Retroreflective surface optimisation for optical cavities with custom mirror profiles

William J. Hughes, Peter Horak · 2025

Read Retroreflective surface optimisation for optical cavities with custom mirror profiles by William J. Hughes, Peter Horak (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-2508-14712.txt.

Introduction

William Hughes and Peter Horak reshape the mirrors of optical cavities to grip light more tightly. A cavity's grip on a trapped emitter sets the strength of the light-matter link used in quantum devices. Round mirrors are standard, but custom shapes could do better; searching shape space directly is a numerical swamp of local optima.

Their two-step method dodges the swamp. First they optimize the cavity's light mode itself for the job at hand. Then they build mirror surfaces that bounce that exact mode back on itself. For one central emitter, grip rose about threefold across many shapes. For several emitters sharing one mode, gains passed tenfold.

Catalog record

CREATORWilliam J. Hughes, Peter Horak
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE44 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2508-14712

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Retroreflective surface optimisation for optical cavities with custom mirror profiles William J. Hughes, Peter Horak Abstract Coupling an emitter to a Fabry-Pérot optical cavity can provide a coherent and strong light-matter interface whose performance in a variety of applications depends critically on the emitter-photon coupling strength. Altering the typically spherical profiles of the cavity mirrors can improve this coupling strength, but directly optimising the mirror shape is numerically challenging as the multidimensional parameter space features many local optima. Here, we develop a two-step method to optimise mirror surface profiles while avoiding these issues. First, we optimise the target cavity eigenmode for the chosen application directly, and second, we construct the mirror surfaces to retroreflect this optimised target mode at both ends of the cavity. We apply our procedure to different emitter-cavity coupling scenarios. We show that mirror shaping can increase the cooperativity of coupling to a central emitter by a factor of approximately 3 across a wide range of geometries, and that, for coupling two or more emitters to a single cavity mode, the improvement factors can far exceed an order of magnitude. I Introduction Fabry Pérot optical cavities can offer greatly-enhanced, and even near-deterministic interfaces between quantum emitters and single photons Keller et al. 2004 ; McKeever et al. 2004 ; Bochmann et al. 2010 ; Grinkemeyer et al. 2025 while preserving high optical access Hartung et al. 2024 . These improvements promise significant performance increases in many applications of quantum technology, such as quantum networking Ritter et al. 2012 ; Reiserer and Rempe 2015 ; Krutyanskiy et al. 2023a , modular quantum computation Monroe et al. 2014 ; Li and Thompson 2024 , and quantum-enhanced sensing Petrak et al. 2014 . These cavities have typically been built with spherical mirrors Law and and 1997 ; Krutyanskiy et al. 2019 , yielding Gaussian cavity modes with well-known properties Siegman 1986 ; Yariv 1991 . However, proposals have suggested that non-spherical mirrors could confer advantageous features to the cavity modes Ferdous et al. 2014 ; Podoliak et al. 2017 ; Karpov and Horak 2022a ; Karpov et al. 2023 . As the field pivots towards microcavities in an effort to reduce footprint, improve robustness and scalability, and to reduce interaction timescales Steiner et al. 2013 ; Saavedra et al. 2021 ; Pfeifer et al. 2022 ; Teller et al. 2023 ; Zifkin et al. 2024 , alternative methods to traditional superpolishing are required to create the highly curved mirrors Steiner et al. 2013 ; Jin et al. 2022 . This makes mirror shaping an increasingly compelling prospect, as micromirror fabrication techniques like focused ion beam milling Dolan et al. 2010 and, to a lesser extent, laser ablation Steinmetz et al. 2006 ; Hunger et al. 2010 offer control over the surface topography Trichet et al. 2015 ; Walker et al. 2021 ; Maier et al. 2025 ; Ott et al. 2016 ; Gao et al. 2025 . However, the optimisation of a cavity mirror surface is numerically difficult because the relation between the mirror topography and the structure of the resonant modes is not straightforward (outside of perturbation assumptions van Exter et al. 2022 and specific geometric regimes Nyman and Szymańska 2014 ). This means that any reasonably flexible parameterisation of the surface features a vast parameter space which typically contains many local optima. It is therefore difficult to design the optimum surface for a given task through direct optimisation of the topography without working in a limited parameter space Karpov and Horak 2022a that may hide the possibility