Case Study 1: Helium — The First Three-Body Problem

Overview

The helium atom is the simplest system in quantum mechanics that cannot be solved exactly. With just two electrons and a nucleus, it represents the quantum mechanical three-body problem — and like its classical gravitational counterpart, it has resisted exact analytical solution for a century. Yet the helium ground state energy is now known to extraordinary precision, entirely through variational methods. This case study traces the history of the helium problem from the early failures of the old quantum theory through Hylleraas's pioneering variational calculations to the modern era of 40-digit accuracy.

Part 1: The Problem That Stumped the Old Quantum Theory

Historical Context

By 1925, the Bohr model and its Bohr-Sommerfeld extension had achieved remarkable success for one-electron systems. Hydrogen energy levels, the Stark effect, even the basics of fine structure — all yielded to the old quantum theory's rules of quantized orbits.

But helium was a disaster.

The helium atom has two electrons orbiting a nucleus of charge $Z = 2$. In the old quantum theory, one would try to find two quantized orbits, account for their mutual repulsion, and predict the energy levels. The problem is that the classical three-body problem has no closed-form solution, so there is no well-defined orbit to quantize.

Multiple physicists attempted the helium problem using the old quantum theory, most notably J.H. Van Vleck (1922) and Max Born and Werner Heisenberg (1923). The results were consistently poor — errors of 5% or more in the ground state energy, and qualitative failures in predicting the excited state spectrum. Born and Heisenberg managed to get a result of about $-5.5 R_\infty$ for the ionization energy (where the experimental value is $-5.8 R_\infty$), but only by using an unsatisfying mixture of classical mechanics and ad hoc quantization rules.

🔵 Historical Note: The failure of the old quantum theory for helium was one of the key motivations for the development of matrix mechanics (Heisenberg, 1925) and wave mechanics (Schrödinger, 1926). The old quantum theory could not handle even the simplest multi-electron system. Something fundamentally new was needed.

The Schrödinger Equation for Helium

Schrödinger's wave mechanics provided the correct equation almost immediately. In atomic units:

$$\hat{H} = -\frac{1}{2}\nabla_1^2 - \frac{1}{2}\nabla_2^2 - \frac{2}{r_1} - \frac{2}{r_2} + \frac{1}{r_{12}}$$

The first two terms are the kinetic energies of the two electrons. The third and fourth terms are the electron-nucleus attractions (with $Z = 2$). The fifth term is the electron-electron repulsion — and it is this term that makes the problem unsolvable.

The electron-electron repulsion $1/r_{12} = 1/|\mathbf{r}_1 - \mathbf{r}_2|$ couples the two electrons' coordinates. The wavefunction $\psi(\mathbf{r}_1, \mathbf{r}_2)$ is a function of six variables (three coordinates for each electron), and the Schrödinger equation cannot be separated into independent equations for each electron.

Why Is It So Hard?

The fundamental difficulty is the same as in the classical three-body problem: the motion of each particle depends on the positions of all other particles, creating a coupled, nonlinear system.

In classical mechanics, the two-body gravitational problem is exactly solvable (Kepler orbits), but the three-body problem is not. Similarly, in quantum mechanics, the two-body problem (hydrogen-like atoms) is exactly solvable, but the three-body problem (helium) is not.

Specifically, the electron-electron repulsion $1/r_{12}$ prevents separation of variables. If we could write $\psi(\mathbf{r}_1, \mathbf{r}_2) = \phi(\mathbf{r}_1)\chi(\mathbf{r}_2)$, the problem would factorize into two independent hydrogen-like problems. But the $1/r_{12}$ term couples the two electrons and forbids this factorization.

Part 2: The Perturbative Approach and Its Limitations

First-Order Perturbation Theory

The natural first attempt is perturbation theory: treat $1/r_{12}$ as a perturbation on the solvable two-electron hydrogen-like system.

The unperturbed ground state is the product of two 1s hydrogen-like orbitals with $Z = 2$:

$$\psi_0(\mathbf{r}_1, \mathbf{r}_2) = \frac{8}{\pi} e^{-2(r_1 + r_2)}$$

The unperturbed energy is $E^{(0)} = -Z^2 = -4$ hartree.

The first-order energy correction is the famous Unsöld integral:

$$E^{(1)} = \left\langle \psi_0 \left| \frac{1}{r_{12}} \right| \psi_0 \right\rangle = \frac{5Z}{8} = \frac{5}{4}$$

This six-dimensional integral was first evaluated by Albrecht Unsöld in 1927 using an expansion of $1/r_{12}$ in Legendre polynomials — a tour de force of mathematical physics for the time.

