Attractive Polaron Transition

• The paper demonstrates a first-order zero-temperature transition in 3D Fermi gases, where an impurity polaron converts to a dressed molecule near a critical interaction (1/(k_F a) ≈ 0.9).
• T‐matrix, FRG, and variational methods reveal how key quasiparticle properties like spectral weight, residue, and effective mass shift across the transition.
• Finite temperature and impurity density smooth the sharp transition into a crossover, resulting in a coexistence of polaronic and molecular spectral features.

Searching arXiv for recent and foundational papers on attractive polaron transitions. Attractive polaron transition denotes a family of closely related reorganizations of an impurity quasiparticle on the attractive branch of an interacting many-body medium. In the canonical three-dimensional Fermi-gas problem, it refers most directly to the zero-temperature single-impurity change from an attractive Fermi polaron to a dressed molecule or dimeron at a critical interaction strength; under experimentally relevant finite temperature and finite impurity density, that sharp transition is replaced by a smooth crossover with coexistence of polaronic and molecular spectral weight (Ness et al., 2020, Parish et al., 2020). In other established usages, the same phrase can denote a temperature- and concentration-driven change in the character of residual polaron-polaron interactions around $T \sim 0.1T_F$ (Hu et al., 2017), a pairing transition between polarons and tightly bound dimers in two dimensions (Koschorreck et al., 2012), or a motion-induced entrance of the attractive branch into a molecule-hole continuum at finite momentum (Hennebichler et al., 19 Jun 2026).

1. Definition and quasiparticle framework

An attractive Fermi polaron is the mobile quasiparticle formed when a minority-spin impurity is immersed in a majority-spin Fermi sea and interacts through an $s$-wave contact attraction characterized by the impurity-medium scattering length $a$. In three dimensions, the interaction is parameterized by $1/(k_F a)$: the BCS side has $a<0$ and $1/(k_F a)<0$, unitarity corresponds to $a \to \infty$ and $1/(k_F a)=0$, and the strong-attraction or BEC side has $a>0$ and $1/(k_F a)>0$ (Hu et al., 2017). In two dimensions, the natural control parameter is $s$0, while in the square-lattice attractive Hubbard model the competition is organized by $s$1 and the majority filling fraction $s$2 (Koschorreck et al., 2012, Pascual et al., 2024).

Microscopically, the attractive branch is defined through poles of the impurity Green’s function. Within ladder $s$3-matrix or FRG treatments, the key observables are the pole energy $s$4, quasiparticle residue $s$5, effective mass $s$6, and decay width or decay rate $s$7. In the many-body $s$8-matrix formulation these are extracted from

$s$9

$a$0

while FRG formulations equivalently identify quasiparticle properties from complex poles of retarded impurity and molecule propagators (Hu et al., 2017, Schmidt et al., 2011).

The competing state is a dressed molecule or dimeron: an impurity-majority bound pair dressed by particle-hole excitations of the Fermi sea. The attractive polaron transition is therefore not merely a shift of a spectral line; it is a reordering of low-lying branches in which the dominant quasiparticle character changes from impurity-like dressing to bound-pair dressing. This suggests that the term is best used as an umbrella label for several related branch-reorganization phenomena rather than for a single universal mechanism.

2. Zero-temperature transition and its microscopic interpretation

In the single-impurity limit at $a$1, short-range interacting three-dimensional Fermi gases are predicted to exhibit a first-order transition between a polaron quasiparticle and a molecule. Diagrammatic Monte Carlo gives the critical interaction $a$2, and FRG with full self-energy feedback finds $a$3, in excellent agreement with diagrammatic Monte Carlo $a$4 (Ness et al., 2020, Schmidt et al., 2011). At unitarity, the attractive polaron energy is near $a$5 in the non-self-consistent $a$6-matrix treatment and $a$7 in the FRG calculation, with $a$8 in FRG and $a$9 in Chevy-type treatments (Hu et al., 2017, Schmidt et al., 2011).

Within one particle-hole variational theory, however, the crossing is shifted to $1/(k_F a)$0, and this mismatch is explicitly attributed to the truncation; improved molecular ansätze and higher-order methods move the crossing closer to $1/(k_F a)$1 (Ness et al., 2020, Parish et al., 2020). The same one particle-hole framework also motivated a reinterpretation of the transition as a first-order change between impurity ground states with different total momenta: a $1/(k_F a)$2 polaron at weak coupling and a $1/(k_F a)$3 finite-momentum polaron that becomes molecule-like at strong coupling. In that picture, two local minima coexist over $1/(k_F a)$4, the minima become degenerate near $1/(k_F a)$5, and the strong-coupling branch is well approximated by the molecule ansatz because the residue at $1/(k_F a)$6 collapses toward zero (Cui, 2020).

