Phase Transitions II Order - Disorder Transition: - Brass, Theory Vs Experiment

Download as pdf or txt
Download as pdf or txt
You are on page 1of 5

Lecture 16: Phase Transitions II

Continuous Phase transitions Aims:


Mean-field theory continued:
Order-disorder transitions: Ferromagnetic transition in zero field. Curie Weiss Law.

Order-disorder transition: -brass, Theory vs experiment


Heat capacity (continued from lecture 15):
A heat capacity anomaly is observed. Mean field theory predicts such behaviour and describes the shape correctly at a qualitative level.
cp/kJ K kg 0 0.5 1.0
-1 -1

Universality. Landau theory, (not for examination).


-2 T=0.6Tc -1

1 T=1.2Tc 2 x

200

400 T/C

600

Increasing order

Note: Calorimetry measurements are often used to show the presence of phase transitions.

Order parameter:
T/Tc

0 0.2 0.4 0.6 0.8 1.0 1.2

High-T: paramagnet
May 05

Low-T: ferromagnet
Lecture 16 1

The order parameter agrees qualitatively with experiment. In all cases, the theory is least good close to Tc, where the effects of fluctuations are most important (N.B. they are ignored by the model).
May 05 Lecture 16 2

Ferromagnetism
Weiss theory of ferromagnetism
The spontaneous polarisation, in zero external field, arises as a result of interactions between spins. The Weiss theory represents the interaction as a mean, internal field. Bint = <>.

Ferromagnetic-paramagnetic transition
Temperature dependence of the magnetic moment
Above Tc there is no magnetic order Below Tc ferromagnetism

We have, from our previous analysis of paramagnetism (lecture 6), a relationship between <> and B, which must be satisfied. = tanh (B kT ) B B We get the behaviour at Bext=0 from a second equation (from A, above, with Bext=0 ):

A A

<>

B = Bint + Bext = + Bext

T/Tc 0 0.2 0.4 0.6 0.8 1.0 1.2

Bint

kT B kT

C C

Above Tc

Equations B and C are the same as we obtained for -brass! The solution is the same

System behaves like a paramagnet Curie-Weiss Law

k (T Tc )

May 05

Lecture 16

May 05

Lecture 16

Universality
Beyond the mean-field theory: meanMean-field theory is relatively simple. It is best used to describe systems where fluctuations are unimportant and/or where there are long range interactions. A good example is the transition in type-1 superconductors. In practice, interactions are often short range and fluctuations near Tc occur on all length scales. Here, the mean-field approach breaks down.

An approach to Landau Theory (Not for examination)


Introduction to non-linear dynamics: nonThe approach to equilibrium in a simple, 1-D system can be described by a non-linear differential equation F (x) = d f d x x = f (x)
Order parameter, x Order parameter, x Thermodynamic potential Thermodynamic potential

Universality:
Very different systems exhibit similar behaviour indicating an underlying universality. This is the subject of Landau theory (and, when fluctuations are included, Ginzburg-Landau theory). Both covered in Part II/III Physics. The phenomenon arises from the limited number of ways in which functions (Thermodynamic potentials) can be modified to create/destroy the minima associated with different phases. See final question on examples sheet 2.
May 05 Lecture 16 5

Equilibrium occurs when x = f ( x ) = 0 Starting with the simplest, generic form for f(x). f(x) = rx , where r is a parameter. Only one equilibrium state exists. .
. r>0 x r <0 x x x

Unstable equilibrium Unstable equilibrium

Stable equilibrium Stable equilibrium

A quadratic f allows for two equilibrium states: f(x) = r + x2. A bifurcation occurs as r changes from negative to positive x at equilibrium
. x . x x* x r<0
Lecture 16

x at equilibrium Unstable Unstable


r

x r >0
May 05

Stable Stable

Pitchfork bifurcations (Not for examination)


The simplest system to show multiple, stable, equilibrium positions is: f(x) = rx-x3 (final question on examples sheet 2)
. x x r >0 r<0 . x x x* r

Discontinuous transitions (Not for examination)


Subcritical pitchfork
In the previous example the cubic term in f(x) acted to stabilise the system. If we reverse its sign then we need to add a further stabilising term in x5. (N.B. the symmetry of F is retained) x = f ( x ) = rx + x 3 x 5

Unstable Unstable

Stable Stable Stable

As r changes sign, we move from a single stable equilibrium to two symmetrical, stable equilibrium positions. It is known as a supercritical pitchfork. Integrating f(x) gives the corresponding potentials. (Recall F(x) = - dF/dx) 1 1 F ( x ) = rx 2 + x 4 + const 2 4
F( x) r>0 x F( x) r=0 x F(x) r<0 x

The potential follows from F(x) = - dF/dx.

1 1 1 F ( x ) = rx 2 x 4 + x 6 + const 2 4 6
r>0 F(x ) x r s<r<0 F(x ) x r<rs<0 F( x) x

These are precisely the forms found in a continuous phase transition.


May 05 Lecture 16 7

The transition is discontinuous and exhibits hysteresis as can be seen from the bifurcations diagram (below).
May 05 Lecture 16 8

Discontinuous transitions (Not for examination)


Discontinuous phase transitions
The behaviour in a sub-critical pitchfork is closely related to that of a class of phase transitions known (for historical reasons) as first-order transitions. Consider moving from A to B and back.
x* A r B

X X

Route A to B is different from B to A. An aside: notice that the bifurcation marked X is of the same type as that on p. 6.

Classification of phase transitions


The examples above show how continuous and discontinuous transitions arise necessarily from the fundamental properties of polynomials. Landau (1937) was the first to express these ideas in a general form, though his approach pre-dates the treatment given above.
May 05 Lecture 16 9

You might also like