|
5fs high power OPCPA
laser for attosecond pulse production
23/01/2006 - P.
Bates
Contents
Introduction
Fig. 1:Comparison of OPA gain spectrum
vs Ti:Sa and hollow fibre broadened Ti:Sa spectra
 |
Recently there has been growing interest in the generation of intense
femtosecond pulses with durations in the few cycle regime. Optical
Parametric Chirped Pulse Amplification (OPCPA) provides a method for
scaling ultrashort pulses to much higher energies than are available
directly from conventional CPA laser systems. In OPCPA, a short pulse
is chirped before being amplified parametrically in a non-linear crystal.
OPAs have extremely large gain bandwidths that can allow amplification
of the full bandwidth of a 5 fs pulse, removing the need for post-amplification
spectral broadening and enabling orders of magnitude higher pulse
energies. Minimal thermal deposition in the OPA medium allows for
high average power operation and high optical quality due to the absence
of thermal lensing.
High average power OPCPA requires a high average and peak power pump
laser and efficient conversion of pump to signal. We are building
a diode-pumped Nd:YLF amplifier which will produce pulse energies
up to ~200mJ with pulse durations of ~40ps, at repetition rates up
to 1 kHz.
We have designed a stretcher and compressor system capable of stretching
a pulse with 400 nm bandwidth out to 20 ps and then recompressing
down to 5 fs. The 20 ps chirped pulse will be amplified in two broad
bandwidth OPA stages.
When complete, this system will be integrated into the attosecond
XUV beamline at Imperal college, enhancing the capabilities of that
experimental facility.
High average power pump laser
While the pulse energy and peak intensity of the pump laser are
not insignificant, the predicted average powers of close to 200W have
been the more stringent specifications and have dictated certain elements
of the laser design. We have chosen ND:YLF as the laser medium due
to its low non-linear index, low thermal lensing and natural birefringence.
The laser takes the form of a front end 1 kHz Nd:YLF oscillator providing
1 mJ 40 ps pulses at 1047 nm, followed by a home-built diode pumped
multipass amplifier.
Fig. 2:The multipass ND:YLF amplifier
 |
Detailed analysis of the operation of the amplifier has been addressed
to enable achievement of a maximum pulse energy without exceeding
the limits set by self-focusing and thermal fracture. The final design
of our amplifier uses five, 1 kW laser diode arrays from Dilas to
pump a 100 x 5 mm deep etched Nd:YLF rod at 15 W cm-1.
Recent tests on the complete amplifier system, with 3-pass amplification
and a non-optimum beam diameter, have produced output pulse energies
of up to 130mJ Further work is under way to achieve the design energy
of >200 mJ.
Optical parametric chirped pulse amplification
Extensive modelling has be undertaken to optimise the design of our
OPCPA stages. The overall system must achieve a saturated gain of
~107 with a reasonably flat profile over a 400nm bandwidth.
Most modelling to date has centred around a two stage amplifier setup
with single-pass small signal gain of ~106 in the first
amplifier and 104 in the second amplifier.
Our simulations show that LBO is a good candidate for the OPA medium,
with a gain bandwidth of 700-1150 nm, corresponding to a pulse duration
of 5.8fs. Predicted extraction efficiency from the pump is close to
30%, which will give few cycle pulses with energies up to 10mJ after
compression.
We are now looking in detail at the exact modelling of our current
setup, and considering adding a third stage to the amplification setup.
We will have initial data from gain testing by the end of February
2006, and this will further guide our OPA stage design.
Stretcher/Compressor development
We have developed a design for a stretcher compressor system capable
of stretching an 850 nm pulse with 200 nm FWHM (~400 nm edge-to-edge)
bandwidth to 20 ps and recompressing to 5 fs. To achieve the fine
balance of positive and negative dispersive elements we have designed
a system which includes a prism stretcher, grating stretcher, grating
compressor and a programmable acousto-optic phase filter (Dazzler).
Testing of this system has shown we can stretch the unamplified pulses
from our current oscillator to 10ps (bandwidth 250nm edge to edge)
and recompress them to within 0.5fs of their 11 fs transform limit.
Fig. 3:One of our transmission gratings
and the grating stretcher
 |
Novel features of this design include the use of transmission gratings
in both the stretcher and compressor which have extremely high efficiency
over the full bandwidth of the pulses. Also, we have chosen to use
an 8 element custom designed lens in the stretcher, as it is extremely
difficult to design a reflective stretcher with low chromatic aberrations
over our large bandwidth.
Future Work
Work will continue on developing and building the full OPCPA laser
system, with amplified low rep rate pulses available by the end of
April 2006. As these will be from our 11fs oscillator we will then
work on producing a spectrum broad enough to seed the full gain bandwidth
of the OPA stages, while in tandem with this the pump laser is brought
up to full rep rate operation. As soon as the system is operational
it will be moved to Imperial College and integrated into the beamline
there as a drive laser for high harmonic generation to produce attosecond
pulses.
Publications and conference presentations
arising from this work
[1] A 5 fs OPCPA laser system for attosecond pulse production
E Springate, Y Tang, P Bates, IN Ross, RA Smith, JWG Tisch and JP
Marangos, CLEO Europe, Munich, Germany (June 2005).
[2] A 5 fs high average power OPCPA laser system for attosecond
pulse production P Bates, E, Springate, Y Tang, IN Ross, RA Smith,
JWG Tisch and JP Marangos, UK High Power Laser Users Meeting,
Abingdon, UK, (Dec 2005).
[3]
High-power Nd:YLF Amplifier Development , Y Tang, M Divall,
IN Ross, E Springate and GJ Hirst, CLF Annual Report , (2004
- 2005)
[4]
Design of a Stretcher-Compressor System for High Energy 5 fs Pulses
, P Bates, E Springate, Y Tang, IN Ross, CLF Annual Report
(2004 - 2005.)
|