PHYSICAL REVIEW APPLIED 14, 014026 (2020)

Information-Efficient Metagrating for Transverse-Position Metrology

Zheng Xi,* Sander Konijnenberg,† and H.P. Urbach
Optics Research Group, Delft University of Technology Delft, Netherlands

(Received 23 April 2019; revised 18 February 2020; accepted 20 May 2020; published 9 July 2020)

We introduce an optimal metagrating design for transverse-position metrology in presence of photon

shot noise. The proposed working principle is closely related to the formation of a phase vortex in the
diffraction orders in the parameter space. Using the topological robustness, we optimize the design and
compress all the transverse-position information around a certain point into a small number of detected
photons, saturating the shot-noise limit had all the photons used for probing the position been detected.
The current scheme avoids the problem of detector saturation in the presence of high probe power while
maintaining all the information detected, allowing one to make full use of the high power that is available.
Besides, the direct link between the resonant property in the unit cell and the conditions to achieve the
bound is given: one with the zeroth dipole resonance and the other one with the anapole condition of
the first dipole. The connection between the metagrating design and the optimization using topological
robustness along with the fundamental precision limit using classical light gives new insights in all of
these fields.

DOI: 10.1103/PhysRevApplied.14.014026

I. INTRODUCTION away information about the displacement. Thus an impor-

tant question to be answered is as follows: Can one recover
One particular important research area in optics is the
all the information contained in a large number of probe
development of sensitive metrology methods to measure
photons (Nprobe ) per unit time by detecting a small fixed
a very small displacement around a certain point [1–14].
number of detected photons (Ndet ) per unit time set by the
This problem is fundamental to pushing the limits of fields
limit of the detector? In other words, how can one develop
such as interferometry [7–10], metrology [6], and superres-
an information-efficient measurement that compresses the
olution microscopy [1–4]. Measuring a small displacement
full measured information into a small number of detection
amounts to a parameter-estimation problem in the presence
events?
of measurement noise, which can be due to either imper-
This question has been answered in the development
fections of the setup or, more fundamentally, the intrinsic
of large-scale interferometers [7,15,16], which are used
fluctuations due to photon
 shot noise using coherent light,
to measure extremely small longitudinal displacement. If
which scales as 1/ Nprobe , with Nprobe being the num-
the operating point of the interferometer is set near the so-
ber of photons used to probe the change in position per
called dark fringe of the Michelson interferometer, almost
unit time. One might intuitively think that the best mea-
all the information about the longitudinal displacement is
surement scheme is to detect all Nprobe photons, since no
compressed into a small number of detected photons (Ndet ),
photons and thus no information about the displacement
allowing one to make full use of the very high power
are lost. But this scheme soon runs into a problem when
inside the interferometer to increase the measurement pre-
one tries to push the precision of the measurement by
cision. Moreover, advanced power-recycling techniques
adding more probe photons: while extremely high opti-
allow one to recycle all the undetected light to increase
cal powers are available (e.g., from several hundred watts
the effective number of probe photons, leading to a sig-
to a few kilowatts at the beams splitter inside an inter-
nificant improvement compared with the case that one
ferometer) [7,15,16], most fast photon detectors do not
detects all the photons directly [18–20]. However, as men-
allow the direct exposure of such high powers [17]. Fur-
tioned, this scheme is used mainly for probing longitudinal
thermore, the detectors are more sensitive at relatively low
displacement along the beam propagation direction.
incoming powers. If one attenuates the number of photons
Recently, a promising approach combing knowledge
Nprobe interacting with the measurement device, one throws
from structured light and nanophotonic resonant struc-
tures was proposed for transverse-position metrology [11–
*
[Link]@[Link] 13,21,22]. A distinct feature of resonant nanostructures
† containing balanced multipole resonances is the effect
On leave from Optics Research Group at Delft University of
Technology, now at ASML Research. of directional scattering, known as the Huygens source

