part with high epoxy equivalent weight (CE 4400) (Table 2). Gen-
erally, more viscous materials generate higher shear in a mixer
when they are kneaded by the rotor blades of the mixer. This
additional shear contributes to increased heat, resulting in shorter
times to reach the set-point temperature. As the CE content
increased, the viscosity of the materials increased, leading to
increased shear which contributed to raised temperatures, resulting
in reduced times to reach 200
C, irrespective of the CE grade.
Moreover, the CE grade with lower epoxy equivalent weight
resulted in higher shear due to its higher reactivity and thus
reached 200
C in shorter time compared to its counterpart with
higher epoxy equivalent weight, regardless of CE content. The
increased viscosity of the materials due to the addition of CE and
the difference between the reactivities of the two CE grades will
be further discussed in the following sections.
The reaction between multifunctional epoxies and polyes-
ters such as PLA has been extensively studied in the literature
[28, 32
–35]. PLA has two distinct functional end-groups, i.e., car-
boxyl and hydroxyl, that can react with the epoxy groups in
multifunctional epoxies. The reaction proceeds via ring-opening
of the epoxide group leading to the formation of new secondary
hydroxyl groups and ester linkages when reaction occurs with
PLA
’s carboxyl groups and/or ether linkages with PLA’s
hydroxyl groups (Fig. 4) [28, 32
–35]. Fourier transform infrared
spectroscopy (FTIR) was used to gain an in-depth understanding
of these reaction mechanisms.
The FTIR spectra of neat PLA, CE 4468, and PLA/CE blends
compounded in an internal mixer are illustrated in Fig. 5. Table 3
summarizes the band assignments for the wavenumbers of
peaks found in these spectra to their corresponding functional
FIG. 3.
Typical curves of stock temperature and mixing torque of PLA blends with CE 4468 [(a) and (c)] as well as
CE 4400 [(b) and (d)] as a function of time generated by a torque rheometer.
TABLE 2. Mixing time of PLA with both grades of multifunctional epoxies in
torque rheometer.
CE (%)
Time to reach 200
C (s)
Total time at 200-211
C (s)
CE 4468
a
CE 4400
a
CE 4468
a
CE 4400
a
0
180.3
5.1
A
180.3
5.1
A
119.8
5.1
A
119.8
5.1
A
0.25
166.1
15.8
B
175.0
8.9
A,B
133.9
15.8
B
125.0
8.9
A,B
0.5
163.5
7.2
B,C
174.6
1.7
A,B
136.5
7.2
B,C
125.4
1.7
A,B
0.75
156.5
2.4
B,C
171.5
0.6
B
143.5
2.4
B,C
128.5
0.6
B
1
150.0
5.4
C
155.5
4.7
C
150.0
5.4
C
144.5
4.7
C
a
Same superscript letters within the same column are not signi
ficantly dif-
ferent based on the ANOVA results at 5% signi
ficance level.
4 POLYMER ENGINEERING AND SCIENCE—2019
DOI 10.1002/pen
groups [4, 32, 33]. Two distinct trends were observed. Firstly, the
peaks at 844, 905, and 1,250 cm
−1
in CE 4468 spectrum, assigned
to the C O stretching of epoxy group, disappeared in the spectra
of PLA compounded with CE, regardless of CE content (Fig. 5).
This result indicates that the CE reacted with PLA through ring-
opening reaction of epoxy groups as illustrated in Fig. 4 [32, 33].
Secondly, the peaks present at 1,489 and 1,597 cm
−1
in CE 4468
(Fig. 5), corresponding to C C stretching in phenyl group
(Fig. 1), merged into a broad shoulder in PLA/CE compounds at
1,513
–1,547 cm
−1
(Fig. 5). It should be noted that the phenyl
groups in CE 4468 do not participate in the reaction with PLA
and could hence be used to detect the presence of CE in the
blends since these groups are absent in neat PLA. Thus, the for-
mation of a broad shoulder from 1,513 to 1,547 cm
−1
in the
blends con
firmed the presence of CE. As expected, this broad
shoulder seemed to intensify as the CE content increased into
PLA matrix probably due to more phenyl groups in the blends
with higher CE contents.
