Europe PMC

This website requires cookies, and the limited processing of your personal data in order to function. By using the site you are agreeing to this as outlined in our privacy notice and cookie policy.

Abstract 


Optical tractor beams capable of pulling particles backward have garnered significant and increasing interest. Traditional optical tractor beams are limited to free space beams with small forward wavevectors, enabling them to pull selected particles. Here, we present a comprehensive theory for the optical force exerted by a surface wave using analytical and numerical calculations, revealing the relationship between the canonical momentum and optical forces. Based on this theory, we demonstrate a general purpose optical surface tractor beam that can pull any passive particle, regardless of size, composition, or geometry. The tractor beam utilizes a surface wave with negative canonical momentum characterized by a single well-defined negative Bloch k vector. The tractor beam relies on a mechanism where the negative incident force always surpasses the recoil force. As such, the tractor beam, when excited on the surface of a double-negative index metamaterial, can pull particles with different morphologies.

Free full text 


Logo of ncommsLink to Publisher's site
Nat Commun. 2024; 15: 6836.
Published online 2024 Aug 9. https://doi.org/10.1038/s41467-024-51100-7
PMCID: PMC11315692
PMID: 39122709

Morphology-independent general-purpose optical surface tractor beam

Associated Data

Supplementary Materials
Data Availability Statement

Abstract

Optical tractor beams capable of pulling particles backward have garnered significant and increasing interest. Traditional optical tractor beams are limited to free space beams with small forward wavevectors, enabling them to pull selected particles. Here, we present a comprehensive theory for the optical force exerted by a surface wave using analytical and numerical calculations, revealing the relationship between the canonical momentum and optical forces. Based on this theory, we demonstrate a general purpose optical surface tractor beam that can pull any passive particle, regardless of size, composition, or geometry. The tractor beam utilizes a surface wave with negative canonical momentum characterized by a single well-defined negative Bloch k vector. The tractor beam relies on a mechanism where the negative incident force always surpasses the recoil force. As such, the tractor beam, when excited on the surface of a double-negative index metamaterial, can pull particles with different morphologies.

Subject terms: Optical manipulation and tweezers, Nanophotonics and plasmonics

Introduction

When a laser beam is scattered by a particle, its momentum exchange, which is mediated by Lorentz force, results in optical forces. In some specially prepared situations, the forward photon momentum increases upon scattering; thus, because of momentum conservation, the particle experiences a negative pulling force17. This specially prepared beam that pulls is also referred to as an optical tractor beam in the literature815. This beam is typically realized in free space1627 and occasionally near a surface2837. However, such pulling can be achieved only when the beam and particle are both carefully prepared to tailor the required light–matter interaction. This makes pulling difficult and severely limits its applicability.

Recently, optical pulling schemes in waveguides3843 and metamaterials4446 have been proposed that partially alleviate the requirements on the particles but do not completely solve the problem. To remedy this issue, we propose a general-purpose optical surface tractor beam (OSTB) that always pulls irrespective of the particle composition and morphology. We stress that the OSTB proposed here relies on a mechanism different from those of previous techniques. Its robustness does not stem from directional conversions between limited modes3841 or topological protection41,47.

On the basis of the electromagnetic (EM) energy–momentum tensor, we formally establish a comprehensive analytical theory for the optical force exerted on an arbitrary particle near a planar surface illuminated by a surface wave (SW) with a single well-defined Bloch k. This theory indicates that the longitudinal force is always parallel to the canonical momentum ([variant Planck's over 2pi]k)48, where [variant Planck's over 2pi] is Planck’s constant. Therefore, when an SW propagates forward with a negative k and carries a negative canonical momentum, it always exerts a pulling force instead of a pushing force. This OSTB can be regarded as an extension of the traditional optical tractor beam, offering enhanced robustness and reduced selectivity in terms of particle and beam properties. The optical force acting on a particle by a beam consists of the incident force and the recoil force. The former is due to the capture of incident photons, whereas the latter is due to the re-emission of photons. Unlike traditional approaches, which typically involve the adjustment of incident and scattering angles to reduce the incident force and increase the recoil force1,815, the optical pulling effect observed in this context is uniquely driven by the negative incident force associated with a negative k value. This characteristic endows it with independence from scattering feature control, resulting in exceptional robustness. Next, we present a demonstration illustrating the generation of such an OSTB by employing a double-negative metamaterial with ε, μ < 0.

Our conclusion regarding the validity of the longitudinal force being always parallel to the canonical momentum applies specifically to the SW characterized by a single, well-defined Bloch k. SWs with multiple Bloch k components possess distinct properties and are not within the scope of this study4952. Compared with optical pulling schemes implemented within waveguides3843 and metamaterials4446, the OSTB is more robust and convenient. This feature is noted because the gradient force exerted by the OSTB automatically captures particles in the transverse direction, thereby facilitating their stable transportation.

Owing to its ability to pull arbitrary particle or particle clusters (where the entire cluster can be viewed as a single particle), the proposed OSTB can be applied to manipulate particles in an optofluidic or lab-on-a-chip environment, where one can simultaneously pull a collection of particles backward without worrying about the particle morphology or distribution.

Results

Deep subwavelength particles

We begin by examining a simple yet instructive example of a small particle positioned above a planar isotropic metamaterial substrate, as depicted in Fig. 1a. The metamaterial is assumed to be lossless and transparent. Let us assume that a SW (indicated by the white arrow) is launched from the left, with a simplified time-independent form given by Einc=E0eikpx, where E0 is the complex amplitude decaying away from the surface and kp is the propagating wavenumber. Upon scattering, the incident SW will be converted into three components: SWs propagating in all directions on the xoz plane (represented by yellow arrows), freely propagating waves (FPWs) in the air (indicated by blue arrows), and FPWs within the metamaterial (illustrated by purple arrows). If the particle is small enough such that the scattering between the particle and the substrate can be safely ignored, the longitudinal (along the x direction) optical force of the particle can be approximated as follows5355:

Fx=12Re[xEinc*(αeEinc)]=12kpE02Im(αe)=12kpE02ε0k0σ,
1

where αe and σ are the polarizability and scattering cross-sections of the particle, respectively; ε0 is the permittivity of vacuum; and k0 is the wavenumber in the surrounding air. Given that σ is always positive for passive particles, when kp<0, the longitudinal optical force is negative, indicating a pulling effect on the particle. Because the canonical (or Minkowski) momentum density of the SW is given by g=kpW/ω56,57, where W is the EM energy density and is always nonnegative, a SW with a negative kp carries a negative canonical momentum. This type of SW is characterized by a single well-defined negative Bloch wavevector k, which is inherently different from the so-called backward surface waves supported on the surfaces of photonic crystals4952. Backward surface waves, in contrast, possess multiple Bloch wavevector components and essentially carry positive canonical momenta. This difference increases the reliance of optical pulling by backward surface waves on the scattering features.

