresh look at the decomposition of calcium oxalate and to warn users of
thermogravimetric analysis against the hasty interpretation of results obtained. Since the pioneer work of Duval 70 years
ago, the scientific community has agreed unanimously as to the decomposition of anhydrous calcium oxalate (CaC2O4) into
calcium carbonate (CaCO3) and CO gas, and that of the calcium carbonate into calcium oxide (CaO), and CO2 gas. We will
demonstrate how these reactions, simple in appearance, in fact result from a succession of reactive phenomena involving
numerous constituents both solid (CaCO3, free carbon) and gaseous (CO2 and CO) produced by intermediary reactions.
The mass losses evaluated in the two distinct domains correspond closely to the molar masses of CO and CO2, respectively.
The simple mathematical calculation of that mass loss has simply concealed the existence of other reactions, and, most
particularly the Boudouard reaction and that of solid phases between CaCO3 and C. It just goes to show that appearances
can be deceiving.

Keywords Thermochemistry  Solid decomposition  Metal oxalate  Whewellite  Mass spectrometry

Introduction oxalate constitutes more than 70% of the mass of a kidney

stone, a third of whose mass is formed of the monohydrate.
An incalculable number of studies have been carried out on An excessive presence of CaCO3 in the body increases the
the synthesis, characterization, and decomposition of cal- risk of urethral stricture, and consequently, the develop-
cium oxalate monohydrate (CaC2O4H2O) and the products ment of infections. It can similarly build up in other parts
of its decomposition, such as calcium carbonate (CaCO3, of the body, developing into pathological calcifications
i.e., calcite or limestone) and calcium oxide (CaO, i.e., (e.g., of the thyroid, breast, or prostate).
quicklime). In architecture, lichens that build up on supports con-
Usage of these calcium-based compounds is widespread, taining limestone cause the weathering of edifices by a
and thus, this research shall be of interest to a great many similar mechanism: These microorganisms secrete organic
sectors. acids that react with the limestone to form a thin, rich layer
Calcium oxalate monohydrate, whose mineralogical of calcium oxalate and thus instigate the degradation of the
name is whewellite, is as commonplace in kingdoms edifice [6].
of man [1], plants [2–5], and animals. If the scientific Another evocative example is the precipitation of cal-
community is very much invested in the study of this cium oxalate in the lime kilns of sugar manufacturers [7].
material, this interest is due to its relevance to problems in This problem is a major headache for the industry, posing
a great many fields. serious delays in production due to repeated and oftentimes
In humans, for example, resulting from crystallization of severe stoppages of long production lines.
mineral salts and organic acids present in the body, calcium In metrology, calcium oxalate is often used to calibrate
thermogravimetric instruments and even for mass spec-
trometers [8–12]. Its thermal decomposition is described by
& Djamila Hourlier three successive stages in the course of a linear increase in
djamila.hourlier@iemn.univ-lille1.fr temperature [13].
1
IEMN (CNRS, UMR 8520), Université Lille 1, Avenue
Poincaré, CS60069, 59652 Villeneuve d’Ascq Cedex, France

