Expand this Topic clickable element to expand a topic
Skip to content
Optica Publishing Group

Gain switch laser based on micro-structured Yb-doped active fiber

Open Access Open Access

Abstract

In this paper a near infrared gain-switched fiber laser based on oscillator stage only design with high peak power is presented. Output pulses reached 2.3 kW of peak power and duration of less than 60 ns. The dependence of the laser pulse duration on operation parameters was measured and theoretically explained. As the setup is based on flexible micro structured single polarization fiber, the laser output exhibits high polarization extinction ratio. Due to the narrow output spectrum the setup is suitable for second harmonic generation.

© 2014 Optical Society of America

1. Introduction

The method of gain switching in pulsed fiber lasers continues to be of interest mainly due to the progress of high power single emitter based laser diode modules. They can provide enough pump power and can be switched on and off fast enough for laser gain switching. Using gain switched approach simple and robust lasers with output pulses in nanosecond region can be built. Regarding simplicity and robustness such lasers can easily compete with the commonly used approach based on Q-switching which usually uses AOM modulators [1] or another alternative technique [2] and can replace them in applications such as metal micromachining and super continuum generation [3].

In the most commonly used wavelength range (near 1064 nm) one of the key issues of the gain switched approach is how to achieve peak power that is high enough together with the desired short pulse duration by using only the oscillator stage, without any amplifying stages.

The gain medium that is typically used in this wavelength range for high power fiber lasers is ytterbium. There are several reports on achieving gain switched operation with ytterbium doped fibers. The first lasers using only oscillator stage had laser pulse duration in the range of 500-100 ns. In 2006 Maryashin et al. [4] achieved laser pulses with duration of 500 ns (tLp) and peak power (PLp) of 120 W, pumped with pump pulse power (Pp) of 20 W. Three years later pulse duration of 125 ns and peak power of 120 W with Pp of 34 W was reported [5]. Another example from 2011 is presenting pulses with duration of 200 ns and peak power of 700 W, where pump pulse power is 100 W [3]. Recently laser pulses under 100 ns were achieved, for example a pulse duration of 90 ns was reached with a pump pulse power of 82 W and corresponding peak power 800 W [6] and even shorter pulse duration of 66 ns and peak power of 700 W with pump pulse power of 210 W [7].

In all of the above mentioned publications the peak power of laser pulses never exceeded 1 kW from oscillator stage only. The mayor improvement was presented in [8] where pulses as low as 27 ns and peak powers up to 2 kW were reached at 330 W pump pulse power. In this paper we present a laser setup using oscillator stage only that uses special fiber to simplify the construction. The laser pulses so achieved had peak power up to 2.3 kW with only 260 W of pump pulse power needed. To the best of our knowledge this is the highest peak power reported from only oscillator stage gain switched fiber laser operated in the 1 µm wavelength range. In our setup we use output wavelength around 1030 nm instead of standard 1064 nm, to keep laser pulses short while increasing the laser active volume to achieve higher peak power. Also the special design of the fiber allows single polarization operation with high extinction ratio.

2. Gain switching

The method of gain switching for producing laser pulses is based on the relaxation oscillations of the photons ϕ inside the oscillator, as they interact with excited ions N2 of the upper laser level, when the pump is switched on. This interaction between the two populations can be described by the rate equations:

N2t=Γσc0Vns(N2N1)ϕN2τ21+w,
ϕt=Γσc0Vns(N2N1)ϕϕτL+βN2τ21.

Equations (1) and (2) are coupled by the first term, describing stimulated emission stating that its rate is proportional to the stimulated emission cross section σ, inversion population N2-N1and overlap integral Γ between laser photons and the fiber core and inversely proportional to the interaction volume V. Speed of light is c0, ns is refractive index of the silica, τL is the laser cavity photon time of flight and τ21 is the upper laser level relaxation time. The latter governs the constant loss rate of excited ions from the upper laser level. Spontaneously relaxed ions emit photons of which only a part is guided in the active fiber core. This part is denoted with β, also called the coupling factor, and is the ratio between the number of spontaneously emitted photons which are coupled into the core and the number of all the spontaneously emitted photons [9].

