You are on page 1of 114

Coupling of

Thermal Analysis and Gas Analysis

Techniques and Applications

ed Gas Ana
lv lys
vo is
E

FTIR MS

Th is
er m s
al Analy
Editors: E. Kaisersberger
E. Kapsch
E. Post
A. Schranner
Coupling of
Thermal Analysis and Gas Analysis

Techniques and Applications

NETZSCH-Gertebau GmbH, 95100 Selb,


2001

This publication, including all of the text and graphics contained in it, is protected by copyright.
All rights thereby established are reserved, in particular, regarding reproduction, distribution, revision,
translation, microfilming, oral presentation and storage, processing or reproduction in databanks
or other electronic media and systems, also with respect to excerpts. The technical descriptions
and representations are made to the best of the authors knowledge.
Responsibility for utilization of such is assumed by the user.
Coupling of Thermal Analysis and Gas Analysis
Techniques and Applications

Table of contents
Page
Preface 1
List of Authors 2

Techniques and Applications


Introduction, Instrumentation TA, MS, FTIR, Coupling techniques , E. Kaisersberger 4
Quantification of Spectroscopic Signals Recorded in TA-MS and TA-FTIR Systems,
M. Maciejewski 19

Polymers, Paints, Organic Chemistry


Curing and Decomposition of Phenolic Resin, A. Schranner 27
Determination of Emission of Volatiles during Paint Drying by TG-FTIR, H. Mhler 32
Recycling of Polyamides, G. Matuschek 41
Characterization of Epoxy-based Electronic Waste, E. Post 47
Development of Undercoating Materials Containing Polyvinylchloride, PVC, A. Rager 51
Softener Identification by TG-FTIR: Fomblin Y 06, A. Rager 55

Pharmaceutics
Decomposition Behavior of Acetylsalicylic Acid (ASA), A. Rager 59

Ceramics, Building Materials

Binder Burnout of an Al2O3 Green Body, E. Post 63


HF Emissions at the Firing Process of Brick Clays, E. Kapsch 67
Characterization of Raw Cement and Cement Mixtures, E. Post 72

Metals and Semiconductors


Identification of Metal Vapors and Metal Salts - Cd, PbCI2, E. Post 75
Application of Thermal Analysis to Material Science Case Study on Hard Metals,
G. Leitner 79
Ternary Semiconductors CuGaSe2, ZnGeP2 -impurities, decomposition-, E. Post 90

Miscellaneous (Inorganic Chemistry, Catalysts, Food Ingredients, superconductors)


Application of Pulse Thermal Analysis (PULSETA), M. Maciejewski 94
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Preface

The analytical information obtained from the separate application of single methods of
thermal analysis is limited and from the early beginning combination of methods was
practiced. The lasting success of simultaneous TG-DTA and nowadays even more of the
simultaneous TG-DSC demonstrates the power of additional information gained through the
combination of methods in hyphenated techniques.
The detection and analysis of evolved volatiles was attempted from the early stages of
widespread application of thermal analysis, often in home made coupling systems for the gas
analysis.
General features of the techniques involved, namely of TG and simultaneous TG-DSC from
thermal analysis and of FTIR and MS from the gas analysis side are discussed by A. Rager
and myself, together with the various coupling techniques for their combination to form one
functional unit. The quantitative work with the coupled techniques is also one of the main
topics of this summary, extensively described by the contribution of M. Maciejewski on pulse
techniques.
The broad scope of application areas for hyphenated techniques can not be covered at all in
one issue of limited page numbers, however windows and ideas can be opened for similar
applications in many fields of research, material development and process optimization.

The contributions are brought to a simple ordering scheme, with polymers at the first
position. As with thermal analysis in general, polymer applications are also dominant with the
hyphenated techniques, A. Schranner, H. Mhler, G. Matuschek, E. Post and A. Rager
present examples for thermoplastics, resins, paints and plasticizers. The famous Aspirin is
introduced by A. Rager for a pharmaceutical example, many further for coupled TG-FTIR and
TG-MS are included in the NETZSCH Annual 2000, Thermoanalytical Characterization of
Pharmaceuticals. Ceramics and building materials, e.g. cement, present one group of
applications where firing processes are involved for the industrial production. The analysis of
volatiles for material and process characterization is additionally important for the aspect of a
possible impact on the environment by evolved gases during the firing process. These topics
are covered by the contributions of E. Post and E. Kapsch. E. Post and G. Leitner highlight
the field of metals, alloys and powder metallurgy for development of hard metals. In a further
contribution, semiconductors are analyzed by E. Post for impurities and production residues.
An overview on quantitative evaluations using the versatile pulse technique is given by
M. Maciejewski with various applications for inorganic materials, ceramic and metal-based
catalysts, food ingredients, ceramic superconductors and zeolites.

Cordial thanks to all authors and co-authors, to the team of editors and colleagues, who did
important work in the background, and to all who took their time to contribute to this volume
in very short term.

Selb, June 2001


In the name of the team of editors:

Erwin Kaisersberger
NETZSCH-Gertebau GmbH

1
Coupling of Thermal Analysis and Gas Analysis
Techniques and Applications

List of Authors

E. Kaisersberger, E. Post, A. Schranner


Netzsch-Gertebau GmbH
Wittelsbacherstrae 42
D-95100 Selb
Germany
Fax: (+49) 9287 881 44
e-mail: at@ngb.netzsch.com
http://www.ngb.netzsch.com

E. Kapsch
Netzsch-Gertebau GmbH
Office Netzsch do Brazil
Rua Michigan 166
Brooklin
BR-04 556-000 Sao Paulo/SP
Brazil
Fax: (+55) 11 55 43 27 16
e-mail: elisabeth.kapsch@ndb.netzsch.com.br
http://www.ngb.netzsch.com

G. Leitner
Fraunhofer-Institut fr Keramische Technologien und Sinterwerkstoffe - IKTS
Winterbergstrae 28
D-01277 Dresden
Germany
Fax: (+49) 351 2553 600
e-mail: leitner@ikts.fhg.de
http://www.ikts.fhg.de

2
Coupling of Thermal Analysis and Gas Analysis
Techniques and Applications

M. Maciejewski
ETH Zrich
Laboratorium fr Technische Chemie
ETH-Hnggerberg, HCI E 139
Wolfgang-Pauli-Strasse 10
CH-8092 Zrich
Switzerland
Fax: (+41) 1 632 1163
e-mail: maciejewski@tech.chem.ethz.ch
http://www.tech.chem.ethz.ch

G. Matuschek
GSF-Forschungszentrum fr Umwelt und Gesundheit GmbH
Institut fr kologische Chemie
Ingolstdter Landstrae 1
D-85764 Neuherberg
Germany
Fax: (+49) 89 3187 3371
e-mail: matuschek@gsf.de
http://www.gsf.de

H. Mhler
Fachhochschule Wrzburg-Schweinfurt
Studiengang Kunststoff- und Elastomertechnik
Rntgenring 8
D-97070 Wrzburg
Germany
Fax: (+49) 931 3511 339
e-mail: moehler@mail.fh-wuerzburg.de
http://www.fh-wuerzburg.de

A. Rager
Bruker Optik GmbH
Rudolf-Plank-Strae 23
D-76275 Ettlingen
Germany
Fax: (+49) 7243 504 651
e-mail: albrecht.rager@bruker.de
http://www.bruker.de

3
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Introduction
Erwin Kaisersberger, NETZSCH-Gertebau GmbH, Wittelsbacherstr. 42, 95100 Selb
Albrecht Rager, Bruker Optik GmbH, Rudolf-Plank-Str. 23, 76275 Ettlingen

Methods of Thermal Analysis, Instrumentation

Thermal Analysis comprises the instrumental techniques in which a physical property of a


sample material is measured as a function of temperature or time. Linear and non-linear
temperature programming, also in the form of a temperature modulated heating, is applied.
Furthermore isothermal measurements at selected temperatures are performed and also
cooling of a sample is possible either by uncontrolled cooling or by cooling under
temperature-controlled conditions.
The main physical parameters measured in Thermal Analysis are calorimetric, gravimetric
and dimensional changes (Fig. 1).

Thermal Analysis (TA)

Differential Thermogravimetry Thermomechanical


Scanning (TG) Analysis
Calorimetry (TMA)
(DSC, DTA) Dilatometry (DIL)
Dynamic-Mechanical
Analysis (DMA)

Thermal effects Mass changes due to Dimensional changes,


related to physical evaporation, deformations,
and chemical decomposition and viscoelastic
processes can be interaction with the properties,
characterized atmosphere transitions, density

Evolved Gas Analysis (EGA)

Fig. 1: Overview of the most common methods of thermal analysis

Mono-functional instruments measure a single physical property, however systems are


available on the market, which combine two and more measuring principles for the analysis
of a sample. These combined instruments exist in two classes, namely simultaneous and
coupled instruments. A typical representative for simultaneous instruments is the
combination of a thermogravimetric system (thermobalance, TG) with a differential thermal
analysis system in simultaneous TG-DSC and TG-DTA.
The connection of a thermobalance with an FTIR spectrometer or of a simultaneous TG-DSC
analyzer with a mass spectrometer via an appropriate interface are systems of choice for
coupled instruments.

4
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Characteristic of a separate application of DSC and DTA, the most frequently applied
thermoanalytical techniques, is the calorimetric analysis of samples up to their limit of
thermal stability; the signal is seldom closely related to the quantity of evolved volatiles,
which are products of evaporation and chemical reactions. Therefore combined gas analysis
has found no great importance for DTA and DSC alone.

Thermogravimetry determines the weight changes of the sample for all solid-gas and liquid-
gas interactions. Single volatiles or adsorbents are quantitatively detected through the
corresponding weight change and for all complex reactions with more than one volatile
product an integral weight change signal is achieved. Especially here combined gas
detection with specific detectors and with universal gas analysis methods found the most
frequent application. The additional calorimetric information of simultaneous TG-DSC and
TG-DTA is clearly of further advantage for interpretation of complex thermal events
measured in these coupled systems.
Thermomechanical methods measuring dimensional changes of solid samples are not a
main target for combination with gas analysis systems. A few special applications however
are found for studying firing processes in ceramics and powder metallurgy as well in
characterization of the gas exchange during heat treatment of metals.

Characteristics of a thermobalance: TG 209 C Iris

Basic instrument: TG 209 C Iris


Options: TG 209 C Iris-ASC (automatic sample changer)
TG 209 C Iris coupled with a Bruker Vector 22 FTIR system,
c-DTA (caloric information, calculated from the measured curve
of the sample temperature)

The TG 209 C Iris is a top-loading thermo-microbalance usable for the detection of small
mass changes.

TG 209 C Iris

Fig. 2: Thermo-microbalance TG 209 C Iris

5
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Top-loading balance:
The sample holder is located above the microbalance, thus simplifying the placement of the
sample crucible and avoiding contamination of the balance with decomposition products from
the sample, which are rapidly carried out to the top of the furnace through the vertical gas
flow.
Acquisition of sample temperature:
The sample temperature is directly measured with a thermocouple of Platinel being in
contact with the sample crucible. The STC function (sample temperature controller)
guarantees that the sample temperature follows exactly the temperature program, except
during transitions and reactions.

Cooling:
The cooling time from maximum temperature down to room temperature is very short
(approx. 15 min) due to the water-cooling of the microfurnace.

Sample atmosphere:
Measurements can be made in either static or dynamic pure gas atmosphere due to the
ability to evacuate the measuring unit. The valves for the gas change and the evacuation unit
AutoVac are activated via the software or manually.

Software:
The 32-bit software NETZSCH PROTEUS, running under Windows 95/98/NT, is optimally
tailored to meet the needs of practically all users.
Manual input of the sample weight is unnecessary (in accordance with Good Laboratory
Practice (GLP)), when using the internal weighing mode of the TG 209 C Iris.

Calibration:
The instrument is delivered with a certified standard weight for mass calibration. A suitable
calibration set for temperature calibration (melting standards for c-DTA) is available on
request.

Technical data:
Temperature range: room temperature ... 1000 C
Heating rates: 0.01 to 100 K min-1
Effective resolution: 1 g (internally: 0.02 g).
Weighing range: 500 mg
Balance capacity: 1 g (max.)
A great variety of sample crucibles is available.

Characteristics of a simultaneous TG-DSC system: STA 449 C Jupiter

Basic instrument: STA 449 C Jupiter


Options: STA 449 C Jupiter with mass spectrometer coupling
STA 449 C Jupiter with FTIR coupling

Mass changes (TG) and thermal effects (DSC or DTA) can be determined for the same
sample under identical conditions in one run using the Simultaneous Thermal Analyzer STA
449 C Jupiter. The electronically compensated top-loading thermo-microbalance stands out
due to its high accuracy and resolution in the sub-g range, as well as its excellent stability.

The STA 449 C Jupiter is designed for both high and low temperature applications. In the
specific temperature range of organic chemistry, biochemistry, pharmaceuticals and food, the
instrument can also be used for measurements in combination with the low temperature
furnace (between -120 C and 750 C).

6
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Cooling:
The low-temperature furnace can be equipped with a liquid nitrogen-cooling device, which
can be controlled automatically according to the selected temperature program.

Gas atmosphere:
Measurements are possible over the entire temperature range in a variety of static or
dynamic gas atmospheres as well as under vacuum down to 10-4 mbar. The vacuum-
tightness of the system allows investigation of materials which are extremely sensitive to
atmosphere (e.g. oxidation).

Sample holder systems:


Each sensor of the easily exchangeable sample holder systems can be configured with
appropriate thermocouples, allowing optimization for sensitivity, time constant and for the
selected atmospheres. The special design of the TG-DSC sensors, which includes
reproducible crucible placement, results in the quantitative determination of specific heat,
transformation temperatures and enthalpies, in addition to the accurate analysis of mass
changes.

Crucibles:
Alumina, aluminum, platinum, gold, fused silica,
zirconia and graphite for sample volumes of 40, 85,
190, 900 l up to 3.4 ml (TG only).

Software:
The NETZSCH PROTEUS Software (MS
Windows), by combining easy-to-understand menus
and automated routines, has provided the STA 449 C
Jupiter with a tool, which is very user-friendly, and, at
the same time, ready for complicated analyses.

Calibration:
The instrument is delivered with a certified standard
weight for mass calibration. Several calibration sets
(temperature and enthalpy calibration) for different
temperature ranges and crucible materials are
available on request.
Fig. 3: STA 449 C Jupiter

Technical data:
Temperature range: -120 ... 1650 C
Heating rates: 0.01 to 50 K min-1
Balance capacity: 5g
Weighing range: 5g
Balance resolution: 0.1 g
DSC resolution: < 0.1 W
DSC noise: < 0.3 W (root mean square (RMS), isothermal)

7
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Gas Analysis Methods

Fourier Transform Infrared Spectrometry

IR spectroscopy is a classical technique which is based upon the interaction of infrared


radiation with the vibrating dipole moments of molecules. It gives, with the exception of
homonuclear diatomics and noble gases, a characteristic spectrum for each substance.

For monitoring continuous processes, the Fourier-Transform (FT) technique offers the
specific advantages of speed and sensitivity.
Fourier-Transform-Infrared (FTIR) spectroscopy is the most advanced method of infrared
spectroscopy and nowadays, more or less, the only one used. The optical system of an FTIR
spectrometer consists essentially of an interferometer. Broad-band infrared radiation is
emitted by a thermal source (S), heated up to approx. 1500 K. The emitted lR light is directed
to a beam splitter (BS), which, ideally, reflects half of the light and transmits the other half.
After traversing a certain distance, the reflected part falls onto a fixed mirror (in the case of a
classical Michelson interferometer). At the mirror the radiation is again reflected and, after
traversing the same distance back, once again falls on the beam splitter surface. The
transmitted radiation follows a similar path, but is reflected by a movable mirror. This mirror
can be moved very precisely back and forth along the optical axis by a certain distance (x)
with accurate registration of position. Since the two back reflected beams are spatially
coherent, they interfere on recombination at the beam splitter. Again, the back reflected light
from the mirrors is divided at the beam splitter into a reflected and transmitted part. It follows
that half of the light emitted by the source is back reflected to the source.
The other half of the beam is modulated by the movable mirror and leaves the interferometer,
passes through the sample cell (Sa) and is finally focused on the detector (D). The signal
registered at the detector, the interferogram, is the radiation intensity, measured as a
function of the displacement of the movable mirror from its zero path difference position.

Fig. 4:
Scheme of a Michelson interferometer:

S: source
M1: fixed mirror
M2: movable mirror
BS: beam splitter
Sa: sample
X: mirror displacement
D: detector
AD: amplifier and AD-converter
FT: FT-computer

8
NETZSCH Coupling of Thermal Analysis and Gas Analysis

The mathematical transformation of the interferogram (Fourier Transformation) by the


computer provides a single channel spectrum, including the sample information, but also the
instrumental characteristics. To account for these instrumental characteristics, this spectrum
is divided by a reference spectrum, measured in the absence of a sample in order to obtain
the IR-spectrum of the sample alone. The IR-spectrum is then stored digitally in the
computer.
This FT-technique possesses two major advantages suitable for monitoring continuous
processes such as evolved gases by TG.
The so-called Jacquinot- or throughput advantage arises from the fact that the circular
apertures used in interferometers have larger area than the linear slits used in grating
spectrometers, thus enabling higher throughput of radiation.
In conventional spectrometers using monochromators, the spectrum is measured directly
by registering the radiation intensity as the wavelength is continuously changed by turning a
grating. Depending on the spectral resolution chosen, only a very small part of the radiation
from the monochromator reaches the IR-detector. In the interferometer all frequencies
emitted by the source reach the detector simultaneously. This is known as the multiplex or
Fellgett advantage. The time to record a full spectrum by the FT-technique is the time
needed for the movement of the mirror by a certain distance proportional to the desired
resolution. As the mirror can be moved very quickly, a complete spectrum can be measured
in a fraction of a second. Nowadays conventional FTIR spectrometers are able to collect
more than 100 spectra per second.

To analyze the evolved gases from a sample in the thermoanalyzer, moderate time
resolution in the range of 5 to 10 spectra per minute is normally sufficient.

The TG-FTIR coupling used here is based on a stainless steel gas cell, which has been
optimally adapted to the IR beam geometry. It has an optical path length of 123 mm and a
volume of just 8.7 ml. On account of the very small active volumes of the sample
compartment of the thermal micro balance (2.4 ml at the TG 209 C Iris), the transfer line
(2.2 ml) and the gas cell, a very low transfer gas flow can be used. This leads to a high
concentration of the evolved gases transported into the gas cell. The excellent sensitivity of
the FTIR spectrometer in combination with a fast and low noise semiconductor IR-detector
(mercury-cadmium-telluride, MCT) in the coupling unit gives an optimized system suitable
even for routine measurement of trace gases. Because of the vacuum-tight construction of
the thermo-microbalance and the gas cell, both systems can be evacuated together. The
much smaller amount of residual gases minimizes their influence on the decomposition
process and their reaction with the released gases.

Heatable gas lines avoid substance condensation and even allow the analysis of relatively
involatile components. Due to the principle of IR spectroscopy, which is a non-destructive
method, single components can be analyzed from the spectra of mixtures. The evaluation of
gas phase IR spectra is simplified because there is only minimal molecular interaction.
Therefore IR spectra of gas mixtures can be resolved into their single components by
subtracting away library spectra of the corresponding pure compounds. In most cases, time
consuming chromatographic separation is not required anymore.