An external file that holds a picture, illustration, etc.
Object name is 41467_2024_51100_Fig1_HTML.jpg
Geometric sketches.

a Schematic illustration of the optical pulling system. The incident surface wave (SW) (white arrow) is scattered by the particle positioned immediately above the metamaterial and converted into scattered SWs (yellow arrows), freely propagating waves (FPWs) in the air (blue arrows) and within the metamaterial (purple arrows). The SWs are confined on the interface (xoy plane), whereas the FPWs can radiate in all directions. b Schematic of the cross section of the system on the x‒z plane. Arbitrary surfaces S1 and S4 are illustrated, along with planar surfaces S2 and S3, which are positioned infinitesimally above and below the interface, respectively. The intersection curves between these surfaces and the x‒z plane are drawn as dashed lines.

General theory of optical force for particles of arbitrary morphology

For an arbitrary particle, the optical force can be rigorously calculated using the formula Fopt=STmaxn^dS, where Tmax is the Maxwell stress tensor, S is any surface in the air that encloses the particle, and n^ is the unit outward normal vector on S. Here, we consider S as the combination of S1 and S2, where S2 is a planar surface immediately above the metamaterial substrate, and S1 is a cap that encompasses the particle when combined with S1, as shown in Fig. 1b. Given that the outward normal vector of S2 points in the negative-z direction, the longitudinal optical force can be alternatively expressed as follows:

Fx=S1x^Tmaxn^dS+S2x^Tmax(z^)dS.
2

Given that, in air, the Maxwell stress tensor is the same as the Minkowski stress tensor Tmin, we can substitute Tmax with Tmin in Eq. (2), and the last term of Eq. (2) can be rewritten as S2TxzmaxdS=S2TxzmindS, where Txzmin=x^Tminz^=12Re(ExDz*+HxBz*). Because Ex,Dz,Hx,Bz are continuous across the interface, Txzmin is also continuous, yielding the following equation:

S2TxzmindS=S3TxzmindS,
3

where S3 is immediately below the interface and mirrors S2 symmetrically; see Fig. 1b. In the case of a lossless homogeneous metamaterial, the integration of Tmin over any closed surface within the metamaterial is zero, for example, the closed surface formed by S3 + S4, as shown in Fig. 1b. As a part of the closed surface S3 + S4, the outward normal of S3 points in the positive-z direction. Therefore,

S3+S4x^Tminn^dS=S3TxzmindS+S4x^Tminn^dS=0.
4

Combining Eqs. (2)–(4), the longitudinal optical force can be calculated as follows

Fx=S1x^Tminn^dSS3TxzmindS+S3TxzmindS+S4x^Tminn^dS=S1+S4x^Tminn^dS.
5

Here, S1 + S4 form an arbitrary closed surface enclosing the particle. Thus, according to Eq. (5), the longitudinal optical force can be calculated by integrating the Minkowski stress tensor over any closed surface enclosing the particle.

Demonstrating the equivalence between Eqs. (2) and (5) constitutes a crucial step. The deliberate selection of the closed surface for integration greatly simplifies the subsequent derivation. Notably, opting for a surface situated at a considerable distance from the particle enables us to circumvent the intricacies associated with the complex near field. Because the Minkowski stress tensor is connected to the canonical momentum flux density of light, Eq. (5) establishes a connection between the longitudinal optical force and canonical momentum. We emphasize that the same treatment can be applied to the lateral optical force Fy but not to the transverse optical force Fz because Txymin is generally discontinuous across the planar interface.

By rigorous derivation (see Supplementary Note 1 for details), Eq. (5) can be further transformed into the following:

Fx=1vpWext1vpWsca(s)cosϕ+1cWsca(f,a)cosϑ1+1vfWsca(f,m)cosϑ2,
6

where Wext represents the extinction rate of the particle in the system; Wsca(s), Wsca(f,a), and Wsca(f,m) represent the scattering rates to the SWs, the FPWs in air, and the FPWs in the metamaterial, respectively; left angle bracketcos ϕright angle bracket, left angle bracketcos [theta]right angle bracket1, and left angle bracketcos [theta]right angle bracket2 represent the corresponding averaged cosines of the scattering angles with respect to the x direction; c represents the light speed in air; and vp and vf represent the phase speeds of the SW and FPW in the metamaterial, respectively. Equation (6) is a key result of our work. It can be regarded as an extension of the optical force by a plane wave in free space expressed as follows1,58,59:

Fx=1c(WextWsca(f,a)cosϑ1).
7

Equation (6) reduces to Eq. (7) when the incident beam is a plane wave and is only scattered into the FPWs within the air. Compared with Eq. (7), in the context of free space, the optical force described by Eq. (6) can be further adjusted by manipulating the phase speed, in addition to the scattering angle. This introduces a different approach to achieving optical pulling. Given that the same treatment can be applied to the lateral optical force, we can obtain a similar expression to Eq. (6) for Fy. This enables us to comprehend and explore the lateral optical force acting on the particle near an interface6064.

On the basis of the relationships among the canonical momenta, energy densities, and energy fluxes of SWs and FPWs, Eq. (6) also reveals that the longitudinal optical force is equivalent to the change in the total canonical momentum of light per unit time (see details in Supplementary Note 2). Like the optical force in free space53,65, the first term in Eq. (6) corresponds to the momentum transfer resulting from directly capturing the incident photons, representing the incident force, whereas the remaining terms correspond to the recoil forces induced by reemitting the captured photons.

In the absence of gain, the extinction rate is given as Wext=Wsca(s)+Wsca(f,a)+Wsca(f,m)+Wabs, where Wabs is the absorption rate of the particle in the system. Importantly, all these rates are nonnegative. When kp is negative, the phase speed of the SW, denoted as vp=ω/kp, is also negative. Note that vp>max{c,vf} (refer to the confirmation in the following section), and the longitudinal optical force satisfies the following equation:

Fx=1vpWabs+1vp(1cosϕ)Wsca(s)+1vpcosϑ1cWsca(f,a)+1vpcosϑ2vfWsca(f,m)<0.
8

Accordingly, any SW with a negative kp will exert a pulling force on any passive particle, thus qualifying it as a general-purpose OSTB. The magnitude of the incident force is clearly always greater than the magnitude of the recoil force. Therefore, when kp<0, the incident force is negative, ensuring that the total longitudinal optical force is pulling. This stands in stark contrast to traditional tractor beams, where optical pulling arises from a stronger recoil force. Similarly, the longitudinal optical force is always pushing when kp>0, as indicated in Eq. (8).

