Algebraic approach to a d d -dimensional matrix Hamiltonian with so( OPEN d + 1 ) d+1) symmetry
C. Quesne · 2025
Read Algebraic approach to a d d -dimensional matrix Hamiltonian with so( OPEN d + 1 ) d+1) symmetry by C. Quesne (2025) 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-2508-07949.txt.
Introduction
Christiane Quesne presents a 2025 arXiv paper in mathematical physics, on a matrix Hamiltonian with rotational symmetry in any dimension. Quesne introduces a spin-extended version of the so(d+1,1) algebra and shows it frames the system neatly. Extra operators complete a basis for one of the algebra's representations.
Inside that frame, the constants of motion of the system in one representation fall out with little effort. Quesne then carries them into the Schrödinger picture, including a Laplace-Runge-Lenz vector with spin. The paper derives that vector and its properties inside the wider algebraic setting, which may aid other symmetric quantum systems.
Catalog record
| CREATOR | C. Quesne |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 33 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2508-07949
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Algebraic approach to a d d -dimensional matrix Hamiltonian with so( OPEN d + 1 ) d+1) symmetry C. Quesne Abstract A novel spin-extended so( d + 1 d+1 ,1) algebra is introduced and shown to provide an interesting framework for discussing the properties of a d d -dimensional matrix Hamiltonian with spin 1/2 and so( d + 1 d+1 ) symmetry. With some d + 2 d+2 additional operators, spanning a basis of an so( d + 1 d+1 ,1) irreducible representation, the so( d + 1 d+1 ,1) generators provide a very easy way for deriving the integrals of motion of the matrix Hamiltonian in Sturm representation. Such integrals of motion are then transformed into those of the matrix Hamiltonian in Schrödinger representation, including a Laplace-Runge-Lenz vector with spin. This leads to a derivation of the latter, as well as its properties in a more extended algebraic framework. Keywords: matrix Schrödinger equation; Laplace-Runge-Lenz vector; so( d+1 ) symmetry 1 Introduction The hydrogen atom is known to have a nice symmetry (for a review see, e.g., [ 1 ] ). Apart from the invariance of the Hamiltonian under rotations, leading to the conservation of the orbital angular momentum, it is also characterized by that of the Laplace-Runge-Lenz (LRL) vector, whose introduction in quantum mechanics dates back to Pauli [ 2 ] and which was then studied by Fock [ 3 ] , Bargmann [ 4 ] , and many other authors. As a result, the hydrogen atom is a maximally superintegrable system, being characterized by an so(4), e(3), or so(3,1) invariance algebra for negative-, zero-, and positive-energy states, respectively. All these algebras may be embedded into an so(4,2) dynamical algebra [ 5 , 6 ] . More generally, the d -dimensional Coulomb