Frame dragging with optical vortices
J. Strohaber
Texas A&M University, Department of Physics, College Station, TX 77843-4242, USA
*
Corresponding author: jstroha1@[Link]
General Relativistic calculations in the linear regime have been made for electromagnetic beams
of radiation known as optical vortices. These exotic beams of light carry a physical quantity
known as optical orbital angular momentum (OAM). It is found that when a massive spinning
neutral particle is placed along the optical axis a phenomenon known as inertial frame dragging
occurs. Our results are compared with those found previously for a ring laser and an order of
magnitude estimate of the laser intensity needed for a precession frequency of 1 Hz is given for
these steady beams of light.
PACS numbers: 04.20.-q, [Link], 42.60-v
In pre-relativity, Newtons law of universal gravitation was used to quantify the gravitational
force between massive point objects. From this elementary theory of gravity, it is well known
that mass is the source of gravitational fields, and those gravitational fields act only upon
massive particles [1,2]. In contrast, in 1804 the German physicist Soldner used Newtons
corpuscular (little particle) theory of light to show that starlight skimming the sun could be
deflected about a straight path. This result is often referred to as the Newtonian prediction
[2,3]. It was not until the birth of General Relativity (GR) that Einstein theoretically
demonstrated that the trajectories of photons could be influenced by a gravitational field and in
the case of the deflection of starlight by the sun he gave the correct result which is twice the
Newtonian prediction [4]. In May of 1919, this deflection was observed by Eddington during a
total eclipse [5]. More recently, the deflection of light around massive objects has been exploited
to detect and investigate astronomical phenomenon by a GR effect known as gravitational
lensing [6].
Even more fascinating, in GR the geometry of spacetime depends on the energy-
momentum configure of a source, and since light carries energy and momentum, light also acts
as a source of gravity. In 1931, Tolman showed that thin pencils of light produce a
gravitational influence on test rays and particles [7]. Scully later extended upon the earlier work
of Tolman et. al., by calculating the gravitational coupling between subluminal laser pulses [8].
More recently, Mallett generalized upon these earlier works by solving Einsteins equations in
the weak-field limit for a ring laser configuration [9]. These calculations demonstrated a
phenomenon known as inertial frame dragging. Of further interest, in a later controversial
publication, Mallet found exact solutions of the Einstein field equations and showed that the
exterior metric contained closed timelike curves (CTCs) [10,11]. These light-induced CTCs are
conjectured by Mallett to be the foundation for a time machine. In this letter, we extend upon the
work of Mallett by calculating the weak gravitational influence from a Laguerre-Gaussian ( LG
)
laser beam on a massive spinning neutral particle. These exotic beams of light known as optical
vortices (OVs) have generated considerable interest in the scientific community because they
carry a physical quantity known as optical orbital angular momentum (OAM) [12]. It is the goal
of this paper to calculate the coupling of this unique form of radiation with a spinning test
particle and to compare the results with those found in the case of a ring laser.
Under the Lorentz gauge condition c
o
A
o
= 0 and for source-free vacuum
electromagnetic radiation, the four-potential satisfies the homogeneous wave equation
c
A
o
= 0. The amplitude of paraxial beams are solutions of the scalar paraxial wave equation
(PWE), which is ultimately derived from the wave equation. To find the electromagnetic fields
of paraxial beams, the scalar and vector potentials are chosen as A
t
= A
0
and
A = A
0
o
e
x
+ |
e
y
( )
respectively [13]. Here o and | are polarization parameters such that
1 o = , 0 | = ( ) 0, =1 o | = correspond to polarization along the -direction x ( ) -direction y , and
is a solution to the PWE. Using Maxwells equations and the Lorentz gauge condition, the
electric and magnetic fields within the paraxial approximation are,
E = E
0
i oe
x
+ |e
y
( )
1
k
o
c
cx
+ |
c
cy
|
\
|
.
