This action might not be possible to undo. Are you sure you want to continue?

BooksAudiobooksComicsSheet Music### Categories

### Categories

Scribd Selects Books

Hand-picked favorites from

our editors

our editors

Scribd Selects Audiobooks

Hand-picked favorites from

our editors

our editors

Scribd Selects Comics

Hand-picked favorites from

our editors

our editors

Scribd Selects Sheet Music

Hand-picked favorites from

our editors

our editors

Top Books

What's trending, bestsellers,

award-winners & more

award-winners & more

Top Audiobooks

What's trending, bestsellers,

award-winners & more

award-winners & more

Top Comics

What's trending, bestsellers,

award-winners & more

award-winners & more

Top Sheet Music

What's trending, bestsellers,

award-winners & more

award-winners & more

P. 1

Weak Measurements of Light Chirality With a Plasmonic Slit, Bliokh 2012|Views: 3|Likes: 0

Published by Shay Maayani

See more

See less

https://www.scribd.com/doc/104386579/Weak-Measurements-of-Light-Chirality-With-a-Plasmonic-Slit-Bliokh-2012

08/30/2012

text

original

r

X

i

v

:

1

2

0

4

.

0

3

7

8

v

3

[

p

h

y

s

i

c

s

.

o

p

t

i

c

s

]

9

J

u

l

2

0

1

2

Weak Measurements of Light Chirality with a Plasmonic Slit

Y. Gorodetski,

1

K. Y. Bliokh,

2, 3

B. Stein,

1

C. Genet,

1

N. Shitrit,

4

V. Kleiner,

4

E. Hasman,

4

and T. W. Ebbesen

1

1

ISIS, Universit´e de Strasbourg and CNRS (UMR 7006),

8 all´ee Gaspard Monge, 67000 Strasbourg, France

2

Advanced Science Institute, RIKEN, Wako-shi, Saitama 351-0198, Japan

3

A. Usikov Institute of Radiophysics and Electronics, NASU, Kharkov 61085, Ukraine

4

Micro and Nanooptics Laboratory, Faculty of Mechanical Engineering and Russell Berrie Nanotechnology Institute,

Technion–Israel Institute of Technology, Haifa 32000, Israel

We examine, both experimentally and theoretically, an interaction of tightly focused polarized

light with a slit on a metal surface supporting plasmon-polariton modes. Remarkably, this simple

system can be highly sensitive to the polarization of the incident light and oﬀers a perfect quantum-

weak-measurement tool with a built-in post-selection in the plasmon-polariton mode. We observe

the plasmonic spin Hall eﬀect in both coordinate and momentum spaces which is interpreted as

weak measurements of the helicity of light with real and imaginary weak values determined by the

input polarization. Our experiment combines advantages of (i) quantum weak measurements, (ii)

near-ﬁeld plasmonic systems, and (iii) high-numerical aperture microscopy in employing spin-orbit

interaction of light and probing light chirality.

PACS numbers: 42.25.Ja, 73.20.Mf, 42.25.Gy, 42.50.Tx

Introduction.— Polarization-dependent transverse

shifts of spatially-conﬁned optical beams, also known

as the spin Hall eﬀect of light (SHEL), has become a

topic of an intensive research since pioneering studies

by Fedorov and Imbert [1, 2] and other important early

works [3–5]. The SHEL manifests itself in opposite

out-of-plane displacements of the trajectories of right-

and left-hand circularly polarized beams reﬂected or

refracted by a plane interface. Fundamentally, this

subwavelength phenomenon stems from a spin-orbit

interaction (SOI) of light, i.e., a weak coupling of photon

spin (helicity or chirality) and the trajectory of light

propagation [6–8]. During the past few years, interest in

spin-dependent transverse shifts has grown intensively

[8–13] (for a review, see [14]), motivated by the rapid

development of spintronics and nano-optics. After

50 years of highly controversial studies, an accurate

theoretical description of the SHEL at a plane dielectric

interface was formulated [9, 10] and completely veriﬁed

in a remarkable experiment by Hosten and Kwiat [11]

(see also [12–14]).

Importantly, the experiments of [11, 13] achieved in-

credible angstrom accuracy in determination of the SHEL

shift owing to the method of quantum weak measurements

[15–17] (for reviews, see [18]). It was shown that purely

classical interaction of a transversely-conﬁned polarized

optical beam with a plane interface can be interpreted as

a quantum weak measurement of the photon spin (helic-

ity) by the transverse proﬁle of the beam, which is de-

scribed by the optical SOI Hamiltonian. Owing to this,

employing almost orthogonal polarizers before and after

the interface (i.e., pre-selection and post-selection of the

spin states), one can enormously magnify the observed

beam shift from the subwavelength to beam-width scale.

This measured shift represents the actual SHEL shift

multiplied by the weak value of photon helicity which

can take large complex values. In this manner, real weak

values correspond to spatial displacements of the beam,

whereas imaginary weak values correspond to angular de-

ﬂections of the beam (i.e., shifts in the momentum space)

[10–14]. So far, weak-measurement SHEL experiments

used only imaginary weak values and angular shifts be-

cause they result in much higher beam shifts in the far-

ﬁeld.

Alongside classical-optics far-ﬁeld systems, the sub-

wavelength nature of the SHEL makes it highly relevant

and attractive for near-ﬁeld optics [19], high-numerical-

aperture (NA) microscopy [20], and, particularly, plas-

monics [21, 22]. In these areas, the optical SOIs dramat-

ically modify distributions of near ﬁelds and oﬀer promis-

ing applications. In the present Letter, we combine the

fundamental advantages of (i) quantum weak measure-

ments, (ii) near-ﬁeld plasmonic system, and (iii) high-NA

microscopy. We show that coupling of a tightly focused

optical beam to surface plasmon polaritons (SPPs) oﬀers

a natural weak-measurement tool with a built-in post-

selection provided by the ﬁxed linear polarization of the

SPP mode [23]. We use slightly tilted linear and slightly

elliptical input polarizations of light which provide both

imaginary and real weak values of spin and measure both

spatial and angular transverse shifts in the SPP beam

launched by a single slit. These measurements are per-

formed on a leakage radiation microscope [24] which al-

lows visualization of a plasmonic beam in real and mo-

mentum (Fourier) spaces.