2331-7019/20/14(1)/014026(8) 014026-1 © 2020 American Physical Society

XI, KONIJNENBERG, and URBACH PHYS. REV. APPLIED 14, 014026 (2020)

fulfilling Kerker’s condition [23,24]. When combined with (a) (b)

structured light illumination, this directional scattering
condition is highly dependent on the local vectorial field
distributions. If designed properly, a sharply switched scat-
tering directionality can be achieved when the nanostruc-
tures are displaced slightly from the origin of interest. It
was demonstrated recently that the far-field directionality
can be used to extract the transverse location of nanos-
tructures [11–13,21,22]. Because this scheme detects only (c)
parts of the totally scattered light, one would expect loss
of information. Thus, an important question remains unan-
swered: How can one use this scheme to optimize the
illumination and the resonances excited to achieve the
shot-noise limit for transverse-position metrology in an
information-efficient way?
In this work, we propose an optimal resonant metagrat-
ing consisting of meta-atoms with specifically engineered
scattering properties [25–28] as an information-efficient
tool for transverse-position measurement. The position FIG. 1. (a) The metagrating design consisting of a periodic
information of the metagrating around a certain point array of cylindrical nanowires. The incident fields are repre-
inside an illumination field is encoded into the reflected sented by the blue arrows with either TE or TM polarization.
diffraction orders. By writing the field scattered by the For each incident field, there are four diffraction orders, as indi-
unit cell of the metagrating as a superposition of fields cated by the dashed green lines. (b) When the metagrating is
produced by a set of multipoles, we show that at the res- placed inside an interference field formed by two plane waves
onance of these multipoles, the reflected order of interest of initial phase difference π , there are two reflected orders P+
has a phase vortex as a function of certain design param- and P− . Depending on the displacement indicated by the dashed
red circles, P+ and P− disappear at +x0 and −x0 , respectively.
eters. We further use the topological stability of the phase
(c) Normalized power difference as a function of displacement.
vortex to tune the design parameters to maximize the infor- The illumination is TM polarization with φin = 60◦ , R = 70 nm,
mation contained in the detected photons, allowing one to and for this design x0 = 22 nm. The green-shaded region d is the
achieve the shot-noise limit set by Nprobe , with Ndet much linear region within which D(x) can be fitted by the red line.
smaller than Nprobe . Through the physical insights gained,
two special resonant conditions are found to be related to
the fundamental precision limit. The connection between is designed to make P+ and P− vanish when displaced by
the formation of a phase vortex due to the multipole res- +x0 and −x0 , respectively. Because the displacement x0 is
onances of the metagrating and the fundamental precision equivalent to adding a phase shift of 2kx x0 to one of the
limit for position metrology opens new possibilities for the two beams, this requirement can be expressed as
use of metadevices in metrology.
r0 − r1 e−2ikx x0 = 0, (1)

II. WORKING PRINCIPLE OF THE where r0 and r1 are the reflection coefficients of the zeroth
METAGRATING and first orders and the minus sign is due to the initial phase
shift of π . Assuming a total input power of P0 , the powers
Without any loss of generality, silicon nanowire (n =
of the two diffraction orders at position x are
3.5) of radius R with electric and magnetic resonances
is chosen as the constituting element for the metagrating
P+ = 2|r0 |2 P0 sin2 kx (x − x0 ),
[26,29–33]. The metagrating is designed such that it has (2)
four diffraction orders as shown in Fig. 1(a). It is placed P− = 2|r0 |2 P0 sin2 kx (x + x0 ).
inside an interference field produced by two plane waves
with an initial phase shift of π as shown in Fig. 1(b). In Fig. 1(c) we show the normalized power difference
The wavelength used is 650 nm throughout this paper. of the two reflected orders defined by D(x) = (P+ −
The illumination of the metagrating is chosen such that P− )/(P+ + P− ). The value of position x can be inferred
the x component of the incident wave vectors satisfies by measuring D(x) without scanning, provided that x is in
kx = π/P, where P is the pitch of the metagrating. Under the interval [−x0 , x0 ]. For a region d shown in Fig. 1(b),
this condition, the two reflected orders P+ and P− (each D(x) can be linearized with the slope given by D (0) =
produced by a combination of the zeroth and first reflection −2/x0 . A smaller x0 leads to a larger slope and thus a
orders) are parallel to the incident angles. The metagrating higher sensitivity.

014026-2
INFORMATION-EFFICIENT METAGRATING... PHYS. REV. APPLIED 14, 014026 (2020)