Metamaterial supporting the OSTB

We demonstrate the feasibility of supporting the OSTB on the surface of a specific type of double-negative index metamaterial. We focus on the TE-polarized OSTB, which is characterized by an electric field polarized in the xoz plane. The corresponding TM-polarized OSTB can be achieved using a duality transformation εμ.

For a TE-polarized surface mode supported on the interface between the metamaterial and air (the interface is at z = 0), considering that the tangential component should be continuous, the magnetic fields at two domains are expressed as follows:

Hy=hyeikpxκ0z,z>0hyeikpx+κz,z<0,
9

where hy is the complex amplitude of the magnetic field at the interface, and κ0 and κ are the decay constants in the air and metamaterial, respectively. In the lossless limit, the propagating wavenumber kp must be real, and the decay constants must be both real and positive. In the air and metamaterial,

kp2κ02=k02,kp2κ2=k02εμ,
10

where ε,μ are the relative permittivity and permeability of the metamaterial, respectively. According to Maxwell’s equation ×H=iωεε0E, the electric fields are expressed as follows:

E=iωε0(κ0hy,0,ikphy)Teikpxκ0z,z>0(ε1κhy,0,ε1ikphy)Teikpx+κz,z<0.
11

The continuity of electric field along the x direction across the interface z = 0 requires

κ0hy=ε1κhy.
12

Combining Eqs. (10) and (12), we obtain the following:

kp2=ε(εμ)ε21k02,κ=ε2(1εμ)ε21k0=εκ0.
13

Here, we have discarded the solutions in which κ,κ0 are negative. Because both κ0 and κ are positive, ε must be negative, according to the second of Eq. (13). Since kp and κ must be real, according to Eq. (13), one can obtain the following:

ε(εμ)ε21>0,1εμε21>0.
14

As previously discussed, the OSTB possesses a negative canonical momentum when it propagates forward. In the context of light propagation, the direction of the total energy flux serves as an indicator. Consequently, the OSTB can be regarded as an SW with a positive total energy flux but a negative propagating wave number kp. Using the expressions of Eqs. (9) and (11), the total energy flux of the TE-polarized SW is given as follows:

P(s)=12ReE×y^Hy*x^=12kpωε0hy20e2κ0zdz+01εe2κzdz=12kpωε0hy2εκ+κ02κκ0ε=14ωκ0ε0hy2ε21ε2kp.
15

Given that κ0 is a positive real value, for kp<0, the total energy flux is positive only when ε21ε2<0, which leads to the following:

ε2<1.
16

Substituting Eq. (16) into Eq. (14), we easily obtain εμ>1. Therefore, we can establish both the necessary and sufficient conditions for supporting the TE-polarized OSTB as follows:

1<ε<μ1<0.
17

Given that the OSTB is characterized by a negative kp, our selection for the propagating wavenumber of the TE-polarized OSTB is solely the negative solution found in the first part of Eq. (13), resulting in the following:

kp=ε(εμ)ε21k0.
18

Conversely, if the SW exhibits a positive canonical momentum, we opt for the positive solution. By duality, the conditions for supporting the TM polarized OSTB are expressed as follows:

1<μ<ε1<0.
19

The propagating wavenumber of a TM-polarized OSTB is expressed as follows:

kp=μ(με)μ21k0.
20

Using Eq. (17) and Eq. (13), it is easy to show that kp2=ω2vp2>k02,kp2>εμk02=ω2vf2, confirming that vp>max{c,vf}. As an example, Fig. 2a depicts the excitation of a TE-polarized SW with a negative kp on the surface of a metamaterial with ε = −0.9 and μ = −1.2. Applying the Fourier transform to the rightward-propagating SW yields the propagating wavenumber kp=1.19k0, which is consistent with Eq. (18), as shown in Fig. 2b.

An external file that holds a picture, illustration, etc.
Object name is 41467_2024_51100_Fig2_HTML.jpg
Excitation of a TE-polarized OSTB on a homogeneous metamaterial surface.

a OSTB and particle. The OSTB is excited by five line magnetic currents with a uniform amplitude of 1 V and position-dependent phases of eikpxi, where kp=1.19k0 and xi are the coordinates marking a point on the line currents along the x direction, as illustrated in the inset. The particle, with a radius of r, is positioned above the surface at a distance of h (from the center of the particle to the interface). The line sources operate at the frequency f = 100 THz (the vacuum wavelength is 3 μm). b Magnitude of the Fourier transform of the z-component electric field Ez of the incident OSTB as a function of the wavenumber. The Fourier transform is calculated as follows: Ez(k)=1xdxsxsxdEz(x,z0)eikxdx, where z0=0 denotes the position of the interface and [xs,xd] represents the interval used for the Fourier transform. (Source data are provided as a source data file.).

Robust and loss-enhanced optical pulling using the OSTB

We now demonstrate the OSTB using numerical simulations. To simplify the computational process, let us consider an infinite cylinder oriented along the y direction as the particle of interest. The longitudinal forces acting on this circular cylinder with a dielectric constant εr and a radius r are plotted in Fig. 3a. As expected from Eq. (8), these forces are consistently negative. Figure 3b further illustrates that the forces remain negative for transparent and lossy cylinders with circular, triangular, or square cross-sections. Notably, traditional optical tractor beams face significant challenges in pulling lossy particles. However, in our approach, we successfully overcome this limitation, enabling unconditional pulling regardless of the particle size, composition, or shape. Given that the longitudinal optical force is solely due to the transfer of canonical momentum, for an OSTB carrying a negative canonical momentum, any absorption will indeed intensify the optical pulling force. On the other hand, as indicated by Eq. (8), loss actually enhances the optical pulling force by increasing the absorption rate, except near resonances where loss can reduce the extinction and scattering rates, as depicted in Fig. 3b.

