Clamped-nuclei approximation An approximate description of the dihydrogen cation starts with the neglect of the motion of the nuclei - the so-called clamped-nuclei approximation. This is a good approximation because the nuclei (proton, deuteron or triton) are more than a factor 1000 heavier than the electron. Therefore, the motion of the electron is treated first, for a given (arbitrary) nucleus-nucleus distance
R. The electronic energy of the molecule
E is computed and the computation is repeated for different values of
R. The nucleus-nucleus repulsive energy
e2/(4
ε0
R) has to be added to the electronic energy, resulting in the total molecular energy
Etot(
R). The energy
E is the eigenvalue of the
Schrödinger equation for the single electron. The equation can be solved in a relatively straightforward way due to the lack of electron–electron repulsion (
electron correlation). The wave equation (a
partial differential equation) separates into two coupled
ordinary differential equations when using
prolate spheroidal coordinates instead of cartesian coordinates. The analytical solution of the equation, the
wave function, is therefore proportional to a product of two infinite
power series. The numerical evaluation of the series can be readily performed on a computer. The analytical solutions for the electronic energy eigenvalues are also a
generalization of the
Lambert W function which can be obtained using a
computer algebra system within an
experimental mathematics approach.
Quantum chemistry and Physics textbooks usually treat the binding of the molecule in the electronic ground state by the simplest possible
ansatz for the wave function: the (normalized) sum of two 1s hydrogen wave functions centered on each nucleus. This ansatz correctly reproduces the binding but is numerically unsatisfactory.
Historical notes Early attempts to treat H2^+ using the
old quantum theory were published in 1922 by
Karel Niessen and
Wolfgang Pauli, and in 1925 by
Harold Urey. In 1928,
Linus Pauling published a review putting together the work of Burrau with the work of
Walter Heitler and
Fritz London on the hydrogen molecule. The complete mathematical solution of the electronic energy problem for in the clamped-nuclei approximation was provided by
Wilson (1928) and
Jaffé (1934). Johnson (1941) gives a succinct summary of their solution. or 1sσg and it is gerade. There is also the first excited state A2Σ (2pσ
u), which is ungerade. Asymptotically, the (total) eigenenergies
Eg/
u for these two lowest lying states have the same asymptotic expansion in inverse powers of the internuclear distance
R: : E_{g/u} = - \frac12 - \frac{9}{4 R^4} + O\left(R^{-6}\right) + \cdots This and the energy curves include the internuclear
1/R term. The actual difference between these two energies is called the
exchange energy splitting and is given by: : \Delta E = E_{u} - E_{g} = \frac{4}{e} \, R \, e^{-R} \left[ \, 1 + \frac{1}{2R} + O\left(R^{-2}\right) \, \right] which exponentially vanishes as the internuclear distance
R gets greater. The lead term was first obtained by the
Holstein–Herring method. Similarly, asymptotic expansions in powers of have been obtained to high order by Cizek
et al. for the lowest ten discrete states of the molecular hydrogen ion (clamped nuclei case). For general diatomic and polyatomic molecular systems, the exchange energy is thus very elusive to calculate at large internuclear distances but is nonetheless needed for long-range interactions including studies related to magnetism and charge exchange effects. These are of particular importance in stellar and atmospheric physics. The energies for the lowest discrete states are shown in the graph above. These can be obtained to within arbitrary accuracy using
computer algebra from the generalized
Lambert W function (see eq. (3) in that site and reference The red solid lines are 2Σ states. The green dashed lines are 2Σ states. The blue dashed line is a 2Π
u state and the pink dotted line is a 2Π
g state. Note that although the generalized
Lambert W function eigenvalue solutions supersede these asymptotic expansions, in practice, they are most useful near the
bond length. The complete Hamiltonian of (as for all centrosymmetric molecules) does not commute with the point group inversion operation
i because of the effect of the nuclear hyperfine Hamiltonian. The nuclear hyperfine Hamiltonian can mix the rotational levels of
g and
u electronic states (called
ortho-
para mixing) and give rise to
ortho-
para transitions.
Born-Oppenheimer approximation Once the energy function
Etot(
R) has been obtained, one can compute the quantum states of rotational and vibrational motion of the nuclei, and thus of the molecule as a whole. The corresponding 'nuclear' Schrödinger equation is a one-dimensional ordinary differential equation, where the nucleus-nucleus distance
R is the independent coordinate. The equation describes the motion of a fictitious particle of mass equal to the reduced mass of the two nuclei, in the potential
Etot(
R)+
VL(
R), where the second term is the centrifugal potential due to rotation with angular momentum described by the quantum number
L. The eigenenergies of this Schrödinger equation are the total energies of the whole molecule, electronic plus nuclear.
High-accuracy ab initio theory The
Born-Oppenheimer approximation is unsuited for describing the dihydrogen cation accurately enough to explain the results of precision spectroscopy. The full Schrödinger equation for this cation, without the approximation of clamped nuclei, is much more complex, but nevertheless can be solved numerically essentially exactly using a variational approach. Thereby, the simultaneous motion of the electron and of the nuclei is treated exactly. When the solutions are restricted to the lowest-energy orbital, one obtains the rotational and ro-vibrational states' energies and wavefunctions. The numerical uncertainty of the energies and the wave functions found in this way is negligible compared to the systematic error stemming from using the Schrödinger equation, rather than fundamentally more accurate equations. Indeed, the Schrödinger equation does not incorporate all relevant physics, as is known from the hydrogen atom problem. More accurate treatments need to consider the physics that is described by the
Dirac equation or, even more accurately, by quantum electrodynamics. The most accurate solutions of the ro-vibrational states are found by applying non-relativistic quantum electrodynamics (
NRQED) theory. For comparison with experiment, one requires differences of state energies, i.e. transition frequencies. For transitions between
ro-vibrational levels having small rotational and moderate vibrational
quantum numbers the frequencies have been calculated with theoretical fractional uncertainty of approximately . Additional contributions to the uncertainty of the predicted frequencies arise from the uncertainties of
fundamental constants, which are input to the theoretical calculation, especially from the ratio of the proton mass and the electron mass. Using a sophisticated ab initio formalism, also the hyperfine energies can be computed accurately, see below. ==Experimental studies==