Vous êtes sur la page 1sur 8

European Polymer Journal 50 (2014) 235242

Contents lists available at ScienceDirect

European Polymer Journal


journal homepage: www.elsevier.com/locate/europolj

Annealing of poly (ethylene terephthalate)


Ziyu Chen, M.J. Jenkins, J.N. Hay
The School of Metallurgy and Materials, College of Engineering and Physical Sciences, The University of Birmingham, Edgbaston, Birmingham B15 2TT, UK

a r t i c l e

i n f o

Article history:
Received 30 August 2013
Accepted 4 November 2013
Available online 14 November 2013
Keywords:
Poly (ethylene terephthalate)
Annealing
FTIR spectroscopy
DSC
Lamellae thickening
Secondary crystallization

a b s t r a c t
Poly (ethylene terephthalate), PET, crystallized isothermally at 234 C has been annealed
above the crystallization temperature in regimes II, below the onset of melting, and in
regime III, in the range of observable melting [1,2]. The mechanism has been analyzed
by differential scanning calorimetry and FTIR spectroscopy and shown to be one of the limited recrystallization nucleated by crystalline regions which survived melting followed by a
process of lamellae thickening.
Annealing above the crystallization temperature in regime II where melting and recrystallization did not occur involved lamellae thickening by growth of the lamellae along the
chain direction. This was a diffusion controlled mechanism involving either diffusion of
non-crystallizable impurities or chain entanglements. The process of stem thickening
was the same as that of secondary crystallization previously observed in isothermal crystallization [3] and which is responsible for the observed melting point of the polymer being
dependent on the annealing time as well as the temperature.
In regime III melting of the polymer occurred initially followed by recrystallization on
the residual lamellae acting as seeds by one dimensional growth along the original growth
direction. This was later accompanied by thickening of the lamellae with the same characteristic time dependence of diffusion control as observed in regime II and with secondary
crystallization. This involves reptation of the molecular chain from the entangled melt onto
the growth face.
2013 Elsevier Ltd. All rights reserved.

1. Introduction
Annealing has a profound effect on the material properties of partially crystalline polymers by increasing the
melting point, fractional crystallinity, stiffness, yield stress
and brittleness, as well as decreasing the elongation to
break, and fracture toughness [4]. It is invariably associated with a change in morphology and a re-organization
of the distribution of lamellar stem lengths which results
in a reduction in chain mobility as measured by NMR line
width and dynamic mechanical thermal analysis [14]. An
understanding of these changes is vital for the end users
point of view if material properties are to be maintained
Corresponding author. Tel.: +44 121 414 4544; fax: +44 121 414
5232.
E-mail address: j.n.hay@bham.ac.uk (J.N. Hay).
0014-3057/$ - see front matter 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.eurpolymj.2013.11.004

and optimized with the correct choice of processing


conditions.
Three distinct molecular mechanisms for annealing have
been suggested [1.2] wmechanisms for annealing have been
suggested s; at low temperatures, well below melting, solid
state diffusion occurs by which the lamellae thicken with
migration of chain defects; the second occurs above but
close to the crystallization temperature in which thin lamellae melt and re-crystallize as thicker lamellae; the third occurs at temperatures within the observable melting region
and involves partial melting of the lamellae surface inwards
towards the core along the chain axis followed by re-crystallization at the annealing temperature producing thicker
lamellae and increased crystallinity. Considerable evidence
has been obtained for all three mechanisms based on electron microscope analysis of the morphologies produced
and SAXS [1] measurements of the lamellae stem size.