|
e
z
(
e
i kzet ( )
B = B
0
i |e
x
oe
y
( )
+
1
k
|
c
cx
o
c
cy
|
\
|
.
|
e
z
(
e
i kzet ( )
. (1)
Three families of solutions have been found by separation of variable. In Cartesian
coordinates the solutions are the Hermite-Gaussian modes, and in cylindrical polar and elliptical
coordinates they are the Laguerre- and Ince-Gaussian modes respectively. For the present work,
the radiation field of the LG
beams will be of interest as the source of the gravitational
influence,
( )
G
2 2
2 1
2 2 0
2 2 2
2 2 1
exp exp
2
i p
i ikz
p
w r r k
L r i r e e e
w w w w R
u
+ + +
| |
| |
| | | |
|
| | |
|
\ . \ .
\ .
\ .
. (2)
Here
2 2
0 0
( ) 1 / w z w z z = + is the beam size, where
0
w and
2
0 0
/ 2 z kw = are the waist and
Rayleigh range respectively.
2
0
( ) / R z z z z = + is the radius of curvature,
0
( ) arctan( / )
G
z z z + = is
the Gouy phase and
p
L
are the associated Laguerre polynomials.
Using Eqs. (1) and (2) and a change of basis, it is straightforward to calculate the
Poynting vector
S =
E
B
( )
/
0
for the LG
beams in cylindrical polar coordinates [12],
S = E
0
2 1
c
0
r
R(z)
e
r
+
1
k
r
o
z
1
2
c
cr
|
\
|
.
|
e
u
+ e
z
2
. (3)
Here the photon helicity parameter is o
z
= i o|
-
o
-
|
( )
, where o 1 is for right and left
circularly polarized light and o
z
= 0 is for linearly polarized light. In addition to the longitudinal
component of energy flow normally encountered with planewaves, paraxial beams in general
have radial and azimuthal components. Integral curves of the Poynting vector spiral around the
optical axis of the beamsuggesting that these curves are the paths traveled by the photons. The
radial component is inversely proportional to ( ) R z , which is an inescapable consequence of the
diffraction of finite-sized beams. For OVs, the azimuthal component is dependent upon the
angular momentum mode number
in addition to the helicity parameter o
z
. In general, the
azimuthal component in Eq. (3) is small compared to the longitudinal component; however, we
will demonstrate that the magnitude presents no difficulties relative to results found in Ref [9],
where the longitudinal components were the sole contributor to the Poynting vectors.
For weak gravitational fields, the full nonlinear Einstein equations can be linearized as
c
h
v
= k T
v
q
v
T / 2
( )
, where T
v
= F
o
F
o
v
q
v
F
o|
F
o|
/ 4
( )
/
0
is the energy-
momentum tensor and F
v
= c
A
v
c
v
A
is electromagnetic tensor. It will be seen that the trace
of the energy-momentum tensor is zero T = q
o|
T
o|
= 0, thereby reducing the linearized field
equation to c
h
v
= kT
v
[8,9]. As a note, in the linear approximation, Einsteins equations
are analogous to potential problems typically encountered in graduate textbooks on
electrodynamics [13,14]. The solutions for the metric perturbation in the linearized version of the
Einstein field equations can be approximated as [79],
( )
3
( ) ( , ) , h x d x G x x T x t x x
v v
k ' ' ' ' =
}
. (4)
Here
4
8 / G c k t = where the gravitational constant is
11 3 2
6.67 10 m /kg/s G
= ,
( )
( , ) 1/ 4 G x x x x t ' ' = is the Greens function and T
v
is the energy-momentum tensor. Using
the results from Eqs. (1) and (3) and noticing that
x y
E cB = and
y x
E cB = , the energy-
momentum tensor for
LG
p
beams with angular and radial quantum numbers
and = 0 is
found in the paraxial approximation to be,
0 0
1
0 0
z x y z
x x
y y
z x y z
S S S S
S S
T
S S
c
S S S S
v
(
(
(
=
(
(
(
. (5)
The energy-momentum tensor in Eq. (5) is symmetric, and conveniently within the paraxial
approximation, its elements can be written in terms of the Cartesian components of the Poynting
vector in Eq. (3).