9
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Quadrupole Mass Spectrometry

Small dimensions, fast and continuous function as gas analyzer and high sensitivity for all
evolved volatiles make the quadrupole mass spectrometers ideally suited for coupling with
thermal analyzers. Four main parameters are important for the characterization of a
quadrupole mass spectrometer:

Vacuum requirements
Ionization technology
Ion separation (mass filter)
Ion detection

One has to recall that the function of the quadrupole mass filter for separation of gases
according to their mass over charge ratio (m/z) requires a high vacuum typically in the range
of 10-5 mbar (10-3 Pa). In this vacuum the ion mobility is not significantly disturbed by the
residual gas and long lifetime of filaments in the ion source and of secondary electron
multipliers (SEM) for ion detection can be expected. The gas density at 10-5 mbar is still quite
high (about 1011 particles in one cm3; in general at P mbar: 2.687* 1019*P*1013-1 particles in
one cm3). The linear movement of particles without collision with others is at this pressure in
the range of meters (mean free path in cm for air: 6.66*10-3 *P-1; P in mbar).

Fig. 5: Scheme of a quadrupole mass spectrometer with the three main function units

The most common way for ion formation is the bombardment of the gas with electrons, which
are accelerated in an electric potential gradient (electron impact ionization). The ion yield is
depending on the number of electrons, their energy and length of interaction with the gas for
analysis:

I = N*Q(T) * p * L * i

I total ion current


N*Q(T) differential ionization, temperature dependent
p pressure in ionization chamber
L length of interaction between electrons and gas
i electron current (emission of cathode-filament)

10
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Quantitative analysis requires proportionality between the total ion current and the pressure
in the ionization chamber.
Highest ion yield and low dependence upon changes of the electron energy is found for most
gases in the range 70 eV to 100 eV for ionization energy. The high electron energy, beyond
the ionization energy of atoms and molecules, leads to complex fragmentation of the gases
to be analyzed (Fig. 6).

A) multiple ion formation of argon

B) fragmentation of a three atomic molecule

Fig. 6: Multiple charged ions A) and fragmentation of molecules B) through the electron
impact ionization process

The quadrupole mass filter separates ions according to their mass over charge ratio. The
applicable range of a mass filter suitable for coupling with thermal analysis can be up to
1024 atomic mass units (1 amu = 1.67*10-24 g). Systems for routine application show a unit
resolution, which means they separate the ions with mass m and m+1 amu.
Detectors with unit amplification (Faraday cup) are stable and robust, but not sensitive
enough for lower ion currents. Secondary electron multipliers (SEM) show amplification
factors up to 106 and are mostly applied for the measurement of low ion currents. A typical
mass spectrum is shown for carbon dioxide in Fig. 7.

11
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Fig. 7: Isotopes and fragmentation lead to a complex spectrum as shown for CO2 (44 amu),
with the key-fragments 28, 16 and 12 amu.

The fragmentation leads to an apparent complication of mass spectra, in most cases it is


however used as a key to identify single components in a mixture of gases. Library spectra
are available and often necessary to support the search for individual components.
Mass spectrometry is used in three operation modes in the coupling with thermal analysis,
which are the SCAN mode to register a certain mass range repeatedly, the MID mode for a
continuous monitoring of the changes of fixed, selectable mass numbers, and the SCAN-
BARGRAPH, which allows to display also quasi MID curves from selected mass numbers of
a scan.

Coupling techniques
In seldom cases one can arrange the quadrupole system directly within a high vacuum tight
thermobalance, e.g. to measure vapor pressures according to the Knudsen effusion method
with simultaneous identification of the vapor phase [1]. All other applications require a
pressure reduction between the thermobalance, operated at atmospheric pressure and the
high vacuum at the mass spectrometer.

Fig. 8: Pressure reduction in a two-step mass spectrometer coupling interface

12
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Various approaches were made to realize this pressure reduction; a comprehensive review is
given by Raemaekers [2]. Own experience was described in [3, 4] with double stage orifice,
Skimmer and capillary couplings, applied during the past 25 years. From the differences in
design one can deduct the differences in function, especially the sensitivity for the gas
detection and the influence of gas losses by condensation during the transfer. The very
perfect gas flow conditions in the Skimmer coupling are described below.

Fig. 9: Vertical arrangement of the two stage orifice-Skimmer-coupling-system at a top-


loading thermobalance and pressures at the different stages

The flow through the divergent nozzles with the formation of a compression zone depending
on pressure, dimensions and temperature is visualized and schematically described in
Fig. 10 A - B.

13
NETZSCH Coupling of Thermal Analysis and Gas Analysis

A) Flow distribution behind first orifice B) Schematical description of flow

Fig. 10: Flow distribution behind a divergent nozzle


A) picture of gas flow
B) scheme of flow

The Skimmer is arranged at the end of the compression zone to collect a parallel oriented
molecular beam out of the gas, which is under reduced pressure in this intermediate vacuum
range. The temperature dependent flow through this nozzle-Skimmer system is
compensated by a pressure regulation device, which keeps the pressure in the high vacuum
range for the ion source and the mass filter at a constant level. The influence of diffusion
between purge gas and evolved gases [5] with the effect of broadening and tailing of
detected peaks is low because of short distance from the place of gas evolution to the
detection.

14
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Fig. 11: Skimmer coupling integrated into the STA 409 furnace (2580014502000C)

The gas flow conditions in the Skimmer coupling lead to much higher sensitivity compared
with other coupling systems when looking on the same mass range; the difference in
sensitivity is increasing with selection of higher mass numbers.

The capillary coupling contains also two stage pressure reduction systems. The first step of
pressure reduction is achieved by pumping of the gas through a capillary of a length of 1 m
or 2 m and of very low diameter (e.g. 20 150 m) by means of a membrane or a rotary
vane pump. Thus 10-1 mbar to 10 mbar are measured at the end of the capillary. A second
stage is formed by an orifice (metal foil) and separates the intermediate stage from the high
vacuum in the quadrupole analyzer.

Fig. 12: Capillary coupling of a quadrupole mass spectrometer to the TG 209 C Iris

15
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Coupling of infrared spectrometers

Fourier-transform infrared spectroscopy is well established for the identification of organic


molecules in the solid phase, and it is therefore sometimes also favored for the identification
of volatiles during decomposition of organic materials. To be used as gas analysis system,
the infrared spectrometer must be equipped with a gas-measuring cell, in which the infrared
radiation comes into an intensive contact with all the gases (see above). The gas transfer
system between the thermobalance and the gas cell must ensure the complete and
unchanged transport of all evolved species; differently from the MS coupling no pressure
reduction is required for the FTIR coupling.

3 3
10 m bar 10 m bar exit

transfer line gas cell


Therm obalance FTIR Spectrometer

Fig. 13: Schematic coupling of an FTIR spectrometer to a thermobalance

Fig. 14: Arrangement of the gas-measuring cell of Bruker Vector 22 FTIR in the coupling with
the NETZSCH TG 209 C Iris

16
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Optimal gas flow conditions, appropriate length and volume of the transfer path, and
consequent heating guarantee a high sensitivity of gas detection also with FTIR. The
selection of purge gases is easier in FTIR couplings, as symmetrical molecules are not
activated by the radiation and will not disturb by a high background signal; only the gases
with a dipole moment are detected. Beside the mentioned application with organic materials,
FTIR couplings have proven to provide also very helpful information for the characterization
of many inorganic materials.

Calibration of coupled gas analyzers

The complex situation with the gas transfer in coupled TG-MS and TG-FTIR, and
temperature dependent flow properties, requires a calibration of the whole coupling and gas
analysis system for quantitative work. Beside the linearity of the gas detection for different
amount of evolved volatiles, the linearity as function of temperature has to be ensured [6].
The unique pulse thermal analysis technique (PTA-PULSETA) [7] is a perfect tool to achieve
the quantitative gas detection in separate calibration runs, or even on-line during a sample
measurement.

Principle of Pulse Thermal Analysis (PTA)

injected gas
Any gas can be injected
to the stream of the

carrier gas

TA MS/FTIR

by the special

pulse device TG-DSC gas composition

Fig. 15: Schematic set-up of the PULSETA

A known amount of gas is injected into the sample gas stream and the registered signal of
the resulting pulse is integrated. The proportionality between the injected gas quantity and
the peak areas was found for the TG-MS and the TG-FTIR couplings in a broad range of
temperatures and gas quantities. The injected gas can be selected according to the expected
evolved gases, so that an on-line calibration is possible for many of the volatiles. The
application of the PULSETA for increasing the sensitivity of thermoanalytical measurements,
for studying gas-solid reaction with stepwise control of the reaction by a pulsed supply of the
reactive gas, and for adsorption/desorption experiments will be described in the separate
contribution for PULSETA in this volume.

17
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Comparison of MS and FTIR for evolved gas analysis in combination with


thermal analysis

Both mass spectrometry and Fourier-transform infrared spectroscopy are very successfully
applied together with thermobalances and simultaneous TG-DTA and TG-DSC instruments.
Looking at the capillary coupling for the MS, this is nearly equally simple like the connection
of the FTIR by a transfer line. The Skimmer coupling requires higher instrumental efforts,
however this results in higher sensitivity compared with the capillary coupling by several
orders of magnitude.

The effect of electron impact ionization in a quadrupole MS contributes to the fragmentation


of molecules much more than the effect of irradiation by the infrared source, see table 1. The
fragmentation into key fragments is on the other hand an additional help in the interpretation
of mass spectra.

MASS SPECTROMETRY INFRARED SPECTROSCOPY


STRUCTURE/ IONISATION WAVE WAVE EQUIVALENT
EFFECT ENERGY/eV NUMBER/cm-1 LENGTH/m ENERGY/eV

ATOMS 5 ... 25 500 20.0 0.062

MOLECULES 8 ... 15 4000 2.5 0.496


ELECTRON 70 ... 100
IMPACT
IONIZATION IN MS

Table 1: Levels of ionization and irradiation energies in MS and FTIR

The tools for interpretation of spectra are relatively useful and advanced with FTIR, although
the number of gas spectra in computer libraries is still small compared with sample spectra
for solids and liquids. For mass spectrometry the usefulness of digital libraries is more
limited, as beside the fragmentation during ionization there is a further contribution to the
intensity of mainly lower mass numbers because of thermal dissociation and decomposition
along the gas path. For the detection of trace amounts of volatiles, MS shows in general the
higher sensitivity with detection limits in the ppb (parts per billion) range.

Literature:
[1] E. Kaisersberger, W. Hdrich, W.-D. Emmerich, Thermochim. Acta, 95 (1985)
331-336
[2] K.G.H. Raemaekers, J.C.J. Bart, Thermochim. Acta, 295 (1997) 1-58
[3] W.-D. Emmerich, E. Kaisersberger, J. Thermal Anal., 17 (1979) 197-212
[4] E. Kaisersberger, W.-D. Emmerich, H. Pfaffenberger, Thermochim. Acta, 88 (1985)
319-322
[5] B. Roduit, J. Baldyga, M. Maciejewski, A. Baiker, Thermochim. Acta, 295 (1997)
59-71
[6] M. Maciejewski, A. Baiker, Thermochim. Acta, 295 (1997) 95-105
[7] M. Maciejewski, C.A. Mller, R. Tschan, W.-D. Emmerich, A. Baiker,
Thermochim. Acta, 295 (1997) 167-182

18
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Quantification of the Spectroscopic Signals Recorded in


TA-MS and TA-FTIR Systems
Marek Maciejewski, Florian Eigenmann, Alfons Baiker
Laboratory of Technical Chemistry, ETH Zurich, 8092 Zrich

The main feature of the coupled techniques TA-MS and TA-FTIR is the identification of the
gaseous products, which together with the thermal effects (DTA) and mass changes (TG),
aids in interpreting the course of the investigated reactions. The qualitative analysis is
routinely done by comparing recorded spectra with key fragment ions and their relative
intensities for known elements and compounds (MS) and with reference spectroscopic
signals (FTIR). Spectroscopic data are available in the large libraries (see e.g. Gas Phase IR
Spectra and Mass Spectra in http://webbook.nist.gov/chemistry/ or the article "Internet
Resources for Mass Spectrometry" published in the J. of Mass Spectrometry, 34, 1 (1999)
available in http://userwww.service.emory.edu/~kmurray/mass-spec-resources.html). The
examples of FTIR and MS spectra of H2O, CO and CO2 are presented in Fig. 1 A and B,
respectively.

1.6 A
absorbance / a.u.

H O
2
1.2
0.8 CO
0.4
CO
2
0.0
4000 3500 3000 2500 2000 1500 1000 500
wavenumber / cm-1
100
B
Relative abundance

80
HO
2
60
CO
40 CO
2
20
m/z
0
10 15 20 25 30 35 40 45 50

Fig. 1: (A) FTIR and (B) MS spectra of water, carbon monoxide and carbon dioxide.

Due to the fragmentation of the investigated atoms and molecules during the ionization
process in the mass spectrometer (see e.g. the same mass-to-charge ratios of m/z=28 for
CO and CO2 in Fig.1B) and possible overlapping characteristic IR bands, the identification of
the gaseous species, especially in multicomponent systems, is sometimes difficult. However,
an even more complicated problem is the quantitative interpretation of spectrometric data,
which needs the calibration of the system, i.e. the determination of the relationship between
the observed intensities of the ion currents (MS) or absorbance (IR-spectra) and the amount
of the analyzed species.

19
NETZSCH Coupling of Thermal Analysis and Gas Analysis

The common method of calibration of spectrometric signals is time consuming and requires
the application of gaseous mixtures with well-defined composition. An additional difficulty is
that such calibration is usually performed at room temperature, while during the thermal
decomposition the temperature of the gas phase often changes considerably what can
render the results obtained by calibration at low temperature doubtful. The method of
calibration described here is based on introducing into TA-MS or FTIR systems a known
amount of the calibration gas. Specific amounts of the probe molecules can be introduced
into the system by two methods: (i) by decomposing solids via well-known, stoichiometric
reaction, or (ii) by injection of calibration gas into the carrier gas stream flowing with constant
rate through the system (PULSETA technique [1], commercially available by Netzsch). Both
methods were applied for quantification of the spectroscopic signals in the systems:TA-MS:
(STA 409-MS, Balzers QMG 420), and TA-FTIR: (TG209-Bruker Vector 22).

Figure 2 presents the 3-D FTIR diagram of the decomposition of sodium bicarbonate
(NaHCO3, gaseous products CO2 and H2O) with a rate of 10 K/min. To calibrate the traces
for evolved CO2, two pulses of CO2 (1 cm3 each) were injected before and after decom-
position process. The relationship between the mass of the reactant and intensity of CO2
signals (peak areas) is shown in Fig.3.

CO pulse
2

CO pulse
2

Fig. 2: Decomposition of NaHCO3 investigated by FTIR, 3-D diagram.

20
NETZSCH Coupling of Thermal Analysis and Gas Analysis

14
50 47.1 mg
12

amount of CO found by FTIR / mg


40 36.2
10
CO (FTIR) / a.u.

30 pulse at 50 C pulse at 250 C 8


Fig.2 Decomposition23.1
of NaHCO3 investigated by FTIR, 3-D diagram.
P P
1 2
20 6

2
10.5
2

4
10
5.15
2.46 2
0
0
0 500 1000 1500 2000 2500 0 10 20 30 40 50
A time / s B sample mass / mg

Fig. 3: A) Relationship between the mass of decomposed NaHCO3 and integral


intensity of CO2 traces recorded in TA-FTIR system
B) amount of CO2 found by FTIR as a function of the reactant mass.

A linear relationship was found between the amount of the analyzed species and the
corresponding spectrometric signals even when the sample mass was increased by a factor
of ten. This conclusion is illustrated by the results shown in Fig. 3B. Similar behaviour was
observed in the TA-MS system. Results presented in Fig. 4A illustrate also the linear
relationship between the integral intensities of MS signals m/z=44 and amount of injected
CO2.

9.5
integral intensity m/z = 44 (a.u)

2.0 9.0
8.5
8.0
7.5
intensity m/z = 44 / a.u.

7.0
6.5
0 1 2
pulse volume (ml) m/z = 44
intensity m/z = 44 / a.u.

1.0

0.5
0.25

0 10 20 30 40 50 0 200 400 600 800 1000


A time / min B temperature / C

Fig. 4: A) The intensities of MS signals as a function of the amount of injected CO2


B) Temperature dependence of the intensities of MS signals.
21
NETZSCH Coupling of Thermal Analysis and Gas Analysis

For capillary coupling the sensitivity of the spectrometric signals does not depend on the
temperature in the range 20 - 1000C. This enables quantitative evaluations of MS and FTIR
spectra obtained at any temperature using a single-point calibration. The dependence of the
integral intensity of the mass spectrometric signals (m/z = 44 resulting from the injection of
1 cm3 pulses of CO2 into the carrier gas helium) upon temperature is presented for the
capillary coupling in Fig.4B. However, for the Skimmer system, calibrations at different
temperatures are required.

The verification of the calibration procedure carried out by PULSETA technique can be done
by applying a decomposition process with well-known stoichiometry. Figure 5 depicts the
calibration of the m/z=44 MS signal using the decomposition of NaHCO3 and injection of
CO2. The reaction proceeds according to the equation:

2NaHCO3 Na2CO3 + H2O + CO2

The theoretical total mass loss during decomposition of NaHCO3 amounts to 36.90 weight%
(wt%) with the contributions of CO2 and H2O being 26.19 wt% and 10.71 wt%, respectively.
The mass of decomposed reactant was 35.66 mg; the mass loss (after correction of TG-
base-line) was 13.01 mg corresponding to 36.48 wt%. (0.99 of the stoichiometric value).
The integral signal of a 1 cm3 pulse of CO2 was 3.30 a.u. (As). The temperature of the
injecting loop was 30C, therefore the mass of injected CO2 amounted to 1.781 mg. The
integral signal m/z = 44 resulting from the evolution of CO2 had an intensity of 17.33 a.u.,
which, after comparing to the integral intensity of the injected pulse, is equivalent to 9.35 mg
of CO2.The stoichiometric amount of CO2 in the sample was 35.660.2619 = 9.34 mg. The
error in the determination of the CO2 content by MS was thus comparable to the error of the
thermogravimetric analysis. In routine experiments the applied method provides an accuracy
of 2-3 % and can be very easily verified, as described above, by comparing results obtained
by conventional thermoanalytical technique (TG) with results obtained by quantification of
spectroscopic signals (MS or FTIR) calibrated by injected pulses of the probe molecules.

162.5
TG
0

160.1

-10
mass loss / %

m/z = 18 HO
-20 2

P
m/z = 44 CO
-30 2

92.5 222.1

-36.5%
-40
50 100 150 200 250 300
temperature / C

Fig. 5: Decomposition of NaHCO3 during heating with a rate of 10 K/min in He


(50 ml/min), P- pulse of 1cm3 of CO2 injected for the calibration of m/z=44 MS
signal.
22
NETZSCH Coupling of Thermal Analysis and Gas Analysis

The comparison of the results obtained by both methods during decomposition of zinc
oxalate dihydrate (gaseous products: water, CO and CO2) illustrates well the advantages and
drawbacks of both, MS and FTIR, techniques in qualitative and quantitative analysis of the
gas phase.

Due to fragmentation of the evolved CO2 the results obtained using the TA-MS system
(Fig. 6) can not be used for unambiguous confirmation of CO formation without calibration of
the mass spectrometric signals. The main mass-to-charge ratio m/z=28 of CO is overlapped
by the same signal formed during ionization of CO2. For CO the abundance of m/z=28 signal
amounts to 100. According to the reference data, the relative abundance of m/z=44 and
m/z=28 signals in the mass spectrum of CO2 are 100 and 11-16 (depending on the ionization
energy), respectively. The interpretation of the m/z=28 signal for mixtures of CO and CO2 can
be thus misleading if the contributions of the CO2 fragments are not taken into account.

TG
0
P
CO
-10 31.30
m/z=18 (H O) P
-20 2 -17.3% CO
2 P
mass change / %

H
24.95 2
18.90
14.97
-30
m/z=28 (CO, CO ) 2.29
2

-40 m/z=44 (CO )


21.76
2
m/z=2 (H ) 3.65
2
-50

-55.3%
-60
0 100 200 300 400 500 600 700
temperature / C

Fig. 6: Thermal decomposition of ZnC2O42H2O investigated by TA-MS in He


atmosphere (50 ml/min). Pulses of injected CO, CO2 (4 cm3) and H2 (1 cm3)
are marked as P. Intensities of MS signals are expressed in arbitrary units.