An external file that holds a picture, illustration, etc.
Object name is 41467_2024_51100_Fig3_HTML.jpg
Optical forces exerted by the OSTB.

a The longitudinal optical force, Fx, is plotted as a function of the dielectric constant, εr, and the radius, r, of the particle. The optical force is expressed in units of μN/m. The distance of the particle is h = 1 μm. b The longitudinal optical force, Fx, is plotted against the real part of the dielectric constant, Re(εr), for particles of different geometries. The black dashed line corresponds to the dashed line in a (r = 0.7 μm, Im(εr)=0). The red solid line denotes the longitudinal optical force when the particle experiences losses, with Im(εr)=0.5. The blue dotted and magenta dash-dotted lines correspond to transparent (Im(εr)=0) particles with triangular (side length equals 3r) and square (side length equals 2r) shapes, respectively. c The transverse optical force, Fz, is plotted against the distance, h, of a circular particle with a radius of r = 0.7 μm when its dielectric constant is set to 3 (black), 9 (red), and 15 (blue). d The trajectory (black dashed line) of the particle (r=0.7μm,εr=3) near the metamaterial is calculated by solving the motion equation. The arrow indicates the direction of movement. (Source data are provided as a source data file.).

In Fig. 3c, we present the transverse optical force Fz acting on the particle as a function of the height h (distance from the particle center to the interface) for different dielectric constants of the particle. The transverse optical force is governed primarily by the optical gradient force, which attracts the particle toward regions of high light intensity where the longitudinal pulling force is strong. Notably, for certain particles (e.g., r = 0.7 μm, εr = 9, as indicated by the red line in Fig. 3c), equilibrium positions exist along the y direction. These particles are suspended above the interface, avoiding contact with the surface and associated friction forces. As a result, the particles are initially drawn toward the interface by the gradient force and then pulled toward the light source, as observed in the particle trajectory shown in Fig. 3d and Supplementary Note 3. This feature provides a significant advantage over other optical pulling schemes in waveguides3943 and metamaterials4446, where the particle must be located inside the waveguides or metamaterial structures using additional means. The method of simulating the particle trajectory is introduced in the “Methods” section.

Optical pulling over a realistic metamaterial structure

We employ a photonic crystal structure to realize the metamaterial supporting the TM-polarized OSTB. The photonic crystal comprises hollow air cores periodically arranged in a square lattice embedded within a high dielectric host medium, as depicted in the inset of Fig. 4a. Figure 4a shows the corresponding band structure for TM polarization, with a focus on the band segment highlighted in red. Using the boundary effective medium approach66, we compute the effective constitutive parameters for this specific band segment and present the results in Fig. 4b. For frequencies below the critical point, the condition for supporting the TM-polarized OSTB, i.e., 1<μe<εe1<0, is satisfied. Using supercell calculations, the surface band dispersion is obtained and depicted by the line of red circles in Fig. 4c. For kp>0, the dispersion line clearly has a negative slope, indicating that the surface state has a negative group velocity. Consequently, a forward-propagating SW corresponds to a negative kp, precisely representing the OSTB.

An external file that holds a picture, illustration, etc.
Object name is 41467_2024_51100_Fig4_HTML.jpg
OSTBs supported on a realistic structure.

a Dispersion relation of the TM-polarized modes. The inset shows the unit cell, where the cyan region represents the background medium characterized by a relative permittivity of εb=120, and the white circle (with a radius of rc=0.1a) represents the air hole. b Effective constitutive parameters corresponding to the highlighted band segment in (a) are presented. c A projection of the bulk bands (depicted by gray disks) and the surface band (highlighted by red circles) is shown. The inset displays the normalized electric field distribution within a supercell, where the black disks and dashed line indicate the air holes and the interface between the photonic crystal and air, respectively. (Source data are provided as a source data file.).

The assumption of a dispersive-less host medium may appear unreasonable, especially in high-frequency regimes. However, we can circumvent this limitation by substituting it with a composite metamaterial incorporating metallic components. Despite the composite metamaterial background exhibiting dispersion, the condition for supporting the TM-polarized OSTB remains satisfied, as elaborated in Supplementary Note 4.

In simulations, the OSTB is excited using 13 line sources embedded inside the host dielectric medium, and the phases of these line sources are adjusted to match the propagating wavenumber kp of the OSTB, as depicted in Figs. 5a and and5b.5b. As shown in Fig. 5b, the line sources, each with an amplitude of 1 mA, are positioned one lattice constant below the metamaterial surface and are equally spaced with a distance a between them. From left to right, the initial phase of the j-th line source is (1-j)×0.2π. The line sources are used to facilitate numerical simulations. Because the surface states are not within a complete band gap, as illustrated in Fig. 4c, the FPWs inside the metamaterial are also excited, which reduces the excitation efficiency. However, because the FPWs decay rapidly as they propagate, whereas the OSTB does not, the OSTB becomes dominant in the region far away from the sources. Applying the Fourier transform to the electric fields on the metamaterial surface yields a negative propagating wavenumber, as shown in Fig. 5d, confirming that the SW is indeed an OSTB.

An external file that holds a picture, illustration, etc.
Object name is 41467_2024_51100_Fig5_HTML.jpg
Optical pulling over a realistic structure.

a TM-polarized OSTB excited on the surface of a metamaterial with a real microstructure. b A zoomed-in view of the source region indicated by the black dashed rectangle in Panel a. The black disks in Panel b represent the air holes, whereas the black crosses denote the line sources (LSs) with an amplitude of 1 mA. The lattice constant is assumed to be a=100 nm, and the vacuum wavenumber of the sources is k0=0.527/a. c OSTB excited using a prism with a refractive index of 1.69. The direction of the incident bean is indicated by the black arrows, which are perpendicular to the surface of the prism. The wavenumber of the incident beam along the x direction is matched with that of the OSTB. The gray regions in Panels a and c correspond to the absorbing metamaterials. d Fourier transform spectrum of the electric field on the metamaterial surfaces (cyan dashed lines in Panels a and c). e The longitudinal optical forces on circular particles of different sizes versus the displacement of the particles along the x-axis, denoted as δ. The relative permittivity of the particles is εr=5, and the gap between the bottom of the particles and the surface of the metamaterial is h=a. f The longitudinal optical forces of the dielectric (DP, εr=5), plasmonic (PP, εr=3+0.5i), and chiral (CP, εr=5 and chirality parameterχ=1.0) particles versus their dimensionless particle sizek0r. All the particles are circular. The optical forces on the plasmonic particles are increased 10-fold. g Pulling forces on circular, square and triangular shaped particles arranged equidistantly. The particles are transparent (εr=9), lossy (εr=9+0.5i), and chiral (εr=9,χ=1.0), respectively, with k0r=1.5. (Source data are provided as a source data file.).

