See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/328959555

Thermogravimetric Analysis of Polymers

Article · November 2018

DOI: 10.1002/0471440264.pst667

CITATIONS READS

6 12,908

6 authors, including:

H.M. Ng Norshahirah mohamad saidi

20 PUBLICATIONS   243 CITATIONS   
University of Malaya
14 PUBLICATIONS   40 CITATIONS   
SEE PROFILE
SEE PROFILE

Rameshkasi Kasi Ramesh T subramaniam

University of Malaya University of Malaya
143 PUBLICATIONS   1,667 CITATIONS    206 PUBLICATIONS   4,221 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Optimized gel polymer electrolytes for dye sensitized solar cell application View project

Biomedical applications of Hydrogels View project

All content following this page was uploaded by Shahid Bashir Baig on 02 August 2020.

The user has requested enhancement of the downloaded file.

THERMOGRAVIMETRIC ANALYSIS OF POLYMERS
1. Introduction and History of Thermogravimetric Analysis

Thermal analysis of polymers typically involves a combination of various tech-

niques, including well-established thermogravimetric analysis (TGA), differential
thermal analysis (DTA), differential scanning calorimetry (DSC), and evolved gas
analysis (EGA) (1).
The earliest TGA experiments were performed by heating the samples to a
specific temperature and then removing the sample from the sampling head at
specific time intervals for weighing (2). In 1912, Urbain and Boulanger (3) built
a breakthrough device that comprised a conventional balance adapted for null-
point electromagnetic compensation, with the sample hanging from the balance
arm into an electrically heated furnace in a gas-tight environment. They also de-
signed it to have an input for the supply of absorbent materials for evolved gases.
In this setup, the sample could be weighed while it was continuously heated. In
1915, Honda took the idea of the initial TG development and produced a more de-
pendable TG (4). His equipment was designed for the heating rate to be gradually
reduced from an initial rate during the occurrence of weight losses. It could also
be called a sample-controlled thermogravimetric system. In this period, TG was
mostly used for the analysis of inorganic compounds, and it reached its pinnacle
in Duval’s monumental description of over 1000 gravimetric precipitates of nearly
80 elements (5).
In modern TGA, it determines the quantity and the frequency of the weight
variation of the samples against temperature and time in a controlled atmosphere
(eg, the purging of nitrogen gas) (6–8). TGA can be used primarily to investigate
the thermal stability (the strength of the material at a given temperature), oxida-
tive stabilities (the oxygen absorption rate on the material), as well as the com-
positional properties (eg, fillers, polymer resin, solvents) of the samples. Besides,
the weight gain/loss of the samples corresponds to different factors. Generally,
weight gain is attributed to the adsorption or oxidation, whereas weight loss is
attributed to decomposition, desorption, dehydration, desolvation, or volatiliza-
tion (9–12). TGA is especially important in applications that use polymeric ma-
terials such as injection molding, coating material for electrical and electronic
components, paints, pharmaceutical, food industries, petroleum, adhesive, and so
forth (13).
Other types of thermal analyses such as dynamic mechanical analy-
sis (DMA), DSC analysis, and temperature-modulated differential scanning
calorimetry (TMDSC) could complement the TGA analysis. The DMA analysis

1
Encyclopedia of Polymer Science and Technology. Copyright c⃝ 2018 John Wiley & Sons, Inc. All rights reserved.
DOI: 10.1002/0471440264.pst667
2 THERMOGRAVIMETRIC ANALYSIS OF POLYMERS

is useful to study the viscoelastic behavior of polymeric materials as a function

of temperature or frequency. The material is subjected to a sinusoidal mechanical
stress, and the strain of the sample is measured. DMA allows to detect thermal
effects from the variations in the complex modulus. TMDSC can complement TGA
by giving more information than conventional DSC. However, the measurement of
the glass transition temperature is limited because the short range of frequency
used. TMDSC analysis can provide a better absolute heat capacity with proper
calibration. It could be used to study the reversible melting and crystallization of
polymer (14,15). Meanwhile, DSC analysis is used to measure energy absorbed or
released by a material as a function of temperature when the material is heated
or cooled. It could provide quantitative and qualitative data on endothermic and
exothermic processes. The properties such as melting temperature, glass transi-
tion temperature, crystalline phase transition temperature, and specific heat or
heat capacity could be obtained from DSC analysis (16).

2. Theory

In general, TGA studies the changes in the weight of a sample while the sam-
ple is being heated or cooled at a controlled temperature, T(t), and the changes
are continuously monitored. The program that controls the temperature can be
isothermal (T (t) is constant) or nonisothermal. Isothermal is a condition when
T(t) is constant, and nonisothermal is a condition when the heating rate is con-
stant (caused by the linear change in temperature with time). The heating rate
(𝛽) can be expressed as

dT
𝛽= (1)
dt

where dT is the change in temperature and dt is the change in time (17).

The mechanism and kinetics of the process linked to the weight changes of
the sample can influence the shape and position of the TG curves. Three major
variables that can influence the kinetics of the thermally simulated processes are
the temperature (T), the extent of the conversion (𝛼), and pressure (P) (18,19). The
variable can be implicitly shown as

d𝛼
= 𝜅 (T ) f (𝛼 ) h (P) (2)
dt

Although pressure can have a large effect on the kinetics processes, h(P) is usually
ignored in the kinetic simulations (where h(P) = constant, that is, the environment
is equilibrium). This happens when the entire reactive gas that reacts with the
sample is removed (either by flowing an excess amount of reactive gas in a gas–
solid reaction or purging high flow rates of inert gas).
By assuming pressure has no effect on the kinetics, the process rate would
become a function with only 2 variables, T and 𝛼:

d𝛼
= 𝜅 (T ) f (𝛼 ) (3)
dt
 THERMOGRAVIMETRIC ANALYSIS OF POLYMERS 3

where 𝜅(T) represents the temperature dependence of the rate and the f(𝛼) is the
conversion dependence by the reaction model. The extent of conversion is deter-
mined from the experiment as a fraction of total weight loss:

mi − m
𝛼= (4)
mi − mf

where mi and mf are the initial and final weight of the sample, respectively. The
𝛼 value can vary from 0 to 1, referring to the overall transformation from the
beginning (materials) to completion (products).
The temperature dependence of the rate constant commonly obeys the Ar-
rhenius equation (20):
( )
−E
k (T ) = A exp (5)
RT

where A and E are kinetic parameters, for example, the preexponential factor
(representing the frequency of the occurrence of a condition that leads to reaction)
and activation energy, respectively. The universal gas constant is denoted as R,
and T is the temperature that is set by the TGA program (in kelvins). Equation (5)
can be described as an isothermal TG curve:
𝜒 ( )
d𝛼 −E
g (𝛼 ) = = A exp t (6)
∫ f (𝛼 ) RT
0

where g(𝛼) is the integral form of the reaction model. The above equation can be
obtained by substituting equation (5) into equation (3), followed by an integration
as shown. In an isothermal condition, normally only 𝛼 would be changing through-
out the experiment. Thus, 𝛼 versus t curve would be defined by reaction models
(21). The reaction models can be separated into three major models: (1) acceler-
ating, (2) decelerating, and (3) sigmoidal. It would look like Figure 1 when it is
plotted on the weight loss curve.

1. Accelerating models can be shown as the weight loss rate increases contin-
uously throughout the entire process, and the model can be shown as below:

g (𝛼 ) = 𝛼 1∕n (7)

where n is a constant value.