The quantitative evaluation of MS signals is possible after injection of CO and CO2 pulses
before (or after) the decomposition of the oxalate. Firstly, quantifying fragmentation of
injected CO2 allows the determination of the ratio between m/z=28 and 44 signals which can
be used for the correction of the intensity of m/z=28 traces resulting from the evolution of CO.
Secondly, the signal intensities of both probe gases, compared to intensities of MS signals of
evolved molecules, allow the determination of the composition of the gas phase. Results of
the evaluation of the mass spectrometric signals are illustrated in the following.

23
NETZSCH Coupling of Thermal Analysis and Gas Analysis

During the decomposition proceeding according to the reaction:

ZnC2O42H2O 2H2O + CO + CO2 + ZnO

the mass losses due to the evolution of water, CO and CO2 amount to 19.00, 14.78 and
23.23 wt%, respectively. The mass of decomposed reactant was 36.95 mg.
The intensities of 4 cm3 pulses of CO and CO2 were 24.95 and 14.97 a.u., respectively. The
ratio of intensities of MS signals, m/z=28 and m/z=44, during injection of CO2 was 0.153. The
intensities of m/z=28 and 44 signals during oxalate decomposition were 31.30 and 18.90
a.u., respectively. After correction of the m/z=28 signal due to CO2 fragmentation, i.e. after
subtracting 0.15318.90 from 31.30 and taking as a reference the intensities of injected
pulses, the amount of CO and CO2 evolved during decomposition was found to be 5.13 and
8.99 mg, which amounts to 0.94 and 1.03 of the stoichiometric values.

CO
CO2

Fig. 7: Thermal decomposition of ZnC2O42H2O investigated by TA-FTIR.


Pulses of CO2 and CO were injected before the decomposition.

The results obtained using the TA-FTIR system agree well with those described above. After
quantification of CO and CO2 traces the ratio between evolved CO and CO2 was found to be
1.07 (1.09 determined by MS). The qualitative interpretation of the spectroscopic signals was
in this case easier than by MS due to the presence of bands characteristic for both gases.
However, the lack of the FTIR technique, to detect homonuclear diatomic molecules, did not
allow confirming the occurrence of the secondary reaction between CO and traces of water.
The formation of hydrogen in this reaction was confirmed using the TA-MS system. Detection
of H2 explained the lower than expected amount of evolved CO and the higher amount of
CO2 due to the reaction: CO + H2O = H2 + CO2.

24
NETZSCH Coupling of Thermal Analysis and Gas Analysis

The application of the described method for quantifying spectroscopic signals is shown in
Fig. 8 depicting the determination of the diamond content in grinding tools [3] and the amount
of organic residues in ZrO2 aerogel [4].

1.6
P C 492 8

intensity of m/z = 44 / a.u.


m/z = 18

int. m/z=18 / A * E-11

m/z = 44
853 CO pulse

int. m/z=44 / A * E-11


P 2
1.4 6
713 492
4
835 CO 1.2
2 357
C 2
709 1.0 m/z = 44
0
20 60
DSC 0 492
16 15.2 TG
mass change / %

-2 357 40
DTG

mass change / %
12 -4 2.5%

T / V
8 714 -6 20

4 +2.7 857 -8 DTA


0
TG -10
0 -1.1 9.6%
-12 -20
0 200 400 600 800 1000 1200 1400 0 200 400 600
A temperature / C B temperature / C

Fig. 8: A) Determination of diamond content by its combustion in air and quantification of MS


signal m/z=44 (CO2),
B) determination of the carbon content in ZrO2 aerogel.

Some of the special grinding tools contain synthetic diamonds embedded in the bonding
matrix made from Ni-Co alloy. During heating in air (Fig. 8A) the metals are oxidised (mass
gain) and diamond reacts with oxygen forming CO2 (mass loss). The formation of metal
oxides is overlapped at higher temperatures by the transformation of Co3O4 into CoO (mass
loss). Also the course of the exothermic reactions, namely the oxidation of the diamond and
the metals is overlapped by the endothermic reaction of Co3O4 decomposition. All these,
simultaneously occurring processes result in quite complicated shape of both, TG and DTA,
curves. The determination of the mass loss from the TG curve is uncertain, therefore the only
method for analysis of the diamond content was the quantification of the evolved CO2. The
comparison of the integral intensities (the area of the MS signal) of the injected pulses (1cm3)
with the CO2 signals resulting from diamond oxidation allowed quantification of the evolved
CO2, which in turn gave the amount of the diamond in the investigated sample. The diamond
content was 12.3 wt%.
Fig. 8B illustrates the application of the PULSETA technique for the determination of the
amount of organic residues in ZrO2 aerogel prepared by acid-catalyzed alkoxide sol-gel route
[4]. The observed mass loss is caused by the evolution of physisorbed water and oxidation of
the organic residues present in the aerogel. The formation of CO2, indicated by the
exothermic effect on the DTA curve and the accompanying signal of m/z=44 centred at
357C, indicates the presence of organic residues in the aerogel matrix. The calibration of
mass spectrometric signal by injection of 1cm3 CO2 at 550C allowed determining the carbon
content (1.9 wt%) which was confirmed by elemental analysis (2.0 wt%).
The investigated sample was X-ray amorphous. Its crystallization is represented by the sharp
exothermic peak centred at 492C. During crystallization, the rest of the residual species
trapped in the amorphous network is evolved. It manifests by the evolution of water (m/z=18)
and a small peak of CO2 (its magnification marked as C is presented in the inset located in
the left upper part of the figure). The amount of the carbon released during the zirconia
aerogel crystallization amounts to 0.007 wt%.
25
NETZSCH Coupling of Thermal Analysis and Gas Analysis

The method of quantification of spectroscopic data based on PULSETA technique has been
applied for the monitoring and determination of:
- CO2 content in ZrO2 and TiO2 supports and carbonaceous residues in gold supported
catalysts [5].
- Content of cobalt oxide in the solid products of cobalt oxalate decomposition [6].
- Course and products distribution of secondary reactions occurring in the gas phase
during CoC2O4 decomposition [6].
- Adsorption of hydrocarbons and their decomposition during interaction with metallic
iridium [7].
- CO oxidation over Au/TiO2 and Au/ZrO2 catalysts [5,8].
- Amount of sulfur and organic matter in the rock samples collected during petrol searching
[3].

Conclusion:

PULSETA enables the introduction of a well-defined amount of probe gas into TA-MS or TA-
FTIR systems at any temperature (non-isothermal) or time (isothermal mode). Since the
calibration can be done during the course of the reaction, the accuracy is significantly
improved compared to off-line calibration methods. Influences of changing experimental
conditions can be accounted for.

The possibility of in situ calibration of the spectroscopic signals with the PULSETA technique
increases greatly the potential of the coupled methods. By conventional analysis, only the
total amount of analyzed species can be measured, and it is impossible to resolve the
species evolved in multistep reactions. The results shown in Fig. 8 indicate that the range in
which the target species can be determined is very broad, between tens and 0.1, or, in
certain cases, even 0.01 wt%.

The method is fast (for capillary coupling only one or two pulses are required), simple
(no extensive calculations) and possesses good reproducibility. Measurements can be easy
confirmed by in situ calibration applying decomposition of solids with well-known
stoichiometry. User libraries of spectroscopic data can be set up which aid in qualitative and
quantitative interpretation of the experiments.

Adaptation of injection system for liquids is possible, as preliminary studies indicate. This
provision will greatly extend the opportunities of PULSETA.

Literature:

1. M. Maciejewski, C. A. Mller, R. Tschan, W.-D. Emmerich and A. Baiker, Thermochim.


Acta, 295 (1997) 167
2. M. Maciejewski and A. Baiker, Thermochim. Acta, 295 (1997) 95
3. M. Maciejewski, A. Baiker, E. Kapsch, M. Hollering and W.-D. Emmerich, GIT-Labor
Fachzeitschrift, June 2001
4. C. Stcker, M. Schneider and A. Baiker, J. Porous Mater., 2 (1966) 323
5. M. Maciejewski, P. Fabrizioli, J.-D. Grunwaldt, O. S. Becker and A. Baiker, Phys. Chem.
Chem. Phys., 3 (2001) 3846
6. M. Maciejewski, E. Ingier-Stocka, W.-D. Emmerich and A. Baiker, J. Therm. Anal. Cal., 60
(2000) 45
7. C. Wgerbauer, M. Maciejewski and A. Baiker, Appl. Catal. B, 34 (2001) 11
8. J.-D. Grunwaldt, M. Maciejewski, O.S. Becker, P. Fabrizioli, and A. Baiker, J. Catal., 186
(1999) 458

26
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Curing and Decomposition of Phenolic Resin


Anton Schranner, Stephan Knappe
NETZSCH-Gertebau GmbH
Wittelsbacherstr. 42, 95100 Selb

Application: Identification, Quality Control and Failure


Analysis

Goal of Analysis Determination of the low molecular weight vola-


tiles during polycondensation and the decom-
position behavior of a non-cured phenolic resin
by TG-FTIR

Instrument: TG 209 C Iris FTIR VECTOR 22

Sample: phenolic resin (powder)

Experimental conditions:

Sample mass: 8.383 mg


Crucible: Al2O3 (open)
Atmosphere: N2 (15 ml/min), ambient pressure
Temperature range: 30 850C
heating rate: 10 K/min
IR spectral resolution of 4 cm-1 at a time resolution
of approx. 19 s

Results:
The TG measurement of a non-cured phenolic resin in the temperature range up to 850C is
presented in fig. 1. For a more detailed discussion of the thermal behavior two regions were
separated, i.e. the range of curing (up to 320C) and the range of decomposition.

Fig. 1: TG measurement of a non-cured phenolic resin, mass loss signal (black) and
decomposition rate (1st derivative, green).
27
NETZSCH Coupling of Thermal Analysis and Gas Analysis

The calculated heat flow curve (c-DTA) in fig. 2 shows clearly an exothermal peak at 148C,
which is due to the curing of the sample. A DSC measurement in a closed, pressure-tight
crucible would provide a 3-step curing reaction in contrary to the obtained 1-step curing
reaction in an open pan. The information of c-DTA in combination with the detected mass
loss exhibits clearly a polycondensation, as expected [1].

The detected mass loss of 0.9% with a maximum decomposition rate at 69C (DTG-peak) is
due to the release of moisture. The curing reaction provides two mass loss steps of 0.8%
(DTG-Peak at 159C) and of 2.8% (209C).

Fig. 2: TG measurement of a non-cured phenolic resin, mass loss signal (black) and
decomposition rate (1st derivative, green). The calculated heat flow signal
(c-DTA, red) indicates an exothermal reaction (polycondensation) at 148C.

Single spectra of the simultaneous FTIR-measurement are recorded at 162C and at 210C
(fig. 3). Both spectra are equal beside differences in intensity. The identification with library
search exhibits water and ammonia, which is a critical gas in respect of health and
environmental aspects. A gas exhaust for the processing is needed. The appearance of the
evolved gases in the temperature range up to 320C is shown in fig. 4. The water trace (blue)
is integrated in the wave number range of 3574 3557 cm-1, the ammonia trace (green) in
the range of 973 952 cm-1.

28
NETZSCH Coupling of Thermal Analysis and Gas Analysis

0.07
0.06
0.05
0.04
Absorbance Units

H2O
0.03
0.02

210C
0.01

162C
0.00

NH3
-0.01

4000 3500 3000 2500 2000 1500 1000


Wavenumber cm-1

Fig. 3: Single spectra recorded at 162C (gray) and at 210C (black); the identification
confirms water (blue) and ammonia (green) as polycondensation products.

Fig. 4: The evolution of water (blue) and ammonia (green) in the temperature range up to
320C; Integration ranges: water (3574 3557 cm-1), ammonia (973 952 cm-1)

29
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Investigations are continued on the decomposition behavior of the (now) cured sample. The
extracted spectrum at 533C (black) in fig. 5 shows the appearance of phenol (blue) and
2,4-xylenol (green). Both the components are characteristic for phenolic resin [2]. The further
identification exhibits methane (red), carbon dioxide (absorbance peak at 2400 2250 cm-1)
and carbon monoxide (2250 2000 cm-1).
0.03 0
0.02 5
0.02 0
Absorbance Units
0.01 5
0.01 0
0.00 5
0.00 0

4000 3500 3000 2500 2000 1500 1000


W avenumber cm-1

Fig. 5: Extracted single spectrum at 533C (black); the identification confirms the
presence of phenol (blue), 2-4-xylenol (green), methane (red), carbon
dioxide (absorbance peak at 2400 2250 cm-1), carbon monoxide
(2250 2000 cm-1) and water (4000 3300 cm-1 and 2150 1250 cm-1)

30
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Fig. 6: Decomposition of the cured phenolic resin, mass loss signal (black)
and decomposition rate (dotted line), traces of the detected gases: phenol
(blue, 1630 1585 cm-1), 2-4-xylenol (green, 2950 2915 cm-1), methane
(magenta, 3025 3000 cm-1), carbon dioxide (red, 2400 2250 cm-1)

The gases evolved during the decomposition process are presented in fig. 6. The main
decomposition step of 30.7% is due to the degradation of the polymer, as detected by FTIR
(phenol and xylenol). The further mass loss of 11% is mainly assigned to the release of
carbon dioxide due to the cracking of the polymer chain.

Conclusion:

Investigations on the curing and decomposition behavior of a non-cured phenolic resin with
simultaneous TG-FTIR technique exhibit an excellent tool for quality control and identification
of the polymer by the release of phenol and xylenol. The obtained results show the behavior
of a typical phenolic resin with amine hardener. Any differences in the obtained spectra can
easily be used for quality control with respect to curing and decomposition.

Literature:

[1] Ehrenstein, G.W., Bittmann, E., Duroplaste, Carl Hanser Verlag, Mnchen, Wien (1997)
[2] Oberbach, K., Kunststoff-Taschenbuch, Carl Hanser Verlag, Mnchen, Wien (1998)

31
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Determination of Emission of Volatiles


during Paint Drying by TG-FTIR
Heinz Mhler, Helmut Walter*, Stephan Knappe**
*Studiengang Kunststoff- und Elastomertechnik, Fachhochschule Wrzburg-Schweinfurt,
Rntgenring 8, 97070 Wrzburg
**Netzsch-Gertebau GmbH, Wittelsbacherstr. 42, 95100 Selb

1. Introduction and Objective

Materials which are being processed under an increased temperature, have a more or less
high potential for emission of volatiles. In ASTM E 1868-97 [1] thermogravimetry is described
as a standard test method for the quantitative determination of the amount of volatile
components evolved from any sample, i.e. also form paints. It is used in processing and
quality control as well as for the determination of product specifications and application
conditions under practical temperature-time profiles. It must be particularly noted that this
method can be used only for the determination of mass loss, not directly for identification of
the sample material.

Nowadays emissions are always related to health and environmental aspects. The
possibilities of the TG-FTIR coupling for identifying such emissions during paint drying with
the example of a powder paint and hydro clear coat will be presented here.

2. Powder Paint

2.1. Measurements and Test Results

Measurements were carried out with a vacuum-tight micro thermobalance (TG 209 C Iris by
Netzsch-Gertebau), coupled to an FTIR spectrometer (Vector 22 by Bruker Optik,
Ettlingen). In addition to using a short, heated transfer line, the small volumes of the TGA
measuring cell and IR gas cell are very important in order to minimize changes of the gas
streams from the TGA cell on the way to the IR gas cell.

A polyurethane (PUR) powder paint system with blocked isocyanate hardener was
investigated. The hardening process has already been described and kinetically analyzed
in [2].

TG-FTIR Test parameters:


Crucible: Al, open
Temperature range: 20C ... 500C
Heating rate: 5 K/min
Atmosphere: N2 (45 ml/min).
The transfer line was heated to 200 C.

For the spectral resolution of the FTIR, 4 cm-1 was chosen. An average over 32 scans
(single spectra) was made with data acquisition every 20 s or 1.7 C, respectively.

32
NETZSCH Coupling of Thermal Analysis and Gas Analysis

2.2 Discussion of the Test Results

As shown in figure 1, the solid black TG curve represents the percentage mass loss in
nitrogen between 20 and 500C whereas the dotted blue curve represents the derivated
TG signal, or DTG curve. DTG peaks are related to the mass losses of the individual
decomposition steps. The percentage mass losses evaluated for the individual steps are
shown at the TG curve.

In the IR spectra of the gas phase (figure 2), measured at temperatures of DTG peaks and
shoulders, the intensity courses of the two adsorption bands (marked with arrows) at approx.
2275 cm-1 and 1735 cm-1 are especially characteristic for emissions from the paint, when
heated up to complete pyrolysis.

TG /% DTG /(%/min)

-0 .19 % -0 .80 %
100
-10 .69 %
0
85 .0 C 203.0 C
80
-1.0
-43 .93 %
60
-2.0
314.7 C

40
-3.0

-40 .57 %
20 -4.0

353 .4 C
410 .8 C
0 -5.0
50 100 150 200 250 300 350 400 450 500
Temperature /C

Fig. 1: TG and DTG curves of the PU powder paint 17.65 mg in a nitrogen atmosphere

33
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Fig. 2: FTIR Spectra recorded at temperatures of the DTG peaks or shoulders


presented in Fig. 1

For identification of the evolved gases the measured spectra were compared with spectra
from the EPA (Environmental Protection Agency) data base. The recorded spectrum at 85C
(lower curve in figures 2 and 3) was in best agreement with library data of methyl
methacrylate (upper curve in figure 3). The band at 1735 cm-1 can be related to the stretching
vibrations of the carbonyl group C=O of the ester group of the methyl methacrylate. The
bands between approx. 1450 cm-1 and 1150 cm-1 derive from the ether (C-O-C-) group and
the bending vibrations of the methyl-(CH3)- and methylene(CH2)-group. At approx. 2900 cm-1
CH2/CH3 stretching vibrations occurred.

34
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Library spectrum:
Methacrylic acid, methyl ester

Fig. 3: Identification of the FTIR spectrum of the gas evolution at 85C (DTG peak)

The measured spectrum at 203C is depicted in the lower part of figure 4. The EPA gas
phase data base [2] yields butyl isocyanate as best agreement with the measured spectrum.
The sharp band at 2275 cm-1 is a reliable indication for the isocyanate (N=C=O) group. The
evolution of both gaseous species agrees well with the composition of the powder paint.

Library spectrum:
Isocyanic acid, butyl ester

Fig. 4: Identification of the FTIR spectrum of the gas evolution at 203 C (DTG peak)

35
NETZSCH Coupling of Thermal Analysis and Gas Analysis

It may be interesting in practice to follow the intensities of both of the marked absorption
bands and thus determining dependency on temperature of the related gas evolution. Figure
5 shows the interesting temperature range from 20 to 250C for application and drying of the
paint. These so-called traces or spectral window plots represent the course of methyl
methacrylate evolution (blue curve) and of isocyanate evolution (purple curve). The TG curve
(black) and DTG curve (dotted black) as well as the DSC curve (red) are also presented in
this figure.

arb. DSC/ DTG/


TG /% units (mW/mg) (%/min)
TG 0.70 0.2
exo 0.25
DTG 0.1
99.5 0.60
0.20 0
99.0
0.50
203.0C -0.1
98.5
0.40 0.15 -0.2
98.0 DSC
0.30 -0.3
97.5 0.10
FTIR Traces: -0.4
97.0 204.9C 0.20
Methacrylic acid
0.05 -0.5
96.5 Isocyanic acid
88.1C 0.10
-0.6
96.0
0 0
0 50 100 150 200 250
Temperature /C

Fig. 5: FTIR traces of the methyl methacrylate band at 1735 cm-1 and of the butyl
isocyanate band at 2275 cm-1 together with the TG, DTG and DSC curves

The emission of methyl methacrylate begins during the glass transition and ends at approx.
110C. The isocyanate emission overlaps the exothermal DSC peak for the curing process.
This proves the existence of an additional reaction step used in the kinetic analysis of the
DSC measurements for curing of this paint [4]. Moreover, the overlapping of paint curing with
its decomposition, starting approx. at 240C, can be seen in the TG curve as well as in FTIR
traces.

