Download as pdf or txt
Download as pdf or txt
You are on page 1of 67

1

Thermal Analysis

Presenter: Dr D. Grooff
2

TABLE OF CONTENTS
1 INTRODUCTION 3
1.1 Definition 3
1.2 Thermal analysis and materials 3
1.3 Thermal analysis equipment4
1.4 Common techniques 5
2 THERMOGRAVIMETRY (TG) 6
2.1 Equipment 6
2.2 Processes amenable to TG 8
2.3 The atmosphere 8
2.4 Sample 10
2.5 Temperature 11
2.6 Rules for using TG analyser 12
2.7 Typical curves 14
2.8 Applications 16
2.9 Controlled rate TG 17
3. DIFFERENTIAL THERMAL ANALYSIS (DTA) AND DIFFERENTIAL SCANNING
CALORIMETRY 20
3.1 DTA 20
3.1.1 Reference materials 21
3.2 Heat flux DSC 22
3.3 Power compensation DSC 23
3.4 Sample containers for DSC 24
3.5 DSC thermal events 25
3.6 Calibration materials 27
3.7 Heating rate 27
3.8 DSC Q100 rules 30
3.9 Applications of DSC 31
4 MODULATED TECHNIQUES 34
4.1 MTDSC 34
4.2 Applications of MTDSC 36
4.2.1 Processes that follow Arrhenius type kinetics 36
4.2.2 Glass transition determination by MTDSC 38
4.2.3 Studies on blend miscibility 42
4.3 Precautions when performing MTDSC experiments 43
4.4 Advantages of MTDSC 43
4.5 Modulated temperature temperature TG 44
5. HEAT FLOW CHARACTERISTICS IN THERMAL ANALYSIS EQUIPMENT 47
6. THE IMPORTANCE OF CALIBRATION OF THERMAL ANALYSIS EQUIPMENT 56
7. THERMAL ANALYSIS EXPERIMENTS 59
7.1 Pharmaceutical crystallization studied with dynamic DSC measurements for estimating Arrhenius
parameters from the Friedman isoconversional method. 59
7.2 Experiment 2: Kinetics of decomposition of calcium oxalate monohydrate, studied by
thermogravimetry 67
3

1. INTRODUCTION

1.1 Definition

Thermal analysis may be defined as the measurement of changes in physical properties of a


substance as a function of temperature, while the substance is subjected to a controlled temperature
programme (Brown, 1985)

Operationally thermal analysis has come to mean the temperature response of small samples in
instrumentation sold by vendors for the express purpose of supplying a set heating programme
(usually an isothermal or constant heating rate experiment) and of measuring a specific physical
property e.g. mass, heat flow or modulus.

1.2 Thermal analysis and materials

Thermal analysis is an ideal technique for the study of materials and composite structures. This is
because:

• the thermal behaviour of a material usually determines the range of operational temperatures
of the material;
• the physical properties measured during a thermal analysis experiment are often directly
related to the performance specifications of that material;
• the heat treatment supplied to a material often determines the properties of a material and this
can accurately be mimicked in thermal analysis equipment;
• the manufacture of materials often involves the judicious supply of energy in the form of heat.

In fact humankind has long been interested in the application of heat to substances to produce new
products. The history of such applications can be traced back nearly 10,000 years to the earliest
forms of pottery. This was followed by the production of metals from ore and later by the production
of glass in the Middle East. The application of heat (via fire) was at the core of the alchemist dreams
of material transmutation. Unfortunately thermal analysis can still not produce gold from base metal.

The theoretical basis of thermal analysis was developed in the 19th century with the introduction of
thermodynamics. It was only at the turn of the 20th century that thermal analysis as we know began
with the introduction of instrumentation to perform controlled experiments. The development of
computers has greatly enhanced thermal analysis and the last 20 years have seen thermal analysis
equipment becoming standard although the level of sophistication is ever increasing.
4

1.3 Thermal analysis equipment

Although many physical properties may be measured according to controlled heating programmes,
all equipment have a number of features in common. The power of computers has meant that the
physical properties recorded need to be converted to an electrical signal by an appropriate transducer.
The figure below illustrates a generalized instrument

Amplifier
Purge gas controllers

Y-signal
Property transducer

Controlled atmosphere

X signal
Sampl
Time, temp e

Computer/controller
device Furnace

Temperature control
circuitry

Figure 1 Generalized thermal analysis instrument

The output of a thermal analysis experiment is usually an X-Y curve where X is usually
time/temperature and Y is the property measured or some function thereof.

Key features in the X-Y curve such as step-discontinuities, peaks, troughs and/or changes in slope
are important for materials characterization. These features are often described as thermal events.

In some instances the instrument measure as absolute value of a sample property such as size or mass
while in other experiments the difference between the property in the sample and a reference material
is important (differential methods). Often the derivative (i.e. rate of change) of the property with
respect to temperature or time is of interest. Such methods are known as derivative methods.
1.4 Common techniques

TABLE I Common techniques

Property Technique Abbreviation


5

Mass Thermogravimetry TG (some TGA for analysis)


Derivative thermogravimetry DTG
Temperature Differential thermal analysis DTA
Heat flow (enthalpy) Differential scanning calorimetry DSC
Mechanical props Thermomechanical analysis TMA
(e.g. modulus) Dynamic mechanical analysis DMA

The above table contains the most common techniques and the are by far the most common
techniques applicable to materials analysis. Nevertheless a wide range of techniques may be applied
and other properties studied include dimensions (dilatometry), optical properties (e.g. birefringence)
and electrical properties (e.g. conductance and dielectric properties).

When the instrumentation is coupled to a device for analyzing any gases produced during the heating
programme, this is known as evolved gas analysis (EGA). Typical detection systems included FTIR
(Fourier Transform Infra-Red) and mass spectrometry (MS). Some techniques which modify the
heating program according to the signal response are known as controlled rate techniques.

Recently many of the above techniques have been modified to allow the superposition of a cyclical
heating programme upon the conventional heating programme (sinusoidal, saw-tooth or step
function). These techniques are known as modulated temperature techniques.
6

2 THERMOGRAVIMETRY (TG)

2.1 Equipment

This is probably the oldest thermal analysis technique where changes in sample mass are measured
using a thermobalance (originally weighing was performed manual on bulk samples). Such a device
is a combination of a microbalance (samples are after all very small) with a programmable furnace.
The balance is typically enclosed in a suitable atmosphere, although vacuum conditions may be
employed. The figure below illustrates a schematic thermobalance.

Electronic
Balance housing weighing
mechanism

Tare

Exhaust Thermocouple

Sample Output device

Furnace
Mass

Purge gas
Temperature
control system Temperature

Figure 2 A schematic thermobalance

Typically the sample is suspended in a cradle into the furnace. All modern instruments may be tared
electronically.

Typically results are reported not as an absolute mass loss/gain but rather as a percentage mass
change. Thermal events are often easier to observe when the derivative thermogravimetric curve
(DTG) is plotted.

It is important not to overload the balance as this may greatly affect sensitivity as well as delay
important thermal events. Balance sensitivity is usually about 1 g (although newer balances have
sensitivity of 0.1 g) with total capacity of around a gram although such large samples are not
recommended.
7

The figure below is a schematic illustration of the TA instruments TGA 500. This is the updated
version of the TGA 2050 currently in the laboratory.

1 Sample pan
2 Thermocouple
3 Tare pan
4 Sample platform
5 Purge gas inlet
6 Purge gas outlet
7 Balance chamber
8 Furnace

Figure 3 Schematic of TA Q500 TG


8

2.2 Processes amenable to TG

A number of processes are amenable to study by thermogravimetry. Some of the typical processes
are detailed below

TABLE II Typical processes studied that can be studied by TG

Process Weight gain Weight loss


Ad- or absorption X
Desorption X
Dehydration/desolvation X
Sublimation X
Vaporization X
Decomposition X
Solid-gas reactions e.g oxidations X X

2.3 The atmosphere

Most thermal analysis experiments are performed in a inert atmosphere. For this purpose high purity
nitrogen is usually used. This is especially important for reactions at high temperature where the
presence of small quantities of oxygen may lead to rapid oxidations.

For studies in nitrogen above 600oC, it is recommended that instrument grade


nitrogen is used. This will require a special cylinder to be supplied by the user.

When oxidative experiments are performed, oxygen or air is typically used. It is very rare in the field
that speciality gases such as H2S are required. Where this is performed special care needs to be taken
so that corrosive gases do not damage the sensitive balance mechanism. Note that such gases may
also be formed in situ.

Care should be taken when performing experiments in vacuum since thermomolecular flow may
result where molecules move in the direction hot to cold, up the suspension wire leading to spurious
results. This can be minimized by rather working in an inert atmosphere or if vacuum is required
performing corrections with an inert sample.
9

Usually experiments are performed in a flowing atmosphere because this prevents condensation on
the cooler parts of the balance mechanism, prevents corrosive gases getting to the mechanism and
cools the balance mechanism thereby improving electronic stability.

We recommend a nitrogen flow rate at 100 mL min-1. This must be adjusted into
a flow of 90 and 10 mL min-1 at the flow regulator. The latter is most important
as it prevents corrosive gases entering the balance mechanism

DO NOT:

Adjust the flow rate at the regulator.

DO:

Close the gas cylinders at the main valve when complete. However, CHECK
that the gas is not being used for a DSC run. If in doubt, leave the nitrogen on.

Most instruments divert some of the purge gas via the balance mechanism so as to protect the
mechanism from corrosive gases formed during decomposition. Where a corrosive gas is used as
purge gas, a separate inert gas stream must be supplied to the balance chamber.

Figure 4 Splitting of gas stream to protect balance in TA Q500 TG


With newer instruments the splitting of the purge gas and gas flow control is performed automatically
10

A good test of the inert atmosphere quality in a thermobalance is to hold a sample of carbon black or
graphite at 1000°C (or as close as the TG will allow), and measure the rate of weight loss. A well-
made instrument, after thorough initial purging with inert gas, should be able to give a result of below
1g/min. For the best results with highly sensitive materials (e.g. finely-powdered metals) it may
well be necessary to deliver the inert gas via metal pipes, and to use an oxygen scrubber.

