Conference PaperPDF Available

Looped Time-Lens Compression for Generation of 3.5 nJ Femtosecond Pulses from a CW Laser

Authors:

Abstract

We generate 516 fs pulses at 3.5 nJ energy from a continuous wave 1.55 mum source without mode-locking. Our system is compact, all-fiber, and allows continuous tuning of pulse width and center wavelength.
Fig. 1. Experimental setup. DFB: distributed feedback laser, MZ: Mach-Zehnder modulator, PC: pulse carver, EDFA: erbium doped fiber
amplifier, PP: pulse picker, EDWA: erbium doped waveguide amplifier, BPF: band pass filter, PM: phase modulator, M: mirror, G: grating.
Looped time-lens compression for generation of 3.5 nJ
femtosecond pulses from a CW laser
James van Howe, Jennifer Lee, and Chris Xu
School of Applied and Engineering Physics, Cornell University, Ithaca, NY 14853
jwv9@cornell.edu
Abstract: We generate 516 fs pulses at 3.5 nJ energy from a continuous wave 1.55 µm source
without mode-locking. Our system is compact, all-fiber, and allows continuous tuning of pulse
width and center wavelength.
© 2007 Optical Society of America
OCIS Codes: (320.7090) Ultrafast lasers; (320.5520) Pulse compression; (060.2380) Fiber optics sources
We demonstrate time-lens compression in a loop configuration that enables us to obtain unprecedented energetic
femtosecond pulses from a CW laser without mode-locking. A time-lens refers to a device that imposes a quadratic
phase in time onto an incoming pulse or temporal profile just as a spatial lens imposes a quadratic phase onto a
spatial profile in space [1]. In our work, a sinusoidally driven, fiber-integrated, LiNbO3 electro-optic phase
modulator serves as the time-lens, where the cusp of the sinusoid is used to approximate the proper quadratic phase.
Just as a spatial lens can expand or focus a beam in space, so can a time-lens broaden or compress a pulse in time.
Time-lens compression using integrated electro-optic modulators was shown in the past to generate picosecond
pulses by applying ~1 W of sinusoidal RF modulation [1-3]. Hero experiments in which kilowatts of RF power were
applied to a bulk modulator achieved 550 fs pulses at 514 nm but with limited pulse energy (~ pJ) and fidelity [4].
More practical time-lens pulse generators have been limited to compressed pulse widths no less than 2 ps. More
recently, time-lenses have been combined with nonlinear compression techniques to surpass this limit [5, 6], though
with the disadvantage of performance being dependent upon optical power. Our method for pulse generation
circumvents these issues by using one time-lens driven at reasonable RF power (~1 W) in a loop to emulate many
stacked lenses. By effectively stacking time-lenses we achieve 516 fs pulses at 3.5 nJ energy starting from a CW
1.55 µm input. Our system is compact, all-fiber, and allows continuous tuning of pulse width and center wavelength.
The experimental setup is shown in Fig. 1. The femtosecond pulse generator consists of a seed source shown
before point A, a time-lens loop shown between points A and B where chirped optical bandwidth is generated,
followed by an amplification and compression stage beyond point B. The general operating principle is to allow
pulses to circulate the loop N number of times where they will acquire bandwidth for every pass of the time-lens.
