Doublydressed states for nearfield trapping and subwavelength lattice structuration
Abstract
We propose a scheme to tailor nanostructured trapping potentials for ultracold atoms. Our trapping scheme combines an engineered extension of repulsive optical dipole forces at short distances and attractive CasimirPolder forces at long distances between an atom and a nanostructured surface. This extended dipole force takes advantage of excitedstate dressing by plasmonicallyenhanced intensity to doubly dress the ground state and create a strongly repulsive potential with spatially tunable characteristics. In this work, we show that, under realistic experimental conditions, this method can be used to trap Rubidium atoms close to surfaces (tens of nanometers) or to realize nanostructured lattices with subwavelength periods. The influence of the various losses and heating rate mechanism in such traps is characterized. As an example we present a nearfield optical lattice with 100nm period and study the tunability of lattice and trapping depths. Such lattices can enhance energy scales with interesting perspectives for the simulation of stronglycorrelated physics. Our method can be extended to other atomic species and to other nearfield hybrid systems where a strong atomlight interaction can be expected.

January 2018
1 Introduction
Various hybrid quantum systems have been developed in the recent years with the idea that the overall combination of systems could exploit the properties of individual systems at best and overcome their individual performances [1]. In this spirit, an intensive research has been carried on various combinations such as for example NV centers or cold atom coupling to superconducting circuits processors for quantum memory applications [2, 3, 4, 5], magnetic sensing [6, 7] or hybrid combinations involving macroscopic quantum systems such as mechanical oscillators [8, 9, 10, 11]. In this family of coupled quantum systems, ultracold atoms are special because of their very wellcontrolled intrinsic properties due to their strong decoupling from the environment and because they stand in gaseous form while other systems are predominantly solidstate devices. Transporting atoms close to a quantum counterpart is crucial for applications in quantum information and quantum simulation [12]. By increasing the coupling strength with, for instance, microwave superconducting photons [13, 14], optical photons [15, 16, 17, 18], slow light [52, 20] or plasmons [21], we can devise efficient quantum memory or engineer surfacemetiated atomatom interactions [22, 23, 24]. Tailoring nanostructured optical lattice quantum simulators with tunable properties (energy scale, geometry, surfacemediated longrange interaction) will allow one to investigate new regimes in manybody physics. In this quest for exotic quantum phases (e.g., quantum antiferromagnetism [25]), the reduction of thermal entropy is a crucial challenge [26, 27, 28, 29, 30]. The price to pay for such low temperature and entropy is a long thermalization time that will ultimately limit the experimental realization. Miniaturization of lattice spacing is a promising solution to speed up this dynamics. There is therefore a wide effort in the community to push the optical trapping technology to the nanometer scale [21, 22, 24, 31, 32, 33] or to generate subwavelength lattice spacing using multiphoton optical transitions [34, 35], timeperiodic modulation [36] or superconducting vortices [37]. Engineering cold atom hybrids offers promising perspectives but requires to interface quantum systems in different states of matter at very short distances, which still remains an experimental challenge.
In this paper, we present a novel trapping method that enables to trap and manipulate cold atoms below 100 nm from surfaces where the electromagnetic environment of cold atoms could be engineered. We show that our method can also be used to create subwavelength potentials with controllable parameters. Our method, that we refer as DoublyDressed State (DDS), is inspired by [24, 33] where the authors take advantage of vacuum forces and material engineering to create nearfield traps. Here, two offresonant laser light fields interacting with the structured surface overcome the CasimirPolder (CP) attraction at short distances [38]. The combination of the DDS repulsion and CP attraction forms a trap that follows the shape of the surface. A modulation of the CP potential can be generated with a periodicallystructured surface, thus resulting in a lattice potential for cold atoms. In particular we study here the generation of a 1D lattice by periodicallyspaced dielectric ridges in the spirit of [24]. Another scheme for the realization of surface traps with subwavelength optical lattices is published backtoback [39].
In section 2 we present the general arguments of the trapping method which combines optical and vacuum forces. In section 3 we apply the method to a stratified surface and characterize the trapping properties. We then extend it to a onedimensional nanostructured surface where longitudinal and transverse trapping in a subwavelength lattice is demonstrated. In section 4 we discuss the advantages and limitations of this method, to finally conclude in section 5 with perspectives.
