Evolved gas analysis of apatite materials using thermochromatography

Article abstract of DOI:10.1016/S0040-6031(98)00478-X

Evolved gas analyses of model inorganic compounds and apatites are described. The measurements are performed on the system - a 'thermochromatograph' (ThGC) - with a low-volume, thermal furnace, interfaced to a capillary gas chromatograph (GC) via a computer-controlled, pneumatic sample inlet device in the 70-600°C region. The 'information content' performance of this inherently simpler system is comparable to that of TG/GC/MS and TG/GC/IR systems when similar materials are analyzed.

Full text of DOI:10.1016/S0040-6031(98)00478-X

Thermochimica Acta 322 (1998) 25±32
Evolved gas analysis of apatite materials using thermochromatography
a,*
Mihkel Koel , Marina Kudrjasova , Kaia ToÄnsuaadu , Merike Peld , Mihkel Veiderma
a
b
b
b
Ï
a Institute of Chemistry, Akadeemia tee 15, Tallinn EE0026, Estonia
b Institute of Chemistry, Tallinn Technical University, Ehitajate tee 5, Tallinn EE0026, Estonia
Received 22 September 1997; accepted 29 June 1998
Abstract
Evolved gas analyses of model inorganic compounds and apatites are described. The measurements are performed on the
system ± a `thermochromatograph' (ThGC) ± with a low-volume, thermal furnace, interfaced to a capillary gas chromatograph
(GC) via a computer-controlled, pneumatic sample inlet device in the 70±6008C region. The `information content'
performance of this inherently simpler system is comparable to that of TG/GC/MS and TG/GC/IR systems when similar
materials are analyzed. # 1998 Elsevier Science B.V.
Keywords: Evolved gas analysis; Carbonate apatite; Gas chromatography; Inorganic salts
1. Introduction
loss curves can be readily reconstructed from EGA
curves (taking care that all evolved components are
detected, and that detected responses to all compo-
nents are known). On the other hand, many EGA
applications ± especially those for inorganic samples
± have relatively simple and predictable evolved gas
composition and the justi®cation for spectrometric
identi®cation tools is not clear. In addition to a high
cost per sample, these seemingly more sophisticated
approaches add several problems: (a) relatively large
TG furnace volume can degrade system performance
for narrow EGA peak shapes; (b) potential sample
component loss in interconnecting lines between TG
and IR or MS instruments because of condensation of
tars and other high-boiling products; and (c) unwanted
selective detection (e.g. not all evolved components
are IR active).
An ideal evolved-gas analysis (EGA) instrument
would provide a time- or temperature-based record
of the appearance and disappearance of all evolved
compounds together with their identi®cation. Most
recently, thermogravimetry-mass spectrometry (TG-
MS) and thermogravimetry-infrared spectrometry
(TG-IR) instruments have become available [1,2].
Although the current instruments can continuously
monitor the evolved products with scan speeds of
fractions of a second, their complete capability may
only be required when the evolved gas composition is
a truly complex mixture. Even then, better resolution,
in terms of non-collinear information content, might be
achieved by a thermogravimetry±gas chromatography±
spectrometry (TG-GC-MS/IR) system [3]. Weight-
Despite its many attractive features, using gas
chromatography (GC) as an EGA detector introduces
at least two more problems. Firstly, the detection of
*Corresponding author.
0040-6031/98/$ ± see front matter # 1998 Elsevier Science B.V. All rights reserved
P I I : S 0 0 4 0 - 6 0 3 1 ( 9 8 ) 0 0 4 7 8 - X

