We propose a new method that transcribes the fermion system to the boson-fermion system by introducing a unitary operator defined in the whole boson-fermion space. We obtain the boson-fermion expansion of the fermion pair operators such that only the collective pair modes are transformed into boson operators while the non-collective modes are unchanged in the lowest order approximation. The odd- A system is treated in the same formalism as the even-even nuclear system. We briefly discuss the relation between our method and the conventional boson expansion theory.
In this paper we show that a system of three fermions is exactly solvable for the case of a single-
j in the presence of an angular momentum-
J pairing interaction. On the basis of the solutions for this system, we obtain new sum rules for six-
j symbols. When the Hamiltonian contains only an interaction between pairs of fermions coupled to spin $J={J}_{\mathrm{max}}=2j\ensuremath{-}1$, the ``non-integer'' eigenvalues of three fermions with angular momentum
I around the maximum appear as ``non-integer'' eigenvalues of four fermions if
I is around (or larger than)
Jmax. This pattern is also found in five and six fermion systems. A boson system with spin
l exhibits a similar pattern.
We predict the existence of a pair-liquid phase in lattice fermion systems with finite-range attractive interactions. This exotic state competes on one side with a normal Fermi liquid of unpaired fermions and on the other side with a phase-separated state where all fermions are coupled into macroscopic clusters. We show that such a phase is absent in bosonic systems and therefore is protected by the exclusion principle. In contrast with zero-range attractive systems where clustering of more than two fermions is directly prohibited, here quantum statistics acts dynamically and in a more subtle way. Since a many-fermion wave function must have nodes, the cluster formation threshold is larger than the pair formation threshold. By directly solving a four-body Schr\"odinger equation on one- and two-dimensional lattices, we map the boundaries of pair stability. The pair liquid phase should be observable in cold-atom systems. We also discuss implications for the preformed-pair mechanism of high-temperature superconductivity.