Experiment and weak-measurement model.— The ex-

perimental setup is schematically shown in Figure 1. We

used a sample consisted of a glass wafer, coated with

a thin layer of gold (the thickness is about 70nm). A

straight 100nm-wide straight slit was milled in the metal

2

using focused ion beam. The slit was illuminated by a fo-

cused laser beam (785nm) prepared using objectives with

the numerical apertures NA = 0.45 and NA = 0.6. Upon

interaction with the slit, the incident optical beam is par-

tially scattered into two SPP beams propagating along

the gold layer orthogonally to the slit (Fig. 1). The SPP

beams were observed using leakage signal collected via

an immersion objective attached to the back side of the

sample. More detailed description of the standard leak-

age radiation microscope setup can be found elsewhere

[24]. Note that the incident beam was focused to a plane

behind the gold layer, so that the actual focal spot occurs

in the secondary, SPP beams (see Fig. 2). Using a po-

larizer with rotating quarter-wave and half-wave plates

at the input of the system, we were able to produce an

arbitrary polarization state of the incident light (Fig. 1).

Remarkably, we observed extraordinary asymmetric

deformations in the SPP beams when the input polar-

ization of light was just slightly oﬀ from being parallel to

the slit, as shown in Fig. 2. For instance, a tiny rotation

of the quarter-wave plate (producing a slightly elliptical

polarization) caused a strong angular deviation of the

beams, whereas a tiny rotation of the half-wave plate

(slightly tilted linear polarization) resulted in a well pro-

nounced spatial displacement of the focal spot. These

anomalous SPP beam shifts represent plasmonic SHEL

and can be associated with “quantum weak measure-

ments” [15–18] of the incident light helicity via spin-orbit

coupling induced by the light-to-SPP transformation at

the slit. First, we interpret the results within a simple

“quantum weak measurements” model of the SOI of light,

and afterwards will give a complete wave description of

the problem.

Let the incident light propagate along the z-axis (z = 0

represents the gold surface), the slit be parallel to the

y-axis, whereas the SPP beams propagate in the ±x di-

rections (from now on we consider only the +x beam),

Fig. 1. According to the “weak measurements” formal-

ism, the external transverse spatial proﬁle of the beam,

Φ(y), and its internal polarization state, |Ψ, represent

“classical measuring subsystem” and “quantum mea-

sured subsystem”, respectively [16]. For simplicity, let

the y-distribution of the incident light be Gaussian in

the focal plane:

Φ

in

∝ exp

_

−y

2

/w

2

0

¸

, (1)

where w

0

∼ (kNA)

−1

is the beam waist and k is the

wavenumber of light. At the same time, using the ba-

sis of linear polarizations |X and |Y along the corre-

sponding axes, and the spin basis of right- and left-hand

circular polarizations, |R and |L, the pre-selected input

polarization of light is chosen to be almost y-linear:

|Ψ

in

≃ |Y −ε |X =

(−i −ε) |R + (i −ε) |L

√

2

, (2)

paraxial light

focused light

polarization

pre-selection

high-NA

objective

glass

SPP beam

gold

leakage radiation

microscopy

slit

x

y

z

FIG. 1: (color online). Conceptual scheme of the experimen-

tal setup. The objective is used to focus the incident light

with pre-selected polarization on a thin layer of gold with a

straight slit along the y axis. The slit generates SPP beams

propagating in the ±x directions, and the leakage SPP signal

is collected by an immersion objective.

Here ε (|ε| ≪ 1) is a complex number, with real and

imaginary ε corresponding to the slightly tilted linear and

slightly elliptical polarizations, respectively (see Fig. 2).

In the spin basis, states |R = (1, 0)

T

and |L = (0, 1)

T

are the eigenvectors of the photon helicity operator ˆ σ

3

=

diag (1, −1) [14].

Interaction of light with the slit and transformation

to SPPs is similar to the beam refraction at a plane in-

terface, and the geometric-phase diﬀerence between the

constituent plane-wave components produce the SOI of

light [11, 14, 25]. One can show [21] that the geometric-

phase factor for the generated SPP waves with diﬀerent

k-vectors is exp(iˆ σ

3

k

y

/k) for k

y

≪k (cf. [11, 14]), which

implies the dimensionless SOI Hamiltonian

ˆ

H

SOI

≃ −λˆ σ

3

k

y

, (3)

where λ = k

−1

is the SHEL shift playing role of the cou-

pling constant [11]. Employing the weak-measurement

interpretation [11, 16], the transverse proﬁle of the beam,

Gaussian-distributed in y and k

y

“weakly measures” its

helicity ˆ σ

3

multiplied by the SHEL constant λ at the

moment of interaction with the interface (slit).

Since the y-component of the incident electric wave

ﬁeld cannot interact with plasmons via the slit, this nat-

urally deﬁnes the post-selected polarization state to be

perpendicular to the slit:

|Ψ

out

= |X =

|L +|R

√

2

. (4)

In fact, this |X state in the local coordinate frame at-

tached to the direction of propagation of the beam corre-

3

sponds to the z-component of the SPP beam propagating

along the x-axis.

In terms of weak measurements, the input and output

polarization states (2) and (4) determine the weak value

σ

w

of the photon helicity [11, 15–18]:

σ

w

=

Ψ

out

| ˆ σ

3

| Ψ

in

Ψ

out

| Ψ

in

=

i

ε

. (5)

Remarkably, this weak value is complex and large, |σ

w

| ≫

1, although the photon helicity eigenvalues are σ = ±1.

It is seen from Eq. (5) that elliptical and linear tilted

pre-selected polarizations yield real and imaginary σ

w

,

respectively. Weak measurement of helicity, Eq. (5),

results in the transverse shift of the “measuring subsys-

tem”, i.e., of the transverse beam proﬁle [16]:

Φ

out

∝ exp

_

−(y −∆)

2

/w

2

0

_

, ∆ ≃ −λσ

w

. (6)

Thus, the output SPP beam undergoes complex trans-

verse shift (6) equaling to the SHEL shift λ multiplied

by the helicity weak value (5). In this manner, real and

imaginary parts of σ

w

produce coordinate shift (displace-

ment) and momentum shift (deﬂection) of the beam pro-

ﬁle [10–14, 17]:

y = Re∆ , k

y

= 2w

−2

0

Im∆ . (7)

Note that the weak-measurement approximation fails in

the vicinity of ε = 0 and is applicable at λ/w

0

≪ ε ≪ 1

[16].