236

Z. Chen et al. / European Polymer Journal 50 (2014) 235242

Alternative mechanisms have also been suggested for


annealing of crystalline polymers [48] without any prior
melting. The embrittlement of poly (3-hydroxybutyrate)
and its copolymers on storage at room temperature about
150 C below the onset of melting, has been attributed to
the crystallization of interlamellar regions by a process in
which thin lamellae are produced. These lamellae underpin the amorphous regions, reduce the mobility of the
amorphous chain segments and increase the stiffness of
the amorphous regions within the boundary of the
spherulites.
Recently [9] we have reported that the crystallinity of
poly (ethylene terephthalate) can be measured directly
by FTIR spectroscopy as a function of temperature and over
extended time periods with sufcient accuracy to enable
both primary and secondary crystallization to be measured
with similar accuracy. The technique has the potential to
follow changes in amorphous and crystalline content during annealing and resolve the problems associated with the
change in mechanisms of annealing.
This paper reports the study of annealing on a PET sample which has been crystallized by a standard procedure to
produce well-developed lamellar morphology which has
not thickened by a secondary crystallization process. It
compares data on annealing this sample obtained by DSC
and FTIR spectroscopy.

2. Experimental
PET lms, 1.513 lm thick, were obtained from Goodfellow Ltd., UK. These were used in spectroscopic analyses
as outlined elsewhere [9]. The lms were dried in a vacuum oven at 100 C for 12 h prior to use. These were stress
relieved by melting between KBr plates and quenched to
produce amorphous samples, as measured by DSC [10,11].
FTIR spectra were measured on a Nicolet model 8700
spectrophotometer with a DTGS-KBR detector on thin lms
samples mounted between KBr plates and contained within the furnace of a Linkam hot-stage [9,12]. The furnace
temperature was controlled by a Unicam PR600 thermal
controller. All spectra were recorded at a resolution of
2 cm1 and total of a 100 scans were accumulated for each
spectrum along with the background. The spectra were
analyzed for peak resolution and area. Thin lms were
cut into small pieces, placed between KBr plates and
melted at 270 C for 1 min. mounted in a Linkam PS600
hot-stage and placed vertically in the IR beam. In this
way, shrinkage and orientation of the lms were removed
and the lms adhered to the KBr disks. The temperature of
the hot-stage was controlled to an accuracy of 0.1 C. The
kinetics of isothermal crystallization have been studied,
previously over the temperature range 230240 C and
the same experimental procedures adopted [9].
A PerkinElmer differential scanning calorimeter, DSC7, interfaced to a computer was used as described elsewhere [10,11] to follow the rate of heat evolution with
time. The temperature scale of the DSC was calibrated from
the melting point of high purity metals; tin (230.06 C) and
indium (155.58 C). Power response of the calorimeter was

calibrated from the enthalpy of fusion of indium, taken to


be 28.47 Jg1.
3. Results
3.1. Initial considerations
A standard crystallization procedure was adopted to
produce reproducible crystalline samples for annealing
studies. This involved heating samples from room temperature to 280 C, maintaining them for 2 min at this temperature to produce an amorphous sample and cooling at
160 C min1 to 234 C, the crystallization temperature.
Crystallization was carried out for 50 min. to allow crystallinity to develop to the end of the primary stage [8] and the
sample heated to the annealing temperature. The fractional crystallinity was followed by changes in absorbance
of the carbonyl bands in the FTIR spectrum with time.
On heating the standard sample in the DSC from the
crystallization temperature at 10 C min1 melting was observed to occur over the temperature range 235260 C. A
broad endotherm was observed with three distinct melting
regions; melting did not begin immediately but corresponds to the calorimeter reaching an equilibrium rate of
heating; the second corresponded to the melting of the
least stable less well-ordered lamellae present; this was
followed by the melting of more stable lamellae. Melting
was completed at 258 C.
Yeh et al. [2] have distinguished three approximate
temperature regions for carrying out annealing studies
where chain mobility and changes to material properties
of polymers have been observed to occur; these are below
the crystallization temperature, above the crystallization
temperature prior to melting and in the observed melting
range. Only the last two regions were studied and annealing temperatures were accordingly selected from 234 to
250 C.
The change in amorphous content and crystallinity was
followed by FTIR spectroscopy by resolving the carbonyl
band into amorphous and crystalline components, and
the fractional crystallinity determined as outlined in
Appendix 1. The change in the absorbance of the amorphous and crystalline bands at 1727 and 1717 cm1
respectively with log time is shown in Fig. 1 for the isothermal crystallization at 234 C up to 50 min; in line with
these assignments, the absorbance at 1727 cm1 decreased
while that at 1717 cm1 increased following an initial
exponential dependence with time as described by the
Avrami equation [5] followed by a linear dependence on
log time. These have been attributed to the presence of a
primary crystallization with an exponential increase in
crystallinity with time and a secondary crystallization corresponding to a logarithmic increase with time. [3].
3.2. Annealing studies
On annealing the changes in the fractional crystallinity,
Xc,t, and amorphous content, Xa,t, were measured as a
function of time, calibrating the relationship between