2. As for the decelerating model, it was represented by the weight loss rate
that decreases throughout the entire process, and the model can be shown
as below:
1 − (1 − 𝛼 )1−n
g (𝛼 ) = (8)
1−n
where n is a reaction order ≠1 (for n = 1, g(𝛼) = −ln(1 − 𝛼)).
4 THERMOGRAVIMETRIC ANALYSIS OF POLYMERS

Weight loss (%)

3

Time (s)

Fig. 1. The reaction models of accelerating (1), decelerating (2) and sigmoidal (3) when it
is plotted into the weight loss curve (17).

3. The Avrami–Erofeev model is applied to the sigmoidal models as this model

shows processes when the rate reaches its maximum point at some inter-
mediate values during the conversion. The Avrami–Erofeev model is shown
below:
[ ]1∕n
g (𝛼 ) = −ln (1 − 𝛼 ) (9)

On the other hand, equation (5) can be substituted into equation (3), fol-
lowed by integration of equation (1), producing an equation that describes
a nonisothermal TG curve as explained below:
T ( )
A −E
g (𝛼 ) = exp dT (10)
𝛽 ∫ RT
0

However, the shape formed from this nonisothermal TG curve cannot be re-
lated to the reaction model 3. For the nonisothermal condition, the heating rate
remains constant, which is caused by simultaneous variation of both T and 𝛼. This
condition contradicts with the condition of reaction model 3 in which the sigmoidal
reaction has an acceleratory phase followed by a decelerator phase.

3. Instrumentation and Operating Principles

TG instrument (Fig. 2) consists of three different components to provide the

flexibility for the accurate analytical data: balance (for measuring weight of
 THERMOGRAVIMETRIC ANALYSIS OF POLYMERS 5

Weighing mechanism

Weight
Furnace Temperature

Sample Display Tare

Temperature
programmer

Gas Gas

Fig. 2. Schematic diagram of TG instrument.

sample), furnace (providing linear heating), temperature programmer (for tem-

perature measurement and control), and recorder (for recording unit for weight
and temperature changes) (10). As the weight variations occur, a balance beam
tends to deviate from its normal position. There are two types of balance can be
used: null or deflection balance (22–25). A null type of balance is commonly used
in which a sensor is employed to sense the deviation of the balance beam from its
null point position. A restoring force is then applied to the balance to restore its
null point. Some systems measure the beam deflection by converting the beam
deflection using different techniques (photographic recording, recording electrical
signals, or using an electrochemical device) into a suitable weight.
There are several variables that can affect the sensitivity of the instrument
and alter the shape of the TG curves (i–v). By considering those variables, the
instruments are required to be calibrated before conducting the experiment.

I. Sample: The chemical description including its source, purity, and pretreat-
ments, the particle size, must be clearly known.
II. Sample holder: The capacity of the holder should be large enough to hold
the whole sample to avoid loss of signal due to spurting and creeping. Then,
to obtain an accurate data or signal, the heat transfer between the furnace
and sample holder needs to be really effective, which is why a metal holder
is better.
III. Crucible: A recommended type of crucible is a platinum crucible, which is
cylindrical and has sufficient height. Other types of the recommended cru-
cible are gauze crucible, polyplate crucible, and labyrinth crucible.
IV. Atmosphere: Choice of different atmospheric environment depending on the
type of gas purged in. It could be inert or oxidative to the sample. The
6 THERMOGRAVIMETRIC ANALYSIS OF POLYMERS

chemical changes due to the oxidative environment could be detected dif-

ferently by the analyzer, and it could become a very useful information re-
garding the characterization of a sample.
V. Heating rate: The rate of decomposition depends on the heating rate. The
higher the heating rate, the higher the decomposition temperature. Usually,
a recommended heating rate is 3.5◦ C/min for reproducible TGA.

Typically, a TGA instrument offers a temperature range from 1000 to 1600◦ C

and the heating rate can be set from a range of a few decimal places to hundreds
of ◦ C/min. However, typical heating rate is between 1 and 20◦ C/min, with very
fast (>20◦ /min) or very slow (<1◦ /min) rates used only for some special reactions.
The resolution of the thermogram can be controlled by the programmed heat rate,
and usually the slower the heating rate the higher the resolution of the curves.
The modern TGA can run on a single temperature program (eg, a constant heat
rate at 5◦ C/min from 30 to 900◦ C) or using a several combination of temperature
programs (eg, isothermal at 100◦ C for 1 h to totally remove all the adsorbed mois-
ture followed by heating from 100 to 900◦ C at a heating rate of 10◦ C/min to study
the thermal properties of the sample).
The sample tested could be either in solid, gel, or liquid form with the weight
ranging from as small as 1 to 100 mg, up to 100 g. Some of the examples of the mass
change processes are decompositions, vaporizations of the solvents, oxidations,
and reductions of the polymer electrolytes at different temperature and differ-
ent atmospheric environment. On the other hand, the TG curves data are usually
coupled with a derivative thermogram (DTG) curve to allow better understand-
ing or resolution of consecutive weight changes. The DTG curve can be extracted
from the TG curve data by deriving the weight data as a function of tempera-
ture. Figure 3 illustrates the TG curves together with its DTG curves of vertically
aligned reduced graphene oxide/polyvinyl alcohol (VArGO/PVA) composite. From
the figure, it can be seen that the first small peak of DTG was associated with the
first weight loss of the composite, followed by the second and third as labeled in
Figure 3.

4. Applications

Thermogravimetry provides information regarding changes in mass of the ma-

terials due to change in temperature or time in a controlled manner. Therefore,
this is specific to the study of decomposition and oxidation reactions and physical
process such as sublimation, vaporization, and desorption. It is especially useful
for the study of polymeric materials, including thermoplastics, thermosets,
elastomers, composites, films, fibers, coatings, and paints. TGA measurements
provide valuable information that can be used to select materials for certain
end-use applications, predict product performance, and improve product quality.
The technique is particularly useful for the following types of measurements.
4.1. Thermal Stabilities Assessment and Moisture Content. Ther-
mal stability is one of the most important properties of polymers to study their ap-
plications in numerous fields. The analysis of thermal stability of polymers using
 THERMOGRAVIMETRIC ANALYSIS OF POLYMERS 7

100 2.0
Thermogravimetric curve
Derivative thermogravimetric curve 1.8
80
1.6
60

Derivative weight (%/°C)

1.4

40 1.2
Weight (%)

1.0
20
0.8
0
0.6
–20 0.4
I II III

–40 0.2

0.0
–60
100 200 300 400 500 600 700 800 900
Temperature (°C)

Fig. 3. TG and DTG analysis of VArGO/PVA composite (26).

thermogravimetry is a prominent application of this technique. Thermogravimet-

ric curves provide decomposition mechanism of the polymers. This decomposition
profile defines the characteristics and identification of the polymer. The onset of
mass loss often defines the upper limit of thermal stability for the material. How-
ever, it should be appreciable that extensive degradation of polymer may already
take place before the point at which detectable changes in mass occur. Polymer
degradation can be categorized through the following mechanism:

a. main-chain scission
b. side group scission
c. elimination
d. depolymerization
e. cyclization, and
f. cross-linking.