The equilibrium torque taken at 300 s in Figs. 2 and 3 (end tor-
que value) is proportional to the apparent viscosity and molecular
weight of the materials and depends on polymer chain structure
such as branching [28
–30]. An increase in the end torque values
indicates an increase in the viscosity and molecular weight due to
the formation of longer and/or branched chains [18, 19, 28
–30].
Consequently, the end torque values could be used to monitor the
chain extension reaction of a polymer.
Figure 6 illustrates the end torque values of PLA chain-
extended
with
low
and
high
epoxy
equivalent
weight
multifunctional epoxies obtained from their torque versus mixing
time curves (Fig. 3) to determine their effectiveness and ef
ficiency
in chain extending PLA. Both multifunctional epoxies chain-
extended PLA effectively since they signi
ficantly increased the
end torque during mixing. However, the ef
ficiency of the
multifunctional epoxy grades as chain extenders varied and was
dependent on the epoxy equivalent weights. Chain extender
’s effi-
ciency is de
fined as the contribution to increase the end torque of
the blend per unit of chain extender [36] as expressed in the fol-
lowing equation:
CE efficiency =
End torque
CE content
ð3Þ
It is evident from Fig. 6 that the additive with lower epoxy
equivalent weight (CE 4468) was more ef
ficient in increasing the
end torque of PLA compared to its counterpart with high epoxy
equivalent weight (CE 4400) due to the smaller amount of CE
needed to increase the end torque at a similar value. For example,
by drawing a horizontal line parallel to the x-axis (chain extender
FIG. 4.
Generalized reaction mechanism of a multifunctional epoxy chain extender with the hydroxyl and carboxyl
end groups in PLA.
FIG. 5.
Infrared spectra of PLA, CE 4468, and PLA blended with different
CE contents in an internal mixer. Absorbance in the y-axis is in arbitrary
units.
TABLE 3. Band assignments for the wavenumbers of peaks used in FTIR
analysis and corresponding functional groups.
Wavenumber (cm
−1
)
Peak assignments
[4, 32, 33]
PLA
CE 4468
PLA blends
with 0.25, 0.5,
0.75, and 1% CE
1,744
1,721
~1,745
Stretching of C O
-
1,489, 1,597
1,513
–1,547
Stretching of C C in phenyl
1,451
1,447
~1,451
Scissoring of CH
3
1,263
-
1,263
Stretching of C O in carboxyl and
C O C stretch
-
1,250
-
Stretching of C O in epoxy
951
-
~952
Rocking of CH
3
and stretching
of C C
-
844, 905
-
Stretching of C O in epoxy
865
-
~865
Stretching of C-O-C
DOI 10.1002/pen
POLYMER ENGINEERING AND SCIENCE—2019 5
content) at an arbitrary end torque value (e.g., 8 Nm in Fig. 6),
the data can produce two intersection points [36]. The
first inter-
section point detected at lower chain extender content for CE
4468 (0.5%) and the next seen at higher chain extender content
for CE 4400 (1%), which clearly indicates the higher ef
ficiency of
CE 4468 in chain extending PLA, as expected from Eq. 3 [36].
This result can be explained by the lower equivalent weight of
CE 4468, resulting in more epoxy groups per chain and thus
higher reactivity. Since the ef
ficiency of CE with a lower epoxy
equivalent weight (CE 4468) was higher than its counterpart with
high epoxy equivalent weight (CE 4400), it was then selected for
further studies.
Chain extension of PLA with the most ef
ficient grade of CE
(CE 4468) was further con
firmed by the values of molecular weights
and dispersity indices of PLA and PLA/CE blends (Table 4). The
addition of CE increased the number (M
n
), weight (M
w
), and viscos-
ity (M
v
) average molecular weights, as well as polydispersity index
(DI) of PLA, due to the formation of longer and/or branched chains.
Similar results have been reported for PLA chain-extended with other
non-food grades of multifunctional epoxies [9, 19, 32] and polyethyl-
ene terephthalate chain-extended with 1,6-Diisocyanatohexane [37],
among others.
