Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
net
ABSTRACT: The influence of molecular weight, M, on the fragility and fast dynamics
in polyisobutylene (PIB) was studied using dielectric and mechanical relaxation spec-
troscopies, calorimetry, and Raman spectroscopy. The measurements indicate a
decrease in fragility with increasing M for shorter chains, in the range of M where
Tg is M-dependent. Such behavior is not observed for other polymers and is at odds
with traditional theoretical models that predict an increase in fragility with chain
length. These results confirm the unusual character of PIB, as evident in various
properties including extremely low gas permeability, a low fragility, and a segmental
relaxation spectrum much broader than expected for a low-fragility material. The
reason for this anomalous behavior remains unclear, but might be related to the sym-
metric structure of the PIB repeat unit, together with comparable flexibility of both
structural components, the backbone and side groups. V C 2008 Wiley Periodicals, Inc.
(iv) a mechanical segmental dispersion much model12 and more recently in the model of Dudo-
broader than expected for such a low fragility wicz et al.14 The common idea of both
material8; and (v) a very unusual spectrum of approaches is that fragility relates to the sever-
the fast dynamics; that is, a ‘‘constant loss’’ re- ity of steric constraints from neighboring seg-
gime where the susceptibility changes negligibly ments, which impede segmental motions12 and/
with frequency.9 These examples illustrate that or frustrate chain packing.14 The packing idea
PIB can be viewed as a very unusual polymer leads to the expectation that packing of rigid
with properties deviating from behavior common chains or chains with bulky side groups will be
for many other polymers. more frustrated as M increases. As a result,
The glass transition in supercooled liquids is such polymers should exhibit a strong increase
associated with rapid changes of the rate of in fragility with increasing molecular weight.
structural relaxation (segmental relaxation pro- Flexible chains should pack better and, accord-
cess in polymers). The segmental relaxation ing to the model,14 should show a much weaker
time, sa , increases by many orders of magnitude increase in fragility with M. No theory predicts
upon cooling only a few degrees at temperatures a decrease of fragility with molecular weight.
close to the glass transition temperature, Tg. sa However, preliminary studies of PIB pre-
in polymers exhibits a strongly non-Arrhenius sented in ref. 15 indicated a decrease in fragility
temperature dependence, which is well-approxi- with increase of the chain length. This would
mated by the Vogel–Fulcher–Tamman (VFT) suggest that PIB again exhibits a trend opposite
equation as follows: to the pattern of other polymers and to the theo-
retical expectation. We present herein a detailed
sa ¼ s0 exp½B=ðT T0 Þ ð1Þ study of the dynamics of PIB using mechanical
and calorimetry measurements, dielectric relax-
where s0 and B are material constants and T0 is ation, and Raman scattering spectroscopy. The
the temperature of an ‘‘ideal’’ glass transition. results demonstrate a decrease of fragility with
The concept of fragility has been introduced to increase in molecular weight for PIB. Possible
characterize the steepness of the temperature reasons for the unusual behavior of PIB are dis-
variations of sa close to Tg. Essentially, fragility cussed.
characterizes the degree of deviation of the tem-
perature dependence of sa from Arrhenius-like
behavior (i.e., deviation from thermally acti-
EXPERIMENTAL
vated behavior with a constant activation
energy). A ‘‘strong’’ material has a near Arrhe-
Materials
nius temperature dependence of sa , whereas a
‘‘fragile’’ material exhibits strongly non-Arrhe- PIB samples were precision synthesized by liv-
nius behavior. To quantify the fragility, Boehmer ing carbocationic polymerization in the Depart-
et al.10 defined the fragility (steepness) index m ment of Polymer Science at the University of
as follows: Akron. The polymerization of isobutylene was
initiated by 2-chloro-2,4,4-trimethyl-pentane/
@ log sa TiCl416 and terminated using trimethyl alumi-
m¼ : ð2Þ
@ðTg =TÞT¼Tg num (Aldrich)17 to obtain PIBs with only C and
H atoms. 1H NMR analysis of the resulting
In general, polymers tend to be more fragile PIBs showed that they were exclusively termi-
than molecular liquids. It has been observed nated by CH3 groups (synthesis details will be
that polymers with a rigid backbone or bulky reported elsewhere). This is important for our
pendant groups, such as polystyrene (PS) and study to exclude any chain-end interactions.
poly(methyl methacrylate) (PMMA), exhibit a Four different molecular weight PIB samples
strong increase in fragility with increase in mo- (Table 1) were synthesized using this procedure.
