27.10.2013 Views

PDF (119 KB) - Fisica

PDF (119 KB) - Fisica

PDF (119 KB) - Fisica

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

Discussion<br />

Comment on ‘‘Decrease in the configurational entropy during a<br />

meltÕs polymerization’’ [Chem. Phys. 305 (2004) 231]<br />

S. Corezzi a,b, *, D. Fioretto b,c , P.A. Rolla d<br />

a Dipartimento di <strong>Fisica</strong>, Università di Roma ‘‘La Sapienza’’, P.le A. Moro 2, I-00185 Roma, Italy<br />

b INFM-CRS Soft, c/o Università di Roma ‘‘La Sapienza’’, P.le A. Moro 2, I-00185 Roma, Italy<br />

c Dipartimento di <strong>Fisica</strong> and INFM, Università di Perugia, Via A. Pascoli, I-06123 Perugia, Italy<br />

d INFM and Dipartimento di <strong>Fisica</strong>, Università di Pisa, Via F. Buonarroti 2, I-56127 Pisa, Italy<br />

In a recent paper, Johari [1] examines the manner in<br />

which, in a paper published by us [2], the reduction in configurational<br />

entropy during a meltÕs polymerization was<br />

connected to the number of chemical bonds, and a relationship<br />

that links the relaxation time to the advancement of<br />

reaction was derived. The author claims to have demonstrated<br />

that the connection we made is untenable on experimental<br />

grounds. He also suggests that the apparent success<br />

in fitting the relaxation time data may just be an artifact of<br />

an improper use of the Adam–Gibbs entropy equation [3].<br />

In this Comment, we discuss the main points of the JohariÕs<br />

argument, concerning the validity of our premise and its<br />

use in combination with the Adam–Gibbs equation.<br />

(1) To investigate the existence of a correlation between<br />

the dynamical slowdown during a meltÕs polymerization<br />

and the loss of configurations caused by the formation of<br />

chemical bonds, we have considered a convenient class of<br />

reactions and chosen convenient experimental conditions,<br />

in which such a correlation could be more clearly exposed.<br />

It is important for the following discussion to recall them<br />

here. We have considered step polymerizations where ring<br />

formation is negligible, and focused on slow reactions (i.e.,<br />

low T) to stay in thermodynamic quasi-equilibrium condition.<br />

A pronounced slowdown of the dynamics usually corresponds<br />

to a number-average degree of polymerization xn<br />

between two and few units (


with s 0 approximately independent of both T and a, and<br />

B =[C/TSc(0)] only dependent on temperature, on the<br />

assumption that the variation of C with a is negligible –<br />

i.e., the relaxation time data vs a should lie on a Vogel–Fulcher-like<br />

curve that extrapolates to infinity at a = a0. Contrary<br />

to what is done by Johari [1], the consequences of Eq.<br />

(1) should be analyzed within the physical conditions that<br />

allow this relation to hold valid. In particular, a tendency<br />

of S c to zero (i.e., of s to diverge) at a finite a 0, as extrapolated<br />

from data at a < a0, does not imply a true vanishing<br />

of Sc (i.e., a divergence of s) ata0. The reaction, in fact, is<br />

not prevented from departing from Eq. (1) and from proceeding<br />

to an extent greater than a0, if the conditions for<br />

Eq. (1) are progressively removed. As a matter of fact<br />

any polymerization can be brought to its full extent by progressively<br />

increasing the reaction temperature, and this is<br />

normally used to estimate the total heat of reaction via calorimetry.<br />

It is also possible when the reaction temperature<br />

is high that the system remains fluid at even full reaction, as<br />

recognized in [1].<br />

While data at a close to or higher than a0 are of no use<br />

for testing a correlation between Sc and s through a test of<br />

Eq. (2), a large amount of relaxation time data in the expected<br />

range of validity of Eq. (1) have been analyzed,<br />

from which an excellent agreement with Eq. (2) has turned<br />

out with the parameter a0 predicted on a molecular basis<br />

[2,6,7]. The observation that this relationship applies to a<br />

wide range of systems investigated, in different conditions,<br />

using different experimental methods, by numerous authors<br />

is hard to be seen as coincidental. Instead, it strongly suggests<br />

that the essence of the relaxational slowdown on polymerization<br />

can be traced back to a bond-controlled<br />

reduction in the number of configurations available to<br />

the system.<br />

Further experiments by our group – in which the characteristics<br />

of the systems investigated and the experimental<br />

conditions have been chosen to meet the requirements for<br />

Eq. (2), mentioned above – have strengthened the basis<br />

for Eq. (1) in combination with the Adam–Gibbs relation,<br />

particularly persuasive being the results for a number of<br />

epoxy–amine mixtures composed of two reagents (DGE-<br />

BA–DETA) in different molar ratios [8]. Because a0 may<br />

be controlled by varying the molar ratio of reagents in<br />

the initial mixture (solid line in Fig. 1), according to Eq.<br />

(2) the extent of reaction at which the relaxation time<br />

behavior appears to diverge would also be varied in the<br />

same way. We have found, for each reaction carried out<br />

isothermally, that Eq. (2) represents the relaxation time<br />

data well, in a wide relaxation range (at least five decades<br />

for s >10 4 s), where xn(a) is few units [8]. The derived values<br />

for the fitting parameter a0 are reported in Fig. 1. Indeed,<br />

Fig. 1 reveals that a0 not only changes with the<br />

mixture composition but (within the experimental error)<br />

actually follows the expected variation with the molar ratio<br />

(N a/N e) of the reacting molecules, in a wide N a/N e range.<br />

Moreover, the same a 0 is obtained by reacting a given mixture<br />

at different temperatures [9], despite the time-depen-<br />

S. Corezzi et al. / Chemical Physics 323 (2006) 622–624 623<br />

α 0<br />

1.0<br />

0.9<br />

0.8<br />

0.7<br />

0.6<br />

10:3<br />

5:2<br />

5:2.8<br />

10:9<br />

0.2 0.3 0.4 0.5 0.6<br />

Na /Ne 0.7 0.8 0.9 1.0<br />

Fig. 1. Variation of the parameter a 0 with the molar ratio N a/N e between<br />

