High signal-to-noise ratio, single-frequency 2 μm

2626
OPTICS LETTERS / Vol. 39, No. 9 / May 1, 2014
High signal-to-noise ratio, single-frequency 2 μm
Brillouin fiber laser
Yongfeng Luo,1 Yulong Tang,1 Jianlong Yang,1 Yao Wang,1 Shiwei Wang,1 Kunyu Tao,2 Li Zhan,1 and Jianqiu Xu1,*
1
Key Laboratory for Laser Plasmas (MOE), Shanghai Jiao Tong University, Shanghai 200240, China
2
The 802nd Research Institute Academy of Spaceflight Technology, Shanghai 200090, China
*Corresponding author: [email protected]
Received January 14, 2014; revised March 25, 2014; accepted March 26, 2014;
posted March 27, 2014 (Doc. ID 204709); published April 22, 2014
A high optical signal-to-noise ratio (OSNR) single-frequency 2 μm Brillouin fiber laser (BFL) with watt-level output
and high transfer efficiency is demonstrated for the first time to the best of our knowledge. The Brillouin pump is
constructed with a two-stage thulium-doped fiber amplifier (TDFA) seeded by a 2 μm laser diode, providing 4.02 W
average power with 1 MHz linewidth. Using an optimized length of 14 m for the Brillouin ring cavity, the BFL works
stably in single-mode region with 8 kHz linewidth because of the linewidth narrowing effect. The transfer efficiency
is 51% with 1.08 W output power and 62 dB OSNR for 3.22 W pump power. © 2014 Optical Society of America
OCIS codes: (060.3510) Lasers, fiber; (140.3510) Lasers, fiber; (260.3060) Infrared; (290.5900) Scattering, stimulated
Brillouin.
http://dx.doi.org/10.1364/OL.39.002626
Single-frequency 2 μm wavelength lasers are of special
importance for applications such as gas detection,
free-space optical communication, and long-range LIDAR
[1], because the wavelength is within the absorption peak
of vapor and the atmospheric transparency window.
A high optical signal-to-noise ratio (OSNR) and narrow
linewidth single-frequency 2 μm lasers are urgently
needed to achieve high detection ranges and resolutions.
To date, different configurations have been reported
for single-frequency fiber lasers operating near 2 μm,
including distributed feedback (DFB) cavity [2,3] and
distributed Bragg reflector (DBR) configuration [4].
Currently, the output power and slope efficiency are still
limited. A DFB single-fiber laser operating at 1.93 μm
with 22 mW maximum output power has been demonstrated in Ref. [2]. The slope efficiency was 10%, which
was limited by the concentration of the thulium-doped
fiber (TDF). The linewidth of this DFB laser was estimated to be between 20 and 50 kHz. A core-pumped
DBR fiber laser operating at 1.95 μm with 3 kHz linewidth
has been demonstrated. The maximum output power was
40 mW for 500 mW pump power with a very low slope
efficiency of 8% [4]. When pumped into the outer clad,
the maximum power was increased to 100 mW, and there
was a mode hop when the pump power was increased
farther [4]. Still, the maximum output power was limited
by the damage threshold of the fiber Bragg grating. The
linewidths of both configurations were limited by the grating and the cavity length used to select single mode. Due
to the thermal fluctuations of the grating and the phase
noise of the pump, it is difficult to achieve a very high
OSNR single-frequency laser.
The use of BFLs has proven to be a potential approach
to achieving a single-frequency laser with high OSNR and
efficiency below the 1.5 um wavelength range [5–8].
However, the performance of BFL is dubious for a
2 μm wavelength, because the Brillouin gain linewidth
is inversely proportional to λ2 [9], beneficial to achieve
a much narrower bandwidth; on the other hand, the
Brillouin gain is low for 2 μm laser due to the high
absorption loss of silica fiber at 2 μm. In this case, a
0146-9592/14/092626-03$15.00/0
brightness pumping source and relatively long fiber
are required for high efficiency, high OSNR, and narrow
linewidth of 2 μm BFLs. In this Letter, we develop a highpower Tm-fiber laser with relatively broad bandwidth
(∼1 MHz) as the pump source, and demonstrate a wattlevel high OSNR single-frequency BFL operating at 2 um
with a long cavity for the first time to the best of our
knowledge.
Figure 1 shows the experimental configuration of the
TDFA pumped BFL, which consists of a fiber-coupled
2 μm DFB laser with linewidth less than 1 MHz and output power of 2 mW, a two-stage TDFA, and a BFL cavity.