Compare now the above weak-measurement model,

based on the SOI of light, with the experimental plas-

monic results presented in Fig. 2. Both coordinate and

momentum shifts are clearly visible in the SPP ﬁelds

in real (a,b) and Fourier (c,d) spaces for two types of

the pre-selected polarization. However, the observed co-

ordinate and momentum shifts of the SPP beams turn

out to be swapped as compared to the model (5)–(7):

Real ε (tilted linear polarization) causes coordinate shift

y = 0, whereas imaginary ε (elliptical polarization) in-

duces momentum shift k

y

= 0. As we show below, the

accurate description of the light-SPP coupling appends

an additional π/2 phase factor that interchanges coor-

dinate and momentum shifts with respect to the real or

imaginary nature of ε. These swapped relations with re-

spect to usual SHEL shifts have to be understood as an

inherent feature of the plasmonic SHEL.

Complete wave theory.— The detailed wave picture of

light evolution in the system can be divided into three

stages: (i) focusing of the initial paraxial polarized ﬁeld

by a high-NA objective; (ii) interaction of the focused

ﬁeld with the slit; and (iii) generation and propagation

of the SPP beams.

First, approximating the incident light by a single

plane wave with the complex electric-ﬁeld amplitude

E

0

∝ (−ε, 1, 0)

T

(in the Cartesian basis), the high-NA fo-

cusing is described by the Debye-Wolf approach [20, 26].

10 µm

xx

yy

(a) (b)

(c) (d)

x

kk

x

y

k

y

0.07i ε = −

x

yyy

x

kk

x

y

k

y

0.07 ε =

y

FIG. 2: (color online). Typical SPP ﬁeld distributions in real

and Fourier spaces (see also [29]). The mean momenta ky

and transverse coordinates y of the beams are schematically

indicated by orange and green arrows, respectively, whereas

pre-selected polarizations of the incident light are shown in

cyan. In the case of elliptical polarization (imaginary ε), pan-

els (a) and (c), the SPP beams demonstrate angular SHEL

deviation which is seen as the momentum shift in the Fourier

space. For tilted linear polarization (real ε), panels (b) and

(d), the SPP beams exhibit coordinate SHEL shift which is

seen only in real space.

It implies that the transverse electric ﬁeld of the wave

is parallel-transported along each geometrical-optics ray,

refracted by the lens, without change of the polarization

state in the local coordinates. Marking the rays by spher-

ical angles (θ, φ) indicating the direction of the wave vec-

tors k = k (cos θ, sin θ cos φ, sin θ sin φ), the plane-wave

spectrum of the focused ﬁeld,

˜

E

lens

(k) ≡

˜

E

lens

(θ, φ), is

given by geometric rotational transformation [20]:

˜

E

lens

(θ, φ) ∝

√

cos θ

ˆ

T

lens

(θ, φ) E

0

. (8)

Here

ˆ

T

lens

=

ˆ

R

z

(−φ)

ˆ

R

y

(−θ)

ˆ

R

z

(φ), with

ˆ

R

a

(α) denot-

ing the SO(3) matrix operator of rotation about the a-

axis by the angle α, and

√

cos θ is the apodization factor

which provides conservation of the energy ﬂow [26]. The

real-space focused ﬁeld is given by the Fourier-type inte-

gral over all plane waves

˜

E

lens

(k), but the coupling with

the SPPs is described in the momentum representation,

and we perform this integration afterwards.

Second, the focused ﬁeld (8) interacts with the slit. In

doing so, only x and z components of the ﬁeld can excite

SPPs, whereas the y component, parallel to the slit, does

not take part in the interaction. In other words, the slit

acts as a polarizer which cuts out the y component of

the ﬁeld. This is described by the projection of the ﬁeld

˜

E

lens

onto the (x, z) plane:

˜

E

slit

=

ˆ

P

y

˜

E

lens

, (9)

where

ˆ

P

y

= diag (1, 0, 1) is the projection operator.

Finally, the ﬁeld

˜

E

slit

can be considered as the source

for the SPPs. Excited SPP ﬁeld can be decomposed into

4

plane waves

˜

E

p

which are characterized by the real wave

vectors k

p

=

_

k

px

, k

py

, 0

_

(dissipation is neglected here-

after), exponential decay away from the metal surface,

exp (κz) at z < 0, and the complex electric-ﬁeld ampli-

tudes

˜

E

p

=

˜

E

p⊥

e

z

+

˜

E

p

k

p

/k

p

. From Maxwell equations

it follows that the longitudinal and transverse ﬁeld com-

ponents are related as

˜

E

p

= −iχ

˜

E

p⊥

, χ = κ/k

p

[27].

Excitation of the SPPs by focused light via the slit is

determined by: (i) the phase matching condition that

provides transformation of the wave momenta k → k

p

and (ii) the coupling eﬃciency, γ, which we model by

the inner product of the electric-ﬁeld amplitudes of the

incident light and SPPs: γ ∝

˜

E

∗

p

·

˜

E

slit

[28]. Since the sys-

tem is homogeneous in the y-direction, the corresponding

momentum component must be conserved. Taking into

account that the SPP wave number k

p

> k is ﬁxed (which

determines the SPP circle in the Fourier space, Fig. 2),

the phase matching condition can be written as

k

py

= k

y

, k

px

=

_

k

2

p

−k

2

y

, (10)

where k

y

= k sin θ sin φ. From here and equations above,

the SPP plane-wave amplitude is

˜

E

p

∝

_

_

−iχ

¸

1 −

_

k

y

k

p

_

2

, −iχ

k

y

k

p

, 1

_

_

T

. (11)

Taking into account the coupling eﬃciency γ, the result-

ing SPP ﬁeld in the momentum and coordinate represen-

tations reads

˜

E

p

∝

_

˜

E

∗

p

·

˜

E

slit

_

˜

E

p

, E

p

∝

2π

_

0

θc

_

0

˜

E

p

e

ikp·r

sinθdθdφ. (12)

Here the second Eq. (12) is the Fourier (Debye) integral

over all incoming plane waves, θ

c

= sin

−1

(NA) is the

aperture angle of the focusing objective, and we consider

only the (x, y)-distribution of the SPP ﬁeld, omitting the

common z-dependence.

Equations (8)–(12) completely describe our system

starting from the incident paraxial ﬁeld E

0

and yielding

the output SPP ﬁeld distribution E

p

on the gold layer.

We have veriﬁed numerically that these equations yield

SPP distributions and shifts corresponding to experimen-

tal Figure 2 [29].

In fact, for our experiment one can use the ﬁrst post-

paraxial approximation, sin θ ≃ θ and cos θ ≃ 1, which

signiﬁcantly simpliﬁes equations and allows analytical so-

lution. Keeping only terms linear in θ and ε, but neglect-

ing εθ-order terms, Eqs. (8)–(12) yield

˜

E

slit

∝

_

−ε, 0, −

k

y

k

_

T

,

˜

E

p

∝

_

−iχ, −iχ

k

y

k

p

, 1

_

T

, (13)

Reε

−3

−2

−1

0

1

2

3

Theory

Exper.