As previously mentioned, chain extension and/or branching
can also be monitored by the zero-shear viscosity (
η
0
) of a poly-
mer melt, which is directly related to its molecular weight (M
w
)
by the Mark-Houwink equation:
η
0
= k
× M
a
w
ð4Þ
with k and a as Mark-Houwink constants.
Figure 7 shows a plot of zero-shear viscosities determined
from the MFI data (Table 4) as a function of weight-average
molecular weights (M
w
) of PLA compounded with different con-
tents of CE (Table 4) in an internal mixer. The viscosity increased
with molecular weight at different extents below and above a
molecular weight value of 475 kDa, which corresponds to CE
content of 0.5% (Table 4). Below 475 kDa, increasing the molec-
ular weight of PLA/CE blends from 212 to 475 kDa gradually
increased their viscosities from 1,086 to 7,270 Pa-s, respectively.
This trend is not clearly seen in Fig. 7 due to a wide range of vis-
cosity values on the y-axis, which was overcome by truncating
and displaying this range of the graph as an insert. In contrast, a
sharp increase in viscosity was observed above 475 kDa, where
the viscosity of the blends increased from 7,270 to 240,727 Pa-s
when the molecular weight varied from 475 to 583 kDa, respec-
tively. This point of in
flection indicates the critical molecular
weight, that is, the onset of chain entanglements attributable to
the high degree of branching. Dalsin and coworkers reported a
similar trend for atactic polypropylene wherein a weak depen-
dence of zero-shear viscosity on its molecular weight was
reported until a critical molecular weight of 8.5 kDa, after which
the zero-shear viscosity increased exponentially with increasing
molecular weight [38].
FIG. 6.
End torque of PLA chain-extended with low (CE 4468) and high
(CE 4400) epoxy equivalent weight multifunctional epoxies as a function of
their contents.
TABLE 4. Molecular weights, dispersity indices, and melt
flow index values
of PLA chain-extended with CE 4468.
CE (%)
Molecular weight (kDa)
DI
MFI (g/10 min)
M
n
M
w
M
v
0
123
5
212
6
198
6
1.7
0.0
11.51
0.30
0.25
139
3
340
34
296
25
2.4
0.2
5.94
0.03
0.5
157
4
475
44
407
59
3.0
0.4
1.62
0.00
0.75
181
12
503
14
433
9
2.8
0.3
0.22
0.00
1
221
10
583
28
469
15
2.6
0.1
0.05
0.00
FIG. 7.
Zero-shear viscosities as a function of weight-average molecular
weights of PLA/CE 4468 blends processed in an internal mixer. Insert shows
the viscosity of these blends at low-molecular-weight values.
FIG. 8.
Effect of CE 4468 content on the crystallinity of PLA processed in
an internal mixer.
6 POLYMER ENGINEERING AND SCIENCE—2019
DOI 10.1002/pen
The degree of crystallinities of PLA and its blends were also
evaluated to corroborate chain branching (Fig. 8) observed
through the molecular weight and viscosity data. Generally, dense
packing of branched polymer chains is dif
ficult, leading to
reduced crystallinity. As expected, the addition of CE into the
PLA matrix reduced its degree of crystallinity indicating
branching and con
firming chain extension, consistent with the
trend reported in the literature [39].
Extrusion-Blown Chain-Extended PLA Films
The feasibility of utilizing the identi
fied most efficient CE
grade (CE 4468) in extrusion-blown PLA
film processing was
assessed by dry-blending it with PLA
first and then attempting
film manufacturing. The melt stock temperature and pressure at
the die generated during processing of the blends with various CE
concentrations (up to 1 wt%) were recorded and listed in Table 5.
The melt experienced a stock temperature higher than the rec-
ommended 200
C (Table 5) due to shear generated during
processing. Notice that
film manufacture was feasible only with
blends containing up to 0.5% CE, becoming unprocessable above
this content due to the increased viscosity at this critical molecular
weight for chain entanglement (Fig. 7), indicated by the increased
melt pressure during processing (Table 5). The melt pressure in the
extruder (Table 5), suggestive of the blend
’s shear stress, increased
with increasing chain extender content attributable to the chain
extension reaction between the PLA and CE [18, 19, 30]. The
increased molecular weight (Table 6) and decreased crystallinity
(Fig. 9) of PLA
films with chain extender content corroborated the
chain extension and branching of PLA. As expected, these results
were similar to those obtained for PLA/CE blends processed in the
internal mixer (Table 4 and Fig. 8).