problem, whose Hamiltonian is defined by H=\frac{1}{2}{\bf p}^{2}+\frac{\alpha}{r} , where {\bf p}^{2}=\sum_{i=1}^{d}p_{i}^{2} , p_{i}=-{\rm i}\partial/\partial x_{i} , r=\left(\sum_{i=1}^{d}x_{i}^{2}\right)^{1/2} , and \alpha is some parameter, can be discussed in terms of an so( d+1 ,2) Lie algebra and its bound states turn out to be a basis of an so( d+1 ) algebra. As recently recalled [ 7 ] , this so( d+1 ,2) algebra is a good starting point for discussing the Coulomb problem in Sturm representation, characterized by the operator K=r\left(\frac{1}{2}{\bf p}^{2}-E\right) and for deriving the invariance algebra of the latter in a very simple way. From these results, it is then straightforward to obtain the LRL vector corresponding to the Schrödinger representation, as well as its properties. In the same study [ 7 ] , it was shown that such an approach can be applied to a generalization of the d -dimensional Coulomb problem to the d -dimensional Dunkl-Coulomb one, wherein the derivatives \partial/\partial x_{i} are replaced by Dunkl operators D_{i} [ 8 , 9 ] . The latter are differential-difference operators, defined by D_{i}=\partial/\partial x_{i}+(\mu_{i}/x_{i})(1-R_{i}) , i=1,2,\ldots,d , where \mu_{i} is some positive parameter and R_{i} is a reflection operator such that R_{i}f(x_{i})=f(-x_{i}) . This has led to some deformed algebras and to a deformed LRL vector. The purpose of the present paper is to apply the same method to another generalization of the d -dimensional Coulomb problem. Some years ago, a system describing a neutral particle with spin 1/2 and a non-trivial dipole momentum interacting with an external field inverse in radius in two [ 10 ] or three dimensions [ 11 ] was shown to be endowed with some generalized LRL vector. A similar study was then carried out for arbitrary spin [ 12 , 13 ] and generalized to a d -dimensional space [ 14 ] . We plan to analyze along the lines of [ 7 ] the d -dimensional matrix Hamiltonian with spin 1/2 considered in [ 14 ] . In Sect. 2, such a Hamiltonian is defined and some examples for low d values are presented. In Sect. 3, a spin-extended so( d+1 ,1) algebra is introduced together with some operators spanning an irreducible representation of the latter. In Sect. 4, such results are used to determine the invariance algebra of the Hamiltonian in Sturm representation. In Sect. 5, this invariance algebra is transformed into that in Schrödinger representation, thereby deriving the LRL vector with spin together with its properties. Finally, Sect. 6 contains the conclusion. 