The pumping rate w states how many ions are excited per unit time and is directly proportional to the absorbed pump power through the quotient of pump wavelength λp and Planck constant h times speed of light:

w=λpPabshc0=λphc0(110αL/10)PP.
The absorbed pump power was expressed in term of absorption coefficient α, length of the fiber L and pump pulse power PP. The last parameter is the most important one in gain switching. If it is switched off after the number of laser photons first reaches the maximum a single output laser pulse is generated, corresponding to the first overshoot of the relaxation oscillations. How intense this first peak of photons is depends on how much the two mentioned populations can oscillate form their steady state value. The magnitude of the pump rate and the stimulated emission terms in Eqs. (1) and (2) defines the characteristics of the laser pulse. The higher the pump rate, the shorter the laser pulse and higher its peak power. The stimulated emission cross section and the inverse interaction volume V have the same effect on pulse parameters as the pump rate

These relations were obtained by analyzing Eqs. (1) and (2) directly. A more precise relation can be acquired analytically by solving Eqs. (1) and (2) around steady state for small perturbations [10]. The result is the relation of laser pulse duration to the afore mentioned parameters [11]:

tLp2π(nsVhΓλp1σPabs)12.

It should be noted that the validity of this equation is related to the assumptions as follows. The optimal pump pulse must be as long as the laser pulse build up time to satisfy the small perturbation around steady state. This complies with the gain switched laser systems having the feedback loop, because the pump is switched off only after the onset of the laser output pulse. The pump pulse duration (tPp) is also related to the pump rate or pump pulse power as described in [6]:

tPpPabs1/2.

Considering the optimal pump pulse duration at the given pumping rate the dependencies can be expressed in terms of the supplied energy [12]. For non-optimal pump pulse durations which end before the laser pulse onset the first overshoot of the relaxation oscillations will have smaller amplitude or even would not occur if the pump pulse is too short. The smaller amplitude of relaxation oscillations means lower laser peak power.

Beside the absorbed pump pulse power in Eq. (4) other parameters influence the gain switched laser pulse duration. One is the interaction volume that is the product of the length of the oscillator (L) and the active fiber core cross section (Aco). This means that longer the oscillator, longer will be the pulses. Expanding the core will have the same effect. In this work we use micro structured active fiber with relatively large core diameter of 40 μm. Comparing this to the more standard values as for example 15 μm core diameter the laser pulses should be longer for more than factor 2.5 if other parameters were kept the same. In our case this is compensated by better absorption of the pump light which shortens the pulse duration.

3. Experimental setup

The key elements of the gain switched laser setup were active fiber and pumping system. The active fiber was double-clad single polarization type. The pump system was based on temperature stabilized 976 nm laser diodes controlled by high speed driver that enables rise time of a few tens of nanoseconds. The maximum peak power of the pump system was more than 300 W in the pulsed pumping mode. The experiments were made at fixed repetition rate of 50 kHz. The system used active feedback loop in order to detect the output pulse from the laser and consequently switch off the pump system. On one side of the laser resonator a grating was mounted while on the other side only Fresnel reflection from the fiber end face was used, as shown in Fig. 1.

 figure: Fig. 1

Fig. 1 A single polarization micro structured ytterbium doped fiber is used. It is pumped with high power 976 m pump diodes, temperature stabilized and controlled through the feedback loop.

Download Full Size | PDF

The fiber allowed only one polarization (slow axis) to be transmitted and amplified. It was micro-structured to achieve larger core area and at the same time retain single mode operation. Both pump and laser light were guided by microstructures. The outer cladding was designed to give the fiber a preferable coiling direction, to minimize stress from torsion. The fiber has a core diameter of 40 μm and inner cladding diameter of 200 μm. The low signal absorption of the pump light at 976 nm is 10 dB/m.

4. Results and discussion

Usually the gain switched ytterbium doped lasers operate at wavelengths around 1064 nm, which is convenient due to available fiber components and also offers less attenuation in ytterbium doped amplifiers because of the negligible reabsorption at this wavelength.