The very weak emission of isocyanate during burn in of this powder paint requires an
exhaust system. Such devices are always necessary in technical processes when toxic
gaseous products are evolved, e.g. formaldehyde during injection molding of POM [5] or HCL
during decomposition of PVC [6] and [7].

36
NETZSCH Coupling of Thermal Analysis and Gas Analysis

3. Hydro Clear Coat


3.1. Measurements and Test Results
Measurements were carried out with the measuring system described in 2.
A two component (2 K) hydro clear coat system was investigated.

TG Test Parameters:
Crucible: Al2O3, open
Temperature range: 30C ... 600C
Heating rate: 5 K/min
Atmosphere: N2 (45 ml/min)
All other parameters as described in 2.

3.2 Discussion of the Test Results


The solid black line in figure 6 depict the mass loss curve (TG) and the dashed green line the
derivated mass loss curve (DTG). Only the temperature range of interest for practice was
evaluated. In the IR spectra recorded in the range of the first mass loss up to approx. 60C,
only the hyperfine structure lines of water vapor in the ranges from 4000 cm-1 to approx.
3400 cm-1 and from approx. 2200 cm-1 to approx. 1300 cm-1 can be recognized (figure 7).
Moreover, rudimentary CH2/CH3 valence vibration bands at approx. 2900 cm-1 and a band at
approx. 1200 cm-1, can be seen. This indicates that mainly water evaporates in this range as
well as a small amount of hydrocarbon-containing species which cannot clearly be identified.
The bands at approx. 2300 cm-1 are due to CO2, however this is not relevant for the FTIR
investigations on this hydro clear coat.

absorbance units

Fig. 6: TG and DTG curves as well as FTIR traces of the hydro clear coat,
sample weight 31.9 mg

37
NETZSCH Coupling of Thermal Analysis and Gas Analysis

3648.98
Absorbance

3500 3000 2500 2000 1500 1000


-1
Wavenumber / cm

Fig. 7: Recorded spectrum at 60C, indicating evolution of mainly water vapor

Looking at the spectrum in the range of the DTG peak at 95C in figure 8, the water vapor
hyperfine structure lines can still be seen, although they are very weak. Apart from the
CH2/CH3 bands at approx. 2900 cm-1, three further bands at 1763 cm-1, 1233 cm-1 and
1127 cm-1 are clearly visible. Due to identifications with the help of the EPA gas phase data
bank [3], which are not demonstrated here, the best hits were an acetate, taking the carbonyl
bands at 1763 cm-1 into consideration, and an aliphatic alcohol without consideration of this
band. Both seem to be reasonable regarding the small amounts of organic solvents used for
water coats, whereas the band at 1233 cm-1 is particularly typical for the alcohol.

38
NETZSCH Coupling of Thermal Analysis and Gas Analysis

1126.5
1763.4 1232.5
absorbance

4000 3500 3000 2500 2000 1500 1000


-1
Wavenumber / cm

Fig. 8: Spectrum at the DTG peak at 95C

For the above-mentioned objectives, it is very interesting to analyze the course of these
emissions as a function of temperature. For the water trace the hyperfine structure line
between 3660 cm-1 and 3638 cm-1 is used, for the acetate trace the carbonyl band between
1800 cm-1 and 1725 cm-1 and for the alcohol trace the band between 1270 cm-1 and
1210 cm-1. As can be seen in figure 6, the water trace (solid blue curve) is in very good
agreement with the DTG peak at 95C. The other two traces, i.e. alcohol (red) and acetate
(purple), show two maxima. Contrary to the maximum of the water trace, the first smaller
peaks are shifted to an approx. 5C higher temperature and the second peaks correlate in
both cases with the DTG peak at 154C. Moreover, compared to the alcohol trace, the
acetate trace shows a clearer shoulder which is in good agreement with the DTG shoulder at
122C. That means these two components are obviously responsible for the open time of
the paint during drying.

It is important to point out that the heights of the peaks of the traces have nothing to do with
the quantitative amount of the emissions. In order to make such a comparison, one should
know the values of the molar extinction (absorbance) coefficients for the bands which were
used for calculation of the traces and then evaluate the traces based on a relative absorption
scale.

The entire emission course in the temperature range of interest for the drying process of this
hydro clear coat can completely be described with the emissions of water and the two
organic solvents. Further, it is important to note that no other emissions can be detected. No
indication was found in the FTIR spectra that would explain the two stages of the alcohol and
acetate emissions.

39
NETZSCH Coupling of Thermal Analysis and Gas Analysis

4. Conclusion:

The TG-FTIR coupling yields very important information on the composition of the gas
phases evolved from materials which are processed at an increased temperature. Coupling
of TG and FTIR is a helpful tool which can be used not only in research and development,
but also in paint application and for the determination of product specifications, especially for
health and environmental reasons. Moreover, it can be employed for simulations with
different temperature/time programs.

5. Literature:

[1] ASTM E 1868-97: Standard Test Method for Loss-On-Drying by Thermogravimetry;


ASTM, West Conshohocken, PA
[2] S. Knappe, H. Mhler, J. Opfermann, H. Walter: Focus on: Thermal Analysis for Paints,
NETZSCH-Gertebau GmbH, Selb, 2001
[3] EPA-Gasphase-Library, Bruker Optik, Ettlingen, 1988
[4] J. Opfermann: Kinetic Analysis using multivariate non-linear regression; Thermal
Anal. Cal. 60 (2000), 641-658
[5] H. Mhler, M. Rhl, K. Siegel: Die instrumentelle Analytik - ein ausgezeichnetes
Werkzeug fr die Qualittssicherung, Schadenanalyse und Entwicklung in der
Kunststoff- und Elastomertechnik; Werkstoff und Technologie Seminar,
Rothenburg o.d.T., 12/2000
[6] H. Mhler, S. Knappe: Characterization of calcium/zinc (Ca/Zn) stabilizers and processing
agents for PVC pipe extrusion using the TG-FTIR coupling system,
Focus on: Thermal Analysis for Polymers, NETZSCH-Gertebau GmbH, Selb, (1998) 35
[7] E. Post, S. Rahner, H. Mhler und A. Rager: Study of recyclable polymer automobile
undercoatings containing PVC using TG-FTIR, Thermochim. Acta, 263 (1995) 1-6

40
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Recycling of Polyamides
Georg Matuschek, Antonius Kettrup, Maclovio Herrera
GSF-Forschungszentrum fr Umwelt und Gesundheit GmbH
Institut fr kologische Chemie, 85764 Neuherberg

PA 6 repeat unit [-NH-(CH2)5-CO-]


PA 66 [-NH-(CH2)6-NH-CO-(CH2)4-CO-]

Application: Recycling of Polyamides (PA). Polyamides are


often used in carpet industry

Goal of Analysis: Determination of different PA types, identification


of impurities through the identification of specific
degradation products by TA-MS/GC-MS.

Instrument: Macro-STA 419 QMG 420

Sample: Powder
Experimental conditions:
Sample mass: 1g
Crucible: Special Macro TA crucible (Al2O3)

Temperature program: 25 C 800 C, 2.6 K/min


Atmosphere: N2; 100 ml/min
Mass spectrometer: For simultaneous MS measurement, two
channels were selected.
Channel 1: scan m/z 10 60; range 10-8 A
Channel 2: scan m/z 59 200, range 10-9 A
GC-MS Analysis Adsorption on XAD-4 Resin, desorption using
cyclohexane, HP6890 GC-MS Chemstation

Results:

Aliphatic polyamides like PA 6 and PA 66 are widely used in the carpet industry. For a
recycling of these materials, a strict separation of the different PA types is necessary. For
this reason, a method for the determination of PA types as well as impurities of other PAs is
needed.
The thermal degradation of aliphatic polyamides was studied in the past systematically and
the detection of NH3, H2O, CO, CO2 as well as HCN was described in the literature [1-10].
The main degradation product of PA 6, -caprolactam, can be achieved in amounts of 85 %
beneath some oligomers, aliphatic nitriles [11-13, 18]. The main degradation product of
PA 66 is cyclopentanone [9, 11-12, 15-17], but also some aliphatic hydrocarbons, nitriles and
dinitriles were reported [11, 14-15, 18]. From PA 12 lauryllactam and some di- and trimers
are the primary degradation products [12, 18-20], and PA 612 produces mainly cyclic
oligomers [12] as degradation products.
The thermoanalytical curves of PA 6 are given in Fig. 1. As it can be seen from these curves,
the degradation starts at 335 C and continues in one step up to 96 % weight loss.
The time/temperature depending occurrence of characteristic mass fragments from the
degradation of PA 6 can be seen in the 3D overview in Fig. 2.
41
NETZSCH Coupling of Thermal Analysis and Gas Analysis

0 0

-5
-40
-10

-15
-80

-20

-120 -25
0 100 200 300 400 500 600 700 800

Fig. 1: TG, DTG and DTA curves of PA 6 in nitrogen

1,20E-08

8,00E-09
Intensity

4,00E-09

re
0,00E+00 600 tu
10 20 400 ra
30 40 50 60 70 80 200 pe ]
90 100 m C
T e [
110 120
m/z

Fig. 2: 3D plot of mass spectra of PA 6 in nitrogen

The mass spectrum of caprolactam as well as that of the degradation products at the DTG
peak (441 C) are given in Fig. 3. The fragments m/z 16 (NH2+), 18 (H2O) und 44 (CO2) show
the highest intensities, but the fragments of caprolactam m/z 31, 42, 55, 67, 84 and 113 are
clearly to identify. Also the chromatogram of the degradation products adsorbed during the
thermal analysis shows the main product caprolactam beside some nitriles (cf. Fig. 4).
42
NETZSCH Coupling of Thermal Analysis and Gas Analysis

120
Mass-Spectrum of Caprolactam
100
1,2E-08
80

Intensity
60

40
Intensity

8,0E-09
20

0
10 20 30 40 50 60 70 80 90 100 110 120
4,0E-09 m/z

0,0E+00
10 20 30 40 50 60 70 80 90 100 110 120
m/z

Fig. 3: Mass spectra of the degradation products of PA 6 and of caprolactam

2,4E+6

2,0E+6
Abundance

1,5E+6
O

NH
1,0E+6

Nitriles

0,5E+6

0
4 8 12 16 20 24 26
Time (min)

Fig. 4: Off-line chromatogram of the degradation products of PA 6

The thermal degradation of PA 66 is quite similar to that of PA 6. As can be seen from Fig. 5,
the degradation starts at about 350 C and continues in one step to 95 % weight loss.

43
NETZSCH Coupling of Thermal Analysis and Gas Analysis

0
0

-5
-60
-10

-15
-120

-20

-180 -25
0 100 200 300 400 500 600 700 800

Fig. 5: TG, DTG and DTA curves of PA 66 in nitrogen

The mass spectrum of the degradation products at the DTG peak (431 C) in Fig. 6 show
clearly the presence of ammonia (NH3+ m/z 17), water and carbon dioxide (m/z 18 and 44)
with high intensities. The fragments at m/z 41, 55 und 84 are due to the evolution of
cyclopentanone. The off-line chromatogram of the adsorbed degradation products shows
cyclopentanone as the main product (Fig. 7).

120

100
Mass-Spectrum of Cyclopentanone
80
Intensity

1,2E-08 60

40

20

8,0E-09
Intensity

10 20 30 40 50 60 70 80 90 100
m/z

4,0E-09

0,0E+00
10 20 30 40 50 60 70 80 90 100
m/z

Fig. 6: Mass spectra of cyclopentanone and of the degradation products of PA 66

44
NETZSCH Coupling of Thermal Analysis and Gas Analysis

3,2E+6

O
2,4E+6
Abundance

1,6E+6

8,0E+5

0
4 8 12 16 20 24
Time (min)

Fig. 7: Off-line chromatogram of the degradation products of PA 66

Due to the sensitivity of MS and GC-MS one can clearly distinguish the different main
thermal degradation products of polyamides and additionally identify further fragments.

Conclusion:

The investigations have shown, that the combination of Macro TA-MS with additional GC-MS
analysis of the degradation products ensure the distinguishing between the two polyamides
PA 6 and PA 66 as well as to determine the contamination of one type with the other.

Literature:

[1] M. Chaigneau, G. Le Moan, Annales pharmaceutiques franaises, 31 (1973) 495-501


[2] N. H. Margossian, Plastiques Modernes et Elastomres, Janvier-Fvrier (1975) 119-121
[3] J. Michal, J. Mitera, J. Kubt, Fire and Materials, 5 (1981) 1-5
[4] W. Kaminsky, Makromol. Chem., Macromol. Symp. 48/49 (1991) 381-393
[5] M. Nielsen, P. Jurasek, J. Hayashi, E. Furimsky, J. Anal. Appl. Pyrolysis, 35 (1995)
43-51
[6] C. Shimasaki, N. Watanabe, K. Fukushima, S. Rengakuji, Y. Nakamura, S. Ono,
T. Yoshimura, H. Morita, M. Takakura, A. Shiroishi,Polym. Deg. Stab. 58 (1997) 171-180
[7] C. Siat, S. Bourbigot, M. Le Bras, Polym. Deg. Stab. 58 (1997) 303-313
[8] M. Herrera, G. Matuschek, A. Kettrup, J. Therm. Anal. Cal. 59 (2000) 385-394
[9] D. O. Hummel, In C.E. Roland Jones, Carl A. Cramers, Analytical Pyrolysis. Elsevier
Scientific Publishing Co., Amsterdam (1976) 117-138

45
NETZSCH Coupling of Thermal Analysis and Gas Analysis

[10] I. Lderwald, F. Merz, M. Rothe, Die Angewandte Makromol. Chemie, 67 (1978)


193-202
[11] H. Ohtani, T. Nagaya, Y. Sugimura, S. Tsuge, J. Anal. Appl. Pyrolysis, 4 (1982) 117-131
[12] A. Ballistreri, D. Garozzo, M. Giuffrida, G. Impallomeni, G. Montaudo, Polym. Deg. Stab.,
23 (1988) 25-41
[13] S. V. Levchik, L. Costa, G. Camino, Deg. Stab., 36 (1992) 229-237
[14] M. Svoboda, B. Schneider, J. tokr, Collect. Czech. Commun., 56 (1991) 1461-1476
[15] A. Ballistreri, D. Garozzo, M. Giuffrida, G. Montaudo, Macromolecules, 20 (1987)
2991-2997
[16] U. Bahr, I. Lderwald, R. Mller, H.-R. Schulten, Die Angewandte Makromol. Chemie,
120 (1984) 163-175
[17] T. B. Blease, G.A. Paterson, J.H. Srivens, B. Polym. Journal, 21 (1989) 37-44
[18] I. Lderwald, F. Merz, Die Angewandte Makromol. Chemie, 74 (1978) 165-185
[19] S. V. Levchik, L. Costa, G. Camino, Polym. Deg. Stab., 43 (1994) 43-54
[20] V. Mailhos-Lefievre, D. Sallet, B. Martel, Polym. Deg. Stab., 23 (1989) 327-336

46
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Characterization of Epoxy-based Electronic Waste

Ekkehard Post
NETZSCH-Gertebau GmbH, Wittelsbacherstr. 42, 95100 Selb

Application: Environmental protection

Goal of Analysis: Comparison of the thermal behavior of raw and


treated electronic waste

Instrument: STA 409 C Skimmer coupling

Sample: Electronic waste material

Experimental conditions:
Sample mass: 10 mg
Temperature programme: RT to < 800 C
Heating rate: 10 K/min
Crucibles: Pt/Rh
Sample holder: TG-DSC type S
Atmosphere: He, 75 ml/min
Mass spectrometer: Scan mode
recipient pressure: 4 x 10-6 mbar

General Comments:

Recycling of electronic waste materials is a problem in the highly industrialized countries.


These materials contain plastics, ceramics, glasses and metals. The noble metals are often
removed by chemical treatment. The polymer components are usually removed by thermal
treatment. During this firing process toxic gases can be produced, especially when there are
aromatic compounds and/or fire retardant substances present. Electronic boards were very
often produced from epoxy resins with bromine as fire retardant additive.

47
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Fig. 1: Pyrolysis of electronic waste material, TG curve and mass spectra

Fig. 2: Electronic waste without treatment, TG curve and quasi-MID MS curves for
Br 79 and 81 amu and also for HBr 80 and 82 amu

48
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Fig. 3: Electronic waste treated with acids, Br and HBr evolution is found already
as of 100 C

Results:

Fig.1 depicts a part of the TG-MS result on a ground electronic waste sample. It can clearly
be seen that aliphatic and aromatic fragments over the mass range of 110-180 amu occur.
Fig. 4 depicts the formula of an epoxy resin and some of the fragmentation possibilities.

Fig. 4: Epoxy resin with some possible fragmentations

49
NETZSCH Coupling of Thermal Analysis and Gas Analysis

During chemical treatment (for example with acids to remove the metals) the polymer can
also be modified. Fig. 2 shows the untreated, only milled raw material. Mass numbers 79, 80,
81, 82 represent mainly the Br or HBr which was used in this material as a flame retardant.
At around 200C small amounts of the flame retardant were evolved followed by the main
decomposition at 300C. The situation for the acid-treated sample is quite different: here
already at 100C the evolution of bromine is seen (fig. 3) due to the physical-chemical
change of the polymer caused by the acid treatment.

Conclusion:

The TG-MS method is very useful tool for studying the pyrolysis of polymers. The thermal
stability and also the kind of evolved gases during the decomposition can be investigated.
Electronic waste is treated with acids to remove noble metals or other valuable materials
before the combustion of the polymeric parts. These treatments often change the behavior of
the polymers which could be studied by TG-MS. Further the evolved gases can be identified,
which is of great interest from the environmental point of view.

Literature:

E. Post, E. Kaisersberger, S. Knappe, W.-D. Emmerich, J. Therm. Anal. Cal., 57 (1999)


265-273

50
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Development of Undercoating Materials Containing


Polyvinylchloride, PVC

Albrecht Rager
BRUKER Optik GmbH, Rudolf-Plank-Str. 23, 76275 Ettlingen

PVC repeating unit

CH CH2

Cl n

Application: Standard thermoplastic. PVC is often used in


automobile undercoating materials.

Goal of Analysis: Determination of the HCl content/release of two


different undercoating materials by TG-FTIR.

Instrument: TG 209 C IRIS FTIR VECTOR 22


Evaluation: Decomposition process by determination of mass
loss and composition of evolved gases by
TG-FTIR.

Sample: Powder, Paste

Experimental conditions:
Sample mass: approx. 10 mg
Crucible: Al2O3, open
Temperature program: 20 min isothermal at 30C;
30C 800C, 15 K/min
Atmosphere: N2; 25 ml/min
FTIR: For simultaneous FTIR measurement, a spectral
resolution of 4 cm-1, in the spectral range of
650 - 5000 cm-1 at a time resolution of approx.
15 s, was selected.

51
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Results:

The decomposition behavior of pure PVC in comparison with two undercoating materials
sample 1 and 2 is presented in Fig.1. Rapid decomposition of the PVC starts at approx.
300C. Two further mass-loss steps can be observed with increasing temperature [1].
Sample 1 already shows a slight mass loss during the isothermal segment at 30C, due to a
loss of water as easy detectable by FTIR. Decomposition starts with a mass loss of 5 % at
200C. This mass loss is followed by a rapid mass-loss step at approx. 300C. The first small
mass-loss step of sample 2 is less expressed compared to sample 1. The rapid mass loss at
approx. 300C is slightly smaller than in pure PVC.