2.4 Sample

The sample is typically placed in a low mass cradle which is made of an inert material such as
platinum or a high temperature ceramic. The sample should be spread out across the cradle so as to
facilitate the removal of any gases that are evolved. This also ensures that the sample geometry is
uniform between experiments.

Sample size should be small. This is because heat transfer is non-uniform in larger samples and self-
heating or cooling may occur within the sample when reactions occur. The exchange of volatile
products with the atmosphere is also inhibited with larger samples. This all lead to irreproducibility.
Small samples also lessen the risk of explosion and reduce the likelihood of mess..

Do you really know your sample composition before testing?

Unfortunately small samples may be problematic when analysing inhomogeneous material e.g.
recycled waste. In these situations it is nevertheless still recommend to use small samples and to
perform replicate analyses for statistical correctness.

The general rules of sample preparation for thermal analysis:

1. Less is better. TGA sample sizes should be 10-15 mg, no more unless really
necessary
2. Preferably grind fine powders before analysis. This provides reproducibility of
surface area.
3. For the same reason try to spread the sample across the bottom of the cradle.
11

2.5 Temperature

The temperature program is most often a linear increase in temperature, but isothermal studies can
also be carried out, when the changes in sample mass with time are followed. The sample temperature
normally lags slightly behind the furnace temperature. Direct sample temperature measurement is
not possible as it affects the weighing process. This lag can be significant in systems under vacuum
or when high heating rates are employed.

In these cases it is important to calibrate for temperature although this is often ignored by many
practitioners. This can be achieved by performing Curie point measurements on ferromagnetic
materials. At the Curie point, ferromagnetic materials loose there magnetism.

Experimentally one needs to heat a ferromagnetic sample in the furnace. Typical such materials are
nickel (354oC) and iron (780oC) and other alloys supplied by commercial vendors. Once the sample
is loaded a magnet is brought into proximity of the furnace (usually placed below). The magnetic
will apply a magnetic force to the sample. This will be reflected by a mass change (gain/loss
depending on the magnet position). At the Curie point the sample will then appear to lose/gain mass
when this magnetic force is dissipated. Curie point measurements may be performed using multiple
standards in the cradle each providing a characteristic temperature reading. Samples are reusable as
the transition is not accompanied by any real mass change.

Temperature calibration is performed for you. Note that the temperature is the
temperature of the thermocouple which is not in contact with the sample. If the
thermocouple touches the sample, spurious mass readings will be obtained. Make
sure the cradle does not touch the thermocouple and that the thermocouple is 2 mm
above the sample in the TGA 2050.

Because temperature calibration is rather crude in a TGA, there is no point in


quoting temperatures to more than zero decimal places. In fact reproducibility is of
the order of 2oC and because of sample effects may be worse.

Choose your heating rate carefully. For compassion with DSC experiments, heating
rates should be matched.
12

A heating rate of 20OC min-1 is useful for preliminary work. It is not acceptable for
scientific reporting. A heating rate of 10oC min-1 is recommended for standard
experiments. For better resolution slower heating rates are recommended.

There is a trade off between resolution and time taken for analysis.

Some authors recommend suspending inert mass pieces hung by links that melt at characteristic
fusion temperatures. Personal experience, however, has shown that this makes a mess of the furnace.
Another secondary calibration method is to make use of the sublimation kink observed at the melting
point of materials that sublimate. Materials that sublimate typically experience a discontinuous
change in mass loss when volatilization changes from sublimation to vaporization at the melting
point.

Nevertheless even careful calibration and thermocouple positioning cannot avoid self-heating or self-
cooling effects that may result from strongly exothermic or endothermic processes occurring in the
sample.

With newer instruments that offer simultaneous DSC/DTA data, calibration may be performed using
melting phenomenon

2.6 Rules for using the thermogravimetric analyser (TGA 2050)

• Do not exceed the operating temperature of the furnace. This will damage the instrument.
On the TGA 2050 this is 1000oC. Do not exceed 950oC.
• Make sure that the thermocouple is in close proximity to the sample. Failure to do so, will
lead to spurious results. Most systems do not make direct temperature measurements on the
sample. The thermocouple should be about 2 mm above the sample. Report bent
thermocouple wires. Do not replace the thermocouple wire yourself.
• Make sure that the cradle/sample does not touch the side of the furnace. Beware bent hand-
down wires. Report such mishaps. One can move the balance mechanism forward or back
to adjust play in the furnace using the appropriate Allen key. Do not replace the hangdown
wire yourself.
• Make sure that exhaust remains free. Because the exhaust is at a lower temperature than the
sample, material that sublimates may easily reprecipitate here. This problem can be
circumvented by using small sample sizes and by cleaning the furnace regularly. An elevated
temperature ‘burn’ in oxygen may be all that is required. Pipe cleaners are supplied. Do not
wash the furnace with solvent.
13

• Always operate the instrument using the flow rates recommended by the manufacturer. In
many instruments the purge flow is split to provide a protective air flow to the balance
mechanism. This is extremely important where corrosive gases are generated. Too low flow
rates may lead to spurious changes where sample condenses in the cooler part of the furnace.
Too high flow rates may ‘blow’ the sample out of the cradle leading to spurious mass changes.
Use a 90/10 mL min-1 split.
• Do not use speciality gases unless authorised.
• Beware of sample reactions with the cradle. Note that at elevated temperatures Al alloys with
Pt and Al DSC pans should not be used in a Pt cradle above 500oC. Use a ceramic cradle in
that case.
• Avoid large samples. Large samples will take longer to lose mass. Explosive gas build up
may occur within the sample causing sample to jump out of the cradle.
• Beware of sample geometry effects – this can often be overcome by crushing powder samples
in a mortar and pestle before analysis.
• Re-calibrate the system for mass when changing the cradle type or size. Follow the
instructions on screen. This will usually be forewarned by an error message indicating sample
or tare weight over or under range. Note that the ceramic pans are heavier than the aluminium
pans.
• Make sure the instrument is calibrated for temperature. Often ignored but not difficult to
perform.
• Like all thermal analysis techniques it is important to beware of leaks into the gas system.
Polymeric tubing can leak air and may need to be replaced by metal tubing. Do not grease
the lines
• Switch off the gases after use.
• Be very careful if you drop a cradle in the furnace. Call for assistance.
• Clean the cradles before and after use. This can often just be a burn in a Bunsen flame. Note
that acid or base treatment may be necessary but also note the tendency of Pt to dissolve in
extreme solvents.
• Be very careful when loading the cradle not to distort the hangdown wire.
• Do not move the instrument unless authorised.
• Fill in the log book.
14

2.7 Typical curves


Below are some typical curves obtained in thermogravimetic experiments

ii

mass
iii

iv

vi

vi

temperature
Figure 5 Typical TG curves (after Duval)

(i) This indicates that this sample appears to be stable over the temperature range. Such result
are desired when performing thermoxidative stability tests. Note that changes of phase or
reactions where products have the same mass as reactants are still possible.
(ii) This is typical of rapid desorption which is characteristic of drying experiments. Such
curves often indicate wet samples or samples containing high quantities of solvent.
(iii) This indicates a single stage decomposition. The onset of decomposition is often used to
indicate the upper limit of stability of the material. The mass remaining after
decomposition may indicate filler composition with polymeric material although caution
should be exercised as some fillers such CaCO3 may undergo their own decompositions.
(iv) Here decomposition occurs in a series of discrete steps. The intermediates formed are
stable over a limited temperature range.
(v) Decomposition is not discrete and is typical of unstable intermediates. Furthermore such
overlap may also be indicative of a multi-component material whose individual
decomposition processes overlap w.r.t. temperature.
(vi) Here mass increases as a result of interaction with the atmosphere, e.g. water adsorption
or oxidation
15

(vii) Here the product of (vi) decomposes at still higher temperature. Many oxides and sulfides
behave in this manner over the temperature range of interest.

Analysis may be facilitated by using DTG curves. This is especially true of samples that display
curves like (v) above. Below are the TG and DTG curves for calcium oxalate monohydrate which
has been well studied.

Figure 6 TG and DTG curves for the decomposition of calcium oxalate monohydrate

From the mass losses a decomposition scheme can often be determined

CaC2O4.H2O → CaC2O4 → CaCO3

The measured losses above agree well with the theoretical losses.
16

2.8 Applications

Obviously applications are restricted to those that involve mass changes. Not all thermal events,
however, are accompanied by mass changes. Judicious coupling of TG results with results from other
instrumentation may lead to better process understanding.

Common applications include

• Optimization of drying conditions e.g. for hydrated salts


• Determination of thermo-oxidative limits of polymeric materials.
• Corrosion studies: TG provides an excellent means of studying oxidation, or reaction with
other reactive gases or vapours.
• By applying appropriate kinetic analyses, lifetimes of polymeric materials (and others such
as drugs) may be estimated. This is achieved by measuring activation energies from
decomposition rates at elevated temperatures. Decomposition rates are then determined by
extrapolation (e.g. using Arrhenius behaviour).
• Determination of filler content/ash residue. Organic vs. inorganic residues may be separated
by switching to an O2 atmosphere.
• Determination of moisture/solvent content.

Figure 7 Analysis of moisture content of an organic salt hydrate by TG

• Determination of plasticizer content.


• Identification of polymeric composite materials by matching against ‘fingerprint’ curves.
Below are sample curves for a variety of common polymers.
17

Figure 8 Sample TG curves for common polymers

• Polymer identification by coupling with evolved gas analysis. Similar results may be obtained
by pyrolysis gas chromatography (Pyr-GC). New instruments here allow programmed
heating sequences which brings this technique into the family of thermal analysis techniques.
• Determination of vapour pressures of materials.

2.9 Controlled rate TG

In many cases the mass loss profiles of polymers are complex and consist of several overlapping
steps. One way of overcoming this is to decrease the heating rate so as to achieve more complete
separation of each stage of the decomposition process. Such an approach naturally increases the time
for the experiment to be carried out. An alternative approach can be used whereby instead of the
usual linear temperature ramp a strategy is adopted whereby the rate of rise of temperature is slowed
or even suspended as soon as some predetermined rate of mass loss is detected. Once this has
occurred, heating is recommenced until a further mass change is detected. This technique is known
as Sample Controlled Thermogravimetry (SCTG).