2 Trapping method
In the absence of external field, an atom in vacuum close to a surface interacts with it because of quantum charge and field fluctuations, also known as the CasimirPolder (CP) interaction [38]. It can be calculated by assimilating the atom to a fluctuating induced dipole with a moment , which creates an electromagnetic field that is reflected by the surface before interacting back with the dipole. This CP interaction is therefore characterized by an energy proportionnal to the atomic polarizability and to the scattering Green tensor . The latter describes the roundtrip propagation (between the atom and the surface) of an electric field at frequency intially generated at the position of the atom . It contains all the optical properties of the surface. The CP force derived from the energy between two electric or paramagnetic objects is always attractive for an atom in the ground state.
We consider here a semiinfinite dielectric surface with a dielectricvacuum interface at . Fig.1(a) sketches three internal states of a atom modulated by the vacuum forces only. In that case, the potentials are attractive and drag the atoms to the surface. We consider now a laser field with an intensity profile decaying from the surface as shown in Fig.1(b), which dresses the state. This laser is tuned a few hundreds of GHz on the blue of the to resonance. The resulting AC Stark shift effect induces a strong repulsive potential for the state while barely affecting the state. The  transition frequency therefore depends on the distance from the surface. Double dressing is achieved by using a second laser at frequency tuned close to the free space transition . The resulting AC Stark shift on the state, is governed by the spatiallydependent detuning (see Fig.1(c)), which cancels at position . This position directly depends on the laser frequency . For positions , the detuning is positive (blue) and the doublydressed potential expels the atoms away from the surface. As depicted in Fig.1(c), a trap is then formed by the contribution of the attractive CP and repulsive doublydressed state potentials.
The trapping potential is calculated by integrating the time derivative of the atomic momentum over the trajectory of the atom [40] :
(1) 
where (resp. ) is the population of the (resp. ) state obtained by solving the optical Bloch equations. This expression for the total potential states that both excited and ground state energy gradient contributions (state population energy gradient) are opposite at the trap position : .
3 Results
3.1 Case of a planar and stratified surface
As shown later, a key ingredient of the DDS trapping method is the amplitude of the excitedstate energy shift. As explained in section 2, the decaying profile of the 1529nmlaser intensity directly imprints the energy shift of the 5P state. The spatial variation of intensity (gradient) must be as strong as possible to create a sufficient barrier to trap the atoms. To create an enhanced exponentiallydecaying field in the vacuum, we take advantage of Surface Plasmon Resonance (SPR) that occurs at a metaldielectric interface of a stratified surface when the angle of incidence is where is the refractive index in the incident medium and is the effective index of the Surface Plasmon Polariton (SPP) mode.
The surface geometry studied in this section is given in Fig.2(a). The geometrical parameters of the stratified surface are chosen by maximizing the 1529nmintensity decay from 50 nm to 100 nm as shown in Fig.2(b). Given the optimization, we take a stratified structure composed of a dielectric layer () with a thickness of 158 nm and a 41nm layer of gold deposited on a substrate. The lasers intensities are determined using transfer matrix method [41] implemented with Reticolo software [42].
The CP potentials are calculated using the multipolar coupling approach [43]. The thermal effects are neglected considering the distances at which the atoms are trapped (, being the atom position and the thermal wavelength). For an atom in a state ( or ), the interaction with the stratified surface is described by a potential that can be split into two terms :
(2) 
where
(3) 
is the nonresonant term. The second term, named resonant term, is nonzero if the atom is excited () and is given by
(4) 
By considering a atom with an isotropic polarizability , these two expressions can be written as
(5) 
and
(6)  
For numerical reasons, the integrations are performed over imaginary frequencies () for the nonresonant term (5) and over imaginary wavevectors () for the resonant term (6). We define longitudinal and transverse parts for the wavevector as follows: where and . (resp. ) is the amplitude reflection coefficient of the stratified surface for (resp. ) polarization defined by the electric (resp. magnetic) field being transverse to the plane of incidence. and are respectively the electric dipole moment and the frequency of the transition ( and nm for ). In this planar and stratified example, the coefficients and are calculated with Fresnel equations. The dynamic polarizability for an isotropic atom in a state at imaginary frequencies is given by [44]
(7) 
where the sum is carried over all possible transitions with a frequency and a reduced dipole moment . denotes the total electronic angular momentum of the state. The calculated CP potentials are plotted in Fig.2(c) and present an oscillation for the excitedstate modulation due to the possible spontaneouslyemitted real photon. Both potentials are attractive for atoms at position nm from the surface.