III. FISHER INFORMATION AND CRAMÉR-RAO Ndet = [P+ (x) + P− (x)]/ω. One can use Fisher infor-
LOWER BOUND FROM THE DETECTED mation to characterize the bound in the uncertainty of the
PHOTONS retrieved position. By applying the definition of Fisher
information and the chain rule [34], one arrives at
Even in the ideal case of no background noise, the
quantum property of light introduces fluctuations in the  2
1 dNdet
detected signal, which sets a limit on the precision that can Idet (x) = . (5)
Ndet dx
be achieved. This limit is often described using a lower
bound on the measurement uncertainty called the quantum If one is interested in knowing the position x = 0 or
Cramér-Rao lower bound (CRLB), which is the inverse of measuring very small displacements around this point,
the square root of quantum Fisher information. For classi-
cal light, the bound is generally expressed as the shot-noise V2 sin2 (2kx x)cos2 01
limit and is optimized over all the measurement schemes lim Idet (x) = lim 8A0 kx2 Nprobe
x→0 x→0 1 − Vcos(2kx x)cos01
and estimators for use of the coherent state.
For our case, because a displacement x from the origin ≤ 16|r0 |2 kx2 Nprobe , (6)
is equivalent to adding a phase shift of e2ikx x to one of the
incoming beams, the shot-noise limit for position estima- which reaches its maximum when V = 1 and 01 = 0. The
tion can be easily derived from the well-known shot-noise uncertainty bound σ CRLB (0) for position estimation around
limit for phase estimation, which is this point is thus
1

1 σ CRLB (0) = √
σxSNL = σφSNL /2kx = , (3) Idet (0)
2kx Nprobe
1
≥  , (7)
where Nprobe = P /ω is the total number of probe pho- 4|r0 |kx Nprobe
tons used per unit time to probe the position. This limit
serves as a criterion to evaluate the performance of differ- which is dependent on the reflection coefficient of the
ent measurement schemes, in our case different metagrat- grating used. Comparing this with the shot-noise limit in
ing designs. The optimal metagrating design should attain Eq. (3), which is the bound irrespective of the specific grat-
this limit by detecting Ndet photons, with Ndet  Nprobe . ing parameters, we have an additional requirement for the
However, as noted, the shot-noise limit is optimized metagrating design such that |r0 | = 1/2.
over all metagrating designs and is thus independent of Through this detailed analysis of the CRLB for different
the detailed design parameters of the metagrating. To gain metagrating designs and its comparison with the shot-noise
physical insights into the links between the detailed prop- limit, which is the optimal limit that can be achieved for
erties of the metagrating and the shot-noise limit, we next any grating design and any unbiased estimators, we sum-
consider the Fisher information contained in the detected marize the two requirements for the metagrating design:
photons and derive the Cramér-Rao bound from these
detected photons, which is dependent on the properties of (a) V = 1 and 01 = 0. This condition is equivalent to
the metagrating. By comparison of the Cramér-Rao bound the requirement of Eq. (1) with x0  kx by noting 2kx x0 =
in the detected photons and the shot-noise limit, knowl- 01 .
edge of the design requirements for the metagrating can be (b) The reflection coefficient should be |r0 | = 1/2.
gained.
We consider the Fisher information contained in the Once these two conditions are fulfilled, we can say the
reflected photons because we detect only them. We first metagrating is optimal for measurement of very small dis-
release the condition that Eq. (1) imposes and consider placement under photon shot noise in a sense that it attains
a generalized grating with reflection coefficients |r0 | and the shot-noise limit given by Eq. (3). Moreover, accord-
|r1 |e−i01 with a phase difference 01 . In this case, the ing to Eq. (2), as x0 becomes smaller, the detected power
powers contained in the two reflected orders are around x = 0 in P+ and P− is much less than the probe
power P . In the case of very large P , the scheme avoids
P+ (x) = A0 P [1 − V cos(2kx x − 01 )], the problem of detector saturation. Thus, it is information
(4) efficient.
P− (x) = A0 P [1 − V cos(2kx x + 01 )],
IV. DESIGN OF THE METAGRATING
where A0 = (r20 + r21 )/2 and V = |2r0 r1 |/(r20 + r21 ) is the
visibility. The question now is how to design a metagrating that
Under photon shot noise, the number of detected pho- fulfills the above requirements. We use multipole scat-
tons per unit time follows a Poisson distribution with mean tering theory to describe our system as it relates directly

014026-3
XI, KONIJNENBERG, and URBACH PHYS. REV. APPLIED 14, 014026 (2020)

the diffraction properties to the different multipole compo- unit cells. From these, one can write down the expres-
nents, which provides valuable insights into the optimized sions for the reflection and transmission coefficients of
design. diffraction order μ:

A. Multipole scattering theory rμ = Cμ (φμ )G(φμ , φin ),

(11)
The total field in the far field of the grating can be written tμ = δμ0 + Cμ (2π − φμ )G(2π − φμ , φin ).
as
It can be seen that the reflection coefficient of a certain
 order is fully determined by the Green’s function along
U=e ikinx x+ikiny y
+ Cμ (φμ )G(φμ , φin )e ikμx x+ikμy y
, (8)
μ
that direction. We substitute the reflection coefficients into
Eq. (1):
where Cμ (φμ ) = 2/k0 Psinφμ , with P being the pitch of
G(φ0 , φin ) − G(φ1 , φin )e−2ikx x0 = 0, (12)
the grating. The first term corresponds to the incident field
with incident wavevectors kin = (kinx , kiny ) and the second
where φ0 = π − φin and φ1 = φin are the angles of the two
term corresponds to the scattered field from the grating. φμ
reflected orders. For sufficiently small radii, one can keep
is the diffraction angle of order μ, and φin is the incident
only the first three terms A−1 , A0 , and A1 in the expansion
angle. G(φμ , φin ) is the multiple-scattering Green’s func-
of G:
tion along the φμ direction and can be expressed using
multipole decomposition:    
01 01
A0 sin − iA1 cos φin − (13)
+∞
2 2
  
G(φμ , φin ) = Am (φin )eimφμ , (9) 01
− iA−1 cos φin + = 0,
m=−∞ 2

with Am corresponding to the multiple-scattered multipole where 01 = 2kx x0 .

coefficients. They can be constructed from the isolated The above derivation works for both TE and TM polar-
scattering amplitude am for a single nanowire: ization. For TM polarization, the term A0 corresponds to
the multiple-scattered linearly polarized magnetic dipole
 +∞

 in the array. Because of the broken symmetry, the A−1 and
Am (φin ) = am e imφin
+ An (φin )L (n, m) , (10) A1 terms are different when φin = 90◦ . Since their scat-
n=−∞ tered fields carry different topological charges −1 and +1
as shown in the insets in Fig. 2(a), they can be associated
where L (n, m) is the lattice sum, which takes full account with the multiple-scattered left-rotating and right-rotating
of the multiple scattering between nanowires at different electric dipoles. For TE polarization, A0 corresponds to the

(a) (b) (c)

FIG. 2. (a) Amplitude of the first three multipole coefficients for φin = 60◦ . The insets show the phase of the scattered field for each
multipole for TM polarization. The reflection coefficient |r0 | is also shown. A clear resonance can be seen around R = 70 nm, around
which |r0 | ≈ 0.5. (b),(c) Absolute value and phase of the complex reflection amplitude of P+ when the metagrating of different radius
R is placed at different locations inside an interference field. A clear amplitude zero and a phase vortex can be seen at the location
of the design fulfilling Eq. (1) with x0 = 22 nm. The results are obtained with use of multipole scattering theory developed in Refs.
[35,36] and are double-checked by COMSOL MULTIPHYSICS simulation.

014026-4
INFORMATION-EFFICIENT METAGRATING... PHYS. REV. APPLIED 14, 014026 (2020)

