第二章正式进入场论的计算。
平衡态巨正则系综可以表示成一个路径泛函。可以由场论方法非常自然地给出统计关系。
Chapter 2 Imaginary-Time Path Integral
2.1 The core formula in quantum statistic physics
The grand partition function is:
And for a quantum system with Halmitonian
For interacting system, it is very hard to calculate all eigen states and eigen energy. Generally, we can calculate the trace under any representation:
Then we can define the density matrix:
And for any operator
The equation (3) and (5) are the core formula in quantum statistic physics.
2.2 Thermodynamics of a harmonic oscillator
Harmonic oscillator is a simple but important model in quantum mechanics. It has a simple Halmitonian and energy levels:
QM = (0+1)-dimensional QFT:
The partition function can be easily calculated. First we use the imaginary-time path integral to get the partition func. Then we can compare it to the result from the direct calculation of (3).
In energy representation,
The thermal avg of energy:
At low
At high
Path integal: using x-basis.
Compare to the propergation amplitude in path integal:
With imaginary time interval
Then we do the trick of inserting orthogonal complete basis:
For each time interval
Note here we have the periodic boundary condition(PBC): $x{N+1}=x{0}=x$. This comes from the Boson statistics. And for Fermions we will later see a anti-PBC.
Now we have to consider elements of the type
Here
So the partition function
The integral over the momenta
Then the partition func becomes:
Note that here we have a factor which is divergent and independent of
Whene
From quantum mechanical path integral
- (i) Wick rotation:
. - (ii) $\mathcal{L}{E}=-\mathcal{L}{M}(\tau \equiv i t)=\frac{m}{2}\left(\frac{d x}{d \tau}\right)^{2}+V(x)$.
- (iii)
. - (iv) Require periodicity:
.
With these steps the int becomes (30), which is called imaginary-time formalism:
E: Euclidean.
M: Minkowski.
For fields with non-zero spin: few modification, still works.
To compute (30), write
with periodic conditions, $\omegan = 2\pi n /\beta \hbar
Here,
Generally with the Fourier sum
Use this the int in
Next we need to consider the
The determinant is hard to compute and could be written into the previous hard-to-compute
Now we have the partition function:
This is a Gaussian integral, and finally:
Calculate