Differential Scanning Calorimetry (DSC)

Name: Edward Flood

Student number: 010929
!ab "artner: #ary S$ortall
Date: 1%0&%'0(
1
)b*tract:
The purpose of this experiment was to use Differential Scanning
Calorimetry to find numerous thermal properties of a number of polymer
samples. The credibility of the machine to be used was tested using Indium
metal as a reference material. Indium was used because its melting
temperature is low enough to coincide with the temperature range in which
most of the polymer transitions that apply to this experiment occur. Also, it
has a welldefined melting point with an accepted melting temperature and
enthalpy of melting of 1!"
o
C and #$.!%&g respecti'ely. The 'alues obtained
using DSC were 1!"
o
C and #1.((!%&g, which shows that the enthalpy
measurement is off by approximately #!) and the temperature
measurements are about right. The polymers analysed were as follows*
1. +olyethylene ,high density- ,.D+/-
#. +oly,0inyl Chloride- high molecular weight ,high 12 +0C-
3. +oly,0inyl Chloride- low molecular weight ,low 12 +0C-
(. +oly,Acrylic Acid- ,+AA-
!. +oly,1ethyl 1ethacrylate- ,+11A-
". +olystyrene
4. +oly,0inyl Alcohol- ,+0A-
As well as melting points and enthalpies of melting, the glass transition
temperatures of the abo'e polymers were found ,where applicable- and
attempts were made to account for any de'iations from ideality that
ine'itably occurred.
+ntroduction and t$eory:
Differential Scanning Calorimetry is a techni5ue used to measure thermal
properties of polymers based on the rate at which they absorb heat energy
compared to a reference material. The techni5ue ta6es ad'antage of the
energy changes in'ol'ed in the 'arious phase transitions of certain polymer
molecules. This allows se'eral properties of the material to be ascertained7
melting points, enthalpies of melting, crystallisation temperatures, glass
transition temperatures and degradation temperatures.
A heat flux differential scanning calorimeter will be used in the experiment.
This is one in which a sample is heated along with a reference material with
#
a 6nown specific heat. 8ne of the criteria of this techni5ue is that the sample
and reference material remain at the same temperature during heating. This
can be achie'ed by Setting the machine to heat both the sample and
reference material at a specific rate ,In this experiment, the rate is set to
19
o
C per minute-. This allows the heat flux or difference in energy input
between the sample and reference to be measured. 1aintaining a constant
supply of heat to both materials would not ,unless the materials ha'e the
same heat capacity at all points, which is unli6ely- maintain a minimal
temperature difference between them. Instead, a computer is connected to
the machine, and using the software and 'arious signals from the
calorimeter, :decides; when to supply heat to either material. This
information is then dealt with by the computer software and presents it as a
graph of the energy changes 'ersus the temperature. A standard DSC cur'e
for a particular polymer is shown in fig. 1.
<ig 1. A standard output for a polymer from a DSC machine.
<rom the diagram, it can be seen that a sudden upward =ump in the cur'e
signifies an exothermic process. A sudden drop in heat flux indicates an
endothermic process.
It is possible to approximate the heat flow into the sample holder using the
following e5uation*
( ) T T K
dT
dQ
b
=
>>>>>..,1-
T is the sample temperature
T
b
* programmed bloc6 temperature
3
?* thermal conducti'ity of the material
qt T T
b
+ =
9 >>>>>..,#-
T
9
is the initial temperature and 5 is the programmed heating rate.
The heat capacity is defined as the amount of heat energy re5uired to raise
the temperature of a body by 1?. <or a substance with a constant heat
capacity*
( )
9
T T C Q
P
=
>>>>>..,3-
It is possible to deri'e an e5uation from e5uations # and 3 that forms the
basis for the DSC experiment. This is as follows*
K
C
q T
P
=
>>>>>..,(-
@T is the difference in temperature between the reference material and the
sample.
The heat capacity is gi'en as
P P
mc C =
>>>>>..,!-
2here c
p
is the specific heat ,Amount of heat re5uired to raise the
temperature of unit mass by 1?-.