26
M. Koel et al. / Thermochimica Acta 322 (1998) 25±32
the evolved compounds in the case of samples
which release only small amount of gas during
their heating (relative to their total mass) as is the
case for many inorganic samples. Secondly, separation
of the evolved components requires careful selection
of the GC column to avoid selective trapping in the
column.
onto the GC column must be achieved at relatively-
high temperatures. Most commercially available,
mechanical sampling valves exhibit accelerated
mean-time-to-failure when maintained at the higher
temperatures needed to avoid condensation and to
match common GC column inlet operating conditions.
An alternative is to have sampling done with a pneu-
matic sampling valve with no moving parts, but at the
expense of some further sample dilution. Such sam-
plers operate on the principle of carefully balancing
the pressures in the sample stream and the carrier gas
¯ow, ®rst described by Deans [4], and is used in this
system to introduce the aliquots of sample head space
into a capillary column. A complete description of the
sampler is given in an earlier work [5]. The chroma-
tograph is a Carlo Erba 4200 GC with a thermocon-
ductivity detector (mod.430) and a porous layer
column 0.53 mm i.d and 25 m in length {NSW-PLOT
(HNU Nordion Oy, Finland)} was used to separate
gaseous products of decomposition.
In EGA by GC, where the reactor size is optimised
according to needs of chromatography, the result is set
of chromatograms of the gases released at the corre-
sponding (sampling) temperatures, i.e. a particular
sample temperature can be associated with each chro-
matogram. If the temperature interval between two
chromatograms is small and a set consists of numerous
recorded chromatograms, it is straightforward to form
an EGA response surface with axes: sample tempera-
ture at injection time vs. chromatographic runtime. By
analogy to other hyphenated analytical techniques and
because of the axis names, this method is called
thermochromatography (ThGC), and the correspond-
ing surface plots are known as thermo(gas)chromato-
grams. If there are several independent reactions
occurring in the sample during the entire course of
heating, then the corresponding thermochromatogram
is `a linear combination of the thermochromatograms
of individual processes'.
2.2. Sampling control and data acquisition
The sampling is controlled by an Apple II computer
and a locally made interface card. The sampling events
occur at intervals equal to the separation time of the
evolved components (total chromatogram runtime in
our case was 110 s) and then chromatograms of
evolved gases are found at certain predetermined
temperatures. The detector signal is converted to
digital form by a 23-bit analog-to-digital converter
(ECTA, Estonia), received by the Apple II control
computer and transferred to a 486-type PC over an
RS232 interface for ®nal data processing. At ®rst,
image processing methods display such EGA response
surfaces as 2D objects either as a contour or mesh
plots. Fig. 2 presents the thermochromatogram of
magnesium carbonate hydroxide as a mesh plot.
Although such plots do not add much to the ®nal
presentation of the EGA results ± the evolution
rates of the released gas components as a function
of sample temperature, they provide for the analyst a
convenient `picture' of the thermal events occurring
in the sample.
The goal of the present work was to demonstrate
ThGC performance in the case where evolution of gas
phase materials occurs from inorganic materials, and
to show ThGC ability to provide much information for
a better understanding of thermal decomposition of
apatites.
2. Experimental
2.1. Furnace, sampler and chromatograph.
A schematic of the overall system is presented in
Fig. 1. The quartz tube furnace which replaces the
injection port of the chromatograph has a volume of
4 ml and the sample is located therein. The furnace/
reactor temperature is controlled by a separate, stand-
alone temperature programmer with an operating
range of 70±6008C, at heating rates of 1±258C/min.
In our experiments, a heating rate of 108C/min was
used in the temperature region from 708 to 6008C, for
all the samples. Automatic injection of the evolved gas
Most of the software used was written by the
authors. Equipment control (sampling and data
recording) software was written in assembly language.
When necessary, the chromatogram pre-processing

M. Koel et al. / Thermochimica Acta 322 (1998) 25±32
27
Fig. 1. Schematic of a thermochromatographic experiment.
(base-line correction, spike removal, digital ®ltering,
etc.) was performed, for which software was written in
C. The thermochromatogram 2D representation sub-
routines are written in MATLAB (Math Works,
Natick, USA).
PO34 ). The formulae of samples calculated by data of
chemical analysis, and according to the electroneu-
trality principle, are as follows:
sample 1 Ca9:60&0:40ꢀPO4†4:82ꢀCO3†
0:80
 ꢀCO3F†0:38F1:30ꢀOH†
0:70
2.3. Samples
sample 2 Ca8:12Mg1:75&0:13ꢀPO4†
4:74
 ꢀHPO4†0:33ꢀCO3†0:93F1:67ꢀOH†
0:33
Two inorganic reagents, MgNH4PO4Á6H2O, Fluka,
Germany, 4MgCO3ÁMg(OH)2Á5H2O, Reachim, Rus-
sia, and three samples of synthetic carbonate apatites,
synthesised at the Tallinn Technical University, were
studied. The apatite samples were synthesised by the
precipitation method described previously [6,7]. The
apatites were identi®ed by XRD and IR spectroscopy
as B-type carbonate apatites (CO23 substituted for
sample 3 Ca6:38Mg2:64ꢀNH4†0:94ꢀHPO4†
0:64
 ꢀPO4†4:98ꢀCO3†0:38F1:47ꢀOH†
0:53
The samples were placed in a quartz vital inside the
reactor and the weight of each sample was ca. 20 mg.
Under thermal in¯uence, all the proved samples
evolve together with H2O and NH3 or CO2.