multiple-scattered linearly polarized electric dipole, while C. Use of topological robustness to minimize x0
A−1 and A1 correspond to the rotating magnetic dipoles. The results of exploring the above possibilities are sum-
On the left-hand-side of Eq. (12) is the complex ampli- marized in Fig. 3(a). On the horizontal axes, the radius R
tude of reflection order P+ at position x = x0 . To make this and incident angle φin are varied and the absolute value of
order vanish at x0 , there should be a proper balance of the the normalized power difference D(x0 ) is plotted as differ-
different multipole components as indicated by Eq. (13). ent colors at different x0 values. Each red point in Fig. 3(a)
The term i in front of A1 and A−1 indicates a phase dif- corresponds to a good metagrating design with |D(x0 )| ≈
ference of π/2 is required. When one of the multipoles 1. By continuous variation of φin , the good design that cor-
experiences a resonance, the phase around this resonance responds to the formation of the phase vortex is always
changes rapidly, by which this requirement can be fulfilled. satisfied for some other R and x0 values.
There are three distinct curves giving good metagrating
B. Topological vortex in the design space designs, which correspond to the trajectories of phase vor-
To verify this, we show in Fig. 2(a) the behavior of texes. The continuity of the curve confirms the topological
different multipoles as a function of the radius R of the robustness in the presence of a small perturbation. The ori-
nanowire for φin = 60◦ , with the insets showing the phase gins of the three curves can be identified by comparing
map of each multipole component. The incident angle φin them with the position of multipole resonances as indicated
is chosen to be 60◦ as in Fig. 1(c). A clear resonance in A0 in the plot. It can be seen that good designs can also be
is seen at R = 70 nm, which corresponds to the multiple- obtained from resonances with m = −1, 1 for TE polariza-
scattered magnetic dipole resonance. In Figs. 2(b) and 2(c) tion. In this case, it is the resonances from the two rotating
the absolute value and phase of the complex amplitude of dipoles that provide the required π/2 phase shift. A small
the reflection order P+ are shown as a function of posi- change of R gives a large change in x0 for the m = −1, 1
tion x and radius R. At x = 22 nm and R = 70 nm, the branch as follows from the dotted line in Fig. 3(a) (which
amplitude is zero, meaning that for these values Eq. (1) is is not continuous due to insufficient sampling in R). We
satisfied, which is in agreement with the zero of D(x) in therefore focus on the two m = 0 branches.
Fig. 1(c). It is particularly interesting to look at the phase Although an explicit expression for the dependence of
distribution around this zero-amplitude point. Because of x0 on R and φin is complicated, to make x0 small, a smaller
the presence of the multiple-scattered magnetic dipole res- R and a relatively large φin are needed. We restrict φin
onance in A0 , the phase changes very rapidly, forming a to be smaller than 70◦ to represent the occurrence of a
phase vortex of topological charge +1 in Fig. 2(c). It is second reflection order. By imposing this constraint, the
at this phase-vortex point that the metagrating satisfies minimum x0 is determined for each polarization. The nor-
the design requirement. Because such a phase vortex is malized power difference D(x) is plotted in Fig. 3(b) for
topologically stable [37–40], it follows that by continu- both polarizations for the design of the minimum x0 . With
ous variation of parameters such as the incident angle φin , TE polarization, a much smaller x0 (0.08 nm) can be
Eq. (1) is again satisfied for some other R and x0 val- achieved than with TM polarization (15 nm) meaning that
ues. Therefore, we can use this topological robustness to a much higher probe power can be used for TE polariza-
minimize x0 . Besides, different incident polarizations have tion than for TM polarization. It is further confirmed that
different multipole coefficients, which also influence how for both cases a phase vortex occurs around the desired
small x0 can be obtained. design shown in the insets in Fig. 3(b).

(a) (b) (c) (d)

FIG. 3. (a) Achievable x0 for different combination of R and φin for both TE polarization and TM polarization using metagrating
designs. (b) Absolute value of the normalized power difference D(x) for TE and TM polarizations for the optimum design with
minimum x0 . The insets show the formation of a phase vortex around the optimum design. (c) Multipole amplitude strength for TE
and TM polarizations at minimum x0 . (d) Achievable σ CRLB (0) for TE and TM polarizations. Designs with minimum x0 saturate the
shot-noise limit.

014026-5
XI, KONIJNENBERG, and URBACH PHYS. REV. APPLIED 14, 014026 (2020)

D. Fulfilling the requirement for reflection coefficient

|r0 | = 1/2
It remains to be checked if the optimized design given
above also fulfills the second requirement: that is, the
reflection coefficient |r0 | = 1/2.
To see this is indeed the case, we refer to Eq. (11). We
note that the optimized design obtained above works at
A0 resonance. We assume at this resonance other higher-
order Am terms are much smaller. In this case, we can
approximate Eq. (11) using the contribution only from A0 :

r0 ≈ C0 (φ0 )A0 (φin ),

r1 ≈ C1 (φ1 )A0 (φin ) = r0 ,
(14)
t0 ≈ 1 + C0 (2π − φ0 )A0 (φin ) = 1 − r0 , FIG. 4. Attaining the shot-noise limit near the anapole condi-
t1 ≈ C1 (2π − φ1 )A0 (φin ) = −r0 . tion. The inset shows the structure of the unit-cell design.