To test for the effects of inertial frame dragging, a spinning test particle will be placed
along the optical axis of the beam [Fig. 1]. With this in mind, it is computationally advantageous
to inspect the form of the General Relativistic spin equations before evaluating the integrals in
Eq. (4). The General Relativistic spin equations in the slow motion and weak field approximates
are given by [9,15]
0 1 0
0 0
k j k l k j i
i k i j il k ik j
ds
c s s v s v c s v v
dt
= I I I I , (6)
where
o|
I are the Christoffel connection coefficients, and
i
s (not to be confused with the
Poynting vector, which uses upper case lettering) and
i
v are the spin and velocity vectors. For a
stationary particle, Eq. (6) reduces to
0
/
k
i i k
ds dt c s = I and the required Christoffel symbols are
given by
( )
0 ,0 0, 0,
2
k kj
i ji j i i j
h h h q I = + . In calculating the energy-momentum tensor given in Eq.
(5), a continuous-wave field was time averaged, and because the integral in Eq. (4) will be
calculated for a steady beam of light between / 2 z L = and / 2 z L = , the connection
coefficients take the simplified form
( )
0 0, 0,
2
k kj
i j i i j
h h q I = . Using these conditions, the spatial
components of Eq. (6) can be written as,
( ) ( )
( ) ( )
0 , 0, 0 , 0,
0 , 0, 0 , 0,
0 , 0 ,
2
2
2
yy zz x
y x x y y z x x z z
y xx zz
x y y x x z y y z z
xx yy z
x z x y z y
ds
h h s h h s
c dt
ds
h h s h h s
c dt
ds
h s h s
c dt
q q
q q
q q
= +
= +
= +
. (7)
The Greens function in cylindrical polar coordinates can be cumbersome to work with
during integration; however, since we are interested in the interior points very near to the optical
axis, the Greens function can be expanded using a Taylor-Maclaurin series expansion about the
optical axis where the test particle will be placed,
( )
0
0 0 0
1 1
( , ) cos
4
n
n
n
r r
G x x P
R R R
u u
t
=
| | | | '
' ' =
| |
\ . \ .
. (8)
Here
2 2
0
(0, 0, ) ( ) R R z r z z ' ' = = + and ( )
n
P x are the Legendre polynomials. The primed
quantities are those of the source of radiation and the unprimed quantities are for the observation
points. Equation (8) simplifies calculation by allowing for the trigonometric function to be made
separable from the radial r and longitudinal z
coordinates.
Even with the simplifications resulting from the use of a steady beam and the expansion
of the Greens function in terms of the Legendre polynomials, the integral in Eq. (4) is
challenging. Up to this point, we are still within the framework of the paraxial and General
Relativistic approximations. Our first approximation is to take the doughnut-shaped intensity
profile of the Laguerre-Gaussian beams as a thin cylindrical shell of radius
0
r and length L
[see
Fig. 2],
( ) 2
0
L
0
( )
2
r r
A z
r
o
t
'
' = [ . (9)
Here
2
0
A r t = is taken to be the transverse area of the beam,
0
( ) r r o is the Dirac delta function,
and
L
H is a top-hat function of length Land centered on 0 z = [14]. The radial position
0
r of the
peak in the intensity profile
2
, which is determined by the condition
2
/ 0 r c c = , is at
0
( ) / 2 r w z = . Because the integral in Eq. (4) is performed with the Dirac delta function of
Eq. (9), we can simplify the derivative term appearing in Eq. (3); this derivative is
( )
2 2
2
/ 2 / 2 / r r r w c c = and upon substituting
0
( ) / 2 r w z = for r into this
expression we find
2
/ 0 r c c = . Lastly, we neglect the effects of diffraction by setting
-dependent z beam parameters to their value at 0 z = , viz.,
0
(0) w w = and (0) R for the
entire track of the beam. The coefficients in Eq. (7) can then be found by performing the
integrals in Eq. (4) with Eqs. (8) and (9), taking the appropriate derivatives in Eq (7), and
evaluating the results at the origin,
0 , 4 2
2 2
0
0
0 , 4 2
2 2
0
0
0, 0, 0 , 0 ,
( 0)
2
( / 2)
( 0)
2
( / 2)
( ) 0
x y L
y x L
x z y z z x z y
G L
h x
c r
r L
G L
h x
c r
r L
h x h h h
t
(
( = =
( +
(
( = =
( +
= = = =
, (10)
Here
L
is the linear radiation density and is related to the volume density by
2
0 L V
r t = .