The FTIR spectra monitored in the rapid decomposition range (300450C) are represented
in the three-dimensional plot (Fig. 2). The rotation lines characteristic for HCl appear in the
wave number range 28502500 cm-1.

Semi quantitative conclusions can be drawn through integration of the HCl rotation lines. The
relative intensities plotted as a function of time are so-called frequency traces. The traces
indicate the evolution of HCl during the decomposition of the pure PVC and the two different
undercoatings (Fig. 3).

TG/ % Temperature /C
100.0 900
PVC pure
90.0 800

80.0 700
70.0
600
60.0
500
50.0
Sample 1
400
40.0

30.0 300
Sample 2
20.0 200

10.0 100

0 10 20 30 40 50
Time /min

Fig.1: TG measurements of pure PVC in comparison with two different undercoatings


sample 1 and sample 2

52
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Fig. 2: Rotation lines of the infrared spectrum of HCl.

PVC pure

Fig. 3: Single spectrum of HCl and HCl traces during the decomposition of pure
PVC and two different undercoating materials, sample 1 and sample 2.

53
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Conclusion:

Simultaneous TG-FTIR measurements can be taken into consideration for the formulation
development of undercoating materials based on PVC. Determination of the plasticizer
system as well as the content of HCl, which is released from the PVC component, can be
investigated by decomposition of the undercoating material. The comparison of the HCl
intensities plotted as a function of time or temperature indicates the differences in the content
of HCl in the materials.
In the investigated PVC and the undercoating samples the lowest content of volatile HCl was
observed for the sample 2.
A quantitative calculation of the HCl content by determination of the mass-loss step at the
maximum decomposition rate temperature (approx. 300C) is not possible due to the
overlapping of the HCl emission with evolution of other gaseous products formed during the
bulk decomposition. A more accurate quantitative result of the produced HCl amount can be
achieved by the PULSETA method [2].

Literature:

[1] E. Post, S. Rahner, H. Mhler, A. Rager, Thermochim. Acta, 263 (1995) 1-6
[2] M. Maciejewski, et al., Thermochim. Acta, 295 (1997) 67

54
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Softener Identification by TG-FTIR: Fomblin Y 06


Measurements under Reduced Pressure

Albrecht Rager
BRUKER Optik GmbH, Rudolf-Plank-Str. 23, 76275 Ettlingen

Fomblin:
CF3 - (O - CF - CF2 )n - (O - CF2 )m - O - CF3
|

CF3

Application: Perfluorinated compounds are often used as high


temperature stable elastomers
Fomblin (supplier: AUSIMONT S. p. A.
Bollante (Mi) / Italy):
Mol. Weight: 1800 g/mol,
vapor pressure: 1.3 x 10-5 mbar at 20C
Fomblin was identified as softener.

Goal of Analysis: Identification and quantification of high boiling,


decomposing softeners by TG-FTIR under
reduced pressure.

Instrument: TG 209 C IRIS FTIR VECTOR 22,


rotary pump with pressure regulation system.
Evaluation: Determination of mass loss and identification of
evolved gases by TG-FTIR.

Sample: slices of fluorinated elastomer / (reference


Fomblin sample as liquid)

Experimental conditions:
Sample mass: 31.98 mg
Crucible: Al2O3, open
Temperature program: 30C 500C, 10 K/min
Atmosphere: N2; 5 ml/min inlet at 1013 mbar,
Internal pressure reduced to appr. 100 mbar.
FTIR: For simultaneous FTIR measurement, a
resolution of 4 cm-1, in the spectral range of
650 5000 cm-1 at a time resolution of approx.
15 s, was selected.

55
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Results:

One of the current limitations of TGA-FTIR coupling systems is the temperature of the
transfer line and of the gas cell in the infrared spectrometer. The temperature of the interface
to the FTIR cannot be as high as the temperature that can be achieved in the furnace, and
this leads to possible condensation of high boiling compounds that are evolved from the
sample. Due to the reduced pressure, the boiling point of the softener is lowered, therefore
its evolution does not overlap the decomposition of the elastomer thus allowing more
accurate determination of the amount of softener from the TG curve. Under ambient pressure
the release of the high boiling (decomposing) softener starts together with the destruction of
the polymer material itself. To determine whether the measured spectra of the released
fragments are related to the softener or the bulk polymer is generally impossible. Under
reduced pressure the softener does not decompose and starts to evolve before the
elastomeric structure breaks down.

DTG /(%/min)
TG /%
Gram Schmidt
100 2
-10 % 0.18
90 0
0.16

80 -2 0.14

0.12
70 -4
ptot = 100 mbar 0.10
60 -6 0.08

50 -8 0.06

0.04
40 -10
0.02
30 -12 0
100 200 300 400 500
Temperature /C

Fig.1: TG, DTG and total IR chromatogram (Gram Schmidt) measured at a


perfluorinated elastomer O-ring under reduced pressure (p = 100 mbar).

56
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Fig. 2: 3D-IR presentation of the perfluorinated elastomer O-ring. Measurement


taken under reduced pressure of 100 mbar.

ptot = 100 mbar

HF T = 460C

Fomblin
T = 370C

4000 3500 3000 2500 2000 1500 1000


-1
Wavenumber/cm

Fig. 3: Single spectra of the evolved gases during the decomposition of a


perfluorinated O-ring. At 370C the softener spectrum can be measured.
For identification a spectrum of pure Fomblin was taken under the same
conditions. At 460C additional bands (marked by arrow) and HF are
identified due to the beginning decomposition of the elastomer.

57
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Conclusion:

Simultaneous TG-FTIR measurements under reduced pressure can be taken into


consideration for the quantification and identification of plasticizers, evolved under ambient
pressure at the starting decomposition of the polymer. By reducing the pressure, the boiling
temperature of plasticizer decreases what leads to better reparation of simultaneously
occuring processes: evaporation of plasticizer and decomposition of elastomer. This allows
more accurate determination of the amount of plasticizer.
Furthermore due to the lowered boiling point, decomposition of large molecules, normally
taken into account under ambient pressure, may be overcome. Additionally, lowering the
pressure prevents the decomposition of the large molecule of the plasticizer.
The additional information obtained in this way can help elucidate both the structure of the
material and the chemistry of the decomposition process [1].

Literature:

[1] R. S. Jackson, and A. Rager, Thermochim. Acta, 367-368 (2001) 415- 424

58
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Decomposition Behavior of Acetylsalicylic Acid (ASA)

Albrecht Rager
BRUKER Optik GmbH, Rudolf-Plank-Str. 23, 76275 Ettlingen

O COCH3

OH

Application: Decomposition behavior and thermal stability of drugs in formulations

Goal of Analysis: Monitoring of decomposition steps and


identification of evolved gases by TG-FTIR.

Instrument: TG 209 C IRIS FTIR VECTOR 22


Evaluation: Decomposition process by determination of mass
loss and evolved gases by TG-FTIR.

Sample: Powder

Experimental conditions:
Sample mass: 9.89 mg
Crucible: Al2O3, open
Temperature program: 30C 600C, 10 K/min
Atmosphere: N2; 45 ml/min
FTIR: For simultaneous FTIR measurement, a spectral
resolution of 4 cm-1, in the spectral range of
650 5000 cm-1 at a time resolution of approx.
15 s, was selected.

Results:

The total IR chromatogram (Gram Schmidt) along with the thermal decomposition behavior
as a function of temperature and time (Fig. 1) shows the hydrolyses of the ester group of the
acetylsalicylic acid into acetic acid and salicylic acid during the first thermal decomposition
step. With the aid of the TG-FTIR coupling (Fig. 2), it is easy to follow this step and to obtain
the IR spectrum of the more readily volatile acetic acid (Fig. 3). A comparison with the
spectral database simplifies interpretation. At higher temperatures, in a further decomposition
step, the decarboxylation of the salicylic acid is recorded with a strong release of phenol and
CO2 (Fig. 4) [1, 2, 3].

These results allow us to monitor the complete decomposition steps of the acetylsalicylic
acid and to set up the decomposition reaction scheme as shown in Fig.5 [4].

59
NETZSCH Coupling of Thermal Analysis and Gas Analysis

TG /% DTG /(%/min) -2
Gram Schmidt *10
100.0
0
90.0 3.0
-34.08 -1.0
80.0
-2.0 2.5
70.0
-3.0
60.0 2.0

50.0 -4.0
1.5
40.0 -5.0
-54.89
30.0 -6.0 1.0

20.0
-7.0 0.5
10.0 176.7
344.6C
-8.0
100 200 300 400 500 600
Temperature /C

Fig. 1: TG, 1st derivative (DTG) and total-IR-chromatogram (Gram Schmidt) of


ASPIRIN decomposition.

Fig. 2: 3D-IR presentation of the ASPIRIN decomposition. The two


decomposition steps can clearly be seen by producing different eluents.

60
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Acetic acid

4000 3500 3000 2500 2000 1500 1000


-1
Wavenumber/cm

Fig. 3: Extracted spectrum at a decomposition temperature of approximately 175 C


(black) versus a reference spectrum of acetic acid out of a digital library (red).

Phenol

Salicylic acid

4000 3500 3000 2500 2000 1500 1000


-1
Wavenumber/cm

Fig. 4: Extracted spectrum at a decomposition temperature of approximately 345 C


versus reference spectra of salicylic acid and phenol out of a digital library. The
carbon dioxide (blue) can be monitored in the range of 2300 - 2400 cm-1.

61
NETZSCH Coupling of Thermal Analysis and Gas Analysis

O COCH3 Acetylsalicylic acid


(Aspirin)
O

OH

+ H2O
- CH3COOH
Salicylic acid
OH
bp. at 1013 mbar, 211C
O

OH

- CO2
OH Phenol
bp. at 1013 mbar, 181C

Fig. 5: Reaction scheme for the decomposition of Acetylsalicylic acid. The water
in the first hydrolysation step is mostly supplied by the formulation agents.

Conclusion:

The supply of medicines is dependent on the storage stability of the active ingredient in the
formulation. Therefore, the decomposition behavior of the active ingredient is very important
in determining the shelf life of a pharmaceutical product. Knowledge of the decomposition
products from the active ingredient is useful in helping to avoid the formation of undesired by-
products during storage. TG-FTIR is excellently suited to identify the products of drug
decomposition.

Literature:

[1] A. Rager, 10th International Conference on FT-Spectroscopy, (1995), Budapest, Hungary


[2] A. Rager, STK Swiss Thermal Society, Valbella meeting [1994], Valbella, Switzerland
[3] A. Rager, and L.A. Giersig, Symposium on Thermal Analysis [1995], Sopron, Hungary
[4] A. Rager, and E. Kapsch, GIT, 4 (2000) 370

62
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Binder Burnout of an Al2O3 Green Body

Ekkehard Post
NETZSCH-Gertebau GmbH, Wittelsbacherstr. 42, 95100 Selb

Application: Characterization of the binder burnout from raw


material for technical ceramics

Goal of Analysis: Identification of the evolved gases before


sintering of an Al2O3 green body

Instrument: STA 409 C Skimmer coupling

Sample: Al2O3 green body

Experimental conditions:
Sample mass: 125.30 mg
Temperature programme: RT to 1400 C
Heating rate: 10 K/min
Crucibles: Pt/Rh, open
Sample holder: TG/DSC type S
Atmosphere: air, 75 ml/min
Mass spectrometer: Scan mode
recipient pressure: 6 x 10-6 mbar

63
NETZSCH Coupling of Thermal Analysis and Gas Analysis

mass / %

Temperature / C

Fig. 1: Evolution of water (17, 18 amu) and carbon dioxide (44 amu) during firing of a
green body

Fig. 2: Al2O3 green body: evolution of water, CO2 and organics (small intensities)

64
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Fig. 3: Single scan during binder burnout at 264C (mass range 50 90 m/z)

Fig. 4: Evolution of organics during the binder burnout from an Al2O3 green body

65
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Results:

Fig. 1 depicts the TG, DSC and the MS intensities of H20 (17,18 amu) and CO2 (44 amu)
versus temperature. During the first two TG steps water is evolved. CO2 is found during
burning of the organics and decomposition of the carbonate content. From the DSC curve
the exothermal combustion energy (200 400C) can be calculated. This exothermal effect
is overlapped by the endothermal release of water. Also the phase transition of the crystalline
quartz content (endothermal peak at about 570 C) is detectable.

In fig. 2 the H2O and CO2 evolution is shown as a 3-D plot. Also organic fragments with lower
intensity were detected. This can clearly be seen in fig. 3. In the upper part of the figure the
mass range 50-90 amu is shown for a scan at about 264 C. With the evolution of the
organics an exothermal effect in the DSC curve is observed (lower part of fig. 3).

Fig. 4 shows a 3-D plot for the mass range (50-90 amu). It can be concluded that during the
binder burnout organic fragments were also evolved in the lower temperature region without
the complete combustion to CO2 although the measurement was performed in air
atmosphere.

Conclusion:

For shaping of parts from non-plastic ceramic powders, organic binders are often employed.
After forming the green body by pressing or casting, the binder must be removed (normally
by pyrolysis) before the sintering process can be started. The binder must be removed
without leaving residue from the binder raw material and also by avoiding the creation of
pores or gas bubbles in the raw material (exception: required porosity, e.g. for catalyst
carriers). Also from environmental aspects, identification of the evolved gas species is very
important since a lot of gaseous products are generally produced during firing of ceramics.
With this knowledge decisions can be made regarding the necessity of filter systems or post-
burning of the gases. By TG-DSC-MS the temperatures, rates of decomposition, calorimetric
data and evolved gases can be determined.

66
NETZSCH Coupling of Thermal Analysis and Gas Analysis

HF Emissions at the Firing Process of Brick Clays


Elisabeth Kapsch* and Ekkehard Post**
*Netzsch-Gertebau GmbH, Office Netzsch do Brazil, BR-04 556-000 Sao Paulo/SP
**Netzsch-Gertebau GmbH, Wittelsbacherstr. 42, 95100 Selb

Application: Brick clays show a complex mass loss behavior


during firing. The detection of evolved gases like
HF has become of actual importance because of
its possible impact on the environment [1, 2].
Building materials.

Goal of Analysis: Detection of the HF release by TA-FTIR/MS.

Instrument: TG 209 IRIS FTIR VECTOR 22


STA 409 C MS ThermoStar
Evaluation: Decomposition process by determination of mass
loss and the evolved gases HF and H2O by
STA-FTIR/MS.

Sample: grinded powder of brick clay, pre-dried

Experimental conditions:
Sample mass: 94.51 mg (TG-FTIR); 2654.99 mg (STA-MS)
Crucible: Pt, open (TG-FTIR);
Al2O3, lid with ca. 0.5mm hole (STA-MS)
Temperature program: 30 1000C, 20 K/min (TG-FTIR)
RT 1250C, 10 K/min (STA-MS)
Atmosphere: air
FTIR: For simultaneous FTIR measurement, a
resolution of 4 cm-1, in the spectral range of
650 5000 cm-1 at a time resolution of approx.
15 s was selected.

Results:
The simultaneous STA-MS measurement shows several mass-loss steps in the temperature
range between RT and 1250C (fig. 1). The total mass loss detected in the TG curve
amounts to nearly 10 %. In order to follow the emission of water and HF, quasi-MID curves
for 18, 19 and 20 amu were calculated from the bargraph scans. By correlation of these MID
curves with the TG curve it can be seen that the release of gaseous species with mass to
charge ratios (m/z or amu) 18, 19 and 20 is contributing to the main mass loss steps during
firing of the brick clay. At 1020C the evolution of a gas contributing to 20 amu increases
again while the levels for 18 and 19 amu remain constant . From this it could be concluded
that above 1000C the signal is due to the release of HF, while below 1000C water is the
evolved species. In order to be able to distinguish between the emissions of H2O and HF a
TG-FTIR measurement was conducted. The TG and DTG curves are presented in fig. 2. A
single FTIR spectrum was extracted at 980C in order to prove emissions of HF (fig. 3). By
subtraction of the spectrum of water (4000 3500 cm-1 and 1900 1400 cm-1) the evolution
of HF can be clearly detected as a comparison with the absorption bands of pure HF (fig. 4)
proves.
67
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Fig. 1: STA-MS measurement of a brick clay, no clear conclusion whether HF (20amu)


emissions occur due to contribution of H2O to the signals at 18, 19 and 20 amu

Fig. 2: TG and DTG curves of the TG-FTIR measurement in the temperature range
30 1000C.

From the FTIR measurement, the evolution of HF and H2O can be shown as function of
time/temperature and correlated to thermal decomposition behavior by setting spectral
windows. For this so-called trace calculation the characteristic bands of HF between 4200
4100 cm-1, which are not overlapped by the bands of water, and these of H2O between 1750
1700 cm-1 were chosen. These traces clearly prove that the emissions of water occur
between 50 and 800C with major peaks at 140C and 570C. The evolution of HF starts at
780C and increases strongly up to 1000C.

68
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Fig. 3: Measured FTIR Spectrum at 980C after subtraction of the H20 spectrum, the HF
spectrum is clearly visible (compare Fig. 4)

Fig. 4: FTIR spectrum of HF, measured at room temperature.


69
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Fig. 5: FTIR traces of HF (blue) and H2O (red) during firing of a brick clay. The release of
HF starts at 780C (top), the evolution of H2O is detected with peaks at 140C and
570C (bottom)

70
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Conclusion:

For the determination of acid gas species, e.g. HF, during the firing process of traditional
ceramics such as brick clays, the method of choice is the TG-FTIR coupling. The separation
of the processes of HF and H2O evolution is clearly possible by FTIR, but nearly impossible
by the TG-MS technique due to overlapping characteristic m/z signals for these two species.

Proper selection of the spectral windows enables precise investigation of the temperature of
the beginning of the HF evolution and its complete emission profile.

The correct identification of air pollutants and detection of their release temperature is of criti-
cal importance. Many environmental toxicology assessment models require an input of the
evolved gases. The key questions of the emerging problems are:
what gas species are present in the exhaust gases and
how do process parameters such as temperature, residence time, and atmosphere affect
the type and quantity of toxic gases present in the exhaust.
These questions can be answered with the help of the hyphenated techniques TG-FTIR and
TG-MS. Correct information will prevent the ceramic industry from unnecessary and un-
realistic regulations [3].

