Point-Interaction Schrödinger Kernels

• Point-interaction Schrödinger semigroup kernels are integral kernels for quantum systems with zero-range interactions, capturing analytic, spectral, and probabilistic properties.
• Explicit formulas in one and two dimensions highlight differences in singularity behavior and validate the kernel’s construction via resolvent analysis and self-adjoint extensions.
• Applications include modeling quantum diffusion, scattering phenomena, and constructing stochastic processes through Doob transforms, bridging PDE techniques with probabilistic insights.

A point-interaction Schrödinger semigroup kernel is the transition kernel (integral kernel) of the contraction or unitary semigroup generated by a Schrödinger operator that includes a zero-range (point) interaction, most notably as a Dirac-δ or δ'-type perturbation. Such kernels allow for the analytic, spectral, and probabilistic investigation of quantum and stochastic systems where all interaction is concentrated at isolated points. The form of these kernels, their regularity, and their behavior, especially near the interaction location and in various time limits, are central to understanding diffusion, transmission, and reflection phenomena in singular-perturbation models. The rigorous construction and explicit forms of these kernels, as derived for one- and two-dimensional cases, provide the foundation for various physical and mathematical applications, including stochastic processes with singular drift and spectral theory of singular perturbations (Hussein et al., 2019, Kovarik et al., 2010, Mian, 24 Jan 2026).

1. Analytic Construction of Point-Interaction Schrödinger Semigroup Kernels

The prototypical point-interaction Hamiltonian in one dimension is constructed as a self-adjoint extension of the Laplacian, typically on a domain such as the real line or a pair of half-lines coupled at a point. For the two half-line setting, all such extensions are parametrized by a matrix $U \in U(2)$, yielding boundary conditions of the form

$(U - I)\Psi(0) + i(U + I)\Psi'(0) = 0,$

where $\Psi(0)$, $\Psi'(0)$ are the boundary value and the outgoing derivatives at the junction respectively. For each $U$, the generator $A_U$ defines a semigroup whose kernel $K_t^U(x, y)$ may be constructed via resolvent analysis, employing the Cayley transform and the method of images, followed by Laplace inversion (Hussein et al., 2019).

In one dimension with two delta interactions at $\pm a$, the kernel is decomposed as a sum of the free kernel and a series of correction terms involving parabolic-cylinder functions, obtained via resolvent kernel expansion and Stone’s formula. For higher-dimensional constructions such as the two-dimensional delta interaction at the origin, the singularity of the delta requires self-adjoint extension or renormalization. The resulting semigroup kernel admits an exact integral representation in terms of the two-parameter Volterra function and the free heat kernel (Mian, 24 Jan 2026).

2. Explicit Forms and Special Cases

A variety of explicit kernel formulas are available for standard types of point interactions:

• One-dimensional settings (two coupled half-lines at the origin):
  • Dirichlet coupling ($U = -I$):

  $$K_t^D(x, y) = \frac{1}{\sqrt{4\pi t}}\left[e^{-\frac{(x-y)^2}{4t}} - e^{-\frac{(x + y)^2}{4t}}\right]$$ - Neumann coupling ($U = +I$):

  $$K_t^N(x, y) = \frac{1}{\sqrt{4\pi t}}\left[e^{-\frac{(x-y)^2}{4t}} + e^{-\frac{(x + y)^2}{4t}}\right]$$ - $\delta$-interaction (coupling by jump condition):

  $$K_t^\delta(x, y) = K_t^N(x, y) - \frac{\alpha}{2} \int_0^t K_s^N(x, 0)K_{t-s}^N(0, y)ds$$

  with a further closed form involving the complementary error function. - $\delta'$-interaction (coupling by derivative jump): a corresponding formula using erfc and exponential terms (Hussein et al., 2019).

• One-dimensional two-delta (double well):

$$K_\alpha(t; x, y) = \sum_{j: E_j < 0} e^{-iE_j t} \phi_j(x)\overline{\phi_j(y)} + U_\alpha(t; x, y)$$

where $U_\alpha$ is an infinite sum over corrections involving parabolic-cylinder functions. The expansion converges uniformly for fixed $t>0$, allowing decomposition into free propagation and multiple-scattering (interaction) contributions (Kovarik et al., 2010).

• Two-dimensional delta interaction:

$$K_\alpha(t; x, y) = g_t(x - y) + 2\pi\alpha \int_{0<r<s<t} g_r(x)\nu'(\alpha(s-r))g_{t-s}(y)ds\,dr,$$

where $g_t(z)$ is the free planar heat kernel and $\nu$ is the Volterra function (Mian, 24 Jan 2026).