In its simplest implementation, the temperature program can be arranged to alternate between linear
heating at a constant rate of temperature rise interspersed by isothermal segments when mass loss is
occurring. This has been termed "step-wise isothermal" heating. The advantage with this approach
is that the temperature profile can be recorded and used to specify conditions for future experiments
on similar samples (for use in quality control). This technique is also useful for evaluation
evaporation rates at different temperatures.
18

Another approach is to use a dynamic heating rate whereby the rate of temperature change is
gradually reduced from an initial rate (possibly even resulting in the sample being cooled) as the rate
of mass loss increases. This approach has been commercialised as "high resolution
thermogravimetry" by TA Instruments (HiRes™ TGA) and by Mettler Toledo as MaxResTM

The curves below show the mass and rate of mass loss of a sample of a laminate of nylon-6 and low
density polyethylene (LDPE). The weight losses of each component are almost completely
overlapping and it is impossible to use the data for a quantitative assay of the sample. The shape of
the DTG curve suggests two components, however.

Figure 9 DTG and TG curves for LDPE/nylon-6 blend

The next set of curves were obtained using a variable heating rate technique. Although there is still
incomplete separation of the degradation of the two components, measurement of the mass change
between the plateaux either side of the saddle in the derivative mass loss curves allows quantitative
analysis of each of the components which agrees well with theoretical calculations based upon the
thickness of each layer. Ancillary experiments on the pure components show that the nylon-6
decomposes before the LDPE.
19

Figure 10 Hi-ResTM TG and DTG curves for the same LDPE/nylon-6 blend

The penalty for this particular approach is the extended experimental time needed to obtain such data
indicated by the curves shown in below comparing the tie taken for each experiment. This is an
extreme case to illustrate the use of sample controlled thermogravimetry. For less overlapping weight
losses often the variable heating rate program can be quicker than a conventional linear ramp and
always with improvements in resolution of weight losses.

Figure 11 Comparison of experiment time for linear and Hi-ResTM experiment


20

3 DIFFERENTIAL THERMAL ANALYSIS (DTA) AND DIFFERENTIAL


SCANNING CALORIMETRY (DSC)

As the names suggest these techniques depend on measuring differences between the sample and an
appropriate reference material.

3.1 DTA

DTA is probably the simplest of all the thermal analysis techniques (TG possibly vies for this title
too). It is also the technique of which DSC grew. As the name suggests the difference in temperature
is measured between the sample and a reference material which are both subjected to the same heating
programme.

In classical DTA, a single block with cavities for sample and reference is heated in a furnace. The
block is chosen to act as a heat-sink and s sample-holder of low thermal conductivity is included
between the block and sample so that an adequate difference in temperature can be measured between
the sample and reference

T

S R

Furnace

Figure 12 Classical DTA apparatus

If an endothermic event (e.g. melting or vaporization) takes place in the sample, the sample
temperature, TS, will lag behind that of the reference, TR. If the difference in temperature (T = TS-
TR) is recorded against the reference temperature (or furnace temperature) a curve such as the one
below would be obtained.
21

Once a thermal event has been completed the thermal conductivity of the system will ensure the lag
between TR and TS is eliminated. The area under the T curve (relative to the baseline) is typically
proportional to the enthalpy of the transition.

T ENDO

TOnset TMax

TR

Figure 13 Typical DTA thermogram

Since the definition of T is arbitrary (it could just as easily be TR-TS), the endo or exothermic
direction should be marked. If exothermic events are studied the curve will move in the opposite
direction.

When reporting such thermal events 2 points are often of interest: TOnset and Tmax. For pure metals
Tonset is often sharp and is best reported. This is the point used for temperature calibration. With
polymeric samples and others which undergo changes over a temperature range Tmax is often reported
for comparative studies. Note that TOnset may still be of interest as it could indicate the temperature
at which a material loses dimensional integrity or becomes processable. If cooling experiments are
performed TOnset will naturally be on the high temperature side.

Usually an inert gas is supplied over the sample and reference although other environments are
possible. The thermocouples are placed inside the samples in classical DTA.

3.1.1 Reference materials

These should undergo no thermal events in the temperature range of interest. That being said, they
must obviously not react with the sample holder (pan). The thermal conductivity and heat capacity
of the reference and sample should be similar. This is often ignored where empty sample holders are
used. Otherwise alumina, Al2O3 or silicone oil are often used as reference materials for inorganic
and organic materials respectively.
22

Improved thermal conductivity and heat capacity matching can be achieved by using the reference
material as a diluent for the sample. This presupposes no reaction between reference and sample.

In many modern DTA instruments thermocouples are no longer directly in the sample, thereby
eliminating reaction with the sample and increasing instrument life.

3.2 Heat flux DSC

Boersma modified the classical DTA by placing the sample and reference in suitable holders (pans)
on thermally conducting bases. Thermocouple junctions are attached to these bases and thus no
longer directly in the sample. The advantage is that the response is less dependant on the thermal
properties of the sample although slower.

This method allows for T measurements which with suitable mathematical manipulation can
provide a measure of the heat flow into or out of the sample. The heat capacities of the components
of the instrument as well as the thermal resistance are necessary. In this form the instrument acts as
a calorimeter and can supply not only the temperature position of thermal events but also the
associated enthalpy. This is useful for purity determinations.

S R

T

Furnace

Figure 14 Heat flux DSC

The sample containers need to make good thermal contact with the cell. The cell is also made of a
material of high thermal conductivity to achieve symmetrical heating of the cell and thus of sample
and reference. The image below is of the inside of the cell in the TA instruments Q100 DSC cell.
23

Figure 15 Inside of Q100 DSC cell

DSC cells are designed that parts that wear out can be replaced. Mettler Toledo have produced a
ceramic sensor for their DSC822e. This has 56 thermocouples for high resolution and sensitivity

Figure 1 Ceramic sensor (inset) of the Mettler 822e

Unfortunately the Q100 DSC requires the entire cell be replaced. This comes at a trade
off, the solid silver block of the cell ensures much better stability and sensitivity, but
means that accidents are significantly more costly. Aply the following rule.

If there is the slightest chance of damage, don’t do it.

3.3 Power compensation DSC

Here the sample and reference material are maintained at the same temperature. The energy
difference supplied to the sample or reference to maintain T = 0 is then recorded.

Although the operational principles of heat-flux DSC and power compensation DSC are different,
history has shown that both types of instruments provide comparable results. Typically most
24

commercial DSC’s are of the heat-flux type in that only one furnace is required which has
implications for cost.

Note that a variety of mathematical expressions are available for determination of heat flows by DSC.
The difference lies in the level of theory used. For accurate comparison of results, the same level of
theory (mathematical algorithm) should be used. Nevertheless for most materials applications low
theoretical levels are suitable.

3.4 Sample containers for DSC

These are typically aluminium pans for applications below 500oC. Above this temperature Al
undergoes softening and its own thermal events which in many DSCs will lead to destruction of the
DSC cell. For most polymeric materials applications, experiments above 500oC are not necessary
(although Au, Pt, graphite, alumina pans may be used in these situations at cost). It is essential to
make sure that there is good contact between cell and sample. This requires flat-based pans.

Since Al can deform easily, watch out for deformed pans. If your pans are deformed
rather re-prepare the sample than waste a useful experiment.

Pans may be used open, closed and hermetically sealed. The choice of pan operation will depend on
whether volatile components are formed. Note that these volatile components (notably HCl) can
destroy DSC cells. Crimping and sealing of pans should always be considered since this process
improves thermal contact between the sample and pan. This is especially important when testing
fibrous materials. Special pans with even better seals may be obtained (at cost) when it is absolutely
essential that volatiles are contained.

Note that hermetic pans will also burst if there is significant pressure build up.

The DSC is not useful for studying evaporations. Use a TGA for that.

Do not run an experiment in a DSC if your TGA trace shows more than a 5% mass
loss. If in doubt, don’t do it.

The following are good practices with respect to pan use:


25

1. The reference pan should be of the same type as the sample plan, i.e use a hermetic
reference with a hermetic sample pan. If the sample is sealed, so should the
reference be.
2. Clean the pans with dichloromethane before use.
3. Do not touch the pans with your fingers.
4. Use the hermetic and standard reference pans provided. This saves cost. Return
them after use.

The purge gas is normally N2 which is inert. DSC may also be operated to low temperatures (-150oC)
where the system is coupled to an appropriate cooling apparatus. The purge gas also serves to
improve thermal conductivity and uniform heating. In certain circumstances high pressure DSCs
may be used with elevated pressures of purge gases. The need for high pressures in the field of
polymeric materials is limited though.

It is often important to determine specific heat flows (i.e. W/g). This necessitates the accurate
weighing of the sample (usually by difference with the pan). For this purpose an accurate
microbalance is required.

It is good practice to weight the sample and pan after an experiment to determine if a
volatile component has been evolved.

3.5 DSC thermal events

These are normally detected by the deviations from a baseline. Because of difference between the
thermal conductivity of reference and sample and mismatches in size between sample and reference
pans it is often difficult to determine baselines. Luckily with modern instrumentation baseline
stability is no longer the fault of the instruments (unless you muck it up).

Modern thermal analysis software provide a number of methods for estimating baseline (linear,
tangential and curve fitting (often sigmoidal) which are used to estimate baselines. Some of these
are illustrated below.

Where the baseline is difficult to establish, these may give significantly different peak areas.
26

Linear

Tangential

Sigmoidal

Figure 2 Some techniques for baseline estimation

The bottom curve above is typical of situations where the heat capacities of the sample before and
post reaction/transition are different.

As with DTA, the DSC curve may be characterized by:

• temperature (onset, end, extrapolated onset and maximum), alternatively time;


• size (energy required for the transition);
• peak shape (the latter is useful for interpreting the kinetics of reaction/process).

The table below details some processes that give enthalpic responses (peaks) which are easily studied
by DSC
27

TABLE III Enthalpic processes that may be studied by DSC

Process Exotherm Endotherm


Solid-state transition X X
Crystallization X
Melting X
Vaporization X
Sublimation X
Adsorption X
Desorption X
Desolvation/Drying X
Decomposition X X
Solid/solid(liquid,gas) reaction X X
Curing X
Polymerization X
Catalytic reactions X

The DSC may also be used to determine second order transitions. These are indicated by abrupt
changes in slope. Examples include Curie point transitions and the glass transition (Tg) of polymers.
Here there is no enthalpy change, H, but there is a change in heat capacity.