the amino and the epoxy monomer constituents of the system DGEBA–<br />

DETA Ne:Na. The experimental points for different compositions are<br />

labelled by their Ne:Na values. Open symbols are used to distinguish a<br />

different reaction temperature. The solid line is a 0 =[1+(N a/N e)]/2,<br />

calculated from the functionality and the molar ratio of the reacting<br />

molecules.<br />

dent evolution of the reaction is greatly affected. This is<br />

shown in Fig. 1 by closed and open symbols. In the same<br />

figure, isoabscissa closed-points refer to reactions repeated<br />

at the same temperature to check reproducibility.<br />

(2) A number of other observations reported by Johari<br />

deserve our comment. It is asserted in [1] that a connection<br />

between the relaxation time and a by using the Adam–<br />

Gibbs equation can be made only by allowing Dl, and<br />

therefore, the quantity B / Dl to change with a, being<br />

Dl a free-energy barrier per cooperative unit. We notice<br />

that this is the opposite of what the same author maintained<br />

in previous studies of polymerization [10–12]. Apart<br />

from this contradiction, we observe that as long as multimers<br />

behave effectively like cooperative units a variation<br />

of Dl with a seems not to be required by the theory [3].<br />

On the other hand, if one would be forcing Eq. (2) with<br />

wrong but constant values of B, to fit the experimental data<br />

for polymerizations with Dl strongly varying with a, the<br />

values a0 which describe the data could scarcely correspond<br />

to the values of a physical quantity, i.e., the molecular<br />

parameter a0 calculated from the functionality of the reagents.<br />

This suggests that a sensible variation of Dl with<br />

a is not required either by the experimental data. A minor<br />

variation can, of course, account for details.<br />

Moreover, it is asserted in [1] that s 0 in Eq. (2) should<br />

necessarily correspond to the time scale of phonons, i.e.,<br />

10 13 –10 14 s. There is no question that in simple glassforming<br />

liquids analyses using the Adam–Gibbs approach<br />

usually do give pre-exponents not too far from the phonon<br />

time scale, though the values obtained notoriously span<br />

several decades (e.g., from 10 10.5 to 10 17.3 sin[14]). We<br />

remark that for it to be a general expectation the pre-exponent<br />

s 0 should represent the true value of s in the limit<br />

T !1, and different spectroscopic techniques (such as<br />

dielectric and photon-correlation spectroscopy) should<br />

measure the same s. Neither of these is the case. It is<br />

4:3


624 S. Corezzi et al. / Chemical Physics 323 (2006) 622–624<br />

well-known that relaxation times measured by different<br />

techniques all sensitive to structural dynamics may even<br />

differ by a few decades, which obviously reflects in different<br />

values of s0. It is also well-known that at some intermediate<br />

temperature on heating above Tg, the relaxation mechanism<br />

becomes different from cooperative transitions governed<br />

by the configurational entropy, and a marked<br />

change in the dynamics is observed [13]. Because of this<br />

change the relaxation time of a liquid at high temperature,<br />

which actually corresponds to the phonon time scale, may<br />

differ from s0 in Eq. (2), which is the value extrapolated<br />

from the cooperative regime. The difference between the<br />

two could be less pronounced if the process under analysis<br />

is a simple physical process. However, if it is a chemical<br />

process like polymerization then chemical step activation<br />

entropies could become involved and then all bets should<br />

be off about the value of the pre-exponent s 0.<br />

(3) In [1], the author reports as a feature of Eq. (1) that<br />

‘‘... CP, c of a polymerizing melt at a fixed T would not<br />

change as a increases from zero to one’’, and similarly,<br />

‘‘... the product bcVc, [where bc and Vc are the configurational<br />