lecular weight.11,12 On the other hand, flexible
polymer chains such as poly(dimethyl siloxane)
Dynamic Mechanical Analysis
(PDMS) and poly(methyl phenyl siloxane)
(PMPS) show less or no increase in fragility Dynamic mechanical data were obtained using a
with molecular weight.13 These results were Bohlin VOR with a parallel plate geometry. Isother-
explained in the framework of the coupling mal measurements were made over a 10–12 8C
Journal of Polymer Science: Part B: Polymer Physics
DOI 10.1002/polb
1392 KUNAL ET AL.
0.51 6 0.01
0.45 6 0.01
0.33 6 0.02
range of temperatures, corresponding typically
(Imin/Imax)
to 0.96 Tg/T 1.01, with 0.1 K control. The
Fast Dynamics
N.A
R1
frequency was varied according to the tempera-
ture within the range 0.0001–2 Hz to obtain the
peak in the loss modulus, G00 , within the mea-
surement window. Strains were between 0.02
(cm1) and 0.5%. Figure 1(a) presents G00 spectra for
20.46
23.65
25.50
26.94
mBP
47
data.
4
4
4
5
Dielectric Spectroscopy
6
6
6
6
DS
44
38
36
36.5
The VFT parameters were obtained from the fitting of the dielectric relaxation data.
56
57
56
56
187
6
6
6
6
128.5
127.5
128.4
121.6
Raman Spectroscopy
340
23
30
56
6
6
6
6
2500 (1.07)
1100 (1.09)
(Mn/Mw)b
GPC.
BTg
m¼ ð4Þ
ðTg T0 Þ2 ln 10
gility with increasing M. PIB is also one of the conformational states (trans and gauche) have
least fragile polymers.10 The reasons for these essentially the same energy.34 As a result,
peculiarities are not understood. Ngai and changes in temperature do not significantly affect
Roland12 proposed that the symmetry of the the population of conformational states of the
monomer and the absence of bulky pendant chain molecule. This peculiar property of PIB
groups play a role in the fragility of polymers might also contribute to its unusual behavior.
through weakened intermolecular cooperativity. We need to emphasize that other polymers
According to the coupling model, polymers with such as PDMS also have a symmetric structural
smooth symmetric chain structure have weak unit and no tacticity. However, PDMS exhibits
intermolecular constraints on local motions that commonly observed polymeric behavior and has
leads to lower m.12 This concept alone cannot a much higher fragility.15 So, the roles of sym-
account for the comparison of PIB to PDMS, metry and tacticity remain to be fully investi-
since the latter has symmetric side groups but a gated. In particular, the conformational energies
much higher fragility. The failure might be of the PDMS rotameric states differ signifi-
related to the need to consider additional ther- cantly, so that the molecule changes its equilib-
modynamic variables, as has been addressed in rium conformation with temperature.34 Al-
recent developments of the coupling model.29,30 though PDMS and PIB have the same methyl
Another possible explanation is the absence of side groups, their backbones have very different
tacticity in PIB chains. Atactic asymmetric poly- flexibilities. For PDMS the characteristic energy
mer chains have many different structures dic- barrier for conformational changes of the back-
tated by the stereoconfiguration of the repeat bone is less than 2 kJ/mol,35 whereas methyl
units. This results in many possible conforma- group rotation has a barrier 7 kJ/mol.36 Thus,
tional states, each with its corresponding config- methyl groups in PDMS can exert steric con-
urational energy, leading to a broad distribution straints on the motions of the (otherwise flexi-
of configurational energy minima. A broad dis- ble) backbone. In contrast, the methyl groups in
tribution of energy minima is usually associated PIB have an energy barrier, 14 kJ/mol,37
with fragile behavior31,32 (although the topology which is comparable to the energy barrier for
of the energy landscape is also strongly influ- conformational transitions of the backbone, E
enced by the heterogeneous molecular packing). 18–25 kJ/mol.35,37 This implies that conforma-
Since PIB has a symmetric repeat unit, the dis- tional transitions of the PIB backbone occur on
tribution of configurational energies might be in- a time scale similar to the side group motions,
dependent of chain length.33 Moreover, it has thus coupling their respective dynamics.38 Evi-
been established from analysis of n-alkanes that dently, the relative flexibility of the backbone
the CC rotational energy barrier increases and pendant groups can exert a significant influ-
slightly with increase in chain length, from ence on the fragility of polymers.