To excite the SW, it is crucial to phase match its negative k vector. This can be achieved using a prism (Fig. 5c) or a series of line sources embedded inside the metamaterial (Fig. 5a). Like other eigenmodes of a photonic crystal, there is no closed-form expression for the SW. However, the beam profile has been calculated numerically, as shown in Fig. 4d, and an enlarged portion of Fig. 5a is also given in Supplementary Fig. 4. For prism excitation of the TM-polarized OSTB, an approximate Gaussian incident wave normally illuminates the right surface of the prism. An incident and prism angle of 45 degrees and a prism refractive index of 1.69 are chosen such that the incident wave vector, after entering the prism, can phase match the SW. If the refractive index of the prism differs from 1.69, then another appropriate incident angle can be adopted to phase match the SW and achieve efficient excitation. Following one of the standard approaches in COMSOL Multiphysics finite element simulation, the incident wave is generated using Gaussian distributed current sources located at the right surface of the prism, i.e., J0exp(rg2/w02), with its center coinciding with the center of the right surface, J0=1mA/m, where w0=15a and rg is the distance from the surface center to the local point on the surface. Then, the focal plane coincides with the prism’s right surface, with a beam waist of ~15a and an electric field amplitude of ~1.37 V/m at the beam focus. In Supplementary Note 5, we explore how the incident angle and polarization of the incident beam affect the excitation efficiency.

As the lattice structure disrupts the continuous translational symmetry of the substrate, the longitudinal optical force is, in principle, no longer independent of the x-coordinate of the particle. However, even when the particle’s radius is slightly smaller than the lattice constant, as shown by the black circle line in Fig. 5e (r=0.95a), the longitudinal optical force exhibits minimal dependence on the position along the x direction. Therefore, if the particle size is not too small, the longitudinal optical force can be considered constant with respect to the x-coordinate of the particle. In Fig. 5f, we illustrate the longitudinal optical forces acting on dielectric (black dashed), plasmonic (red solid), and chiral (blue dash-dotted) circular particles as functions of their dimensionless size parameters k0r. It is evident that the longitudinal optical forces remain negative across all particle sizes, spanning from Rayleigh to Mie sizes. In Supplementary Fig. 6, we present the longitudinal optical forces on particles of various shapes and constitutive parameters, which consistently exhibit negativity.

Finally, we explore the optical pulling phenomenon involving multiple particles. We consider three particles of different shapes arranged equidistantly along the x direction, as illustrated in the inset of Fig. 5g. These particles are transparent, lossy, and chiral, separately, with identical real parts of relative permittivity. All three particles experience optical pulling forces, regardless of their separation. When the particle sizes and constitutive parameters are varied, some of the particles may experience a pushing optical force for certain separations due to optical binding, as discussed in detail in Supplementary Note 6. However, the total longitudinal optical force acting on the three particles remains negative. This phenomenon occurs because, collectively, the total longitudinal optical force on all the particles is still determined by Eq. (8). Therefore, the OSTB can exert a pulling force on a cluster of particles.

Previous studies, such as references38,67, have demonstrated the use of photonic crystal structures to induce optical pulling forces. However, our approach is fundamentally different. Owing to the lattice constant being much smaller than the wavelength, the photonic crystal we employ functions effectively as a double-negative-index metamaterial. This characteristic allows the SW to be well described by a single negative Bloch wavevector. This unique property of SW enables a dominant negative incident force, ensuring the optical pulling of arbitrary particles along the surface of the photonic crystal structure. Unlike previous works38,67, our approach does not rely on any directional waveguide mode conversions, eliminating the need to bound the particle on both sides.

Discussion

We emphasize that our theory regarding longitudinal and lateral optical forces is rigorously established because it originates from a well-defined optical force derivation and introduces no unjustified approximations. Although SWs do propagate partially within media and an ongoing debate exists regarding the definition of light momentum within media, commonly known as the Abraham–Minkowski controversy48,68, it is essential to emphasize that the derivation of Eq. (6) is independent of any specific choice or definition of optical momentum. As a result, our derivations remain unaffected by the complexities surrounding the Abraham–Minkowski controversy. The sole impractical assumption made in practice pertains to the metamaterial being lossless, a requirement essential for the derivations. Nevertheless, in Supplementary Note 7, we showed that the OSTB remains effective even when the metamaterial substrate exhibits some degree of loss. As shown in Supplementary Fig. 8, when the relative permittivity and permeability of the metamaterial become ε=0.9+0.02i,μ=1.2+0.02i, the particle continues to experience a pulling force. However, given that the OSTB decays as it propagates, the optical pulling range is shortened. Our theory, on the one hand, reveals the underlying physics of optical forces mediated by SWs. On the other hand, it paves the way for distinctive approaches to customize optical forces, extending beyond the sole objective of optical pulling.

Moreover, Eq. (6) establishes a direct connection between the optical force and the energy of the scattered photons. Because the phase speed vp of the SW is typically slower than the speed of light c, a comparison between Eqs. (6) and (7) reveals that the SW has the potential to be more efficient than an FPW in free space for generating optical force when an equal number of photons are scattered. This, to some extent, highlights the advantages of SWs in optical manipulation.

Compared with traditional tractor beams and other optical pulling techniques, the OSTB not only employs SWs but, more notably, leverages a mechanism that is entirely different from those of previous techniques. This mechanism centers around manipulating the phase speed of the SW to generate a negative incident force, as opposed to modifying the scattering characteristics. Consequently, there is no need to customize particle properties such as size, shape, or composition, and limited waveguide modes or topological materials are not needed to regulate scattering paths. Notably, this mechanism is almost exclusive to the OSTB. In free space, the speed of light for FPW remains constant; within media, although the phase speed of FPW can be adjusted according to the refractive index, a credible theory of optical force is lacking owing to the ongoing Abraham–Minkowski controversy.

In summary, we have introduced a general-purpose OSTB utilizing SWs with negative canonical momenta that can pull passive particles, regardless of their size, composition, or geometry. Our analytical theory, derived from the EM energy–momentum tensor, reveals the underlying mechanism of this pulling phenomenon, which distinguishes it from traditional tractor beams. Furthermore, we have demonstrated that the OSTB can be implemented on the surface of double-negative index metamaterials.

Methods

Numerical simulations

Full-wave numerical simulations were performed using the finite element package COMSOL Multiphysics. The optical forces shown in Figs. 3 and and55 are calculated rigorously by integrating the time-averaged Maxwell stress tensor over a closed surface embedded in air. For the results presented in Figs. 2 and and3,3, we utilized the free triangular mesh and refined it thrice. Conversely, for the results depicted in Fig. 5, we employed the free triangular mesh with two refinements. In all the simulations, we utilized the Direct solver.

