--- title: "Non Gaussian Photonic Crystal Surface Modes for Robust Trapping of Ultracold Atoms" authors: ["Valery N. Konopsky"] affiliation: "Institute of Spectroscopy, Russian Academy of Sciences, Troitsk, Moscow region, Russia." journal: "Optical and Quantum Electronics" year: 2026 volume: "" issue: "" article_number: "" pages: "" doi: "" type: journal-article site_group: "" url_abstract: "" url_pdf: "https://valery.konopsky.com/kvnlocal/Konopsky4OQE2026_preprint.pdf" language: en source_tex: "Z:\\ValeryData\\Valery_New\\my_articles\\OQE2026\\OQE\\manuscript\\Konopsky4OQE2025.tex" source_pdf: "Z:\\ValeryData\\Valery_New\\my_articles\\OQE2026\\OQE\\published\\Konopsky4OQE2026_preprint.pdf" --- ## Abstract We present a photonic-crystal-based rib waveguide design for atom-chip applications that enables robust trapping of ultracold atoms. Using the generalized Effective Index Method (gEIM) together with concepts from ``scale-invariant'' non-Gaussian photonics, we design a geometry that supports a flat-top attractive mode and an edge-enhanced repulsive mode. The structure, fabricated on a 1D photonic crystal substrate, sustains surface modes at 850\,nm (red-detuned, attractive) and 640\,nm (blue-detuned, repulsive) with effective refractive indices near unity, yielding extended evanescent penetration into vacuum. This configuration enhances lateral confinement, mitigates atomic losses near the waveguide edges, and removes the need for an auxiliary transverse-control laser. Full-vectorial simulations confirm a stable trap for ultracold rubidium atoms with a depth of 170\,\textmu K at a distance of over 865\,nm from the surface, minimizing surface-induced effects and providing a compact platform for quantum sensing applications. # Introduction Compact quantum sensors require robust platforms for trapping and guiding ultracold atoms. These sensors aimed at chip-scale implementations of atomic-interferometer-based devices, including gravimeters [6, 8, 24, 10], gradiometers [17], gyroscopes [7, 9], and other quantum measurement systems [3, 4]. Atom chips [12, 18, 16, 1], which integrate optical and electronic components for atom manipulation, offer a viable pathway toward this miniaturization. A key challenge in miniaturizing these systems is to engineer compact and efficient light-atom interfaces. These must maximize atom-photon interactions while minimizing loss mechanisms, such as Casimir-Polder forces and surface-induced heating. One promising approach for miniaturization is to use evanescent optical fields from dielectric waveguides to trap atoms above surfaces using red-detuned (attractive) and blue-detuned (repulsive) lasers [2, 5, 20, 19]. However, this approach using conventional waveguides has several critical limitations. First, the evanescent field’s short penetration depth — typically tens of nanometers — concentrates atoms too close to the surface. This leads to increased surface-induced losses and reduced trap stability. Second, conventional waveguides exhibit Gaussian-like field profiles, leading to weak lateral confinement and atomic loss near waveguide edges — a serious limitation for miniaturized quantum sensors. The first limitation can be overcome with a photonic crystal (PC) waveguide operating near cutoff. This produces surface modes with extended penetration depths. As a result, atoms can be trapped at greater distances from the dielectric boundary, alleviating the constraint [13]. The photonic bandgap of the 1D PC isolates the surface mode from substrate leakage, enabling modes with effective indices near unity ($n_\mathrm{eff}\!\approx\!1$) and unprecedented evanescent field penetration into vacuum. Specifically, a PC-based rib waveguide was introduced, which sustains two surface modes (850 nm red-detuned and 640 nm blue-detuned) to achieve large evanescent decay lengths, enabling cold rubidium atoms to be trapped at distances exceeding 600 nm from the surface. To mitigate the second limitation (atom leakage at waveguide boundaries), a third auxiliary laser (633 nm, blue-detuned) for lateral confinement was proposed in [13]. While effective, this approach complicates experimental implementation, increases power dissipation, and introduces additional alignment challenges. Here, we introduce a new rib waveguide design with customized transverse field distributions inspired by recent advances in non-Gaussian photonics. Rodrigues et al. [22] demonstrated “scale-invariant” waveguides where light confinement in low-index materials exhibits uniform field profiles independent of waveguide geometry. While [22] originally applied to 1D dielectric slot waveguides, we adapt this principle to 2D surface modes in rib waveguides. By carefully shaping the rib waveguide’s cross-sectional profile, we achieve a specific geometry. It produces a near-uniform, flat-top field profile for the attractive (red-detuned) light at 850 nm. In contrast, the repulsive (blue-detuned) light at 640 nm exhibits maxima near the waveguide edges. This departs from the traditional Gaussian profile and enhances lateral confinement, preventing atom leakage at the waveguide edges