237

Z. Chen et al. / European Polymer Journal 50 (2014) 235242


1

T = 234 C
c

Fractional Crystallinity, Xc,t

0.6

0.4

1717 cm-1
0.2

0.8

0.8

0.6

T = 250 C

0.6

0.4

0.4
o

T = 250 C
c

0.2

0.2
o

T = 234 C

2.5

3.5

4.5

2.5

log (time / s)
1

) and

absorbance and fractional crystallinity at each temperature, as outlined in Appendix 1.


At 244 C on heating to the annealing temperature the
crystallinity increased initially by 5.3 1.3% from the last
value measured at the crystallization temperature to the
rst steady value at the higher temperature. No melting
of the sample occurred and annealing was accompanied
by a progressive increase in crystallinity with time, see
Fig. 2. The corresponding decrease in amorphous content
followed a similar but decreasing trend with time.
However, on annealing at 250 C partial melting of the
sample occurred which was not completed until about
5 min after rst equilibrating at the annealing temperature, see Fig. 3. On annealing melting of the sample was
apparent from the reduction in fractional crystallinity to
0.15 0.02% and was followed by re-crystallization; this
exhibited a lower time dependence than that observed in
primary crystallization, implying a change in crystallization mechanism. This in turn was followed by period in

which the crystallinity increased linearly with log time,


usually taken to be symptomatic of secondary crystallization (see Fig. 4).
At 252 C melting was much more apparent and occurred over rst 1530 min after equilibrating at the
annealing temperature and a low residual crystallinity,
0.04 0.02, was achieved, see Fig 4. This was followed by
a slow recrystallization which appeared to accelerate with
log time. The nal degree of crystallinity achieved, 0.18,
was less than that of the original sample but limited by
the timescale chosen for the experiment, see Fig. 5.
3.3. Kinetic analyses
The development of crystallization with time for each
annealing temperature as shown in Fig. 5 was analyzed
using a modied form of the Avrami equation [13]

0.4

T = 244 C
c

0.2

0.2
o

T = 234 C
c

3.5

4.5

1
o

234 C

Fractional Crystallinity, Xc,t

Fractional Crystallinity, Xc,t

0.6

T = 244 C

2.5

252 C

0.8

0.8

0.6

0.6

0.4

0.4
o

234 C

0.2

0.2

252 C

2.5

3.5

4.5

Fractional Amorphous Content, Xa,t

0.8

0.8

4.5

Fractional Amorphous Content, Xa,t

Tc = 234 C

Fig. 3. The change in fractional crystallinity, Xc,t, and amorphous content


Xa,t, with time on crystallizing at 234 C for 50 min and then after partial
melting and re-crystallizing at 250 C (time was reset to zero on heating
to 250 C).

0.4

0
3.5

log (time / s)

Fig. 1. The change in the absorbance of the crystalline (at 1717 cm


amorphous bands (at 1727 cm1) on crystallization at 234 C.

0.6

Fractional Amorphous Content, Xa,t

1727 cm -1

0.8

Absorbance

Tc = 234 C

log (time / s)

log (time / s)
Fig. 2. The change in fractional crystallinity, Xc,t, and amorphous content,
Xa,t, with time on crystallizing at 234 C for 50 min and then on annealing
at 244 C (time was reset to zero on heating to 244 C).