3. Boundary Conditions, Self-Adjoint Extensions, and Spectral Structure

The theory of point-interaction semigroups is fundamentally governed by the parameterization of self-adjoint extensions. In one dimension, the $U(2)$-parametrization encodes all possible self-adjoint extensions of the Laplacian coupled at a point. The specific form of $U$ prescribes physical conditions such as continuity, jump, and reflection-transmission behavior at the junction. For example, the standard $\delta$ and $\delta'$ interactions arise from continuity and jump (or their dual) constraints.

In two dimensions, the delta perturbation is too singular to be treated by straightforward potential theory. The self-adjoint extension, or equivalently a logarithmic boundary condition at the origin, prescribes the zero-range interaction in terms of an asymptotic relation between coefficients of the singular and regular parts of the wavefunction near the interaction point (Mian, 24 Jan 2026). The coupling constant $\alpha$ then determines the linear relation between these coefficients.

The spectral decomposition for such singular Hamiltonians typically features an absolutely continuous spectrum (e.g., $[0, +\infty)$), with the possibility of negative eigenvalues (bound states) when the interaction is attractive. These bound states contribute discrete terms to the time-evolution/semigroup kernel, while the absolutely continuous spectrum underlies the integral (diffusive) part (Kovarik et al., 2010).

4. Singularity Structure, Asymptotics, and Scaling Limits

The singularity structure of the kernel near the interaction point is a characteristic feature:

• One dimension: The image term (method-of-images formula) is exponentially suppressed away from the junction, but dominates as $x, y \to 0$ or $t \to 0^+$. As $t \to 0^+$, all kernels recover the free diffusion singularity.
• Two dimensions: As $x$ or $y \rightarrow 0$, the kernel diverges logarithmically, in agreement with the logarithmic boundary condition on the domain. The perturbative component to the kernel behaves as $C(t, y)\ln|x|^{-1} + O(1)$ for some positive $C$ and fixed $y \neq 0$.
• Time asymptotics: As $t \to \infty$, decay on the diagonal is uniformly $O(t^{-1/2})$ in the one-dimensional point-interaction case, but the sign and size of the kernel are modulated by the choice of boundary conditions (matrix $U$) (Hussein et al., 2019). In two-dimensional cases, the point-interaction effect persists at all times through the logarithmic singularity, and the kernel decays as $t^{-1}$ for $x \neq y$ (Mian, 24 Jan 2026).
• Scaling limits: When parameters such as the separation between two delta interactions ($a$) or coupling strength ($\alpha$) are taken to limiting values, the kernel converges to well-known forms—free, hard-wall (Dirichlet), or specific one-delta point-interaction kernels (Kovarik et al., 2010).

5. Applications: Markov Diffusions and Doob Transforms

Point-interaction semigroup kernels serve as fundamental objects in the construction of stochastic processes with singular drift. In two dimensions, performing a Doob–h-transform with the ground-state (generalized eigenfunction) of the point-interaction Hamiltonian yields a continuous Markov process, the so-called “skew-product diffusion,” governed by the SDE: $dX_t = dW_t + \nabla\ln\phi_\alpha(X_t)dt,$ where $\phi_\alpha(x) = K_0(\sqrt{2\alpha}|x|)$ is the modified Bessel function of order zero (Mian, 24 Jan 2026). The kernel of the corresponding semigroup can be explicitly written as: $$K_\alpha^{\rm Doob}(t; x, y) = e^{t\alpha}\frac{\phi_\alpha(y)}{\phi_\alpha(x)}K_\alpha(t; x, y).$$ Further, an axiomatic framework characterizes what driving families $h_t(x)$ lead to legitimate such diffusions, specifying regularity, harmonicity, and small-ball occupation controls to ensure the process does not “stick” at the singularity and admits SDE representations and explicit transition densities.

A plausible implication is that point-interaction semigroup kernels provide direct analytic routes to model processes conditioned on delayed absorption (conditioning on survival up to time $T$), time-inhomogeneous diffusions, and systems with explicit interaction at the origin.

6. Extensions, Generalizations, and Outlook

The resolvent and kernel-based construction extends to multiple point interactions at arbitrary positions, with the key algebraic quantities reducing to finite-dimensional determinants in the transfer-matrix approach. In such settings, the propagator splits into a sum of the free kernel and a finite (or infinite) sum of “multiple-scatter” terms, each expressible in terms of parabolic-cylinder or error functions (Kovarik et al., 2010). This suggests a robust and algorithmic approach to calculating transition kernels for arbitrary finite configurations of point interactions, with applications across quantum transport and stochastic processes with singular perturbations.

In summary, point-interaction Schrödinger semigroup kernels serve as sharp analytic tools for quantifying the dynamics, spectral properties, and probabilistic behavior of quantum and diffusive systems with singular local interactions. The explicit constructions in various dimensions, together with the spectrum of admissible boundary and matching conditions, underpin a comprehensive theory that connects functional analysis, PDE, probability, and mathematical physics (Hussein et al., 2019, Kovarik et al., 2010, Mian, 24 Jan 2026).