To ensure that the melts experienced the recommended 200
C
within 120 s for 99% reaction to occur between PLA and CE, the
residence times of processable blends inside the extruder were
recorded. Figure 10 shows typical RTD curves of PLA
films
blended with 0.25 and 0.5% CE. The residence time, peak time,
and end time were the three parameters extracted from each RTD
curve generated (Table 5). The residence time is de
fined as the
TABLE 5. Processing parameters recorded during extrusion-blown
film and residence times as a function of CE 4468 content in PLA.
CE (%)
Melt temperature
a
(
C)
Pressure
a
(10
4
Pa)
Residence time
b
(s)
Peak time
b
(s)
End time
b
(s)
0
211
0
465
3
-
-
-
0.25
211
0
496
33
132
6
156
10
246
6
0.5
211
0
571
10
176
24
228
30
290
28
0.75
NOT PROCESSABLE
1
a
Average values of at least 45 data points.
b
Average values of three replicates.
TABLE 6. Molecular weights and dispersity indices of PLA
films
chain-extended with CE 4468.
CE (%)
Molecular weight (kDa)
DI
M
n
M
w
M
v
0
105
7
183
2
170
1
1.7
0.1
0.25
112
5
232
2
208
2
2.1
0.1
0.5
127
9
243
16
219
12
1.9
0.3
0.75
NOT PROCESSABLE
1
FIG. 9.
Effect of CE 4468 content on the crystallinity of PLA
films.
FIG. 10.
Typical residence time distribution curves of PLA with (a) 0.25% CE 4468 and (b) 0.5% CE 4468.
DOI 10.1002/pen
POLYMER ENGINEERING AND SCIENCE—2019 7
first time at which a tracer signal of 5% of the maximum peak
height was detected. It relates to the minimum exposure time and
duration for PLA to react with CE in the extruder. The peak time,
where the maximum signal occurred, represents the average tran-
sit or mean processing time. The end time, the time for detection
of the last signal of 5% of the maximum peak height, is linked to
the maximum time the blends resided in the extruder [24, 25].
PLA/CE blends experienced residence times longer than the
recommended 120 s at 200
C, suggesting 99% completion of
reaction. The residence time of PLA blended with 0.5% CE was
longer than its counterpart with 0.25% (Table 5), due to its higher
viscosity (Fig. 7), which provided more resistance to
flow
resulting in slower movement of the blend through the extruder;
thus higher peak and end times (Table 5 and Fig. 10). It is also
worth mentioning that the maximum time that the materials stayed
in the extruder (end time) was more than twice the 120 s and the
melt experienced a temperature of 11
C above the 200
C rec-
ommended by the manufacturer (Table 5 and Fig. 10). This
clearly indicates that the blends were exposed to extrusion
processing conditions (heat and time) that favored chain extension
reactions in the
films.
Effect of Chain Extender Content on the Dart Impact Strength of PLA
Films
Figure 11 shows the failure mass, indicative of the falling dart
impact strength of PLA
films chain-extended with various
amounts of CE 4468 (low epoxy equivalent weight). Chain exten-
sion of PLA
film was found beneficial in overcoming its brittle-
ness. As seen in Fig. 11, the impact strength of PLA
film
increased almost linearly with the chain extender content, imply-
ing that the
film became more ductile by branching the polymer
chain and reducing its crystallinity (Fig. 9). Interestingly, the addi-
tion of only 0.5% chain extender into the PLA
films increased its
failure mass by ~56%, due to the chain extension reaction. These
results validate the need for chain extenders to reduce the brittle-
ness of PLA
films even though neat PLA films without any chain
extenders have been successfully manufactured in this and our
previous studies [3
–6].