Fig. 4. The change in fractional crystallinity, Xc,t, and amorphous content,


Xa,t, with time on crystallizing at 234 C for 50 min. and then after partial
melting and re-crystallizing at 252 C (time was reset to zero on heating
to 252 C).

238

Z. Chen et al. / European Polymer Journal 50 (2014) 235242


0.6

log (- ln (1- (X -X )/(X -X ))))

0.5

Ta = 244 C

inf

0.4

0.3

Fractional Crystallinity, Xt

0.5

Ta = 250 C

0.2

0.1

Tc = 252 C

2.5

3.5

4.5

Ta = 252 C

-1
o

Ta = 250 C

-1.5

-2

Ta = 244 C

-0.5

2.5

log (time / s)

3.5

4.5

log (time / s)

Fig. 5. Effect of annealing temperature on the development of crystallinity with time.

X t X o X 1  X o 1  exp Zt n

Fig. 6. Fit of the modied Avrami equation to the development of


crystallinity with time.

log ln1  X t  X o =X 1  X o n log t log Z

1.4
1.2

0.8
0.6
0.4
0.2

the rate parameters as listed in Table 1 were obtained from


a plot of log (ln(1((Xt  Xo)/(X1  Xo)))) against log t,
see Fig. 6, and the half-lives, t1/2, determined from the time
corresponding to

0.2

0.4

0.6

0.8

Extent of Crystallization
1
0.8

A linear dependence was observed for Eq. (2) and n values in the range 0.7 0.1 at 244 and 250 C and 1.0 0.2 at
252 C.
Deviations from linearity were observed towards the
end of the annealing process and in order to determine if
this was due to the presence of two consecutive processes
the data was re-analyzed to determine instantaneous n
values as a function of the extent of crystallization, see
Fig. 7, using the differential form of Eq. (2),

0.6

n Value

X t X 1  X 0 =2

250 C

n Value

where Xt is the crystallinity at time t, Xo the fractional crystallinity initially present and X1 the nal crystallinity
achieved; n is the Avrami exponent whose value depends
on the crystallization mechanism [13], and Z the composite
rate constant.
Since

244 C

0.4
0.2
0
0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

Extent of Crystallization

Fig. 7. Change in instantaneous n values with extent of annealing. j


Averaged and actual I dXt/dt values with error bars.

where dXt/dt is the instantaneous rate of crystallization at


time t.
dXt/dt is the greatest source of error in determining
instantaneous values of n by Eq. (4) due to the errors in

measuring Xt as outlined in Appendix 1. Accordingly


smoothened values of dXt/dt were used by tting best t
values of the equation,

n t  dX t =dt=1  X t ln1  X t

dXt=dt aX b
t

Table 1
Avrami rate parameters for annealing.
Annealing temperature/C

244

250

252

n Value 0.2
Rate constant Z/sn 104
Xo
X1
Half life t1/2/s  103

0.67
19.5
0.31
0.49
18.2

0.79
9.59
0.14
0.39
4.28

1.03
0.00756
0.047
0.23
64.3

where a and b are numeric constants to the rate-time data.


Values of a and b were determined for each annealing temperature and used to determine dXt/dt as a function of Xt.
These were used to calculate n as a function of Xt as shown
in Fig. 9 along with the individual data points.
The kinetic data at 252 C had too much experimental
scatter for any meaningful conclusions to be made and

239

Z. Chen et al. / European Polymer Journal 50 (2014) 235242


40
35

100 min

30
10 min

25

1.5 min

20

3.4. Effect of annealing on melting

15
10
0 min

5
0
235

240

245

250

255

260

265

270

Temperature / C
Fig. 8. Change in melting endotherm with time at 250 C (0 min refers to
the starting material crystallized at 234 C for 50 min).