Furthermore, energy conservation requires that chosen to be nAg = 0.052 225 + 4.4094i and nSiO2 = 1.5.
We keep the incident angle φin = 5◦ and the outer radius
|r0 |2 + |r1 |2 + |t0 |2 + |t1 |2 = 1, (15) RSiO2 = 50 nm fixed and vary the inner radius of the Ag
which leads to |r0 | = 1/2. core. We show in Fig. 4 the multipole coefficient A1 . An
In Fig. 3(d), we plot the ratio of the shot-noise limit anapole excitation is clearly seen at RAg = 25 nm. We fur-
σ SNL (0) given by Eq. (3) and the achievable uncertainty ther plot the ratio of the shot-noise limit and the achievable
bound σ CRLB (0) for the above designs under different uncertainty bound as a function of RAg . It can be seen at
polarizations at φin = 70◦ . Because of the large |A0 /A1 | the exact anapole condition that the uncertainty in retriev-
ratio shown in Fig. 3(c), the shot-noise limit is achieved for ing x = 0 is much larger than the shot-noise limit because
TE polarization at R = 22 nm and approximately achieved the phase requirement is not fulfilled. It is also important
for TM polarizaton at R = 62 nm. The fact that the shot- to have the detected power slightly above the noise caused
noise limit is achieved proves the design is optimal. by other sources, thus minimizing the effect of imperfect
measurement and maximizing the amount of information
detected. If one moves away slightly from the anapole con-
V. SATURATING THE SHOT-NOISE LIMIT NEAR
dition, the A1 coefficient is still very small but the phase
THE ANAPOLE CONDITION
condition is fulfilled. The shot-noise limit is achieved as
Already from Eq. (13) one can gain some knowledge indicated by the optimal point in the plot.
of the optimal design. We assume x0  kx is fulfilled and
A1 ≈ A−1 , and then Eq. (13) can be simplified as VI. POWER RECYCLING
kx x0 ≈ 2iA1 /A0 . (16) Finally we discuss briefly the technique of power recy-
cling and show how is it possible to combine it with our
There are two ways to make x0 small. The first one is to metagrating scheme to increase the measurement preci-
work around the resonance of the zeroth dipole term A0 as sion. This technique was firstly introduced by Drever [18]
shown in the above analysis. The second one is to work for the detection of gravitational waves. The key idea is
around the point where A1 becomes very small. We now basically to recycle all the noninformative photons that are
discuss the second case. not used for detection in the so-called bright port back
Looking at Eq. (10), we note when the scattering coeffi- into the interferometer and send them into a Fabry-Perot
cient a1 becomes zero, the effect of the lattice sum becomes cavity enclosing the intereferometer which can be used
negligible. For the isolated nanowire, a1 = 0 corresponds for probing the phase shift again. In this way, the power
to the excitation of the anapole at which destructive inter- inside the interferometer is boosted by the effective time
ference of the dipole and the toroidal moments happen. At that the photons are trapped inside the cavity, which is on
this condition A−1 = A1 = 0, which makes the ratio A1 /A0 the order of several hundred times due to typical loss of the
zero. However, it does not satisfy the π/2 phase require- cavity [16]; therefore, the number of effective probe pho-
ment of Eq. (13). We therefore expect to see an optimal tons is increased, allowing one to achieve an even-lower
design slightly away from the anapole condition. CRLB permitted by the shot-noise limit than in the case
We redesign our unit cell using Ag-core, SiO2 -shell without power recycling. In our metagrating scheme, since
nanowire as shown in Fig. 4. The refractive indexes are we detect only the reflected photons, which are designed

014026-6
INFORMATION-EFFICIENT METAGRATING... PHYS. REV. APPLIED 14, 014026 (2020)