The categories (a)–(d) usually result in the evolution of unstable volatile

products, hence mass change occurs. The assessment of thermal properties
depends on the environment during the analysis. The polymers quantitatively
degrade into parent monomers in an inert atmosphere. However, oxidation of
polymer into oxides of the constituent elements occurs in the presence of air. For
example, halogen containing polymers yield hydrogen halides (H-X) and nitrogen
containing polymers form hydrogen cyanide or ammonia. Moreover, cyclization
and cross-linking rarely occur unless in conjunction with (a)–(d). These two
cannot be detected easily by thermogravimetry.
Figures 4a and 4b show the thermogravimetric curves of polypropylene after
passing through nitrogen and oxygen, respectively. The degradation behavior is
8 THERMOGRAVIMETRIC ANALYSIS OF POLYMERS

105
(a)
90

75 PP x0
PP x1

Mass (%)
60
PP x2
45 PP x3
PP x4
30 PP x5
PP x6
15

0
300 350 400 450 500

105
(b)
90

75 PP x0
PP x1
PP x2
Mass (%)

60
PP x3
45 PP x4
PP x5
30
PP x6
15

0
200 250 300 350 400
Temperature (°C)

Fig. 4. Thermogravimetric curves of polypropylene in (a) nitrogen and (b) oxygen (27).

clearly different under nitrogen and oxygen atmosphere. The systematic degrada-
tion is observed under a nitrogen atmosphere, which is due to the chain scission of
polypropylene. However, the oxidation temperature dropped sharply for samples
under an oxygen atmosphere.
Furthermore, thermogravimetric curves of polyethylene in nitrogen and oxy-
gen are shown in Figures 5a and 5b. The curves for nitrogen show only slight
differences in the degradation temperatures; however, they do not change sys-
tematically, but they do decrease with the number of extrusion cycles, again, due
to the scission of the polyethylene chains, which become shorter and are prone to
thermal degradation at lower temperatures. The curves for oxygen are very dif-
ferent from each other and show an initial mass increase owing to oxygen uptake
and formation of peroxy radicals and hydroperoxides, which are not stable toward
higher temperatures and rapidly convert into other labile products and are sub-
sequently evolved. The mass increase is higher for the degraded samples since
shorter chains react promptly with oxygen. After reacting with oxygen, the repro-
cessed samples start losing mass faster since volatile compounds are evolved (27).
Similarly, the same study reveals the systematic decrease in degradation
temperature of polypropylene under nitrogen atmosphere, which might be due to
 THERMOGRAVIMETRIC ANALYSIS OF POLYMERS 9

105
(a)
90

75
PE x0
PE x1
Mass (%)

60
PE x2
45 PE x3
PE x4
30 PE x5
PE x6
15

0
400 450 500
Temperature (°C)

105
(b)
90 PE x0
PE x1
75 PE x2
PE x3
Mass (%)

60 PE x4
PE x5
45 PE x6

30

15

0
150 300 450 600
Temperature (°C)

Fig. 5. Thermogravimetric curves of polyethylene in (a) nitrogen and (b) oxygen (27).

scission of polypropylene chains. Although it is widely accepted that polypropy-

lene degrades via chain scission, polyethylene mainly undergoes branching
and cross-linking, in addition to chain scission. The results presented above
demonstrate that polyethylene and polypropylene under the extrusion conditions
undergo chain scission, which is consistent with the results published by some
authors (28–30).
Figure 6 displays the thermal degradation curves of microcrystalline cellu-
lose (MCC) and cellulose tri-stearate (CTs) under argon environment. The curve
shows the series of thermal degradation such as dehydration at 100◦ C (evapora-
tion of moisture contents) and thermal pyrolysis of cellulose backbone at 300◦ C.
However, the thermal degradation of CTs shows different thermal curve as com-
pared MCC. CTs did not show significant weight loss around 100◦ C because of
hydrophobic nature. The temperature corresponding to maximum thermal degra-
dation of MCC was 344.6◦ C whereas 371◦ C for CTs. The thermal stability of CTs
is higher than MCC ,which is due to the substituted groups. The long-chain fatty
acid groups on the cellulose chain can be in a regular arrangement and form an
10 THERMOGRAVIMETRIC ANALYSIS OF POLYMERS

100

80

(1–α)/% 60
MCC
40 CTs

20

0
100 200 300 400
Temperature/°C

Fig. 6. Thermal degradation curves of MCC and CTs (31).

ordered structure. Such work is useful in predicting the thermal stability of poly-
mer skeleton after substitution (31).
The degradation of block and graft copolymers rarely corresponds to the
sum of their parent polymers. Copolymerization may result in a material that
is thermally more stable than either of the parent polymers. TGA of pure
high-fluidity polypropylene (HF-PP), metallocene poly(ethylene-butene-hexene)
copolymer (mEBHC), and their blends at a heating rate of 10◦ C/min under a nitro-
gen atmosphere was investigated. The thermal degradation of HF-PP shows one
stage of weight loss, which exhibits the onset degradation temperature (Tonset , at
5% weight loss) at 373◦ C, whereas the Tonset of mEBHC is 368◦ C. Incorporation of
mEBHC was found to increase the thermal stability of HF-PP with the increase
in Tonset . Overall, the more mEBHC is incorporated, the higher the Tonset of the
blend (32).
4.2. Filler Content in Polymers. Fillers are particles added to the ma-
terial (plastics, composite material, concrete) to lower the consumption of more
expensive binder material or to better some properties of the mixture material.
Worldwide, more than 53 million tons of fillers with a total sum of approximately
EUR16 billion are used every year in different application areas, such as paper,
plastics, rubber, paints, coatings, adhesives, and sealants. As such, fillers, produced
by more than 700 companies, ranked among the world’s major raw material and
are contained in a variety of goods for daily consumer needs (33).
Formerly, fillers were used predominantly to cheapen end products, in which
case they are called extenders. Among the 21 most important fillers, calcium car-
bonate holds the largest market volume and is mainly used in the plastics sec-
tor (34). While the plastic industry mostly consumes ground calcium carbonate,
the paper industry primarily uses precipitated calcium carbonate that is derived
from natural minerals. Wood flour and saw dust are used as filler in thermosetting
plastic.
 THERMOGRAVIMETRIC ANALYSIS OF POLYMERS 11

In some cases, fillers also enhance properties of the products, for example,
in composites. In such cases, a beneficial chemical interaction develops between
the host material and the filler. As a result, a number of optimized types of fillers,
nano-fillers, or surface-treated goods have been developed.
Polymers reinforced with nanofillers are promising candidates for various
applications requiring lightweight and high-performance, multifunctional mate-
rials. The basic characteristics of nanosized materials that are responsible for
their unique properties are the nanometer length scales and enormously large sur-
face area-to-volume ratio. These features can enhance the overall material perfor-
mance by synergistically producing unique properties resulting from phenomena
that occur only when the morphology and the physics coincide at the nanoscale
(35). The main challenges in utilizing these materials as polymer fillers are to
control and optimize their dispersion within the polymer matrix and their inter-
face chemistry, and maintain their nanoscale morphology during the manufac-
turing process (36). In addition to the filler and manufacturing process employed
to fabricate the composites, the resulting properties strongly depend on the poly-
mer used as a matrix. Petroleum-based polymers such as polypropylene, nylons,
polystyrene, and epoxies have commonly been used (37).
TGA is the most common technique for determining the quantity of a filler
within a polymer. The product is placed on a balance inside of a furnace and heated
to burn away all of the organic material until the filler content is all that remains.
TGA provides a percentage of mass that remains from the initial amount of mate-
rial tested. Additionally, a muffle furnace can be used to determine the ash content
(filler) as well.
Rheem and co-workers prepared thermally expandable microspheres of acry-
lonitrile, methyl methacrylate, and methacrylonitrile containing a blowing agent
through suspension polymerization. These microspheres showed two-step decom-
position in TGA as shown in Figure 7. The first stage represented the outburst
and evaporation of the blowing agent, and the second stage was due to the decom-
position of polymers (38).
Thermal analysis of the polyolefines (polypropylene and polyethylene) and
fillers was performed in nitrogen and air atmosphere. From Figure 8, TGA curves
of the fillers in nitrogen as well as air atmosphere showed a slight weight loss
between 40 and 100◦ C, indicating the vaporization of water. A second weight loss
from approximately 150–500◦ C was due to the decomposition of the three ma-
jor constituents of biofillers, namely cellulose (275–350◦ C), hemicellulose (150–
350◦ C), and lignin (250–500◦ C). At 700◦ C, rice husk and saw dust left a higher
amount of char content, decomposing only by about 65% and 80% in nitrogen at-
mosphere. The ash in the rice husk was mainly composed of silica (96%). In an
air atmosphere, rice husk degradation shifted to lower temperatures and split
into two processes (320 and 438◦ C). The second might be associated with ther-
mal oxidation degradation of char. Therefore, the residues of rice husk filler at
700◦ C in the air atmosphere (20%) were lower than that in a nitrogen atmosphere
(36%) (39).
4.3. Compositional Analysis of Multicomponent Polymers. One
of the most important applications of TGA is the assessment of the
compositional analysis of polymeric blends. Among thermal analysis techniques,
thermogravimetry can be used for precise and accurate analysis of composition
12 THERMOGRAVIMETRIC ANALYSIS OF POLYMERS