We now illuminate the rear side of the surface at the plasmonic angle with laser light at a wavelength nm, bluedetuned by 26 pm from the transition. As shown in Fig.2(d), using an incident power of 400 mW and a waist of 200 m, we obtain a maximum SPP intensity at the dielectricvacuum interface () of (100 SPP enhancement factor) corresponding to a maximal AC Stark shift of the state of 31 GHz. This AC Stark shift is calculated using strongfield approximation (fine structure basis), justified by the SPP enhancement. The laser at nm, i.e. GHz, with a power of 200 mW in a 200 waist is retroreflected on the surface (see inset in Fig.2(e)). The detuning is zero at distance nm (resonant position). In this simulation we use a spatially homogeneous Rabi frequency MHz and we obtain the total potential represented in Fig.2(e), with a trapping depth MHz and a trapping position nm. The geometrical properties of the trap such as the position and the depth can be adjusted with the powers of both lasers (1529 and 780 nm) and the frequency of the 780nm laser (Fig.3).
As in [33], we evaluate the performance of our configuration by considering the four following characteristic times : the exit time , the tunneling time , the antidamping time and the trap oscillation time . We neglect the effect of transient heating [33]. The exit time is due to scattering heating. It is defined as where is the energy change for a duration due to the absorption by the atom with a mass of a photon with an effective momentum , at a scattering rate . is the groundstatebinding energy determined along with the wavefunction by using imaginary time propagation method. The tunneling time is due to the possibility for the atom to escape the trap by tunneling towards the surface. This characteristic time is calculated within the WKB approximation [45]. The antidamping time corresponds to the time for the energy increase due to bluetransition heating to equal the energy . In our case it is defined by where is the antidamping rate with and the recoil energy [33]. We note and the position and momentum standard deviations of the trapping ground state respectively. The trap oscillation time represents a characteristic adiabatic time.
In Fig.4, we plot the trapping depth and barrier height as well as the characteristic evolution timescales for two trapping positions (corresponding to the dashed and full lines in Fig.3(a)) as a function of the 1529nmlaser power. The trapping depth, which is mostly controlled by the groundstate () CP potential, strongly increases while reducing the traptosurface distance (Fig.4(b)). As shown in Fig.4(c),(e),(f), all contributions to the trap lifetime increase together with the 1529nmlaser radiation. The 1529nm power () is therefore a key parameter that tunes the energy gradient of the state and optimizes the trap characteristics. Experimentally, the optimal power (maximal) will be given by the heat capacity and absorption of the surface at 1529nm. In the range of experimental parameters studied here and for atomtosurface distances around nm, the exit and antidamping times are the limiting trapping times, which increase together with . At higher intensities and for both trapping positions presented here, antidamping heating becomes the limiting lifetime effect.
Within the WKB approximation, the tunneling time is exponentially dependent on the barrier height and width. For similar barrier heights (Fig.4(a)), the reduction of barrier width at shorter distances has a strong impact on the tunneling time (20ordersofmagnitude decrease) as we can see in Fig.4(e). At short atomtosurface distances, tunneling through the barrier will therefore become the limiting lifetime effect.
3.2 Case of a 1D grating
In this section, we extend the previously described method to a nanostructured surface which modulates the CP potentials in the transverse direction (i.e. () plane), therefore tailoring the trapping potential. Hereafter, we restrict our example to a 1D periodic nanostructuration, sketched in Fig.5(a), to create a lattice potential with spacing fixed at nm. Using the same optimization procedure as in the case of the planar stratified structure, we optimize the geometry by maximizing the intensity decay of the 1529nm laser in front of a ridge (Fig.5(b)). This led to a dielectric thickness of 500 nm, a ridge width of 25 nm and a metal thickness of 10 nm (optimization not shown).