Where peak areas are measured, suitable calibration with an appropriate material, allows for the
determination of enthalpy changes, H, associated with a reaction or process e.g. a melt. This is
performed using the analysis software.

3.6 Calibration materials

A variety of materials are commonly used for calibrating DSC cells for temperature and enthalpy. It
is essential that these substances be pure to provide sharp melting endotherms. The melting point Tm
is conventionally obtained by extrapolation of the sharp change in slope to the baseline.

Impure substances have broader melting endotherms as well as experiencing changes in Tm. This,
however, can be used to estimate purity.
28

TABLE IV Recommended inorganic materials used for temperature calibration

Substance Tm (/oC) Substance Tm (/oC)


NH4H2PO4* -121.4 SiO2 (quartz form) 573
NH4SO4* -48.8 K2SO4 583
KNO3 127.7 K2CrO4 665
KClO4 299.5 BaCO3 810
Ag2SO4 412 SrCO3 925

* These are in fact solid-solid transitions that are useful for low temperature calibration

Often pure metals are used for calibration. Unfortunately many alloy with Al and so other pans
should be considered, or Al pans (and lids) should be annealed above 400oC in air before use.

TABLE V Pure metals recommended for enthalpy and temperature calibration

Metal Melting point (/oC) Hfus(/J g-1)


Hg -38.83 11.47
Ga 29.76 79.88
In 156.60 28.62
Sn 231.30 7.17
Bi 271.40 53.83
Pb 327.46 23.00
Zn 419.53 108.6
Al 660.32 398.1

Unfortunately most of these transitions are at elevated temperatures. For calibration at low
temperatures a variety of organic materials may also be used.
29

TABLE VI Organic materials for low temperature calibration (melting points)

Substance Tm (/oC) Substance Tm (/oC)


Toluene -95.01 Biphenyl 68.93
Cyclohexane (solid-solid) -56.8 Napthalene 80.23
Octane -56.8 Acetanilide 114.34
Decane -29.6 Benzoic Acid 122.34
Dodecane -9.6 Anisic Acid 183.28
Water 0.00 2-Chloroanthraquinone 209.83
Cyclohexane (solid-liquid) 6.7 Caffeine 236.1
Diphenyl Ether 26.87 Carbazole 284.52
4-Nitrotoluene 51.61

Vaporizations do not make suitable transitions for calibration because evaporation begins below the
boiling point especially if the pan is open. For sharp vaporizations, pressure DSC cells are required
such as the Mettler Toledo DSC 827e below

Figure 3 Mettler Toledo Pressure DSC 827e

DSCs can also be used to determine heat capacities. The introduction of a sample causing a shift in
the baseline in the endothermic direction proportional to the heat capacity of the sample. The
proportionality constant may be determined by using an appropriate standard material, usually
sapphire whose heat capacities have been well determined, scanned under similar conditions to the
sample.

Do not re-calibrate the instrument. If you suspect there is a problem report it.
30

Note that if you want low temperature calibration it is performed not as a freezing
experiment but rather as a melt to prevent super-cooling effects.

3.7 Heating rate

Note that slow heating rates are not necessarily best. Slow heating rates are good for
resolution but come at the expense of reduced sensitivity. If in doubt use a heating
rate of 5oC min-1.

3.8 Rules for using the DSC Q100

• Make sure that the instrumentation is properly calibrated. Do not use high temperature
calibration points for low temperature readings. Make sure that calibration data occur on
either side of thermal events of interest. Also avoid only using a single temperature
calibration point. If you suspect a problem, report it
• Beware super-cooling when calibrating low temperatures. Rather freeze the samples and use
the melts.
• Only use pure calibration standards. Impure standards will have lower melting temperatures.
Do not tamper with our calibration standards
• The reference material is often an empty pan. Beware of matching reference pans carefully
to those used for the sample. If the sample is closed, so should the reference be. Poor
matching of sample and reference pans can lead to drifting baselines. It is also important to
make sure that pans make good contact with the cell, i.e. have a flat base.
• Do not exceed the temperature range of the DSC cell. This will damage the cell. Note that
DSC cells have much lower operational ranges than TG instruments. Do not run experiments
above 300oC without permission.
• Make sure that the cell is clean. If it is not in a clan state, report it so we can perform an
oxygen clean. Do not perform this yourself.
• Beware spurious results when gaseous products are formed in sealed pans. Do not study
evaporations in our DSC.
• Where analysis depends on enthalpy measurements, the instrument must also be calibrated
for heat flow.
• Be very careful when performing analysis on halogen containing materials. HCl may be
evolved and this can destroy instrumentation.
31

• Make sure that for comparative analysis, consistent baseline techniques are used. Beware of
uncertainties in peak area measurements.
• Where calibration materials are stored for repeated use, watch out for contamination. This is
especially true of the adsorption of water by inorganic materials. Also beware alloying of the
standard with Al pans.
• Beware spurious results from interactions between sample and pan. Al is susceptible to attack
by acid (passivity is effectively removed at higher temperatures). In these cases other pans
need to be considered. Note also that a material itself may not be acidic but its decomposition
products might be.
• Do not scratch the surface of the cell with any sharp instrument.
• Make sure the LNCS level is above 20% before beginning an experiment.
• Remove your samples after use.
• Fill in the log book.

3.9 Applications of DSC

Common applications include:

• The determination of glass transition temperatures (Tg) for polymers. This temperature is
useful for predicting where a material will become brittle. Note that for many polymeric
samples this will necessitate cooling. These temperature measurements may be used to
determine the effectiveness of a plasticizer or its content.
• The determination of blend components, based on Tg measurements.
• DSC may be used to ‘fingerprint’ materials. Note, however, that the transition temperatures
of pure materials are more susceptible to the presence of impurities than TG thermal events.
Such ‘fingerprinting’ should, only be used on pure materials or where the thermal event is a
strong indicator e.g. the cold crystallization of PET.
• With 100% crystalline materials, peak area may be used to estimate material composition.
• The determination of sample purity (usually > 98%). This is a routine procedure for
pharmaceuticals and fine chemicals.
• The determination of key phase transitions. Melting temperatures may be useful to indicate
processing temperatures in a variety of applications.
• The determination of percentage crystallinity of polymers. This may be useful for predicting
ultimate properties. Allied with this DSC allows for investigations into the effect of thermal
history on polymers. Various heating/cooling regimes can be mimicked accurately in the
DSC. The figure below illustrates typical events observed for polymeric samples. Note that
not all events will be observed with all polymers.
32

Crystallization/Curing

Glass transition Oxidation

W/g

Degradation

Melting
ENDO

Tg

Temperature

Figure 4 Typical DSC thermogram for an organic polymer

By use of an appropriate models DSC may also be used to estimate rates of crystallization
which can be used to inform processing operations.
• DSC may be used also to determine oxidation stabilities (oxidation induction time (OIT),
ASTM Ds895) and lifetimes in the isothermal mode. This is analogous the results obtained
from TG experiments. The relative oxidative stability of materials is an important area of
application, and can be used to determine the effectiveness of an antioxidant, or assess the
amount present in e.g. oils, fats, waxes and polymers. The preferred method consists of
raising the sample temperature to a pre-determined level, while in an inert atmosphere, then
switching the atmosphere to air or oxygen. The time to the onset of exothermic reaction is
measured. The experiment can be carried out under pressure (up to 100 bar) to avoid
volatilisation of some materials in an appropriate DSC apparatus. The figure below illustrates
one such OIT experiment.
33

Figure 5 Determination of OIT

• The curing rates (and temperatures) of epoxy resins and other curable polymers. The figure
below illustrates typical features for an epoxy cure

Curing

First scan
W/g

Second scan

ENDO

Tg
Temperature

Figure 6 Typical DSC thermograms for epoxy cures

• The identification of liquid crystalline transition temperatures.


• Construction of phase diagrams.
• Because of the small samples used in DSC, potentially hazardous materials/processes can be
identified. This is usually based on the observation of an unwanted exotherm.
• The identification of typical polymers in waste materials based on their melting temperatures.
The use of heat capacity measurements to predict thermal conductivities.
34

4 MODULATED TECHNIQUES
4.1 Modulated temperature DSC (MTDSC)

Modulated Temperature DSC is a variant of conventional DSC in which the conventional linear
heating program has a modulation superimposed upon it. The modulation is typically sinusoidal but
can be a combination of sine waves (which, in certain combinations, will result in a square wave or
a saw-tooth configuration). This modulation in heating rate results in a corresponding modulation in
heat flow. The response of the sample can then be considered to be made up of two components, the
response to the underlying linear heating rate and the response to the modulation. These two
components can separated by use of an averaging procedure combined with a Fourier transform.

The result is three signals, the underlying or average response that is equivalent to a conventional
DSC at the same underlying heating rate, the amplitude of the modulation and the phase lag. The
phase lag can be used to separate the response to the modulation into an in and out of phase
component if desired.

Figure 22 Undeconvoluted heat flow signals

Above are the undeconvoluted signals for a PET samples. Careful observation will reveal the glass
transition temperature, cold crystallization and melting.
35

For transitions other than melting, the modulation enables the reversing heat capacity to be measured
which can be transformed into a reversing heat flow by multiplying it by the average heat rate. The
term reversing means reversible at the time, temperature and frequency at which the measurement is
made. By subtracting the reversing heat flow from the underlying heat flow the non-reversing heat
flow can be determined (non-reversing meaning not reversible at the time, temperature and frequency
at which the measurement is made). In this way enthalpies of reactions can be separated from heat
capacity changes occurring at the same time. Also enthalpy relaxation at a glass transitions can be
separated, with due allowance for the effective frequencies of the measurements, from the change in
heat capacity at the glass transition. This means that glass transitions that are obscured by enthalpy
relaxation effects or by the cold crystallization of other components of the sample can be measured.
This is not the case with conventional DSC. As with conventional DSC, heat capacity measurements
are also possible and these can be separated into reversing and non-reversing components.