Literature:

[1] H. Kochmeier, Emissions, Ziegelindustrie Int., 38 (1991) 544 548


[2] H. Kochmeier, Emissions and Control in the Brick and Tile Industry, Ziegelindustrie Int.,
33 (1986) 516 530
[3] D. Brosnan, Use of Thermal Analysis in Identification of Air Pollutants in Manufacturing
of Traditional Ceramic Products, Proceedings of the 3rd SKT 2000
Hyphenated Techniques in Thermal Analysis (2000) 97 - 110

71
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Characterization of Raw Cement and Cement Mixtures

Ekkehard Post
NETZSCH-Gertebau GmbH, Wittelsbacherstr. 42, 95100 Selb

Application: Building industry

Goal of Analysis: Compositional analysis, identification of volatiles

Instrument: STA 449-MS capillary coupling


(Balzers ThermoStar)

Sample: Cement raw meal, mixtures with additives

Experimental conditions:
Sample mass: 45 mg
Temperature programme: RT 1450C
Heating rate: 10 K/min
Crucibles: Pt/Rh, lids with hole ca. 0.5 mm
Sample holder: TG-DSC type S
Atmosphere: synthetic air, 50 ml/min
Mass spectrometer: SCAN Bargraph

DTG /(%/min)
TG /% Ion Current *10-12 /A
[2]
100 -2.04 % 1.0
6.0
95 0
5.0
90
-1.0 4.0
486.7 C
85
-33.61 % 1412.4 C -2.0 3.0
125.5 C 343.9 C H2 O
80
2.0
440.0 C -3.0
75
1.0
70
SO2 -4.0
CO2 0
-0.98 %
65 -5.0 -1.0
200 400 600 800 1000 1200 1400
Temperature /C

Fig. 1: TG-MS results of a raw cement mixture

72
NETZSCH Coupling of Thermal Analysis and Gas Analysis

DSC /(mW/mg)
TG /% DTG /(% /min)
-1.42 % 853.7 C = ex
100 -0.56 % 3.0
2.0
2.0
95
1.0
1.5
90 0
146.3 C 451.1 C -25.74 % -1.0
85 1.0
-2.0
80
453.5 C -44.0 J/g! 0.5 -3.0
856.0 C
149.1 C 575.5 C
75 -4.0
-0.52 %
9.1 J/g! 17.6 J/g! 0.8 J/g! 946.3 J/g! 1337.3 C
0 -5.0
200 400 600 800 1000 1200 1400
Temperature /C

Fig. 2: TG and DSC analysis of a mortar mixture (cement, sand)

TG /% Ion Current *10-9 /A


-1.42 % [3]
100 -0.56 % 854.0 C 6.0

5.0
95
-25.74 %
4.0
90 451.7 C
CO2 3.0
147.3 C 1326.6 C
85
SO2 2.0

H2O -4. - 9 x 10-12 1.0


80

0
75
0 - 3 x 10-9 -0.52 %
-1.0
200 400 600 800 1000 1200 1400
Temperature /C

Fig. 3: Water, carbon dioxide and sulphur dioxide evolution from the mortar sample
(compare Fig. 2)

73
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Results:

A variety of different mineral binders, such as gypsum, lime and cement and as well polymer-
based binders and pure organic adhesives are used in the building industry. There is a wide
range of different cements (DIN 1164), although Portland cement is the one which is most
frequently used. Portland cement is made of calcareous materials and clays, which are
ground and then heated in a rotary kiln to yield so-called clinkers. After the firing process the
clinker must be ground again. It contains as typical main phases tri-calcium silicate (C3S), tri-
calcium aluminate (C3A), etc.

Figure 1 depicts the TG, DTG and MS curves of a raw mixture for Portland cement. After
several dehydration steps (more detailed in figure 2), decomposition of the carbonate content
starts at approximately 600C. This can be concluded from the start of decomposition shown
by the TG curve and the increase of the CO2 intensity (MS signal 44 amu).
The carbonate decomposition follows the reaction:
CaCO3 CaO solid + CO2 gas
Pure CaCO3 shows a weight loss of 43.97 % for the CO2 evolution. The raw cement mixture
shows a weight loss of 33.61 %. Dividing by the theoretical value of 43.97 % for pure CaCO3
it indicates a content of 76.3 % CaCO3 in the raw cement mixture.
It seems that this material contains also small amounts of sulphate (gypsum or anhydrite
(CaSO4) and other sulphates). This is shown in the curve of mass number 64 by peaks at
440 C and 1412C, which indicate SO2 evolution.

In the figures 2 and 3 the TG, DTG, DSC and MS results of a cement mortar is shown. After
the dehydration and dehydroxylation steps (see fig. 3, evolution of water) also the carbonate
decomposition starts. The CaCO3 content is clearly lower than that of the previous sample
(approx 58.5 % CaCO3). The endothermal DSC peak at 571C indicates a small amount of
crystalline quartz. According to the enthalpy value of 8.62 J/g taken from literature [1] for the
phase transition of pure quartz a crystalline quartz amount of 10.8 % can be calculated. It
seems that this mixture contains also small amounts of sulphates (gypsum, anhydrate), since
at approximately 1350C a small TG step of 0.52 % weight loss was detected with
SO2 evolution (figure 3).

Conclusion:

The decomposition of complex cement mixtures can be studied with Simultaneous Thermal
Analysis quite well. However, for the complete characterization of the decomposition
behavior very often the information about the evolved gases is additionally required. By a
mass spectrometer coupling (or FTIR coupling) the released gases can be identified and also
in case of overlapping mass loss steps the individual gases can be separated. In this case
the extension of the STA-MS with PULSETA would be also very useful in order to quantify
the evolved gases.

The measurements performed clearly show the capability of the STA-MS coupling for the
analysis of cement mixtures for quality control, research and development.
The firing process can be studied under the aspect of clinker formation, quartz and carbonate
content, etc. as well as under environmental aspects, which is only possible with the
additional information from evolved gas analysis.

Literature:

[1] F. Gronvold, S. Stolen, S. R. Svendsen, Thermochim. Acta, 139 (1989) 225-243


74
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Identification of Metal Vapors and Metal Salts


Cd, PbCl2

Ekkehard Post
NETZSCH-Gertebau GmbH, Wittelsbacherstr. 42, 95100 Selb

Application: Detection of heavy metals

Goal of Analysis: Detection of metal vapors, investigation of


dissociation and fragmentation of metal salts

Instrument: STA 409 C Skimmer coupling

Sample: Cadmium metal, lead dichloride (powder)

Experimental conditions:
Sample mass: 4 mg Cd
4.3 mg PbCl2
Temperature program: RT 800 C
Crucibles: Al2O3, open
Sample holder: TG-DSC
TG
Atmosphere: He, 75 ml/min
Mass spectrometer: Scan mode
MID mode
recipient pressure: 4 x 10-6 mbar

75
NETZSCH Coupling of Thermal Analysis and Gas Analysis

20 3

-20 2

Cd 110 12.39%
-40 Cd 112 24.07%
mass change / %

-9
Intensity / A*10
-60 1

112
-80
-5
10.1 * 10 A*s/g
TG
-100 0
110
-120
5.36 * 10-5 A*s/g

-140 -1
300 350 400 450 500 550 600 650 700

Temperature / C

Fig. 1: Detection of Cd isotopes at 110 and 112 amu

Fig. 2: TG curve for calcination of PbCl2 and MS scan at 584C showing its fragmentation

76
NETZSCH Coupling of Thermal Analysis and Gas Analysis

m/z Natural abundance [%]


Pb 208 52.3
206 23.6
207 22.6
Cl 35 75.5
37 24.5
PbCl 243 (208 + 35) 39.5
243 (206 + 37) 5.8
242 (207 + 35) 17.8
241 (206 + 35) 17.0
245 (208 + 37) 12.8
244 (207 + 37) 5.5
PbCl2 278 (208 + 35 + 35) 30.0
280 (208 + 35 + 37) 9.7
280 (208 + 37 + 35) 9.7
etc....

Table 1: List of dominant fragments of PbCl2


mass change / %

30
PbCl 2
60 278
243
208
90
TG

120
0 100 200 300 400 500 600 700 800

Temperature / C

Fig. 3: TG curve and MID signals for Pb, PbCl and PbCl2 during the evaporation of PbCl2

77
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Results:

Fig. 1 shows the complete evaporation of a cadmium metal. The intensities for two of the
eight Cd isotopes are followed by MID curves versus temperature. Starting at 350 C one
can observe the beginning of the mass loss due to the evaporation of cadmium. The natural
abundance of the cadmium isotopes at 110 and 112 amu is 12.39 % and 24.07 % of all the
Cd isotopes. Integration of the peak areas in MID curves allows comparing the amount of
evolved gases or vapors and is also a basis for calibration of MS signals and therefore for
precise quantitative evaluations.

During the evaporation of metal salts both the molecule ion and fragment ions are detected.
This is due to thermal dissociation in the thermobalance and transfer system and due to the
electron impact ionization in the mass spectrometer.
During evaporation of PbCl2, not only the complete molecule PbCl2, but also PbCl, Pb and
Cl2 (not shown) are detected (fig. 2). The low intensity of the molecule ion compared with the
fragment ions shows the effect of thermal dissociation and ionization as discussed above.
Additional peaks in the MS scan are registered for the rhenium cathode material in the ion
source of the mass spectrometer (Re 185, Re 187).
The fragmentation pattern partly shown in fig. 2 is due to the combinations of the Pb isotopes
(204, 206, 207, 208 amu) and the Cl isotopes (35 and 37 amu). Table 1 shows some of the
combination possibilities. In fig. 3 the intensities of the fragments 208 (Pb), 243 (PbCl) and
278 (PbCl2) are shown by the corresponding MID curves together with the TG curve.

Conclusion:

Detection of evaporating metals or metal salts by TG-MS is usually only possible with the
Netzsch Skimmer coupling. The Skimmer coupling parts are located in the hot zone of the
STA furnace and have approximately the same temperature as the sample and the evolved
metal vapor, respectively. This means condensation of the evolved gases and vapors can
practically be avoided and the detection of evaporating metals and metal salts becomes
possible.

78
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Application of Thermal Analysis to Material Science -


Case Study on Hard Metals
Gert Leitner, Klaus Jaenicke-Rler, Tim Gestrich
Fraunhofer-Institut fr Keramische Technologien und Sinterwerkstoffe
IKTS Dresden
Winterbergstrasse 28, 01277 Dresden

Abstract

Thermoanalytical methods are used for investigation of outgassing and sintering of products
made from hard metals. Shrinkage (DILA), mass loss (TG), gas evolution (EGA - mass
spectrometry) and thermal effects (DSC) allow to describe sintering processes. The results
may be applied for a better understanding of technological procedures, e.g. for
improvements of temperature-time-atmosphere cycles in the production scale.

1. Introduction

In this paper topical results of the application of thermoanalytical methods for the in-
vestigation of outgassing and sintering of hard metals are demonstrated. For a better
understanding information on hard metals is presented first. The simulation of technological
processes such as sintering requires TA methods which can be used at high temperatures
(most of the apparatus at the IKTS can reach 2000C, a temperature far above that of the
usual working conditions of most of the other TA labs).
At the IKTS, research work on hard metals by means of TA methods is focused on four
special topics (Fig. 1):

Fig. 1: Materials characterization by TA methods

79
NETZSCH Coupling of Thermal Analysis and Gas Analysis

y Technological questions (solid-state sintering - liquid-phase sintering, effect of


impurities, role of the carbon balance)
y Materials development (effect of doping elements, development of fine-grained
hard metals)
y Theoretical aspects (basic mechanisms)
y Data bases (thermophysical properties).

A few characteristic results are presented in this paper. More detailed discussions are given
in references [1] to [4].

2. Hard metals - an old and a new material

Hard metals have been used for a long time. The first hard metals were introduced on the
market in 1924 by Krupp-Widia. Since then, the materials and the technology have been
continually improving. Nowadays much effort is still being put into improving this old and new
materials.
Hard metals are used

y for cutting, milling, and drilling;


y as wear parts; and
y for many other applications.

Hard-metal products are produced in a powder metallurgical process, i.e. the production
route (Fig. 2) begins with powder which is compacted by pressing. The pressed sample with
a porosity of 40 to 50 vol.% must be densified by sintering to full density with zero porosity.
Normally organic pressing aids like paraffin or others are used in order to improve the
compaction. These organic compounds must be burned out before closing the pores during
sintering.

Fig. 2: Scheme of production cycle of WC-Co hard metals

80
NETZSCH Coupling of Thermal Analysis and Gas Analysis

The simplest type of hard metals consist of

y WC as the hard material and


y Co as the ductile binder.

The phase diagram (Fig. 3) shows that during heating of the powder mixture there is a
solution of WC in the Co phase. At about 1300C a eutectic is formed. In the literature the
sintering of hard metals is described as being a prototype for liquid-phase sintering.
However, we often find effects during sintering which are far away from the equilibrium states
shown in the phase diagrams. There is a considerable interplay between thermodynamics
and kinetics.

Modern developments in hard metal research tend towards

y new materials,
y new technologies,
y improved properties, and
y reduced costs.

What can be done with TA in the field of hard metal research?

Fig. 3: Phase diagram of WC-Co [5]

In Fig. 4 the basic information found from the application of dilatometry DILA (length change),
thermogravimetry TG (mass change), differential scanning calorimetry DSC (caloric effects),
and evolved gas analysis EGA (gas evolution) is shown for the investigation of outgassing
and sintering of hard metals.

81
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Fig. 4: Thermal analysis of hard metals

3. Thermal behavior of hard metals

3.1. Technological approach

First we look for shrinkage and liquid phase formation (Fig. 5). The material is a normal-
grained WC-Co hard metal with 10 wt% Co and WC with a mean grain size of 1.3 m
(WC DS130 from H.C. Starck / Germany). All investigations were carried out under argon
flow and with a heating rate of 10 K/min.
The black curve is the DSC curve. It shows the formation of the liquid phase - an
endothermic effect - at about 1350C. If we look at the gray curve - the shrinkage curve - we
see that the greater part of the shrinkage takes place in the solid state before melting of the
eutectic. The shrinkage rate curve (dotted curve) indicates the change of the sintering
mechanism through the liquid phase formation in a close relation to the start of the
endothermic effect in the DSC curve.
Let us look at the outgassing behavior at lower temperatures (Fig. 6). The TG curve shows
the mass loss during heating of the pressed sample to 1000C. For better interpretation of
the effects this sample was pressed without any organics. We find several steps in the mass
loss curve which can be correlated with mass spectrometric curves of the evolved gas
analysis. The evolution of CO and CO2 is shown in the curves of ion current intensities for
mass numbers m/z 28 and m/z 44. The effects at about 350C and 520C result from the
reduction of oxide impurities of the cobalt powder by the carbon in the mixture. Also the
larger effect at about 750C results from the reduction of the oxygen containing phases of the
hard material WC. These phases are formed during the technological pretreatment of the
mixture, especially during milling of the components.

82
NETZSCH Coupling of Thermal Analysis and Gas Analysis

For technical purposes it is important to realize that the carbon balance is responsible for the
quality of the final hard metal. Technologists have to take care to control the carbon balance
in the production cycle.
Fig. 7 shows an overview of the results of a complex thermal analysis of this material. For
comparison the ordinate is given in arbitrary units. Shrinkage, thermal effects, mass loss and
gas formation show characteristic effects which may be better understood by synergetic
interpretation of all curves.

Fig. 5: Shrinkage and liquid phase formation, measured by dilatometry


and DSC

Fig. 6: Mass loss and evolution of CO and CO2

83
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Fig. 7: Outgassing and sintering of WC-10 wt% Co

3.2. Materials development

Fine-grained hard metals are produced by doping the mixture with additional carbides such
as vanadium carbide (VC) or chromium carbide (Cr3C2), which prevent grain growth during
sintering. Fig. 8 shows the influence of doping on the liquid phase formation. Small contents
of less than 1 wt% change the eutectic temperatures and the widths of the temperature
range of liquid phase formation.
The correlation of these results with the corresponding influence on shrinkage and shrinkage
rate - which is not discussed here - gives very important information about the efficiency of
the doping procedure.

84
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Fig. 8: Influence of doping on liquid phase formation

3.3. Basic mechanisms

For some reasons, especially for modeling of sintering, one must know the temperature at
which sintering begins. At first we look at the shrinkage experiment. Enlarging the scale and
correcting for the thermal expansion, the shrinkage curve shows a relatively sharp bend
above 700C (Fig. 9), characterizing the beginning of sintering.

Fig. 9: Beginning of sintering (thermal diffusivity, shrinkage)

85
NETZSCH Coupling of Thermal Analysis and Gas Analysis

But there is another possibility for identifying the change in the contact area between the
particles. A very sensitive method for the investigation is the determination of the thermal
diffusivity measured by the laser flash technique [6]. If the contact area between the particles
is enlarged by diffusion-controlled processes we observe an increase in the thermal
diffusivity. The measuring curve shows a real increase in the contact area. At first particle
approach takes place. At higher temperatures particle rearrangement is relevant.
A theoretical description is given in reference [3].
Shrinkage and increase in thermal diffusivity begin at the same temperature range, between
700 and 750C. Both effects are based on the same mechanisms. The beginning of
shrinkage is made possible by the fact that at this temperature particle surfaces are cleaned
by reduction processes resulting in evolution of CO and CO2. On these activated and
reactive surfaces diffusion processes are now enhanced and lead to mass transport, particle
approach, shrinkage, reduction of porosity, and finally to sintering.

3.4. Data bases

If we consider the application of hard metals for cutting, we know that the temperature at the
edge of a cutting tip reaches values of about 800C to 900C (Fig. 10).
For describing this process it is necessary to know the heat conductivity of such materials not
only at room temperature but also at the temperature of the application.
A normal hard metal (discussed previously) shows a heat conductivity of about 80 W/mK at
room temperature; the value at the edge of the cutting tip is reduced during cutting by 40 %
(Fig. 11). The tool designer must consider this change in order to optimize tool geometry and
cutting conditions.

Fig. 10: Application of hard metals cutting

86
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Fig. 11: Thermal properties of hard metals and cermets


(ceramic-metal-composites)

4. Up-scaling and outlook

Many valuable results can be achieved using methods of thermal analysis for
characterization of materials and technological procedures. Our main aim is the description
of outgassing and sintering of high-temperature materials. In our case, in-situ sintering
means the simulation of sintering using methods of thermal analysis. However, samples in
thermal analysis are very small (on the order of 100 mg or smaller). If we want to transfer the
results to a production scale we have to develop methods of up-scaling (which are also
extremely necessary for successful partnership with industrial colleagues).
An example of up-scaling experiments is given in Fig. 12. In a macro thermobalance (sample
weight < 450 g) the mass loss by outgassing is shown for a ceramic tile. The results can be
used in the development of improved temperature-time-atmosphere cycles on a production
scale.

87
NETZSCH Coupling of Thermal Analysis and Gas Analysis

In Fig. 13 important future fields of application of thermoanalytical investigations are


summarized. Development of new materials and improvements in production technology are
the main aims of TA simulation. Results for hard metals given in this paper seem to be a
suitable example for demonstrating the potential of TA methods for solving materials
problems.

