Non-relativistic limit of generalized relativistic Pauli operators by Feynman-Kac formulae
Soichiro Sakamoto · 2026
Read Non-relativistic limit of generalized relativistic Pauli operators by Feynman-Kac formulae by Soichiro Sakamoto (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-06099.txt.
Introduction
Soichiro Sakamoto studies the non-relativistic limit of a generalized relativistic Pauli operator using Feynman-Kac formulae. It extends the relativistic Pauli operator through the Bernstein function setting, with parameters tied by a constraint. Its heat flow can be written as an average over Brownian motion joined with a subordinator and a Poisson process.
Using that form, Sakamoto proves the flow converges strongly to a simpler non-relativistic one as the speed of light grows without bound. The limiting generator has an explicit form built from the Pauli matrices, the vector potential, and the scalar potential. The paper also treats the non-relativistic limit of a generalized relativistic Schrödinger operator by the same approach.
Catalog record
| CREATOR | Soichiro Sakamoto |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 30 KB |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2605.06099
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Non-relativistic limit of generalized relativistic Pauli operators by Feynman-Kac formulae Soichiro Sakamoto Abstract The non-relativistic limit of a generalized relativistic Pauli operator H c S , α = ( 2 c β ( σ ⋅ ( − i ∇ − a ) ) 2 + ( m c γ ) 2 / α ) α / 2 − m c γ + V H_{c}^{S,\alpha}=\left(2c^{\beta}\bigl(\sigma\cdot(-i\nabla-a)\bigr)^{2}+(mc^{\gamma})^{2/\alpha}\right)^{\alpha/2}-mc^{\gamma}+V on L 2 ( ℝ 3 , ℂ 2 ) L^{2}(\mathbb{R}^{3};\mathbb{C}^{2}) is investigated under the constraint 2 α = γ β + γ 2 2\alpha=\gamma\beta+\gamma^{2} . This operator generalizes the relativistic Pauli operator within the framework of Bernstein functions. The associated heat semigroup e − t H c S , α e^{-tH_{c}^{S,\alpha}} admits a Feynman–Kac representation involving Brownian motion, a subordinator, and a Poisson process. Using this representation, we prove that the semigroup e − t H c S , α e^{-tH_{c}^{S,\alpha}} converges strongly to e − t H S , α e^{-tH^{S,\alpha}} as c → ∞ c\to\infty , where the limiting generator is given by H S , α = α 2 m 2 α − 1 ( σ ⋅ ( − i ∇ − a ) ) 2 + V . H^{S,\alpha}=\frac{\alpha}{2m^{\frac{2}{\alpha}-1}}\bigl(\sigma\cdot(-i\nabla-a)\bigr)^{2}+V. The non-relativistic limit of a generalized relativistic Schrödinger operator is also investigated. 1 Introduction In this paper, we study the relativistic Schrödinger operator with a vector potential a , H_{c}=\sqrt{c^{2}(-i\nabla-a)^{2}+m^{2}c^{4}}-mc^{2}+V acting on L^{2}(\mathbb{R}^{d}) , and the relativistic Pauli operator H_{c}^{\mathrm{S}}=\sqrt{c^{2}\bigl(\sigma\cdot(-i\nabla-a)\bigr)^{2}+m^{2}c^{4}}-mc^{2}+V acting on L^{2}(\mathbb{R}^{d};\mathbb{C}^{2}) . Here \sigma=(\sigma_{1},\sigma_{2},\sigma_{3}) denotes the Pauli matrices. The operator H_{c}^{\mathrm{S}} incorporates