to have most of the information, the transmitted photons Thomas Adams, Paolo Addesso, and R. X. Adhikari, , et al.,
carry negligible information about the transverse position. Observation of Gravitational Waves from a Binary Black
Those photons in the transmitted orders can then be sent Hole Merger, Phys. Rev. Lett. 116, 061102 (2016).
back to the metagrating again for power recycling, thereby [8] J. R. Gardner, M. L. Marable, G. R. Welch, and J. E.
Thomas, Suboptical Wavelength Position Measurement of
increasing Nprobe to gNprobe , where g is the gain factor intro-
Moving Atoms Using Optical Fields, Phys. Rev. Lett. 70,
duced by the cavity enclosing the metagrating [18–20]. 3404 (1993).
Thus, the sensitivity of the scheme is increased further. [9] David M. Giltner, Roger W. McGowan, and Siu Au
We emphasize this power-recycling scheme can be another Lee, Atom Interferometer Based on Bragg Scattering
major advantage of the current scheme besides efficient from Standing Light Waves, Phys. Rev. Lett. 75, 2638
information compression. (1995).
[10] M. Granata, C. Buy, R. Ward, and M. Barsuglia, Higher-
VII. CONCLUSION Order Laguerre-Gauss Mode Generation and Interferome-
try for Gravitational Wave Detectors, Phys. Rev. Lett. 105,
In summary, we propose an optimal metagrating for 231102 (2010).
high-precision transverse-position metrology under Pois- [11] Martin Neugebauer, Paweł Woźniak, Ankan Bag, Gerd
son shot noise. By exploring the resonance and topological Leuchs, and Peter Banzer, Polarization-controlled direc-
features of the metagrating, we show that the shot-noise tional scattering for nanoscopic position sensing, Nat. Com-
limit can be achieved by detecting only a small number of mun. 7, 11286 (2016).
[12] Zheng Xi Lei Wei, Aurle Joseph Louis Adam, H. P.
photons used for probing the position, making the scheme Urbach, and Luping Du, Accurate Feeding of Nanoantenna
information efficient. It is found that around the zeroth by Singular Optics for Nanoscale Translational and Rota-
dipole resonance and the anapole condition of the first tional Displacement Sensing, Phys. Rev. Lett. 117, 113903
dipole, the shot-noise limit is achieved. We believe our (2016).
metagrating design using topological robustness to achieve [13] Ankan Bag, Martin Neugebauer, Paweł Woźniak, Gerd
the shot-noise limit in an information-efficient way will Leuchs, and Peter Banzer, Transverse Kerker Scattering for
result in new insights in both the field of nanophotonics Angstrom Localization of Nanoparticles, Phys. Rev. Lett.
and the field of nanometrology. 121, 193902 (2018).
[14] Lei Wei, Anatoly V. Zayats, and Francisco J. Rodríguez-
Fortuño, Interferometric Evanescent Wave Excitation of a
ACKNOWLEDGMENT Nanoantenna for Ultrasensitive Displacement and Phase
This work was funded through the EMPIR project Metrology, Phys. Rev. Lett. 121, 193901 (2018).
17FUN01-BeCOMe. The EMPIR initiative is cofunded by [15] Junaid Aasi, J. Abadie, B. P. Abbott, Richard Abbott, T.
D. Abbott, M. R. Abernathy, Carl Adams, Thomas Adams,
the European Union Horizon 2020 research and innovation Paolo Addesso, and R. X. Adhikari, et al., Enhanced sen-
program and the EMPIR participating states. sitivity of the ligo gravitational wave detector by using
squeezed states of light, Nat. Photonics 7, 613 (2013).
[16] J. Abadie, Benjamin P. Abbott, R. Abbott, Thomas D.
Abbott, M. Abernathy, Carl Adams, R. Adhikari, Christoph
[1] Stefan W. Hell and Jan Wichmann, Breaking the diffrac- Affeldt, B. Allen, and G. S. Allen, et al., A gravitational
tion resolution limit by stimulated emission: Stimulated- wave observatory operating beyond the quantum shot-noise
emission-depletion fluorescence microscopy, Opt. Lett. 19, limit, Nat. Phys. 7, 962 (2011).
780 (1994). [17] Hans-Albert Bachor, Timothy C. Ralph, St Lucia, and Tim-
[2] Michael J. Rust, Mark Bates, and Xiaowei Zhuang, Sub- othy C. Ralph, A Guide to Experiments in Quantum Optics
diffraction-limit imaging by stochastic optical reconstruc- (Wiley Online Library, Weinheim, 2004), Vol. 1.
tion microscopy (storm), Nat. Methods 3, 793 (2006). [18] R. W. P. Drever, Gravitational Wave Detectors Using Laser
[3] Eric Betzig, George H. Patterson, Rachid Sougrat, O. Wolf Interferometers and Optical Cavities: Ideas, Principles and
Lindwasser, Scott Olenych, Juan S. Bonifacino, Michael Prospects (Springer, Boston, MA, 1983).
W. Davidson, Jennifer Lippincott-Schwartz, and Harald F. [19] Brian J. Meers, Recycling in laser-interferometric
Hess, Imaging intracellular fluorescent proteins at nanome- gravitational-wave detectors, Phys. Rev. D 38, 2317
ter resolution, Science 313, 1642 (2006). (1988).
[4] Stefan W. Hell, Far-field optical nanoscopy, Science 316, [20] Kevin Lyons, Justin Dressel, Andrew N. Jordan, John C.
1153 (2007). Howell, and Paul G. Kwiat, Power-Recycled Weak-Value-
[5] Norman Bobroff, Recent advances in displacement measur- Based Metrology, Phys. Rev. Lett. 114, 170801 (2015).
ing interferometry, Meas. Sci. Technol. 4, 907 (1993). [21] Martin Neugebauer, Sergey Nechayev, Martin Vorndran,
[6] Arie J. den Boef, Optical wafer metrology sensors for Gerd Leuchs, and Peter Banzer, Weak measurement
process-robust cd and overlay control in semiconductor enhanced spin hall effect of light for particle displacement
device manufacturing, Surface Topography: Metrol. Prop. sensing, Nano Lett. 19, 422 (2018).
4, 023001 (2016). [22] Lei Wei, Anatoly V. Zayats, and Francisco J. Rodríguez-
[7] Benjamin P. Abbott, Richard Abbott, T. D. Abbott, M. R. Fortuño, Interferometric Evanescent Wave Excitation of a
Abernathy, Fausto Acernese, Kendall Ackley, Carl Adams, Nanoantenna for Ultrasensitive Displacement and Phase