A further dynamical refinement is that the transition is not exhausted by a level crossing of energies; it also restructures spectral weight and lifetimes. FRG calculations show that the quasi-particle weight shifts from the attractive to the repulsive branch across the transition, while the molecular bound state carries a very small residue characteristic of a composite particle, with $1/(k_F a)$7 at unitarity. On the polaronic side, the excited molecular branch has a decay width that obeys $1/(k_F a)$8, where $1/(k_F a)$9 (Schmidt et al., 2011).

3. Finite temperature, impurity density, and the crossover regime

Experiments are not performed in the single-impurity, zero-temperature limit, and the most robust modern conclusion is that finite temperature and finite impurity density smooth the sharp polaron-molecule transition into a crossover. In a $a<0$0 spin-imbalanced gas at $a<0$1, with a global impurity fraction $a<0$2, trap-averaged local fraction $a<0$3 rising to $a<0$4 at strong coupling, and typical local impurity density $a<0$5, Raman spectroscopy shows that the coherent polaron weight $a<0$6 decreases continuously from $a<0$7 on the BCS side to $a<0$8 at unitarity and $a<0$9 on the BEC side, vanishing smoothly beyond the transition region (Ness et al., 2020).

The corresponding theory treats polarons as fermionic quasiparticles and molecules as bosonic quasiparticles occupied according to

$1/(k_F a)<0$0

with a single impurity chemical potential $1/(k_F a)<0$1 fixed by the total impurity density. Under these conditions, both branches are thermally populated across the entire transition region, so even where the molecule is the ground state, polaron states remain occupied, and vice versa (Ness et al., 2020). A closely related finite-temperature variational analysis likewise shows that the $1/(k_F a)<0$2 kink in the impurity free energy becomes a smooth crossover already by $1/(k_F a)<0$3 because of thermal population of excited states and thermal smearing of the majority Fermi surface (Parish et al., 2020).

Finite temperature and finite impurity concentration also deform the attractive branch itself. Non-self-consistent and self-consistent many-body $1/(k_F a)<0$4-matrix calculations show weak $1/(k_F a)<0$5 and $1/(k_F a)<0$6 dependence on the BCS side but sizable effects toward and beyond unitarity: $1/(k_F a)<0$7 rapidly approaches the bare mass, $1/(k_F a)<0$8 increases with temperature, and $1/(k_F a)<0$9 grows roughly as $a \to \infty$0 at low temperature, reaching $a \to \infty$1 on the BEC side by $a \to \infty$2 (Hu et al., 2017). In the same work, the impurity chemical potential is expanded as

$a \to \infty$3

and the Landau parameter $a \to \infty$4 changes sign near $a \to \infty$5: at low $a \to \infty$6, $a \to \infty$7–$a \to \infty$8 near unitarity, consistent with repulsive residual interactions, whereas at $a \to \infty$9 in the unitary or BEC regimes the polaron energy decreases with $1/(k_F a)=0$0, implying effectively attractive residual polaron-polaron interactions (Hu et al., 2017). The paper explicitly states that this “transition” is not a true phase transition but a qualitative change in effective residual interactions.

4. Experimental observables and spectroscopic diagnostics

The attractive polaron transition is diagnosed through branch energies, coherent spectral weight, linewidths, and continuum structure rather than through a single universal order parameter. In three dimensions, Raman spectroscopy with large momentum transfer isolates the coherent quasiparticle contribution and avoids the thermal peak shift that affects rf spectroscopy. For the $1/(k_F a)=0$1 experiment, the coherent Raman peak obeys

$1/(k_F a)=0$2

with $1/(k_F a)=0$3, and the line shape is fit by a coherent contribution plus an incoherent background whose high-frequency tail behaves as $1/(k_F a)=0$4, thereby providing access to the Tan contact (Ness et al., 2020).

In two dimensions, momentum-resolved photoemission spectroscopy directly measures the impurity spectral function $1/(k_F a)=0$5. The attractive polaron regime is characterized by a sharp, symmetric quasiparticle peak with nearly parabolic dispersion, while the molecular regime shows a weak, asymmetric feature with a lower-energy tail and broad incoherent background. The most sensitive indicators are not the quasiparticle energy itself, which varies only weakly through the transition, but the effective mass and the spectral weight: $1/(k_F a)=0$6 diverges for $1/(k_F a)=0$7, and linear extrapolation of the $1/(k_F a)=0$8 spectral weight gives $1/(k_F a)=0$9, compared with the homogeneous strict-2D theoretical prediction $a>0$0 (Koschorreck et al., 2012).

Momentum-resolved Raman injection spectroscopy also reveals finite-temperature remnants of the transition in the single-impurity spectral function. When the Raman momentum transfer is chosen near $a>0$1, the molecule-hole continuum at $a>0$2 is enhanced relative to the $a>0$3 attractive polaron. This provides a direct way to visualize the lowering of the molecule-hole onset across the crossover even when thermodynamics is already smooth (Parish et al., 2020).

Thermodynamics supplies an independent signature. The impurity free-energy shift satisfies

$a>0$4

and the Tan contact follows from

$a>0$5

The resulting $a>0$6 is non-monotonic on the polaron side because thermal occupation of higher-contact molecule-hole states initially increases the contact, whereas on the molecule side $a>0$7 decreases monotonically (Parish et al., 2020). This makes the contact a thermodynamic remnant of the zero-temperature transition even where spectroscopy already shows coexistence.