the spin- \tfrac{1}{2} degree of freedom, whereas H_{c} does not. Both H_{c} and H_{c}^{\mathrm{S}} describe relativistic particles minimally coupled to the vector potential a . The function V represents an external potential, c>0 is the speed of light, and m>0 is the particle mass. The purpose of this paper is to investigate the non-relativistic limit of the semi-groups e^{-tH_{c}^{\#}} as c\to\infty , where the symbol \# stands for either the scalar or spin case. Indeed, since \sqrt{c^{2}u+m^{2}c^{4}}-mc^{2}=\frac{1}{2m}u+O\!\left(\frac{u^{2}}{m^{3}c^{2}}\right), we formally expect that e^{-tH_{c}^{\#}}\;\longrightarrow\;e^{-tH^{\#}} in the strong sense as c\to\infty . Here \displaystyle H=\frac{1}{2m}(-i\nabla-a)^{2}+V, \displaystyle H^{\mathrm{S}}=\frac{1}{2m}\bigl(\sigma\cdot(-i\nabla-a)\bigr)^{2}+V, are the non-relativistic Schrödinger and Pauli operators, respectively. In this paper we further introduce a class of generalized operators. Consider the function \Psi_{c}(u)=(2c^{2}u+m^{2}c^{4})^{1/2}-mc^{2}, which is a Bernstein function, i.e, (-1)^{n+1}\Psi_{c}^{(n)}(u)\geq 0 for all u\geq 0 . Also, \Psi_{c}(0)=0 is satisfied. Then the operators H_{c} and H_{c}^{\mathrm{S}} can be expressed as \displaystyle H_{c}=\Psi_{c}\!\left(\tfrac{1}{2}(-i\nabla-a)^{2}\right)+V, \displaystyle H_{c}^{\mathrm{S}}=\Psi_{c}\!\left(\tfrac{1}{2}(\sigma\cdot(-i\nabla-a))^{2}\right)+V. We generalize this construction by replacing \Psi_{c} with a family of Bernstein functions \Psi_{\alpha,\beta,\gamma,c}(u)=\bigl(2c^{\beta}u+(mc^{\gamma})^{2/\alpha}\bigr)^{\alpha/2}-mc^{\gamma},\qquad 0<\alpha<2. Under the condition 2\alpha=\beta\gamma+\gamma^{2} , one verifies that \Psi_{\alpha,\beta,\gamma,c}(u)\longrightarrow\frac{\alpha}{2m^{\frac{2}{\alpha}-1}}\,u as c\to\infty . The original function \Psi_{c} corresponds to the special case \alpha=1 and \beta=\gamma=2 . Accordingly, we introduce the generalized operators: \displaystyle H_{c}^{\alpha}=\Psi_{\alpha,\beta,\gamma,c}\!\left(\tfrac{1}{2}(-i\nabla-a)^{2}\right)+V, \displaystyle H_{c}^{{\mathrm{S}},\alpha}=\Psi_{\alpha,\beta,\gamma,c}\!\left(\tfrac{1}{2}(\sigma\cdot(-i\nabla-a))^{2}\right)+V. The main purpose of this paper is to prove that e^{-tH_{c}^{\#,\alpha}}\;\longrightarrow\;e^{-tH^{\#,\alpha}} as c\to\infty , where \displaystyle H^{\alpha}=\frac{\alpha}{2m^{\frac{2}{\alpha}-1}}(-i\nabla-a)^{2}+V, \displaystyle H^{\mathrm{S},\alpha}=\frac{\alpha}{2m^{\frac{2}{\alpha}-1}}\bigl(\sigma\cdot(-i\nabla-a)\bigr)^{2}+V. Our approach is based on the Feynman–Kac formula (FKF) for semi-groups e^{-tH_{c}^{\#,\alpha}} . The FKF for e^{-tH_{c}} and e^{-tH_{c}^{\mathrm{S}}} were derived in [ JHV2020 ] . The corresponding formulae for e^{-tH_{c}^{\alpha}} and e^{-tH_{c}^{{\mathrm{S}},\alpha}} can also be established by a slight modification of these arguments. In particular, the FKF for e^{-tH_{c}^{{\mathrm{S}},\alpha}} involves three independent stochastic processes: a Brownian motion, a subordinator associated with the Bernstein function \Psi_{\alpha,\beta,\gamma,c} , and a spin process driven by a Poisson process. Finally, we briefly review related results. A path integral representation of the semi-group e^{-tH} is known as the Feynman–Kac–Itô formula [ Sim1979 ] . Ichinose and Tamura extended this framework to the relativistic Hamiltonian H_{c} in [ IT1986 ] , although their representation differs from the one employed in the present work. Based on these developments, Ichinose proved the non-relativistic limit in [ Ichi1987 ] in the spinless case. The path integral representations of the heat semi-groups for the Pauli operator and the relativistic Pauli operator are given by [ GGM1983 , BJ1981 ] and [ GAM1991 ] , respectively, and it is generalized in [ HIL2009 ] . The non-relativistic limit of the Dirac operator has also been investigated from an operator-theoretic perspective in [ Hun1975 , Tha1992 ] . Acknowledgments The author would like to express his deepest gratitude to Professor Fumio Hiroshima for his invaluable comments on the structure of this paper and for meticulously pointing out several mathematical errors in an earlier draft. His suggestions were instrumental in improving both the logical clarity and the mathematical rigor of this work. 2 \alpha/2 -relativistic subordinator 2.1 Bernstein functions and subordinators We introduce an \alpha/2 -relativistic subordinator, which plays an important role in the subsequent stochastic arguments. Bernstein functions introduced below are treated in [ App2009 , JHV2020 ] . Definition 2.1 . (Bernstein function) Let \mathscr{B}=\biggl\{f\in C^{\infty}((0,\infty))\bigg|f(x)\geq 0\;\text{and}\;(-1)^{n}\frac{d^{n}}{dx^{n}}f(x)\leq 0\;\text{for}\;\text{all}\;n\in\mathbb{N}\biggr\}. An element of \mathscr{B} is called a Bernstein function. Furthermore, we define a subset of \mathscr{B} : \mathscr{B}_{0}=\{f\in\mathscr{B}\;|\lim_{x\to 0+}f(x)=0\}. We introduce a set of measures. Let \mathscr{L} be the set of Borel measures \lambda on \mathbb{R}\backslash\{0\} such that \lambda((-\infty,0))=0 and \int_{\mathbb{R}\backslash\{0\}}(y\wedge 1)\lambda(dy)<\infty. In particular, \lambda\in\mathscr{L} is a Lévy measure, i.e, \int_{\mathbb{R}\backslash\{0\}}(y\wedge 1)^{2}\lambda(dy)<\infty. Proposition 2.2 . Let \Psi\in\mathscr{B}_{0} . Then there exists a unique (b,\lambda)\in[0,\infty)\times\mathscr{L} such that \Psi(u)=bu\;+\;\displaystyle\int_{0}^{\infty}(1-e^{-uy})\lambda(dy). (2.1) Conversely, the right-hand side of (2.1) is in \mathscr{B}_{0} for each (b,\lambda)\in\mathbb{R}_{+}\times\mathscr{L} . There exists a fundamental relationship between subordinators and \mathscr{B}_{0} . This relationship is indispensable for establishing the path integral representation of generalized relativistic Pauli operators. Let (X_{t})_{t\geq 0} be a random process on a probability space (\Omega,\mathcal{F},P) . From now on, we write \mathbb{E}_{P}^{x}[f(X_{t})]=\int_{\Omega}f(X_{t}(\omega)+x)dP(\omega). Proposition 2.3 . Let \Psi\in\mathscr{B}_{0} , and (b,\lambda)\in\mathbb{R}_{+}\times\mathscr{L} be