123
 D. Hourlier

1. the hydrated salt (CaC2O4H2O) yields anhydrous salt Citing the works of Simon and Newkirk [22] on the
(CaC2O4) and water vapor decomposition of the oxalate under two different gases (N2
CaC2 O4  H2 O ! CaC2 O4 þ H2 O ð1Þ and O2), Vallet [27] has the distinction of having raised in
his book some pertinent questions, concerning mass vari-
2. the anhydrous salt (CaC2O4) forms solid calcium ation in the second domain in particular, where one notes a
carbonate (CaCO3) and CO gas rather marked shift in the relative curvature toward lower
CaC2 O4 ! CaCO3 þ CO ð2Þ temperatures in oxygen atmosphere when compared to
results obtained under a current inert atmosphere. The
3. the solid calcium carbonate forms solid calcium oxide authors attribute this shift to the heat given off by the
(CaO) and CO2 gas oxidation of CO produced by the reaction (2). This heat
CaCO3 ! CaO þ CO2 ð3Þ would help to raise the oxalate to a more elevated tem-
perature and would therefore accelerate decomposition of
These three distinct, and apparently simple, steps con- the oxalate CaC2O4.
stitute the model by which students are introduced to the In the last domain, at temperatures over 550 C, the shift
reactivity of solid and gaseous phases. Likewise, the is reversed and the relative curvature for treatment in
retailers of instruments of thermogravimetric instruments oxygen is slightly offset toward higher temperatures.
proudly display the quality and precision of their systems, Among the leads suggested, the pressure of CO2 would be
and reproducibility of results obtained thereby, presenting responsible for the delay in the formation of CaO. The
the thermogravimetry analysis (TGA) graph as a model of authors have suggested possible leads without really set-
the decomposition of the oxalate, with the three reactions tling on any explanation in particular, and thus, the ques-
enumerated above—it is even their selling point. tion lingers without a satisfactory response.
Yet on closer examination of the curves of thermo- Following this observation, we ask ourselves if we
gravimetric analysis and analyzing the gases of calcium might take advantage of the performance of modern devi-
oxalate decomposition, or of any other oxalate, certain ces, notably their sensitivity, to better understand the
doubts arise with regard to the simplicity of the reaction (2) decomposition of the anhydrous oxalate and to proffer an
evoked in the second domain of mass loss, between 350 explanation for the emission in the second domain of CO
and 550 C, which would correspond to the elimination of and CO2, often observed between 200 and 500 C but, alas,
CO and formation of CaCO3. Many studies [14–18] as yet poorly interpreted.
otherwise reveal an emission of CO2, almost concomitant
with that of CO. The different stages of CO2 evolution can
be clearly distinguished on the remarkable thermo- Materials and methods
gravimetry analysis coupled mass spectrometry (TGA/MS)
curves reported by Frost [19]. The work of Price [15], Thermogravimetry analysis (TG) (Netzsch STA449F3
Manley [17], and also that of Frost [19] are good references Jupiter apparatus) coupled with a quadrupole mass spec-
to throw light on the emission of CO2 in the temperature trometry (MS) (Aëolos QMS403D, 70 eV, Electron
range from 350 C to 550 C. Certain researchers Ionization) via a heated capillary system, was applied to
[17, 20–22] attribute the formation of CO2 to the dispro- monitor the decomposition of material during the annealing
portionation of CO following the reac- process and to determine the evolved gaseous species.
tion 2CO ? C ? CO2. Certainly, the presence of free Before each experiment, the TG system is first evacuated
carbon has been proven [15, 22], and the powder, initially and then flushed with ultrahigh purity of the gas that will be
white, become greyish at temperatures less than 600 C. used for the thermal treatment. The experiments were
Another explanation that has been advanced for the for- carried out under dynamic inert or reactive gas (helium:
mation of CO2 in the second domain of mass loss is the 99.999% purity or 2% O2/He: 99.999% purity) at a flow
oxidation of CO [16, 23, 24]. Oxygen is possibly arising rate of 90 mL min-1. The sample mass (CaC2O4H2O) of
from a leakage in the system. Other researchers [25] simply 30 mg, the same for all cycles of pyrolysis, is placed in a
ignore this CO2, doing away with their own experimental cylindrical aluminum crucible with a diameter of 17 mm
data in order to better harmonize their conclusions with the and a total height of 20 mm. The heating is achieved by a
accepted wisdom on this subject. In fact, in practice, the current of gas flow (90 mL min-1) at a heating rate of
disproportionation reaction of CO is revealed to be a very 15 C min-1.
slow reaction [26]. Some carbon monoxide, which, when
pure, is metastable between 350 and 550 C, can exist in
the presence of CO2 in this temperature range.

123
Thermal decomposition of calcium oxalate: beyond appearances

Results and discussion 0 0.0

– 10
Characterization of the starting calcium oxalate – 0.1

monohydrate powder – 20

DTG/% °C–1
Mass loss/%
– 0.2
– 30
The starting calcium oxalate monohydrate has been ana-
–12.2 – 0.3
lyzed using Raman spectroscopy and energy-dispersion – 40
0.02

–12.3
x-ray spectroscopy (EDX). 0.00 – 0.4
–12.4
From Fig. 1 (inset EDX analysis), it can be observed – 50
–12.5
that there is no impurity. –0.02
– 0.5
– 60 –12.6 1st run
According to the Raman assignment proposed by –0.04 2nd run
225 250 275 300 325 350 375 400 425
Petit et al. [28] for pure calcium oxalate monohydrate, the – 70
Temperature/°C
– 0.6
100 200 300 400 500 600 700 800
main characteristic bands are observed: two strong bands at Temperature/°C
1491 cm-1 and 1463 cm-1, and 1397 cm-1, 897 cm-1
corresponding to different ms (CC) ? ms (CO) on different Fig. 2 Thermogravimetry analysis: curve for powdered (CaC2O4H2-
planes with nearest neighboring water molecules. The O) in flowing inert gas (He) at the heating rate of 15 C min-1
bands at 597 cm-1, 522 cm-1, and 503 cm-1 are due to m

°C
41 °C
(CC) ? m (Ca–O) in two different planes. EDX analysis

44 °C
°C
5
4
9
3
51
37
(inset of Fig. 1) of the powder revealed that the only 0
10–8
detectable elements were Ca, C, and O, and no other ele-
ments higher than atomic number of calcium. – 10
10–9
Before presenting the results of thermogravimetric