5. Dimensionality, lattice effects, and moving polarons

Dimensionality and band structure strongly modify the fate of the attractive branch. In a quasi-2D $a>0$8 gas, the ever-present two-body bound state implies that sufficiently strong attraction favors tightly bound dimers, and the observed pairing transition is identified by vanishing spectral weight and a diverging effective mass rather than by a large energy jump (Koschorreck et al., 2012). By contrast, in the two-dimensional attractive Hubbard model on a square lattice, no polaron-to-dimeron transition is observed for $a>0$9 between $1/(k_F a)>0$0 and $1/(k_F a)>0$1: both a one-particle-hole variational approach and a sign-problem-free determinant diagrammatic Monte Carlo algorithm find $1/(k_F a)>0$2 for all tested couplings up to $1/(k_F a)>0$3 in Monte Carlo and to $1/(k_F a)>0$4 variationally, with a finite quasiparticle residue throughout (Pascual et al., 2024).

A further extension is the moving Fermi polaron. In a $1/(k_F a)>0$5-$1/(k_F a)>0$6 mixture, low momentum obeys the usual Fermi-liquid form

$1/(k_F a)>0$7

but at intermediate momentum the attractive branch becomes non-monotonic and broadens sharply. The transition condition is kinematic: when the attractive polaron dispersion enters the molecule-hole continuum, $1/(k_F a)>0$8, it is no longer the ground state. Experiment and $1/(k_F a)>0$9-matrix theory identify the onset around $s$00–$s$01 for $s$02, accompanied by rapid linewidth growth and a kink in the dispersion; this is explicitly interpreted as a motion-induced polaron-molecule transition (Hennebichler et al., 19 Jun 2026).

Setting	Control parameter	Diagnostic signature
3D continuum Fermi gas	$s$03, $s$04, $s$05	Smooth reduction of $s$06, coexistence of polarons and molecules
2D continuum Fermi gas	$s$07	$s$08 divergence below $s$09, $s$10 at $s$11
Moving Fermi polaron	$s$12	Non-monotonic $s$13, sharp growth of $s$14
2D attractive Hubbard model	$s$15, $s$16	No $s$17 crossing in the explored regime

These results show that the attractive polaron transition is not dimension-independent. The continuum three-dimensional problem supports a well-established zero-temperature transition, the quasi-2D continuum shows a pairing transition with experimentally shifted critical coupling, the square-lattice Hubbard model can suppress the transition over a broad density range, and finite momentum opens a distinct kinematic route to molecularization even when the zero-momentum attractive polaron remains the ground state.

6. Thermodynamics, phase separation, and theoretical caveats

At finite impurity density, the transition cannot always be reduced to single-polaron physics. Near the polaron-molecule boundary, a low-temperature spin-imbalanced Fermi gas is argued to undergo phase separation because the effective dimer-dimer interaction becomes attractive through Fermi-sea mediation. In the Bose-Fermi mixture description of dressed dimers and excess majority fermions,

$s$18

with $s$19, so $s$20. The resulting tricritical point occurs at $s$21 for the broad-resonance, equal-mass case (Parish et al., 2020). This means that, for finite impurity density and low temperature, the single-impurity transition line is not the full thermodynamic phase boundary.

Methodologically, the field remains sensitive to approximation scheme. In the finite-temperature $s$22-matrix analysis, the non-self-consistent scheme reproduces zero-temperature benchmarks for attractive polaron energy and residue, whereas the self-consistent scheme more reliably locates the polaron-molecule threshold near $s$23 but tends to underestimate the attractive polaron energy and residue at low temperature (Hu et al., 2017). FRG with full self-energy feedback captures lifetimes and incoherent continua but still shows a unitarity energy offset relative to diagrammatic Monte Carlo, attributed to neglected particle-hole and atom-dimer vertices (Schmidt et al., 2011). Finite-temperature variational approaches reproduce the smoothing of the transition and the thermodynamic signatures but inherit the one-particle-hole shift of the zero-temperature critical coupling (Parish et al., 2020).

The resulting picture is therefore internally differentiated rather than contradictory. A zero-temperature, single-impurity, three-dimensional continuum problem supports a first-order polaron-molecule transition near $s$24; finite temperature and finite impurity density smooth that transition into a crossover with coexistence; two-dimensional continuum systems show a pairing transition with different diagnostics; certain lattice systems show no transition in the explored regime; and finite momentum can itself drive the attractive branch into the molecule-hole continuum (Ness et al., 2020, Pascual et al., 2024, Hennebichler et al., 19 Jun 2026). A plausible implication is that “attractive polaron transition” is most precise when accompanied by its control axis—interaction, temperature, density, dimensionality, lattice filling, or momentum—because the underlying branch reorganization is generic but its manifestation is highly context dependent.