The enthalpy change for a gi'en phase transition may be found by
integrating o'er the area in which the transition is seen to occur on the DSC
plot. This change may be described by the following integral*

= =
f
i
f
i
T
T
T
T
P
dT
q
T K
dt C H >>>>>..,"-
2here the limits of integration T
i
and T
f
are the initial and final temperatures
o'er which the graph is integrated. Therefore it is the area under the cur'e
that gi'es information about the enthalpy changes in'ol'ed in the 'arious
transitions.
(
Some transitions can be identified at certain temperatures, and the following
paragraphs include short descriptions of these transitions, the order in which
they occur, and their significance at a molecular le'el.
#la** tran*ition tem"erature, -
#
*
This is the point at which, on heating, an amorphous polymer changes from
being hard, brittle and glass li6e to being a soft rubber li6e substance. In a
way, this is analogous to a percolation threshold. As the polymer is cooled,
molecules ha'e less freedom and become bonded to each other. As the
percentage of bonded molecules increases, there comes a point at which
longrange connecti'ity occurs ,the :percolation threshold;- and the
polymer re'erts to a glass li6e phase again. This point may be identified as a
dip on a graph of heat flux 'ersus increasing temperature, this is due to the
molecule ha'ing a greater degree of freedom and absorbing energy to
maintain the same rate of heating as the reference. It can then be concluded
that glass transition is an endothermic process. Anli6e a transition from a
solid to a li5uid, or li5uid to gas, the energy change in'ol'ed in glass
transition ,a second order phase change- is 5uite small. <ully crystalline
polymers ha'e no glass transition and their structure remains intact until the
melting point. The following table is a list of factors relating to polymers
and their effect on the glass transition temperature*
Decreasing T
g
Caused by*
Increasing T
g
Caused by*
1ain chain flexibility 1ain chain rigidity
<lexible side chains Bul6y or rigid side chains
Increase in Tacticity Increased cohesi'e energy density
Increased symmetry Increased molecular weight
Branching Increased polarity
Addition of
plasticisers
8r diluents
Cross lin6ing
-em"erature of cry*talli*ation, -
C
*
Continuing to heat a polymer past its Class transition temperature e'entually
leads to another transition. Before this point in the thermal analysis, the
!
molecules are arranged in a random fashion and are coiled around each other
in an unfa'ourable manner. A transition occurs when molecules ac5uire
enough freedom to mo'e into a more energetically stable phase, i.e. a
crystalline state. This would indicate that the phase following the glass
transition is metastable and when sufficient energy is supplied, its molecules
adapt a more stable ,lower energy- arrangement. Because the molecules in a
crystalline solid ha'e less freedom than those in a flexible rubbery one, the
transition between these two states is exothermic and may be seen as a brief
upward :=ump; in the graph.
-em"erature of melting, -
.
:
This is the point at which the ,now crystalline- polymer molecules ha'e
gained enough 'ibrational freedom to brea6 free from the solid binding
forces and form a li5uid. Due to the increased freedom of these molecules,
the DSC graph should ta6e a sudden dip at this temperature to indicate the
endothermic nature of the process, which is a first order transition. The
melting point of any substance can be gi'en by*
m
m
m
S
H
T

=
>>>>>.>.,4-
@.
m
is the melting temperature and @S
m
is the change of entropy in'ol'ed in
the melting phase ,entropy of fusion-.