The CP potentials for a structured surface are calculated using equations (3) and (4) in which the scattering Green tensor for a onedimensional grating is given by
(8) 
where is the atom position. is the unit vector describing the field that propagates along in the order with a polarization , and scatters out of the surface in the order with polarization , being the reflection coefficient. The reflected field is then described by and propagates along . The integration over is performed in the first Brillouin zone where is defined as the period of the grating. The amplitude reflection coefficients at imaginary frequencies for the ground state CP potential are calculated by implementing Fourier modal method [46, 47, 48] and the ones at real frequencies for the excited state CP potential are calculated using Reticolo software.
The optical scheme uses here two laser beams at nm : one being shined from the back at frequency that excites SPPs (noted ’’) and one shined from the front at frequency (noted ’’). As shown in Fig.5(c) top, the Back laser does not create a decaying intensity profile in front of a groove. This is detrimental to trap atoms at this position. The Front laser produces the inverted profiles (Fig.5(c) middle). By having a slight frequency mismatch the interferences can be timeaveraged and the total contribution to the potential is given by the sum of both intensities. As shown in Fig.5(c) bottom, the total intensity has a rapidlydecaying profile at any transverse position. We define as the ratio of power between the Front and Back lasers.
The profile of the total trapping potential is plotted in Fig.6(a) with longitudinal and transverse crosssections in Fig.6(b) and (c) respectively. Here, the parameters are mW, , GHz and MHz. In this configuration, the atoms are trapped in front of a ridge at nm and the modulation of the groundstate CP potential creates lattice and trap depths of MHz ( with the lattice recoil energy) and a period of nm. The trapping frequencies are MHz. In Fig.6(d)(f), the parameter is scanned to find the optimal intensity profiles that create a trap both above a ridge and a groove (Fig.6(d),(e)).^{2}^{2}2We note that the existence of a lattice potential requires the formation of a trap in front of a ridge and the presence of a barrier in front of a groove. The final lattice depth, which corresponds to the energy difference between the groove and ridge energy minima, is plotted in Fig.6(f). As shown in Fig.6(d)(f), for the lattice depth is little sensitive to an increase of the 1529nmforwardlaser power (). To minimize the heating effect due to laser power, we choose , which fixes the spatial profile of the light intensity while keeping its amplitude as a free parameter via the power of the 1529nmbackwardlaser beam ().
As demonstrated before, the parameter is key to increase the atom lifetime in the DDS trap (Fig.4). In Fig.7(a) we show the dependence of the lattice depth as a function of . The 780nmlaser beam can also be used to change the trap parameters by controlling the coupling between the ground () and excited () states. As shown in Fig.7(b), by tuning both the frequency and the power of the 780nm laser, the onedimensionnal lattice depth can be tuned from 0 to MHz (610 ). Although the contribution of these parameters to the lattice depth seems equivalent (symmetry), their influence on the traptosurface distance is very different (Fig.7(c)). Indeed, a scan of at constant has little effect on (dotted line) while a scan of at constant shifts the trap from 17 to 60 nm (dashed line).
4 Discussion
We have described and analyzed an experimental method to trap atoms in the near field of planar and nanostructured surfaces. The properties of the trap formed with this method can be tuned in lattice and trapping depths over relevant scales to simulate Hubbard Hamiltonians (0  600 ). The lattice parameters are experimentally adjusted by tuning the amplitude and frequency of laser beams, which are variables very wellcontrolled in laboratories. The surface geometry studied here led to a onedimensional lattice. The idea can be straightforwardly extended to generate 2D subwavelength lattices at the cost of an increased numerical complexity for the computations of the optical response of the 2D structured surface. The algorithm for the 2D calculations based on Rigorous CoupledWave Analysis (RCWA) has already been developed [49, 50] and is available on software such as Reticolo [42].
In [33], a similar trapping method has been proposed in which the repulsive shape of the 6P state of Cesium was induced by a Drude material with plasmon resonance frequency corresponding to the atomic transition and a Qfactor of . This condition is extremely stringent on material engineering and a practical implementation seems hardly possible. Our method is an interesting compromise where the excitedstate properties are tuned by simple optical means which are little influenced by the material properties. Regardless of experimental implementation issues, we point out that [33] benefits from a divergence in potential close to the surface, which allows to access a diverging gradient of the excited state close to the surface. On the contrary, our method has an exponentiallydecaying intensity profile and its slope is therefore upperbounded by its value at the dielectricvacuum interface. To trap atoms at distance below 15 nm, our method will require unrealistic optical power or close to transition frequency that will make it experimentally inappropriate.