Ion current/A
Mass loss/%

– 20
analysis coupled with mass spectrometry (TGA/MS), it is
– 30
always important to specify the experimental conditions m/z 18 10–10
employed. It is equally important to insist on the perfect – 40 m/z 28

replicability of thermogravimetric experiments completed m/z 17

– 50
in the same conditions, as shown in Fig. 2. m/z 44 10–11
– 60
m/z 12
Thermogravimetric analysis coupled with mass 1.5 × 10–12
100 200 300 400 500 600 700 800
spectrometry (TGA/MS) Temperature/°C

Profiles of the thermogravimetric analysis curves coupled Fig. 3 Thermogravimetry coupled with mass spectrometry analysis
of powdered (CaC2O4H2O) in flowing helium atmosphere, at the
with mass spectrometry (Fig. 3) show that the decompo-
heating rate of 15 C min-1 (the signal intensity of the ions is on the
sition proceeds in three distinct stages: log scale)

The first step runs between 45 and 250 C and corre-

sponds to 12.33% of mass loss. This loss of mass is prin-
EDX
25 1463 cipally linked to the loss of H2O (reaction 1) as evidenced
20
by analysis of the gases by mass spectrometry, with the
appearance of the parent peak at m/z = 18 (H2O?) and the
Intensity/arbitrary units
cps/ev

C
15 O Ca
ca
fragment peaks at m/z = 17, 16 (see Table 1). For
10
1491 D1G = 0, PH2 O = 1 bar at T = 317 K (41 C), the water
5

0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5
K/eV Table 1 Relative intensities of the molecular ion and fragments [30]
1630
897
Molecule m/z
503
599 863 1397 12 14 16 17 18 22 28 29 44

H2O 1.0 21.0 100.0

400 600 800 1000 1200 1400 1600 1800
CO 5.0 1.0 2 100.0 1.2
Raman shift/cm–1
CO2 6.7 9.5 1.0 8.2 1.0 100.0
Fig. 1 Raman spectra and EDX (inset) acquired from starting calcium
oxalate monohydrate powder

123
 D. Hourlier

would apparently leave at a temperature just above 41 C, The CO2 is evacuated by the purge gas (helium or
which is confirmed by mass spectrometry where one nitrogen), while a new solid carbon phase is formed within
observes the starting signal emission from the peak at m/ the powder. This easy reaction, even at temperatures below
z = 18 toward 48 C. 250 C, would also explain the slight emission of CO2
CaC2 O4  H2 O ! CaC2 O4 þ H2 O ð1Þ observed during the dehydration of (CaC2O4H2O) occur-
red in the first step. The emission of CO2 at m/z = 44 slows
D1 G ðKJ mol1 Þ ¼ 36400114:8T
toward 374 C, and this is expressed by a first shoulder on
The maximum rate of loss of H2O stands toward 187 C. the profile of the mass spectrometry curve. This slowdown
One similarly notes a feeble variation in signals at masses is likely due to a chemical disequilibrium caused by the
m/z = 12, 28, 44. These masses correspond to a very slight appearance of a gas generated by a new chemical process,
emission of CO2; parent peak m/z = 44 and its fragment which we shall develop upon below.
peaks at m/z = 12, 16, and 28. Some traces of hydrogen at From 374 to 419 C, the release of CO2, after a first
m/z = 2 are also observed. Recently [29], it was shown that slowdown, starts rising again, as does the release of CO.
other molecules such as formic acid also are formed from The intensity of the peak at m/z = 28 rises and surpasses
the degradation of the oxalate counterion in the presence of the intensity expected for simple fragmentation of CO2.
H2O at approximately 200 C. The carbon monoxide CO is therefore a product generated
Next comes the second step which runs from 250 to by the sample within the furnace and not formed by the
550 C, during the course of which there is a mass loss of simple fragmentation of CO2 in the ionization chamber
19.32%, often attributed to the loss of CO. Certainly, the after electron bombardment. In other words, parallel to
simple mathematical calculation shows that the percentage reaction (4), the new chemical process producing CO could
of 19.32% is approximately equal to the mass of one mole correspond to the reaction (2), which is well known and
of CO. This is but a coincidence, as the gases of decom- often referenced in the literature.
position are not uniquely CO (m/z = 28); one also notes an CaC2 O4 ! CaCO3 þ CO ð2Þ
emission of CO2 (m/z = 44).
D2 G ðKJ mol1 Þ ¼ 80000152:8T
On closer inspection of the profile of CO2 at m/z = 44,
one sees that in reality the emission of CO2, already After 400 C, one again notes a slowing of CO and CO2
observed in the first domain during the dehydration of the emissions as evidenced by the shouldering toward 419 C.
oxalate, continues throughout the heating with cycles of This slowdown is once again caused by the existence of
acceleration and slowing. another chemical process, which comes to disturb the first
To better understand these reactional phenomena, it two processes (reactions 4 and 2).
would be judicious to break this domain down into a From 419 to 443 C, after the slowing observed, one
number of smaller temperature ranges: notes a light resumption toward 424 C in the emission of
Between 250 and 374 C, the overall TGA profile of species CO and CO2 and intensities then attenuate toward
Fig. 2 might appear to plateau and thus give the impression 440 C and rise again toward 443 C. This gentle slowing
that nothing is happening. In reality, by expanding the scale that covers a rather narrow temperature interval (of the
(curve inserted in Fig. 2) one sees that the oxalate con- order of 25 C) could fit the Boudouard reaction (reac-
tinues to lose mass. This loss is only very slight tion 5) between the gaseous phase constituted of CO and
(* 0.22%), and during this period only CO2 (m/z = 44) is CO2, and the solid phase C (free carbon).
detected by mass spectrometry. The weak variation of the 0:5ðCO2 þ CÞ ! CO ð5Þ
peak at m/z = 28 corresponds not to a principal product of
D5 G ðKJ mol1 Þ ¼ 8551587:33T
the decomposition of the oxalate but to the fragment peak
of CO2 generated in the ionization chamber after electron From 443 to 550 C, the emissions of gases CO and CO2
bombardment. reach a maximum rate of emission toward 515 C and then
This emission of CO2 could be explained by the attenuate progressively until the disappearance of CO
reaction (4): toward 550 C, while there remains a notable partial
CaC2 O4 ! CaCO3 þ 0:5ðCO2 þ CÞ ð4Þ pressure of CO2.
From 550 to 800 C, the emission of CO2 progressively
D4 G ðKJ mol1 Þ ¼ 551565:5T
increases until 850 C and is explained by the decompo-
In effect, this reaction, thus far ignored and never sition of the carbonate CaCO3 which leads to the formation
mentioned, is thermodynamically favorable, regardless of of calcium oxide CaO and CO2 as described in reaction (3).
temperature, as evidenced by the values of D4G. CaCO3 ! CaO þ CO2 ð3Þ
D3 G ðKJ mol1 Þ ¼ 168500143:5T