Using combinations of the quantities found in Eq. (10), the time rate of change of the spin
vector with respect to time is found to satisfy the cross product d dt =
S / S O , where
( )
OV
0, 0,
z
O
O=
is the precession frequency. The spinning test particle placed at the center of the
OV will have a rate of precession given by
OV 3 2
0
( ) /
z L
G f L c r t O = , where
2 2
0
( ) / 4 f L L r L = + . Our first observation is that the precession frequency in Eq. (11) is
quantized with the angular momentum mode number of the OV and depends on its sign.
Second, for an infinitely long beam track L the precession frequency
OV
z
O
is finite. This is
expected since the strength of the influence decreases from a finite source. For example, in Ref
[9], we find that the strength of the influence on a particle placed at a position z is
( )
3 2 2 2 2
2 / 2 4 4 a a z a z
(
+ +
, which falls off as
3
1/ z . For OVs, the strength of the influence is
2 2
0
/ 4 f L r L = + . Letting the beam track be equal to twice the Rayleigh range (diffraction
length)
0
2 L z = and allowing the transverse beam size to be equal to , the strength reaches
~ 95% of its maximum value. This justifies neglecting diffraction and allows us to simplify the
expression for the rate of precession as
OV 1 2 3
0
/
z L
G r c t
O ~ .
The rate of precession in the case of the ring laser was found to be
ring 3
8 2 /
z L
G ac O = ,
where a is the length of each side of the ring laser. From the previous discussion on OVs, it
was shown that the influence is a collective effect from the beam track, while the influence for
the ring laser is from the pointing vector being in the plane of the laser. Both configurations
benefit from decreasing the radius of their rings. However, optical vortices can be tightly focused
to roughly a wavelength so that
OV 1 3
/
z L
G c t
O ~ . Ultimately, The ring laser dimensions are
limited by the damage threshold (
12 2
~10 W/cm ) of the optical material used in its construction:
making the ring laser smaller requires making the beam smaller, which results in a greater peak
laser intensity.
In conclusion, General Relativistic calculations in the weak-field limit have shown that
optical vortices can exhibit frame dragging on a spinning test particle. As a result of the
azimuthal component of the Poynting vector, OVs do not require optical constraints to circulate
the radiation. Most importantly their ring size is limited only by the wavelength of light. To get a
feel for the magnitude of the frame-dragging effect, an estimate of the peak laser intensity
needed for a 1 Hz precession frequency is made for these steady configurations. By focusing
an 800 nm, single cycle pulse down to a wavelength, the required intensity is
28 2
~10 W/cm ,
which is close to the Schwinger intensity of
29 2
10 W/cm
S
I needed for electron-positron pair
production [16]. Currently, the worlds most intense laser pulses are produced by the Hercules
laser system at the University of Michigan, which deliverers 30 fs laser pulses to a focal spot-
size of about a micron to yield an intensity of
22 2
~ 2 10 W/cm [17].