0 0

Relative mass change rate d ( m/m0 )/dT in % / min


-1 -0.05
Relative mass change ( m/m 0 ) in %

-0.1
-2
-0.15
-3
-0.2
-4
-0.25
-5
-0.3
-6
-0.35
-7
-0.4
-8
-0.45
-9
0 250 500 750 1000
Temperature in C
Sample weight: 21 g
Atmosphere: Flowing air
5 K/min 15 K/min
TG TG
DTG DTG

Fig. 12: Up-scaling with a macro thermobalance

Fig. 13: Outlook - Application of TA

88
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Literature:

[1] G. Leitner, K. Jaenicke-Rler, T. Gestrich, T. Breuning, "Thermal analysis aids Sintering


of hard metals", Metal Powder Report 52 (1997) 32-37
[2] G. Leitner, K. Jaenicke-Rler, T. Gestrich, T. Breuning, "Entbindern und Sintern von
Keramik - Optimierung durch thermoanalytische Simulation", cfi/Ber. DKG 76 (1999),
in press.
[3] G. Gille, G. Leitner, W. Hermel, "Schwindungsverhalten von WC-Co Hartmetallen",
Z. Metallkd. 89 (1998) 73-76
[4] G. Leitner, G. Gille, T. Gestrich, "Shrinkage, liquid phase formation and gaseous
reactions during sintering of WC-Co hard metals and correlation to the WC grain size", in:
Proc. 14th Plansee Seminar, ed. by G. Kneringer, P. Rdhammer, P. Wilhartitz,
Metallwerk Plansee, Reutte, Vol. 2 (1997) 86-99
[5] W. Schedler, "Hartmetall - Aufbau, Herstellung, Eigenschaften und industrielle
Anwendung einer modernen Werkstoffgruppe", VDI-Verlag, Dsseldorf (1988) 125
[6] M. Perl, G. Leitner, "Temperatur- und Wrmeleitfhigkeit im Hochtemperaturbereich",
J. Thermal Anal., 47 (1996) 643-650

89
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Ternary Semiconductors CuGaSe2, ZnGeP2


-impurities, decomposition-

Ekkehard Post
NETZSCH-Gertebau GmbH, Wittelsbacherstr. 42, 95100 Selb

Application: Characterization of semiconductors

Goal of Analysis: Detection of impurities, investigation of thermal


stability during melting

Instrument: STA 409 CSkimmer coupling

Sample: CuGaSe2, ZnGeP2

Experimental conditions:
Sample mass: 300 mg CuGaSe2
3 mg ZnGeP2
Temperature program: 251050 C, 10 K/min
Crucibles: Al2O3, open
Sample holder: TG-DTA
Atmosphere: He, 75 ml/min
Mass spectrometer: Scan mode
recipient pressure: 4 x 10-6 mbar

General Comments:

The most important and well-known semiconductor is silicon (Si). This material is used for
the production of diodes, transistors or microprocessors. For special applications (i.e. IR
detectors, extremely fast microchips) the so-called II-VI or III-V semiconductors were em-
ployed. The crystal structure of these semiconductors can be related to the silicon structure.
Other semiconductors belong to the family of the I-III-VI2 or II-IV-V2 compounds with the
chalcopyrite structure. These semiconductors have a direct band gap, contrary to silicon with
an indirect band gap (indirect band gap limits the velocity of the microchips). By proper
selection of the elements the band gap can be optimized for the required application. Of
course there are also some disadvantages mainly effected by the physico-chemical
properties and the factor stoichiometry, which makes the crystal growth and/or the
manufacturing of epitaxial layers relatively complicated compared with the crystal growth of
silicon.
For the study of the physico-chemical properties (e.g. melting point, evaporation, character-
ization of gas species during heating, determination of impurities or temperature of in-
congruent melting) the TG-DSC(DTA)-MS analysis is a very helpful tool and can deliver a lot
of important information. The application of TA will be illustrated in the investigation of
CuGaSe2 and ZnGeP2, both examples of ternary semiconductors of the chalcopyrite type.
90
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Fig.1: TG-MS analysis of CuGaSe2, detection of impurities

Results:

The CuGaSe2 material was synthesized from the elements Cu, Ga, Se taken in stoichio-
metric amounts. The STA-MS analysis however shows around 450 C selenium evaporation,
detected in the MS scan through the isotope distribution for Se3 between 230 and 245 amu
(fig. 1). This demonstrates that there was no pure CuGaSe2 synthesized. Also iodine is
detected during this TG step. This means either one or more of the elements were not pure
or iodine was used as a so-called mineralizer for the synthesis.
The further occurrence of selenium at temperatures higher than 900 C is due to the thermal
degradation of CuGaSe2. This is, of course, also important for the crystal growth. This means
regulation of the Se vapor pressure is required to control the stoichiometry.

Very important is also the control of the vapor pressure during the crystal growth of ZnGeP2.
During heating of this semiconductor zinc and phosphorous will evaporate (figures 2 and 3).
As the vapor pressure of germanium is very low, only Ge remained as residue from ZnGeP2
in the crucible after the measurement. Producing of epitaxial thin films might therefore be
very difficult by PVD (Physical Vapor Decomposition). In this case CVT (Chemical Vapor
Transport) or MOCVD (Metal Organic Chemical Vapor Decomposition) are the more
successful methods.

91
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Fig. 2: Evolution of phosphorus (multiples of P=31 amu) during thermal decomposition of


ZnGeP2

Temperature / C

Fig. 3: Detection of Zn isotopes (64, 66, 68 amu) during thermal degradation of ZnGeP2

92
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Conclusion:

For the understanding of crystal growth and the production of high-quality semiconductor
crystals the knowledge of the phase diagrams is extremely important. TG-MS investigations
deliver important information about thermal stability of the materials, especially sublimation,
evaporation, decomposition and impurities. With calorimetric techniques, like Differential
Scanning Calorimetry (DSC), the melting behavior can be analyzed and decisions about the
crystal growth method can be taken. Furthermore the sub-solidus regions can be studied
for modification changes.
A very important aspect for the crystal growth of compound semiconductors is the control of
the atmosphere to prevent evaporation. This can only be examined by monitoring of the
evolved gases. However, very often the released gas molecules may show condensation in
the heated interface of a commercial TG-MS capillary coupling. For these cases only the
Skimmer coupling is useful because of the special arrangement of the coupling parts
directly in the hot furnace, where they have the temperature of the sample under test.

Literature:

E. Post, E. Kaisersberger, S. Knappe, W.-D. Emmerich, J. Therm. Anal. Cal., 57 (1999)


265-273

93
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Application of Pulse Thermal Analysis (PULSETA)


for Characterization of Materials
Marek Maciejewski, Alfons Baiker
Laboratory of Technical Chemistry, ETH Zurich, 8092 Zrich

Thermal analysis (TA) is generally used for characterizing gas-solid reactions or the
decompositions of solids in the range from zero to full conversion, i.e. the entire process is
monitored. In conventional TA, the experiments are carried out isothermally or with a linear
temperature change. In isothermal experiments, some time is necessary for the system to
reach the desired temperature. During this heating period the reaction may start, rendering a
proper investigation of the initial transformations almost impossible. For the same reasons, it
is impossible to quench the reaction at a desired time (or at a desired extent of the reaction).
Also, the most frequently used non-isothermal mode of thermal analysis does not allow
controlled decomposition of the sample to a certain degree or quenching reaction at a
specified time.

The novel pulse thermal analysis PULSETA method [1] eliminates, or at least reduces, the
difficulties mentioned above. The PULSETA method is based on the injection of a specific
amount of the gases or liquids into the inert carrier gas stream and monitoring of changes in
the mass, enthalpy and gas composition, resulting from the incremental reaction extent.
Because a known amount of the selected gas, which can be used for calibration, is injected
into the system, the method is also suitable for quantification of the evolved gas by MS or
FTIR. In contrast to conventional TA and all its modifications, the reaction is controlled not
only by the temperature, but also by a distinct change in the composition of the reactive
atmosphere.
PULSETA offers three principle opportunities of thermoanalytical studies, depending on the
kind of injected gas:

(i) injection of a gas which does not react with the sample. This procedure
facilitates quantitative calibration of the mass spectrometric signals and
increases the sensitivity of TA measurements so that species in amounts
lower than 0.01 wt% can be detected.
(ii) injection of a gas which reacts with the solid sample. This mode provides
the opportunity investigating all types of gas-solid reactions.
(iii) injection of a gas which adsorbs onto the sample surface. This mode
offers the study of adsorption phenomena under atmospheric pressure and at
specified temperature.

The most distinct feature of PULSETA is that changes of the gaseous atmosphere and the
solid sample occur during a well-defined period of time after each injected pulse of the gas
or liquid. The transient character of the pulse technique offers interesting opportunities for
PULSETA applications, which will be illustrated using examples from various fields such as
catalysis and material science.

94
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Experimental

Experiments were carried out isothermally or non-isothermally (generally heating rates were
in the range of 5-10 K/min) on a Netzsch STA 409 simultaneous thermal analyzer equipped
with a gas pulse device (PULSETA box, Netzsch), enabling injection of a certain amount of
two different pure gases or gaseous mixtures into the system. The amount of injected gas
could be varied from 0.01 to 10 cm3. Primarily, volumes of 0.25, 0.5 and 1.0 cm3 were used.
Gases evolved during reaction and/or injected into the system were monitored on-line with a
Balzers QMG 420 quadrupole mass spectrometer, connected to the thermal analyzer by a
heated (ca. 200oC) capillary.

Results:

(i) Injection of a gas which does not react with the sample.

The opportunity of injecting a known amount of gas or liquid into the carrier gas stream
provides a quantitative calibration by relating the MS-signal to the injected quantity of probe
gas. More information concerning quantification of the spectroscopic signals recorded with
TA-MS and TA-FTIR systems are given in another contribution contained in this volume.
Previously published results [2,3] indicate that the kind of carrier gas and temperature do
only affect the shape of the MS and FTIR signals but not their integral intensities.

The application of the spectroscopic signals for quantitative analysis with combined TA-MS
and TA-FTIR techniques is especially useful in the case of multicomponent systems, where
two or more processes can overlap due to the occurrence of simultaneous reactions.
A typical example is shown in Fig. 1, which depicts the calcination of calcium carbonate and
-alumina mixtures, possessing well-known compositions. Due to the continuous mass
change resulting from the evolution of water from alumina, the exact determination of the
mass loss due to CaCO3 decomposition is uncertain. This leads to an imprecise
determination of the amount of CaCO3 in the mixture. The onsets of the TG curve taken for
the determination of the evolution of CO2 have to be chosen arbitrarily, what leads to
substantial errors in the calculation of the real composition of the analyzed samples.
The smaller the content of CaCO3 the greater is the error of thermogravimetric analysis.
For the sample containing only 0.21 wt% carbonate the difficulties in assessing the
beginning and the end of the CO2 evolution cause a gross overestimation of the calcium
carbonate content. The results obtained by quantification of the MS signals are much more
accurate.

95
NETZSCH Coupling of Thermal Analysis and Gas Analysis

0 TG 22.71 wt% TG 2.18 wt%


0

-4
MS 23.10% MS 2.17% -2

mass change / wt%


TG 23.81% TG 3.64% m
-8
mass change / wt%

m/z = 44 -4
-12 m/z = 44

1
TG 0.46 wt% TG 0.21 wt%
0 0

-1
MS 0.44% MS 0.23%
-2 TG 1.55% TG 1.32% -2

-3
m/z = 44 m/z = 44
-4 -4
0 200 400 600 800 1000 0 200 400 600 800 1000
temperature / C

Fig. 1: Determination of the CaCO3 content in the mixture with -Al2O3 by TG and MS.
The mass spectrometric signals of evolved CO2 (m/z=44) were calibrated by
few pulses of CO2 injected before and/or after carbonate decomposition. The
real composition of the samples and results of TG and MS analysis are given in
yellow and blue rectangles, respectively.

The quantitative interpretation of the TG curve is even more uncertain when the temperature
ranges of the various decomposition stages overlap so that the specific stages of the
decomposition are not discernible. Also, when the total mass change is below 1-2 wt%, its
quantitative interpretation is difficult due to the buoyancy effect. Fig. 2 depicts the
quantitative determination of the carbon containing species in supported gold catalysts
Au/ZrO2 and Au/TiO2 [4]. During calcination in 20 vol% O2 the water and carbon dioxide
evolved simultaneously and the observed mass loss was very small and amounted to
1.2 and 1.4 wt%. The quantification of the mass spectrometric signal of evolved CO2 by
injection of 1 cm3 calibration pulses allowed determination of the organic residue in the
catalysts. The carbon content in Au/ZrO2 and Au/TiO2 samples was 0.04 and 0.05 wt%,
respectively.

96
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Au/ZrO P
2
P
Au/TiO
2
309
215
426 CO
2 312
1.5
252
mass change / wt%

CO

mass change / wt%


303 2
1.0 78 1.0
HO 98
2
0.5 0.5 HO
2
0.0
0.0 -0.3%
-0.5
-1.0 -1.4 % TG -0.5 TG
-1.5
-1.0 -1.2%
-2.0
100 200 300 400 500 600 700 800
A temperature / C 0 100 200 300 400 500 600
B temperature / C

Fig. 2: Calcination of Au/ZrO2 (A) and Au/TiO2 (B) catalysts in 20 vol%O2/He. The 1 cm3
pulses of CO2 (P) were injected to calibrate MS signals of m/z=44 (CO2).

The PULSETA technique provides the only in situ method of quantitative analysis of TA
signals in coupled MS or FTIR systems when analyzing more complicated systems,
containing several components. Figure 3 presents the results of the determination of the
sodium bicarbonate content in a commercial sweetener (Sucramid).

P 163 467
int. m/z=18,44,64 / a.u.

P
CO 164
2

HO 305 461
2
SO
2
DTG
461
0 TG 163 60
-10
mass change / %

-8.9% 40
-20
282
T / V

-30 20
-40 DSC
0
-50 -57.6%
-60 164 283 -20

0 200 400 600


temperature / C

Fig. 3: Thermal decomposition of a commercial sweetener (Sucramid). Amount of


NaHCO3 was determined by quantifying CO2 signal centred at 163C.
97
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Sucramid contains two active components, namely cyclamat (sodium salt of


cyclohexylsulfamic acid) and saccharin (sodium salt of 1,1-dioxo-1,2-dihydro-1-
benzoisothiazol-3-one). As prepared in the form of effervescent tablets, it contains also the
tartaric acid, which during dissolution of the tablet, reacts with the sodium bicarbonate,
producing gas bubbles (CO2). The main gaseous products of Sucramid decomposition in air
are water and CO2, which are formed during the oxidation of the organic components. The
additional gaseous product found was sulfur dioxide (m/z=64). The low-temperature
evolution of water and CO2, centred at 163C, is due to the decomposition of NaHCO3.The
quantification of the m/z=44 signal in the range 125-220C allowed determining the NaHCO3
content in the tablet, which was 24.8 wt%.

During decomposition of Sucramid the sulfur content has not been analyzed, but it can be
simply done by injecting SO2 into the carrier gas stream and comparing the intensities of
m/z=64 or 48 of the injected pulse to intensities resulting from the decomposition. The
example of the determination of the sulfur and carbon content relevant for petrol
searching [5] is presented in Fig. 4.
int. of m/z=18, 44, 64 / a.u.

406 452
SO
2

P P
372 469
CO
2

temperature difference / V
118 349 HO
2
353
0 TG 451 40
-4
20
mass change / %

-8 DSC
-12 0

-16 -18.2% -20


-20
0 200 400 600 800
temperature / C

Fig. 4: Determination of the sulfur and carbon content in the sample collected during
petrol searching by its calcination in air followed by the quantification of CO2
and SO2 mass spectrometric signals.

The rock samples collected during drilling for petroleum contain, among others, two
important components, being decisive for the characterization of the quality of the
investigated materials: the organic matter and sulfur. Due to the large variety of compounds
containing carbon and sulfur in the rock, the exact quantitative determination of these
elements is difficult and time consuming. The use of an oxidative atmosphere caused the
oxidation of the carbon- and sulfur- containing phases to CO2 and SO2, respectively. The
evolution of carbon dioxide (m/z = 44) and sulfur dioxide (m/z = 64) was observed in the
range of ca. 250-650C. The calcination process was highly exothermic as evidenced by the
DSC signal. This exothermic reaction of sulfur and carbon containing species with oxygen
was overlapped by the endothermic dehydration, therefore the conventional interpretation of
TA and MS signals could be used only for the qualitative characterization of the samples.

98
NETZSCH Coupling of Thermal Analysis and Gas Analysis

The quantification of MS signals was done by injecting 1 cm3 of CO2 and 1 cm3 of a mixture
containing 20 vol% of SO2 in He. The carbon and sulfur content were 2.5 and 3.7 wt%,
respectively.

The possibility of exact calibration of the MS signal by means of PULSETA significantly


increases the potential of the coupled TA-MS and TA-FTIR techniques. Using this method to
determine the content of certain species in the investigated system has considerable
advantages over conventional elemental analysis measurements. With conventional
analysis, only the total amount of analyzed species can be measured, and it is impossible to
resolve the species evolved in multistage reactions.
This benefit of the PULSETA is illustrated by the results of the quantification of the evolved
gases during TA-MS measurement of a cuprate superconductor of the Y2Ba4Cu6+nO14+n
family [6]. These superconductors contain often a remarkable quantity of CO32-ions, which
can strongly influence their properties. In the studies of the carbon dioxide incorporation
process a cardinal question is whether it is possible to remove CO2 from the superconductor
without its decomposition. Also the knowledge of the substitution of certain species in the
crystallographic lattice of the superconductor by CO32- can help in the interpretation of
YBaCuO properties in relation to preparation method.

706 921
int. m/z=32,44 / a.u.

CO pulse 830
2
m/z = 32

A 861
m/z = 44
DTA
1 0
TG
mass change / %

0
-1 A 705 -20
T / V

-2 m/z = 44 -40
-3 1.93
931 -60
-4
-5 0.36 % 0.005% -80
5.34
-6
200 400 600 800 1000
temperature /C

Fig. 5: Calcination of the Y2Ba4Cu6+nO14+n superconductor in argon. Inset A presents


the magnification of the m/z=44 signal. The pulse of CO2 was used for the
calibration.

The MS signals clearly indicate that evolution of carbon dioxide occurs at temperatures
higher than that of superconductor decomposition, this is reflected by the temperature of the
beginning of oxygen evolution. The 1 cm3 pulse of CO2, used later for the calibration of the
m/z=44 signal, was injected before the carbon dioxide release, starting at ca. 600C.
As emerges from the above results, the application of PULSETA allowed to distinguish
signals of CO2 with intensity ratios as small as ca. 1:70 (0.36 vs. 0.005 wt%) and helped to
quantify two steps of CO2 evolution.

99
NETZSCH Coupling of Thermal Analysis and Gas Analysis

(ii) Injection of the gas which reacts with the solid

This procedure enables investigations of the changes of the solid phase and gas
composition resulting from the injected pulse. The interesting feature of PULSETA is its
suitability for investigating complex reaction systems at any desired temperature. Artifacts
caused by temperature and gas phase composition settlings are omitted which facilitates
better distinction of concomitantly occurring processes. Very often e.g. the reduction of the
oxides in multicomponent systems is overlapped by evolution of physisorbed or chemisorbed
species, or occurs already at room temperature, during first contact of the solid sample with
the reducing agent. Due to the possibility of carrying out the reaction at a well-defined
temperature, PULSETA allows the separation of the reactions occurring simultaneously. This
feature is illustrated in Fig. 6 by a study of the determination of the copper content in alumina
supported catalyst by the reduction of CuO with hydrogen.

50 100 150 200 250 300


20
0.0 A TG
197 temperature / oC
15
1

T / V
0.5 10
5
1.0 DTA 0
m = 1.21%
mass loss / %

1.5 1-2 -5
m/z=18 -10
2
0.6

intensity m/z=2 / a.u.


T = 350C m/z=2
0.8 0.00

1.0
1.2
1.4
1.6 TG -1.09%
1.8 B
25 50 75 100 125 150 175
time / min

Fig. 6: Reduction of alumina supported CuO with hydrogen carried out by (A)
conventional TA, and (B) PULSETA at 350C.

The reduction of CuO investigated by conventional TA is shown in Fig. 6A. TG, DTA, and
m/z=18 curves indicate that two processes occur simultaneously. Dehydration of alumina
begins already at room temperature as emerges from the continuous weight loss starting at
30oC and the accompanying weak endothermic effect observed by DTA, and the broad peak
of the m/z=18 curve. Starting from about 150oC, the dehydration of the support is overlapped
by the exothermal reduction of CuO (note the exothermal effect in DTA together with the
weight loss and evolution of water). The weight loss due to CuO reduction calculated from
the conventional TG curve (1.21 wt%) is ca. 10% greater than that obtained by PULSETA
because calculation of the exact weight loss resulting from the reduction is uncertain,
resulting in an imprecise determination of the amount of CuO. Using the PULSETA method
(Fig. 6B), the sample was dehydrated at 350oC till the weight loss caused by the water
evolution was negligible, and reduced progressively by 1 cm3 pulses of hydrogen. The
observed weight loss due to the pulses of hydrogen results from the reduction only and the
determination of the Cu content in the sample is much more accurate.

100
NETZSCH Coupling of Thermal Analysis and Gas Analysis

In some cases, when gas-solid reactions (reduction, oxidation) occur already at room
temperature, the proper application of conventional TA can only be made when starting the
experiment below ambient temperature. Using the PULSETA technique one can carry out
the reaction at any temperature because gas-solid reactions, when investigated by
PULSETA, generally occur only during the pulse. This feature is illustrated by the
determination of the IrO2 content in zeolite supported iridium catalyst. The comparison of the
results obtained by conventional TA and PULSETA is depicted in Figs. 7A and 7B,
respectively.

