Ultrafast CO2 laser technology: Application in ion acceleration
I. Pogorelsky1, V. Yakimenko1, M. Polyanskiy1, P. Shkolnikov2, M. Ispiryan2, D. Neely3,
P. McKenna4, D. Carroll4, Z. Najmudin5, and L. Willingale5.6
1
ATF, Brookhaven National Laboratory, Upton, NY 11973-5000, USA
of Electrical and Computer Engineering, SUNY at Stony Brook, NY
11794-2350 , USA
3 Central Laser Facility, Rutherford-Appleton Laboratory, Chilton, Oxon, OX11 0QX,
U.K.
4 SUPA Department of Physics, University of Strathclyde, Glasgow G4 0NG, U.K.
5 Blackett Laboratory, Imperial College London, London SW7 2BZ, U.K.
6 Center for Ultrafast Optical Science, University of Michigan, Ann Arbor, Michigan,
48105, USA
2 Department
Abstract
We review principles of picosecond CO2
lasers, operating at 10 m wavelength,
and their applications for strong-field
physics research. Such laser has been
used in a number of BNL experiments
that explore advanced methods of particle
acceleration and x-ray generation. We
illustrate merits of the wavelength scaling
from optical to mid-IR region by the
examples of ion/proton acceleration and
report the first experimental results that
confirm the expected wavelength scaling
of the process.
INTRODUCTION
Mainstream experimental research in
strong-field physics capitalizes so far on
the chirped pulse amplification (CPA)
solid-state lasers that have reached
petawatt peak power and 1021 W/cm2
intensities. Concurrently, there is interest
in exploring the capabilities of the CO2
gas lasers. While the peak power of CO2
lasers can hardly compete with that of
solid-state lasers, relativistic intensities
are already available, and longer
wavelength (~10 m) may offer
significant advantages in applications, as
well as a window into new areas of highintensity
laser-matter
interactions.
Presently, Neptune Laboratory at UCLA
and the Accelerator Test Facility (ATF) at
BNL conduct strong-field physics
experiments with CO2 lasers.
The following summary of the key
potential benefits of high-intensity CO2
lasers for R&D on advanced accelerators
and radiation sources is based on the
ATF’s 15-year experience in using longwavelength laser radiation combined with
a 70-MeV high-brightness electron linac.
Our first premise is the ease of scaling of
structure-based laser accelerators, and
electron phasing into the laser field, as
illustrated by STELLA, the first staged
monoenergetic laser accelerator founded
on the principle of the inverse free
electron laser (IFEL) [1]. In STELLA, the
laser and electron beams co-propagate
through two successive wigglers. In the
first wiggler (the “buncher”), the laser
provides periodical energy modulation of
the electron beam, which subsequently
divides into femtosecond microbunches at
the location of the second wiggler
(“accelerator”). Being periodically spaced
exactly to the laser’s wavelength, the
microbunches
in
the
accelerator
experience a uniform acceleration when
phased to the laser’s maximum amplitude,
as has been demonstrated in our
experiment. Such wavelength-accuracy is
difficult to achieve with the much shorterwavelength optical radiation of CPA
lasers.
Other applications gain from the
proportionally larger number of photons
per joule of laser energy at longer
wavelength. For example, a higher x-ray
yield in inverse Compton scattering is
achieved from counter-propagating the
electron- and CO2 laser-beams [2]. This
demonstration strongly argues for the
excellent prospects of ultra-bright laser
synchrotron sources for multi-disciplinary
applications [3].
At the center of our research reported
here is another favorable wavelength
scaling,
that
of
the
electron
ponderomotive potential in a laser field:
 pond 
1 e  2
It ensures that
E0
4 m L2
relativistic quiver motion  pond  mc 2
is
reached at a hundred times lower
laser intensity at 10 m than at 1 m.
(This relativistic condition is usually
expressed as a0=1 via dimensionless laser
strength a 0  0.89 I18  , where I18 is the
laser beam’s intensity in units of 1018
Wcm-2 and  in m.) The favorable
wavelength scaling was a factor that
enabled our direct single-shot imaging of
the 2nd harmonic in inverse Compton
scattering [4]. As the collective ion
motion in laser fields is usually driven by
ponderomotively accelerated plasma
electrons, we expect this scaling to
benefit laser-driven ion acceleration. In
particular,
the
main
mechanism
responsible for the observed laser
acceleration of protons and ions by lasers
interacting with thin foils, TNSA, relies
on relativistic electrons accelerated by the
laser [5].
Since our laser is inherently circularly
polarized, proton acceleration in our
experiments may occur also by another
mechanism,
Radiation
Pressure
Acceleration (RPA) [6]. A simplified
theoretical model for RPA yields the
maximum proton energy and accelerated
photon number at
p
MeV   nc ne a02 ,
E max
N  Sni 0  / 4 ,
(1)
where ne is the electron density in the area
where the acceleration takes place;
nc=π/(re λ2) is the critical electron density;
re≈2. 810-13 cm is the classical electron
radius, S is the laser focus spot area, and
ni0 is the ion density in this area.