Degradation -em"erature, -
D
*
The final transition on a DSC graph is the degradation temperature, T
D
. At
this point in the heating cycle, indi'idual bonds between atoms start to brea6
as the 'ibrations become more and more fierce until e'entually, indi'idual
polymer molecules decompose into their components. Depending on the
nature of the substance under in'estigation, this process can be either
endothermic or exothermic
In this experiment, the thermal properties of se'en polymers will be
in'estigated along with those of a reference material, which, in this case,
will be indium metal. These are as follows*
1. Indium
#. +olyethylene ,high density- ,.D+/-
"
3. +oly,0inyl Chloride- high molecular weight ,high 12 +0C-
(. +oly,0inyl Chloride- low molecular weight ,low 12 +0C-
!. +oly,Acrylic Acid- ,+AA-
". +oly,1ethyl 1ethacrylate- ,+11A-
4. +olystyrene
$. +oly,0inyl Alcohol- ,+0A-
The transitions listed abo'e will, where applicable, be obser'ed and the
enthalpy changes in'ol'ed in these processes will be measured by
integration of areas underneath the start and endpoints of any pea6s or dips
on the DSC graphs. Computer software accompanying the differential
scanning calorimeter will enable the enthalpy changes to be measured
without the need for e'aluating large numerical integrals by hand.
)""aratu*:
Shown in fig.# is a diagram of the wor6ings of a differential scanning
calorimeter.
<igure #, the wor6ings of a DSC machine*
The reference pan is to remain empty at all times. Signals from both pans
pro'ide information to the computer. This allows it to regulate the
temperature and pro'ide a constant heating rate. Dot shown are the gas inlets
and outlets, using these, a chemically inert gas such as Argon or nitrogen
may be pumped through the machine in order to pre'ent the accumulation of
dust. The gas also ser'es to remo'e any 'olatile chemicals that may be
present.
4
/rocedure:
<irstly, samples were prepared for analysis. In order to a'oid confusion, this
was performed as the samples were re5uired as opposed to preparing all
samples at once. /xtreme care was ta6en not to contaminate samples or
sample pans with either other chemicals or matter from human s6in. This
meant that all e5uipment had to be handled with tweeEers. Samples were
prepared as follows*
The tare weight of the sample pan and lid were obtained. Sample pans with
flat undersides were selected so as to ma6e the maximum contact possible
with the bottom of the calorimeter. <ollowing this, between ! and 19
milligrams of polymer were placed in the sample pan and the mass of the
filled sample pan and lid were found. This was the mass of polymer to be
used. Again using tweeEers, the sample pan and lid were mo'ed to the
crimping press whereupon crimping the lid sealed the pan. Another sample
pan was weighed along with its lid. This, howe'er, was left empty and
sealed using the crimping de'ice, forming the reference pan to be heated
along with the sample.
The first sample to be obser'ed was the Indium reference material. This was
placed in the left hand side of the furnace in the calorimeter while the
reference pan was placed in the right hand position, where it was to remain
for the duration of the experiment.
.eating cycle for Indium*
Ditrogen atmosphere, with a gas flow rate of 19Cm
3
&min, heating rate of
19
o
C&min from 1#9
o
C to 1$9
o
C.

<or the polymer samples, the heating cycle proceeded as follows*
The same gas flow rate was maintained, along with the heating rate of
19
o
C&min. The only change was that the range of temperature was modified
so that it ranged from 39
o
C to #!9
o
C. <ollowing this, the furnace was
allowed cool to room temperature before the next sample could be
introduced.
$
The resulting graphs are then analysed and information about onset, pea6
and endpoint temperatures and changes in enthalpy for the 'arious processes
can be obtained using the computer software.
0e*ult* 1 analy*i*:
Shown in table 1 are the masses of each of the polymer samples ta6en along
with the mass of Indium. Also temperatures of glass transition,
crystallisation, melting and degradation are shown along with the enthalpy
changes associated with each of these.
Table 1*
Sample Dame 1ass in
mg
T
C
,
o
C- T
1
,
o
C-
Indium 1#.4 n&a 1!".F
+olyethylene
,high density-
4.1 n&a 1#(.F#
o
C
+oly,0inyl Chloride-
.igh 12
F.# $!.3$! n&a
+oly,0inyl Chloride-
Gow 12
F.! $1.#!# n&a
+oly,Acrylic Acid- F.4 139
+oly,1ethyl
1ethacrylate-
19.3 193.3#
+olystyrene F.F 41.!!"
+oly,0inyl Alcohol- $.1 F9 1"#.1$
The resultant graphs are shown on the next few pages along with the
enthalpy changes and temperature range of each transition. The first of these
graphs is for the Indium reference material.