2 d -dimensional matrix problem with spin 1/2 Let us consider the d -dimensional Hamiltonian [ 14 ] H=\frac{1}{2}\mathbf{p}^{2}+\frac{\alpha}{r^{2}}\boldsymbol{\gamma}\cdot\mathbf{x}, (2.1) where \mathbf{p}^{2}=\sum_{i=1}^{d}p_{i}^{2} , p_{i}=-{\rm i}\partial/\partial x_{i} , r=\left(\sum_{i=1}^{d}x_{i}^{2}\right)^{1/2} , \boldsymbol{\gamma}\cdot\mathbf{x}=\sum_{i=1}^{d}\gamma_{i}x_{i} , and \alpha is some constant. Here \gamma_{i} , i=1,2,\ldots,d are matrices that are basis elements of the Clifford algebra \mathit{Cl}_{d} , with defining relations \gamma_{i}\gamma_{j}+\gamma_{j}\gamma_{i}=2\delta_{i,j}. (2.2) In terms of these \gamma_{i} ’s, the matrices S_{ij}=-\frac{\rm i}{4}(\gamma_{i}\gamma_{j}-\gamma_{j}\gamma_{i}). (2.3) satisfy the so( d ) commutation relations [S_{ij},S_{kl}]={\rm i}\left(\delta_{i,k}S_{jl}+\delta_{i,l}S_{kj}+\delta_{j,k}S_{li}+\delta_{j,l}S_{ik}\right). (2.4) As explained in [ 14 ] , Hamiltonian ( 2.1 ) may be interpreted as describing a particle of spin 1/2. We plan to deal with the Schrödinger equation H\Psi(\mathbf{x})=E\Psi(\mathbf{x}), (2.5) corresponding to ( 2.1 ), as well as the eigenvalue problem K\Psi(\mathbf{x})=-\alpha\Psi(\mathbf{x}), (2.6) for the operator K defined by K=(\boldsymbol{\gamma}\cdot\mathbf{x})\left(\frac{1}{2}\mathbf{p}^{2}-E\right) (2.7) and corresponding to the so-called Sturm representation. Since from ( 2.2 ), it follows that (\boldsymbol{\gamma}\cdot\mathbf{x})^{2}=r^{2}, (2.8) the operators H and K are related through the equations K=(\boldsymbol{\gamma}\cdot\mathbf{x})(H-E)-\alpha,\qquad H=\frac{\boldsymbol{\gamma}\cdot\mathbf{x}}{r^{2}}(K+\alpha)+E. (2.9) For some low d values, we may assume for instance the following \gamma_{i} ’s: \begin{split}&\bullet\quad d=2:\gamma_{1}=\sigma_{1},\quad\gamma_{2}=\sigma_{2},\\ &\bullet\quad d=3:\gamma_{1}=\sigma_{1},\quad\gamma_{2}=\sigma_{2},\quad\gamma_{3}=\sigma_{3},\\ &\bullet\quad d=4:\gamma_{i}=\begin{pmatrix}0&{\rm i}\sigma_{i}\\ -{\rm i}\sigma_{i}&0\end{pmatrix},\quad i=1,2,3,\quad\gamma_{4}=\begin{pmatrix}0&I\\ I&0\end{pmatrix},\\ &\bullet\quad d=5:\gamma_{i}=\begin{pmatrix}0&{\rm i}\sigma_{i}\\ -{\rm i}\sigma_{i}&0\end{pmatrix},\quad i=1,2,3,\quad\gamma_{4}=\begin{pmatrix}0&I\\ I&0\end{pmatrix},\\ &\hphantom{\bullet\quad d=5:{}}\;\;\gamma_{5}=\begin{pmatrix}I&0\\ 0&-I\end{pmatrix},\end{split}. (2.10) where \sigma_{1} , \sigma_{2} , \sigma_{3} denote the Pauli spin matrices and I is the 2x2 unit matrix. For such choices, Hamiltonian ( 2.1 ) is a 2x2 or a 4x4 matrix Hamitonian if d=2,3 or d=4,5 , respectively. The corresponding so( d ) generators ( 2.3 ) read \begin{split}&\bullet\quad d=2:S_{12}=\tfrac{1}{2}\sigma_{3},\\ &\bullet\quad d=3:S_{ij}=\tfrac{1}{2}\epsilon_{ijk}\sigma_{k},\quad i,j,k=1,2,3,\\ &\bullet\quad d=4:S_{ij}=\tfrac{1}{2}\epsilon_{ijk}\begin{pmatrix}\sigma_{k}&0\\ 0&\sigma_{k}\end{pmatrix},\quad i,j,k=1,2,3,\\ &\hphantom{\bullet\quad d=5:{}}\;\>S_{i4}=\tfrac{1}{2}\begin{pmatrix}\sigma_{i}&0\\ 0&-\sigma_{i}\end{pmatrix},\quad i=1,2,3,\\ &\bullet\quad d=5:S_{ij}=\tfrac{1}{2}\epsilon_{ijk}\begin{pmatrix}\sigma_{k}&0\\ 0&\sigma_{k}\end{pmatrix},\quad i,j,k=1,2,3,\\ &\hphantom{\bullet\quad d=5:{}}\;\>S_{i4}=\tfrac{1}{2}\begin{pmatrix}\sigma_{i}&0\\ 0&-\sigma_{i}\end{pmatrix},\quad i=1,2,3,\\ &\hphantom{\bullet\quad d=5:{}}\;\>S_{i5}=-\tfrac{1}{2}\begin{pmatrix}0&\sigma_{i}\\ \sigma_{i}&0\end{pmatrix},\quad i=1,2,3,\\ &\hphantom{\bullet\quad d=5:{}}\;\>S_{45}=\tfrac{{\rm i}}{2}\begin{pmatrix}0&I\\ -I&0\end{pmatrix},\\ \end{split}. (2.11) where \epsilon_{ijk} stands for the totally antisymmetric tensor with summation over repeated indices. 