without the need for an additional laser wavelength. The innovation of this work lies in the first adaptation of scale-invariant non-Gaussian field profiles — originally demonstrated in 1D dielectric slot waveguides [22] — to 2D rib waveguides on 1D photonic crystal substrates for surface modes. This approach surpasses our prior design [13] by eliminating the need for an auxiliary blue-detuned laser (633 nm) for lateral confinement, reducing the system from three to two wavelengths. This simplification minimizes experimental complexity, alignment challenges, and power dissipation. Additionally, the flat-top attractive and edge-enhanced repulsive profiles enhance lateral confinement without extra optics, enabling more robust traps for quantum sensing with fewer components. The resulting potential landscape provides more robust lateral confinement, directly mitigating the primary loss mechanism (edge escape) of all previous waveguide designs. # Methods ![Schematic of the generalized Effective Index Method (gEIM) applied to the rib-on-1D-PC waveguide structure. Bottom: cross-section showing lateral slices with widths wH, wS and with heights dH, dL, dS on the photonic crystal substrate. In the first step, for each lateral slice, the effective refractive index neff, i is computed by solving a 1D mode equation along the vertical x-axis under the assumption of infinite lateral width of each slice (wi = ∞). The corresponding mode wavevectors are ki = neff.i ω/c. In the second step, lateral guidance is modelled by solving the second 1D problem (along y-axis, top view) with the precomputed indices (neff.i), yielding the global effective index of the 2D waveguide mode (Neff).](media/OQE2026/Fig1.eps) *Schematic of the generalized Effective Index Method (gEIM) applied to the rib-on-1D-PC waveguide structure. Bottom: cross-section showing lateral slices with widths wH, wS and with heights dH, dL, dS on the photonic crystal substrate. In the first step, for each lateral slice, the effective refractive index neff, i is computed by solving a 1D mode equation along the vertical x-axis under the assumption of infinite lateral width of each slice (wi = ∞). The corresponding mode wavevectors are ki = neff.i ω/c. In the second step, lateral guidance is modelled by solving the second 1D problem (along y-axis, top view) with the precomputed indices (neff.i), yielding the global effective index of the 2D waveguide mode (Neff).* Figure 1 shows the waveguide design aimed at producing the intended non-Gaussian field profiles for robust atom trapping. The structure consists of a 1D PC substrate topped with a rib waveguide of variable widths ($\mathrm{w_H}$, $\mathrm{w_S}$) and heights ($\mathrm{d_H}$, $\mathrm{d_L}$, $\mathrm{d_S}$), where the subscripts denote regions of high, low, and intermediate effective indices. For the initial design phase, we employed the generalized Effective Index Method (gEIM) [14, 15]. This method extends the traditional EIM by accurately handling waveguides operating near cutoff, a regime essential for achieving large evanescent field penetration into the vacuum. Unlike conventional EIM, gEIM corrects the effective indices in cutoff regions, enabling precise modeling of 2D waveguides on 1D PC substrates. The gEIM (analogous to the EIM) determines the effective refractive indices $N_\mathrm{eff}$ of the 2D waveguide modes by solving two sequential 1D problems: vertical confinement, followed by lateral guidance. In the first stage, vertical confinement is solved and modal effective indices ($n_\mathrm{eff.i}$) calculated for each slice. In the second stage, lateral guidance is determined using the modal effective indices obtained earlier (see the top plane in Fig. 1, where the structure is analyzed from a top-down perspective). Now we adapted the concept introduced by Rodrigues et al. [22] into this gEIM algorithm. Only instead of using the bulk refractive indices ($n_\mathrm{L}$, $n_\mathrm{H}$, $n_\mathrm{S}$) of a dielectric slot waveguide, we operate with the modal effective indices ($n_\mathrm{eff.L}$, $n_\mathrm{eff.H}$, $n_\mathrm{eff.S}$) obtained from the first stage of gEIM. By implementing this concept in our waveguide design, we expect specific field profiles. For the attractive optical field at 850 nm, it is flat-top (or only slightly convex). For the repulsive optical field at 640 nm, it is concave (i.e., edge-enhanced with maxima near the edges). To achieve this, we must determine the top layer thicknesses $d_\mathrm{L}$, $d_\mathrm{H}$, and $d_\mathrm{S}$ corresponding to a critical point condition for the effective indices $n_\mathrm{eff.L}$, $n_\mathrm{eff.H}$, and $n_\mathrm{eff.S}$. The critical point is reached when the effective index of the middle slice ($n_\mathrm{eff.S}$) equals the effective index of the 2D mode ($N_\mathrm{eff0}$) of a waveguide without the middle slice, i.e., $N_\mathrm{eff0}=n_\mathrm{eff.S}$ , when $\mathrm{w_S}=0$. It also corresponds to the cutoff frequency of the asymmetric 2D waveguide formed by half of the symmetric structure in Fig. 1 when the size of the middle slice tends to infinity, $\mathrm{w_S}=\infty$.