FDTD

k

y

−3 −2 −1 0 1 2 3

(a)

Imε

−1 −0.5 0 0.5 1

−0.2

−0.1

0

0.1

0.2

Theory

Exper.

FDTD

/

y

k

k

(b)

FIG. 3: (color online). Theoretically calculated (Eqs. (15)

with χ = 1 [28]) and experimentally measured SHEL shifts of

the SPP beams versus complex pre-selection parameter ε (see

also [29]). (a) – imaginary ε generating the momentum shift

ky and (b) – real ε producing the coordinate shift y. For

experimental convenience, the real space shifts were measured

with NA=0.45 and Fourier-space shifts with NA=0.6.

where k

y

/k ≃ θ sin φ. This results in the following cou-

pling coeﬃcient and the SPP ﬁeld:

γ ≃ −iχε −

k

y

k

,

˜

E

p

∝

_

−iχε −

k

y

k

_

(−iχ, 0, 1)

T

. (14)

It is seen from here that the imaginary longitudinal SPP

ﬁeld plays crucial role in the coupling with light. In-

deed, the weak component of the input polarization,

ε, is multiplied by −iχ = −iκ/k

p

because of the cou-

pling between the x-components of light and SPP ﬁelds.

Therefore, imaginary transverse wave number of SPPs,

k

pz

= −iκ, eﬀectively swaps the real and imaginary parts

of ε in the coupling process. Hence, in order to make

the weak-measurement formalism (1)–(7) consistent with

our plasmonic system, one has to substitute ε → −iχε

or

ˆ

H

SOI

→ iχ

−1

ˆ

H

SOI

in Eqs. (3)–(7). This modiﬁca-

tion immediately ascertains perfect agreement between

the weak-measurement model and experimental results

in Fig. 2.

To compare experiment and theory quantitatively, we

calculate the SPP beam centroids in the coordinate and

momentum spaces. Taking into account that the posi-

tion operator is ˆ y = i∂/∂k

y

in the momentum repre-

sentation, we deﬁne y =

_

˜

E

p

¸

¸

¸ ˆ y

¸

¸

¸

˜

E

p

___

˜

E

p

¸

¸

¸

˜

E

p

_

and

k

y

=

_

˜

E

p

¸

¸

¸ k

y

¸

¸

¸

˜

E

p

___

˜

E

p

¸

¸

¸

˜

E

p

_

, where the inner prod-

uct implies the scalar product of the complex vector am-

plitudes and the integration in the momentum space:

_

˜

E

p

¸

¸

¸

˜

E

p

_

∝

2π

_

0

θc

_

0

_

˜

E

∗

p

·

˜

E

p

_

θdθdφ. Performing these cal-

culations with the SPP Fourier spectrum (14), we ﬁnally

arrive at

y=

1

kθ

c

2χθ

c

Reε

2χ

2

|ε|

2

+θ

2

c

/2

, k

y

=−kθ

c

χθ

c

Imε

2χ

2

|ε|

2

+θ

2

c

/2

.

(15)

5

We emphasize that Eqs. (15) are valid in the whole range

of values of ε. In the weak-measurement range λ/w

0

≪

ε ≪ 1, they are precisely equivalent to Eqs. (7) with

modiﬁcation ε → −iχε, i.e., ∆ → iχ

−1

∆ = (χkε)

−1

,

and w

0

= 2 (kθ

c

)

−1

. Comparisons of the experimentally

measured coordinate and momentum transverse shifts

(as dependent on the complex polarization parameter ε)

with the theoretical results (15) are shown in Figure 3.

Moreover, we performed a ﬁnite diﬀerence time domain

(FDTD) simulations for two input polarizations, and the

resulting shifts are also presented in Fig. 3. Evidently, the

experiment, wave theory, weak-measurement interpreta-

tion, and FDTD simulations are all in perfect agreement

[29].

Conclusion.— We have observed and examined in de-

tail an extraordinary plasmonic SHEL appearing in the

interaction of focused light with a straight slit on the

metal surface. Remarkably, this simple system oﬀers a

perfect weak-measurement tool where ﬁxed polarization

of SPPs provides a built-in post-selection. Tiny varia-

tions of the input polarization of light bring about huge

SHEL shifts of the SPP beams in coordinate and momen-

tum spaces, which correspond to the imaginary and real

parts of the weak value of the helicity of light. The pre-

sented results demonstrate the unique ability of surface

waves to perform as a post-selecting measuring device

which might be potentially useful for various sensing ap-

plications involving the chirality of light.

We acknowledge fruitful discussions with Y. P. Bliokh

and A. Y. Nikitin, and the ﬁnancial support from the

ERC (Grant 227577) and the European Commission

(Marie Curie Action).

[1] F. I. Fedorov, Dokl. Akad. Nauk SSSR 105, 465 (1955).

[2] C. Imbert, Phys. Rev. D 5, 787 (1972).

[3] H. Schilling, Ann. Phys. (Berlin) 16, 122 (1965); J. Ri-

card, Nouv. Rev. Opt. 5, 7 (1974); J. P. Hugonin and R.

Petit, J. Opt. 8, 73 (1977).

[4] M. A. Player, J. Phys. A: Math. Gen. 20, 3667 (1987); V.

G. Fedoseyev, J. Phys. A: Math. Gen. 21, 2045 (1988).

[5] L. Dutriaux, A. Le Floch, and F. Bretenaker, Europhys.

Lett. 24, 345 (1993); F. Pillon, H. Gilles, and S. Girard,

Appl. Opt. 43, 1863 (2004).

[6] V. S. Liberman and B. Y. Zel’dovich, Phys. Rev. A 46,

5199 (1992).

[7] K. Y. Bliokh and Y. P. Bliokh, Phys. Lett. A 333, 181

(2004); K. Y. Bliokh, A. Niv, V. Kleiner, and E. Hasman,

Nature Photon. 2, 748 (2008).

[8] M. Onoda, S. Murakami, and N. Nagaosa, Phys. Rev.

Lett. 93, 083901 (2004).

[9] K. Y. Bliokh and Y. P. Bliokh, Phys. Rev. Lett. 96,

073903 (2006).

[10] K. Y. Bliokh and Y. P. Bliokh, Phys. Rev. E 75, 066609

(2007).

[11] O. Hosten and P. Kwiat, Science 319, 787 (2008).