120
V-70 30 °C
V-70 40 °C
100 V-70 50 °C
V-65 50 °C
V-65 60 °C
80
V-65 70 °C
Weight % (%)

V-59 60 °C
V-59 70 °C
60
V-59 80 °C
AIBN 60 °C
40 AIBN 70 °C
AIBN 80 °C

20

0
0 100 200 300 400 500 600
Temperature (°C)

Fig. 7. TGA thermograms of expandable microcapsules prepared with different initiators

and temperature (38).

100

80

60
Weight (%)

40 RH-N2
RH-air
SD-N2
20
SD-air

0 100 200 300 400 500 600 700

Temperature (°C)

Fig. 8. TGA curves of stock materials: fillers saw dust and rice husk as well as polyolefines
PP and PE in nitrogen and air atmospheres (39).

and for the identification of polymers from their decomposition pattern and, there-
fore, it can be used for quality and process control. Compositional analysis of mul-
ticomponent polymers can be done using TGA–FTIR (Fourier infrared spectrom-
etry), high-resolution TGA, auto stepwise TGA, and sample-controlled thermal
 THERMOGRAVIMETRIC ANALYSIS OF POLYMERS 13

analysis (SCTA). The information about the composition of the polymer blend can
be developed through these numerous types of TGA, where the TGA heats up the
polymer samples and holds it under isothermal condition and instrument detects
weight loss. Fernández-Berridi and co-workers used high-resolution TGA to mon-
itor the blend composition of rubber (tire). In hi-resolution TGA, the heating rate
of the sample material is dynamically and continuously modified in response to
changes in the rate of decomposition of the material so as to maximize weight
change resolution. This approach allows very high heating rates to be used in re-
gions where no weight changes occur, whereas the heating rate is decreased when
a weight change is detected. This method gives narrower and more intense deriva-
tive weight loss peaks with a greater resolution from competing and overlapping
weight losses, and even with a substantial reduction of the experimental time.
This approach was used to analyze the main components of rubber under nitro-
gen atmosphere. Thermogram of TGA showed different weight losses at a different
temperature, which indicated the components of rubber. The first weight loss un-
der nitrogen flow, between 200 and 300◦ C, corresponds to 9% of the total weight of
the tire sample and is attributed to the volatilization of processing oil or any other
low boiling point components. The next mass loss of 37% observed in nitrogen flow,
which has a maximum rate at 350◦ C, due to the decomposition of natural rubber,
whereas the weight loss of 25% at 424◦ C corresponds to styrene butadiene rubber.
These temperatures have been assigned according to the obtained results from the
pure component TGA. Finally, when air atmosphere is activated, oxygen reacts
with carbon black and leads to the fourth weight loss at 563◦ C. The nonvolatile
residue (6%) corresponds mainly to inorganic fillers (40). Auto-stepwise TGA can
be used to observe the composition of polymer blends. It was used to observe the
composition of ABS polymer alloy composed of a styrene acrylonitrile copolymer
with butadiene. The weight loss transition of the butadiene rubber component oc-
curred extremely close to the decomposition of the SAN copolymer. Standard TGA
cannot even separate the two events. However, auto-stepwise TGA potentially an-
alyzed the quantitative compositional analysis of ABS matrix. The butadiene was
nicely separated from SAN decomposition, and this provided excellent informa-
tion, which is shown in Figure 9.
Conventional TGA has limitations in the case of closely occurring events
that happen at a similar temperature range and the overlapping of decomposition
events occur. Attempts to improve the ability to separate closely occurring event,
by changing the heating rate in conventional heating schedules, might yield better
results but are not always sufficient for a proper discrimination.
Plastisols are composed of vinyl resins mainly polyvinyl chloride (PVC) and
plasticizers that increase the flexibility. The plasticizer contents have severe side
effects on human health and on the environment. The contents of plasticizer
can be determined using special TGA. Conventional TGA cannot determine the
plasticizer contents as the decomposition of PVC overlaps with the release of
plasticizer. SCTA is a type of TGA that provides much higher resolving power and,
therefore, has an ability to successfully overwhelm the problems of conventional
heating rates for the separation of closely occurring decomposition events. In
thermal analysis, the resolving power concerns the extent to which adjacent, or
partially overlapping, thermally induced processes are separated. Despite this
ability, the potential of SCTA for analytical studies has not been explored so far.
14 THERMOGRAVIMETRIC ANALYSIS OF POLYMERS

Weight loss (%) ABS SAN

Auto stepside TGA copolymer

Butadiene Carbon

Temperature (°C)

Fig. 9. Auto-stepwise TGA results for ABS polymer alloy (41).

In SCTA experiments what is controlled is not the temperature–time profile, as

in conventional analysis, but rather the evolution of the reaction rate with the
time. Most usually, the temperature is controlled in such a way that the reaction
rate is maintained at a constant value previously selected by the user along the
entire process. In such a case, the technique is known as constant rate thermal
analysis (CRTA). The enhanced control over the decomposition process provided
by CRTA makes it a much more effective tool than conventional methods for
the discrimination of overlapping processes. Furthermore, the temperature–time
profile, which is not known in advance and depends on the evolution of the
reaction with temperature, can provide a good deal of information about the
nature of the process studied, since it can be related to the mechanism that
drives the reaction. Sánchez-Jiménez and co-workers used the sample-controlled
thermal analysis to determine the composition of plastisols containing PVC and
plasticizers, a system where the individual components have very similar thermal
stabilities, thereby rendering useless thermogravimetric experiments run under
conventional conditions. Different SCTA procedures, such as CRTA, which has
received special attention, and the results obtained have been compared with
the linear heating rate technique. It has been proven that CRTA can be used
effectively to determine the exact composition of the blend (42).
4.4. Measurement of Low Level of Volatiles. Another important ap-
plication of TGA is the quantitative measurement of a low level of volatile matter
(typically with the weight loss less than 1%) evolved by samples. Examples of
volatile matter are water, residual solvent, residual monomer, and low molecular
weight polymer. In the pharmaceutical sector, the existence of moisture, which
may come from a bulk drug, inactive excipients, and environmental conditions,
can greatly affect the product stability, tablet compaction, microbial growth, and
wet granulation (43). Likewise in injection molding technology that use poly-
mers (eg, thermoplastics, thermosets, and elastomers), the measurement of the
 THERMOGRAVIMETRIC ANALYSIS OF POLYMERS 15