Fig2.eps (1,97)(a) (1,45)(b)

Transverse field profiles for λ1 = 850 nm (red solid curves) and λ2 = 640 nm (blue dash-dotted curves), calculated using (a) the semi-analytical gEIM method, yielding Neff1 = 0.997 and Neff2 = 1.003, and (b) the full-vectorial FDE solver, yielding Neff1 = 0.998 and Neff2 = 1.004. The black dotted line marks the boundary of the waveguide’s final SiO2 layer.
We set $\mathrm{w_H} = 1$ μm and $\mathrm{w_S} = 3$ μm, with layer thicknesses $d_\mathrm{L} = 162$ nm, $d_\mathrm{H} = 207$ nm, and $d_\mathrm{S} = 195$ nm. This choice of $d_\mathrm{S}$ results in $n_\mathrm{eff.S}$ being close to the critical point for $\lambda_1 = 850$ nm (yielding the flat-top profile) and slightly below it for $\lambda_2 = 640$ nm (yielding the edge-enhanced profile). The sensitivity of the field and potential profiles to variations in the intermediate layer thickness $d_\mathrm{S}$ is examined in Appendix A. The photonic structure otherwise matches [13], designed to sustain surface modes at 850 nm (red-detuned, attractive) and 640 nm (blue-detuned, repulsive) with effective refractive indices close to unity. This configuration ensures extended penetration depths into the vacuum, facilitating atom trapping at safe distances from the surface. The PC structure follows: ``` math \text{substrate}/(LH)^{15} L'' H L' /\text{vacuum}\, , ``` where $L$ is SiO$_2$ (170.73 nm), $H$ is TiO$_2$ (86.46 nm), $L''$ is an extended SiO$_2$ layer (304 nm), and $L'$ represents the final SiO$_2$ layers with thicknesses $d_\mathrm{L}$, $d_\mathrm{H}$, and $d_\mathrm{S}$ specified above. The refractive indices at $\lambda_1 = 850$ nm and $\lambda_2 = 640$ nm are 1.4666 and 1.4679 for SiO$_2$, and 2.3137 and 2.3674 for TiO$_2$, respectively. # Results Since the preliminary design and determination of the appropriate $d_\mathrm{S}$ were performed using the gEIM method, we first present the field profiles for $\lambda_1 = 850$ nm and $\lambda_2 = 640$ nm obtained with this approximation. In Fig. 2(a) results are presented for calculated 2D waveguide modes with effective refractive indices $N_\mathrm{eff_1} = 0.997$ for $\lambda_1 = 850$ nm and $N_\mathrm{eff_2} = 1.003$ for $\lambda_2 = 640$ nm. The corresponding effective indices obtained through full-vectorial simulations with a Finite Difference Eigenmode (FDE) solver [25] in Lumerical Inc. MODE software, yield $N_\mathrm{eff_1} = 0.998$ and $N_\mathrm{eff_2} = 1.004$ for $\lambda_1 = 850$ nm and $\lambda_2 = 640$ nm, respectively. The field profiles for these wavelengths simulated via the FDE solver are displayed in Fig. 2(b). The 850 nm mode exhibits a nearly uniform profile across the waveguide due to operation near the critical point, whereas the 640 nm mode (operating below the critical point) shows two distinct maxima near $y = \pm 1.5$ μm, which provide the repulsive potential required for lateral atom confinement. The complete 2D field distributions for $\lambda_1 = 850$ nm and $\lambda_2 = 640$ nm, obtained using a full-vectorial FDE solver, are shown in Fig. 3(a) and Fig. 3(b), respectively.