[12] A. Aiello and J. P. Woerdman, Opt. Lett. 33, 1437

(2008); A. Aiello, M. Merano, and J. P. Woerdman, Phys.

Rev. A 80, 061801(R) (2009); M. R. Dennis and J. B.

G¨otte, arXiv:1204.0327v1.

[13] Y. Qin, Y. Li, H. He, and Q. Gong, Opt. Lett. 34, 2551

(2009); Y. Qin, Y. Li, X. Feng, Y.-F. Xiao, H. Yang, and

Q. Gong, Opt. Express 19, 9636 (2011).

[14] K. Y. Bliokh, A. Aiello, and M. A. Alonso, in The Angu-

lar Momentum of Light, edited by D. L. Andrews and M.

Babiker (Cambridge University Press, 2012, to appear).

[15] Y. Aharonov, D. Z. Albert, and L. Vaidman, Phys. Rev.

Lett. 60, 1351 (1988).

[16] I. M. Duck, P. M. Stevenson, and E. C. G. Sudarshan,

Phys. Rev. D 40, 2112 (1989).

[17] R. Jozsa, Phys. Rev. A 76, 044103 (2007).

[18] Y. Aharonov, S. Popescu, and J. Tollaksen, Physics To-

day 63, 27 (2010); A.G. Kofman, S. Ashhab, and F. Nori,

arXiv:1109.6315; Y. Shikano, arXiv:1110.5055v3.

[19] D. Haefner, S. Sukhov, and A. Dogariu, Phys. Rev. Lett.

102, 123903 (2009); Y. Zhao, J. S. Edgar, G. D. M. Jef-

fries, D. McGloin, and D. T. Chiu, Phys. Rev. Lett. 99,

073901 (2007).

[20] O. G. Rodr´ıguez-Herrera, D. Lara, K. Y. Bliokh, E.

A. Ostrovskaya, and C. Dainty, Phys. Rev. Lett. 104,

253601 (2010); K. Y. Bliokh, E. A. Ostrovskaya, Miguel

A. Alonso, O. G. Rodr´ıguez-Herrera, D. Lara, and C.

Dainty, Opt. Express 19, 26132 (2011).

[21] K. Y. Bliokh, Y. Gorodetski, V. Kleiner, and E. Hasman,

Phys. Rev. Lett. 101, 030404 (2008); Y. Gorodetski, A.

Niv, V. Kleiner, and E. Hasman, Phys. Rev. Lett. 101,

043903 (2008).

[22] N. Shitrit, I. Bretner, Y. Gorodetski, V. Kleiner, and E.

Hasman, Nano Lett. 11, 2068 (2011).

[23] W. L. Barnes, W. A. Murray, J. Dintinger, E. Devaux,

and T.W. Ebbesen, Phys. Rev. Lett. 92, 107401 (2004).

[24] A. Drezet, A. Hohenau, D. Koller, A. Stepanov, H. Ditl-

bacher, B. Steinberger, F. R. Aussenegg, A. Leitner, and

J. R. KrennGordon, J. MSEB 149, 220 (2008).

[25] High-NA focusing also induces SOI and partial spin-to-

orbital angular momentum conversion in the incident

light [19, 20]. However, eﬃciency of this conversion is

estimated as NA

2

/4 ≃ 0.05 here and can be neglected

against the strong SOI (∝ NA) induced by the slit.

[26] B. Richards and E. Wolf, Proc. Soc. London A 253, 358

(1959); P. T¨or¨ok, P. D. Higdon, and T. Wilson, Opt.

Commun. 148, 300 (1998).

[27] F. Yang, J. R. Sambles, and G. W. Bradberry, Phys. Rev.

B 44, 5855 (1991).

[28] This coupling coeﬃcient γ implies equal eﬃciency of cou-

plings for the x and z components of the ﬁelds. In general,

this is not so because of the strong anisotropy of the sys-

tem in the (x, z) plane. The coupling anisotropy can be

taken into account by modifying coeﬃcient χ relating the

longitudinal and transverse SPP ﬁelds. Therefore, we re-

gard χ as a free parameter, which is chosen in Fig. 3 to

ﬁt the experimental data.

[29] Details of the centroid calculations and measurements,

as well as experimental, theoretical, and FDTD distribu-

tions of the SPP ﬁelds at diﬀerent input polarizations are

given in Supplementary Material.

2). SPP beams (see Fig. For simplicity. Conceptual scheme of the experimental setup. The SPP beams were observed using leakage signal collected via an immersion objective attached to the back side of the sample. 2 (4) In fact. we observed extraordinary asymmetric deformations in the SPP beams when the input polarization of light was just slightly oﬀ from being parallel to the slit. At the same time. Since the y-component of the incident electric wave ﬁeld cannot interact with plasmons via the slit. According to the “weak measurements” formalism. Upon interaction with the slit. First. Note that the incident beam was focused to a plane behind the gold layer. whereas a tiny rotation of the half-wave plate (slightly tilted linear polarization) resulted in a well pronounced spatial displacement of the focal spot. These anomalous SPP beam shifts represent plasmonic SHEL and can be associated with “quantum weak measurements” [15–18] of the incident light helicity via spin-orbit coupling induced by the light-to-SPP transformation at the slit. −1) [14]. 14]). Employing the weak-measurement interpretation [11.and left-hand circular polarizations. a tiny rotation of the quarter-wave plate (producing a slightly elliptical polarization) caused a strong angular deviation of the beams. 1). and the spin basis of right. −1 polarization pre-selection high-NA objective paraxial light focused light slit x gold glass leakage radiation microscopy SPP beam z y FIG. Gaussian-distributed in y and ky “weakly measures” its helicity σ3 multiplied by the SHEL constant λ at the ˆ moment of interaction with the interface (slit). the pre-selected input polarization of light is chosen to be almost y-linear : |Ψin ≃ |Y − ε |X = (−i − ε) |R + (i − ε) |L √ . 16]. represent “classical measuring subsystem” and “quantum measured subsystem”. T T In the spin basis. Here ε (|ε| ≪ 1) is a complex number. 2. More detailed description of the standard leakage radiation microscope setup can be found elsewhere [24]. The objective is used to focus the incident light with pre-selected polarization on a thin layer of gold with a straight slit along the y axis. 1). The slit generates SPP beams propagating in the ±x directions. Using a polarizer with rotating quarter-wave and half-wave plates at the input of the system. states |R = (1. we interpret the results within a simple “quantum weak measurements” model of the SOI of light. which σ implies the dimensionless SOI Hamiltonian ˆ σ HSOI ≃ −λˆ3 ky . (2) 2 where λ = k −1 is the SHEL shift playing role of the coupling constant [11]. with real and imaginary ε corresponding to the slightly tilted linear and slightly elliptical polarizations. respectively (see Fig. |Ψ . and its internal polarization state. the transverse proﬁle of the beam.2 using focused ion beam. 25]. 1: (color online). One can show [21] that the geometricphase factor for the generated SPP waves with diﬀerent k-vectors is exp (iˆ3 ky /k) for ky ≪ k (cf. and afterwards will give a complete wave description of the problem. 1. Remarkably.45 and NA = 0. 1) are the eigenvectors of the photon helicity operator σ3 = ˆ diag (1. we were able to produce an arbitrary polarization state of the incident light (Fig. The slit was illuminated by a focused laser beam (785nm) prepared using objectives with the numerical apertures NA = 0. Fig. [11. the external transverse spatial proﬁle of the beam. For instance. (3) (1) where w0 ∼ (kNA) is the beam waist and k is the wavenumber of light. 2). and the geometric-phase diﬀerence between the constituent plane-wave components produce the SOI of light [11. Φ(y). the slit be parallel to the y-axis. using the basis of linear polarizations |X and |Y along the corresponding axes. 0) and |L = (0.6. 14. respectively [16]. this naturally deﬁnes the post-selected polarization state to be perpendicular to the slit: |Ψout = |X = |L + |R √ . whereas the SPP beams propagate in the ±x directions (from now on we consider only the +x beam). this |X state in the local coordinate frame attached to the direction of propagation of the beam corre- . Let the incident light propagate along the z-axis (z = 0 represents the gold surface). so that the actual focal spot occurs in the secondary. |R and |L . and the leakage SPP signal is collected by an immersion objective. let the y-distribution of the incident light be Gaussian in the focal plane: 2 Φin ∝ exp −y 2 /w0 . as shown in Fig. Interaction of light with the slit and transformation to SPPs is similar to the beam refraction at a plane interface. the incident optical beam is partially scattered into two SPP beams propagating along the gold layer orthogonally to the slit (Fig.