28
M. Koel et al. / Thermochimica Acta 322 (1998) 25±32
Fig. 2. Mesh plot of magnesium carbonate hydroxide degradation products: peak at retention time 45 s corresponds to CO2, at 60 s
corresponds to H2O.
3. Results and discussion
ꢁ MgNH4PO4)MgHPO4‡NH3
ꢁ 2MgHPO4)Mg2P2O7‡H2O
3.1. Magnesium ammonium phosphate hexahydrate
ThGC offers a good chance to check the proposed
scheme. On the evolving curve of ThGC (Fig. 3), one
can follow three peaks: (1) temperature region 90±
1908C, when water is evolving (73.7% of the total
amount); (2) temperature region 240±3208C, when
water (23.1% of the total amount) and ammonia are
simultaneously evolving; and 3) wide temperature
region 450±5708C, when water is evolving (3.2%).
From these data, one can conclude that the release
of ®ve water molecules and the decomposition of the
intermediate MgNH4PO4ÁH2O became well separated
and a part of crystal water remained and simulta-
neously evolved with ammonia. This evolution is
taking place in the 240±3208C range. By peak height,
the amount of water evolving in that region is approxi-
mately 112 molecules and corresponds to the following
main reactions:
During thermal analysis of ammonium salts, ammo-
nia is evolved along with water as part of the total
weight loss at the same temperature. If each evolution
process is to be distinguished, there are distinct advan-
tages in obtaining separate evolution curves. In this
case, ThGC offers a simpler solution to the detection
of simultaneous evolving gases than, for example, the
approach Paulik et al. [8] used for thermo-gas titri-
metry (TGT).
A relatively complex process of degradation takes
place in case of magnesium ammonium phosphate
hexahydrate. For MgNH4PO4Á6H2O, the following
scheme of thermal decomposition is presented as
follows:
ꢁ MgNH4PO4Á6H2O)MgNH4PO4‡6H2O

M. Koel et al. / Thermochimica Acta 322 (1998) 25±32
29
Fig. 3. Thermal decomposition of magnesium ammonium phosphate (MgAP) and carbonate apatite sample 3 (S3): (A) evolution of NH3; (B)
evolution of H2O.
MgNH4PO4 Á H2O ) MgNH4PO4 ‡ H2O
or
2MgNH4PO4 ) Mg2P2O7 ‡ 2NH3 ‡ H2O
4MgNH4PO4) MgH2P2O7 ‡ Mg3ꢀPO4†
‡4NH3 ‡ H2O
2
The existence of the third step at higher tempera-
tures (450±5708C) shows the evolution of some part
of remaining constitutional water. It means that the
hydrous intermediate(s) are formed in the second step
And these intermediates decompose at higher tem-
peratures and give evolution of water:
2MgHPO4 ) Mg2P2O7 ‡ H2O
MgNH4PO4 ) MgHPO4 ‡ NH3

30
M. Koel et al. / Thermochimica Acta 322 (1998) 25±32
Fig. 4. Thermal decomposition of magnesium carbonate hydroxide: (A) 1, H2O; 2, CO2; (B) 3, total evolving curve; 4 ± total differential
evolving curve.
or
3.2. Magnesium carbonate hydroxide hydrate
2MgH2P2O7 ) Mg2P4O12 ‡ 2H2O
From the evolved gas analysis alone, it is impos-
sible to identify the reactions and one must use other
analytical methods to con®rm the ®nal products.
4MgCO3ÁMg(OH)2Á5H2O belongs to group of
hydrated oxides with a complicated crystal structure
and they commonly exhibit a multi-stage decomposi-
tion [9]. The curve (Fig. 4) of total evolution for ThGC