123
Thermal decomposition of calcium oxalate: beyond appearances

CaC2O4 = CaCO3 + 0.5CO2 + 0.5C

The asymmetry of the peak in CO2 suggests that the 150 CaC2O4 = CaCO3 + CO
reaction (3) is not the only process operating in the third 0.5C + 0.5CO2 = CO
0.5CaCO3 + 0.5C = 0.5CaO + CO
domain of mass loss, but that other reactions may be taking 100
CaCO3 = CaO + CO2
place, bringing reaction (5) into play, or the solid CaCO3 CO = 0.5CO2 + 0.5C

Δ G°/kJ mol –1
and free carbon according to reaction (6). 50

CaCO3 þ C ! CaO þ 2CO ð6Þ 0

D6 G ðKJ mol1 Þ ¼ 339530318:16T
– 50
The solid phase known as free carbon is also present
long after the second domain of mass loss, even unto the – 100
end of the thermal cycle as revealed by Raman spec-
troscopy (Fig. 4). The two broad humps at D * at – 150
1360 cm-1 and G * at 1585 cm-1 are the signature of 200 400 600 800 1000 1200

amorphous disordered carbon structure. The G and D bands Temperature/°C

are more easily detectable when the material is heated in Fig. 5 Temperature relationships of Gibbs free energy for the various
3%H2–argon atmosphere than in helium (see Fig. 4). The reactions involved in the decomposition of calcium oxalate
presence of carbon in the solid residue has already been monohydrate
reported by earlier workers [15, 22].
The variations in Gibbs energy of the reactions pro- From the Gibbs energy curve in Fig. 5, one sees that, in
posed, by way of explanation for the decomposition of the fact, if the decomposition of the oxalate into CaCO3 with
anhydrous oxalate and that of calcium carbonate, are rep- the emission of CO following the reaction (2) is easy, that
resented in the same graph (Fig. 5). Thus, one can gauge the emission of CO2 following the reaction (4) is even
the relative ease of carrying out such a reaction and com- easier. Moreover, until around 450 C, CO is not stable and
pare it to alternatives. would disproportionate into CO2 and C, though we know
It should be noted that there is very little thermodynamic that the disproportionation of CO is a slow reaction. Under
data in the available literature that describe calcium oxa- a current of the inert purge gas (helium), CO would be
late. The rare values found are those published by the evacuated before having the time to disproportionate and
Polish researchers [31, 32], and it is these, their theoretical produce CO2 and free carbon. One can thus affirm that the
values, that we have used, while taking care to eliminate carbon formed before 500 C cannot result from the dis-
the more fantastical interpretations which might have led to proportionation of CO, as Dollimore would suggest [33].
an excessive pressure of CO, for example. The carbon is not a by-product of the reaction, but rather a
product wholly of the reaction described by reaction (4).
We would not have given such importance to the
decomposition of the calcium oxalate if we had not the
G
coupling TGA/MS and the sensitivity of the modern
apparatus at our disposal, as well as our concern for both
D the precision of measurements and the calibration of our
systems of measurement. In effect, this reaction has been
widely studied and the reactional mechanisms are supposed
Intensity/arbitrary units