z) plane: ˆ ˜ ˜ Eslit = Py Elens . does not take part in the interaction. but the coupling with the SPPs is described in the momentum representation. the focused ﬁeld (8) interacts with the slit.— The detailed wave picture of light evolution in the system can be divided into three stages: (i) focusing of the initial paraxial polarized ﬁeld by a high-NA objective. 15–18]: σw = Ψout | σ3 | Ψin ˆ i = . (ii) interaction of the focused ﬁeld with the slit. without change of the polarization state in the local coordinates. the ﬁeld Eslit can be considered as the source for the SPPs. (5). whereas the y component. the SPP beams demonstrate angular SHEL deviation which is seen as the momentum shift in the Fourier space. approximating the incident light by a single plane wave with the complex electric-ﬁeld amplitude E0 ∝ (−ε. 0)T (in the Cartesian basis). sin θ cos φ. 2. First. Eq. with the experimental plasmonic results presented in Fig. Second. |σw | ≫ 1. the output SPP beam undergoes complex transverse shift (6) equaling to the SHEL shift λ multiplied by the helicity weak value (5). respectively. Ψout | Ψin ε (5) (a) 10 µm (b) y x (c) y x (d) Remarkably. and cos θ is the apodization factor which provides conservation of the energy ﬂow [26]. 26]. 17]: y = Re∆ . (9) ˆ where Py = diag (1. 2: (color online). whereas pre-selected polarizations of the incident light are shown in cyan.3 sponds to the z-component of the SPP beam propagating along the x-axis. based on the SOI of light. FIG. However. (8) ˆ ˆ ˆ ˆ ˆ Here Tlens = Rz (−φ) Ry (−θ) Rz (φ). is given by geometric rotational transformation [20]: √ ˆ ˜ Elens (θ. Complete wave theory. φ). φ) indicating the direction of the wave vectors k = k (cos θ. In terms of weak measurements. refracted by the lens. It implies that the transverse electric ﬁeld of the wave is parallel-transported along each geometrical-optics ray. ∆ ≃ −λσw . φ) E0 . Both coordinate and momentum shifts are clearly visible in the SPP ﬁelds in real (a. In doing so. 1) is the projection operator.d) spaces for two types of the pre-selected polarization. sin θ sin φ). real and imaginary parts of σw produce coordinate shift (displacement) and momentum shift (deﬂection) of the beam proﬁle [10–14. the high-NA focusing is described by the Debye-Wolf approach [20.07i k x ε = 0.. Excited SPP ﬁeld can be decomposed into . only x and z components of the ﬁeld can excite SPPs. As we show below. this weak value is complex and large. Weak measurement of helicity. with Ra (α) denoting the SO(3) matrix operator of rotation about the a√ axis by the angle α. parallel to the slit. the accurate description of the light-SPP coupling appends an additional π/2 phase factor that interchanges coordinate and momentum shifts with respect to the real or imaginary nature of ε. (6) Thus. These swapped relations with respect to usual SHEL shifts have to be understood as an inherent feature of the plasmonic SHEL. and we perform this integration afterwards. panels (a) and (c). Compare now the above weak-measurement model. In other words. and (iii) generation and propagation of the SPP beams. This is described by the projection of the ﬁeld ˜ Elens onto the (x. the plane-wave ˜ ˜ spectrum of the focused ﬁeld. whereas imaginary ε (elliptical polarization) induces momentum shift ky = 0. the slit acts as a polarizer which cuts out the y component of the ﬁeld. Marking the rays by spherical angles (θ. (5) that elliptical and linear tilted pre-selected polarizations yield real and imaginary σw . 1. panels (b) and (d).e. −2 ky = 2w0 Im∆ . Elens (k) ≡ Elens (θ. ˜ Finally. the SPP beams exhibit coordinate SHEL shift which is seen only in real space. although the photon helicity eigenvalues are σ = ±1.b) and Fourier (c. φ) ∝ cos θ Tlens (θ. the observed coordinate and momentum shifts of the SPP beams turn out to be swapped as compared to the model (5)–(7): Real ε (tilted linear polarization) causes coordinate shift y = 0. i. Typical SPP ﬁeld distributions in real and Fourier spaces (see also [29]). results in the transverse shift of the “measuring subsystem”. respectively. 0. For tilted linear polarization (real ε). The mean momenta ky and transverse coordinates y of the beams are schematically indicated by orange and green arrows. In this manner. In the case of elliptical polarization (imaginary ε). (7) Note that the weak-measurement approximation fails in the vicinity of ε = 0 and is applicable at λ/w0 ≪ ε ≪ 1 [16]. The real-space focused ﬁeld is given by the Fourier-type inte˜ gral over all plane waves Elens (k). It is seen from Eq. the input and output polarization states (2) and (4) determine the weak value σw of the photon helicity [11.07 . of the transverse beam proﬁle [16]: 2 Φout ∝ exp − (y − ∆) /w0 2 ky ky k x ε = −0.