100

99.95 Low level of volatiles from PET resin

at 130 °C
Weight % (%)
99.90
Delta Y = 0.219%

99.85

99.80

99.75
0 100 200 300 400 500 600
Time (min)

Fig. 10. Low level of volatiles from a single pellet of PET under isothermal condition (41).

variation of volatiles has become increasingly critical. High sensitivity and ex-
cellent stability of TGA is very useful to determine very small weight changes of
materials. This is due to the fact that the blow molding process can be affected
by the existence of small amounts water, solvent, or monomer. As a consequence,
bubbles or imperfections may be produced during the process.
Another example for PET is shown in Figure 10 in which the measurement
was taken during the sample was kept under isothermal conditions at 130◦ C. The
weight loss observed was only 0.219% at 10 h (41).
4.5. Sample-Controlled Thermogravimetry. Generally, thermo-
gravimetry analysis is conducted under isothermal or linear rising temperature
(linear nonisothermal) conditions. In isothermal, this practice may contribute
to errors because the samples usually do not heat instantaneously when a high
temperature is applied. While for linear nonisothermal, it takes a long duration
of analysis when need to adjust the temperature to limit the maximum rate
of reaction. Thus, there is another alternative that enables a precise control of
the reaction environment. That is the rate of increase of induced temperature
is monitored (either by slowing down or suspending) once some predetermined
weight loss is detected. This smart temperature control technique is called
sample-controlled thermogravimetry. Most often, the technique is implemented
in the industrial sector to achieve a desired characteristic profile, high quality, or
specific sample properties.
Two types of control heating approach have been designed: stepwise isother-
mal and dynamic rate (45). Stepwise isothermal (proposed by Sorenson) is used
when the rate of temperature is governed by alternating the induce temperature
variation between linear rising temperature with isothermal segments. During
analysis, the temperature is increased when the reaction rate drops below the
minimum until the rate exceeds the maximum. Then, the induced temperature
is held constant until the reaction rate again drops below the minimum. At this
point, the temperature is again increased causing the whole steps to be performed
16 THERMOGRAVIMETRIC ANALYSIS OF POLYMERS

Fig. 11. TG curve of low-density polyethylene/nylon-6 laminate obtained at a dynamic

heating rate (47).

over and over until all weight loss variations have been obtained. The main benefit
of this approach is the determination of steps used for multiple step decomposi-
tion from a single TGA analysis. In addition, these steps can be reused for the
future analysis on similar samples.
While for dynamic heating rate (a novel method patented by Sauerbrunn and
Gill), the rate of heating is sigmoidally linked to the rate of weight loss (46). Dur-
ing decomposition, the rate of temperature variation is slowly reduced from the
initial rate as the rate of weight loss increases. This approach results in high res-
olution of particular thermal events. For instance, the heating rates can be set up
in the range from a fraction to hundreds ◦ C/min, the most commonly used range is
1–20◦ C/min. A TGA run can include a single temperature program, for example,
heating from 25 to 600◦ C at 10◦ C/min, or a combination of several programs, for
example, heating from 25 to 100◦ C at 100◦ C/min, followed by an isothermal hold
at 100◦ C for 60 min, then heating from 100 to 600◦ C at 10◦ C/min. The measure-
ments are conducted in controlled gaseous atmosphere that can be either static
or dynamic. The most commonly used controlled static atmosphere is a gas at
reduced pressure that can be as low as 10−3 to 10−4 Pa. On the other hand, for
dynamic atmosphere, the measurement can be accomplished by running a gas
around the sample at a certain flow rate (50–100 ml/min). The gas can be either
inert (nitrogen, or more rarely, argon or helium) or reactive (oxygen, hydrogen,
carbon dioxide). The example of measurement that employed the variable heat-
ing rate method is depicted in Figure 11. The figure shows the derivative mass loss
curves for a bilayer polyethylene/nylon-6 film. Although the degradation of each
component in the film (ie, polyethylene and nylon-6) is not well separated, quan-
titative analysis of the components still can be measured from the mass change
between the plateau curves.
4.6. Lifetime Prediction and Degradation Kinetics. In polymer ap-
plications (eg, wire insulation, scaffolds, prosthetics), the ability to estimate the
lifetime of the product is very important. Lifetime prediction is useful in the
 THERMOGRAVIMETRIC ANALYSIS OF POLYMERS 17

development of polymers for long-term usage in various applications. Tradition-

ally, the lifetime of polymers can be estimated through oven aging analysis, which
is time consuming especially for stable polymeric materials. The test typically
takes weeks to months of exposure at elevated temperatures until voltage break-
down occurs. Then several tests need to be carried out at different oven temper-
ature to obtain a semilogarithmic plot versus failure temperature. The obtained
data sometimes do not span the temperature range of interest and require fur-
ther long experiment at a lower temperature to be conducted. This method fits
first-order kinetics, and the lifetime can be estimated using extrapolation.
Nevertheless, the estimation of polymer lifetimes following first-order kinet-
ics is unpredictable because many polymers tend to decompose. The sample usu-
ally has started to decompose before being carried to the desired temperature.
One strategy to overcome this problem is by adjusting a rate-controlling variable
during an experiment. This method is known as a parameter jump method, or
specifically, in this case, it is called temperature jump TG (48). From this practice,
the rate of decomposition can be determined from the change in temperature (T1
or T 2 ) and the Ea can be obtained from equation (11)
( )
⎡ dm ⎤[ ]
⎢ dt T1 ⎥ 1 1
Ea = R ln ⎢ ( ) ⎥ − (11)
⎢ dm ⎥ T1 T2
⎣ dt T2 ⎦

where (dm∕dt)T1 and (dm∕dt)T2 are the rate of weight loss at T1 and T2 .
Degradation kinetics can be analyzed using data from linear rising temper-
ature TG. Several measurements are performed at different heating rates, and
the temperatures at which weight loss happened is observed. Figure 12 shows the

110

100

90
Weight % (%)

80

70

60

50

40
200 300 400 500 600 700 800 900
Temperature (°C)

5°C/min 10°C/min
15°C/min 20°C/min

Fig. 12. TGA curve of PEEK at different heating rates (49).

18 THERMOGRAVIMETRIC ANALYSIS OF POLYMERS

1.4

1.2

log (β)(°C/min)
1

0.8

0.6
1.14 1.16 1.18 1.2 1.22 1.24
1000/T (1/K)
2.5% 5.0% 7.5%
10.0% 12.5% 15.0%

Fig. 13. Kinetic analysis of results obtained from Figure 12 to calculate Ea (49).