Fig3.eps (1,45)(a)

Fig4.eps (1,45)(b)

(a) Electric field intensity distribution for red-detuned attracting light beam, λ1 = 850 nm, Neff1 = 0.998 (b) Electric field intensity distribution for blue-detuned repulsive light beam, λ2 = 640 nm, Neff2 = 1.004.
Using the field distributions obtained, we calculated the total potential $U_\mathrm{total}(x,y)$ acting on ultracold rubidium atoms. This potential combines the attractive optical potential from 850 nm radiation, the repulsive optical potential from 640 nm radiation, the Casimir-Polder potential, and the gravitational potential: ``` math \begin{equation} U_\mathrm{total}(x,y)= \sum_{\omega=\omega_1}^{\omega_2} U_\mathrm{dip}(x,y,\omega) - \frac{C_3 \lambda_\mathrm{eff}}{x^3 (x + \lambda_\mathrm{eff})} - mgx \; , \end{equation} ``` where $C_3 = 5.7\times 10^{-49}$ J m$^3$ denotes the van der Waals coefficient, $\lambda_\mathrm{eff} = 710/(2\pi)$ nm represents the reduced wavelength [23, 20] characterizing the atom-surface attractive interaction, and $g=9.8$ m/s$^2$ is the gravitational acceleration. For the optical dipole potential $U_\mathrm{dip}(x,y,\omega)$, applicable under large detunings (significantly exceeding the D-line doublet splitting), we employ the approximation [11]: ``` math \begin{equation} U_\mathrm{dip}(x,y,\omega)=\frac{3\pi c^2}{2 \omega_D^3} \frac{\gamma_D}{(\omega-\omega_D)}I(x,y) \; , \end{equation} ``` where $\omega_D = 2\pi \times 380.77$ THz corresponds to the D-line center frequency of the $^{87}$Rb atom ($\lambda_D\simeq787$ nm), and $\gamma_D = 37$ MHz denotes the average natural linewidth of the $D_1$ and $D_2$ transitions. Both the red and blue detunings (listed in Table 1) significantly exceed the D-line doublet splitting ($\simeq7$ THz), justifying the use of the large-detuning approximation. The resulting trap potential is shown in Fig. 4(a) (in units of temperature, $U/k_B$), while Fig.  4(b) presents a vertical cross-section at $y=0$, revealing a trap minimum approximately 865 nm above the surface with a depth of 170 μK. The potential exhibits strong lateral confinement, effectively preventing atom leakage at the waveguide edges.

Fig5a.eps (1,45)(a)

Fig5b.eps (1,45)(b)