Assuming that the deposition of laser
energy and ion acceleration mainly occur
near the critical plasma density, ne≈ni0
≈nc , Eqs. (1) are simplified to
p
MeV   a02 ,
E max
N p ~  re . (2)
p
Thus, both E max
and N p scale favorably
with λ. A similarly simple wavelength
scaling can be derived for TNSA
mechanism with the difference of
p
MeV   a0 .[5]
Emax
Another factor to consider is the
hundredfold lowering of the critical
plasma density when changing the laser’s
wavelength from 1 m to 10 m.
Overall, little experimental evidence
has been accumulated so far regarding
high-intensity, ultrafast laser-plasma
interaction at longer laser wavelengths to
confirm theoretical wavelength scaling. In
view of that, we began exploring
proton/ion acceleration by the ATF CO2
laser in collaborative effort that includes:
SUNY at Stony Brook, USA; Rutherford
Appleton Laboratory, UK; University of
Strathclyde, UK; Imperial College, UK;
and BNL, USA.
EXPERIMENT
The ion acceleration experiment has
required a number of modifications in the
BNL CO2 laser operations. First of all, the
efficiency of laser radiation in terms of
producing intense ion beams depends
significantly upon the laser’s contrast
factor, because a pre-pulse produces a
shock wave in the foil target that melts
and blurs the sharp solid-vacuum
interface at the rear surface of the target,
which is essential for TNSA. The prepulse control was not vital for our earlier
ATF experiments wherein the laser was
used primarily for interacting with the ebeam in a vacuum (e.g., for electron
acceleration, or inverse Compton
scattering). Therefore, at the initial stage
of our ion-acceleration experiment, we
made considerable effort to bring the laser
to the acceptably high contrast level. This
included blocking the picosecond prepulses, emerging due to power circulation
in a regenerative amplifier cavity, from
their leakage from the cavity and further
amplification in the final amplifier. We
accomplished this using a Pockels cell
switch between crossed polarizers, so
ensuring a power contrast at 104 between
the main pulse and a picosecond pre-pulse
that precedes the main pulse by 30 ns
(round trip time in the regenerative
amplifier cavity). No detectable ASE
pedestal as well as no pre-plasma at the
target’s surface have been observed.
Another fundamental problem arises from
the erosion of the spectral envelope of a
picosecond CO2 laser by the rotational
structure of molecular spectrum in the
amplifier. Simulations as well as optical
diagnostics revealed that the Fourier
transform of such a spectrum results in a
train of pulses (see Fig.1). These same
tools guided us to switch from the
conventional P-branch of the CO2 laser
spectrum to the R-branch. In the future,
we will change to using a multi-isotope
mixture
that
ensures
single-pulse
amplification. Meantime, partial pulsesplitting, as is shown in Fig. 1 (7.5 atm,
R-branch), remains a factor that may
influence our results.
During the ATF experiment reported
here, we focused a 5-ps CO2 laser pulse
of 3-J energy via an F#=2 off-axis
parabolic mirror on a 6-12 m thick Al
foils at an 450 incidence angle into a spot
with w0=65 m. This configuration
yielded an intensity ~1016 W/cm2, and
a 0  1 . The laser beam was polarized
circularly which allows us to reduce
parasitic back-reflections from the target
plasma into the laser system.
Fig.2a illustrates the optical arrangement
and ion-beam diagnostic inside the
vacuum interaction chamber. The simple
diagnostic shown schematically in Fig. 2b
includes a 100-m slit, a compact 5 kG
magnet spectrometer, and a metalized
scintillator plate imaged by a CCD
camera. The observed deflection of
particles from the direction normal to the
target surface implies that they are
positively charged ions.
To identify the nature of these ions, a
simple magnet spectrometer was modified
into the Thomson parabola configuration
by changing a slit into a pinhole, and
adding two internal electrodes statically
charged to  1.5 kV. Fig. 3 demonstrates
that the superposition of the magnetic and
electric fields separates the energy spectra
of different ion species by the degree of
their ionization, mass, and energy. The
sensitivity of the scintillator/CCD
diagnostic was insufficient to allow
single-shot observations of particles
transmitted through the 150-m diameter
pinhole and split into multiple traces by
the magnetic and electric fields.
Therefore, we employed the more
sensitive CR39 plastic plates in place of
the scintillator. For the first tests reported
here, we primarily were interested in
exploring features of a proton beam
which is normally released via TNSA
from the water and oil impurities at the
target’s surface and produces easy
recognizable tracks on CR39 plates. The
proton beam’s imprint, clearly visible on
the CR39 plate (Fig. 3b), was identified
and characterized by fitting to the
simulated dispersion curve (Fig. 3c). The
deviation of multiple-charged oxygen
ions from a theoretical parabola observed
at the periphery of the spectrometer’s
field of view could be due to the
combination of various effects such as
partial
recombination,
electric
breakdown, or field imperfection at the
electrode proximity.