1.00E+11

1.00E+09

1.00E+07
Lifetime (years)

1.00E+05

1.00E+03

1.00E+01 1.05 year

1.00E–01

1.00E–03
0 100 200 300 400 500 600
Temperature (°C)

Fig. 14. A predicted lifetime of PEEK versus temperature (49).

example of TG curve for polymer poly(ether ether ketone) (PEEK) at the heating
rates of 5, 10, 15, and 20◦ C/min. The temperature values are then plotted as a
function of logarithms of heating rate 𝛽, and Ea can be calculated (Fig. 13).
Using the Ea obtained, it is possible to estimate the lifetime of the sample as
a function of temperature. The example as displayed in Figure 14 explains that
the time for PEEK decomposition in 5% is about 1.05 year if it is exposed to the
temperature of 350◦ C (49).
There is another method for analyzing sample degradation kinetics called
modulated temperature TG. The temperature profile for this method superim-
poses a sinusoidal temperature fluctuation with a conventional linear rising
temperature program as shown in Figure 15a. The figure illustrates a modulated
temperature TG curves in which the experiment was done on an ethylene-co-vinyl
acetate copolymer. The plot exhibits the temperature modulation (solid time) and
 THERMOGRAVIMETRIC ANALYSIS OF POLYMERS 19

Fig. 15. (a) Modulated temperature TG curve of poly(ethylene-co-vinyl acetate). repre-

sents the temperature curve, whereas —– represents the dm/dT curve. (b) The correspond-
ing plot of weight and Ea versus temperature. (47).

the effect of the modulation on the rate of weight loss (−dm/dt) (dotted line).
These data allow for the determination of Ea , which can be calculated according
to the following equation:
[ ( )2 ]
2 − 0.5T
R Tav L
amp
Ea = (12)
Tamp

where Tav is the average thermodynamic temperature, Tamp is the amplitude of

the temperature modulation, and L is the logarithm of the amplitude of the rate
of the mass loss over one modulation. The calculated Ea is useful to be studied for
plotting another type of data as shown in Figure 15b. Typically, when no mass loss
is observed, the value of Ea is set to zero. However, in this case, the Ea changes
throughout the copolymer mass decomposition.
4.7. Combined Techniques.
4.7.1. Thermogravimetry/Differential Scanning Calorimetry. DSC is a
thermal analysis technique, which provides quantitative and qualitative
information about physicochemical changes in the materials. For example, this
technique can be used to calculate the difference in heat flow between a sample
and inert reference substance when the temperature is varied. Hence, the changes
in heat capacity of the sample, the activity that consumes heat (endothermic), and
release heat (exothermic) can also be measured (50). Other properties that are
mostly measured and studied using the DSC technique are glass transition tem-
perature, the degree of crystallinity, melting/crystallinity (J/g), cross-linking reac-
tions, polymorphic transformation/stability, protein denaturation, decomposition
behavior, thermal/oxidation stability, boiling points, and sample purity. Figure 16
displays the possible transitions in a DSC curve (51).
Figure 17 illustrates the setup of TGA–DSC to measure the heat flux.
Briefly, the sample is placed inside a crucible, which is then placed inside the
measurement cell (furnace) of the DSC system along with a reference hematic
aluminum pan, which is usually empty. The pans are then placed on a metal or
ceramic plate containing thermocouples. Thermocouples are served to detect the
20 THERMOGRAVIMETRIC ANALYSIS OF POLYMERS

exo
Cross-linking
(Cure)

Heatflow

Oxidation
or
Glass *Crystallization decomposition
transition *Polymorphic *Melting
*Polymorphic conversion
conversion *Denaturation
endo

Temperature

Fig. 16. The possible transitions in a DSC curve (51).

Cylindrical
Reference pan Sample pan furnace

Sample

Thermoelectric
disk

Measurement thermocouples

Fig. 17. TGA–DSC illustrates setup.

difference in temperature between these two pans and convert this to a thermal
energy difference (50).
This approach has some advantages over conventional DSC and TG through
the saving in the experimental time needed to acquire two sets of data (52). This
technique is worthwhile to be able to determine energy changes related to ther-
mal decomposition reactions. Most of the chemical reactions in polymer systems,
for example, curing of thermosets are condensation reactions, water, or ammonia
may be released during the process of the reaction. This will result in a change
in mass of the sample in addition to the heat of reaction, which can be measured
by DSC. Normally, the physical changes in polymer structure such as those that
 THERMOGRAVIMETRIC ANALYSIS OF POLYMERS 21

TGA
100

80
Weight loss (%)

60

DSC

40

20
0 100 200 300 400 500 600
Temperature (°C)

Fig. 18. TGA–DSC curve of the BaZrO3 (53).

accompany the glass transition, crystallization, or melting cannot be detected by

TGA, but with the extra information provided by DSC will be useful. Figure 18
shows the example of TGA–DSC graph for zinc oxide (ZnO) nanofibers. Three
weight losses can be observed at the TG curve, whereas the DSC curve exhibits
the first endothermic peak at 105◦ C and three exothermic peaks at 255, 285, and
500◦ C (53).
4.7.2. Thermogravimetry/Evolved Gas Analysis. TGA is capable of mea-
suring the weight variation of a sample against temperature. However, in some
cases, the curves can be complex and especially for multiple decomposition
products. In addition, TGA is only capable of measuring the weight changes, which
are difficult to separate and identify the individual degradation products (off-
gases). Another type of thermal analysis, called evolved gas analysis or EGA, is a
type of thermal analysis in which the amount of gaseous volatile products evolved
by samples is measured as a function of temperature. By coupling EGA with TGA,
the capabilities owned by EGA and TGA can be utilized in one instrument. Two
types of analysis techniques that employ evolved gases are FTIR or mass spec-
trometry (MS) (54–56).
Infrared spectroscopy measures the light absorbed by different types of vi-
brations in molecules. Infrared radiation from the light sources is divided into
two beams by the beam splitter. Each of the beams will be reflected onto a moving
mirror and also onto a stationary mirror, respectively. Both beams are then re-
combined and pass through the sample toward the detector. Fourier transforma-
tion of the resulting interferogram yields a mid-infrared transmission spectrum.
Figure 19 illustrates the schematic diagram of FTIR spectrometry.
22 THERMOGRAVIMETRIC ANALYSIS OF POLYMERS

Detector

Sample

Beam
splitter Moving mirror
Light source

Fixed mirror

Fig. 19. Schematic diagram of FTIR spectrometry.

Detector
Resonant ion
Nonresonant ion

Source

DC and AC voltage

Fig. 20. Schematic diagram of MS.

Figure 20 shows the schematic diagram of MS where it characterizes sub-

stances by identifying and measuring the intensity of molecular fragment ions of
different mass-to-charge ratios (m/z). At first, the incoming gas molecules will be
ionized in the ion source. Then, it will form positive molecular and fragment ions.
Both types of ions are separated corresponding to their m/z values by a combina-
tion of magnetic and electrostatic fields. A mass spectrum is recorded by scanning
the field strength so that ions of increasing m/z ratio arrive at the detector.
When gases are evolved, the FTIR or MS can track their evolution profile.
This is because infrared spectra and mass spectra are substance specific. Thus,
the spectra can be used to characterize the substance or substance class via
spectral interpretation and comparison with database reference spectra. As a
result, the decomposition pathway can be elucidated. This combination technique
is able to use in research and development, in quality control, and to investigate
the failure of the materials; for example, characterization of starting material
 THERMOGRAVIMETRIC ANALYSIS OF POLYMERS 23

Table 1. Advantages of TGA–FTIR and TGA–MS technique

Technique Features Benefits
TGA–FTIR High chemical TGA–FTIR is able to characterize
specificity substances by determining their
and fast compounds and functional groups.
measurement This technique is ideal for the
online measurement of substances
that indicate medium to strong
infrared absorption
TGA–MS High sensitivity TGA–MS is only able to detect
coupled with extremely small amounts of
very fast substances. This technique is ideal
measurement for the online characterization of
all types of volatile compounds

TG analyzer Purge Transfer line Gas analyzer

Purge gas +
gas products
Balance Furnace Heater MS or FTIR

Computer

Fig. 21. Workflow of operating procedure of TGA–FTIR and TGA–MS.