for performance improvements. In this manuscript, we present a two-stage procedure to optimise a cavity for a user-specified goal. The first step is to optimise the cavity mode, instead of the mirror surface, to take advantage of the (typically) simpler optimisation landscape. The second step is to construct the cavity surface to retroreflect this target mode. We show that this approach finds cavities whose performance exceeds any cavity with spherical mirrors, with the potential improvements particularly stark for cavities that are designed to couple to multiple emitters simultaneously. II Mode parameterisation and optimisation Our method will optimise the optical field inside Fabry-Pérot cavities, and we therefore need to describe a general intracavity field. We assume that this intracavity field is monochromatic and not highly divergent from the cavity axis, which is generally a good assumption for Fabry-Pérot cavities that are typically longer than they are wide. Under these assumptions, the intracavity field can be described using the paraxial equation. Formally, the intracavity electric field is a superposition of forward and backward running waves, \bm{E}^{\pm}(\bm{r},t) at spatial position \bm{r}=(x,y,z) and time t , which are specified through a scalar field \epsilon^{\pm}(\bm{r}) Barré et al. 2017 that satisfies \displaystyle\bm{E}^{\pm}(\bm{r},t) \displaystyle=\bm{{\epsilon}}^{(u)}\epsilon^{\pm}(\bm{r})\exp(\mp ikz)\exp(i\omega t), (1) \displaystyle\partialderivative{z}\epsilon^{\pm}(\bm{r}) \displaystyle=\mp\frac{i}{2k}\left(\partialderivative[2]{x}+\partialderivative[2]{y}\right)\epsilon^{\pm}(\bm{r}). (2) with \omega the angular frequency of the field, k=\omega/c the wavevector, \bm{{\epsilon}}^{(u)} the fixed, unit-magnitude polarisation of the field, which must be perpendicular to the z -axis, and \pm indicating propagation towards or away from positive z respectively. The paraxial equation (Eq. 2 ) is satisfied by Gaussian beam solutions, which are particularly convenient for Fabry-Pérot cavities because they are transversely localised and satisfy the reflection boundary conditions imposed by spherical mirrors. We will be concerned with radially symmetric mirrors, and will therefore use the Laguerre-Gauss modes, which may be written \displaystyle\epsilon^{\pm}_{l,m}\left(\bm{r}\right)= \displaystyle\sqrt{\frac{2l!}{\pi\left(l+\absolutevalue{m}\right)!}}\frac{1}{w(z)}\left(\frac{\sqrt{2}r}{w(z)}\right)^{\absolutevalue{m}}L^{\absolutevalue{m}}_{l}\left(\frac{2r^{2}}{w^{2}(z)}\right) (3) \displaystyle\exp\left\{-\frac{r^{2}}{w^{2}(z)}-i\left(m\phi+\frac{kr^{2}}{2R(z)}-\left(2l+\absolutevalue{m}+1\right)\psi_{G}\right)\right\}, where r , \phi , and z are the conventional coordinates of a cylindrical polar system, with \displaystyle w(z)=w_{0}\sqrt{1+\left(\frac{z}{z_{0}}\right)^{2}}, \displaystyle z_{0}=\frac{\pi w_{0}^{2}}{\lambda}, (4) \displaystyle R_{u}(z)=z\left(1+\left(\frac{z_{0}}{z}\right)^{2}\right), \displaystyle\Phi_{G}(z)=\arctan\left(\frac{z}{z_{0}}\right), where \lambda=2\pi/k is the wavelength, L^{\alpha}_{n} are the associated Laguerre polynomials, and z_{0} is the Rayleigh range of the beam. The coordinate origin has been fixed to lie at the centre of the smallest waist of the beam. The Laguerre-Gauss modes are indexed by radial index l\in\mathbb{N}_{0} and azimuthal index m\in\mathbb{Z} . They form a complete basis ( {\left\{\epsilon_{l,m}\right\}} ) Nienhuis and Allen 1993 , which means that any field satisfying the paraxial equation can be written as a linear superposition of these basis modes, \epsilon^{\pm}(\bm{r})=\sum_{s=0}^{S}C_{s}\epsilon^{\pm}_{s}(\bm{r}), (5) provided the basis size S+1 is sufficiently large, where (l,m) is mapped, for notational convenience, to a single index s , and C_{s} is the complex coefficient of the basis state \epsilon^{\pm}_{s} . We can thus identify any solution to the paraxial equation with a complex vector \vec{C} \vec{C}=\{C_{0},C_{1},...,C_{S}\}^{T}. (6) Our method will optimise the cavity mode for a given task by optimising the coefficients of this vector. The process by which the coefficients of \vec{C} are optimised depends upon the metric of performance that is being optimised. For the general case where only the performance metric (but not the gradient) can be calculated for a given set of coefficients, one may use local or global optimisation. For a substantial number of interesting cases, the performance metric may be written as the product and quotient of expectation values of Hermitian matrices. In this case, the gradient can be evaluated directly (see App. A ), which allows for faster local optimisation. For simplicity, we will only optimise modes with no azimuthal dependence ( m=0 ) in this manuscript. Additionally, we will only be interested in metrics that are symmetric upon reflection through the central transverse plane of the cavity. We

Read the complete text