splitvalence

Split-valence basis sets are a class of Gaussian-type orbital basis sets used in computational chemistry to describe molecular orbitals. In these sets the valence electrons are described with more flexibility than the core electrons by splitting the valence basis into multiple parts, providing a higher level of accuracy for chemical bonding while keeping computational cost reasonable.

In practice, split-valence sets describe core orbitals with a relatively tight, single contracted basis and describe valence orbitals with two or more contracted functions of differing flexibility. This “split” of the valence region allows valence electrons to adapt more readily to changes in bonding and molecular environment than a single, rigid valence description.

Common examples include the 6-31G family, where core functions are contracted from several Gaussians and the valence is described by two sets of functions with differing flexibility. Extensions such as 6-31G(d), 6-31G(d,p), and 6-31G** add polarization functions (for example, d orbitals on heavy atoms and/or p orbitals on hydrogens) to improve angular flexibility and describe non-spherical electron distributions more accurately. Other related families, like 6-311G, extend the valence treatment to triple-zeta while retaining the split-valence concept.

Split-valence basis sets are widely used with ab initio and density functional theory methods due to their favorable balance of accuracy and cost. They may be insufficient for systems requiring diffuse or highly correlated descriptions unless augmented with diffuse functions (e.g., aug- basis sets) or larger, more flexible basis sets.