and end products, detection of additives in a matrix, and investigation of the

chemical reaction such as catalysis, synthesis, and polymerization.
Besides the high resolution of TGA–FTIR and TGA–MS, which can interpret
qualitatively the overlapping weight losses, they also offer other specific features
and advantages as shown in the Table 1.
Figure 21 shows the flow of operating procedure of TGA–FTIR and TGA–
MS. The total volume of purge gas and gaseous decomposition products formed
in the TGA can be investigated online by using the combination of TGA and MS
techniques.
In TGA-FTIR technique, the gases are transferred through a heated glass-
coated steel transfer capillary line into a heated gas cell in the FTIR spectrometer.
Usually, nitrogen, which does not exhibit IR-absorption, is used as purge gas. The
FTIR spectrometer determines the spectra of the gases in the gas cell. Figure 22a
shows TGA curves of phenoxy resin at different heating rates under N2 atmo-
sphere. TGA curves show only one step of weight loss. The thermal decomposition
occurs between 325◦ C and 475◦ C depending on heating rate. In Figure 22b, TGA
curves for phenoxy resin in air atmosphere are shown. The TGA curves exhibited
mainly two decomposition steps located between 300◦ C and 450◦ C for the first
step and between 450 and 600◦ C for the second step depending on the heating
rate, being their weight losses around 70 and 30%, respectively. The analysis of
24 THERMOGRAVIMETRIC ANALYSIS OF POLYMERS

(a) (b)

(c) (d)

Fig. 22. (a) TGA curves of phenoxy resin under nitrogen at indicated heating rates.
Dots: selected temperatures for FTIR analysis of residue (b) TGA curves of phenoxy resin
under air atmosphere at different heating rates. Dots: Selected temperatures for FTIR
analysis of residue. (c) FTIR spectra of phenoxy resin at room temperature and solid
residues obtained in nitrogen atmosphere at selected temperatures (see dots in a). (d)
FTIR spectra of phenoxy resin residue obtained in air atmosphere at selected temperatures
(see dots in b) (56).

solid residues formed at various temperatures can also provide direct information
on the changes that occurs in the chemical composition of the polymeric material
during thermal degradation process. In Figure 22c, it is shown the infrared spec-
tra in the range 4000–700 cm−1 of phenoxy resin at room temperature in nitro-
gen atmosphere as well as of the solid residues obtained at higher temperatures
which were selected along the decomposition range observed on TGA curves reg-
istered at 10◦ C min−1 . The selected temperatures are marked with dots in TGA.
Figure 22d shows the FTIR spectra of solid residues obtained at various temper-
atures (marked with dots in Figure 22b for phenoxy resin under air atmosphere.
Observing the evolution of FTIR spectra with temperature it can be clearly seen
the appearance of a new peak at 1730 cm−1 at 404◦ C this peak is attributed to the
presence of carbonyl groups that can appear as consequence of thermooxidative
 THERMOGRAVIMETRIC ANALYSIS OF POLYMERS 25

10–9 m/z

44

10–10
Ion Current / A

10–11
64
48

10–12

96

10–13
0 1000 2000 3000 4000 5000 6000 7000
Time / s

Fig. 23. TGA/MS curve of microcrystalline cellulose modified with (3mercaptopropyl)

trimethoxysilane (57).

reactions undergone by the resin. This band persists in all the spectra of the
residue from 358 to 503◦ C (56).
While, in TGA-MS technique, the TGA is coupled to the MS via a silica cap-
illary tube heated at 200◦ C to prevent condensation. Part of the gases evolved
from the sample in the TGA is sucked into the vacuum in the MS. The MS repeat-
edly measures the whole mass spectrum of monitor m/z values characteristic for
specific structural features. Since MS generally has high sensitivity, only ∼1% of
the effluent gas (argon or nitrogen) is required. The TGA-MS curves of hydropho-
bic polymer matrix and cellulosic fibers modified with silane coupling agents ((3-
mercaptopropyl) trimethoxysilane) is shown in Figure 23. Based on the figure,
the sample shows the generation of furfural (m/z = 96), carbon dioxide (m/z = 44),
thiols (m/z = 64) and sulfur oxide (m/z = 48). These elements demonstrated the
presence of thiol functionalities and thereby prove the formation of solixane layer
onto cellulose surface (57).

5. Conclusion

TGA is a reliable technique that can be used to study the thermal stability
and decomposition of polymeric materials. It determines the rate of change in
weight of the material, whereas the material is being heated or cooled at a
controlled temperature. Specifically, this technique allows for the assessment
of the compositional analysis of the polymer blend. TGA can also be applied to
measure the filler content in polymer systems. Furthermore, TGA can be used for
26 THERMOGRAVIMETRIC ANALYSIS OF POLYMERS

measurement of low level of volatile matter evolved in polymer systems such as

water and residual solvent. The TGA analysis can be conducted under different
heating rate approaches: isothermal or dynamic rate. This analysis is also able to
predict and estimate the lifetime of the materials. The instrument manufactures
have upgraded the fundamental design of TGA instruments by combining with
other techniques such as DSC and EGA in one system. The combination of these
techniques can shorten the experiment duration as it utilizes different analysis
principles during a single measurement.

ACKNOWLEDGMENT

This work was supported by University of Malaya Research Grant (UMRG:

RG382-17AFR).

BIBLIOGRAPHY

1. G. Santhana Krishnan, N. Murali, and A. Jafar Ahamed, J. Therm. Anal. Calorim. 129,
821–832 (2017).
2. C. J. Keattch, Introduction to Thermogravimetry, Heyden, London, 1969.
3. G. Urbain and C. Boulanger, Compt. Rend. 154, 347–349 (1912).
4. K. Honda, Sci. Rep. Tohuku Univ, 97–103 (1915)
5. C. Duval, Inorganic Thermogravimetric Analysis, 2nd ed, Elsevier, Amsterdam,
1963.
6. J. D. Ferry, Viscoelastic Properties of Polymer John Wiley & Sons, Inc., New York,
1980.
7. J. W. Dodd and K. H. Tonge, Thermal Methods: Analytical Chemistry by Open Learning,
John Wiley & Sons, Ltd., Chichester, UK, 1987.
8. M. E. Brown, Introduction to Thermal Analysis, Kuwer Academic Publishers, Dor-
drecht, the Netherlands, 1998.
9. E. L. Charsley, ThermalAnalysis—Technique & Applications, The Royal Society of
Chemistry, Cambridge, UK, 1992.
10. P. J. Haines, in P. J. Haines, ed., Thermal Methods of Analysis. Principles, Applications
and Problems, Blackie, Glasgow, UK, 1995, pp. 22–62.
11. E. A. Turi, ed., Thermal Characterizationof Polymeric Materials, 2nd ed., Academic
Press, San Diego, Calif., 1997.
12. P. K. Gallagher, in M. E. Brown, ed., Handbook of Thermal Analysis and Calorimetry,
Elsevier Science B. V., Amsterdam, 1998, pp. 225–278.
13. L. Cahn and H. E. Shultz, in R. F. Walker, ed., Vacuum Microbalance Techniques,
Plenum Press, New York, 1962, pp. 7–18.
14. S. L. Simon, Thermochim. Acta 374, 55–71 (2001).
15. J. E. K. Schawe, Thermochim. Acta 657, 151–155 (2017).
16. A. Hammer, Thermal Analysis of Polymers: Selected Applications, Mettler Toledo,
Columbus, Ohio, 2013.
17. S. Vyazovkin, in Elton N. Kaufmann, ed., Characterization of Materials, 2nd ed., Vol. 1,
pp. 471–483, 2012.
18. M. E. Brown, Introduction to Thermal Analysis, 2nd ed., Kluwer Academic Publishers,
Dordrecht, the Netherlands, 2001.
19. V. B. F. Mathot, Calorimetry and Thermal Analysis of Polymers, Hanser Publishers,
Munich, West Germany, 1994.
 THERMOGRAVIMETRIC ANALYSIS OF POLYMERS 27