The dynamic trajectory of the particle near the metamaterial substrate is obtained by iterating the motion equation with time as follows:

md2dt2r=Foptγddtr,
22

where m is the mass of the particle, r is the location of the mass center, Fopt is the optical force and γ is the ambient damping constant. Because the particle is immersed in air or a vacuum, the ambient damping can be ignored for a large optical force. However, for calculation purposes, we assume a damping constant of γ=20pN/μm2 per 1μm length along the y direction. This can be approximated as the damping contributed by a sphere with a radius of 0.7μm in the air at room temperature. Additionally, the mass density of the particle is assumed to be ρ=2500kg/m3, which is on the same order of magnitude as that of most solid materials. Distinct mass densities may lead to slight variations in trajectories, but the fundamental properties remain unchanged. A restitution coefficient is introduced to account for the collision between the particle and the metamaterial substrate. In Fig. 3d, a value of 0.7 is employed for this coefficient.

Supplementary information

Peer Review File(3.8M, pdf)

Acknowledgements

This work is supported by the Key Project of the National Key Research and Development Program of China (2022YFA1404500), the National Natural Science Foundation of China (12074267, 12174263, 12074169) and the Guangdong Province Talent Recruitment Program (2021QN02C103).

Author contributions

N.W. and J.N. conceived and designed the research. G.P.W. oversaw and directed the entire project. All the authors discussed the results and contributed to the preparation of the manuscript.

Peer review

Peer review information

Nature Communications thanks the anonymous reviewer(s) for their contribution to the peer review of this work. A peer review file is available.

Data availability

Source data are provided with this paper. The source data generated in this study have been deposited in the figshare under accession code 10.6084/m9.figshare.25605495.

Competing interests

The authors declare no competing interests.

Footnotes

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Contributor Information

Jack Ng, nc.ude.hcetsus@3hzuw.

Guo Ping Wang, nc.ude.uzs@gnawpg.

Supplementary information

The online version contains supplementary material available at 10.1038/s41467-024-51100-7.