Craph 1*
F
1ass of sample* 1#.4mg
/nthalpy of melting* #1.((!1.94%&g
8nset temperature* 1!".F
o
C
+ea6 melting temperature* 1!$.(1
o
C
/olyet$ylene*
Craph #*
1ass of sample* 4.1mg
19
/nthalpy of melting* 11$.3!.F#%&g
8nset temperature* 1#(.F#
o
C
+ea6 melting temperature* 133.4$
o
C
/oly()crylic )cid)*
Craph 3
1ass of polymer ta6en* F.4mg
T
g
*
/nthalpy of melting* 44.F!F3.F%&g
8nset temperature* 139
o
C
+ea6 melting temperature* 1(F.11
o
C
/oly(.et$ylmet$acryalte): Craph (*
11
1ass of polymer ta6en* 19.3mg
T
g
* 193.3#
o
C
/nthalpy of melting*
8nset temperature*
+ea6 melting temperature*
Degradation onset* 1F1.(
o
C
+ea6 degradation temperature* #91.F
o
C
/nthalpy* 1.(F44%&g
/oly(2inyl )lco$ol):
Craph !*
1#
1ass of polymer ta6en* $.1mg
T
g
* F9
o
C
/nthalpy of melting* (3.4$#.#%&g
8nset temperature* 1"#.1$
o
C
+ea6 melting temperature* ###.1!
o
C
/oly*tyrene:
Craph "
13
1ass of polymer ta6en* F.Fmg
T
g
*
/nthalpy of melting* $.#"$"9.(%&g
8nset temperature* 41.!!"
o
C
+ea6 melting temperature* 43.$(
o
C
3ig$ .4 /oly(2inyl C$loride): Craph 4*
1(
1ass of polymer ta6en* F.#mg
T
g
* $!.3$!
o
C
/nthalpy of melting*
8nset temperature*
+ea6 melting temperature
!ow .4 /oly(5inyl C$loride):
Craph $*
1ass of polymer ta6en* F.!mg
T
g
* $1.#!#
o
C
/nthalpy of melting*
8nset temperature*
+ea6 melting temperature
1!
Di*cu**ion:
In this section, the results will be discussed and any discrepancies between
theory and experiment will be addressed*
+ndium:
A welldefined transition is seen to occur between the temperatures of 1!"
and 1"9
o
C ,centred on a pea6 of 1!$.(1
o
C- with an enthalpy change for the
transition of #1.((!1.94%&g. The actual melting point of Indium metal is
1!".9
o
C and its enthalpy of fusion is #$."%&g, so it can be concluded that the
transition on the graph is the melting of the sample.
/olyet$ylene:
+olyethylene is a simple straight chain polymer composed of repeating units
of the ethylene molecule. Its structure is shown below*
The output graph from the DSC machine showed one distinct transition with
an onset temperature of 1#(.F#
o
C and a pea6 of 133.4$
o
C with an associated
enthalpy change of 11$.31!.F%&g. By comparison, the melting point of
highdensity polyethylene is 133.!
o
C. It can be seen that this is in 5uite close
agreement with accepted theoretical 'alues. Because +olyethylene consists
mainly of saturated carbon bonds and due to the lac6 of anything other than
wea6 Gondon interactions to hold its molecules together, polyethylene has
low 'alues for both the glass transition and melting temperatures.