20. T. A. Ozawa, Bull. Chem. Soc. Jpn. 38, 1881–1886, (1965).

21. J. P. McCullough and W. D. Scott, eds., Experimental Thermodynamics, Vol-
ume 1: Calorimetry of Non-Reacting Systems, IUPAC, Plenum Press, New York,
1968.
22. R. S. Milhail and E. Robens, Microstructure and Thermal Analysis of Solid Surfaces
John Wiley & Sons, Inc., New York, 1983.
23. L. E. Nielsen, Mechanical Properties of Polymers, Reinhold, NY. 1962.
24. O. T. Sorensen and J. Rouquerol, Sample Controlled Thermal Analysis: Origin, Goals,
Multiple Forms, Applications and Future, Springer-Verlag, Berlin, 2003.
25. S. Materazzi and S. Vecchio, Appl. Spectr. Rev., 45, 241–273 (2010).
26. D. Han, Y.-H. Zhao, Y.-F. Zhang, and S.-L. Bai, RSC Adv, 5, 94426–94435 (2015).
27. W. Camacho and S. Karlsson, Polym. Degrad. Stab. 78, 385–391 (2002).
28. J. Pospı́šil, Z. Horák, Z. Kruliš, S. Nešpůrek, and S. I. Kuroda, Polym. Degrad. Stab. 65,
405–414 (1999).
29. E. Epacher, J. Tolvéth, K. Stoll, and B. Pukánszky, J. Appl. Polym. Sci. 74, 6, 1596–1605
(1999).
30. H. Hinsken, S. Moss, J. R. Pauquet, and H. Zweifel, Polym. Degrad. Stab. 34, 279–293
(1991).
31. F. Huang, Polymers (Basel) 4, 1012–1024 (2012).
32. J. Wang, H. Jia, Y. Tang, X. Xiong, and L. Ding, Thermochim. Acta 647, 55–61
(2017).
33. C. Delides, in e-RA 11, Athens, Greece, September 21–23, 2016.
34. H. Elias, Plastics, General Survey in Ullmann’s Encyclopedia of Industrial Chemistry,
Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim, Germany, 2005.
35. R. Vaia, DTIC Document, 2002.
36. K. Kalaitzidou, H. Fukushima, and L. T. Drzal, Composites, Part A, 38, 1675–1682
(2007).
37. M. Tait, A. Pegoretti, A. Dorigato, and K. Kalaitzidou, Carbon N. Y. 49, 4280–4290
(2011).
38. M. Rheem, H. Jung, J. Ha, S.-H. Baeck, and S. Shim, Colloid Polym. Sci. 295, 171–180
(2017).
39. T. Doan, H. Brodowsky, and E. Mäder, in M. Poletto, ed., Composites from Renewable
and Sustainable Materials, IntechOpen, London, 2016, pp. 1–24, DOI: 10.5772/65264.
40. N. Fernández-Berridi, M. J. González, A. Mugica, and C. Bernicot, Thermochim. Acta
444, 65–70 (2006).
41. W. J. Sichina, “Characterization of Polymers Using TGA,” 2000.
42. P. Sánchez-Jiménez, L. Pérez-Maqueda, J. Crespo-Amorós, J. Loópez, A. Perejón, and J.
Criado, Anal. Chem. 82, 8875–8880 (2010).
43. A. H. Gerhardt, GXP Compliance 13, 58–66 (2009).
44. G. Dallas, “Taking Thermogravimetric Analysis to a New Level of Performance and
Convenience,” Jan-.
45. M. Reading, in Handbook of Thermal Analysis and Calorimetry Volume 1, Elsevier
Science, New York, 1998, pp. 423–443.
46. S. Sauerbrunn and P. Gill, Am. Lab. (Shelton, Conn.), 26, 29–30,32,34 (1994).
47. D. M. Price and D. J. Hourston, in R. A. Meyers and T. Provder, eds., Encyclope-
dia of Analytical Chemistry, John Wiley & Sons, Inc., Hoboken, N.J., 2006, pp. 8094–
8105.
48. J. H. Flynn and B. Dickens, Thermochim. Acta 15, 1–16 (1976).
49. G. da C. Vasconcelos, R. L. Mazur, and B. Ribeiro, Mater. Res. 17, 227–235
(2014).
50. E. L. Charsley, in E. L. Charsley and S. B. Warrington, eds., Thermal Analysis – Tech-
niques & Applications, The Royal Society of Chemistry, Cambridge, UK, 1992.
28 THERMOGRAVIMETRIC ANALYSIS OF POLYMERS

51. L. C. Thomas, An Introduction to the Techniques of Differential Scanning Calorimetry

(DSC) and Modulated DSC.
52. A. L. C. S. Do Nascimento, J. Anal. Appl. Pyrolysis 119, 242–250 (2016).
53. K. Thangavel, A. Balamurugan, T. Venkatachalam, and E. Ranjith Kumar, Superlat-
tices Microstruct. 90, 45–52 (2016).
54. A. Shiono, A. Hosaka, C. Watanabe, N. Teramae, N. Nemoto, and H. Ohtani, Polym. Test.
42, 54–61 (2015).
55. T. U. Han, Y. M. Kim, A. Watanabe, N. Teramae, Y. K. Park, and S. Kim, Korean J. Chem.
Eng. 34, 1214–1221 (2017).
56. M. A. Corres, M. Zubitur, M. Cortazar, and A. Múgica, J. Anal. Appl. Pyrolysis, 92, 407–
416 (2011).
57. D. Loof, M. Hiller, H. Oschkinat, and K. Koschek, Materials, 9, 415 (2016).

Glossary

A Preexponential factor
dT/dt Heating rate
–dm/dt Rate of weight loss
E Activation energy
m Weight
P Pressure
R Universal gas constant
T Temperature
𝛼 Extent of the conversion
𝛽 Heating rate
𝜅( T ) Temperature dependence of the rate
CRTA Constant rate thermal analysis
CTs Cellulose tri-stearate
DSC Differential scanning calorimetry
DTA Differential thermal analysis
DTG Derivative thermogram
EGA Evolved gas analysis
FTIR Fourier infrared spectrometry
GCC Ground calcium carbonate
HF-PP High-fluidity polypropylene
MCC Microcrystalline cellulose
MS Mass spectrometry
mEBHC Metallocene poly(ethylene-butene-hexene) copolymer
m/z Mass-to-charge ratio
NR Natural rubber
PCC Precipitated calcium carbonate
PEEK Poly (ether ether ketone)
PET Polyethylene terephthalate
PVC Polyvinyl chloride
SBR styrene butadiene rubber
SCTA Sample-controlled thermal analysis
 THERMOGRAVIMETRIC ANALYSIS OF POLYMERS 29

TG Thermogravimetric
TGA Thermogravimetric analysis
VArGO/PVA Vertically aligned reduced graphene oxide/polyvinyl alcohol

H. M. NG
NORSHAHIRAH M. SAIDI
FATIN SAIHA OMAR
K. RAMESH
S. RAMESH
Center of Ionics University of Malaya,
Department of Physics, University of Malaya,
Kuala Lumpur, Malaysia
SHAHID BASHIR
Department of Chemistry, University of Malaya,
Kuala Lumpur, Malaysia

View publication stats