The terms non-reversing and reversing were coined to denote that at the time and temperature that
measurements are made the processes observed are not reversing but might be reversible. Similarly,
at the glass transition, the reversing signal does not include all the enthalpy that is reversible because
this is frequency dependent, thus it comprises only that part that is reversing at the chosen modulation
frequency. For this reason the term reversing is preferred over reversible.

Figure 23 Deconvoluted signals for PET thermogram presented earlier


In the diagram above, the four deconvoluted signals from the data presented earlier are given, in this
case expressed as heat capacities. Note that the glass transition appears in all four signals, the cold
36

crystallisation appears in all except the reversing signal, while the melt again features in all four
signals.

4.2 Applications of MTDSC


4.2.1 Processes that follow Arrhenius type kinetics e.g. curing of epoxy resins

The Arrhenius equation is generally of the form

d − Ea
= f ( ) A e RT
dt

where  = the extent of the reaction, f() = some function of extent of reaction, A = the pre-
exponential constant, Ea = the activation energy, R = the gas constant, T = absolute temperature.

This type of behaviour is typical for thermally activated processes. In this model, a material changes
from one form to another more thermodynamically stable form but must first overcome an energy
barrier that requires an increase in potential energy. Only a certain fraction of the population of
reactant molecules have sufficient energy to do this and the extent of this fraction and the total number
of reactant molecules determine the speed at which the transformation occurs. The fraction of
molecules with sufficient energy is dependent upon the temperature in a way given by the form of
the Arrhenius equation, thus this must also be true for the transformation rate. The types of process
that can be modelled using this type of expression include chemical reactions, diffusion controlled
processes such as the desorption of a vapour from a solid and some phase changes such as
crystallisation.

By suitable mathematical manipulation it is possible to separate the signals as before into reversing
and non-reversing components. In particular we able to decouple heat capacity from the enthalpy of
chemical reactions. This extremely is useful for studying cross-linking reactions such as epoxy cures
where the crosslinking reaction causes a change in heat capacity.

Typical results from this type of system are shown below.


37

Figure 24 Illustration of separation of curing and vitrification

The non-reversing signal shows the progress of the chemical reaction while the complex heat capacity
shows a glass transition as the reaction reaches a certain stage. This occurs because the Tg of the
system increases as the cross-link density increases until it equals the cure temperature. When this
happened the mobility of the system decreases dramatically and consequently the reaction rate
decreases as it becomes diffusion controlled. This is seen as a decrease in the vibrational heat
capacity and the system is said to have vitrified. As the temperature increases (because this is a rising
temperature experiment) this tends to increase mobility thus increasing reaction rate thereby
increasing the glass transition temperature. As this process continues it gives rise to a very extended
period over which the reaction proceeds very slowly under quasi-diffusion control. This effect can
be very important in determining the cure behaviour of a wide range of materials. Conventional DSC
does not allow this phenomenon to be studied in any detail. Furthermore the enthalpies of curing are
able to be determined much more accurately.
38

4.2.2 Glass transition determination by MTDSC

Glass transition phenomena have been observed to occur gradually and may occur over 10oC or more.
The position of the Tg also varies with heating rate (and with frequency in MTDSC) which reveals
that it is a kinetic phenomenon. The co-operative motions that enable large-scale movement in
polymers have an activation energy in a way that is similar to (but not the same as) the energy barrier
model discussed above for Arrhenius processes. Thus, as the temperature is decreased, they become
slower until they appear frozen. There is some part of the heat capacity that is associated with these
motions therefore, as they freeze, these large-scale motions are no longer possible and consequently
the material appears glassy (rigid) and the heat capacity decreases. In reality, whether a polymer
appears glassy depends on how rapidly the observer is attempting to deform it. Thus, if the polymer
is being bent at a frequency of several times a minute, it may be springy and return to its original
shape when let go. If it being bent at several times a year it may well behave like a pliable material
that retains the shape it is given by the deformation when it is released. There is a parallel dependence
of the heat capacity on how rapidly one is attempting to put heat into or take it out of the sample, thus
the position of the Tg changes with heating and cooling rates.

Figure 25 Illustration of non-equilibrium states produced in polymers on cooling

The diagram above gives the enthalpy diagram for glass formation. The enthalpy gained or lost by a
sample is determined by integrating the area under heat flow curve. Above the T g the sample is in
equilibrium (provided no other processes such as crystallization are occurring). Consequently, this
line is fixed regardless of the thermal treatment of the sample and a given temperature corresponds
to a unique enthalpy stored within the sample. As the sample is cooled there comes a point at which
39

the Cp changes as it goes through the glass transition, thus heat capacity changes and so does the
slope of the enthalpy line. At different cooling rates the temperature at which this happens changes,
thus a different glass with a different enthalpy is created. Above the transition the sample is at
equilibrium, below it is at some distance from this equilibrium line but is moving toward it very
slowly, thus glasses are generally meta-stable. If the glass is annealed at temperatures near the glass
transition then it looses enthalpy relatively rapidly and becomes a different glass as it moves toward
the equilibrium line. At temperatures far below the Tg the rate of enthalpy loss becomes very slow
and effectively falls to zero. In the diagram below, the results on heating are shown for a sample
cooled at the same rate it was heated and a sample cooled much slower than it was heated. The
characteristic relaxation peak is seen in the slow cooled, fast heated example. The same peak is
observed for samples annealed so that they experience an enthalpy loss.

Figure 26 Enthalpic relaxation experienced on heating polymeric samples


40

Figure 27 Effect of annealing on DSC signals

The figure above shows typical results for a glass transition for a sample that has been annealed for
different lengths of time. It can be seen that, as expected, the total signal is the same as that observed
for a conventional DSC experiment. As annealing increases, the characteristic endothermic peak at
the glass transition increases. In contrast, the reversing and kinetic signals are largely unaffected by
annealing thus the non-reversing signal shows an increasing peak with annealing time. The use of
MTDSC then seems to eliminate the influence of annealing and enables the relaxation endotherm to
be separated from the glass transition itself. To a first approximation this is true but this must be
understood within the context of the frequency dependence of the glass transition. It is well know
that the temperature of the glass transition is frequency dependent from measurements made with
DMA and DEA. This same frequency dependence is seen in MTDSC. The plots below show the
results for polystyrene at a variety of frequencies. The cooling rate dependence for glass transition
temperatures measured with conventional DSC is also well established, as explained above.
41

Figure 28 Effect of modulation frequency on Tg

When we consider a cooling experiment with MTDSC we have both a cooling rate, , and a frequency
(the frequency of the modulation, ). The result seen in the reversing signal is largely independent of
the cooling rate once it is slow enough to ensure many modulations over the transition region (where
there are not many modulations the result is meaningless). This is extremely important.

Thus, as we keep the frequency of the modulation constant and vary the underlying cooling rate, the
Tg seen in the average signal changes while the Tg seen in the reversing signal remains the same.
Similarly, as shown above, if we keep the cooling rate the same and vary the frequency, the
underlying signal remains constant while the reversing signal changes. In an MTDSC experiment,
the average signal will always give a lower Tg than the reversing signal because the underlying
measurement must, in some sense, be slower (i.e. on a longer time scale) than the reversing
measurement. This is because of the requirement that there be many modulations over the course of
the transition. As the cooling rates become slower, in other words as the time scale of the
measurement becomes longer, the Tg moves to a lower temperature. Similarly, as the frequency
decreases, the Tg moves to lower temperatures. As a consequence of this there is a peak in the non-
reversing signal as the sample is cooled that is clearly not related to annealing but is a consequence
the difference in effective frequency between the average measurement and that of the modulation.

On heating the non-reversing signal, is related to the amount of annealing and also must contain the
effects of the different effective frequencies used in the measurement. We can consider these effects
to be additive thus the non-reversing signal gives a measure of the enthalpy loss on annealing with
an offset due to the frequency difference. In a sense it is non-reversing in the same way that a
42

chemical reaction or crystallization is. This simple picture is only a first approximation but it will be
adequate in many cases. In particular the non-reversing peak at the glass transition can be used to
rank systems in terms of degree of annealing.

4.2.3 Studies on blend miscibility

MTDSC provides a method for studying the structure of polymer blends. When two polymers are
blended but do not mix at the molecular level, the blend will have two distinct glass transitions that
are the same temperature and shape as they would be in the pure materials. When they mix
completely, a single glass transition will be seen at a temperature intermediate between the
temperatures of the two constituents. When, as is most often the case, there is partial mixing the
glass transition temperatures tend to shift towards each other. This can be characterised using
MTDSC. This is illustrated for a polystyrene polyurethane interpenetrating network (IPN).

Figure 29 Effect of miscibility on Tg

The curves above are in fact derivatives of the heat capacity. Glass transitions thus appear as peaks
in the curve. The pure materials can be seen to be completely different from the blend. The
processing has increased the temperature of the glass transition for the polystyrene while, more
typically, increasing that for the PU due to intermixing with the higher Tg phases. The complex
structure of the blend can best be modelled assuming 4 Gaussian peaks corresponding to 4 phases.
This type of detailed structural information can be difficult to obtain by other means.
43

4.3 Precautions when performing MTDSC experiments

All the precautionary notes about DSC experiments apply. Furthermore:

• The frequency of modulation must be such that many modulations occur over the region of
interest.
• Do not use too high a heating rate. Use 5oC min-1 or less.
• Make sure that the amplitude and frequency is such that a negative instantaneous heating rate
is not achieved.