70
0.4

50

250 C

40
30

0.2

252 C
20

Fractional Crystallinity , Xt

Heat of Fusion / J g-1

60

10
0

0.5

1.5

2.5

0.0

log (time / min)


Fig. 9. The change in crystallinity with time on annealing.

the n value on recrystallization at 252 C was considered to


be constant at 1.03 0.1 over the time range measurements were made. As shown in Fig. 7, the n values at
250 C started above 1.0 and fell with conversion to about
0.5 0.1 after 2050% conversion; it remained more or less
constant after this. The overall change in the fractional
crystallinity, 0.13, in the initial region corresponded closely
to the amount of material which had melted 0.12 0.02
prior to annealing and the initial process is associated with
the recrystallization of this material.
An n value of 1.0 in terms of the mechanism of crystallization adopted by Avrami [13] is consistent with one
dimensional growth of a constant number of rods and so
is due to the growth of pre-existing lamellae which survived the melting process at the annealing temperature.
The growth appears to be restricted to the volume created
on melting. The nal n value of 0.5 0.1 was essentially
constant from 20% to 80% of the process. At 244 C the value of n was effectively constant at 0.5 from the start and
there appears to be no recrystallization of melted material.
Values of n = 0.5 have been reported for diffusion controlled crystal growth [14].

The rates of crystallization above 240 C were too low to


be determined directly by DSC as quenching from the melt
produced no discernible crystallization exotherms but only
a return of the calorimeter to the baseline. In order to repeat the above annealing studies PET was quenched to
234 C from the melt and crystallized for 50 min until the
primary stage had been completed. It was then heated to
250 and 252 C, the annealing temperatures, for various
periods of time, although partial melting occurred within
13 min no re-crystallization was directly observed by
DSC. Instead the development of crystallinity with time
was measured from changes in the melting endotherms,
see Fig. 8, and from the increase in heat of fusion as shown
from Fig. 11; in this the heat of fusion of crystalline PET
was assumed to be 150 Jg1 [7] and the fractional crystallinities were similar to those determined by FTIR
spectroscopy.
The melting endotherm of the original sample is included in Fig. 8 for comparison. On heating to 250 C part
of the original sample had melted leaving a residue which
melted above 254 C within a narrow range of about 5 C.
Further annealing produced more crystalline material
which melted at progressively higher temperatures and
over broader melting ranges, from 5 to 10 C. Little of the
original lamellae persist in samples annealed to the full extent and most melt some 79 C higher.
The temperature corresponding to the last trace of crystallinity was used as a measure of the melting point, Tm.
The melting points were observed, see Fig. 10, to increase
rapidly with time over the initial period of melting and
recrystallization but became linear with time after
150 min corresponding to the change in mechanism

275

252 C
270

Melting Point / C

Relative Heat Flow, mW mg-1

There appear to be two crystallization mechanisms


associated with annealing; an initial one whose extent is
limited and which appears to be due to the recrystallization of the material melted prior to annealing. This is followed by a second process exhibiting the characteristic of
diffusion controlled growth.

1000 min

250 C

265

260

255

200

400

600

800

1000

Time / min
Fig. 10. Change in melting point with time on annealing at 250 C.

240

Z. Chen et al. / European Polymer Journal 50 (2014) 235242


25

10 min
3 min

Relative Weight

20
1.5 min

15

0 min

10

10

12

14

16

Stem Length / monomer units


Fig. 11. The change in stem length with time at 250 C on partial melting
and annealing over a short period of time (0 min refers to the starting
material crystallized at 234 C for 50 min).

observed above. It was also observed that material annealed at higher temperatures had consistently higher
melting points at the same annealing time and at each
temperature the melting point increased with time.
In a previous publication [3] we have attributed the
thermal stability of the lamellae to their thickness, in that
the thicker the lamellae the higher the melting point, Tm In
particular,

T m T om 1  2RT m lnn=nDh . . .