to be well established to the point that no-one contests the

H2 facts.
Following our study, a number of issues have been
raised, of which the two principal questions are:
1. Why the reaction (4) is never evoked while it appears
1086 cm– 1
CaCO3 to be easier than reaction (2), proposed in unanimity by
researchers?
He 2. Why have the mechanisms themselves never been
brought back into question when there is much
1000 1200 1400 1600 experimental evidence that would prove the presence
Raman shift/cm–1 of CO2 during the second domain of mass loss, at the
same time and even earlier than the emission of CO?
Fig. 4 Raman spectra acquired from (CaC2O4H2O) heated at
15 C min-1 in He up to 550 C, and in 3%H2/argon up to 730 C

123
 D. Hourlier

One can envisage at least two explanations: Effect of the oxidizing atmosphere
The first is perhaps linked to the representation of mass
spectra. In effect, a linear representation of the intensities The treatment of the powder has not been achieved in an
of mass spectrometry curves (Fig. 6) totally changes the inert atmosphere but in an oxidizing atmosphere (a mix of
perception of the decomposition of the material, compared 2% O2/He). All reactions mentioned above to explain the
with the logarithmic representation of curves (Fig. 3). decomposition of (CaC2O4H2O) in nitrogen remain valid
In each mass loss domain, one volatile compound is in oxygen atmosphere. However, one remark should be
dominant (Fig. 6), and its more prominent signal intensity made: There are subtle differences in their curves between
may mask that of the other species that may be released. the two atmospheres (Fig. 7). Whatever the atmosphere
On the other hand, the logarithmic representation (nitrogen or oxygen) in which the powder is heated, the
(Fig. 3) offers a finer description of the process of gas reaction (4) starts first and leads to two solids CaCO3 and C
formation brought on by the thermal decomposition of (free carbon), and also CO2 gas. The reaction (4) is then
(CaC2O4H2O), in which the importance of the species is followed by the well-known reaction (2) that gives off CO
weighted and the variations are more perceptible. and CaCO3. As has already been observed by Simon and
The other reason holds that in the second domain Newkirk [22], in the second domain there is a shift in the
between 400 to 550 C, the simple mathematical calcula- relative curvature toward lower temperatures in oxygen
tion for the mass loss coincides with the molar mass of CO. when compared to results obtained under a current of
It is true that the calculation for the mass loss, when con- nitrogen.
sidering the two reactions (4) and (5), comes to the same This shift, between 400 and 550 C, is explained by the
result as the consideration of reaction (2) alone. effect induced by the oxidation of CO given off CO2

Generally speaking, the thermal decomposition of oxa- (CO ? ‘O2 ? CO2, DHr ¼ 280:3 kJ mol-1 at 700 K).
late depends on many parameters such as the purity of the Indeed, the heat generated by the exothermic oxidation of
powder, its aging, and experimental conditions (crucible, CO brings the sample to a temperature higher than it would
flow rate of carrier gas, the nature of atmosphere, density of be in the absence of oxidation reaction of CO, and there-
gas (argon, helium, or nitrogen), heating rate, calibration of fore, the degradation of oxalate according to reaction (2)
analytical systems, and so on). Even though the TGA goes faster.
profile looks the same in most published studies, the It is interesting to point out the presence of another
evolved gases detected by mass spectrometry may have slight shift observed in the third domain of mass loss
different profiles and shapes. For all these reasons, the between 600 and 800 C in oxidizing atmosphere. The
results are hardly reproducible and led to misinterpretation. same shift was also obtained in other works by Simon [22]
All these parameters will be discussed soon in a future and Vallet [27]. This shift is related to the oxidation
paper. reaction of CO, as well.
In this third domain, as in nitrogen atmosphere, the two
solid phases CaCO3 and free carbon remaining in the powder
can react among themselves following the reaction (6) and

4.0 × 10–9
0 0
2% O2 in helium
Helium
– 10 – 10
m/z 3.0 × 10–9
12
Mass loss/%

– 20 17
– 20
Ion current/A

18
28
– 30 44
Mass loss/%

2.0 × 10–9 – 30
– 40
– 40
– 50 1.0 × 10–9
– 50
– 60

– 70 – 60
100 200 300 400 500 600 700 800
Temperature/°C – 70
100 200 300 400 500 600 700 800
Fig. 6 Thermogravimetry coupled with mass spectrometry analysis Temperature/°C
of powdered (CaC2O4H2O) in flowing helium atmosphere, at the
heating rate of 15 C min-1 (the signal intensity of the ions is on the Fig. 7 Thermogravimetry analysis: curve for powdered (CaC2O4H2-
linear scale) O) in flowing gas at the heating rate of 15 C min-1