In our method, the constraint for the choice of materials are guided by: a high refractive index in order to generate strong surface plasmon waves, very low absorption at 1529 nm so as to handle high optical powers and the existence of nanofabrication protocole to generate the desired nanostructures. In this work we have shown the predicted potentials for and materials, which have negligible absorption at 1529 nm^{3}^{3}3We could not find any measured absorption at this wavelength in the literature where it is estimated to have zero absorption. as well as a reasonably high refractive index (). In addition, the geometry of nanostructures that we have detailed in this work can be realized by Metal Assisted Etching (MAE). This technique allows for creating nanopillar array with dimensions down to a few tenth of nanometers in diameters and up to a micron in height [51]. MAE can be extended to other types of material. Further improvement can be carried on the material choice to optimize the trapping geometry.
We have seen in section 3.1 that the atomic lifetime in the trap was of the order of 100 ms for the chosen range of parameters. This time was influenced by the exponentiallydecaying optical profile from the dielectricvacuum interface. In the 1D grating case, this exponentiallydecaying profile was found to be absent in front of a groove and was compensated by the forward laser. The latter reduces the slope in front of a ridge that in turn limits the antidamping time to 1 ms at best for our parameter range. Improvement in the lifetime in the lattice trap requires further investigation. In particular, the optimization of the geometry was carried by maximizing the intensity gradient in front of a ridge, which is crucial to increase the lattice depth. In future work, it would be interesting to consider as well the gradient of intensity in front of a groove and to look at different angles of incidence for the laser beams to improve the overall slope and consequently the lifetime of atoms in this subwavelength lattice. Another possibility to improve the lifetime would be to add some laser tunability to the longrange CP attraction force. For example, this can be achieved by a reddetuned evanescent field [39] that would help to increase the lattice depth at large distances to maintain long lifetimes.
5 Conclusion and perspectives
In this work, we have presented the doublydressed state trapping method and applied it to trap atoms in the near field of various surfaces. We emphasized specifically on planar stratified structure and on 1D periodicallystructured surface to generate periodic subwavelength potentials. For both cases, we have shown that the trap characteristics (traptosurface distance, depth and trapping frequencies) could be adjusted on a large range by simply changing the frequency and/or power of the two dressing lasers. In the case of the stratified medium, we simulated the different contributions to the trap loss and showed that trap lifetime of 100 ms could be expected. For the case of 1D nanostructures, we calculated the modulation of the CasimirPolder potential and took advantage of this modulation to generate a subwavelength lattice. We showed that the lattice and trap depths could be adjusted by controlling the dressinglaser parameters. We anticipate that our demonstration of a 100 nm lattice spacing can be extended to smaller sizes as low as a few tenth of nanometers below which the optical power to form a trap will become irrelevant for applications. Our study was developed and adapted to Rb but it can be straightforwardly applied to most alcaline atoms for which the excitedstate transition lies in the range 11.6 m where the absorption of is low. Other materials such as could also be exploited as it has a high refractive index and a very low absorption above 550 nm. In addition, materials such as could be interestingly used to include other optical functionalities on the surface such as optical waveguides for readout and addressing or surface guided modes for opticallymediated longrange interactions [23, 24]. The reduction of lattice spacing is one key possibility to increase energies and to speed up dynamics in future coldatom lattice experiments.
In the present work, the numerical tools that have been used are well adapted to periodic structures but the use of CasimirPolder potential structuration goes beyond this periodic frame. By nanoshaping the surface, we can access a large variety of lattice geometries. The possibilities include the design of pillar defects to simulate the role of impurities in solidstate physics, the introduction of a controlled disorder or confined optical modes [52]. The DoublyDressed State method can be directly applied to tapered nanofibers, hollowcore fibers and slow light waveguides for which nearfield trapping will strongly increase the atomphoton coupling strength.
Acknowledgements
We thank Romain Pacanowski for numerical assistance. J.Z. acknowledges support from China Scholarship Council (CSC) and La Fondation FrancoChinoise pour la Science et ses Applications (FFCSA). P.B. thanks the région Aquitaine : chaire of excellence.
Competing financial interests
The authors declare no competing financial interests.