The first-order result: $E \approx -4 + 1.25 = -2.75$ hartree. The experimental value is $-2.9037$ hartree. The error is 5.3%.

Higher-Order Perturbation Theory

Second-order perturbation theory for helium requires summing over all excited states of the unperturbed system — an infinite sum that includes both bound and continuum states. This was computed by Hylleraas (1930), who found:

$$E^{(2)} = -0.1574 \text{ hartree}$$

The second-order result: $E \approx -2.75 - 0.158 = -2.908$ hartree. This is remarkably close to experiment, but the calculation is substantially more difficult than first order.

The perturbation series for helium converges, but slowly. The electron-electron repulsion is not "small" in any meaningful sense — it is 31% of the total potential energy. The fact that perturbation theory works at all is somewhat fortunate; for systems where the perturbation is comparable to the unperturbed energy, convergence is not guaranteed.

The Limitation

Perturbation theory for helium has a fundamental limitation: it treats the electrons as if they occupy independent hydrogen-like orbitals, with corrections computed order by order. But the electrons are strongly correlated — when one electron is close to the nucleus, the other tends to be far away (and vice versa). This electron correlation is poorly captured by the independent-particle starting point.

Part 3: Hylleraas and the Variational Revolution

The Screening Approach (1928)

The breakthrough came from the variational method. In 1928, several physicists independently realized that a simple modification of the unperturbed wavefunction could dramatically improve the energy.

The key insight: each electron partially screens the nuclear charge from the other. Instead of using $Z = 2$ in the exponential, use an effective charge $Z'$ as a variational parameter:

$$\psi(\mathbf{r}_1, \mathbf{r}_2; Z') = \frac{(Z')^3}{\pi} e^{-Z'(r_1 + r_2)}$$

The variational calculation (detailed in Section 19.3 of the main chapter) gives $Z'_{\text{opt}} = 27/16 = 1.6875$ and $E_{\text{min}} = -2.8477$ hartree — a 1.9% error, beating first-order perturbation theory with far less effort.

Hylleraas's Breakthrough (1929)

Egil Hylleraas, a Norwegian physicist working in Göttingen under Max Born, made the crucial advance: he introduced the inter-electron distance $r_{12}$ explicitly into the trial wavefunction.

Hylleraas defined coordinates:

$$s = r_1 + r_2, \quad t = r_1 - r_2, \quad u = r_{12}$$

and used the trial function:

$$\psi = e^{-Z's} \sum_{l,m,n} c_{lmn} \, s^l \, t^{2m} \, u^n$$

Note the crucial features: - The function depends on $t$ only through $t^2$, ensuring exchange symmetry $\psi(\mathbf{r}_1, \mathbf{r}_2) = \psi(\mathbf{r}_2, \mathbf{r}_1)$ (required for the spin-singlet ground state). - The explicit $u = r_{12}$ dependence captures electron correlation: the wavefunction can increase or decrease when the electrons are close together or far apart.

With just 3 parameters ($Z'$, $c_{01}$ for the $u$ term, and $c_{10}$ for the $s$ term), Hylleraas obtained $E = -2.9025$ hartree — accurate to 0.04%.

With 6 parameters, he achieved $E = -2.90324$ hartree — accurate to 0.016%.

📊 By the Numbers: Hylleraas's Progressive Results (1929)

Parameters	Energy (hartree)	Error
1 ($Z'$ only)	$-2.8477$	1.9%
3	$-2.9025$	0.04%
6	$-2.90324$	0.016%
10	$-2.90349$	0.007%
14	$-2.90360$	0.003%

Why $r_{12}$ Matters

The dramatic improvement from including $r_{12}$ has a deep physical reason. The electron-electron cusp condition requires:

$$\left.\frac{\partial \psi}{\partial r_{12}}\right|_{r_{12} = 0} = \frac{1}{2} \psi(r_{12} = 0)$$

This is analogous to the Kato nuclear cusp condition $\psi'(0)/\psi(0) = -Z$ at a nucleus. Any trial function that does not include $r_{12}$ explicitly cannot satisfy this condition, leading to a logarithmic singularity in the local energy at the electron-electron coalescence point.

Including even a single linear $r_{12}$ term in the wavefunction satisfies the cusp condition approximately and removes the worst errors. This is why Hylleraas's 3-parameter function (which includes a term linear in $u = r_{12}$) is so much better than the 1-parameter screening function.