123
Thermal decomposition of calcium oxalate: beyond appearances

CO This information brings the answer to the question

1.0
Ptotal = 1 bar raised by Simon and Newkirk [22], and Vallet [27], who
0.1 bar provide data with due care, precision, and details. Their

Metastability of CO2 and CO

0.8
0.01 bar important works remain forever a model for users of

in the presence of carbon

PCO ÷ (PCO+PCO2)

0.001 bar
0.6
thermogravimetry analysis.
Metastability of CO

0.4
Conclusions
0.2
Contrary to conventional wisdom, calcium oxalate (CaC2-
0.0 O4) does not decompose to give calcium carbonate
CO2 + C
(CaCO3) and carbon monoxide (CO), as has been sug-
100 200 300 400 500 600 700 800 900 1000 gested until now. Its decomposition is more complex than
Temperature/°C
may appear and consists of a multitude of cascading
Fig. 8 Boudouard equilibrium at various total pressures reactions, of which the end result is a mass loss that is
effectively equivalent to the mass of one molecule of CO.
generate also CO. CO is, in turn, oxidizes to CO2. As a result,
• From around 48 C, CaC2O4H2O loses its water
the partial pressure of CO2 increases in the chamber reactor,
content which escapes as water vapor. Some traces of
and therefore impedes the decomposition of CaCO3 into
CO2 are also detected. The quantity of CO2
CaO and CO2. According to Le Châtelier’s principle, addi-
increases progressively with increasing temperature.
tional reactant as CO2 in the chamber reactor will shift the
• Next comes the main step of decomposition of CaC2-
equilibrium toward the side of the reactant (CaCO3) to annul
O4CaC2O4 decomposes to form two solid phases
the change. The presence of carbon and its reactivity with
(CaCO3 and free carbon) and a gaseous phase CO2
CaCO3 giving off CO that oxidizes into CO2 explain the
following reaction (4):
slight shift toward higher temperatures corresponding to the
delay in the decomposition of CaCO3. At high temperatures, CaC2 O4 ! CaCO3 þ 0:5ðCO2 þ CÞ ð4Þ
it seems that the effect induced by the presence of CO2 in the
• CO2 is metastable in the presence of free carbon phase.
system overcomes the temperature effect induced by the
The two species (CO2 and free carbon) would have to
exothermic oxidation reaction of CO.
react following the Boudouard reaction in order to form
It should be noted that the free carbon is present even
CO, as illustrated in Fig. 8. However, this reaction
when the material is heated in oxygen, where the dispro-
cannot proceed since the CO2 produced according to
portion of CO is repressed. That is to say that the free
reaction (4) is evacuated away from the reaction
carbon is not a by-product resulting from the disproportion
chamber by the stream of helium used during the heat
reaction of CO as believed so far, but it is the product of the
treatment. Thereby, CO2 has a short residence time to
reaction (4) that starts before the reaction (2).
react with free carbon. By contrast, the free carbon solid
phase accumulates within the powder.
• Beyond 374 C and up to 550 C, the well-known
0.2 reaction for the decomposition of the oxalate with
emission of CO occurs (reaction 2).
0.0
CaC2 O4 ! CaCO3 þ CO ð2Þ
– 0.2
Mass change/ng

– 0.4
• At that stage of the process, CO2, C, and CO are present
in the reaction chamber. We now have all components
– 0.6
of the Boudouard reaction (5). The system will try to
– 0.8 adjust in the direction of achieving equilibrium accord-
– 1.0
ing to reaction (5).
– 1.2 0:5CO2 þ 0:5C ! CO ð5Þ
– 1.4 • The disproportionation of CO is not the source of free
0 200 400 600 800
carbon, as suggested by many scientists. It is generated
Temperature/°C
from the decomposition of CaC2O4 following reaction
Fig. 9 Temperature dependence of the mass change following the (4).
decomposition of 2.7 ng calcium oxalate monohydrate [35]