T om

where
is the equilibrium melting point, Dh the heat of
fusion per monomer unit, R the gas constant and n the
number of monomer units per lamellae stem. For each value of Tm in the melting range n was determined as outlined before [3] and the melting endotherms converted
into a weight distribution of lamellae stem lengths, see
Figs. 11 and 12. It can be seen from these distributions that
the sample crystallized at 234 C for 50 min. had stem
lengths from about 5 to 14 monomer units and on heating
to 250 C the lamellae had lengths from about 5 to 11. On
subsequent crystallization and annealing the lamellae
thickened progressively but the change with time was
more obvious above 200 min.
The increase in stem length with increasing fractional
crystallinity is shown in Fig. 13. There are clearly two

1000 min

Relative Weight

40

200 min

500 min

100 min

30

50 min

20
0 min

10

10

15

20

25

30

3.5. Diffusion control


Lamellar thickening during secondary crystallization in
PET [3] has been shown to be diffusion controlled since
lamellar stem growth was dependent on the square root
of time and the rate increased with temperature consistent
with a thermally activated process obeying an Arrhenius
relationship. The measured stem lengths, in nm, plotted
against the square root of the annealing time is shown in
Fig. 14 and exhibits the same dependence suggesting that
annealing is occurring by the same mechanism.
The second stage of annealing, as dened by the linear
portion of the extent of crystallization against log time,
as shown in Figs. 3 and 5, were also analyzed as diffusion
30

Stem Length / monomer units

50

processes present with very different behavior with time;


an initial process in which the crystallinity develops with
little change in stem thickness and a secondary one in
which stem thickening is more rapid with increasing crystallinity and indeed was sufcient to account for the increase in crystallinity. The initial process is attributed to
recrystallization nucleated on the lamellar residue left
after partial melting and limited by the amount of melting
which had taken place.
This growth occurred in the plane of the lamellae and
perpendicular to the thickness of the lamellae as observed
in primary stage of crystallization and associated with the
formation of spherulitic structures. The lamellae thickness
is then determined by nucleation control and in particular
by the degree of supercooling, i.e. 1=T om  T A , where T om is
the equilibrium melting point and TA the annealing temperature. Since the temperature was increased from 234
to 2502 C the lamellae will be thicker than in the original
sample. The increase in crystallinity is due to the extension
in the length along the plane of the lamellae and limited by
the volume of amorphous material produced on partial
melting. This proceeds up to 0.3 when the fractional crystallinity prior to partial melting is achieved.
In the later process the lamellae thicken with time and
the increase in the average stem length more than accounts for the increase in crystallinity observed.

35

40

Stem Length / monomer units

25

20

15

10
0.15

0.2

0.25

0.3

0.35

0.4

0.45

Fractional Crystallinity
Fig. 12. The change in stem length with time at 250 C on further
annealing (0 min refers to the starting material crystallized at 234 C for
50 min).

Fig. 13. The changes in average stem lengths with increasing crystallinity, on annealing at 250 C.

Z. Chen et al. / European Polymer Journal 50 (2014) 235242


35

during the secondary crystallization of the isothermal crystallization at 234 C. The data was displaced by the amount
of fractional crystallinity corresponding to the onset of the
stem thickening process (see Fig. 15).

30

Stem Length / nm

252 C
25

4. Conclusions

20
o

250 C
15

10

10

15

20

25

30

35

(Crystallization Time)1/2 / min1/2


Fig. 14. The dependence of the lamellar stem length on square root of
annealing time.

controlled process by plotting the change in fractional


crystallinity and amorphous content against the square
root of time in Fig. 15. The growth of fractional crystallinity
and the decrease in amorphous content rates, measured as
the slope, see Table 2, increased with temperature and
implied that it was a thermally activated process. The rates
were also comparable with the rate of change in the
fractional crystallization with the square root of time

Table 2
Diffusion control.
Annealing
temperature/C

Intercept fractional
crystallinity

Rate constant
104/s

Degree of
t R2

250

0.777a
0.254b
0.661a
0.377b
0.727a
0.307b

8.46
6.67
6.77
4.83
6.59
4.51

0.995
0.992
0.995
0.996
0.987
0.988

244
234
a

Fractional amorphous content.


