The Grand Partition Function and Potential

1. The Grand Potential as a Thermodynamic Potential

Before we dive into the statistical mechanics, let's first establish the identity of the grand potential from a purely thermodynamic perspective. You'll recall from your studies of thermodynamics and classical mechanics that Legendre transformations are a powerful mathematical tool used to switch between different sets of independent variables, thereby generating different energy-like functions or potentials, each suited to specific physical constraints.

The internal energy $U(S,V,N)$ is the natural potential for an isolated system, with its differential given by the fundamental thermodynamic relation:

For a closed system at constant temperature, it is more convenient to work with the Helmholtz free energy, $F(T,V,N)$, which is obtained by a Legendre transform that replaces the extensive variable $S$ with the intensive variable $T$:

Now, for an open system, we also want to replace the extensive particle number $N$ with its conjugate intensive variable, the chemical potential $\mu$. We achieve this with a second Legendre transformation, this time on the Helmholtz free energy:

Grand Potential: $\Phi (T,V,\mu )=F-\mu N$

By substituting $F=U-TS$, we can also write it as $\Phi =U-TS-\mu N$. Taking the differential gives us:

This confirms that $\Phi$ is indeed the natural thermodynamic potential for a system whose state is specified by the variables $(T,V,\mu )$—precisely the constraints of the grand canonical ensemble. At equilibrium, for fixed $T$, $V$, and $\mu$, this potential is minimized.

For a clear summary of how the grand potential (denoted $\Omega$ in the text) fits into the family of thermodynamic potentials, please review the following brief reading.

2. The Bridge to Statistical Mechanics: $\Phi =-k_{B}T\ln \Xi$

Having defined $\Phi$ thermodynamically, we now derive its connection to the grand partition function $\Xi$. We will explore two distinct but complementary derivations that highlight different aspects of this fundamental relationship, appealing to both mathematical rigor and physical intuition rooted in the thermodynamic limit.

Derivation I: Comparison of Derivatives

This first method is an elegant proof that relies on comparing the functional forms of derivatives. We will compute the average energy $⟨E⟩$ and average particle number $⟨N⟩$ in two ways: first from statistical mechanics using $\ln \Xi$, and second from thermodynamics using $\Phi$. The consistency between these two approaches will reveal the link we seek.

The derivation is laid out with exceptional clarity in the following notes from the University of Rochester. We will follow its logic step-by-step.

This elegant comparison establishes the central result:

This equation is the grand canonical analogue of $F=-k_{B}T\ln Z$ and serves as the master equation from which all thermodynamic properties of an open system can be derived.

Derivation II: Saddle-Point Approximation

An alternative, more physically motivated derivation highlights the importance of the thermodynamic limit and provides an application of the saddle-point approximation you requested to cover.

Recall from the previous lesson that $\Xi$ can be written as a sum over the canonical partition functions $Z(N)$:

In the thermodynamic limit, we know $Z(N)\approx e^{-\beta F(N)}$. Substituting this gives:

We now have a sum of exponential terms. For a macroscopic system where $N$ is very large (on the order of Avogadro's number), the function in the exponent, $F(N)-\mu N$, will be a sharply peaked function of $N$. The saddle-point method tells us that such a sum is overwhelmingly dominated by its largest term. The largest term occurs at the value of $N=N^{\ast }$ that minimizes the quantity in the exponent, $F(N)-\mu N$. This value $N^{\ast }$ is, of course, the average or equilibrium particle number, $⟨N⟩$.

Therefore, we can approximate the entire sum by this single, maximal term:

Taking the logarithm of both sides:

Recognizing the term in the parenthesis as our thermodynamic definition of the grand potential, $\Phi =F-\mu N$, we arrive at the same conclusion:

The following video walks through this exact argument.

3. The Power of the Grand Potential: $\Phi =-pV$

We have established the link $\Phi =-k_{B}T\ln \Xi$. What makes this potential so particularly useful? One of its most powerful properties comes from the Euler homogeneous function theorem. For a function of extensive variables $U(S,V,N)$, Euler's theorem implies the relation:

This is known as the Gibbs-Duhem equation in an integrated form. Now, look at our definition of the grand potential:

Substituting the Euler relation into this definition gives a remarkably simple and profound result:

This means the grand potential is simply the negative of the pressure times the volume. This provides an incredibly direct route to calculating the equation of state for a system. By calculating the grand partition function $\Xi$, we immediately get the pressure:

The following reading demonstrates how this result emerges naturally in the context of an ideal gas.

This connection between the grand potential and pressure is a key reason why the grand canonical ensemble is often the most convenient framework for studying gases, liquids, and quantum fluids.

Conclusion

In this lesson, we have formally bridged the gap between the microscopic details of an open system and its macroscopic thermodynamic behavior.

Key Takeaways:

• The grand potential, $\Phi (T,V,\mu )$, is the characteristic thermodynamic potential for systems at constant temperature, volume, and chemical potential. It is defined via a Legendre transform: $\Phi =F-\mu N$.
• The fundamental connection to statistical mechanics is given by the "master equation" for the grand canonical ensemble:
  $\Phi =-k_{B}T\ln \Xi$
• We derived this relationship using two different approaches: a rigorous comparison of derivatives and a physically intuitive saddle-point approximation.
• A crucial thermodynamic identity reveals that $\Phi =-pV$, providing a direct path from the grand partition function to the system's equation of state.

In our next lesson, we will put this powerful machinery to work. We will systematically use the relation $\Phi =-k_{B}T\ln \Xi$ and the differential $d\Phi =-SdT-PdV-Nd\mu$ to derive expressions for all the key thermodynamic observables, including average energy, entropy, pressure, and particle number, as derivatives of the grand partition function.