4.4 Advantages of MTDSC

These include

• The separation of events that might obscure each other. Typically enthalpic relaxation (the
loss of stresses as a result of thermal history) at the Tg can often obscure the glass transition
measurement by DSC. With conventional DSC this requires a heating programme which first
releases these stresses and then a measurement of the Tg in a second scan. MTDSC can arrive
at this result in a single scan. Conventional DSC programmes may also cause
curing/degradative processes which can affect Tg. The fact that the reversing signal is
approximately independent of thermal history also simplifies the detection and quantification
of different phases.
• The step-change in heat capacity at the glass transition may be used to quantify amorphous
phases.
• Glass transitions obtained from reversing signals correspond better with DMA results than
conventional DSC.
• In curing samples the progress of the cure reaction can be separated from the progress of
vitrification. This allows for much more accurate determination of enthalpies of reactions.
• Much greater sensitivity than conventional techniques. The glass transitions of minor
components can be accurately measured where they would be lost in conventional DSC. The
limit of detection is thus lowered. When studying amorphous blends, the reversing signal has
a much higher signal to noise and greater resolution that conventional DSC to the glass
transition. The figure below illustrates the detection of the glass transition in the reversing
signal for ABS in an ABS/PET blend. In ordinary DSC the ABS Tg would be obscured by
the PET cold crystallization
44

Figure 30 MTDSC curves showing Tgs in a PET/ABS blend

• More accurate values for polymer crystallinity can be obtained. It is possible for
rearrangement and recrystallization to occur during melting. The reversing signal is sensitive
to these processes that are largely undetected by conventional DSC. This fact assists in
measuring initial crystallinity correctly since reversible melting can be detected.

4.5 Modulated temperature TG

Modulated Temperature Thermogravimetry is a way of obtaining information about the kinetics of


thermal degradation processes.

The temperature dependence of chemical processes may be readily expressed in terms of the
Arrhenius equation:

− Ea
RT
k=A e

where k is the rate constant, R the gas constant and T the absolute temperature. Values of the
Arrhenius parameters (Ea and A) provide measures of the magnitude of the energy barrier to reaction
(the activation energy, Ea) and the frequency of the occurrence of a condition that may lead to
reaction (the frequency factor, A). The rate constant k is defined by the relationship between the
rate of reaction (d/dt) and the extent of conversion or fraction reacted (). Polymer decompositions
are generally heterogeneous reactions since the sample is solid (or molten) and the products are gases.
45

A useful method of determining Ea is temperature jump thermogravimetry. The rate of


decomposition is measured either side of a change in temperature and the activation energy
determined by:

d m d t (T1 )  1 1
Ea = R ln  − 
d m d t (T2 )  T1 T2 

Where dm/dt(T1) and dm/dt(T2) are the rates of mass loss at T1 and T2 either side of the temperature
jump.

Another approach is to studying degradation kinetics using data from conventional linear rising
temperature thermogravimetry. Many such methods have been proposed, but the most popular
strategy is that described by Ozawa, Flynn and Wall which has been incorporated into an ASTM
standard method. Essentially, separate measurements are carried out at different linear heating rates
and the temperatures at which a set percentage mass loss occurs noted. These are then plotted as a
function of heating rate (=dT/dt) and the activation energy determined by an iterative process. These
algorithms have been incorporated into a number of commercially available software packages,
although the user should always question the predictions of such "black box" methods especially
since they often assume, without appropriate justification, that the polymer decomposition reactions
follow a particular reaction mechanism e.g. first order

Modulated temperature thermogravimetry (MT-TG) employs a temperature profile in which a


sinusoidal temperature fluctuation is superimposed upon a conventional linear rising temperature
program. The raw data from such an experiment are shown below for an ethylene-co-vinyl acetate
copolymer. The curves showing the heating rate () and rate of mass loss (-dm/dt) make the effect
of the temperature modulations apparent.
46

Figure 31 Raw MT-TG signals

Ea for the chemical decomposition is then calculated according to:

2
Tamp
Tav −
2

Ea = R 4 f ( m)
Tamp

where Tav. is the average temperature, Tamp. is the amplitude of the temperature modulation and f(m)
is the logarithm of the amplitude of the rate of mass loss over one modulation. Plots of mass and Ea
as a function of temperature from the data are shown below.
47

Figure 32 Activation energies obtained from MT-TG

5. HEAT FLOW CHARACTERISTICS IN THERMAL ANALYSIS EQUIPMENT

The following terms (see figure below) apply to heat flow in the sample and reference materials in a
calorimeter.
Th = temperature of the heat source
Tsm = temperature indicated by measuring thermocouple on sample side
Trm = temperature indicated by measuring thermocouple on reference side
Ts = actual temperature of the sample and its pan
Tr = actual temperature of the reference pan

Rs = thermal resistance to heat flow between Th and Tsm


R’s = thermal resistance to heat flow between Tsm and Ts
Rr = thermal resistance to heat flow between Th ans Trm
R’r = thermal resistance to heat flow between Trm and Tr

Cs = total heat capacity of sample and its pan


Cr = total heat capacity of reference pan
48

Figure 33: General considerations for almost any type of thermal analysis instrument.

The heat capacities and thermal resistances are considered constant which is a valid approximation
over a narrow temperature range in the determination of , for example solidus or liquidus
temperatures. The thermal resistances on both the sample and reference sides can be considered
equal in terms of Rs = Rr and R’s = R’r. It should be noted that thermal resistance between heat
source and thermocouple will differ from thermal resistance between thermocouple and pan system
for both sample and reference (R = R’). If heat flow is assumed to follow a simple one dimensional
law:
𝑑𝑞
=∝ ∆𝑇
𝑑𝑡
1
dq/dt represents the rate of heat exchange between two bodies that differ in temperature by ∆T. The
heat transfer coefficient α can be equated to 1/R. The heat exchanges for the respective sample and
reference sides can be further examined by assuming that heat flow from the heat source (at
temperature Th) to the thermocouple (Tsm or Trm) equals heat flow from thermocouples (Tsm or
Trm) to the pan system (Ts or Tr). This assumption applies to conditions where there is no other heat
transfer processes in the system.
The following equation can be written for a sample subjected to a linear heating rate (dTs/dt) in the
sample compartment:
49

𝑑𝑞𝑠 1 1
= (𝑇ℎ − 𝑇𝑠𝑚 ) = ′ (𝑇𝑠𝑚 − 𝑇𝑠 )
𝑑𝑡 𝑅𝑠 𝑅𝑠
2
And
𝑑𝑞𝑠 𝑑𝑇𝑠
= 𝐶𝑠
𝑑𝑡 𝑑𝑡
3
Similarly for the reference experiencing a linear heating rate (dTr/dt):

𝑑𝑞𝑟 1 1
= (𝑇ℎ − 𝑇𝑟𝑚 ) = ′ (𝑇𝑟𝑚 − 𝑇𝑟 )
𝑑𝑡 𝑅𝑟 𝑅𝑟
4
And
𝑑𝑞𝑟 𝑑𝑇𝑟
= 𝐶𝑟
𝑑𝑡 𝑑𝑡
5
The abovementioned model when applied to classical DTA, heat flux and power compensation
DSC will be adjusted to account for differences in their design features.
In classical DTA (see figure below), the thermocouples are embedded within the sample.
50

Figure 34: Schematic diagram of furnace of DTA, where both sample and standard are heated by
central furnace.

This implies that the thermocouple temperatures (Tsm, Trm) can be equated to the temperatures in
the pan systems (Ts, Tr), i.e. Tsm = Ts and Trm = Tr. The thermal resistance (R’) can therefore be
equated to zero.

Subtracting equation 4 from equation 2 and assuming that Rs = Rr = R, an equation for the
measured temperature difference (Tr-Ts) can be derived.

𝑑𝑞𝑠 𝑑𝑞𝑟
𝑇𝑟 − 𝑇𝑠 = 𝑅 ( − )
𝑑𝑡 𝑑𝑡
6
By combining equations 3, 5 and 6 the measured temperature difference can be expressed in terms
of the sample and reference heat capacities (Cs and Cr) and the heating rate (dT/dt):
𝑑𝑇𝑠 𝑑𝑇𝑟
𝑇𝑟 − 𝑇𝑠 = 𝑅 (𝐶𝑠 − 𝐶𝑟 )
𝑑𝑡 𝑑𝑡
7
51

During a programmed temperature scan at a rate of dT/dt, is steady state is attained at which the
sample and reference temperatures are changing at the same rate (dTs/dt = dTr/dt = dT/dt).
𝑑𝑇
𝑇𝑟 − 𝑇𝑠 = 𝑅 (𝐶 − 𝐶𝑟 )
𝑑𝑡 𝑠
8
The measured signal is thus seen to be proportional to the difference in heat capacities, but the value
of R is necessary to calculate the heat capacity of the sample. Since the thermocouples are embedded
in the sample, R will be influenced by sample characteristics such as thermal conductivity. This
complicates quantitative measurements (example measurement of sample heat capacity) as this
implies in principle further calibrations for temperature ranges of studies and sample configurations.
Power compensation DSC utilizes separate heaters for the sample and reference, each with an
associated thermometer (see figure 35).

Figure 35: Schematic layout of furnace of power compensation DSC (left) and ‘Boersma’ DTA
(right), also known as heat-flux DSC.

For the power compensation design, a simple assumption can be made that the thermal resistance R
(Rs and Rr) be equated to zero (Rs = Rr = 0). The temperature-measuring system controls power to
the two heaters to maintain the sample and reference at the same temperature i.e Tsm = Trm. The
measured signal in this method is the difference in power between the sample and reference sides.
Thus, again assuming that steady-state heating rates have been attained i.e heating rates at sample
and reference sides are identical (dTs/dt = dTr/dt = dT/dt):

𝑑𝑞𝑠 𝑑𝑞𝑟 𝑑𝑇
− = (𝐶 − 𝐶𝑟 )
𝑑𝑡 𝑑𝑡 𝑑𝑡 𝑠
9
52

The measured signal is again proportional to the difference in heat capacity, but the proportionality
constant only includes the operator –selected heating rate (dT/dt) and not a heat transfer coefficient.
Quantitative work is therefore considerably more reliable.
The same set of equations can be applied to heat-flux DSC (illustration above). Since the measuring
thermocouples are mounted on a plate beneath the sample and reference pans, it can be assumed that
the temperatures in the pans (Ts and Tr) are different from the thermocouple temperatures (Tsm and
Trm). The equation for the measured signal :

𝑑𝑇
𝑇𝑟𝑚 − 𝑇𝑠𝑚 = 𝑅 (𝐶 − 𝐶𝑟 )
𝑑𝑡 𝑠
10
R is an instrument constant and independent of the nature of the sample. Indeed, some
manufacturers include the calibration in the electronics of the instrument so that the user obtains a
differential power output.
In both power-compensated and heat-flux DSC there will be a temperature difference between Ts
and Tsm. From equations 2 and 3:

𝑑𝑇𝑠
𝑇𝑠𝑚 − 𝑇𝑠 = 𝑅′𝑠 𝐶𝑠
𝑑𝑡
11
Example for Perkin-Elmer DSC-2:
R’s ≈ 0.06 K.s/ mJ, Cs ≈ 50 mJ/ K, dTs/ dt ≈ 0.167 K/ s gives a calculated temperature lag of ~0.5
K.
For the case in which an endothermic or exothermic process occurs, it is no longer acceptable to
assume that dTs/dt = dTr/dt. The relevant equations are a straightforward extension of the
derivation given above. These equations are given by Gray (1968), among others, and are simply
reproduced here without detailed derivation. For a sample generating heat at a rate dH/dt, we have,
for classical DTA or heat-flux DSC,

𝑑𝐻 1 𝑑𝑇𝑟 𝑑
= + (𝐶𝑠 − 𝐶𝑟 ) + 𝐶𝑠 (𝑇𝑠 − 𝑇𝑟 )
𝑑𝑡 𝑅(𝑇𝑠 − 𝑇𝑟 ) 𝑑𝑡 𝑑𝑡
12
For power-compensated DSC, assuming that the reference temperature changes at the same rate as
the heat source, we have
53

𝑑𝐻 𝑑𝑞𝑠 𝑑𝑞𝑟 𝑑𝑇𝑟 𝑑 𝑑𝑞𝑠 𝑑𝑞𝑟


= −( − ) + (𝐶𝑠 − 𝐶𝑟 ) − 𝑅𝐶𝑠 ( − )
𝑑𝑡 𝑑𝑡 𝑑𝑡 𝑑𝑡 𝑑𝑡 𝑑𝑡 𝑑𝑡
13
Equations 12 and 13 are quite similar. Both contain three terms involving the signal:
Ts-Tr or dqs/dt-dqr/dt, a heat capacity difference Cs-Cr, and a derivative of the signal. Note that it
is the negative of the signal that is the first term in equation 13. Of primary importance is the
presence of a thermal resistance factor in the first term of equation 12, but the factor only appears in
the third term of equation 13.

Modulated DSC (MDSC) is a relatively new technique that improves upon the classical DSC
performance in terms of increased ability to analyse complex transitions, higher sensitivity, and
higher resolution. Classical DSC uses a linear heating or cooling program of the form:
𝑇 = 𝛼 + 𝛽𝑡
14a
𝑑𝑇
=𝛽
𝑑𝑡
14b
Modulated DSC employs a linear heating or cooling program with a small sinusoidal component
superimposed:
𝑇 = 𝛼 + 𝛽𝑡 + 𝐴𝑡 sin(𝜔𝑡)
15a
𝑑𝑇
= 𝛽 + 𝐴𝑇 𝜔 cos (𝜔𝑡)
𝑑𝑡
15b
The heating rate now oscillates between β+ATω and β-ATω with an average underlying heating rate
β. Here, AT is the temperature modulation amplitude (in Kelvin) and ω is the angular frequency of
the temperature modulation (in radians per second) equal to 2π/P, where P is the temperature
modulation period (in seconds). To illustrate the MDSC capabilities, it is possible to recast equation
13 as follows:

𝑑𝑞𝑠 𝑑𝑞𝑟 𝑑𝑇𝑟 𝑑𝐻 𝑑 𝑑𝑞𝑠 𝑑𝑞𝑟


( − ) = (𝐶𝑠 − 𝐶𝑟 ) − − 𝑅𝐶𝑠 ( − )
𝑑𝑡 𝑑𝑡 𝑑𝑡 𝑑𝑡 𝑑𝑡 𝑑𝑡 𝑑𝑡
16
54

which upon combining the last two terms on the right-hand side, denoting differences between sample
and reference quantities with ∆, and replacing dTr/dt with dT/dt transforms into

𝑑∆𝑞 𝑑𝑇
= ∆𝐶 − 𝐹(𝑇, 𝑡)
𝑑𝑡 𝑑𝑡
17
In modulated DSC terminology, the term on the left hand side is the total heat flow, while the first
term on the right hand side is the reversing heat flow or heat capacity component and the second term
on the right hand side is the nonreversing heat flow or kinetic component. Classical DSC (both power
compensated and heat flux) measures and puts out the total heat flow. Modulated DSC is different in
that it measures and puts out the instantaneous heat flow (see figure below), which takes on the
modulated character of the temperature. From the modulated heat flow, it is then possible to obtain
both the total heat flow (roughly by averaging or, more precisely, using Fourier transform analysis
(Figure below) and its heat capacity component (roughly by taking the ratio of the modulated heat
flow amplitude Aq and AT or, more precisely, using Fourier transform analysis). An algebraic
difference of these two quantities, yields the kinetic component of the total heat flow. This
significantly improves the ability to separate overlapping thermal events with different behavior
example crystallization (non-reversing) and glass transition (reversing) in polymer blends. Sensitivity
and resolution are also improved. In addition, the presence of noninstrumental phase lag between the
modulated temperature and the modulated heat flow leads to the formulation of complex heat capacity
Cp*, which can be split into its real (Cp’) and imaginary (Cp’’) components and analyzed separately.
55

Figure 36: MDSC Heating profile (courtesy Journal of Thermal Analysis and Calorimetry 1993 (40),
931-939).

Figure 37: MDSC of quenched cooled polymer (PET). Raw signals include the modulated heat flow
and modulated heating rate (courtesy Journal of Thermal Analysis and Calorimetry 1993 (40), 931-
939).
56

Figure 38: MDSC of quenched cooled polymer: obtaining the kinetic (non-reversing) component
from the total heat flow and the heat capacity (reversing) component (courtesy Journal of Thermal
Analysis and Calorimetry 1993 (40), 931-939).

6. THE IMPORTANCE OF CALIBRATION OF THERMAL ANALYSIS


EQUIPMENT

Quantitative DSC/DTA data can only be obtained if the instrument has been properly calibrated with
respect to both temperature (DSC/DTA) and heat flow or enthalpy. Calibrants should ideally be
certified reference materials (values of transition temperature and/or transition enthalpy have been
determined for a particular lot of material) or at least high purity materials for which literature values
exist. Temperature calibration is a process of determining the difference, δT, between the actual
sample temperature, Ts, and the indicated temperature, Tind, and then either incorporating it into the
final data treatment or eliminating it by instrument control adjustment. The temperature difference
is, in general, a function of the indicated temperature and the heating rate:
δT = Ts - Tind = f(Tind, β)
18
The Tind dependence of δT should be determined with at least two calibrants bracketing the
temperature range of interest (two-point calibration) as close as possible. Using one or more
calibrants within the temperature range of interest is always beneficial (multipoint calibration) for
checking and, if applicable, correcting the linearity assumption of the two point calibration. The
dependence of δT on β is due to the increasing thermal lag of Ts behind Tind at higher heating
57

rates. Isothermal calibration furnishes the value of δT for a zero heating rate at δT0, while
calibration at the scanning rate of the experiment furnishes the thermal lag, δTβ. Hence, δT can be
calculated as

δT = δT0 + δTβ
19
At a given temperature, δT0 might be of either sign but δTβ is always negative in heating and positive
in cooling. A simpler method of temperature calibration is sometimes suggested by instrument
manufacturers that omits the isothermal calibration and yields a universal temperature correction
δT’’. This method is strictly valid only for very small samples.
Heat flow or enthalpy calibration is a process of determining the signal-to-heat-flow conversion
factor or the area-to-enthalpy conversion factor, respectively, under scanning conditions. Heat flow
calibration is used for heat capacity measurements by the scanning method (known heating rates, see
below). Enthalpy calibration is used for measurements of heat capacity by the enthalpy method (see
below) and for general enthalpy measurements. Heat flow calibration is almost universally performed
with two calibrants, namely α-alumina (sapphire) for superambient operation and benzoic acid for
subambient operation. Heat capacities for both materials have been accurately measured by adiabatic
calorimetry, and both are available as standard reference materials. Heat flow dH/dt is assumed to be
proportional to the signal S(t):

𝑑𝐻
= 𝐾(𝑡)𝑆(𝑡)
𝑑𝑡
20
where K is the signal-to-heat-flow conversion factor (K is ideally temperature independent but in
practice varies slightly with temperature and this dependence needs to be determined). The relation
between heat capacity Cp(t) and S(t) is then

𝑑𝐻 𝑑𝑇
𝑐𝑝 (𝑡) = = 𝐾(𝑡)𝑆(𝑡)( )−1
𝑑𝑇(𝑡) 𝑑𝑡

21
so that K(t) can be determined from the known calibrant heat capacity 𝑐𝑝𝑐𝑎𝑙 (𝑡), scan rate dT/dt, and
calibrant signal 𝑆 𝑐𝑎𝑙 (𝑡) as
58

𝑑𝑇
𝑐𝑝𝑐𝑎𝑙 (𝑡)( )
𝐾(𝑡) = 𝑑𝑡
𝑆 𝑐𝑎𝑙 (𝑡)
22
Enthalpy calibration is usually performed with one or ideally more calibrants (high purity materials
or certified reference standards). Rearranging equation 20 leads to a relationship between the enthalpy
and the integrated peak area (A) at some temperature of interest,

𝐻 = 𝐾 ∫ 𝑆(𝑡)𝑑𝑡 = 𝐾𝐴

23
so that K can be determined from the known transition enthalpy 𝐻 𝑐𝑎𝑙 and the integrated calibrant peak
area 𝐴𝑐𝑎𝑙 as

𝐻 𝑐𝑎𝑙
𝐾=
𝐴𝑐𝑎𝑙
24
Note that the signal-to-heat-floe conversion factor and the area-to-enthalpy conversion factor are
identical. This is only the case if the signal is considered a function of time, S(t). For the signal as a
function of temperature, S(T), the signal-to-heat-flow conversion factor K’ and the area-to-enthalpy
conversion factor K’’ are not identical.