014026-7
XI, KONIJNENBERG, and URBACH PHYS. REV. APPLIED 14, 014026 (2020)

Metrology, Phys. Rev. Lett. 121, 193901 [31] Vladislav Popov, Fabrice Boust, and Shah Nawaz Burokur,
(2018). Controlling Diffraction Patterns with Metagratings, Phys.
[23] Arseniy I. Kuznetsov, Andrey E. Miroshnichenko, Mark Rev. Appl. 10, 011002 (2018).
L. Brongersma, Yuri S. Kivshar, and Boris Luk’yanchuk, [32] Alex M. H. Wong and George V. Eleftheriades, Perfect
Optically resonant dielectric nanostructures, Science 354, Anomalous Reflection with a Bipartite Huygens’ Metasur-
2472 (2016). face, Phys. Rev. X 8, 011036 (2018).
[24] Milton Kerker, D.-S. Wang, and C. L. Giles, Electro- [33] V. S. Asadchy, A. Díaz-Rubio, S. N. Tcvetkova, D.-H.
magnetic scattering by magnetic spheres, JOSA 73, 765 Kwon, A. Elsakka, M. Albooyeh, and S. A. Tretyakov,
(1983). Flat Engineered Multichannel Reflectors, Phys. Rev. X 7,
[25] Vladislav Popov, Fabrice Boust, and Shah Nawaz Burokur, 031046 (2017).
Constructing the near Field and far Field with Reactive [34] Steven M. Kay, Fundamentals of Statistical Signal Process-
Metagratings: Study on the Degrees of Freedom, Phys. Rev. ing (Prentice Hall PTR, Upper Saddle River, New Jersey,
Appl. 11, 024074 (2019). 1993).
[26] Younes Ra’di, Dimitrios L. Sounas, and Andrea Alù, Meta- [35] Victor Twersky, On the scatttering of waves by an infi-
gratings: Beyond the Limits of Graded Metasurfaces for nite grating, IRE Trans. Antennas Propagation 4, 330
Wave Front Control, Phys. Rev. Lett. 119, 067404 (2017). (1956).
[27] Zi-Lan Deng, Junhong Deng, Xin Zhuang, Shuai Wang, [36] Victor Twersky, On scattering of waves by the infinite grat-
Tan Shi, Guo Ping Wang, Yao Wang, Jian Xu Yaoyu Cao, ing of circular cylinders, IRE Trans. Antennas Propagation
and Xiaolei Wang, et al., Facile metagrating holograms 10, 737 (1962).
with broadband and extreme angle tolerance, Light: Sci. [37] Bo Zhen, Chia Wei Hsu, Ling Lu, A. Douglas Stone, and
Appl. 7, 78 (2018). Marin Soljačić, Topological Nature of Optical Bound States
[28] Ariel Epstein and Oshri Rabinovich, Unveiling the Prop- in the Continuum, Phys. Rev. Lett. 113, 257401 (2014).
erties of Metagratings via a Detailed Analytical Model for [38] Ling Lu, John D. Joannopoulos, and Marin Soljacic, Topo-
Synthesis and Analysis, Phys. Rev. Appl. 8, 054037 (2017). logical photonics, Nat. Photonics 8, 821 (2014).
[29] Jon A. Schuller, Rashid Zia, Thomas Taubner, and Mark [39] Chia Wei Hsu, Bo Zhen, Jeongwon Lee, Song-Liang Chua,
L. Brongersma, Dielectric Metamaterials Based on Elec- Steven G. Johnson, John D. Joannopoulos, and Marin Sol-
tric and Magnetic Resonances of Silicon Carbide Particles, jacic, Observation of trapped light within the radiation
Phys. Rev. Lett. 99, 107401 (2007). continuum, Nature 499, 188 (2013).
[30] Wei Liu, Generalized Magnetic Mirrors, Phys. Rev. Lett. [40] M. Z. Hasan and C. L. Kane, Colloquium: Topological
119, 123902 (2017). insulators, Rev. Mod. Phys. 82, 3045 (2010).

014026-8