(a) Simulated optical trapping potential for cold rubidium atoms above the waveguide generated by combined 850 nm (red-detuned, attractive) and 640 nm (blue-detuned, repulsive) fields. (b) Vertical cross-section along the y = 0 plane, quantifying the trap depth and surface-to-minimum distance.
Table 1 summarizes the calculated mode parameters for the rib waveguide, including the effective indices, propagation losses, and the optical power required to achieve the 170 μK trapping potential shown in Fig. 4. The required optical intensities ($1.4~\mathrm{GW/m^2}$ at $850~\mathrm{nm}$ and $8.3~\mathrm{GW/m^2}$ at $640~\mathrm{nm}$) may induce heating in the SiO$_2$/TiO$_2$ multilayer structure, potentially affecting long-term stability. However, as discussed in Section 3.3 of [13], several mitigation approaches can be adopted to suppress thermal accumulation, including the use of high-thermal-conductivity substrates (e.g., sapphire, $\kappa\approx\mathrm{30\,W/(m\cdot K)}$), insertion of thin conductive interlayers such as ITO ($\kappa \approx \mathrm{10\,W/(m\cdot K)}$), and the integration of narrow Au heat-spreading stripes along the waveguide flanks. Furthermore, in the considered 1D photonic-crystal geometry, a substantial fraction of the optical energy propagates in vacuum, which helps to reduce the temperature load on the core region. The constituent oxide materials are thermally stable (typical glass transition temperature $T_g>500^{\circ}\mathrm{C}$), although a dedicated experimental evaluation of steady-state heating and long-term coating durability will be necessary in future work. \|C\|C\|CCC\| **wavelength** &detuning from D-line &$\mathrm{n_{eff}}$ &loss, dB/cm & intensity in Fig.4, GW/m$^2$\ **850 nm** &$-28$ THz &0.998 & 1.45 & 1.4\ **640 nm** &$+88$ THz &1.004 & 5.77 & 8.3\  for $\mathrm{Im}(n_j)=10^{-5}, \forall j$ # Discussion Our results show that the engineered rib waveguide, preliminarily optimized with the gEIM, supports the desired non-Gaussian field profiles for both the attractive (850 nm) and repulsive (640 nm) wavelengths. Full-vectorial FDE simulations confirm robust atom trapping with enhanced lateral confinement, enabled by the flat-top attractive mode and edge-enhanced repulsive mode. This removes the need for the third auxiliary laser used in [13], significantly simplifying the experimental implementation. The combined action of the two PC-supported near-cutoff modes yields a trap depth of 170 μK at distance about 865 nm above the surface (Fig. 4). The gEIM proved essential for modeling waveguide behavior near cutoff, where the conventional EIM becomes inaccurate. By adapting the scale-invariant principle to effective indices, we optimized the rib geometry to support desired non-Gaussian field profiles. This strategy can be extended to other optical systems, including purely photonic traps for laser cooling [21]. # Conclusion We designed a photonic-crystal rib waveguide that generates tailored non-Gaussian optical fields, providing lateral stability for ultracold-atom trapping. This approach offers a compact and efficient platform for atom-chip-integrated quantum sensors, combining robust confinement with reduced experimental complexity. This work was supported by the Institute of Spectroscopy RAS under state task No. FFUU-2025-0004 # Declarations The author declares no conflicts of interest. # Robustness of Field and Potential Profiles to Intermediate Layer Thickness Variations As explained in Section 2, the intermediate layer thickness, $d_\mathrm{S}$, should be chosen such that the modal effective index of this slice, $n_\mathrm{S}$, is close to the critical point for a wavelength of 850 nm (to obtain a flat-top profile) and slightly below the critical point for a wavelength of 640 nm (to obtain an edge-enhanced profile). For practical implementation of this approach, it is useful to estimate how sensitive the system under consideration near the critical point is to $d_\mathrm{S}$ variations. The modern thin-film coating facilities provide optical film thickness deposition accuracy much better than 1%. We consider thickness variations of up to $\pm2.5\%$ near $d_\mathrm{S} = 195$ nm, which is equivalent to $d_\mathrm{S}\pm5$ nm. %subsection$-5$ nm thickness variation from $d_\mathrm{S} = 195$ nm ## Performance with $d_\mathrm{S} = 190$ nm ($-5$ nm variation) Figure A1 demonstrates transverse field profiles (like in Fig. 2) and trapping potential for cold rubidium atoms (like in Fig. 4) for the intermediate layer thickness, $d_\mathrm{S}=190$ nm.

FigA1.eps (1,92)(a) (1,40)(b)

FigA1c.eps (1,41)(c)

FigA1d.eps (1,40)(d)

Transverse field profiles and trapping potential with dS = 190 nm (−5 nm variation). (a) The semi-analytical gEIM method, yielding Neff1 = 0.995 and Neff2 = 1.001, and (b) the full-vectorial FDE solver, yielding Neff1 = 0.9968 and Neff2 = 1.002. Black dotted line indicates waveguide boundary and arrows designate −5 nm thickness reduction. (c) Optical trapping potential for rubidium atoms. (d) Vertical cross-section at y = 0 showing trap depth of 158 μK (for corrected intensity).
While maintaining the intensities specified in Table 1, the blue-detuned repulsive laser becomes insufficient to fully repel rubidium atoms from the surface. However, increasing the intensity from 8.3 to 11.6 GW/m$^2$ (Table A1) restores the optical trapping potential to nearly original performance levels, despite the moderately distorted field profiles. \|C\|CCC\| **wavelength** &$\mathrm{n_{eff}}$ &loss, dB/cm & intensity in Fig.4, GW/m$^2$\ **850 nm** &0.9968 & 1.2 & 1.4\ **640 nm** &1.002 & 5.27 & 11.6\  for $\mathrm{Im}(n_j)=10^{-5}, \forall j$ ## Performance with $d_\mathrm{S} = 200$ nm ($+5$ nm variation) Figure A2 demonstrates that even with the more Gaussian-like field profiles resulting from increased $d_\mathrm{S}$, viable trapping potentials can still be formed. The lateral support becomes less ideal, requiring a modest increase in blue-light intensity from 8.3 to 8.7 GW/m$^2$ (Table A2) to prevent edge-escape and maintain full atom confinement.