M. Koel et al. / Thermochimica Acta 322 (1998) 25±32
31
of magnesium carbonate hydroxide hydrate has three
distinct steps, all proceeding consecutively with some
overlap:
during the heating of precipitated apatites are water
and carbon dioxide, as well as ammonia.
The water evolving during thermal treatment of
synthesized apatites is of three different origins:
adsorbed water, structurally incorporated water and
water evolving as a result of interactions of the apatite
constituents [9]:
1. A range 220±3358C, when mainly water and some
CO2 are evolved.
2. A range 350±4508C, where CO2 and a small part of
H2O is evolved.
2HPO24 ˆ P2O47 ‡ H2O
3. A range 450±5508C, when only CO2 is evolved.
‡
2NH4 ‡ 2HPO24 ˆ 2NH3 ‡ H2O ‡ H2P2O72
As ThGC permits one to follow the evolution of
gaseous products independently, it is possible to
measure amounts of a given gas at distinct tempera-
ture intervals. This way, it is found that ®ve parts
of water evolve during the ®rst step, and one part
of water during the second step. This corresponds
to the water of hydration (®ve molecules) and
hydroxide (one group). Amounts of evolved CO2 in
different stages suggests that there are distinct
intermediates undergoing decomposition in each
range. The major phase of CO2 evolution occurs
during steps 2 and 3. Apparently, equal parts of
carbonate degrades at these steps. These data could
be interpreted as the following reactions are occurring
in the sample
H2P2O27 ‡ 2CO32 ˆ 2PO34 ‡ 2CO2 ‡ H2O
The adsorbed water is loosened at 100±1608C for
all samples (Fig. 5). Evolution of the lattice water
from ¯ourocarbonate apatite without substitutions in
the cationic site of the structure (sample 1) occurs in a
wide temperature interval, up to 5008C, with max-
imum in the 300±3208C region. In substituted mag-
nesium carbonate apatite (sample 2), the evolution of
lattice water proceeds at lower temperatures because
of weakening of the crystal structure, with a maximum
at ca. 2008C. The loss of water, resulting from the
above-mentioned reactions also proceeds in the tem-
perature range up to 6008C.
The loss of CO2 on heating carbonate apatites,
investigated by gas chromatography and FTIR analy-
sis of evolved gases, occurs between 4508 and 9508C,
with two or three maxima [7,10,11]. On account of
ThGC of the ¯ourocarbonate apatite (sample 1), there
is a slight loss of CO2 on heating up to 6008C, making
up to 13% of its total content. On the contrary,
evolution of CO2 from magnesium-substituted carbo-
nate apatite proceeds at lower temperatures ± to a
greater extent ± in amounts of 72%, with a maximum
at 5808C. The release of CO2 gives rise to thermally
induced rearrangements in the apatite structure lead-
ing to more stable crystalline phase of ¯uoroapatite.
The evolution of ammonia from magnesium±
ammonium carbonate apatite (sample 3) proceeds in
ThGC experiments to completion, mainly in the tem-
perature interval 220±3008C, that is as by heating of
MgNH4PO4 (Fig. 3). The possibility of the existence
of MgNH4PO4 as a separate phase in sample 3 is not
excluded.
ꢁ 4MgCO3ÁMg(OH)2Á5H2O)4MgCO3ÁMg(OH)2
‡5H2O" (220±3358C)
ꢁ 4MgCO3ÁMg(OH)2)2MgCO3Á3MgO‡H2O"
‡2CO2" (350±4508C)
ꢁ 2MgCO3Á3MgO)5MgO‡2CO2" (450±5508C)
Eventually, it was concluded that the partial pro-
cesses (degradation of carbonate and hydroxide) were
not separated. Thus, some other processes must be
responsible for the temporary ceasing and subsequent
gradual change of the thermal decomposition and this
could be the cessation of re-crystallisation of the new
phase [9].
3.3. Carbonate apatites
The thermally induced changes in synthetic carbo-
nate apatites are rather complex, involving overlap-
ping reactions with release of several gaseous species.
Their behavior on thermal in¯uence depends much on
the pattern of synthesis and conditions of heating.
Using ThGC equipment, it was possible to cover the
main temperature regions up to 6008C, where evolu-
tion of gases takes place. The main products evolved
Thus, the ThGC can be successfully used for
obtaining information on the composition and thermal
stability of carbonate apatites. Also, these are the
examples where EGA must be applied together with

32
M. Koel et al. / Thermochimica Acta 322 (1998) 25±32
Fig. 5. Evolving of gases from the precipitated apatites ± sample 1 (S1) and sample 2 (S2): (A) evolution of CO2; and (B) evolution of H2O.
[4] D.R. Deans, J. Chromatogr. 289 (1984) 43±51.
other thermoanalytical methods like DSC, and others,
to obtain a full description of the processes.
[5] M. Koel, K. Urov, Oil Shale 10 (1993) 261±269.
[6] M. Peld, K. ToÄnsuaadu, M. Veiderma, K. Rimm, Trans. of
Tallinn Technical University, G., Chem. 742 (1994) 4±16.
È
È
[7] K. ToÄnsuaadu, M. Peld, T. Leskela, R. Mannonen, L. Niinisto,
M. Veiderma, Thermochim. Acta 256 (1995) 55±65.
[8] J. Paulik, F. Paulik, L. Erdey, Microchim. Acta (1966) 886±
893.
References
[1] C.M. Earnest, Anal. Chem. 56 (1984) 1471A±1486A.
[2] W.J. Young, X.X. Quan, J.T. Zheng, Thermochim. Acta. 111
(1987) 325±333.
[9] F. Paulik, Special Trends in Thermal Analysis; John Wiley &
Sons, 1995, p. 460.
[3] W.H. McClennen, R.M. Buchanan, N.S. Arnold, J.P. Dwor-
zanski, H.L.C. Meuzelaar, Anal. Chem. 65 (1993) 2819±
2823.
[10] J.C. Elliot, Structural Chemistry of Apatite and Other
Calcium Ortophosphates, Elsevier, 1994, p 487.
[11] M. Jemal, J. Khattech, Thermochim. Acta 152 (1989) 65±76.