3.0 kcal/mol in ethane to 3.6 kcal/mol in butane. This idea agrees with the recent theoretical
A similar increase of rotational energy barrier work14 that relates fragility to polymer chain
can also be expected in PIB. We speculate that packing: polymers that pack well (e.g., flexible
the slight increase of rotational barrier with chain chains without bulky side groups) are expected
length without any increase in the distribution of to be strong, whereas polymers with poor pack-
energy minima might lead to a slight decrease in ing (e.g., rigid chains and those with bulky side
fragility with molecular weight in PIB. This spec- groups) are expected to be fragile. The packing
ulation is supported by the observation that the model does not take into account the chain tac-
VFT temperature T0 appears to be rather inde- ticity, which can also affect packing. A stereore-
pendent of molecular weight (Table 1). Thus, an gular polymer should also pack better, since ran-
increase in molecular weight leads to an increase dom arrangement of pendant groups would
in Tg without significant change in T0. This increase disorder in the system. The packing of
results in an ‘‘effective’’ decrease in fragility with PIB is known to be extremely high (which con-
M (eq 4), suggesting the unusual molecular tributes to its low air permeability). In particu-
weight dependence of fragility in PIB might be lar, Van Krevelen and Hoftyzer39 calculated the
related to the unusual independence of T0 on M. density of many polymers using the group incre-
An important thermodynamic property of PIB ment method and compared it with experimen-
that once again differentiates this material from tally measured density. They applied this
many other polymers is the fact that different approach to 85 amorphous polymers and found
Journal of Polymer Science: Part B: Polymer Physics
DOI 10.1002/polb
1398 KUNAL ET AL.
satisfactory agreement with experimental data, tional states may all contribute. These factors
with the mean deviation between calculated and might give rise to a distribution of configura-
experimental values being 1.3%. However, in tional energies that is independent of chain
their analysis PIB stands out having the strong- length, and thus the anomalous molecular weight
est deviation: its measured density (0.913 g/cm3) dependence of the fragility of PIB. It would be
is 7% higher than the calculated value (0.851 g/ important to discover another polymer with simi-
cm3). Thus, the packing of PIB appears to be lar properties, to clarify the microscopic origin of
much higher than the expected from model cal- the behavior of PIB. At least one conclusion is
culations, which provide reasonable results for obvious: PIB is an ironic choice to illustrate the
many other polymers. general properties of polymers. However, con-
We speculate that the similar flexibility of the trasting the unusual behavior of PIB to other pol-
PIB side groups and backbone and the lack of ymers helps to develop a deeper understanding
atacticity result in the extremely tightly packed of the regular behavior of polymeric materials.
structure. Usually, packing efficiency decreases
with M. This frustration in packing, according The authors thank P. G. Santangelo (NRL) for experi-
to the model,14 would cause the increase in fra- mental assistance, and C. G. Robertson (Bridgestone
gility with M. However, PIB may be exceptional, Americas) and C. Alba-Simionesco (University Paris
because its packing actually improves with Sud) for many helpful and stimulating discussions.
increasing molecular weight. This idea is sup- The Akron University team thanks the NSF Polymer
ported (indirectly) by the observed variations in program for financial support (grant DMR-0605784).
M. Paluch acknowledges the financial support of the
the fast dynamics [Fig. 2(b)]. A decrease in the
Committee for Scientific Research, Poland (KBN,
QES intensity (relative to the boson peak inten- Grant No. N202 147 32/4240) and C. M. Roland the
sity) is usually ascribed to a decrease in the am- support of the Office of Naval Research.
plitude of fast conformational fluctuations (rat-
tling in a cage formed by neighboring mono-
mers).24 Thus, the observed decrease of R1 with
molecular weight in PIB is consistent with REFERENCES AND NOTES
tighter packing of PIB chains.
1. Marvin, R. S. In Proceedings of the 2nd Interna-
tional Congress on Rheology; Harrison, V. G. W.,
Ed.; Butterworths: London, 1954; pp 156–164.
CONCLUSIONS 2. Ferry, JD. In Viscoelastic Properties of Polymers;
Wiley: New York, 1980; p 482.
This analysis demonstrates the decrease in fra- 3. Boyd, R. H.; Pant, P. V. K. Macromolecules 1991,
gility with increasing molecular weight for PIB 24, 6325–6331.
with M 2500 g/mol, that is, in the range 4. Plazek, D. J.; Zheng, X. D.; Ngai, K. L. Macromo-
where Tg exhibits strong dependence on molecu- lecules 1992, 25, 4920–4924.
lar weight. This behavior is opposite to that 5. Ngai, K. L.; Plazek, D. J. Rubber Chem Tech Rub-
ber Rev 1995, 68, 376–434.
commonly observed in polymers and is not pre-
6. Frick, B.; Richter, D.; Trevino, S. Phys A 1993,
dicted by any theory. The deviation from the 201, 88–94.
usual pattern adds to other known peculiarities 7. Ngai, K. L.; Plazek, D. J Macromol 1990, 23,
in the properties of PIB: low gas permeability, 4282–4287.
one of the lowest fragility among polymers, a 8. Santangelo, P. G.; Ngai, K. L.; Roland, C. M. Mac-
segmental relaxation dispersion that is much romolecules 1993, 26, 2682–2687.
broader than expected for strong glass-formers, a 9. Sokolov, A. P.; Novikov, V. N.; Kisliuk, A.; Ngai,
small difference between the time–temperature K. L. Phys Rev B 2001, 63, 172204.
shift factors for segmental and chain relaxations, 10. Boehmer, R.; Ngai, K. L.; Angell, C. A., Plazek, D. J.
and so forth. These observations suggest that J Chem Phys 1993, 99, 4201–4209.