After generating the desired bandwidth, pulses are ejected from the loop, amplified and then dechirped. The seed
source begins with 20 mW from a CW DFB laser at 1.55 µm which is pulse carved into a 33 % duty cycle 10 GHz
pulse train by a Mach-Zehnder modulator. Pulses are amplified and then picked by an intensity modulator to a lower
duty cycle that is an integer multiple of loop lengths. To allow only one pulse to circulate the loop at a time, the
input seed source must have a repetition rate equal to that of Loop
f
N. A pulse pattern generator clocked at 10
GHz is used to drive the pulse picker as well as input and output switches in the loop. Pulses are amplified by an
erbium-doped waveguide amplifier (EDWA) which gives amplification over a very short length and then are filtered
to clean up amplified spontaneous emission (ASE) noise. Keeping a short optical length between components
facilitates matching the timing of RF to optical signals. Next, pulses are injected into the loop. The electrical gates
which open and close the input and output loop switches are delayed such that one pulse enters the loop as the
CW DFB
EDWA
PM
To Autocorrelator
G2
M1
G1
M2
MZ
MZ
1x2 Switch
Pulse Pattern
Generator
5 GHz x2
MZ MZ
EDWAEDFA
EDFA
PPPC BPF
BPF
1x2 Switch
A
B
Loop
Circulation
Time-lens
~
~
~
~
~
~
CW DFB
EDWA
PM
To Autocorrelator
G2
M1
G1
M2
MZ
MZ
1x2 Switch
Pulse Pattern
Generator
5 GHz x2
MZ MZ
EDWAEDFA
EDFA
PPPC BPF
BPF
1x2 Switch
A
B
Loop
Circulation
Time-lens
~
~
~
~
~
~
~
~
~
~
~
~
a2799_1.pdf
CFF2.pdf
1535 1540 1545 1550 1555 1560 1565 1570
-50
-40
-30
-20
-10
0
Wavelength (nm)
(a) (b)
-15 -10 -5 0 5 10 15
0
0.2
0.4
0.6
0.8
1.0
Delay (ps)
IAC FWHM=697 fs
Normalized Intensity
1535 1540 1545 1550 1555 1560 1565 1570
-50
-40
-30
-20
-10
0
Wavelength (nm)
(a) (b)
-15 -10 -5 0 5 10 15
0
0.2
0.4
0.6
0.8
1.0
Delay (ps)
IAC FWHM=697 fs
Normalized Intensity
Fig. 2 (a) Experimental optical spectrum of pulses ejected from the time-lens loop for nine trips. Inset shows the calculated optical spectrum.
Both traces were taken at 0.2 nm resolution bandwidth. (b) Experimental intensity autocorrelation with a pulse width of 697 fs. Right-hand inset
shows ex
p
erimental interferometric trace. Lef
t
-hand inset shows calculated intensit
y
autocorrelation with a
p
ulse width of 600 fs.
previous exits. Inside the loop resides another EDWA to account for loss of the switches and phase modulators
(~17 dB total) as well as a filter to clean up ASE. Note that one phase modulator is drawn for clarity, though there
are actually two modulators inside the loop. Each modulator is driven at approximately 1.0 W of RF power by a 10
GHz sinusoid that is frequency-locked to the low-duty cycle repetition rate of the seed. The total phase modulation
is approximately 10 π radians per pass. In this work we allow pulses to circulate the loop nine times (eight passes
through the phase modulator) in order to achieve 80 π radians phase shift. The fundamental loop frequency is
approximately 28 MHz (7 m of fiber) so that nine loops corresponds to roughly a 3 MHz seed frequency. The input
and output switches in the loop serve to further clean up unwanted energy between pulses. After ejection from the
loop the pulses are amplified to 11 mW and compressed by a grating pair which gives approximately 2.0 ps2 of
anomalous dispersion. The entire loop is made of polarization maintaining components to provide superb stability.
Figure 2 (a) shows the experimental spectrum generated after nine trips through the loop. Each pass through the
modulator builds 2.5 nm of bandwidth for a total of 20 nm. The excellent match from the calculated trace shown in
inset demonstrates the predictability of our system. The expected square top feature of the spectrum is indicative of
aberration in the phase drive from the ideal quadratic. Figure 2(b) shows the experimental second-order intensity
autocorrelation with the experimental interferometric trace shown in the right inset and calculated intensity
autocorrelation in the left inset. The intensity autocorrelation gives a pulse width of 697 fs. Taking into account the
deconvolution factor calculated for this pulse shape gives a final width of 516 fs. Once again we note the remarkable
quantitative and qualitative match between experimental and calculated traces. Currently, our pulse energy is limited
by the nonlinearity generated in the output amplifier. The pulse exiting the amplifier onto the gratings is
approximately 3.5 nJ with a 15 ps pulse width. Note that the pulses exiting the loop are slightly compressed from
the initial 33 ps loop input due to non-negligible anomalous dispersion inside the loop while the pulses accumulate
bandwidth. For larger amplification, chirping the pulse even more via dispersion compensating fiber may be
necessary to prevent deleterious effects from nonlinearity. Additionally, changing the chirping sign of the time-lens
so that normal dispersion would be required for dechirping may increase the system tolerance to interactions with
nonlinearity. Regardless, the clean output pulses obtained in our experiments show promise that our system could be
easily incorporated into a chirped amplification system to achieve even larger pulse energy. We note that the pulse
quality can be further improved using a correction time-lens to correct aberration in the non-ideal phase drive [1].