References
References
 [1] G. Kurizki, P. Bertet, Y. Kubo, K. Molmer, D. Petrosyan, P. Rabl and J. Schmiedmayer, ”Quantum technologies with hybrid systems”, PNAS 112, (2015).
 [2] R. Amsüss, C. Koller, T. Nöbauer, S. Putz, S. Rotter, K. Sandner, S. Schneider, M. Schramböck, G. Steinhauser, H. Ritsch, J. Schmiedmayer and J. Majer, ”Cavity QED with Magnetically Coupled Collective Spin States”, Phys. Rev. Lett. 107, (2011).
 [3] Y. Kubo, F. R. Ong, P. Bertet, D. Vion, V. Jacques, D. Zheng, A. Dréau, J.F. Roch, A. Auffeves, F. Jelezko, J. Wrachtrup, M. F. Barthe, P. Bergonzo and D. Esteve, ”Strong Coupling of a Spin Ensemble to a Superconducting Resonator”, Phys. Rev. Lett. 105, (2010).
 [4] I. Diniz, S. Portolan, R. Ferreira, J. M. Gérard, P. Bertet and A. Auffèves, ”Strongly coupling a cavity to inhomogeneous ensembles of emitters: Potential for longlived solidstate quantum memories”, Phys. Rev. A 84, (2011).
 [5] S. Bernon, H. Hattermann, D. Bothner, M. Knufinke, P. Weiss, F. Jessen, D. Cano, M. Kemmler, R. Kleiner, D. Koelle and J. Fortágh, ”Manipulation and coherence of ultracold atoms on a superconducting atom chip”, Nature Communications 4, (2013).
 [6] P. Weiss, M. Knufinke, S. Bernon, D. Bothner, L. Sárkány, C. Zimmermann, R. Kleiner, D. Koelle, J. Fortágh and H. Hattermann, ”Sensitivity of Ultracold Atoms to Quantized Flux in a Superconducting Ring”, Phys. Rev. Lett. 114, (2015).
 [7] A. Bienfait, J. J. Pla, Y. Kubo, M. Stern, X. Zhou, C. C. Lo, C. D. Weis, T. Schenkel, M. L. W. Thewalt, D. Vion, D. Esteve, B. Julsgaard, K. Molmer, J. J. L. Morton and P. Bertet, ”Reaching the quantum limit of sensitivity in electron spin resonance”, Nature Nanotechnology 11, (2016).
 [8] O. Arcizet, V. Jacques, A. Siria, P. Poncharal, P. Vincent and S. Seidelin, ”A single nitrogenvacancy defect coupled to a nanomechanical oscillator”, Nature Physics 7, (2011).
 [9] S. Kolkowitz, A. C. Bleszynski Jayich, Q. P. Unterreithmeier, S. D. Bennett, P. Rabl, J. G. E. Harris and M. D. Lukin, ”Coherent Sensing of a Mechanical Resonator with a SingleSpin Qubit”, Science 335, (2012).
 [10] K. Fang, M. H. Matheny, X. Luan and O. Painter, ”Optical transduction and routing of microwave phonons in cavityoptomechanical circuits”, Nature Photonics 10, (2016).
 [11] A. Vochezer, T. Kampschulte, K. Hammerer and P. Treutlein, ”LightMediated Collective Atomic Motion in an Optical Lattice Coupled to a Membrane”, Phys. Rev. Lett. 120, (2018).
 [12] I. Bloch, J. Dalibard and S. Nascimbene, ”Quantum simulations with ultracold quantum gases”, Nature Physics 8, (2012).
 [13] K. Henschel, J. Majer, J. Schmiedmayer and H. Ritsch, ”Cavity QED with an ultracold ensemble on a chip: Prospects for strong magnetic coupling at finite temperatures”, Phys. Rev. A 82, (2010).
 [14] H. Hattermann, D. Bothner, L. Y. Ley, B. Ferdinand, D. Wiedmaier, L. Sárkány, R. Kleiner, D. Koelle and J. Fortágh, ”Coupling ultracold atoms to a superconducting coplanar waveguide resonator”, Nature Communications 8, (2017).
 [15] A. Pappa, N. Kumar, T. Lawson, M. Santha, S. Zhang, E. Diamanti and I. Kerenidis, ”Nonlocality and Conflicting Interest Games”, Phys. Rev. Lett. 114, (2015).