The determination of the mass change due to the reduction of iridium oxide with hydrogen is
impossible in a conventional TA measurement. The reduction of IrO2 under the applied
conditions (20 vol% H2, balance He, heating rate of 10 K/min) starts already at room
temperature and occurs in a wide temperature range. Very small mass loss due to the
reduction (0.15 wt%, marked in the plot 7A) is overlapped by the mass loss due to
desorption of water from the support. The exact determination of the IrO2 content was
possible after removal of all water under inert atmosphere at 250C followed by hydrogen
pulses (Fig. 7B). The observed mass loss due to the hydrogen pulse originated from the
reduction only. The reduction was complete after the first pulse, the following H2 injections
did not change the mass of the sample anymore.

temperature / C
50 100 150 200 250 300 350 400

0.0 20 vol% H
TG 2

-0.5

-1.0 A
mass change / %

-0.15 %
-1.5
-1.67%
-2.0
TG
0.0
-0.15 %
-0.2
-0.4 B
-0.6 m/z = 2
-0.8
250C, He
-1.0
20 25 30 35 40 45 50 55 60
time / min

Fig. 7: Reduction of zeolite supported iridium catalyst by (A) conventional TA


(10 K/min, H2 20 vol%/He) and (B) PULSETA (250C, 1 cm3 hydrogen pulses).

PULSETA allows dosing the reacting gas in small quantities facilitating the study of gas-solid
reactions in differential mode. This feature provides interesting opportunities for investigating
gas-solid reactions, as e.g. the preparation of solids with well defined extend of reaction.
Depending on the temperature and pulse volume, any required reaction progress can be
achieved. Figure 8 presents the results of the reduction of CuO with hydrogen at two
different temperatures.

101
NETZSCH Coupling of Thermal Analysis and Gas Analysis

50 100 150 200 250

0 time / min
mass loss / %

2
250oC

4 A 450oC

0.328
6
0.332
CuO
450 Cu2O
intensity / a.u.

Cu

B

250

34 36 38 40 42 44 46
two theta / o

Fig. 8: A) Reduction of CuO with H2 pulses at 250 and 450C to the same reduction
extent (marked on TG curves),
B) XRD patterns of the reduced samples.

There is a controversy in the literature concerning the course of the reduction of copper (+2)
oxide with hydrogen. Generally it is assumed that CuO is directly reduced to metallic copper
without intermediates [7], on the other hand, there are suggestions that Cu2O is formed in
the first stage of the reduction [8]. In order to compare the phase composition of the
reduction products, two samples of CuO were reduced to exactly the same extent (ca 0.33)
by pulses of hydrogen at 250 and 450C, respectively. Due to the much lower rate of the
reduction each pulse at 250C resulted in ca. 1.5% change of the reaction extent only. The
same reduction progress was reached after 22 injections at 250C and nine pulses at 450C
(see Fig.8A). The results of the XRD analysis of both reduction products presented in
Fig. 8B confirmed that despite the same overall degree of reduction (one third of the amount
of oxygen present in copper oxide was removed) the phase composition of the products
differed significantly. The sample prepared at lower temperature contained much less
metallic copper and more, poorly crystalline (broad XRD reflections) Cu2O.

This unique feature of PULSETA, which allows to stop reaction at any point between the
pulses, facilitates the elucidation of the relationship between the composition of the solid and
the reaction extent, as also illustrated by the reduction of Mn-oxides by CO pulses shown
in Fig. 9.

102
NETZSCH Coupling of Thermal Analysis and Gas Analysis

time / min.
100 110 120 130 140 150 160 170
1

0 Mn O
mass change / %

2 3

-1 d
TG e
-2 c Mn O
2 3
-3
Mn O Mn O
-4 3 4 3 4 -3.38 %
Mn O
3 4
Mn O
intensity / a.u.

2 3
MnO
c
XRD d
e
30 32 34 36 38 40 42
two-theta /

Fig. 9: Reduction of Mn2O3 by 1 cm3 pulses of CO at 500C and XRD patterns of the
products corresponding to different extent of the reduction taken at the points
marked by 1, 2 and 3.

The reduction of Mn2O3 proceeds in two steps: (i) formation of Mn3O4 (mass loss 3.38 wt%)
and (ii) formation of MnO (mass loss of 6.99 wt%). The aim of the investigation was to check
whether pure Mn3O4 is formed when the stoichiometric mass loss of 3.38% is reached during
the reduction process. The phase composition of the products was checked after 2, 6 and 12
pulses, which afforded the samples labeled as 1, 2 and 3, respectively. The XRD patterns
presented in the lower part of Fig. 9 indicate that sample 3, corresponding to the
stoichiometric weight loss for reduction of Mn2O3 to Mn3O4, contains three phases: Mn2O3,
Mn3O4 and MnO. Even at very low degree of reduction (sample 1, mass loss 0.67 wt%), the
final reduction product, MnO, was already formed. The preparation of samples
corresponding to such a low and well-defined reaction progress is extremely demanding by
any conventional, iso- or non-isothermal TA technique.

The comparison between conventional and differential characterization of gas-solid


processes is illustrated by the investigation of the redox properties of two different CeO2
samples (Figs. 10 and 11). The redox behavior of this oxide applied in automotive exhaust
catalysts, i.e. the ease of its reduction and reoxidation, distinctly influences its application as
an oxygen buffer during catalytic reaction [9].

As the automotive exhaust gas has a cyclic lean-rich composition fluctuation, an oxygen
storage component, which readily undergoes red-ox cycles, can provide oxygen for CO and
hydrocarbon oxidation in the fuel rich region. In the reduced state it can remove oxygen from
the gas phase when the exhaust gas cycles into the lean region. Thus, the cerium oxide is
not only promoting the oxidation activity of the catalyst, but also widens the air-fuel ratio
composition range where all three major pollutants, CO, hydrocarbons, and NO can be
removed. The difficulty in applying conventional TA techniques for investigating redox
phenomena in CeO2 is caused by the fact that only very small changes of the composition of
ceria occur during the catalytic process.

103
NETZSCH Coupling of Thermal Analysis and Gas Analysis

The results of the conventional investigation of the reduction and reoxidation of pure,
labelled as C, and silica doped CS ceria are presented in Figs. 10A and 10B,
respectively.

temperature / C
400 500 600 700 800 900 1000 1100
0

-1 C
CS
-2 A CeO
1.77
mass change / %

2.17%
-3 CeO
1.67
2.89%
-4
3 2.87%
2.20%
2 CS C
B
1

0 200 400 600 800 1000 1200


temperature / C

Fig. 10: A) Reduction in 20 vol. % H2, and


B) re-oxidation in 20 vol% of O2 of pure C and silica doped CS ceria
samples.

The reduction of the silica doped sample (CS) occurs at lower temperature, beginning at ca.
450C and being complete at ca. 900C. The mass loss due to the oxygen removal indicates
that the product of the reduction has the composition of CeO 1.67. The reduction of pure ceria
sample starts at higher temperature (above 600C) and is not complete at 1080C, where
the TG still indicates a mass change. Oxidation of the reduced samples (Fig. 10B) was
carried out immediately after cooling under reducing atmosphere. With pure ceria
(sample C) the beginning of oxidation is very fast and almost 60% of the total mass gain is
reached below 200C. However, the total oxidation of this sample is barely achieved even at
temperatures as high as 1120C. The silica doped sample CS behaves differently: the
oxidation is slower at low temperatures but is already complete at ca. 600C.

The composition of the ceria during catalytic reactions only slightly changes due to
subsequent reduction and reoxidation cycles, and consequently the characterization of the
redox processes of ceria by investigating the total reduction and oxidation processes are of
little value.
Much better characterisation of the redox behaviour of the catalyst, carried out in
incremental steps at temperatures at which the catalytic reactions occur, was done by
applying the PULSETA technique. The results presented in Fig. 11 indicate that after each
hydrogen pulse there is a distinct mass loss due to ceria reduction marked as H2.

104
NETZSCH Coupling of Thermal Analysis and Gas Analysis

temperature / C
720 740 760 780 800
0.1
TG O
2
0.0
O
H 2

intensity m/z = 2, 32 / a.u.


-0.1 2

-0.2 A
mass change / g

m/z = 32
-0.3
-0.4 m/z = 2

TG
0.0
H
-0.1 2

-0.2 B
-0.3 m/z = 32
-0.4 m/z = 2
-0.5
760 770 780 790 800 810 820 830
temperature / C

Fig. 11: Changes of the mass due to 1 cm3 pulses of hydrogen and oxygen for (A)
silica-doped and (B) pure, conventional ceria.

Both samples behave differently during oxygen pulse. With the silica doped sample (CS), a
distinct oxygen uptake is observable, leading to a mass gain greater than the loss that
occurs during reduction. After ca. 15 min the excess oxygen, O2, is desorbed from the ceria
and the sample mass returns to its initial value. The investigation of the changes of the
composition of the ceria would not have been possible by conventional TA technique. The
PULSETA experiments show that for silica doped ceria the change of the composition from
Ce 2.00 (fully oxidized sample) to CeO 1.99 occurs at temperature ca. 90C lower than for pure
ceria. This result revealed the remarkable enhancement of the oxygen storage capacity of
this sample [9].

Investigating gas-solid processes in differential mode, i.e. during very small reaction extents
is especially important when studying catalytic reactions and can help in elucidation of their
mechanism. The usefulness of the pulse technique for investigation of the factors influencing
the activity of unsupported iridium catalysts in the reduction of NOx by hydrocarbons
(HC-SCR process) is shown next.

It was found [10] that conversion of NO to N2 under lean conditions as well as oxidizability
and reducibilty of the Ir-IrO2 system strongly depend on the iridium crystallite size. When
increasing the crystallite size selectivity to N2 was greatly enhanced, mainly by suppressing
reaction pathways leading to the production of NO2. The study of the influence of Ir crystallite
size on the oxidation rate is illustrated in Fig. 12. Samples of iridium with different crystallite
sizes, obtained by controlled sintering in He at different temperatures, were oxidized and
reduced by following pulses of oxygen and hydrogen. The course of TG and MS signals for
the sample with a crystallite size of 27 nm is shown in Fig. 12A. The shape of the TG curve
unambiguously indicates that at this temperature freshly formed iridium oxide is fully reduced
by the next pulse of hydrogen, as indicated by the same mass of the sample after the cycle
of O2-H2 pulses. Fig. 12 B depicts the mass changes due to pulses of H2 and O2 admitted to
samples with different crystallite size. Large Ir crystallites are much less oxidized, which is a
prerequisite for high reaction selectivity towards nitrogen. In contrast, small Ir crystallites are
easily oxidizable and seem to favor oxidation of the reductanats (propene, CO) [11].

105
NETZSCH Coupling of Thermal Analysis and Gas Analysis

time / min
40 50 60 70 80 90
TG
0.4 A
0.2
0.0 O
2

-0.2
mass change / %

H
2
-0.4
1.2
B 4
350C
TG
0.8 13

0.4 27
46
0.0

40 50 60 70 80 90
time / min

Fig. 12: Influence of iridium crystallite size (marked in nm on curves) on the redox
behavior at 350C. The method of investigation is shown in (A) for the
sample with a crystallite size of 27 nm. (B) mass changes due to oxidizing
and reducing pulses for Ir with different crystallite size.

(iii) Injection of a gas which adsorbs onto the sample surface

Due to simultaneous monitoring of mass and thermal effects, PULSETA can be applied for
investigating adsorption phenomena occurring at atmospheric pressure. An example
showing the adsorption of ammonia at 200C on HZSM-5 zeolite is presented in Fig. 13.

1.0
T = 200 C
0.8
mass change / %

TG
0.6
0.4
intensity m/z = 15 / a.u.

0.2
0.0
m/z = 15
10
8
6
T / V

4
2
DTA
0
-2
40 60 80 100 120 140 160 180
time / min

Fig. 13: Adsorption of ammonia (1cm3 pulses) on HZSM-5 zeolite at 250C.


106
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Both, reversible (physisorption) and irreversible adsorptions occur on the zeolite surface.
The irreversible adsorption (chemisorption) occurs during the first pulse. The next pulses of
ammonia are only weakly adsorbed. After ca. 20 minutes the sample mass returns to its
initial value before pulsing, indicating that no further ammonia is adsorbed. The observed
exothermal effect in the DTA curve, in conjunction with the mass gain resulting from the NH3
chemisorption, allows the determination of the heat of adsorption per mole of adsorbed
ammonia. During weak, reversible adsorption the exo- effect on the DTA curve is followed by
the endothermic ammonia desorption. Decreasing the pulse volume gives access to the
determination of the differential heat of adsorption [12].

The application of PULSETA allows not only determining the amount of strongly adsorbed
adsorbate, but also quantification of the desorption process and characterization of the
strength and kind of active sites. Such an example is presented in Fig. 14 depicting the
adsorption of ammonia on titania-silica aerogels.

0 50 100 150 200 250 300


3.5
time / min
3.0 A CMA-13
2.5 2.32%
mass gain / %

2.0
CMA-18
1.5
1.0 1.23%
0.5
0.0 T = 50C
intensity of m/ =15 / a.u.

B CMA-13

CMA-18

50 100 150 200 250 300 350


temperature / C

Fig. 14: A) Adsorption of NH3 on modified titania-silica aerogels at 50C.


B) Desorption of ammonia during heating with a rate of 10 K/min.

Mesoporous titania-silica mixed oxides containing different amount of covalently bound


methyl groups were prepared in a sol-gel process followed by low temperature supercritical
extraction with CO2. The catalysts were used for the epoxidation of various olefins and allylic
alcohols [13]. The olefin epoxidation activity was generally lowered by the methyl
modification, whereas in the epoxidation of allylic alcohols a maximum in activity was
observed with increasing methyl content. In order to gain more insight into the role of methyl
groups in the catalytic behaviour, the acid properties of the surface were characterised by
the adsorption of ammonia at 50C. Methyl groups located at the surface replace part of the
silanol groups, resulting in a decreased ability to adsorb ammonia. The unmodified aerogel
(CMA-13) adsorbed twice the amount of NH3 compared to CMA-18, a catalyst containing 2.4
mg of methyl groups per one gram of catalyst. The shape of the desorption signals
(Fig. 14B) confirmed that in both samples, Brnsted and Lewis acid sites are present: the
desorption of ammonia occurred in the same temperature range what confirms similar
distribution and strength of active sites. The amount of desorbed ammonia was additionally
quantified by injection of a NH3 pulse (not shown in Fig. 14).
107
NETZSCH Coupling of Thermal Analysis and Gas Analysis

PULSETA technique can be applied not only for determining the strongly adsorbed
(chemisorbed) species, but also for investigating weak, reversible adsorption. Results of a
study of weakly adsorbing gases are presented in Fig. 15 depicting the interaction between
CO, O2 and CO2 and an Au/ZrO2 catalyst [4].

10 20 30 40 100 110 120 130 140 150

intensity m/z=28,44 / a.u.


0.20
A time / min B
mass change / mg

0.15 m/z=44
m/z=44
0.10
m/z=28
0.05 m/z=28
12 TG 14 13 TG
0.00 37 39 38

P
intensity m/z=44 / a.u.

C m/z=44

50 100 150 200


temperature / C

Fig. 15: Admission of 1 cm3 pulses of CO at 30C to Au/ZrO2 catalyst calcined at


200C (A) and 475C (B). Plot C depicts the formation of CO2 due to
injection of 1 cm3 pulses of CO during heating the catalyst pre-calcined at
200C (A) and 475C (B) with a rate of 2C/min. 1 cm3 pulses of CO2 (P)
were used for determining the extent of CO oxidation.

The catalytic properties of the supported gold catalysts were found to depend on calcination
temperature. During calcination organic residues present in the catalyst after synthesis
(c.f. Fig. 2) are removed. Also the adsorption of CO was strongly dependent on this
pretreatment. The catalyst precalcined at 200C adsorbed much less CO than that calcined
at 475C. The CO uptake shown on the TG curves is expressed in g. The adsorption of CO
is weak and fully reversible at 50C, CO desorbs after 10-15 minutes. All experiments shown
in Fig. 15 were carried out in an atmosphere of 20 vol% O2/He, therefore simultaneous with
adsorption, catalytic oxidation of the injected CO occurred and was monitored. As emerges
from the mass spectrometric signals of produced CO2 (m/z=44), also the catalytic activity of
Au/ZrO2 was strongly influenced by the catalyst pretreatment. At room temperature
(Fig. 15B), as well as during slow temperature change (Fig. 15C), the catalyst
calcined at 475C (traces B) was much more active than that calcined at 200C only
(trace A in Fig. 15C).

108
NETZSCH Coupling of Thermal Analysis and Gas Analysis

Conclusion:

Pulse thermal analysis extends the versatility of conventional thermoanalytical methods by


providing a means for studying small reaction progresses. This advantage is combined with
all the opportunities of thermogravimetry, differential thermal analysis or differential scanning
calorimetry and evolved gas analysis. The primary benefits of the new method are:

- Injection of a known amount of gas into the system during measurement allows in situ
quantitative calibration of the MS or FTIR signals, which greatly increases the opportunities
especially for the analysis of complex and multicomponent systems.
- Monitoring gas-solid processes corresponding to a specific extent of reaction at desired
temperature is possible. The reaction can be stopped at any point between pulses,
enabling to unveil the relationship between the composition of the solid and the reaction
progress.
- The potential for monitoring simultaneous changes in mass, thermal effects, composition
and amount of gaseous reactants and products under pulse conditions. This feature
enables to gain simultaneously information concerning both gas and solid phases.

Literature:

1. M. Maciejewski, C. A. Mller, R. Tschan, W.-D. Emmerich and A. Baiker, Thermochim.


Acta, 295 (1997) 167
2. M. Maciejewski and A. Baiker, Thermochim. Acta, 295 (1997) 95
3. B. Roduit, J. Baldyga, M. Maciejewski and A. Baiker, Thermochim. Acta, 295 (1977) 59
4. M. Maciejewski, P. Fabrizioli, J.-D. Grunwaldt, O. S. Becker and A. Baiker, Phys. Chem.
Chem. Phys., 3 (2001) 3846
5. M. Maciejewski, A. Baiker, E. Kapsch, M. Hollering and W.-D. Emmerich, GIT-Labor
Fachzeitschrift, June 2001
6. M. Maciejewski, A. Baiker, K. Conder, Ch. Krger, J. Karpinski and E. Kaldis,
Physica C227 (1994) 343
7. H. H. Voge and L. T. Atkins, J. Catal., 1 (1962) 171
8. M. J. Murray, J. Phys. Chem., 35 (1931) 1011
9. E. Rocchini, A. Trovarelli, J. Llorca, G. W. Graham, W.H. Weber, M. Maciejewski and
A. Baiker, J. Catal., 194 (2000) 461
10. C. Wgerbauer, M. Maciejewski, A. Baiker and U. Gbel, Top. Catal., 16/17 (2001) 181
11. C. Wgerbauer, M. Maciejewski, and A. Baiker, J. Catal., in press
12. F. Eigenmann, M. Maciejewski, and A. Baiker, Thermochim. Acta, 359 (2000) 131
13. C.A. Mller, M. Maciejewski, T. Mallat, and A. Baiker, J. Catal., 184 (1999) 280

109
Reply

To NETZSCH-Gertebau GmbH
Fax No.: (+49) 9287 / 881 44 D-95100 Selb/Germany

Coupling of
Thermal Analysis and Gas Analysis

Techniques and Applications

Editors: E. Kaisersberger, E. Kapsch, E. Post, A. Schranner

My address:

Company:

Name: Dept.:

Street:

City/Post Code:

Phone: Fax:

VAT No.: e-mail:

Date/Signature FL-Coupling/8-01
NETZSCH-Gertebau GmbH

Wittelsbacherstrae 42, D-95100 Selb/Germany


phone +49 9287 8810, fax +49 9287 88144
e-mail: at@nbg.netzsch.com
http://www.ngb.netzsch.com

You might also like