3 Spin-extended so( d+1 ,1) algebra Let us now introduce \tfrac{1}{2}(d+1)(d+2) operators defined by \begin{split}J_{ij}&=L_{ij}+S_{ij}=-J_{ji},\qquad L_{ij}=x_{i}p_{j}-x_{j}p_{i},\\ A_{i}&=\frac{1}{2}x_{i}\mathbf{p}^{2}-\left(\mathbf{x}\cdot\mathbf{p}-{\rm i}\frac{d-1}{2}\right)p_{i}-\frac{1}{2}x_{i}+S_{ij}p_{j},\\ M_{i}&=\frac{1}{2}x_{i}\mathbf{p}^{2}-\left(\mathbf{x}\cdot\mathbf{p}-{\rm i}\frac{d-1}{2}\right)p_{i}+\frac{1}{2}x_{i}+S_{ij}p_{j},\\ T&=\mathbf{x}\cdot\mathbf{p}-{\rm i}\frac{d-1}{2},\end{split} (3.1) where i and j run over 1, 2, …, d , \mathbf{x}\cdot\mathbf{p}=\sum_{i=1}^{d}x_{i}p_{i} , and there is a summation over repeated indices. It is straightforward to show that these operators satisfy the following commutation relations \begin{split}&[J_{ij},J_{kl}]={\rm i}\left(\delta_{i,k}J_{jl}+\delta_{i,l}J_{kj}+\delta_{j,k}J_{li}+\delta_{j,l}J_{ik}\right),\\ &[J_{ij},A_{k}]={\rm i}\left(\delta_{i,k}A_{j}-\delta_{j,k}A_{i}\right),\\ &[J_{ij},M_{k}]={\rm i}\left(\delta_{i,k}M_{j}-\delta_{j,k}M_{i}\right),\\ &[J_{ij},T]=0,\\ &[A_{i},A_{j}]=-[M_{i},M_{j}]={\rm i}J_{ij},\\ &[A_{i},M_{j}]={\rm i}\delta_{i,j}T,\\ &[A_{i},T]=-{\rm i}M_{i},\qquad[M_{i},T]=-{\rm i}A_{i}.\end{split}. (3.2) Here, as in the case where S_{ij}=0 , these spin-dependent operators are the generators {\cal L}_{ab}=-{\cal L}_{ba}={\cal L}_{ab}^{\dagger} of an so( d+1,1 ) algebra, whose commutation relations are given by [{\cal L}_{ab},{\cal L}_{cd}]={\rm i}(g_{ac}{\cal L}_{bd}+g_{ad}{\cal L}_{cb}+g_{bc}{\cal L}_{da}+g_{bd}{\cal L}_{ac}), (3.3) where g_{ab}={\rm diag}(1,1,\ldots,1,-1) . The identifications are the following ones: \begin{split}&{\cal L}_{ij}=J_{ij},\qquad i,j=1,2,\ldots,d,\\ &{\cal L}_{i,d+1}=A_{i},\qquad{\cal L}_{i,d+2}=M_{i},\qquad i=1,2,\ldots,d,\\ &{\cal L}_{d+1,d+2}=T.\end{split} (3.4) The second-order Casimir operator of this algebra reads Q_{2}=\mathbf{J}^{2}+\mathbf{A}^{2}-\mathbf{M}^{2}-T^{2}. (3.5) As shown in appendix A, it reduces to the constant Q_{2}=-\tfrac{1}{8}(d-1)(d+2), (3.6) instead of Q_{2}=-\frac{1}{4}(d-1)(d+1) , obtained in the S_{ij}=0 case. Let us now introduce d+2 additional operators, defined by \begin{split}&\Gamma_{0}=\tfrac{1}{2}(\boldsymbol{\gamma}\cdot\mathbf{x})(\mathbf{p}^{2}+1),\\ &\Gamma_{d+1}=\tfrac{1}{2}(\boldsymbol{\gamma}\cdot\mathbf{x})(\mathbf{p}^{2}-1),\\ &\Gamma_{i}=(\boldsymbol{\gamma}\cdot\mathbf{x})p_{i},\qquad i=1,2,\ldots,d.\end{split} (3.7) As in the standard d -dimensional Coulomb problem, where \boldsymbol{\gamma}\cdot\mathbf{x} is replace by r in ( 3.7 ) and S_{ij}=0 in ( 3.1 ) (see [ 1