23.11.2014 Views

Ph.D. thesis (pdf) - dirac

Ph.D. thesis (pdf) - dirac

Ph.D. thesis (pdf) - dirac

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.

78 Alpha Relaxation<br />

for the 1 s isochrone in [Paluch et al., 2003 a]. Along this 100 s isochrone we calculate<br />

α τ and determine<br />

d log e(ρ)<br />

dlog ρ<br />

from<br />

d log e(ρ)<br />

d log ρ<br />

= 1/(T g |α τ |). From this we determine<br />

ln(e(ρ)) (up to an additive constant) by integrating dln(e(ρ))<br />

ln ρ<br />

d(ln(ρ)), and ln(e(ρ))<br />

is finally converted by e(ρ) = exp(ln(e(ρ))). This gives e(ρ) with an arbitrary<br />

multiplicative constant. The quality of the data collapse is independent of this<br />

constant, as it only depends on the density dependence of e(ρ) not on its absolute<br />

value. In figure 5.7 we show the data as a function of e(ρ)/T, using the e(ρ) found<br />

from this construction. The constructed e(ρ) has an apparent “power-law” exponent<br />

x(ρ) = dloge(ρ)/dlogρ that increases from 1.5 to 3.5 with density in the range<br />

considered. It is seen that this makes the higher density data fall reasonably well<br />

on the master curve and that gives an even better collapse in the low density region<br />

than with the scaling with a simple power law.<br />

As a last note regarding the e(ρ)/T-scaling in figure 5.7, we want to stress that we<br />

cannot test the scaling (Eq. 3.2.1) in the density range above 1.25 g/cm 3 where<br />

there is only one set of data. Indeed, with a unique set of data in a given range of<br />

density it is always possible to construct e(ρ) in this range to make the data overlap<br />

with data taken in other density ranges.<br />

We have determined the ratio between the isochoric fragility and the isobaric fragility<br />

at atmospheric pressure by calculating α τ along the isochrone of 100 s and inserting it<br />

in equation 3.1.5. This leads to m P /m ρ ≈ 1.2, when m P is the atmospheric-pressure<br />

fragility. In figure 5.8 we show the isobaric data taken at atmospheric pressure and<br />

at 230 MPa scaled by their respective T g (P). No significant pressure dependence of<br />

the isobaric fragility is observed when going from atmospheric pressure to 230 MPa,<br />

which is consistent with the result of Sekula et al. [2004]. The pressure independence<br />

of m P is connected to the relatively low value of m P /m ρ = 1.2, (typical values<br />

are 1.2-2 [Alba-Simionesco et al., 2002]); m ρ is pressure independent and the ratio<br />

m P /m ρ cannot be lower than one (see Eq. 3.1.5), so that m P can at most decrease<br />

by 20% from its atmospheric-pressure value. Moreover, the increase in dln(e(ρ))<br />

ln ρ<br />

with<br />

density will tend to cancel the decrease in α P T g , which is usually responsible for the<br />

decrease in fragility with increasing pressure.<br />

5.2.2 m-Toluidine<br />

The glass transition temperature at atmospheric pressure is T g = 187 K and the<br />

isobaric fragility based on dielectric spectra is reported to be m P = 82±3 [Mandanici<br />

et al., 2005; Alba-Simionesco et al., 1999]. (There has been some controversy on the<br />

dielectric relaxation in m-toluidine at atmospheric pressure, see [Carpentier et al.,

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

Saved successfully!

Ooh no, something went wrong!