Unique System Evolution Determined by Initial and Final States under Specific Conditions

Scientists investigate the long-standing problem of uniquely determining a Hamiltonian from initial-to-final state evolution in quantum systems. Manuel Cañizares, Pedro Caro, Ioannis Parissis, and Thanasis Zacharopoulos, all researchers in the field, demonstrate that uniqueness holds for time-independent potentials under considerably weaker conditions than previously established. This research builds upon earlier work by the same authors concerning polynomial decay, now extending to potentials exhibiting -type decay and allowing for -type singularities, a significant advancement achieved through a refined Kenig-Ruiz-Sogge resolvent estimate. By focusing on the time-independent scenario, the team circumvented the need for complex geometrical solutions and stringent decay assumptions, offering a more general and potentially applicable framework for understanding quantum evolution.

Scientists have achieved an advance in determining the potential energy landscape of quantum systems, addressing a fundamental inverse problem: predicting a quantum particle’s complete behaviour given only its initial and final states. Researchers demonstrate that, under specific conditions, the potential governing a particle’s evolution is uniquely defined by this initial-to-final-state map. The study focuses on time-independent potentials, a crucial simplification enabling a more precise understanding of the underlying physics and circumventing complexities inherent in time-dependent scenarios. By refining a mathematical tool known as the Kenig-Ruiz-Sogge resolvent estimate, the team relaxed previous constraints on the types of potentials that can be uniquely determined. Specifically, the research establishes uniqueness for potentials exhibiting a combination of L1-type decay at infinity and Lq-type singularities, representing an improvement over earlier work requiring stronger decay assumptions. The ability to handle potentials with singularities is valuable, as many real-world systems exhibit such irregularities. The research demonstrates that if two potentials, both belonging to the mathematical class L1(Rn) ∩Lq(Rn) where q 1 if n = 2 or q ≥n/2 if n ≥3, produce identical initial-to-final-state maps, then the potentials themselves must be identical. A refined Kenig-Ruiz-Sogge resolvent estimate underpins the methodological approach to discerning time-independent potentials from initial-to-final-state maps. The study considers the Schrödinger equation in high dimensions, establishing a direct problem wherein, given an evolution map from initial to final states at a fixed time, the goal is to uniquely determine the Hamiltonian generating this evolution. Solutions, termed ‘physical solutions’, are constructed by evolving initial conditions within the continuous function space C[0, T]; L2(Rn), where T represents the fixed time and L2(Rn) denotes a space of square-integrable functions, and are used to define the initial-to-final-state map, a bounded linear operator. To probe the uniqueness of potential recovery, the research leverages an Alessandrini-type orthogonality relation, initially valid only for physical solutions, connecting the difference between two potentials with a specific integral involving pairs of solutions. Extending this identity beyond physical solutions necessitates the construction of ‘stationary states’, which are time-harmonic solutions perturbed by the potential itself, obtained by inverting a resolvent-type operator and building correction terms using a Neumann series. The core innovation lies in addressing limitations at the critical endpoint where classical resolvent estimates fail to provide sufficient decay in the energy parameter. To overcome this, a novel scale of Banach spaces is introduced, enabling the recovery of a decaying factor even in the absence of quantitative decay at the endpoint, allowing for an improved Kenig-Ruiz-Sogge estimate. This estimate is crucial for constructing stationary states for potentials with singularities and establishing the uniqueness of the inverse problem, deliberately avoiding complex geometrical solutions employed in time-dependent cases, streamlining the analysis and reducing the need for strong decay assumptions at infinity. Uniqueness of the Hamiltonian is established when potentials V1 and V2, belonging to both L1(Rn) and Lq(Rn), yield identical initial-to-final-state maps U1T and U2T, provided q exceeds 1 if n equals 2, or q is greater than or equal to n/2 if n is 3 or larger. This result represents a significant refinement over previous work requiring stronger decay conditions at infinity for both time-dependent and time-independent potentials, successfully relaxing these assumptions to allow for Lq-type singularities while maintaining L1-integrability. The core of the proof relies on extracting information about the difference between potentials, V1 − V2, from the equality of their respective initial-to-final-state maps. An Alessandrini-type orthogonality relation is derived, revealing that the integral of (V1 − V2) multiplied by pairs of associated solutions vanishes over the time-space domain Σ. Extending this relation beyond physical solutions necessitates the introduction of stationary states, constructed through inversion of a resolvent-type operator and utilising Neumann series corrections, expressed as e−i|κ|2t e−iκ·x plus a correction term wcor. The construction of wcor involves inverting an operator of the form Id − Pλ ◦ V, where Pλ represents the solution operator for the Helmholtz equation, requiring that Pλ ◦ V is sufficiently small, a condition achieved through the refined Kenig, Ruiz, Sogge resolvent estimate ensuring decay in the energy parameter λ, particularly important at the critical endpoint q = n/2. Scientists have long grappled with time-independent potentials, with the more realistic problem of time-dependent potentials remaining largely untouched. Future work will likely focus on extending these results to time-dependent scenarios and exploring the implications for data-driven modelling, potentially allowing for the design of novel materials with tailored quantum properties.