Humid ageing of polyetherimide: Chemical and physical interactions with water - Polyimides and Other High Temperature Polymers - page 262

that the temperature dependence of D can be described by two plateaus, at T ≥

80°C and T 40°C separated by a transition zone. There is, however, no physical

argument for this interpretation. The second one consider that non-linearities re-
sult from experimental uncertainties and that D obeys, in fact, the Arrhenius law.
A complementary test in an autoclave at 120°C, 0.2 MPa gave results in favor
of the latter hypothesis (Fig. 5). The regression coefficient is R = 0.9861. The ap-
parent activation energy is: E D = 43 kJ.mol -1 .
3.2.3. Discussion
The heat of dissolution H s is made up of two contributions: heat of mixing H m and
heat of condensation H p . Two extreme hypotheses can, therefore, be envisaged:
(a) Water occupies only the free volume. In this case, the “dissolution” is essen-
tially due to the re-formation of water-water bonds and, then, H m = 0 and H s =
H p . There is, in fact, no dissolution of water in the matrix.
(b) Water is entirely dissolved into the matrix. In this case, the dissolution is es-
sentially due to the formation of water-polymer bonds, and there are no water-
water bonds so that H p = 0, H m = H s .
The fact that H s ~ H p results from the similarity of water-water and water-
polymer bonds.
Certain authors, for instance Adamson [14] have suggested that water absorp-
tion by polar polymers such as epoxies would correspond to case (a); however,
the following arguments are rather in favor of case (b):
(i) The fact that dissolution is very exothermic (H m closer to H p than to 0) can be
predicted from the structure-property relationships available for polymer-
solvent interactions [12]. The hypothesis that H m = 0 would disagree with
these relationships.
(ii) H s increases, as expected in case (b), with the polymer polarity [10, 11, 13].
In the case where water occupies only the free volume, one would expect H s =
H p , independently of the polymer structure.
(iii) In case (a), water and polymer would form two phases; in this case, DSC is
expected to show the exotherm (or endotherm) of water crystallization (or
melting) at 0°C. This exotherm was never observed in our experiments.
(iv) In case (b) water plasticizes PEI and the extent of plasticization, as deter-
mined from the T g variations, agrees with the hypothesis that all the water and
polymer form one phase into the matrix [8]. In case (a) there would be no
plasticization.
(v) More generally, hypothesis (a) fails to explain why certain materials in which
the free volume fraction is high, for instance apolar liquids, are almost hydro-
phobic.
It can thus be concluded that water is essentially dissolved into the matrix and
that the dissolution is due to the strong polarity of the water molecules. In other
words, there is a critical polymer polarity, for which H s = H p . For polymers of

≤