In summary we demonstrate an all-fiber system for generating femtosecond pulses of a few nJ from CW light
without mode-locking. The flexibility of this new system allows continuous tuning of bandwidth by electronically
changing the number of round trips through the loop (the number of picked pulses), adjusting the phase modulator
drive, or by offsetting the phases of loop modulators. The center wavelength can be changed by simply switching
the input CW DFB source of the seed laser within the limits of the electro-optic components and amplifiers.
Integrated fiber switches, phase modulators and amplifiers are readily available at 1.06 µm and 1.3 µm where
spectral bandwidth is even easier to generate due to the lower V
π
of the phase modulator. We anticipate that further
improvements to this robust, compact, all-fiber system will contribute to true turn-key ultrashort laser systems.
[1] J. van Howe and C. Xu, J. Lightwave Technol., vol. 24, pp. 2649-2662, (2006).
[2] B. H. Kolner, Appl. Phys. Lett., vol. 52, pp. 1122-1124, (1988).
[3] A. A. Godil and D. M. Bloom, Appl. Phys. Lett., vol. 62, pp. 1047-1049, (1993).
[4] T. Khayim, M. Yamauchi, D. Kim, et al., IEEE J. Quantum Electron., vol. 35, pp. 1412-1418, (1999).
[5] M. Hanna, P. Lacourt, S. Poinsot, et al., Opt. Express, vol. 13, pp. 1743-1748, (2005).
[6] C. B. Huang, Z. Jiang, D. E. Leaird, et al., Electron. Lett., vol. 42, pp. 1114-1115, (2006).
a2799_1.pdf
CFF2.pdf
ResearchGate has not been able to resolve any citations for this publication.
Article
A resonant microwave optical phase modulator in LiNbO 3 with multiple passes, based on an off‐axis path in a stable optical resonator, is demonstrated as a time‐lens. With 1 W of cw microwave power at 5.2 GHz, 45 ps pulses at 1.06 μm were temporally focused to 6.7 ps (FWHM). Increasing the drive power to 13 W, at 10% duty cycle, produced 1.9 ps pulses. The aperture of the time‐lens is about 31 ps. This is the first demonstration of a useful time‐lens.
Article
We describe a pulse compression technique that uses integrated electro‐optic phase modulators to linearly chirp optical pulses for compression by a dispersive delay line. In contrast to passive chirp techniques such as self‐phase modulation, this approach of active pulse compression does not depend on the optical power and thus shows promise for application to low‐power solid‐state lasers as an alternative to mode locking for realizing compact picosecond sources. We have demonstrated this technique by compressing the pulses from a mode‐locked neodymium:yttrium aluminum garnet laser from 100 to 45 ps.
Article
A simple, stable, and tunable optical pulse source emitting 3.2 ps pulses at a 10 GHz repetition rate is presented. The pulses are obtained through soliton-assisted time-lens compression in a standard single mode fiber, and are fully characterized by wavelength-conversion frequency-resolved optical gating. The use of nonlinear effects relaxes the constraint of the high driving voltage on the phase modulator usually required in this type of source.
Article
We report a novel electrooptic phase modulator featuring lens modulation and its application to the electrooptic chirping compression method for ultrashort optical pulse generation. In the pulse generation experiment, pedestal free optical pulses of 550 fs in width with a 16.25-GHz repetition rate were obtained by the group delay dispersion of a grating pair from a CW Ar laser. Furthermore, the electrooptic chirping compression method with the electrooptic phase modulator featuring lens modulation is applicable to the negative group delay dispersion as well as positive group delay dispersion
  • M Hanna
  • P Lacourt
  • S Poinsot
M. Hanna, P. Lacourt, S. Poinsot, et al., Opt. Express, vol. 13, pp. 1743-1748, (2005).
  • T Khayim
  • M Yamauchi
  • D Kim
T. Khayim, M. Yamauchi, D. Kim, et al., IEEE J. Quantum Electron., vol. 35, pp. 1412-1418, (1999).