References

1. Chen, J., Ng, J., Lin, Z. & Chan, C. T. Optical pulling force. Nat. Photon5, 531–534 (2011).10.1038/nphoton.2011.153 [CrossRef] [Google Scholar]
2. Dogariu, A., Sukhov, S. & Sáenz, J. Optically induced ‘negative forces’. Nat. Photon.7, 24–27 (2013).10.1038/nphoton.2012.315 [CrossRef] [Google Scholar]
3. Ding, K., Ng, J., Zhou, L. & Chan, C. T. Realization of optical pulling forces using chirality. Phys. Rev. A89, 063825 (2014).10.1103/PhysRevA.89.063825 [CrossRef] [Google Scholar]
4. Wang, N., Chen, J., Liu, S. & Lin, Z. Dynamical and phase-diagram study on stable optical pulling force in bessel beams. Phys. Rev. A87, 063812 (2013).10.1103/PhysRevA.87.063812 [CrossRef] [Google Scholar]
5. Canaguier-Durand, A. & Genet, C. Chiral route to pulling optical forces and left-handed optical torques,. Phys. Rev. A92, 043823 (2015).10.1103/PhysRevA.92.043823 [CrossRef] [Google Scholar]
6. Lee, E., Huang, D. & Luo, T. Ballistic supercavitating nanoparticles driven by single gaussian beam optical pushing and pulling forces. Nat. Commun.11, 2404 (2020). 10.1038/s41467-020-16267-9 [Europe PMC free article] [Abstract] [CrossRef] [Google Scholar]
7. Ali, R., Pinheiro, F., Dutra, R. & Neto, P. M. Tailoring optical pulling forces with composite microspheres,. Phys. Rev. A102, 023514 (2020).10.1103/PhysRevA.102.023514 [CrossRef] [Google Scholar]
8. Sukhov, S. & Dogariu, A. On the concept of “tractor beams”. Opt. Lett.35, 3847–3849 (2010). 10.1364/OL.35.003847 [Abstract] [CrossRef] [Google Scholar]
9. Sukhov, S. & Dogariu, A. Negative nonconservative forces: optical “tractor beams” for arbitrary objects,. Phys. Rev. Lett.107, 203602 (2011). 10.1103/PhysRevLett.107.203602 [Abstract] [CrossRef] [Google Scholar]
10. Saenz, J. J. Laser tractor beams. Nat. Photon.5, 514–515 (2011).10.1038/nphoton.2011.201 [CrossRef] [Google Scholar]
11. Ruffner, D. B. & Grier, D. G. Optical conveyors: a class of active tractor beams. Phys. Rev. Lett.109, 163903 (2012). 10.1103/PhysRevLett.109.163903 [Abstract] [CrossRef] [Google Scholar]
12. Novitsky, A., Qiu, C.-W. & Wang, H. Single gradientless light beam drags particles as tractor beams. Phys. Rev. Lett.107, 203601 (2011). 10.1103/PhysRevLett.107.203601 [Abstract] [CrossRef] [Google Scholar]
13. Novitsky, A., Qiu, C.-W. & Lavrinenko, A. Material-independent and size-independent tractor beams for dipole objects. Phys. Rev. Lett.109, 023902 (2012). 10.1103/PhysRevLett.109.023902 [Abstract] [CrossRef] [Google Scholar]
14. Brzobohatý, O. et al. Experimental demonstration of optical transport, sorting and self-arrangement using a ‘tractor beam’. Nat. Photon7, 123–127 (2013).10.1038/nphoton.2012.332 [CrossRef] [Google Scholar]
15. Wang, N., Lu, W., Ng, J. & Lin, Z. Optimized optical “tractor beam” for core–shell nanoparticles. Opt. Lett.39, 2399–2402 (2014). 10.1364/OL.39.002399 [Abstract] [CrossRef] [Google Scholar]
16. Shvedov, V., Davoyan, A. R., Hnatovsky, C., Engheta, N. & Krolikowski, W. A long-range polarization-controlled optical tractor beam. Nat. Photon8, 846–850 (2014).10.1038/nphoton.2014.242 [CrossRef] [Google Scholar]
17. Chen, H., Liu, S., Zi, J. & Lin, Z. Fano resonance-induced negative optical scattering force on plasmonic nanoparticles. ACS Nano9, 1926–1935 (2015). 10.1021/nn506835j [Abstract] [CrossRef] [Google Scholar]
18. Zemánek, P., Volpe, G., Jonáš, A. & Brzobohatý, O. Perspective on light-induced transport of particles: from optical forces to phoretic motion. Adv. Opt. Photonics11, 577–678 (2019).10.1364/AOP.11.000577 [CrossRef] [Google Scholar]
19. Li, X., Chen, J., Lin, Z. & Ng, J. Optical pulling at macroscopic distances. Sci. Adv.5, eaau7814 (2019). 10.1126/sciadv.aau7814 [Europe PMC free article] [Abstract] [CrossRef] [Google Scholar]
20. Li, H. et al. Optical pulling forces and their applications. Adv. Opt. Photonics12, 288–366 (2020).10.1364/AOP.378390 [CrossRef] [Google Scholar]
21. Lepeshov, S. & Krasnok, A. Virtual optical pulling force. Optica7, 1024–1030 (2020).10.1364/OPTICA.391569 [CrossRef] [Google Scholar]
22. Gao, D. et al. Optical manipulation from the microscale to the nanoscale: fundamentals, advances and prospects. Light.: Sci. Appl.6, e17039 (2017). 10.1038/lsa.2017.39 [Europe PMC free article] [Abstract] [CrossRef] [Google Scholar]
23. Rodrigo, J. A., Martínez, Ó. & Alieva, T. Helix-shaped tractor and repulsor beams enabling bidirectional optical transport of particles en masse. Photonics Res.10, 2560–2574 (2022).10.1364/PRJ.468060 [CrossRef] [Google Scholar]
24. Hadad, B. et al. Particle trapping and conveying using an optical archimedes’ screw. Optica5, 551–556 (2018).10.1364/OPTICA.5.000551 [CrossRef] [Google Scholar]
25. Damková, J. et al. Enhancement of the ‘tractor-beam’ pulling force on an optically bound structure. Light.: Sci. Appl.7, 17135–17135 (2018). 10.1038/lsa.2017.135 [Europe PMC free article] [Abstract] [CrossRef] [Google Scholar]
26. Fernandes, D. E. & Silveirinha, M. G. Optical tractor beam with chiral light. Phys. Rev. A91, 061801 (2015).10.1103/PhysRevA.91.061801 [CrossRef] [Google Scholar]
27. Guo, G., Feng, T. & Xu, Y. Tunable optical pulling force mediated by resonant electromagnetic coupling. Opt. Lett.43, 4961–4964 (2018). 10.1364/OL.43.004961 [Abstract] [CrossRef] [Google Scholar]
28. Maslov, A. Resonant pulling of a microparticle using a backward surface wave. Phys. Rev. Lett.112, 113903 (2014). 10.1103/PhysRevLett.112.113903 [Abstract] [CrossRef] [Google Scholar]
29. Intaraprasonk, V. & Fan, S. Optical pulling force and conveyor belt effect in resonator–waveguide system. Opt. Lett.38, 3264–3267 (2013). 10.1364/OL.38.003264 [Abstract] [CrossRef] [Google Scholar]
30. Kajorndejnukul, V., Ding, W., Sukhov, S., Qiu, C.-W. & Dogariu, A. Linear momentum increase and negative optical forces at dielectric interface. Nat. Photon7, 787–790 (2013).10.1038/nphoton.2013.192 [CrossRef] [Google Scholar]
31. Lu, J. et al. Light-induced pulling and pushing by the synergic effect of optical force and photophoretic force. Phys. Rev. Lett.118, 043601 (2017). 10.1103/PhysRevLett.118.043601 [Abstract] [CrossRef] [Google Scholar]
32. Wang, N. et al. Optical pulling using topologically protected one way transport surface-arc waves. Phys. Rev. B105, 014104 (2022).10.1103/PhysRevB.105.014104 [CrossRef] [Google Scholar]
33. Jin, R., Xu, Y., Dong, Z.-G. & Liu, Y. Optical pulling forces enabled by hyperbolic metamaterials. Nano Lett.21, 10431–10437 (2021). 10.1021/acs.nanolett.1c03772 [Abstract] [CrossRef] [Google Scholar]
34. Ferrari, H., Zapata-Rodríguez, C. J. & Cuevas, M. Giant terahertz pulling force within an evanescent field induced by asymmetric wave coupling into radiative and bound modes. Opt. Lett.47, 4500–4503 (2022). 10.1364/OL.460202 [Abstract] [CrossRef] [Google Scholar]
35. Qiu, C.-W. et al. Photon momentum transfer in inhomogeneous dielectric mixtures and induced tractor beams. Light.: Sci. Appl.4, e278 (2015).10.1038/lsa.2015.51 [CrossRef] [Google Scholar]
36. Yang, L.-F. & Webb, K. J. Pushing and pulling optical pressure control with plasmonic surface waves. Phys. Rev. B103, 245124 (2021).10.1103/PhysRevB.103.245124 [CrossRef] [Google Scholar]