Fractional crystallinity.

Amorphous
Fraction, Xa

0.7
o

250 C

0.6

234 C
o

244 C

0.5

Crystalline
Fraction, Xt

241

0.4

244 C
o

234 C
o

250 C

0.3
160

180

200

220

240

260

280

Square Root of Time, s1/2


Fig. 15. Increase in crystalline and decrease in amorphous fraction with
square root of time during the nal stages of annealing.

Yeh et al. [2] have listed three well separated temperature ranges in which different mechanisms are operational
in the annealing of partially crystalline polymers. These are
rstly refolding of the polymer crystals in the solid state by
which the lamellae thicken by migration of chain defects
within thin lamellae, secondly melting and recrystallization of the thicker lamellae already present in the original
sample and thirdly partial melting of the lamellae surface
inwards towards the core along the chain axis followed
by recrystallization at the annealing temperature producing thicker lamellae. We have not observed these different
mechanisms.
Ageing partially crystalline polymers [4,12] at temperatures well below their crystallization temperature was observed not to increase .the melting point of the polymer, as
would be apparent if stem thickening had occurred, but instead the increase in crystallinity was due to the production of thin lamellae which melted at temperatures
below the observed melting of the sample created in the
original crystallization. Moreover the melting point of the
thin lamellae increased with annealing temperature in line
with Hoffman-Weekes nucleation theory [15].
Annealing above but close to the crystallization temperature did not result in any detectible melting of the polymer
and was accompanied by a progressive increase in crystallinity due entirely to the thickening of the lamellae along
the chain direction. This process was diffusion controlled
in that growth increased linearly with square root of the
annealing time. The thickness increased with annealing
temperature and with time and had all the characteristics
of secondary crystallization observed previously in PET [3].
Annealing within the temperature range of melting was
accompanied rst by melting of the polymer and then by
recrystallization limited in extent by the amount of melting which had occurred. The rate to which this developed
was dependent on the amount of residual lamellae which
survived at the annealing temperature as these acted as
seeds. The initial n value observed in the kinetics analysis
was 1.0 0.1 and consistent with the re-growth of lamellae
in their original form [16,17] but with increased thickness
consistent with the increased temperature. This re-growth
is followed by one of thickening of the lamellae as outlined
above and with diffusion control of growth.
The dependence of the increase in stem length on the
square root of the time implies increasing restriction to
growth as the lamellae develop into adjacent amorphous
regions and reduction in mobility of the chains as the
lamellae surfaces approach one another. Translational
mobility in the amorphous inter-lamellar regions of a
chain is restricted by entanglements with other chains
and according to reptation theory [1821] can only escape
from these entanglements by Brownian motion within a
virtual tube. The mobility of the chain, S, reptating in a

242

Z. Chen et al. / European Polymer Journal 50 (2014) 235242


1.03

0.7
o

244 C

1.02
o

234 C

0.5
1.01
0.4

Xc,t + Xa,t

Crystalline Absorbance, Ac,t

0.6

250 C

0.3

0.99

0.2

0.98

0.1
0
0.5

0.55

0.6

0.65

0.7

0.75

0.8

0.85

0.97
0.4

0.9

Amorphous Absorbance, Aa,t

1/2

1-dimensional tube is, S = (2kTft) 6. where k is Boltzman


constant, T the temperature, f the coefcient of friction and
t time. Translational mobility along the chain will be
increasingly hindered by the approach of crystalline regions and entanglements and non-crystallizable species
will be slower to diffuse away from the growth face leading
to diffusion control of the rate.
The lamella thickening observed during annealing has
all the characteristics of secondary crystallization in PET
and it would be interesting to see how general it is in other
polymer systems.
Acknowledgment
The Authors acknowledge the technical support of Mr.
F. Biddlestone.
Appendix A

The fractional crystallinity was calculated from the


measured absorbances of the crystalline and amorphous
bands at 1717 and 1717 cm1 at each temperature by
assuming a two phase model of amorphous and crystalline
domains with weight fractions Xc,t and Xa,t respectively at
time t, such that

1a

and that Beers-Lambert law is obeyed by each absorption


band, i.e.,

2a

where Ac,t and Aa,t are the measured absorbances of the


crystalline and amorphous bands at time t, e1 and e2 are
the relevant extinction coefcients, C the relevant concentrations and l the sample thickness.

