In 1984 Frenkel and Ladd proposed the “Einstein crystal
method”, a novel scheme to compute the free energy
of a solid [1]: the method is based on a thermodynamic
integration of the Helmholtz free energy in the canonical
ensemble along a reversible path between the system of
interest and an ideal Einstein crystal with the same structure
as the real solid, whose Helmholtz free energy can be
analytically computed. In the ideal Einstein crystal, particles
are attached to their lattice positions via harmonic
springs (of constant ΛE). More recently, some of us have
proposed the “Einstein molecule method” [2, 3], a variant
of the Einstein crystal to compute the free energy
of molecular solids. The main difference between both
methods is the choice of the reference system. In the
Einstein crystal the reference system is an ideal Einstein
crystal with the constraint of the center-of-mass of the
system kept fixed (to avoid a quasi-divergence of the integral
of the free energy change from the reference system
to the real solid). In the Einstein molecule the reference
system is an ideal Einstein crystal with the constraint of
the position of one particle kept fixed.