CONCLUSIONS
The effectiveness and ef
ficiency of FDA-approved food grade
polymeric chain extenders (CE) with low (CE 4468) and high
(CE 4400) epoxy equivalent weights in chain extending PLA
were studied using a torque rheometer in the
first part of this
work. Based on the experimental results, the following conclu-
sions were drawn:
1. The residence time and melt stock temperature exceeded at
least 120 s recommended by the manufacturer for the reaction
between PLA and CE to be over 99% complete at 200
C,
clearly indicating that the blends were processed at conditions
favoring chain extension reaction. This chain extension
proceeded via the ring opening reaction of epoxide groups in
the CE with PLA
’s hydroxyl and/or carboxyl groups, as con-
firmed by Fourier transform infrared spectroscopy.
2. Both CE grades were effective at chain extending PLA because
they signi
ficantly increased PLA’s torque during mixing. The
torque increase was related to increased melt viscosity caused
by molecular weight increase. Additionally, the reduced crys-
tallinity of the blends compared to neat PLA indicated chain
branching of PLA. Nevertheless, the CE with lower epoxy
equivalent weight (CE 4468) was more ef
ficient in increasing
the torque of PLA compared to its counterpart with high epoxy
equivalent weight (CE 4400) due to its higher reactivity.
Secondly, the processabilities of PLA
films chain-extended
and branched with various amounts of the most ef
ficient CE were
assessed, and the following conclusions were drawn:
1. Film manufacture was only feasible with blends containing up
to 0.5% CE but dif
ficult above this content due to chain entan-
glement leading to signi
ficant increase in melt viscosity. Blends
with 0.25 and 0.5% CE experienced residence times and melt
temperatures of at least 132 s and 211
C in the extruder,
respectively. Exposure to these processing conditions favored
99% completion of the chain extension reaction, con
firmed by
the increased melt pressure in the extruder as well as the
increased molecular weight and decreased crystallinity of
PLA/CE blends compared to neat PLA.
2. Chain branching reduced crystallinity and improved the ductil-
ity of PLA
films. The addition of only 0.5% CE into the PLA
films increased its failure mass by ~56%.
The developed ductile PLA/CE
films have tremendous poten-
tial for food packaging applications.
ACKNOWLEDGMENTS
This work is supported by the USDA National Institute of
Food and Agriculture, McIntire Stennis, Project # 1017725.
REFERENCES
1. J. Muller, C. Gonzalez-Martinez, and A. Chiralt, Materials, 10,
952 (2017).
2. National Geographic. Available at: https://news.nationalgeographic.
com. Accessed October, 2019.
3. S.S. Karkhanis, N.M. Stark, R.C. Sabo, and L.M. Matuana,
J. Appl. Polym. Sci., 134, 45212 (2017).
4. S.S. Karkhanis, N.M. Stark, R.C. Sabo, and L.M. Matuana,
Polym. Eng. Sci., 58, 1965 (2017).
5. S.S. Karkhanis, N.M. Stark, R.C. Sabo, and L.M. Matuana, Com-
pos. Part A. Appl. Sci. Manuf., 114, 204 (2018).
6. S. Vijayarajan, S.E.M. Selke, and L.M. Matuana, Macromol.
Mater. Eng., 299, 622 (2014).
7. C. Xing and L.M. Matuana, J. Appl. Polym. Sci., 133, 43201
(2016).
FIG. 11.
Failure mass of PLA
films with various CE 4468 contents.
8 POLYMER ENGINEERING AND SCIENCE—2019
DOI 10.1002/pen
8. L.M. Matuana, S.S. Karkhanis, N.M. Stark, and R.C. Sabo, Bio-
tech, Biomater. Biomed. TechConnect Briefs, 3, 1 (2016).
9. B. Mallet, K. Lamnawar, and A. Maazouz, Polym. Eng. Sci., 54,
840 (2014).
10. B. Mallet, K. Lamnawar, and A. Maazouz, Key Eng. Mater., 554,
1751 (2013).
11. M. Mihai, M.A. Huneault, and B.D. Favis, Polym. Eng. Sci., 50,
629 (2010).