The first-stage amplifier consists of a 793 nm fibercoupled diode laser with a maximum output power of
6 W, a 2 1 × 1 combiner, 2 m PM-TDF-10P/130-HE
TDF, while the second stage amplifier has the same configuration except for two pump diode lasers with a total
of 12 W maximum output power. A 2 μm polarizationdependent isolator is inserted between the two amplifiers. From the first and second stage of TDFA, the laser
seed is amplified to 120 mW and 4.02 W, respectively.
Then the output from the TDFA is used as the Brillouin
pump (BP) which is injected into the Brillouin singlepass ring cavity clockwise. The BFL cavity consists of
a polarization controller (PC), an optical circulator with
20% power loss from port 1 to 2 and port 2 to 3, and a
Fig. 1. Experimental configuration of the BFL. ISO, isolator;
OC, optical circulator; TDF, thulium-doped fiber; and SMF,
single mode fiber.
© 2014 Optical Society of America
May 1, 2014 / Vol. 39, No. 9 / OPTICS LETTERS
60/40 polarization independent coupler, in which the recoupling coefficient is 40%. The cavity length is optimized
to 14 m, corresponding to 15 MHz free spectral range.
Since the Brillouin gain spectrum is narrower in 2 μm
wavelength, the cavity length can be longer comparatively to improve the transfer efficiency. Although the
threshold of this nonresonant BFL is higher than that
of the resonant ones, the simplicity and stability are
appropriate for highpower single-frequency lasers.
The Brillouin Stokes (BS) output power and residual
BP as functions of the pump power are shown in Fig. 2.
Here, the pump power is measured from port 2 of the OC.
The threshold power of the pump is 1.04 W. The output
power of the BFL reaches 1.08 W at 3.22 W pump power
without gain saturation. Considering 20% loss of the OC,
the transfer efficiency from the input to the output is near
26%. The laser has a slope efficiency of 51% beyond the
threshold, which is comparatively lower than that obtained at 1.5 μm [8]. This indicates that the SBS gain is
lower at 2 μm than at 1.5 μm. A higher slope efficiency
can be achieved with a low recoupling coefficient coupler but higher threshold. The PC is used to control
the polarization state of the Stokes. In the experiment,
the output power fluctuations from adjusting the PC
can be ignored because the cavity fiber is carefully coiled
without twisting to reduce stress-induced birefringence.
The calculated threshold in the single-pass cavity is
938.7 mW, which approaches the experimental measurement (1.04 W) [8,10]. The discrepancy might result from
the mismatch between the cavity mode and the peak of
the Brillouin gain.
In the BFL, the pump fluctuation is smoothed by the
stimulated Brillouin scattering. Furthermore, the noise
in the pump is naturally separated from the Stokes wave
due to the opposite translating direction of the pump and
the Stokes wave. The ultrahigh OSNR can be achieved in
the BFL as shown in Fig. 3, where the OSNR is as high as
62 dB. The peak wavelength of the BFL and pump wave
are 2003.26183 and 2003.14599 nm, respectively. The
frequency shift between them is 8.66 GHz.
The linewidths of the BP and BS are measured with the
beat method.
First, we beat the BP with the BS. The beat frequency
of the BP and the BS is centered at 8.38 GHz, which is
frequency difference between the BP and the BS. The
Brillouin frequency shift measured with spectrometer
and the beat method differs by 0.0037 nm, which is in
the range of experimental error. The width (FWHM) of
Fig. 2.
Output power as the function of pump power.
Fig. 3.
2627
Spectrum of the BFL.
the beat frequency equals the sum of the BP and the
BS linewidth. Because the linewidth of the BS is far
narrower than that of the BP [11], the linewidth of the
BP can be retrieved from the width of the beat frequency.
Second, we measure the linewidth of the BS by constructing an additional BFL. The output of the first
BFL is injected into the second BFL as the Brillouin
pump, and benefits from the high power and high transfer
efficiency of the BFL. The BS linewidth of the second
BFL is estimated to be only several tens of hertz, which
is much narrower than that of the first BFL. Then the
beating between the first and second BFL reveals the
linewidth of the first BFL.
Figure 4 shows the spectrum of the beat frequency for
(a) the Brillouin pump and (b) the Brillouin Stokes. After
two-stage amplification, the linewidth of the Brillouin
pump is still less than 1 MHz. The broadening by the
amplified spontaneous emission (ASE) in the TDFA, is
several tens of kilohertz [12], which is less than several
percent of the linewidth of laser seed. Because the
linewidth of the BS is much narrower than that of the
Fig. 4. Spectrum of beat frequency for the (a) BP and (b) BS.
CF: center frequency.
2628
OPTICS LETTERS / Vol. 39, No. 9 / May 1, 2014
Fig. 5. Power fluctuation in 3 h at 1 W output power.
BP due to the linewidth narrowing effect [6], the broadening caused by amplification of the TDFA is negligible in
the finial output.