37. Petrov, M. I., Sukhov, S. V., Bogdanov, A. A., Shalin, A. S. & Dogariu, A. Surface plasmon polariton assisted optical pulling force. Laser Photonics Rev.10, 116–122 (2016).10.1002/lpor.201500173 [CrossRef] [Google Scholar]
38. Zhu, T. et al. Mode conversion enables optical pulling force in photonic crystal waveguides. Appl. Phys. Lett.111, 061105 (2017).10.1063/1.4997924 [CrossRef] [Google Scholar]
39. Zhu, T. et al. Self-induced backaction optical pulling force. Phys. Rev. Lett.120, 123901 (2018). 10.1103/PhysRevLett.120.123901 [Abstract] [CrossRef] [Google Scholar]
40. Li, H. et al. Momentum-topology induced optical pulling force. Phys. Rev. Lett.124, 143901 (2020). 10.1103/PhysRevLett.124.143901 [Abstract] [CrossRef] [Google Scholar]
41. Wang, N., Zhang, R.-Y. & Chan, C. T. Robust acoustic pulling using chiral surface waves. Phys. Rev. Appl.15, 024034 (2021).10.1103/PhysRevApplied.15.024034 [CrossRef] [Google Scholar]
42. Kostina, N., Petrov, M., Bobrovs, V. & Shalin, A. S. Optical pulling and pushing forces via bloch surface waves. Opt. Lett.47, 4592–4595 (2022). 10.1364/OL.464037 [Abstract] [CrossRef] [Google Scholar]
43. Webb, K. J. Negative electromagnetic plane-wave force in gain media. Phys. Rev. E84, 057602 (2011).10.1103/PhysRevE.84.057602 [Abstract] [CrossRef] [Google Scholar]
44. Wang, S., Ng, J., Xiao, M. & Chan, C. T. Electromagnetic stress at the boundary: photon pressure or tension? Sci. Adv.2, e1501485 (2016). 10.1126/sciadv.1501485 [Europe PMC free article] [Abstract] [CrossRef] [Google Scholar]
45. Shalin, A. S., Sukhov, S. V., Bogdanov, A. A., Belov, P. A. & Ginzburg, P. Optical pulling forces in hyperbolic metamaterials. Phys. Rev. A91, 063830 (2015).10.1103/PhysRevA.91.063830 [CrossRef] [Google Scholar]
46. Lee, E. & Luo, T. Long-distance optical pulling of nanoparticle in a low index cavity using a single plane wave. Sci. Adv.6, eaaz3646 (2020). 10.1126/sciadv.aaz3646 [Europe PMC free article] [Abstract] [CrossRef] [Google Scholar]
47. Wang, D., Qiu, C., Rakich, P. T. & Wang, Z. Guidewave photonic pulling force using one-way photonic chiral edge states, CLEO: QELS_Fundamental Science. (Optica Publishing Group, 2015)
48. Barnett, S. M. Resolution of the Abraham-Minkowski dilemma. Phys. Rev. Lett.104, 070401 (2010). 10.1103/PhysRevLett.104.070401 [Abstract] [CrossRef] [Google Scholar]
49. Foteinopoulou, S., Kafesaki, M., Economou, E. & Soukoulis, C. Backward surface waves at photonic crystals. Phys. Rev. B75, 245116 (2007).10.1103/PhysRevB.75.245116 [CrossRef] [Google Scholar]
50. Barvestani, J., Kalafi, M., Soltani-Vala, A. & Namdar, A. Backward surface electromagnetic waves in semi-infinite one-dimensional photonic crystals containing left-handed materials. Phys. Rev. A77, 013805 (2008).10.1103/PhysRevA.77.013805 [CrossRef] [Google Scholar]
51. Narimanov, E. E. Photonic hypercrystals. Phys. Rev.X 4, 041014 (2014).10.1103/PhysRevX.4.041014 [CrossRef] [Google Scholar]
52. Wang, D. & Wang, Z. Optical pulling force in periodic backward-wave waveguides, CLEO: QELS_Fundamental Science. (Optica Publishing Group, 2017).
53. Chaumet, P. C. & Rahmani, A. Electromagnetic force and torque on magnetic and negative-index scatterers. Opt. Express17, 2224–2234 (2009). 10.1364/OE.17.002224 [Abstract] [CrossRef] [Google Scholar]
54. Nieto-Vesperinas, M., Sáenz, J., Gómez-Medina, R. & Chantada, L. Optical forces on small magnetodielectric particles. Opt. Express18, 11428–11443 (2010). 10.1364/OE.18.011428 [Abstract] [CrossRef] [Google Scholar]
55. Jiang, Y., Chen, H., Chen, J., Ng, J. & Lin, Z. Universal relationships between optical force/torque and orbital versus spin momentum/angular momentum of light, Preprint at https://arxiv.org/abs/1511.08546 (2015).
56. Bliokh, K. Y., Bekshaev, A. Y. & Nori, F. Optical momentum, spin, and angular momentum in dispersive media. Phys. Rev. Lett.119, 073901 (2017). 10.1103/PhysRevLett.119.073901 [Abstract] [CrossRef] [Google Scholar]
57. Wang, N. et al. Optical forces on a cylinder induced by surface waves and the conservation of the canonical momentum of light. Opt. Express29, 20590–20600 (2021). 10.1364/OE.428134 [Abstract] [CrossRef] [Google Scholar]
58. Bohren, C. F. & Huffman, D. R. Absorption And Scattering Of Light By Small Particles (John Wiley & Sons, 2008).
59. Radescu, E. & Vaman, G. Exact calculation of the angular momentum loss, recoil force, and radiation intensity for an arbitrary source in terms of electric, magnetic, and toroid multipoles. Phys. Rev. E65, 046609 (2002).10.1103/PhysRevE.65.046609 [Abstract] [CrossRef] [Google Scholar]
60. Wang, S. & Chan, C. T. Lateral optical force on chiral particles near a surface. Nat. Commun.5, 3307 (2014). 10.1038/ncomms4307 [Europe PMC free article] [Abstract] [CrossRef] [Google Scholar]
61. Rodríguez-Fortuño, F. J., Engheta, N., Martínez, A. & Zayats, A. V. Lateral forces on circularly polarizable particles near a surface. Nat. Commun.6, 8799 (2015). 10.1038/ncomms9799 [Europe PMC free article] [Abstract] [CrossRef] [Google Scholar]
62. Hayat, A., Mueller, J. B. & Capasso, F. Lateral chirality-sorting optical forces. Proc. Natl Acad. Sci.112, 13190–13194 (2015). 10.1073/pnas.1516704112 [Europe PMC free article] [Abstract] [CrossRef] [Google Scholar]
63. Shi, Y. et al. Chirality-assisted lateral momentum transfer for bidirectional enantioselective separation. Light.: Sci. Appl.9, 62 (2020). 10.1038/s41377-020-0293-0 [Europe PMC free article] [Abstract] [CrossRef] [Google Scholar]
64. Li, T. et al. Reversible lateral optical force on phase-gradient metasurfaces for full control of metavehicles. Opt. Lett.48, 255–258 (2023). 10.1364/OL.478979 [Abstract] [CrossRef] [Google Scholar]
65. Wu, Y., Li, J., Zhang, Z.-Q. & Chan, C. T. Effective medium theory for magnetodielectric composites: beyond the long-wavelength limit. Phys. Rev. B74, 085111 (2006).10.1103/PhysRevB.74.085111 [CrossRef] [Google Scholar]
66. Wang, N., Zhang, R.-Y., Chan, C. T. & Wang, G. P. Effective medium theory for a photonic pseudospin-1/2 system. Phys. Rev. B102, 094312 (2020).10.1103/PhysRevB.102.094312 [CrossRef] [Google Scholar]
67. Zhu, T. M. et al. Optical pulling using evanescent mode in sub-wavelength channels. Opt. Express24, 18436–18444 (2016). 10.1364/OE.24.018436 [Abstract] [CrossRef] [Google Scholar]
68. Kemp, B. A. & Sheppard, C. J. Electromagnetic and material contributions to stress. energy, momentum metamaterials, Adv. Electromagnetics6, 11–19 (2017). [Google Scholar]

Articles from Nature Communications are provided here courtesy of Nature Publishing Group

Citations & impact 


This article has not been cited yet.

Impact metrics

Alternative metrics

Altmetric item for https://www.altmetric.com/details/166144740
Altmetric
Discover the attention surrounding your research
https://www.altmetric.com/details/166144740

Similar Articles 


To arrive at the top five similar articles we use a word-weighted algorithm to compare words from the Title and Abstract of each citation.

Funding 


Funders who supported this work.

National Natural Science Foundation of China (National Science Foundation of China) (1)