/olyacrylic )cid:
In this case, two distinct transitions may be seen. The first of these occurs at
!$.$#
o
C with a change in enthalpy of ".F((#9.3!%&g and it can be seen that
the transition occurs o'er a relati'ely narrow range of temperature. The
other transition, howe'er, occurs o'er a much broader range from
approximately 139
o
C to #99
o
C with a pea6 temperature of 1(F.11
o
C and an
1"
enthalpy change of 44.F!F3.F%&g. The structure of +oly,Acrylic Acid- is
shown here*
The presence of the hydroxyl group as a substituent on the molecule should
ensure that +AA has a higher melting point ,due to .ydrogen bonding- than
that of the more structurally simple polyethylene. This is indeed the case as
the melting of this substance occurs at approximately 1!9
o
C ,this is the pea6
'alue-. .owe'er, the melting occurs o'er a large range of temperatures
which could be explained in a number of ways. The most probable reason
for this is that there were impurities present in the sample. 1ost li6ely a
foreign ob=ect was pic6ed up during the weighing process as the analytical
balance may not ha'e been cleaned beforehand. Also the sample may ha'e
been contaminated in the handling process, by either the tweeEers or
accidental contact with human s6in. Another possible explanation could
apply if there was a large distribution of molar masses of the polymer
present in the sample. Anfortunately this cannot be 'erified so the blame for
this must lay at human error. Also, in theory, a glass transition should occur
in the region of 19!
o
C, and in the graph, no such transition is seen, pointing
further at sample contamination.
/oly(.et$yl .et$acrylate):
Two transitions are e'ident from the graph7 firstly an onset temperature of
193.3
o
C leads to a pea6 at 11#.4
o
C and an enthalpy change of #.41#9.1(%&g.
Secondly, an exothermic transition is seen to occur, beginning at 1F1.(
o
C,
running through a pea6 of #9(.F
o
C with a small enthalpy change of
1.(F449.94!%&g. The structure of +oly,methylmethacrylate- ,+11A- is
shown below*
14
The melting point of +11A is ta6en from literature as 34$? or 19!
o
C,
indicating the first transition on the graph to be a melting point. The second
transition is much more interesting, howe'er. Due to an exothermic pea6 at
such a high temperature it may be concluded that this is the degradation
temperature.
/oly(2inyl )lco$ol):
The most noticeable thing about the output from the sample of +0A is the
apparent presence of two pea6s. The first pea6 may be due to impurities in
the material as there are no transitions in its immediate :neighbourhood;.
The accepted melting point of this polymer is approximately #39
o
C ,fully
hydrolysed- and no other transitions are 6nown occur in the area of 13F to
1$9
o
C. The second transition occurs from #9!
o
C with a pea6 at ###.1!
o
C and
an enthalpy change of (3.4$$#.#%&g, meaning this is the point at which the
sample melted. Also, according to references, a glass transition occurs at
approximately $!
o
C, this can be seen as a miniscule dip in the graph at
approximately F9
o
C. %ust by loo6ing at the structure of +0A, it becomes
clear why it has such a high boiling point. The molecule itself has minimal
branching and can ma6e the most of any Gondon interactions between
molecules. These forces of interaction donHt dominate, howe'er. .ydrogen
bonding is by far the dominant interaction in this particular polymer and
goes some way into explaining its thermal beha'iour.
/oly*tyrene:
+olystyrene is accepted to ha'e a glass transition at approximately 343? or
199
o
C. A pea6 is seen to occur at approximately 4$
o
C but the nature of this
transition conflicts with that of a glass transition. Because the graph is seen
to dip sharply and all but rise bac6 up to its original :height;, the possibility
of glass transition at this point can safely be ruled out. Because +olystyrene
comes in a massi'e 'ariety of molar masses, large 'ariations in properties
are not uncommon ,'iscosity etc.-, therefore, the only logical conclusion to
draw is that the sample of polystyrene used for this test has a 'ery low
1$
molecular weight or is 'ery highly branched. ,The small range in
temperature o'er which melting occurs displays that chain lengths of this
sample of polystyrene are 'ery similar-. This would lead to the transition in
5uestion being deemed a melting point. Structure of polystyrene*
Iunning the sample through a 'iscometer in an appropriate sol'ent could
'erify a low molecular weight. The relati'e bul6 of the phenyl group
attached to the polymer also contributes another 'ariable to the materialthe
tacticity of polystyrene can dictate whether it can pac6 into a more ordered
solid structure with more stable thermal properties. The more ordered the
+henyl groups lie along the chain, the better able they are to pac6 in an
ordered fashion and this case is syndiotactic polystyrene, the structure of
which is shown below*
This is ob'iously more suited to pac6ing than atactic polystyrene ,formed by
random polymerisation-, which has its phenyl groups randomly oriented.