FigA2.eps (1,92)(a) (1,40)(b)

FigA2c.eps (1,41)(c)

FigA2d.eps (1,40)(d)

Transverse field profiles and trapping potential with dS = 200 nm (+5 nm variation). (a) The semi-analytical gEIM method, yielding Neff1 = 0.99995 and Neff2 = 1.0055, and (b) the full-vectorial FDE solver, yielding Neff1 = 0.9997 and Neff2 = 1.006. Black dotted line indicates waveguide boundary and arrows designate +5 nm thickness increase. (c) Optical trapping potential for rubidium atoms. (d) Vertical cross-section at y = 0 showing trap depth of 154 μK (for corrected intensity).
These results demonstrate that minor thickness variations during deposition/etching can be effectively compensated through laser intensity adjustments, highlighting the practical robustness of our design approach. \|C\|CCC\| **wavelength** &$\mathrm{n_{eff}}$ &loss, dB/cm & intensity in Fig.4, GW/m$^2$\ **850 nm** &0.9997 & 1.77 & 1.4\ **640 nm** &1.006 & 6.38 & 8.7\  for $\mathrm{Im}(n_j)=10^{-5}, \forall j$ ## References 1. barticle Clad'e, P., Guellati-Kh'elifa, S., Schwob, C., Nez, F., Julien, L., Biraben, F.: A promising method for the measurement of the local acceleration of gravity using Bloch oscillations of ultracold atoms in a vertical standing wave. Europhysics Letters 71(5), 730 (2005) barticle endbibitem %%% 2. 2. barticle Ferrari, G., Poli, N., Sorrentino, F., Tino, G.: Long-lived Bloch oscillations with bosonic Sr atoms and application to gravity measurement at the micrometer scale. Physical Review Letters 97(6), 060402 (2006) barticle endbibitem %%% 3. 3. barticle Tino, G.M.: Testing gravity with cold atom interferometry: results and prospects. Quantum Science and Technology 6(2), 024014 (2021) barticle endbibitem %%% 4. 4. barticle Ge, G., Chen, X., Li, J., Zhang, D., He, M., Wang, W., Zhou, Y., Zhong, J., Tang, B., Fang, J., betal: Accuracy improvement of a compact $^85$Rb atom gravimeter by suppressing laser crosstalk and light shift. Sensors 23(13), 6115 (2023) barticle endbibitem %%% 5. 5. botherref Mao, D.-K., Deng, X.-B., Luo, H.-Q., Xu, Y.-Y., Zhou, M.-K., Duan, X.-C., Hu, Z.-K.: A dual-magneto-optical-trap atom gravity gradiometer for determining the Newtonian gravitational constant. Review of Scientific Instruments **92(5) (2021) botherref endbibitem %%% 6. 6. barticle Dutta, I., Savoie, D., Fang, B., Venon, B., Alzar, C.G., Geiger, R., Landragin, A.: Continuous cold-atom inertial sensor with 1 nrad/sec rotation stability. Physical Review Letters 116(18), 183003 (2016) barticle endbibitem %%% 7. 7. botherref Garrido Alzar, C.L.: Compact chip-scale guided cold atom gyrometers for inertial navigation: Enabling technologies and design study. AVS Quantum Science **1(1) (2019) botherref endbibitem %%% 8. 8. barticle Bongs, K., Holynski, M., Vovrosh, J., Bouyer, P., Condon, G., Rasel, E., Schubert, C., Schleich, W.P., Roura, A.: Taking atom interferometric quantum sensors from the laboratory to real-world applications. Nature Reviews Physics 1(12), 731--739 (2019) barticle endbibitem %%% 9. 9. barticle Bongs, K., Bennett, S., Lohmann, A.: Quantum sensors will start a revolution --- if we deploy them right. Nature 617(7962), 672--675 (2023) barticle endbibitem %%% 10. 10. barticle Keil, M., Amit, O., Zhou, S., Groswasser, D., Japha, Y., Folman, R.: Fifteen years of cold matter on the atom chip: promise, realizations, and prospects. Journal of Modern Optics 63(18), 1840--1885 (2016) barticle endbibitem %%% 11. 