As shown in Fig. 4(a), there is only one peak of beat
frequency in the 160 MHz sweep range. Because the
beating is carried between the BP and the BS, the
multi-longitudinal mode of the BS results in multiple
peaks in the beat frequency. The single peak in the beat
frequency indicates the BFL works in single-longitudinal
mode. Because the free spectral range of the BFL cavity
is 15 MHz, if the second-longitudinal mode occurred, its
beat frequency peak should appear in position, which
apart from the central mode is 15 MHz away as marked
in Fig. 4(a) with A and B.
The linewidth of the BS is proportional to the linewidth
of the BP by the factor 1∕K [6], in the form of
Δνs Δνp ∕K, where Δνs is the linewidth of the BS,
Δνp is the linewidth of the BP, and K 1 πΔνB ∕
−c ln R∕nL2 . For the 2 μm wavelength laser, the
FWHM of the gain curve of SBS ΔνB 15 MHz is narrower than that in 1.5 μm wavelength. c is the speed
of light in vacuum. n 1.45 is the refractive index of
fiber core. R 0.32 is the recoupling coefficient of the
cavity. L 14 m is the cavity length. We can calculate
the coefficient K to be 14.5 and the linewidth of the
BS to be 69 kHz when the linewidth of the BP is
1 MHz. However, the linewidth of the BS is also modified
by the cavity loss and the matching condition between
the SBS gain peak and eigenmode frequency. Due to
the large cavity loss, long cavity length, and low SBS gain
in the 2 μm wavelength range, the BFL working in 2 μm
wavelength shows much narrower linewidth than the
shorter wavelength range. We achieved the linewidth
of 8 kHz for the BS output.
The stability of output power from the BFL is monitored. The ambient temperature varies around 2°C.
Whole laser is cooled by the air convection and no
special temperature control is applied. The RMS
fluctuation is 1.35% at the full output power of 1.07 W.
Figure 5 shows a typical measurement during 3 h. The
main fluctuation comes from the temperature variety.
Applying temperature control to the laser should reduce
the fluctuation of output power greatly.
In conclusion, we have demonstrated a singlefrequency 2 μm fiber laser, which can achieve a high
OSNR, high power, high transfer efficiency and very narrow linewidth simultaneously. The laser system includes
a two-stage TDFA seeded from a 2 μm diode, and a BFL
with a long nonresonant cavity pumped with the TDFA.
The 2 μm single-frequency laser has generated 1.08 W
average power with 62 dB OSNR and 8 kHz linewidth.
As narrow as several tens of hertz linewidth can be
achieved with the second BFL. It is believed that this high
OSNR single-frequency 2 μm BFL might be applied in
long-range wind LIDAR-measurements and free-space
optical communication.
This work was supported by the National Natural
Science Foundation of China (Grant Nos. 61275136
and 61138006), the Research Fund for the Doctoral
Program of Higher Education of China (Grant
No. 20120073120085), and the State Key Program for
Basic Research of China (Grant No. 2010CB630703).
References
1. K. Scholle, E. Heumann, and G. Huber, Laser Phys. Lett. 1,
285 (2004).
2. D. Gapontsev, N. Platonov, M. Meleshkevich, O.
Mishechkin, O. Shkurikhin, S. Agger, P. Varming, and J.
H. Poylsen, Lasers and Electro-Optics, CLEO 2007, Baltimore, Maryland (2007), pp. 1–2.
3. S. Agger and J. H. Povlsen, Opt. Lett. 29, 1503 (2004).
4. J. Geng, Q. Wang, T. Luo, S. Jiang, and F. Amzajerdian, Opt.
Lett. 34, 3493 (2009).
5. J. Geng and S. Jiang, Opt. Lett. 32, 11 (2007).
6. A. Debut, S. Randoux, and J. Zemmouri, Phys. Rev. A 62,
023803 (2000).
7. S. P. Smith, F. Zarinetchi, and S. Ezekiel, Opt. Lett. 16, 393
(1991).
8. G. Wang, L. Zhan, J. Liu, T. Zhang, J. Li, L. Zhang, J. Peng,
and L. Yi, Opt. Lett. 38, 19 (2013).
9. L. F. Stokes, M. Chodorow, and H. J. Shaw, Opt. Lett. 7, 509
(1982).
10. Y. Azuma, N. Shibata, T. Horiguchi, and M. Tateda,
Electron. Lett. 24, 250 (1988).
11. F. Mihlic, D. Bacquet, J. Zemmouri, and P. Szriftgiser, Opt.
Lett. 35, 432 (2010).
12. G. J. Cowle, P. R. Morkel, R. I. Loming, and D. N. Payne,
Electron. Lett. 26, 424 (1990).