Part 4: The Modern Era — Precision Beyond Experiment

Pekeris and the 1078-Term Calculation (1959)

Chaim Pekeris at the Weizmann Institute, using one of the first electronic computers (the WEIZAC), extended Hylleraas's approach to 1078 basis functions. His result:

$$E = -2.903724375 \text{ hartree}$$

This was accurate to 9 significant figures — far beyond the experimental precision available at the time.

Drake and Yan (2000s)

Gordon Drake and Zong-Chao Yan pushed the Hylleraas expansion to thousands of terms, incorporating logarithmic terms $(\ln r_{12})$ that are required by the exact structure of the wavefunction. Their results achieved 20+ digit accuracy.

Nakashima and Nakatsuji (2007)

Using the "free iterative complement interaction" (free ICI) method — a systematic extension of the Hylleraas approach — Hiroshi Nakatsuji and colleagues computed the helium ground state energy to 40 significant figures:

$$E = -2.9037243770341195982991 \ldots \text{ hartree}$$

This is far more precise than any conceivable experimental measurement of the helium ionization energy. At this level, the non-relativistic Schrödinger equation is no longer the limiting factor — relativistic corrections, quantum electrodynamic effects (Lamb shift), and nuclear size effects all contribute at the level of the last known digits.

💡 Key Insight: The history of helium calculations perfectly illustrates the power of the variational method: systematic improvement, guaranteed convergence from above, and no fundamental limitation on accuracy. With enough variational freedom, the variational method can achieve arbitrary precision.

Current Status

Today, the helium atom serves as a benchmark for:

1. Computational methods: Any new quantum chemistry method is tested on helium first.
2. QED calculations: The comparison between theory and experiment for helium is used to test quantum electrodynamics at low energies.
3. Nuclear charge radius: Precise helium spectroscopy combined with high-accuracy calculations constrains the alpha-particle charge radius.
4. Fundamental constants: Helium measurements contribute to the determination of the Rydberg constant and fine-structure constant.

Part 5: Lessons and Legacy

What Helium Teaches About the Variational Method

1. Physical insight matters. The jump from 5.3% error (perturbation theory) to 1.9% error (screening) came from a single physical idea — that electrons screen each other. The further jump to 0.04% (Hylleraas) came from another physical idea — that the wavefunction depends on the inter-electron distance.

2. The right coordinates matter. Hylleraas's $s, t, u$ coordinates were not an arbitrary choice. They are the natural coordinates for the helium problem because $u = r_{12}$ directly captures the electron-electron physics that makes the problem hard.

3. Systematic improvement is possible. The Hylleraas expansion provides a systematic way to improve accuracy: add more terms. The variational theorem guarantees that each additional term either improves the energy or confirms convergence. This is not true for perturbation theory, where higher orders can diverge.

4. The cusp conditions are important. Getting the behavior of the wavefunction right at singular points ($r_1 = 0$, $r_2 = 0$, $r_{12} = 0$) is crucial for rapid variational convergence. Functions that violate cusp conditions converge slowly.

5. Computers changed everything. Hylleraas did his 14-parameter calculation by hand. Pekeris used 1078 terms on a 1950s computer. Today's calculations use millions of terms on modern hardware. The variational method scales beautifully with computational power.

The Broader Impact

The variational approach pioneered on helium became the foundation of quantum chemistry. The Hartree-Fock method, configuration interaction (CI), coupled cluster theory, and density functional theory all use variational or variational-like principles. Every quantum chemistry calculation performed today — from drug design to materials science — traces its intellectual lineage back to Hylleraas's helium calculation of 1929.

Discussion Questions

1. Why was helium such a critical test case for quantum mechanics in the 1920s? What would it have meant for the theory if the variational method had failed to produce accurate results?

2. Compare the "brute force" approach (adding more and more terms to the Hylleraas expansion) with the "clever ansatz" approach (choosing the right functional form with fewer parameters). What are the advantages and disadvantages of each?

3. The helium ground state energy is now known to 40 digits, far beyond experimental measurement. Is there scientific value in such extreme precision, or is it merely a computational exercise? What physics can be extracted from these calculations?

4. Hylleraas's key insight was to include $r_{12}$ explicitly. What other multi-electron systems might benefit from explicit inter-electron coordinates? Why is this approach (called "explicitly correlated methods") computationally expensive for larger systems?

5. The variational method finds upper bounds. Can you think of ways to establish lower bounds on the helium ground state energy? (Research hint: look up the Temple lower bound and the Bazley-Fox method.)