123
 D. Hourlier

• Beyond 550 C, CaCO3 is metastable in the presence of negligence of its users, leads to the utmost commitment to
carbon. In contact, the two solids react to form solid uncertain, even fantastical, conclusions.’’
calcium oxide (CaO) and a gaseous phase CO. To our knowledge, this is the only ultrasensitive ther-
CaCO3 þ C ! CaO þ 2CO ð6Þ mogravimetric system that has allowed for the demarcation
of these different inflection points, in the second domain of
• In parallel to the reaction (6), the decomposition of mass loss between 200 and 500 C in particular. These
CaCO3 takes place according to the well-known results corroborate with the gaseous emissions detected by
reaction (3): CaCO3 ? CaO ? CO2 (3) mass spectrometry and the reactions proposed to finally
The existence of these two reactions may explain the explain the decomposition of calcium oxalate
dissymmetry of the emission profiles of CO2 observed by monohydrate.
mass spectrometry.
Acknowledgements I gratefully acknowledge discussions with Pro-
• Gaseous CO2 is present in the reactional environment fessor P. Perrot, concerning thermodynamic calculations. My sincere
thanks to C. Bonhomme, D.Laurencin, S. Venkatachalam, A. Bleu-
throughout the heating process of the initial powder zen, and J. N. Jaubert for their encouragements. Further, thanks to Dr.
(CaC2O4H2O). H. H. Fabian and R. Berger for giving me permission to use their
• Under an oxidizing atmosphere, the CO from the TGA curve for comparison. Thanks also to the people from Netzsch
decomposition reaction of CaC2O4 (Eq. 2) burns out company; Dr Juergen Blumm and his team, as well as, Thierry
Choucroun and Jean-Christophe Jullien, for their technical support by
and the delivered heat of the exothermic oxidation providing ultrasensitive TGA/MS devices.
reaction of CO brings the sample to a higher temper-
ature. As a result, the decomposition of CaC2O4
according to Eq. 2 goes faster and thus explains the References
shift of the TGA curve toward lower temperatures in
oxygen atmosphere, in comparison with results 1. Parmar MS. Kidney stones. Br Med J. 2004;328:1420–4.
obtained under an inert atmosphere. 2. Verrecchia EP, Braissant O, Cailleau G. The oxalate–carbonate
pathway in soil carbon storage: the role of fungi and oxalotrophic
• The oxidations tests have contributed to a better bacteria. In: Gadd GM, editor. Fungi biogeochemical cycles
understanding of the suggested reactions and the [Internet]. Cambridge: Cambridge University Press; 2006.
reactivity of solid phases in presence. p. 289–310.
3. Ahmed J, Ojha K, Vaidya S, Ganguli J, Ganguli A. Formation of
Before closing this study, we could not ignore the work calcium oxalate nanoparticles in leaves: significant role of water
conducted by researchers at the University of Basel content and age of leaves. Curr. Sci. India. 2012;103:293–8.
4. Franceschi VR, Nakata PA. Calcium oxalate in plants: Formation
[34, 35]. They have fabricated an ultrasensitive microbal- and function. Annu Rev Plant Biol. 2005;56:41–71.
ance MMTG (micromechanical thermogravimetry) com- 5. Ciccarone C, Pinna D. Calcium oxalate films on stone monu-
posed of a beam sensor, whose size measures in microns ments—Microbiological investigations. Aerobiologia.
and of which the upper surface is covered in a resistive 1993;9:33–7.
6. Del Monte M, Sabbioni C, Zappia G. The origin of calcium
layer that enables heating. The beam is connected to a oxalates on historical buildings, monuments and natural outcrops.
piezoelectric actuator controlled by a frequency generator. Sci Total Environ. 1987;67:17–39.
It monitors the shift in the oscillation frequency of the 7. East CP, Doherty WOS, Fellows CM, Yu H. Formation of ther-
beam, which is proportional to the mass variation of the modynamically unstable calcium oxalate dihydrate in sugar mill
evaporators. Proceeding 32nd Aust Soc Sugar Cane Technol
sample. ASSCT Conf. Bundaberg, Queensland: ASSCT; 2010. p. 522–33.
The mass loss profile (Fig. 9) obtained by heating a 8. Peltier S, Duval C. Sur la thermogravimétrie des précipités ana-
powder of calcium oxalate monohydrate, with a mass of lytiques. Anal Chim Acta. 1947;1:345–54.
2.7 ng, clearly shows the inflection points in the second 9. Müller-Vonmoos M, Kahr G, Rub A. DTA-TG-MS in the
investigation of clays: quantitative determination of H2O, CO
domain of mass loss. These points correspond to the dif- and CO2 by evolved gas analysis with a mass spectrometer.
ferent phenomena evoked in our study and corroborate the Thermochim Acta. 1977;20:387–93.
proposed reactions. The authors did not doubt the point at 10. Wang J, McEnaney B. Quantitative calibration of a TPD-MS
which their measurement was precise, they even thought system for CO and CO2 using calcium carbonate and calcium
oxalate. Thermochim Acta. 1991;190:143–53.
that the inflexion points were in slight disagreement with 11. Antal MJ, Varhegyi G, Jakab E. Cellulose pyrolysis kinetics:
the theoretical reactions typically published. revisited. Ind Eng Chem Res. 1998;37:1267–75.
For us this recalls the remarks made by André Chrétien, 12. Cao Y, Casenas B, Pan W-P. Investigation of chemical looping
professor of mineral chemistry at the Sorbonne, in the combustion by solid fuels. 2. Redox reaction kinetics and product
characterization with coal, biomass, and solid waste as solid fuels
preface to the precious book devoted to the thermo- and CuO as an oxygen carrier. Energy Fuels. 2006;20:1845–54.
gravimetry analysis by Vallet [27]: ‘‘Thermogravimetry is
one of the methods that, by the ignorance or by the