11. botherref McGilligan, J.P., Moore, K., Dellis, A., Martinez, G., Clercq, E., Griffin, P., Arnold, A., Riis, E., Boudot, R., Kitching, J.: Laser cooling in a chip-scale platform. Applied Physics Letters **117(5) (2020) botherref endbibitem %%% 12. 12. barticle Li, D., He, W., Shi, S., Wu, B., Xiao, Y., Lin, Q., Li, L.: Review of atom chips for absolute gravity sensors. Sensors 23(11), 5089 (2023) barticle endbibitem %%% 13. 13. barticle Afanasiev, A., Skakunenko, P., Bykova, D., Kalmykov, A., Balykin, V.: Atom chip. Uspekhi Fizicheskikh Nauk 194(11), 1146--1158 (2024) barticle endbibitem %%% 14. 14. barticle Barnett, A., Smith, S., Olshanii, M., Johnson, K., Adams, A., Prentiss, M.: Substrate-based atom waveguide using guided two-color evanescent light fields. Physical Review A 61(2), 023608 (2000) barticle endbibitem %%% 15. 15. barticle Burke Jr, J.P., Chu, S.-T., Bryant, G.W., Williams, C.J., Julienne, P.S.: Designing neutral-atom nanotraps with integrated optical waveguides. Physical Review A 65(4), 043411 (2002) barticle endbibitem %%% 16. 16. barticle Ovchinnikov, Y.B., Ayi-Yovo, F.E.: Towards all-optical atom chips based on optical waveguides. New Journal of Physics 22(5), 053003 (2020) barticle endbibitem %%% 17. 17. barticle Ovchinnikov, Y.B.: A perspective on integrated atomo-photonic waveguide circuits. Applied Physics Letters 120(1), 010502 (2022) barticle endbibitem %%% 18. 18. barticle Konopsky, V.: Photonic crystal surface modes for trapping and waveguiding of ultracold atoms. Sensors 23(21), 8812 (2023) barticle endbibitem %%% 19. 19. barticle Rodrigues, J.R., Dave, U.D., Mohanty, A., Ji, X., Datta, I., Chaitanya, S., Shim, E., Gutierrez-Jauregui, R., Almeida, V.R., Asenjo-Garcia, A., betal: All-dielectric scale invariant waveguide. Nature Communications 14(1), 6675 (2023) barticle endbibitem %%% 20. 20. barticle Konopsky, V.: Cutoff-inclusive generalized effective index method for waveguide design. Applied Optics 64(24), 7230--7234 (2025) barticle endbibitem %%% 21. 21. botherref Konopsky, V.N. https://www.pcbiosensors.com/1DPC4all.htm, Version 2.0.9415.36854 (2025) botherref endbibitem %%% 22. 22. barticle Zhu, Z., Brown, T.G.: Full-vectorial finite-difference analysis of microstructured optical fibers. Optics Express 10(17), 853--864 (2002) barticle endbibitem %%% 23. 23. barticle Thompson, J.D., Tiecke, T., Leon, N.P., Feist, J., Akimov, A., Gullans, M., Zibrov, A.S., Vuleti'c, V., Lukin, M.D.: Coupling a single trapped atom to a nanoscale optical cavity. Science 340(6137), 1202--1205 (2013) barticle endbibitem %%% 24. 24. bchapter Grimm, R., Weidem"uller, M., Ovchinnikov, Y.B.: Optical dipole traps for neutral atoms. In: Bederson, B., Walther, H. (eds.) Advances in Atomic, Molecular, and Optical Physics vol. 42, pp. 95--170. Elsevier, Amsterdam (2000) bchapter endbibitem %%% 25. 25. barticle Prudnikov, O., Ilenkov, R.Y., Taichenachev, A., Yudin, V., Bagaev, S.: Deep macroscopic pure-optical potential for laser cooling and trapping of neutral atoms. Physical Review A 108(4), 043107 (2023) barticle endbibitem thebibliography. --- Generated: 2026-08-30 from Z:\ValeryData\Valery_New\my_articles\OQE2026\OQE\manuscript\Konopsky4OQE2025.tex, validated against Z:\ValeryData\Valery_New\my_articles\OQE2026\OQE\published\Konopsky4OQE2026_preprint.pdf.