Accordingly; X c;t Ac;t =Ac;o and X a;t Aa;t =Aa;o

0.7

0.8

0.9

Fig. A2. Internal consistency of the analysis. Xc,t + Xa,t = 1.00.

where Ac,o and Aa,o are the absorbances of crystalline and


amorphous bands for completely crystalline and completely amorphous samples respectively.
Substituting Eqs. (3a)into (1a)and rearranging gives,

Ac;t Ac;o  Aa;t Ac;o =Aa;o

4a

and a plot of Ac,t against Aa,t is linear with slope equal to


(Ac,o/Aa,o) and intercept Ac,o. Plots for annealing at all the
temperatures studied are shown in Fig. A1 from which
the temperature dependent constants, Ac,o and Ac,o, were
determined from the intercept and slope.
They were used to determine the fractional crystallinity
and amorphous content in the annealing experiments.
The consistency of the analysis was measured by the
accuracy with which Eq. (1a)was fullled, i.e. the sum of
the amorphous and crystalline fractions should be 1.00.
Over the range of temperatures studied, 234 to 252 C this
was consistent to 2%, see Fig. A2.
References

A.1. The determination of fractional crystallinity

Ac;t e1 C Tc;t l and Aa;t e2 C a;t l

0.6

Amorphous Fraction, Xa,t

Fig. A1. Relationship between absorbances of crystalline and amorphous


bands on annealing at various temperatures.

X c;t X a;t 1:0

0.5

3a

[1] Fischer EW. Pure Appl Chem 1972;31:113.


[2] Yeh GSY, Horsemann R, Loboda-Cackovic J, Cackovic H. Polymer
1976;17:309.
[3] Chen Z, Jenkins MJ, Hay JN. Eur Polym J.
[4] Biddlestone F, Harris A, Hay JN, Hammond T. Poly Int J 1996;39:221.
[5] Barham PJ, Keller A. J Polym Sci 1986;24:69.
[6] Scandola M, Ceccorulli G, Pizzoli M. Macromolecules 1992;25:6441.
[7] de Koning GJM, Lemstra P. Polymer 1992;33:3295. 1993; 34: 4089.
[8] de Koning GJM, Scheeren AHC, Lemstra PJ, Peeters M, Reynaers H.
Polymer 1994;35:4598.
[9] Chen Z, Hay JN, Jenkins MJ. Eur Polym J 2013;49:1723.
[10] Lu XF, Hay JN. Polymer 2001;42:9423.
[11] Kong Y, Hay JN. Polymer 2003;44:623.
[12] Chen Z, Hay JN, Jenkins MJ. Eur Polym J 2012;48:1586.
[13] Avrami M. J Chem Phys 1939;7:1102. 1940; 8: 212; 1941; 9: 177.
[14] Schultz J. Polymer materials science. Englewood Cliffs, NJ: PrenticeHall; 1974.
[15] Hoffman JD, Weeks JJ. J Res Nat Bur Stand 1962;A73:64.
[16] Hoffman JD, Weeks JJ, Murphey WM. J Res Nat Bur Standards, A
1959;63:67.
[17] Banks W, Gordon M, Sharples A. Polymer 1964;4:289.
[18] Lauitzen JT, Hoffman JD. J Res Nat Bur Standards, A 1960;64:73.
[19] De Gennes PG. J Chem Phys 1971;55:572.
[20] Edwards SF, and Doi M, The Theory of Polymer Dynamics, Oxford
University Press. Oxford 1988.
[21] Doi M, Edwards SF. J Chem Soc, Faraday Trans 2, 1978;74:1978.

Vous aimerez peut-être aussi