𝑑𝑇 −1
𝐾 ′′ = 𝐾′( )
𝑑𝑡
25
59

7. THERMAL ANALYSIS EXPERIMENT


7.1 Experiment 1
Pharmaceutical crystallization studied with dynamic DSC measurements for estimating
Arrhenius parameters from the Friedman isoconversional method

Introduction: Kinetic analysis of nonisothermal experiments

Solid-state reactions are generally more complex than reactions in solution. In general a process
would be studied by measuring the reaction rate as a function of temperature or time. The next step
would then be to fit the data to an appropriate rate equation or kinetic model (f(α)) which may be
used to calculate a value for the rate constant (k):
dα/ dt = k f(α)

Knowledge of the rate constant would then provide information on the Arrhenius parameters (Ea and
A):

k = Ae-(Ea/RT)

dα/ dt = Ae-(Ea/RT) f(α)

where A is the frequency factor and Ea the activation energy. T and R are the absolute temperature
(K) and gas constant (8.314 J.mol-1.K-1) respectively. Knowledge of the activation energy, pre-
exponential factor and kinetic model (kinetic triplet) would enable predictions concerning kinetic
behaviour at other conditions of time and temperature [Maciejewski, Thermochemica Acta 2000).
Solid processes (decomposition, crystallization) may be very complex for example: intermediates
produced during decomposition, multiple solid-solid transformations etc. with more than one
activation energy involved. A particular kinetic model which only permits the calculation of a single
activation energy would therefore be inadequate for describing the system. Isoconversion methods is
usefull in this regard as these methods are model independent and permits the estimation of E as a
function of α. The variation in E as a function of α could therefore provide information on the
complexity of the process.
60

The first step to studying phase transformation kinetics would require a reliable estimate of α. For
crystallization studied by DSC (Differential scanning calorimetry) α can be estimated from the area
under a crystallization peak:
α = s/ S
where s and S is the partial and total area under a crystallization peak respectively.

In a nonisothermal kinetic experiment the temperature is changed according to a defined heating


programme, usually a linear increase of temperature with time. Nonisothermal rate expressions that
describe reaction rate as a function of temperature at a constant heating rate, β can be derived from:
dα/dt = dα/ dT (dT/dt)
where dα/ dT is the nonisothermal reaction rate and dT/dt is the heating rate β.

A reaction studied for example by DSC at a known heating rate, β, would be converted to values of
α and/ or dα/dt at temperatures T. The following general equation is applicable to isoconversional
methods:
ln[(d α/ dT)β] = ln(A/ f(α)) – Ea/ RT
Friedman’s method [J. Polym. Sci. 6C (1963) 183 – 195.]

involves a plot of ln(d α/ dt)i against 1/Ti, at the same value of αi at different heating rates βi. This
isoconversional method will be utilized to study the complexity of crystallization of amorphous
nifedipine. It is known that the amorphous modification will crystallize to an metastable form which
will ultimately convert to a stable modification. This study will investigate the amorphous →
metastable conversion.

Obtain DSC thermograms at various heating rates (β)


• Heat rates 0.5, 1, 1.5, 2.0, 3.0, 3.5 will be employed.
61

Construction of α – T curves:

1.2
0.5 C/min 2.0 C/min 4.0 C/min
1

0.8

0.6
alpha

0.4

0.2

0
60 70 80 90 100 110
-0.2
Temperature (C)

• Running integral (J.g-1) computed for each DSC curve


• Alpha (α) = fraction converted = partial integral/ total integral

ANF010
1.0°C.min-1
Deriv.
Heat Heat
Temperature Flow Flow Integral
°C W/g W/g/°C J/g alpha
-
71.01 0.03695 -0.00196 0.009561 0.00022
-
71.03 0.03705 -0.0018 0.01038 0.0002
-
71.06 0.03713 -0.00175 0.01091 0.0003
-
79.19 0.00628 0.01045 3.931 0.0903
-
79.22 0.00576 0.00928 3.985 0.0916
-
79.23 0.00556 0.009079 4.014 0.0922
-
79.25 0.00524 0.007833 4.043 0.0929
-
79.27 0.00495 0.007017 4.072 0.0936
-
79.28 0.00469 0.006799 4.124 0.0948
-
79.31 0.00436 0.006871 4.184 0.0961
62

79.33 -0.0041 0.006957 4.244 0.0975


-
79.34 0.00402 0.007302 4.281 0.0984
-
79.37 0.00371 0.007314 4.338 0.0997
-
79.38 0.00353 0.007207 4.395 0.101
-
79.4 0.00327 0.007301 4.418 0.1015
-
79.41 0.00313 0.007056 4.449 0.1022
-
79.43 0.00285 0.006408 4.496 0.1033
-
79.46 0.00242 0.006132 4.555 0.1047
-
79.49 0.00199 0.004535 4.598 0.1057
-
79.5 0.00182 0.00401 4.622 0.1062
-
79.51 0.00168 0.003943 4.646 0.1068
-
79.52 0.00146 0.003884 4.67 0.1073
-
79.54 0.00126 0.004075 4.695 0.1079
-
79.56 0.00101 0.004443 4.719 0.1084
-
79.58 0.00073 0.005616 4.743 0.109
-
79.59 0.00048 0.006495 4.772 0.1097
63

Construction of Arrhenius type plot for alpha = 0.1


• Calculation of dα/ dT
• Determine values of ln[(dα/dT)β] and 1/T (K-1) for each heating rate (β)

alpha = 0.1
1 1.5 2
K K K
79.19 352.34 0.0903 83.24 356.39 0.089 86.42 359.57 0.0909
79.37 352.52 0.0997 83.49 356.64 0.1009 86.59 359.74 0.1005
79.59 352.74 0.1097 83.62 356.77 0.1098 86.73 359.88 0.1093

dalpha/Dt
dalpha/Dt : 0.0485 : 0.0547 0.0594
beta fn T (K) 1/T(K-1) ln[fn]
0.0385 0.5 0.01925 344.13 0.002906 -3.9502
0.0526 0.6 0.03156 348.89 0.002866 -3.4559
0.051 0.8 0.0408 350.05 0.002857 -3.1991
0.0458 0.9 0.04122 350.99 0.002849 -3.1888
0.0485 1 0.0485 352.52 0.002837 -3.0262
0.0547 1.5 0.08205 356.64 0.002804 -2.5004
0.0594 2 0.1188 359.74 0.00278 -2.1303
0.0551 3 0.1653 359.86 0.002779 -1.8000
0.051 4 0.204 363.37 0.002752 -1.5896

• Plot graph of ln[(dα/dT)β] versus 1/T (K-1)

0.0000
0.0027
-0.5000 0.00275 0.0028 0.00285 0.0029 0.00295

-1.0000
-1.5000
-2.0000
ln fn

-2.5000
-3.0000
-3.5000
-4.0000
y = -15873x + 42.089
-4.5000
R2 = 0.9828
1/T (K-1)

• slope = -E/R
• Estimation of uncertainties
64

alpha =0.1
SUMMARY OUTPUT

Regression Statistics
0.99134219
Multiple R 5
0.98275934
R Square 7
Adjusted R 0.98029639
Square 6
0.11211670
Standard Error 8
Observations 9

ANOVA
Significance
df SS MS F F
5.01570729 399.017101
Regression 1 5.015707294 4 3 1.97153E-07
0.01257015
Residual 7 0.087991093 6
Total 8 5.103698387

Standard Lower Upper


Coefficients Error t Stat P-value Lower 95% Upper 95% 95.0% 95.0%
42.0888490 18.7435016 3.05603E- 47.3986527 36.7790454 47.3986527
Intercept 7 2.245516868 2 07 36.77904543 1 3 1
- - - - -
15873.0640 19.9754124 1.97153E- - 13994.0624 17752.0656 13994.0624
X Variable 1 5 794.6301043 2 07 17752.06567 4 7 4

slope(-) intercept E /kJ.mol-1 STD Err Ea R2


15873 40.089 131.97 6.606554687 0.9828 y = -15873x + 42.089
R2 = 0.9828

Materials

Nifedipine was purchased from Sigma-Aldrich, and used without further purification.

DSC
65

Differential scanning calorimetry (DSC) was performed on a TA Instruments Pont 910 Standard
DSC module, connected to a TA 2000 Thermal Analyzer. High purity nitrogen, at a flow rate of 65
mL min-1, was used as the purge gas. Samples for DSC were encapsulated using aluminium pans and
lids. The instrument was calibrated for temperature using indium and zinc standards. The cell constant
was calculated from the heat of fusion for indium.

Evaluation of the crystallization kinetics for the amorphous to form B

Nifedipine samples must be placed on light-weight aluminium foil and heated on a hot plate to
190°C. Remove the sample after 1 minute and allow to cool to ambient temperature. The sample,
now a solidified glass, must be accurately weighed in a DSC alumina pan. The samples of
approximately 3 mg will be DSC heated in a nitrogen atmosphere at rates that vary from 0.5 to 3.5°C
min-1, from 60 to 190°C. The heat rates should be slow enough to ensure good resolution for the
amorphous to form B and form B to form A crystallization processes.

Friedman Isoconversion method for evaluation of crystallization processes

The crystallization exotherms obtained from the dynamic DSC heating experiments will be studied
using the Friedman isoconversion method [12]. The procedure permits the estimation of activation
energy (E) as a function of the extent of reaction (α). The DSC crystallization exotherms must be
analysed using Universal Analysis software, provided by Setpoint instruments. Integration of the
crystallization exotherm (amorphous nifedipine to metastable form B) must be performed using a
sigmoidal baseline. Construct Alpha (α) temperature (T) curves using the running integral over the
crystallization temperatures. The value of the integral (J g-1) at a specified temperature divided by the
total integral for the crystallization exotherm would provide an estimate of α at that specified
temperature. The α-T data obtained at various heating rates (β) for the crystallization processes must
be used to calculate dα/ dT values for α values from 0.05 to 0.9. Construct plots of ln[(dα/dT)β]
versus 1/T (K-1) for each α value. The slopes = -E/R and the intercepts = ln[Af(α)]. If a constancy in
E is obtained as a function of α then the value of frequency factor (A) can be estimated by
extrapolation of a plot of the intercept against αi to αi = 0.

Report

• Calculate E from α = 0.1 to 0.9


66

• Comment on the complexity of the crystallization process from the calculated data.
67

Reference

M. E. Brown : Introduction to Thermal analysis : Techniques and applications, Chapter 13

You might also like