12. L.M. Matuana and C.A. Diaz, Ind. Eng. Chem. Res., 49, 2186
(2010).
13. R. Al-Itry, K. Lamnawar, and A. Maazouz, Polym. Degrad.
Stabil., 97, 1898 (2012).
14. L. Lim, T. Cink, and K.T. Vanyo, Poly(Lactic Acid) Synthesis,
Structures, Properties, Processing, and Applications, John
Wiley & Sons Publication, Hoboken, NJ, 191 (2010).
15. S. Pilla, S.G. Kim, G. Auer, K.S. Gong, and C.B. Park, Polym.
Eng. Sci., 49, 1653 (2009).
16. Y. Di, S. Iannace, E. Di Maio, and L. Nicolais, Macromol.
Mater. Eng., 290, 1083 (2005).
17. J. Tuominen, J. Kylma, and J. Seppala, Polymer, 43, 3 (2002).
18. W. Dong, B. Zou, Y. Yan, P. Ma, and M. Chen, Int. J. Mol. Sci.,
14
, 20189 (2013).
19. A. Jaszkiewicz, A.K. Bledzki, and A. Meljon, Polym. Degrad.
Stabil., 101, 65 (2013).
20. United States Department of Agriculture. Available at: www.ers.
usda.gov. Accessed October, 2016.
21. E. Brun, Expert Forecast on Emerging Chemical Risks Related to
Occupational Safety and Health, Of
fice for Official Publications
of the European Communities, Luxemburg, 33 (2009).
22. O. Faruk and L.M. Matuana, J. Vinyl Addit. Techn., 14, 60 (2008).
23. L.M. Matuana and J.W. Kim, J. Vinyl Addit. Techn., 13, 7 (2007).
24. J.
Wang,
C.
Thurber,
M.
Bishop,
D.
Monaenkova,
E. Marchbanks, H. Kim, E. Keene, and X. Chen, SPE Antec
2017
– Anaheim, California, 1168 (2017).
25. J. Wesholowski, A. Berghaus, and M. Thommes, Pharmaceutics,
10
, 49 (2018).
26. R. Levinson, P. Berdahl, and H. Akbari, Sol. Energy Mater. Sol.
Cells, 89, 351 (2005).
27. Y. Liu and L.M. Matuana, J. Appl. Polym. Sci., 136, 47594
(2019).
28. L.C. Arruda, M. Magaton, R.E.S. Bretas, and M.M. Ueki, Polym.
Test., 43, 27 (2015).
29. J. Liu, L. Lou, W. Yu, R. Liao, R. Li, and C. Zhou, Polymer, 51,
5186 (2010).
30. F.N. Cavalcanti, E.T. Teo
filo, M.S. Rabello, and S.M.L. Silva,
Polym. Eng. Sci., 47, 2155 (2007).
31. Y.B. Tee, R.A. Talib, K. Abdan, N.L. Chin, R.K. Basha, and K.
F.M. Yunos, Agric. Agri. Sci. Proc., 2, 289 (2014).
32. V. Ojijo and S.S. Ray, Polymer, 80, 1 (2015).
33. Y. Wang, C. Fu, Y. Luo, C. Ruan, Y. Zhang, and Y. Fu,
J. Wuhan University of Technology-Mater. Sci. Ed., 25, 774
(2010).
34. Y. Tanaka and H. Kakiuchi, J. Polym. Sci. A1., 2, 3405 (1964).
35. X. Wang, J. Mi, and H. Zhou, J. Mater. Sci., 54, 3863 (2019).
36. K.A. Afrifah and L.M. Matuana, Macromol. Mater. Eng., 295,
802 (2010).
37. P. Raffa, M.B. Coltelli, S. Savi, S. Bianchi, and V. Castelvetro,
React. Funct. Polym., 72, 50 (2012).
38. S.J. Dalsin, M.A. Hillmyer, and F.S. Bates, Macro. Lett., 3, 423
(2014).
39. J. Ludwiczak and M. Kozlowski, J. Polym. Environ., 23, 137
(2015).
DOI 10.1002/pen
POLYMER ENGINEERING AND SCIENCE—2019 9
Dostları ilə paylaş: |