 [16] A. Goban, C.L. Hung, S.P. Yu, J.D. Hood, J.A. Muniz, J.H. Lee, M.J. Martin, A.C. McClung, K.S. Choi, D.E. Chang, O. Painter and H.J. Kimble, ”Atomâlight interactions in photonic crystals”, Nature Communications 5, (2014).
 [17] C. Sayrin, C. Clausen, B. Albrecht, P. Schneeweiss and A. Rauschenbeutel, ”Storage of fiberguided light in a nanofibertrapped ensemble of cold atoms”, Optica 2, (2015).
 [18] J. D. Thompson, T. G. Tiecke, N. P. de Leon, J. Feist, A. V. Akimov, M. Gullans, A. S. Zibrov, V. Vuletić and M. D. Lukin, ”Coupling a Single Trapped Atom to a Nanoscale Optical Cavity”, Science 340, (2013).
 [19] J. S. Douglas, H. Habibian, C.L. Hung, A. V. Gorshkov, H. J. Kimble and D. E. Chang, ”Quantum manybody models with cold atoms coupled to photonic crystals”, Nature Photonics 9, (2015).
 [20] X. Zang, J. Yang, R. Faggiani, C. Gill, P. G. Petrov, J.P. Hugonin, K. Vynck, S. Bernon, P. Bouyer, V. Boyer and P. Lalanne, ”Interaction between Atoms and Slow Light: A Study in Waveguide Design”, Phys. Rev. Applied 5, (2016).
 [21] C. Stehle, C. Zimmermann and S. Slama, ”Cooperative coupling of ultracold atoms and surface plasmons”, Nature Physics 10, (2014).
 [22] M. Gullans, T.G. Tiecke, D.E. Chang, J. Feist, J.D. Thompson, J.I. Cirac, P. Zoller and M.D. Lukin, ”Nanoplasmonic lattices for ultracold atoms”, Phys. Rev. Lett. 109, (2012).
 [23] J. D. Hood, A. Goban, A. AsenjoGarcia, M. Lu, S.P. Yu, D. E. Chang and H. J. Kimble, ”Atomâatom interactions around the band edge of a photonic crystal waveguide”, PNAS 113, (2016).
 [24] A. GonzalezTudela, C.L. Hung, D.E. Chang, J.I. Cirac and H.J. Kimble, ”Subwavelentgh vacuum lattices and atomatom interactions in photonic crystals”, Nature Photonics 9, (2015).
 [25] I. Bloch, J. Dalibard and W. Zwerger, ”Manybody physics with ultracold gases”, Rev. Mod. Phys. 80, (2008).
 [26] J.S. Bernier, C. Kollath, A. Georges, L. De Leo, F. Gerbier, C. Salomon and M. Köhl, ”Cooling fermionic atoms in optical lattices by shaping the confinement”, Phys. Rev. A 79, (2009).
 [27] R. A. Hart, P. M. Duarte, T.L. Yang, X. Liu, T. Paiva, E. Khatami, R. T. Scalettar, N. Trivedi, D. A. Huse and R. G. Hulet, ”Observation of antiferromagnetic correlations in the Hubbard model with ultracold atoms”, Nature 519, (2015).
 [28] L. W. Cheuk, M. A. Nichols, M. Okan, T. Gersdorf, V. V. Ramasesh, W. S. Bakr, T. Lompe and M. W. Zwierlein, ”QuantumGas Microscope for Fermionic Atoms”, Phys. Rev. Lett. 114, (2015).
 [29] D. Greif, M. F. Parsons, A. Mazurenko, C. S. Chiu, S. Blatt, F. Huber, G. Ji and M. Greiner, ”Siteresolved imaging of a fermionic Mott insulator”, Science 351, (2016).
 [30] A. Mazurenko, C. S. Chiu, G. Ji, M. F. Parsons, M. KanÃ¡szNagy, R. Schmidt, F. Grusdt, E. Demler, D. Greif and M. Greiner, ”A coldatom FermiâHubbard antiferromagnet”, Nature 545, (2017).
 [31] C. Stehle, H. Bender, C. Zimmermann, D. Kern, M. Fleischer and S. Slama, ”Plasmonically tailored micropotentials for ultracold atoms”, Nature Photonics 5, (2011).
