Optimality of the free quantum evolution: the general case with nodes

In the landscape of quantum mechanics, the Schrödinger equation dictates how a wave function evolves over time. While we often treat this evolution as a deterministic unfolding of waves, a deeper question remains. Is the way a quantum particle moves actually the most "efficient" way possible?

Current theoretical frameworks suggest that quantum evolution can be viewed through the lens of optimal transport. In this view, particle movement minimizes a specific cost functional (a mathematical function used to measure the "effort" of a path). However, a significant mathematical hurdle has long stood in the way: the existence of "nodes." Nodes are points in space where the probability density of a particle becomes exactly zero. In stochastic interpretations of quantum mechanics, these zero-points create singularities (points where a value becomes undefined or infinite). These divergences make it difficult to prove that a stable, well-behaved evolution even exists. This paper addresses that gap. It shows that even when these problematic nodes are present, the free evolution of a wave function remains an optimal path, albeit a local one.

The Singularity of Zero Probability

To understand why nodes are a problem, one must look at Nelson’s Stochastic Mechanics. In this framework, a quantum particle is modeled as undergoing a Markov diffusion process. This is essentially a specialized form of Brownian motion (random jittering) guided by a drift field (the force pushing the particle along a path).

This drift field is composed of two parts: the particle's velocity field and a term involving the gradient of the logarithm of its probability density ($\rho$). The mathematical breakdown occurs because the term $\nabla \ln \rho$ involves dividing by the density itself. At a node, where $\rho = 0$, this term diverges toward infinity. For decades, this posed a severe challenge. If the instructions guiding the particle's random walk become infinitely large at certain points, can we guarantee that a valid path exists?

Previous work by Morato (2022) successfully demonstrated that quantum evolution is a global minimum for the cost functional. However, that work only applied under the strict assumption that the wave function has no nodes. The presence of a single zero-point in the probability distribution breaks the global optimality and threatens the mathematical consistency of the stochastic model.

Bridging Fluid Dynamics and Stochastic Paths

The author overcomes this singularity by shifting the perspective from individual particle paths to a fluid dynamics formulation. Instead of tracking a single erratic particle, the paper treats the evolving probability density as a continuous fluid. This fluid is governed by two key equations: the continuity equation (which ensures mass is conserved) and the Hamilton-Jacobi Madelung (HJM) equation (which governs the phase and energy of the flow).

The methodology follows a structured path to handle the nodes:

1. Defining Proper Characteristics: The author introduces a distinguished subset of "proper characteristics," denoted as $\Xi$. This subset restricts the allowed density and velocity fields. Specifically, nodes must be isolated points or regular surfaces. Additionally, the velocity must vanish at the node boundaries to prevent infinite drifts.
2. Constructing a Convex Representation: Because the actual cost functional $A_Q(\rho, v)$ is non-convex (meaning it has multiple "valleys" and "peaks"), it is difficult to optimize directly. The author constructs an ancillary convex functional, $\hat{F}_\infty$, in a linear space of stochastic processes. This acts as a mathematically smoother proxy that represents the original cost.
3. Applying Nelson’s Renormalization: Using Nelson’s renormalization formula, the author links discrete-time particle displacements to the continuous-time integral of the action functional. This allows the transition from a sum of random steps to the smooth evolution described by the Schrödinger equation.

Local vs. Global Optimality

The core result of the paper is a distinction between two types of stability based on the topology of the wave function. The authors demonstrate that the free evolution of a quantum wave function is always an extremum (a stationary point) of the cost functional.

Specifically, the paper finds that if the wave function contains no nodes, the evolution is a global minimum. Among all possible ways the density could move from point A to point B, the Schrödinger evolution is the absolute most efficient. However, the authors report that when nodes are present, the evolution becomes a local minimum. Within the subset of allowed characteristics $\Xi$, the Schrödinger evolution is still the best path in its immediate mathematical neighborhood. It is no longer guaranteed to be the absolute best path across the entire landscape of possibilities.

This shift from global to local optimality provides a physical insight. The existence of nodes introduces a form of intrinsic instability. If a system is in a state with nodes, it sits in a local valley that is not the lowest possible point in the universe of possible evolutions.

Constraints of the $\Xi$ Framework

While the proof is mathematically rigorous, it has specific limits. The optimality result is strictly confined to the subset $\Xi$ of "proper characteristics." This is a significant caveat. The paper proves that Schrödinger evolution is optimal among a specific class of well-behaved flows.

The $\Xi$ framework requires nodes to be isolated points or regular curves/surfaces. Consequently, the results may not apply to systems where nodes form more complex, irregular structures. Furthermore, the proof relies on the assumption that the solution $\psi$ is unique up to a constant phase. In complex systems where multiple degenerate solutions (different states with the same energy) exist, the relationship between the cost functional and the specific path taken may be more complicated.

The Verdict

The paper provides a definitive mathematical answer to the problem of nodal singularities in stochastic quantum mechanics. By moving from the chaotic view of individual particles to the structured view of fluid dynamics, the author proves that the Schrödinger equation remains an optimal descriptor of reality.

Is the free quantum evolution optimal? Yes, but with a topological asterisk. If the particle's path is smooth and unbroken, it is globally optimal. If the path is interrupted by nodes, it is only locally optimal. This suggests that nodes are not just mathematical curiosities. They are markers of a system that is fundamentally poised for change.