contribution to the thermal expansion coefficient<br />

and the volume of the melt] would not change’’. On the contrary,<br />

we note that Eq. (1) yields a factor in C P, c and b c V c<br />

explicitly dependent on a [e.g., C P, c(a) =T(oS c(0)/oT) P<br />

(1 a/a0) =CP, c(0)(1 a/a0)], which factor has been erroneously<br />

omitted in [1]. Therefore, our Sc(a) relation does<br />

not conflict with the observation that changes in CP, b,<br />

and V during a meltÕs polymerization are also to be ascribed<br />

to changes in the configurational contribution to<br />

these quantities.<br />

(4) In our opinion, the calculations of S c performed by<br />

Wang and Johari [15] in a lattice–hole model are interesting<br />

but cannot prevail over the indications coming from experiments,<br />

as we have described. In addition, we observe that<br />

Sc, calculated with plausible values of the adjustable<br />

parameters, in the Wang and JohariÕs model would be<br />

expected to initially increase and then decrease with a in<br />

such a way that the value at the full extent of polymerization<br />

is nearly the same as (or even higher than) in the<br />

monomeric liquid (see Fig. 2 in [15]). This result can hardly<br />

be reconciled with an entropy-based picture of the process<br />

of polymerization; which picture, instead, Johari upholds.<br />

On the other hand, different lattice models exist and provide<br />

different results for the statistics of polymer chains.<br />

Interestingly, in a Flory–Huggins model of equilibrium<br />

polymerization [16] the average length of polymer chains<br />

scales inversely with the excess entropy of the polymer<br />

solution, similarly to our premise Sc / 1/xn.<br />

In conclusion, we believe there are sufficient experimental<br />

indications that Eq. (2) captures the principal physics<br />

behind the dynamical slowdown that occurs in a wide class<br />

of polymerizations, and therefore, goes beyond a mere phenomenological<br />

description of the process. We stress once<br />

more that it is appropriate in a limited range of a and T<br />

but this range usually covers the most part of relaxation<br />

data from experiments.<br />

Acknowledgement<br />

S.C. acknowledges financial support from MIUR-<br />

FIRB.<br />

References<br />

[1] G.P. Johari, Chem. Phys. 305 (2004) 231.<br />

[2] S. Corezzi, D. Fioretto, P. Rolla, Nature 420 (2002) 653.<br />

[3] G. Adam, J.H. Gibbs, J. Chem. Phys. 43 (1965) 139.<br />

[4] xn represents the average number of monomers per molecule in the<br />

system. a, also called the chemical conversion, provides a measure of<br />

the extent of reaction.<br />

[5] High temperature also tends: (i) to favor cyclization, which changes<br />

the variation of xn with a (e.g., xn would grow less rapidly and tend to<br />

diverge at a > a 0), and (ii) to favor departure from the quasiequilibrium<br />

condition.<br />

[6] S. Corezzi, L. Comez, D. Fioretto, J. Non-Cryst. Solids 345–346<br />

(2004) 537.<br />

[7] S. Corezzi, AIP Conf. Proc. 708 (2004) 604.<br />

[8] S. Corezzi, D. Fioretto, J.M. Kenny, Phys. Rev. Lett. 94 (2005)<br />

065702.<br />

[9] For all the polymerizations studied, the temperature was near the<br />

onset of the exothermic release in a calorimetric scan at the heating<br />

rate of 10 °C/min.<br />

[10] G.P. Johari, C. Ferrari, G. Salvetti, E. Tombari, Phys. Chem. Chem.<br />

Phys. 1 (1999) 2997.<br />

[11] E. Tombari, C. Ferrari, G. Salvetti, G.P. Johari, J. Phys. Condens.<br />

Matter 9 (1997) 7017.<br />

[12] C. Ferrari, E. Tombari, G. Salvetti, G.P. Johari, J. Chem. Phys. 110<br />

(1999) 10599.<br />

[13] F. Stickel, E.W. Fischer, R. Richert, J. Chem. Phys. 104 (1996) 2043.<br />

[14] R. Richert, C.A. Angell, J. Chem. Phys. 108 (1998) 9016.<br />

[15] J. Wang, G.P. Johari, J. Chem. Phys. 116 (2002) 2310.<br />

[16] J. Dudowicz, K.F. Freed, J.F. Douglas, J. Chem. Phys. 111 (1999)<br />

7116.

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!