Actually, through the literature researches and group discussion, we find that Curie-Weiss Law is an empirical formula to determine the phase transition order based on the fitting values of To and C (To and C are the CW temperature and CW constant). There is no a clear line to determine the first and second order phase transition. Sometimes, it can give a vague conclusion. And the valid temperature region of CW law is difficult to define.
Instead, we find that the method using slopes of the 1/Er-T curve below and above Curie temperature (denoted as S- and S+) is much easier to determine the order of phase transition: if S-/S+=-4, it is the first order; if the value equals -2, it's the second order.
That's our understanding about the phase transition order.
[i] In some ferroelectric substances, the temperature dependence of ε can be accurately represented by Curie- Weiss [CW] law up to reasonable accuracy as:
-ε= ε0+C/(T-T0); C( Curie’s constant) and To(CW temperature), which , in general, differs from Curie’s temperature(Tc)
[ii] ε becomes very high when it reaches T0
[iii] In some ferroelectric substances,Tc= To and the phase transition is of the second order while In others , Tc and To differ very largely so that the phase transition is of the first order.
You may find the following reference of some worth regarding temperature range for applying Curie-Weiss Law to ferroelectrics
by M Trainer - 2000 - Cited by 15 - Related articles
susceptibility of strontium titanate in the 90–300 K temperature range. By measuring ... This is the same form of expression as the Curie–Weiss law ... doing this was to set up a Microsoft Excel spreadsheet and use the Linest or Polyfit function.
Click/ Find:
Eur. J. Phys. 21 (2000) 459–464. Printed in the UK PII: S0143-0807(00)12475-4
Ferroelectrics and the Curie–Weiss law
Matthew Trainer
Kelvin Building, Department of Physics and Astronomy, University of Glasgow, Glasgow