11. Casalini, R.; Roland, C. M.; Capaccioli, S. J Chem
PIB may be the most unusual polymer, although
Phys 2007, 126, 184903.
the reasons for its peculiar behavior require fur-
12. Ngai, K. L.; Roland, C. M. Macromolecules 1993,
ther study. The similar flexibility of the backbone 26, 6824–6830.
and side groups, the absence of conformational 13. Roland, C. M.; Ngai, K. L. Macromolecules 1996,
effects due to the symmetric structure of the 29, 5747–5750.
repeat unit and lack of tacticity, and the near 14. Dudowicz, J.; Freed, K. F.; Douglas, J. F. J Phys
equivalence of the energy for different conforma- Chem B 2005, 109, 21285–21292.
Journal of Polymer Science: Part B: Polymer Physics
DOI 10.1002/polb
PROPERTIES OF POLYISOBUTYLENE 1399
15. Ding, Y.; Novikov, V. N.; Sokolov, A. P.; Cailliaux, 27. Sokolov, A. P.; Novikov, V. N.; Ding, Y. J Phys:
A.; Dalle-Ferrier, C.; Alba-Simionesco, C.; Frick, Condens Matter 2007, 19, 205116.
B. Macromolecules 2004, 37, 9264–9272. 28. Qin, Q.; McKenna, G. B. J Non-Cryst Solids 2006,
16. Puskas, J. E.; Lanzendörfer, M. G. Macromole- 352, 2977–2985.
cules 1998, 31, 8684–8690. 29. Ngai, K. L.; Casalini, R.; Roland, C. M. Macromo-
17. Kennedy, J. P.; Sivaram, S. J Macromol Sci, 1973, lecules 2005, 38, 4363–4370.
A-7, 969–989. 30. Ngai, K. L. J Non-Cryst Solids 2005, 351, 2635–2642.
18. Robertson, C. G.; Santangelo, P. G.; Roland, C. M. 31. Zheng, Q.; Durben, D. J.; Wolf, G. H. Angell, C. A.
J Non-Cryst Solids 2000, 275, 153–159. Science 1995, 254, 829–832.
19. Puskas, J. E.; Chen, Y.; Kulbaba, K.; Kaszas, G.; 32. Debenedetti, P. G.; Stillinger, F. H. Nature 2001,
Soleymannezhad, A. J Polym Sci Chem 2006, 44, 410, 259–267.
1770–1776. 33. Suter, U. W.; Saiz, E.; Flory, P. J Macromolecules
20. Roland, C. M.; Paluch, M.; Rzoska, S. J. J Chem 1983, 16, 1317–1328.
Phys 2003, 119, 12439–12441. 34. Erman, B.; Mark, JE. In Structures and Proper-
21. Saltzman, E. J., Schweizer, K. S. J Chem Phys ties of Rubberlike Networks; Oxford University
2004, 121, 1984–2000. Press: New York, 1997; pp 126.
22. Ngai, K. L.; Sokolov, A. P.; Steffen, W. J Chem 35. Somoza, M. M.; Sluch, M. I.; Berg, M. A. Macro-
Phys 1997, 107, 5268–5272. molecules 2003, 36, 2721–2732.
23. Sokolov, A. P.; Rdossler, E.; Kisliuk, A.; Quit- 36. Mukhopadhyay, R.; Alegra, A.; Colmenero, J.;
mann, D. Phys Rev Lett 1993, 71, 2062–2065. Frick, B. Macromolecules 1998, 31, 3985–3993.
24. Novikov, V. N.; Ding, Y.; Sokolov, A. P. Phys Rev 37. Karatasos, K.; Saija, F.; Ryckaert, J. P. Physica B
E 2005, 71, 061501. 2001, 301, 119–125.
25. Roland, C. M.; Casalini, R. J Chem Phys 2003, 38. Karatasos, K.; Ryckaert, J. P.; Ricciardi, R.; Lau-
119, 1838–1842. pretre, F. Macromolecules 2002, 35, 1451–1462.
26. Santangelo, P. G.; Roland, C. M. Macromolecules 39. Krevelen, D. W.; Hoftyzer, P. J. J Appl Polym Sci
1998, 31, 4581–4585. 1969, 13, 871–881.