Observation of large “fast light” pulse advancement
without distortion
Michael D. Stenner and Daniel J. Gauthier
Duke University, Department of Physics, Box 90305, Durham, NC 27708
Voice: (919) 660-2511, FAX: (919) 660-2525, gauthier@phy.duke.edu
Abstract: We observe pulses advanced by 15% of their width and experiencing only minor
distortion using laser-driven potassium atoms in a novel configuration that avoids competing
nonlinear optical effects.
c 2002 Optical Society of America
°
OCIS codes: (190.5530) Pulse propagation and solitons; (030.1670) Coherent optical effects
There is currently great interest in tailoring the dispersion of gasses of atoms using intense laser fields. For
example, it is possible to create spectral regions of large anomalous dispersion so that the group velocity
is greater than c or less than zero (“fast light”). One experimental technique for studying fast light pulses
uses a continuous-wave bichromatic Raman pump laser beam to tailor the dispersion properties of an atomic
vapor [1]. It does this by creating two gain features with similar frequencies. Between these gain features,
the phase velocity dispersion is anomalous and the group velocity dispersion can be minimized. As a result,
a pulse traveling through the medium will experience “fast light” propagation with little distortion if its
frequency falls between these two gain features.
While a recent experiment has achieved extremely fast group velocities [1], the pulse advancement was
small (less than 2% of the pulse width). Large advancement is necessary to carefully study the propagation
of discontinuities imposed on the pulse envelope. These discontinuities are thought to be the carriers of
information and are therefore expected to propagate with velocity c [2], but this has not been verified
experimentally in “fast light” optical pulses.
In an effort to obtain large pulse advancement, we have found that the induced modulation instability
(IMI) causes temporal modulation of the pumping field. As the bichromatic gain becomes larger (leading to
greater “fast light” pulse advancement), the IMI leads to Raman-pump distortion, which in turn leads to
pulse distortion [3].
In order to achieve large advancement with little distortion, we employ a new experimental method in which
the pulse passes through through two spatially distinct gain regions [4]. Each of these regions is illuminated
by only one of the Raman pumping beams, as shown in Fig 1a. By passing through both regions, the pulse
experiences the required anomalous dispersion. However, because the two Raman pumps never exist in the
same location, the IMI can be avoided.
Using this new technique, we can now achieve unprecedented “fast light” pulse advancement while introducing only minor distortion. In preliminary experiments, we have advanced pulses by 15% of their width,
introducing very little distortion, as shown in Fig. 1b. The large advancement and small distortion of this
technique make it possible for the first time to explore experimentally the propagation of discontinuities on
optical pulses with velocities faster than c.
We will discuss “fast light” pulse propagation and its limitations, and present our progress in observation
of the propagation of discontinuities on dramatically advanced optical pulses in an effort to test proposed
definitions of the information velocity [2].
We gratefully acknowledge the financial support of the National Science Foundation, Grant #0139991.
References
1.
2.
3.
4.
L.J. Wang, A. Kuzmich, and A. Dogariu, ‘Gain-assisted superluminal light propogation,’ Nature 406, 277-279 (2000).
R.Y. Chiao and A.M. Steinberg, in Progress in Optics XXXVII, edited by E. Wolf, (Elsevier, Amsterdam, 1997), p. 396.
M.D. Stenner and D.J. Gauthier, ‘Instability Limits to “Fast Light” Pulse Propagation,’ (submitted for publication).
This method was suggested by L.J. Wang in private communication.
Stenner et al., Observation of large “fast light” pulse advancement without distortion
QELS2003/2002 Page
pump
ν1
39
input pulse
high-speed
photoreceiver
pump
ν2
K
39
K
intensity (arb units)
3
advanced pulse
2.5
2
1.5
vacuum pulse
1
0.5
0
-400
-200
0
time (ns)
200
Fig. 1. (a) Dual-zone setup used to generate “fast light” pulses. (b) Plot of pulse intensity as a function of
time for vacuum and advanced pulses. The vertical lines mark the centers of the pulses. The pulses have
been averaged 25 times.
400
2