/2C:
.igh molecular weight +0C differs from the low molecular weight 'ariety
only in terms of chain length and both compounds ha'e the same structure*
1F
The difference between the two forms is merely in the 'alue of :n;. The
similarity, howe'er, does not end there. The lower temperature regions of
the graphs in each case are stri6ingly similar in shape with glass transitions
occurring at, $1.#!#
o
C and $!.3$!
o
C for low and high molecular weight
respecti'ely. This is 5uite a high 'alue for a glass transition and can be
attributed to the Cl

group attached to the main Carbon chain ta6ing a greater

share of the electrons in its bond with Carbon. This has the effect of setting
up an electric dipole that may interact with the other molecules 'ia a dipole
dipole interaction. A second transition is e'ident in the higher temperature
regions of the graph for low density +0C. Starting at 14!.4!
o
C with a pea6
at 1$4.!3
o
C and an endothermic enthalpy change of (.F!9.#!%&g, it has far
too small an energy to be a melting point and is possibly a degradation
temperature. Do information was a'ailable as regards degradation
temperatures, howe'er, so to err on the side of caution, this will be put down
to impurities present in the material. An increase in the 'alue of n for a
straight chain polymer causes an increase in the length o'er which a gi'en
molecule can interact with its nearest neighbours by the 0an der 2aals
force, thus increasing the temperature at which 'arious transitions occur.
Conclu*ion*:
Analysis of the polymers was preceded by the analysis of a material with a
welldefined melting point and enthalpy of melting ,Indium-. The result of
this measurement compared with the theoretical 'alues as follows* Accepted
enthalpy of melting of Indium* #$.!%&g 0alue obtained through DSC*
#1.((!1.94%&g. Accepted melting point of Indium* 1!"."
o
C. 0alue obtained
through DSC* 1!".F ,onset-. These results display two important points.
Temperature 'alues for transitions are 5uite accurate and agree to within a
fraction of a percent. .owe'er, 'alues gi'en for enthalpy transitions can be
ta6en with a pinch of salt with our 'alue being about #!) out from what can
be called acceptable. The following table ser'es to compare 'alues obtained
in this experiment with uni'ersally accepted 'alues*
1aterial
T
C
,
o
C- T
1
,
o
C- Theory T
g
Theory T
m
Indum
n&a 1!".F
1!"
.D+/
n&a 1#(.F#
139
#9
.igh 12 +0C
$!.3$! n&a $!
Gow 12 +0C
$1.#!# n&a $1
+AA
139 19"
+11A
193.3# 119
+olystyrene
41.!!" 19(
+0A
F9 1"#.1$ $!
#99
+re'ious page* Table #comparison of 'alues obtained from the experiment 'ersus accepted 'alues.
All errors in enthalpies were specified by the e5uipment manufacturer at
!) of measured 'alues.
)dditional e6cerci*e*:
+ercentage crystallinity of polyethylene sample*
g J & "1 . #F3 ) 199 =
The percentage crystallinity is therefore*

) 1 . # # . (9 199
"1 . #F3
3 . 11$
=
Thermogram of polymer with the following transitions*
/ndothermic transition at $9
o
C, T
g
at 1(!
o
C, T
m
of #3!
o
C and a
decomposition temperature of (99
o
C
Is shown in figure 3*
<igure 3*
#1
All other re5uested information was discussed earlier in this report.
0eference**
www.fsref.com/Fatal/FEA00002.SHTML
www.pslc.ws/mactest/dsc.htm
ois.nist.gov/nistpus/technipus/ recent/search.cfm"diid#$2%2&
irc.nrccnrc.gc.ca& cbd&cbd1!(e.html
repositor'.matdl.org/retrieve/ 2%($/pol'mer)final)research)article.pdf
www.chem.ox.ac.u6&...& Ga'ender&la'ender.html
www.er*ol.com/eng/+haracteristics.htm
##
#3