a pair satisfying (2.1). Then there exists a unique subordinator (T_{t})_{t\geq 0} on a probability space (\Omega,\mathcal{F},P) such that \mathbb{E}_{P}^{0}[e^{-uT_{t}}]=e^{-t\Psi(u)}. (2.2) Conversely, let (T_{t})_{t\geq 0} be a subordinator on (\Omega,\mathcal{F},P) . Then there exists a unique \Psi\in\mathscr{B}_{0} satisfying (2.2). Proof. See [ JHV2020 , Proposition 3.99] . ∎ 2.2 \alpha/2 -relativistic subordinator Let 0<\alpha<2 , and choose \beta,\gamma>0 such that 2\alpha=\beta\gamma+{\gamma}^{2} . Let \Psi_{c}^{\alpha}(u)=(2c^{\beta}u+(mc^{\gamma})^{2/\alpha})^{\alpha/2}-mc^{\gamma},\quad u\geq 0. (2.3) We can directly see that \Psi_{c}^{\alpha}\in\mathscr{B}_{0} for each c>0 . Let (\Omega,\mathcal{F},P) be a probability space. We call (T_{t}^{c})_{t\geq 0} satisfying (2.4) below an \alpha/2 -relativistic subordinator parametrized by c>0 : \mathbb{E}_{P}^{0}[e^{-uT_{t}^{c}}]=e^{-t\Psi_{c}^{\alpha}(u)}. (2.4) The right-hand side of (2.4) can be represented by \exp(-t\Psi_{c}^{\alpha}(u))={\exp}\Biggl({\frac{{\alpha}(2c^{\beta})^{\frac{\alpha}{2}}t}{2\Gamma(1-{\alpha/2)}}\displaystyle\int_{0}^{\infty}(1-e^{-uy})e^{\frac{-(mc^{\gamma})^{2/\alpha}}{2c^{\beta}}}\frac{dy}{y^{1+\alpha/2}}\Biggr)}. (2.5) See [ Bog2009 , Example 5.9] . Since T_{t}^{c} is a Lévy process, we can see that Lévy triplet of (T_{t}^{c})_{t\geq 0} is given by the form (b,0,\mu) . Here b{\in}\mathbb{R} and \mu is a Lévy measure. By Lévy–Khintchine formula, we have \mathbb{E}_{P}^{0}[e^{iuT_{t}^{c}}]=\exp\Biggl(itbu+t\displaystyle\int_{\mathbb{R}\backslash\{0\}}(e^{iuy}-1-iuy\mathbbm{1}_{\{|y|\leq 1\}})\mu(dy)\Biggr). (2.6) The function \mathbb{R}{\ni}u\mapsto\mathbb{E}_{P}^{0}[e^{iuT_{t}^{c}}] can be analytically continued into the region \{iu\in\mathbb{C}\;|\;u>0\} . Hence we can obtain the Lévy triplet of (T_{t}^{c})_{t\geq 0} by the analytical continuation of both sides of (2.6) and comparing the resulting expression with (2.5): (b,0,\mu)=\left(\displaystyle\int_{0}^{1}y\nu_{c}(dy),0,\nu_{c}\right), (2.7) where \nu_{c}(dy)=(2c^{\beta})^{\frac{\alpha}{2}}\frac{\alpha}{2\Gamma\left(1-\frac{\alpha}{2}\right)}\exp\left(-\frac{(mc^{\gamma})^{2/\alpha}}{2c^{\beta}}y\right)\frac{1}{y^{1+\alpha/2}}dy. By (2.7) and Lévy-Itô decomposition, we can obtain a stochastic integral representation of T_{t}^{c} by T_{t}^{c}=t\displaystyle\int_{0}^{1}z\nu_{c}(dz)+\displaystyle\int_{1}^{\infty}zN_{c}(t,dz)+\displaystyle\int_{0}^{1}z\tilde{N_{c}}(t,dz), (2.8) where N_{c}(t,dz) denote a Poisson random measure and \tilde{N_{c}}(t,dz) a compensated Poisson random measure associated with (T_{t}^{c})_{t\geq 0} . Note that (N_{c}(t,\cdot))_{t\geq 0} is the Poisson process with intensity \nu_{c}(A)=\mathbb{E}_{P}^{0}[N_{c}(1,A)] for each A\subset[0,\infty) . Furthermore, \tilde{N_{c}}(t,A)=N_{c}(t,dz)-t\nu_{c}(A) , and (\tilde{N}_{c}(t,A))_{t\geq 0} is a martingale for each A\subset[0,\infty) . We investigate the exponent of (T_{t}^{c})_{t\geq 0} and limiting behavior of its expectation as c\to\infty . This computation is used