χ = κ/kp [27]. 1 kp T . To compare experiment and theory quantitatively. FDTD 3 2 (b) Theory Exper. the phase matching condition can be written as kpy = ky . We have veriﬁed numerically that these equations yield SPP distributions and shifts corresponding to experimental Figure 2 [29]. 0 (dissipation is neglected hereafter). the SPP plane-wave amplitude is ky kp 2 where ky /k ≃ θ sin φ. is multiplied by −iχ = −iκ/kp because of the coupling between the x-components of light and SPP ﬁelds.1 1 Re ε FIG. in order to make the weak-measurement formalism (1)–(7) consistent with our plasmonic system. −iχ . for our experiment one can use the ﬁrst postparaxial approximation. eﬀectively swaps the real and imaginary parts of ε in the coupling process. FDTD ky / k 0 0 −1 −2 −0. Eqs. 2 kθc 2χ2 |ε|2 + θc /2 ky = −kθc χθc Imε . the resulting SPP ﬁeld in the momentum and coordinate representations reads 2π θc ˜ ˜∗ ˜ ˜ Ep ∝ E p · Eslit E p . 0. Keeping only terms linear in θ and ε. θc = sin−1 (NA) is the aperture angle of the focusing objective. For experimental convenience. 2). 0.2 (a) Theory Exper. Excitation of the SPPs by focused light via the slit is determined by: (i) the phase matching condition that provides transformation of the wave momenta k → kp and (ii) the coupling eﬃciency. (8)–(12) yield ky ˜ Eslit ∝ −ε. but neglecting εθ-order terms. γ. (12) Here the second Eq. we deﬁne y = Ep y Ep Ep Ep and ˜ ˜ ˜ ˜ ky = Ep ky Ep Ep Ep . kpy . 3: (color online). (15) with χ = 1 [28]) and experimentally measured SHEL shifts of the SPP beams versus complex pre-selection parameter ε (see also [29]).6. exponential decay away from the metal surface. one has to substitute ε → −iχε ˆ ˆ or HSOI → iχ−1 HSOI in Eqs. −iχ . 2 2χ2 |ε|2 + θc /2 (15) ky ˜ . In fact. sin θ ≃ θ and cos θ ≃ 1. (10) where ky = k sin θ sin φ. imaginary transverse wave number of SPPs. where the inner product implies the scalar product of the complex vector amplitudes and the integration in the momentum space: ˜ ˜ Ep Ep ∝ 2π θc 0 0 ˜p ˜ E∗ · Ep θdθdφ. kp (11) Taking into account the coupling eﬃciency γ. 1) . we calculate the SPP beam centroids in the coordinate and momentum spaces.1 −0. (13) . Taking into account that the SPP wave number kp > k is ﬁxed (which determines the SPP circle in the Fourier space. − k T It is seen from here that the imaginary longitudinal SPP ﬁeld plays crucial role in the coupling with light. Performing these cal- culations with the SPP Fourier spectrum (14).5 1 −3 −3 −2 −1 0 1 2 3 Im ε k y 0. From here and equations above.4 ˜ plane waves E p which are characterized by the real wave vectors kp = kpx . Indeed. Ep ∝ −iχε − k k (−iχ.2 −1 −0. and we consider only the (x. which we model by the inner product of the electric-ﬁeld amplitudes of the ˜∗ ˜ incident light and SPPs: γ ∝ E p · Eslit [28]. Theoretically calculated (Eqs. exp (κz) at z < 0. Since the system is homogeneous in the y-direction.5 0 0. kpz = −iκ. and the complex electric-ﬁeld ampli˜ ˜ ˜ tudes E p = Ep⊥ ez + Ep kp /kp . we ﬁnally arrive at y = 1 2χθc Reε . the weak component of the input polarization. (14) T ˜ E p ∝ −iχ 1 − T ky . ε. the real space shifts were measured with NA=0. 1 . (a) – imaginary ε generating the momentum shift ky and (b) – real ε producing the coordinate shift y . Therefore. kpx = 2 2 kp − ky . Hence. 0. This results in the following coupling coeﬃcient and the SPP ﬁeld: γ ≃ −iχε − ky ˜ ky . omitting the common z-dependence. y)-distribution of the SPP ﬁeld. (3)–(7). the corresponding momentum component must be conserved. Fig. From Maxwell equations it follows that the longitudinal and transverse ﬁeld com˜ ˜ ponents are related as Ep = −iχEp⊥ . Equations (8)–(12) completely describe our system starting from the incident paraxial ﬁeld E0 and yielding the output SPP ﬁeld distribution Ep on the gold layer. E p ∝ −iχ. Taking into account that the position operator is y = i∂/∂ky in the momentum repreˆ ˜ ˆ ˜ ˜ ˜ sentation.45 and Fourier-space shifts with NA=0. This modiﬁcation immediately ascertains perfect agreement between the weak-measurement model and experimental results in Fig. (12) is the Fourier (Debye) integral over all incoming plane waves. which signiﬁcantly simpliﬁes equations and allows analytical solution. Ep ∝ 0 0 ˜ Ep eikp ·r sinθdθdφ. 2.