REFERENCES
References with Titles
1) Fowles G R and Cassidy G 1990 Analytical Mechanics 6th ed. (Fort Worth: Saunders)
2) Coles, Peter, Einstein, Eddington, and the 1919 Eclipse, in V.J. Martinez, V. Trimble
& M.J. Pons-Borderia, eds., Proceedings of International School on the Historical
Development of Modern Cosmology, Valencia 2000 (Astronomical Society of the
Pacific, San Francisco, CA, 2001) pp. 2141.
3) J. G. von Soldner, On the deflection of a light ray from its rectilinear motion, B. A. J.
161-172 (1804).
4) A. Einstein, Die Grundlage der allgemeinen Relativitatsheorie, Ann. Phys. (Leipzig)
49, 769 (1916).
5) F. W. Dyson, A, S, Eddington and C. Davidson, A determination of the deflection of
light by the suns gravitational field, from observation made at the total eclipse of May
29, 1919, Phil. Trans. R. Soc. Lond. A 220, 291-333 (1920).
6) S. S. Doeleman, et. al., Event-horizon-scale structure in the supermassive black hole
candidate at the Galactic Center, Nature 455, 78 (2008).
7) R. C. Tolman, P. Ehrenfest, and B. Podolsky, On the gravitational field produced by
light, Phys. Rev. 37, 602 (1931).
8) M. O. Scully, General-relativistic treatment of the gravitational coupling between laser
beams, Phys. Rev. D 19, 3585 (1979).
9) R. L. Mallet, Weak gravitational field of the electromagnetic radiation in a ring laser,
Phys. Lett. A 269, 214-217 (2000).
10) R. L. Mallett, The gravitational field of a circulating light beam, Found. Phys. 33, 1307
(2003).
11) K. D. Olum and A. Everett, Can a circulating light beam produce a time machine,
Found. Phys. Lett. 18, 379 (2005).
12) L. Allen, S. M. Barnett, and M. J. Padgett, eds., Optical Angular Momentum (IOP,
2003).
13) D. J. Grifths, Introduction to Electrodynamics (Prentice Hall, Upper Saddle River, NJ,
1999), 3rd ed.
14) J. D. Jackson, Classical Electrodynamics (Wiley, New York, 1999), 3rd ed.
15) M. P. Hobson, G. P. Efstathiou, and A. N. Lasenby, General Relativity: An Introduction
for Physicists (Cambridge University Press, Cambridge, England, 2006).
16) T. Tajima and G. Mourou, Zettawatt-exawatt lasers and their applications in ultrastrong-
field physics, PRST-AB 5, 031301 (2002).
17) V. Yanovsky, et. al., Ultra-high intensity-300-TW laser at 0.1 Hz repetition rate, Opt.
Express 3, 2109 (2008).
Figure Captions
Fig. 1. (color online). Artistic rendition of a massive spinning neutral particle placed at the center
of an optical vortex beam. The multicolored spiral staircase surface is the phasefront of an
optical vortex beam. The intensity profile of the doughnut-shaped beam is encoded in the
transparency of this surface. The blue sphere at the center of the image is a spinning test particle.
Its spin vector is shown by the red arrow. The Pointing vector (not shown) follows integral
curves which spiral in an opposite sense to that in which the phasefront appears to spiral. In this
image the spin precession is in a clockwise sense when viewed from above.
Fig. 2. (color online) A thin cylindrical shell representation of a steady Laguerre-Gaussian
beam of length L and radius
0
r centered on the origin. The transverse beam profile is
approximated using a Dirac delta function and the length of the cylinder is approximated using a
top-hat function. The optical axis of the beam is along the -direction z , and the transverse
dimensions of the beam are in the - plane x y . The x'
vector labels the coordinates of the source
of radiation on the cylinder. The vector x
gives the coordinates of the observation points and the
vector R
is the displacement vector between the source and observation points.
Figures
Figure 1
Figure 2