 [32] J. D. Thompson, T. G. Tiecke, N. P. de Leon, J. Feist, A. V. Akimov, M. Gullans, A. S. Zibrov and V. Vuleti and M. D. Lukin, ”Coupling a Single Trapped Atom to a Nanoscale Optical Cavity”, Science 340, (2013).
 [33] D.E. Chang, K. Sinha, J.M. Taylor and H.J. Wolf, ”Trapping atoms using nanoscale quantum vacuum forces”, Nature communications 5, (2014).
 [34] B.Dubetsky and P. R Berman, ”period optical potentials”, Phys. Rev. A 66, (2002).
 [35] G. Ritt, C. Geckeler, T. Salger and G. Cenn, ”Fourier synthesis of optical potentials for atomic quantum gases”, Phys. Rev. A 74, (2006).
 [36] S. Nascimbene, N. Goldman, N. R. Cooper and J. Dalibard, ”Dynamic Optical Lattices of Subwavelength Spacing for Ultracold Atoms”, Phys. Rev. Lett. 115, (2015).
 [37] O. RomeroIsart, C. Navau, A. Sanchez, P. Zoller and J. I. Cirac, ”Superconducting Vortex Lattices for Ultracold Atoms”, Phys. Rev. Lett. 111, (2013).
 [38] H.B.G. Casimir and D. Polder, ”The Influence of Retardation on the Londonvan der Waals Forces”, Phys. Rev. 73, (1948).
 [39] M. Mildner, A. Horrer, M. Fleischer, C. Zimmermann and S. Slama, ”Plasmonic surface traps with arbitrary shape for cold atoms”, QST …, (2018).
 [40] J. Dalibard and C. CohenTannoudji, ”Dressedatom approach to atomic motion in laser light: the dipole force revisited”, J. Opt. Soc. Am. B 2, (1985).
 [41] P. Yeh, ”Optical waves in layered media”, John Wiley and Sons, NewYork, (1998).
 [42] J.P. Hugonin and P. Lalanne, ”Reticolo software for grating analysis”, Institut d’Optique, Palaiseau, (2005).
 [43] S. Y. Buhmann, L. Knöll, D.G. Welsch and H. T. Dung ”CasimirPolder forces : A nonperturbative approach”, Phys. Rev. A 70, (2004).
 [44] B. Arora, M. S. Safranova and C. W. Clark, ”Magic wavelengths for the transitions in alkalimetal atoms”, Phys. Rev. A 76, (2007).
 [45] D. J. Griffiths, ”Introduction to quantum mechanics”, Prentice Hall editions, (1995).
 [46] R. Guérout, J. Lussange, H.B. Chan, A. Lambrecht and S. Reynaud, ”Thermal Casimir forces between nanostructured surfaces”, Physical Review A 87, (2013).
 [47] A. M. ContrerasReyes, R. Guérout, P.A. Maia Neto, D.A.R. Dalvit, A. Lambrecht and S. Reynaud, ”CasimirPolder interaction between an atom and dielectric grating”, Physical Review A 82, (2010).
 [48] S. Y. Buhmann, V. N. Marachevsky and S. Scheel, ”Impact of anisotropy on the interaction of an atom with a onedimensional nanograting”, Int. J. Mod. Phys. A31, (2016).
 [49] M. G. Moharam, E. B. Grann, D. A. Pommet and T. K. Gaylord, ”Formulation for stable and efficient implementation of the rigorous coupledwave analysis of binary gratings”, J. Opt. Soc. Am. A 12, (1995).
 [50] P. Lalanne and M. P. Jurek, ”Computation of the nearfield pattern with the coupledwave method for transverse magnetic polarization”, Journal of Modern Optics 45, (1998).
 [51] S. A. Kara, A. Keffous, A. M. Giovannozzi, A. M. Rossi, E. Cara, L. D’Ortenzi, K. Sparnacci, L. Boarino, N. Gabouze and S. Soukane, ”Fabrication of flexible silicon nanowires by selfassembled metal assisted chemical etching for surface enhanced Raman spectroscopy”, RSC Advances 6, (2016).
 [52] J. S. Douglas, H. Habibian, C.L. Hung, A. V. Gorshkov, H. J. Kimble and D. E. Chang, ”Quantum manybody models with cold atoms coupled to photonic crystals”, Nature Photonics 9, (2015).