Fig.4 depicts the spectrum of proton
energy, built from the count of
microscopic pit density along the track.
A quick comparison of this spectrum with
typical results from solid-state lasers of
~1018 W/cm2 intensity [7,8] discloses
similar features that confirm scaling in the
maximum proton energy and the beam
luminosity with the laser wavelength, as
we expected from Eq. 3.
CONCLUSIONS
We plan to continue optimization, and to
analyze more accurately effects of
wavelength scaling by detailed parametric
comparisons with earlier results from
solid-state lasers.
We also plan to replace the foil target
with a gas jet so that we can study laser-
plasma interactions closer to critical
conditions. This approach proved
beneficial for acceleration with solid-state
lasers [9]. Furthermore, we expect that
this change will allow us, for the first
time, to implement optical probing of
over-critical plasma interactions.
Simultaneously, we will improve the
laser’s peak intensity by at least an orderof-magnitude by employing a short focal
length parabola with F#=1, as well as via
increase of the laser peak power. These
steps include producing the 1-ps CO2
laser pulse and its amplification in multiisotope medium. A further increase in the
laser’s strength will be attained through
frequency chirping and compression;
these advances might bring the laser to
300 fs pulse-duration and 10 TW peakpower.
Summarizing, we report the first
observation of 1-MeV proton beam
produced by the interaction of a
picosecond CO2 laser with metal foils.
The intensity of the CO2 laser needed to
reach the same high-energy cut-off in the
proton spectrum was 100 times less than
that of a solid-state laser.
This is
consistent with the anticipated favorable
wavelength scaling, and highlights the
potential of long-wavelength CO2 lasers
as drivers for high-luminosity proton- and
ion sources.
Okugi, Y. Liu, P. He, and D. Cline, Phys.
Rev. ST Accel. Beams 3, 090702 (2000).
[3] V. Yakimenko and I. V. Pogorelsky,
Phys. Rev. ST Accel. Beams 9, 091001
(2006).
[4] M. Babzien, I. Ben-Zvi, K. Kusche, I.
V. Pavlishin, I. V. Pogorelsky, D. P.
Siddons, V. Yakimenko, D. Cline, F.
Zhou, T. Hirose, Y. Kamiya, T. Kumita,
T. Omori, J. Urakawa, and K. Yokoya,
Phys. Rev. Lett. 96, 054802 (2006).
[5] J. Fuchs, P. Antici, E. d'Humières, E.
Lefebvre, M. Borghesi, E. Brambrink, C.
A. Cecchetti, M. Kaluza, V. Malka, M.
Manclossi, S. Meyroneinc, P. Mora, J.
Schreiber, T. Toncian, H. Pépin and P.
Audeber, Nature Physics 2, 48 (2006).
[6] A. Macchi, F. Catani, T.V. Liseykina,
and F. Cornolti, Phys. Rev. Lett 94,
165003 (2005).
[7] P. McKenna, F. Lindau, O. Lundh, D.
Neely, A. Persson and C.G. Wahstrom,
Phil. Trans. R. Soc. A 364, 711 (2006).
[8] S. Nakamura, Y. Iwashita, A. Noda,
T. Shirai, H. Tongu, et al, Jap. J. of Appl.
Phys. 45, L913 (2006).
[9] L.Willingale, S.P.D. Mangles, P.M.
Nilson, R.J. Clarke, A.E. Dangor, M.C.
Kaluza, S. Karsch, K.L. Lancaster, W.B.
Mori, Z. Najmudin, J. Schreiber, A.G.R.
Thomas, M.S. Wei, and K. Krushelnick,
Phys. Rev. Lett. 96, 245002 (2006).
FIGURE CAPTIONS
AKNOWLEDGEMENTS
This work is supported by the US DOE
Grant #DE-FG02-07ER41488
REFERENCES
[1] W. D. Kimura, M. Babzien, I. BenZvi, L. P. Campbell, D. B. Cline, C. E.
Dilley, J. C. Gallardo, S. C. Gottschalk,
K. P. Kusche, R. H. Pantell, I. V.
Pogorelsky, D. C. Quimby, J. Skaritka, L.
C. Steinhauer, V. Yakimenko, and F.
Zhou, Phys. Rev. Lett. 92, 054801 (2004).
[2] I. V. Pogorelsky, I. Ben-Zvi, T.
Hirose, S. Kashiwagi, V. Yakimenko, K.
Kusche, P. Siddons, J. Skaritka, T.
Kumita, A. Tsunemi, T. Omori, J.
Urakawa, M. Washio, K. Yokoya, T.
Figure 1. Simulated spectral- and temporalmodulation of a 5-ps Gaussian CO2 laser pulse during
1000- times energy amplification.
Figure 2. Picture and schematic of the principles of
the experiment setup.
Figure 3. Proton- and ion-traces on a CR39 plate (b)
obtained with a Thomson parabola (a), and a trace fit to
the simulated curves (c).
Figure 4. Proton energy spectrum.