I.0327v1. Appl. J. Rev. and C. He. Fedoseyev. and E. Bliokh. Gen. J. 787 (1972). S. [9] K. Dogariu. Y. and FDTD distributions of the SPP ﬁelds at diﬀerent input polarizations are given in Supplementary Material. and J. Barnes. O. Y. and E. Opt. Albert. Li. Therefore. Rev. Phys. Bliokh. and F. this is not so because of the strong anisotropy of the system in the (x. and G. 465 (1955). Gen. A. [16] I. [21] K. G. A: Math. wave theory. P. Rev. Jeffries. 083901 (2004). Phys. Hohenau. Hugonin and R. and S. Jozsa. 96. Y. Chiu. E. The presented results demonstrate the unique ability of surface waves to perform as a post-selecting measuring device which might be potentially useful for various sensing applications involving the chirality of light. Nikitin. 92. E 75. Phys. The coupling anisotropy can be taken into account by modifying coeﬃcient χ relating the longitudinal and transverse SPP ﬁelds. [14] K. Kleiner. (15) are valid in the whole range of values of ε. [8] M. 2012. 27 (2010). Imbert. Murakami. [25] High-NA focusing also induces SOI and partial spin-toorbital angular momentum conversion in the incident light [19. and N. Bliokh. A. 21. Lett. M. Richards and E. Ashhab. and L. and E. 061801(R) (2009). Zel’dovich. Gorodetski. Phys. this simple system oﬀers a perfect weak-measurement tool where ﬁxed polarization of SPPs provides a built-in post-selection. Popescu. We acknowledge fruitful discussions with Y. 253601 (2010). Babiker (Cambridge University Press. Y. V. [7] K. Bretenaker. and J. 1351 (1988).e. V. and the resulting shifts are also presented in Fig. A. Lett. A. in The Angular Momentum of Light. A. V. Opt. Rodr´ ıguez-Herrera. 104. Koller. Y. Phys. Tiny variations of the input polarization of light bring about huge SHEL shifts of the SPP beams in coordinate and momentum spaces. Drezet. Stepanov. [26] B. Lett. Opt. to appear). P. F. Kwiat. Y. 2. MSEB 149. A: Math. D 40. we performed a ﬁnite diﬀerence time domain (FDTD) simulations for two input polarizations. and E. Phys. Evidently. Phys. Hosten and P. Sukhov. [15] Y.6315. Phys. D. 073903 (2006). Dainty. [19] D. In the weak-measurement range λ/w0 ≪ ε ≪ 1.05 here and can be neglected against the strong SOI (∝ NA) induced by the slit. Lett. Vaidman. Y. [2] C. A. oo Commun. J. Sudarshan. A 76. R. London A 253. Leitner. −1 and w0 = 2 (kθc ) . 20]. 5199 (1992). and D. and Q. Lara. o [13] Y. Ostrovskaya. Qin. M. T¨r¨k. Kleiner. In general. Opt. Lett. G. Nouv. Lett. 030404 (2008). 93. and A. A. 2045 (1988). Bliokh. Aussenegg. 122 (1965). Stevenson. 20. Aiello. Dutriaux. [22] N. C. Physics Today 63. E. Phys. 2112 (1989). Gorodetski. Dainty. P. 123903 (2009). L. Gorodetski.W. [12] A. 148. W. Niv. Europhys. 2551 (2009). A 333. Phys. 3 to ﬁt the experimental data. and Q. Bliokh and A. Woerdman. Fedorov. S. 220 (2008). Alonso. Lett. and T. D. Higdon. B. T. K. Y. D. and T. S. [6] V. Rev. [3] H. and FDTD simulations are all in perfect agreement [29]. 181 (2004). 24. Li. Nano Lett. Ebbesen. 073901 (2007). Y. Schilling. 748 (2008). Rev. J. Qin. Phys. Merano. (7) with modiﬁcation ε → −iχε. Kleiner. D. Z. Le Floch. Lett. J. z) plane. F. Nature Photon. 300 (1998). 044103 (2007). Phys. Ricard. Moreover. Haefner. arXiv:1110. Niv. Rev. D. [18] Y. D. Science 319. Y. Phys. Aharonov. Opt. Gilles. Zhao. 107401 (2004). P. Tollaksen. 60. and M. A. Devaux. Dennis and J. Remarkably. Hasman. 787 (2008). A 80. G. Murray. [1] F. Comparisons of the experimentally measured coordinate and momentum transverse shifts (as dependent on the complex polarization parameter ε) with the theoretical results (15) are shown in Figure 3. D 5.. A. KrennGordon. 43. V. Sambles. Y. which is chosen in Fig. Liberman and B. Opt. [24] A. Rev. Rev. Y. Rev. Y. Bliokh and Y. M. Lett. Rev. Phys. . However.5055v3. Lara. Aharonov. Rev. 9636 (2011). Lett. (Berlin) 16. Wilson. Express 19. 8. Ditlbacher. [10] K. Bliokh and Y. 5. [4] M.— We have observed and examined in detail an extraordinary plasmonic SHEL appearing in the interaction of focused light with a straight slit on the metal surface. Bliokh and Y. Aiello and J. 358 (1959). Express 19. and the ﬁnancial support from the ERC (Grant 227577) and the European Commission (Marie Curie Action). J. the experiment. 3. 99. H. P. Y. 5855 (1991). Nagaosa. Miguel A. Onoda. Bliokh. Edgar. R. Duck. Rev. Wolf. 11. 043903 (2008). Rodr´ ıguez-Herrera. Phys. W. Ann. Y. 101. K. M. Yang. K. Lett. Rev. as well as experimental. L. 33. R. Opt. 066609 (2007). [27] F. [20] O. Y. [11] O. 1437 (2008). 73 (1977). Lett. Shikano.-F. G. P. Girard. Proc.G. [17] R. Hasman. Petit. 34. Xiao. and F. Phys. P. we regard χ as a free parameter. Player. and J. Phys. arXiv:1109. Opt. [28] This coupling coeﬃcient γ implies equal eﬃciency of couplings for the x and z components of the ﬁelds. Soc. Phys. Y. Lett. Nori. I. 2068 (2011). E. Dokl. P. Bliokh. Alonso. Woerdman. 1863 (2004). Ostrovskaya. Rev. Nauk SSSR 105. Gong. Akad. R. weak-measurement interpretation. Feng. ∆ → iχ−1 ∆ = (χkε)−1 . Aiello. they are precisely equivalent to Eqs. 101. S. H. S. H. A. A. X. Shitrit. S. [29] Details of the centroid calculations and measurements. 3667 (1987). B 44. D. Rev. eﬃciency of this conversion is estimated as NA2 /4 ≃ 0. Bradberry. Phys. 345 (1993). Kleiner. Conclusion. and E. Andrews and M. P.5 We emphasize that Eqs. Gong. theoretical. McGloin. 102. Pillon. M. Y. [5] L. A. Rev. H. A. Y. J. V. Steinberger. 26132 (2011). and C. Bliokh. Hasman. [23] W. which correspond to the imaginary and real parts of the weak value of the helicity of light. A 46. J. Dintinger. Yang. P. Hasman. A. arXiv:1204. Rev. edited by D. B. Bretner. Bliokh. G¨tte. i. Phys. 7 (1974). J. G. Kofman.

- Read and print without ads
- Download to keep your version
- Edit, email or read offline

Are you sure?

This action might not be possible to undo. Are you sure you want to continue?

CANCEL

OK

You've been reading!

NO, THANKS

OK

scribd