123
Thermal decomposition of calcium oxalate: beyond appearances

13. Haines PJ. Thermogravimetry: thermal methods of analysis— 24. Hotová G, Slovák V. Quantitative TG-MS analysis of evolved
principles, applications and problems [Internet]. Dordrecht: gases during the thermal decomposition of carbon containing
Springer; 1995. p. 22–62. solids. Thermochim Acta. 2016;632:23–8.
14. Gallagher PK. An evolved gas analysis system. Thermochim 25. Arii T. Evolved gas analysis-mass spectrometry (EGA-MS) using
Acta. 1978;26:175–83. skimmer interface system equipped with pressure control func-
15. Price D, Dollimore D, Fatemi NS, Whitehead R. Mass spectro- tion. J Mass Spectrom Soc Jpn. 2005;53:211–6.
metric determination of kinetic parameters for solid state 26. Boudouard O. Les phénomènes de combustion dans les foyers
decomposition reactions. Part 1. Method; calcium oxalate industriels. Rev Phys Chim. 1901;25:6.
decomposition. Thermochim Acta. 1980;42:323–32. 27. Vallet P. Thermogravimétrie: étude critique et théorique: utili-
16. Szekely T, Varhegyi G, Till F, Szabo P, Jakab E. The effects of sation, principaux usages [Internet]. New York: Gauthier-Villars;
heat and mass transport on the results of thermal decomposition 1972.
studies: Part 1. The three reactions of calcium oxalate monohy- 28. Petit I, Belletti GD, Debroise T, Llansola-Portoles MJ, Lucas IT,
drate. J Anal Appl Pyrolysis. 1987;11:71–81. Leroy C, et al. Vibrational signatures of calcium oxalate poly-
17. Manley TR. Thermal analysis of polymers. Pure Appl Chem. hydrates. ChemistrySelect. 2018;3:8801–12.
2018;61:1353. 29. Pinto BV, Ferreira APG, Cavalheiro ETG. Thermal degradation
18. Giron D. Thermal analysis and calorimetric methods in the mechanism for citalopram and escitalopram. J Therm Anal
characterisation of polymorphs and solvates. Pharm Therm Anal. Calorim. 2018;133:1509–18.
1995;248:1–59. 30. McLafferty FW, Gore J. Spectrographie de masse: introduction a
19. Frost RL, Weier ML. Thermal treatment of whewellite—a ther- l’interpretation des spectres de masse. Ed. francaise dirigee par
mal analysis and Raman spectroscopic study. Thermochim Acta. jacques gore. Ediscience; 1969.
2004;409:79–85. 31. Rak J, Skurski P, Gutowski M, Bła_zejowski J. Thermodynamics
20. Kociba KJ, Gallagher PK. A study of calcium oxalate monohy- of the thermal decomposition of calcium oxalate monohydrate
drate using dynamic differential scanning calorimetry and other examined theoretically. J Therm Anal. 1995;43:239–46.
thermoanalytical techniques. Thermochim Acta. 32. Bła_zejowski J, Zadykowicz B. Computational prediction of the
1996;282–283:277–96. pattern of thermal gravimetry data for the thermal decomposition
21. Slager TL, Prozonic FM. Simple methods for calibrating IR in of calcium oxalate monohydrate. J Therm Anal Calorim.
TGA/IR analyses. Thermochim Acta. 2005;426:93–9. 2013;113:1497–503.
22. Simons EL, Newkirk AE. New studies on calcium oxalate 33. Dollimore D. Thermal analysis. Anal Chem. 1996;68:63–72.
monohydrate. Talanta. 1964;11:549–71. 34. Berger R. Fabian JH. Thermal analysis of nanogram quantities
23. Le Parlouër P. Thermal analysis and calorimetry techniques for using a micromechanical cantilever sensor. In: North American
catalytic investigations. In: Auroux A, editor. Calorimetry and Thermal Analysis Society Proceedings, Pittsburgh; 2002.
thermal methods in catalysis [Internet]. Berlin: Springer; 2013. 35. PhD thesis of Jan-Henning Fabian, Mikromechanische Oszilla-
p. 51–101. toren für die thermisch-gravimetrische Analyse, University of
Basel; 2001.

123