Dyson expansion for form-bounded perturbations and applications to the polaron problem

Technical Summary: Dyson Expansion for Form-Bounded Perturbations and Applications to the Polaron Problem

Problem Statement
The paper addresses the mathematical analysis of polaron-type models, which describe the interaction between a charged particle and a bosonic field (e.g., the Fröhlich model for electron-phonon coupling or the Nelson model in quantum electrodynamics). The central object of study is the Hamiltonian $H(P)$ restricted to a sector of total momentum $P \in \mathbb{R}^d$:
$$H(P) = |P - P_f|^2 + d\Gamma(\omega) + \Phi(v)$$
where $P_f$ is the field momentum, $d\Gamma(\omega)$ is the field energy, and $\Phi(v)$ is the interaction field operator. A significant challenge in this setting is that the interaction term $\Phi(v)$ is often unbounded and only relatively form-bounded with respect to the non-interacting Hamiltonian, rather than operator-bounded. This low regularity prevents the direct application of standard Dyson series expansions, which typically require the perturbation to be a bounded operator or at least operator-bounded relative to the unperturbed part.

Methodology
The authors develop a rigorous framework to justify a Dyson expansion for perturbations that are merely form-bounded. Their approach consists of three main components:

1. Abstract Dyson Expansion (Section 2):
   The authors establish a norm-convergent expansion for the semigroup $e^{-t(A+B)}$, where $A \geq 0$ is self-adjoint and $B$ is a symmetric quadratic form form-bounded with respect to $A$ with a relative bound $\lambda < 1$.

   • Instead of the formal integral representation involving products of unbounded operators, they utilize a contour integral representation of the resolvent.
   • By defining a bounded operator $C = (A+a)^{-1/2}(A+B)(A+a)^{-1/2} - 1$ (where $a$ is a shift parameter), they expand the resolvent $(z - A - B)^{-1}$ as a Neumann series in powers of $C$.
   • This yields an expansion $e^{-t(A+B)} = e^{-tA} + \sum_{n \geq 1} D_n$, where $D_n$ are bounded operators defined via contour integrals. Crucially, while the individual terms $D_n$ are well-defined, they do not necessarily take the form of the standard time-ordered integrals of the interaction $\Phi(v)$ unless $B$ is operator-bounded.

2. Application to Polaron Models (Section 3):
   The abstract result is applied to the polaron Hamiltonian by identifying $A = |P - P_f|^2 + d\Gamma(\omega) + a$ and $B = \Phi(v)$.

   • The authors utilize the Lieb–Yamazaki commutator method (recalled in Appendix A) to confirm that under specific assumptions on the interaction potential $v$ and dispersion $\omega$ (Assumption 1), the interaction is infinitesimally form-bounded.
   • They analyze the vacuum expectation value $\langle \Omega | e^{-tH(P)} \Omega \rangle$. By exploiting the fact that odd terms in the expansion vanish due to the unequal number of creation and annihilation operators, they restrict the sum to even orders.
   • Using Wick's rule and the pull-through formula, they map the abstract operator terms to explicit integrals over momentum space involving Wick pairings (permutations in $W_{2n}$) and Dyck paths.

3. Derivation of Renewal Equations and Monotonicity (Section 4):

   • The authors derive a renewal equation for the vacuum expectation value by decomposing the set of Wick pairings into "interlacing" (irreducible) pairings and general pairings. This mirrors probabilistic approaches using Wiener integrals but is derived here via operator theory.
   • They introduce Assumption 2, requiring radial symmetry and that a specific integral involving $|v(k)|^2$ is a completely monotone function of $|P|^2$. This assumption holds for standard Fröhlich and Nelson models (with appropriate ultraviolet cutoffs).

Key Results

1. Convergent Dyson Expansion (Theorem 1a):
   For a large class of polaron models satisfying Assumption 1, the authors prove the existence of a convergent expansion for the vacuum expectation value:
   $$\langle \Omega | e^{-tH(P)} \Omega \rangle = e^{-t|P|^2} + \sum_{n \geq 1} \sum_{\pi \in W_{2n}} \int_{\Delta_{2n}^t} \int_{\mathbb{R}^{dn}} e^{-\sum t_j E_P^{(\pi,j)}(k)} \prod |v(k_j)|^2 dk_j dt$$
   This expansion is valid despite the interaction being only form-bounded.

2. Renewal Equation (Theorem 1b):
   A renewal equation is established relating the full expectation value to a convolution involving only "interlacing" Wick pairings ($W_{2n}^0$). This structure is essential for analyzing the long-time behavior and convexity properties.

3. Complete Monotonicity and Concavity (Corollary 1):
   Under Assumptions 1 and 2:

   • The function $|P|^2 \mapsto \langle \Omega | e^{-tH(P)} \Omega \rangle$ is completely monotone for any $t > 0$.
   • Consequently, the ground state energy $E_0(P) = \inf \text{spec } H(P)$ is a concave function of $|P|^2$. If the interaction is non-trivial ($v \not\equiv 0$), this concavity is strict on the set where a ground state exists.

Significance and Claims
The paper claims to provide a rigorous operator-theoretic justification for Dyson expansions in settings where the perturbation is not operator-bounded, a common scenario in quantum field theory models like the polaron.

• Methodological Contribution: The work bridges the gap between abstract functional analysis (form-bounded perturbations) and concrete physical models, offering a tool to derive series expansions without relying on probabilistic representations (such as Wiener integrals) which were previously the primary method for proving these specific properties.
• Physical Implications: The results confirm the concavity of the ground state energy with respect to the square of the momentum for Fröhlich and Nelson models. This property is noted as a crucial ingredient in recent proofs regarding the validity of the Landau–Pekar formula for the effective mass in the strong coupling limit.
• Generality: The abstract framework (Theorem 2) is presented as a general tool applicable to any system where the perturbation is form-bounded, potentially extending beyond polaron problems to other quantum field theoretic models.

The authors explicitly state that their renewal equation approach recovers results previously obtained via probabilistic methods (specifically referencing [14] for the Fröhlich model) but does so through a deterministic, operator-based Dyson expansion, thereby validating the expansion's utility in low-regularity settings.