According to Eq. (4) shorter gain switched pulse durations can be achieved with higher stimulated emission cross section, Fig. 2. This parameter is wavelength dependent in a way that is the characteristic for the ytterbium gain media [13].

 figure: Fig. 2

Fig. 2 Measured a) laser pulse peak power and b) pulse duration at different wavelengths in strongly pumped fiber. Solid line is the theoretical prediction in this wavelength range.

Download Full Size | PDF

The maximum value for stimulated emission cross section for ytterbium emission peak is reached around 1030 nm (~0.6 pm2) and is more than twice the value compared to the value at 1064 nm. This means that the gain switched laser operating at this wavelength exhibits shorter laser pulses. The improvement in the shortening of the pulse was already reported in [6] where the numerical model was used to compare the influence of different parameters on laser pulse duration. To confirm this, a set of measurements was made using the laser wavelengths near 1030 nm. Using three pump pulse powers of 160 W, 230 W and 330 W for a 0.7 m long active fiber, the laser pulse peak power and duration were measured as shown in Fig. 2. The trend for pulse duration can be observed as predicted by Eq. (4) which is calculated and shown as the solid lines in Fig. 2(b). Using wavelength of 1040 nm instead of 1030 nm reflects in 12% longer pulses.

The change in laser pulse duration is not the only effect when changing the wavelength. The peak power is also affected. Moving away from the ytterbium emission peak, the transient of the relaxation oscillation is smaller, resulting in lower peak powers as shown in Fig. 2(a).

Three instances of simultaneous measurement of time dependence of the laser and corresponding pump pulse powers are shown for laser operation wavelengths of 1015 nm, 1030 nm and 1040 nm in Fig. 3. The power achieved at 1030 nm wavelength exhibits the highest peak. Because the stimulated emission at 1030 nm laser light has the fastest dynamics of the three, it also has the shortest laser pulse build up time and the corresponding pump pulse is the shortest.

 figure: Fig. 3

Fig. 3 Comparison of three laser pulses at different operation wavelengths of 1015 nm, 1030 nm and 1040 nm. The corresponding pump pulses with pump pulse power of 330 W are also shown.

Download Full Size | PDF

Regarding the pulse peak power and pulse duration the gain switched laser operating near 1030 nm has most favorable characteristics. For example, the laser pulse at 330 W pump pulse power had the peak power over 2 kW and pulse duration of 52 ns. The 0.7 m fiber offered adequate absorption of the pump light (81% for low signal) due to large active core.

In order to achieve even higher peak powers from the oscillator stage setup, longer active fiber was used. Using the same type of active fiber but 1.2 m long, the laser pulse duration and its peak power were measured in regard to pump pulse power at 1030 nm, Fig. 4.

 figure: Fig. 4

Fig. 4 Figure a) shows the measured laser pulse duration and peak power in dependence of pump pulse power. The data (circles) are connected with solid line corresponding to linear fit and inverse square root fit as obtained from theoretical predictions. In b) the laser pulses and pump pulses energies are shown at corresponding pump pulse powers.

Download Full Size | PDF

In accordance with Eq. (4), the inverse square root relation between laser pulse duration and pump pulse power is observed in Fig. 4(a). Therefore to decrease the pulse duration by a factor of two the absorbed pump pulse power should be increased four times. The peak powers of the output pulses are presented in the same figure. They increase approximately linearly with the pump power. In Fig. 4(b) the energy of the pump and laser pulses is presented together with the optical to optical efficiency taking the coupling losses into account. The lower efficiencies at smaller pump pulse powers are due to operation near threshold for stable pulsed operation.

The longer fiber (1.2 m) experiment was performed with pump pulse powers ranging from 40 W to 260 W. The latter value is lower than the highest pump pulse power used in the shorter fiber experiment. This is due to the specifics of the pump system and the maximum pump pulse duration that the system allows at given pump pulse power. Also the longer fiber needs longer pump pulses to achieve optimal pumping. The pump pulse duration which was automatically adjusted by the feedback loop decreased from 2.8 µs to 1.1 µs with increasing pump pulse power. The duration of the required pump pulse therefore depends on pump power, because in the case of higher power the condition for the relaxation oscillations in the laser resonators is met sooner due to faster pumping rate.

As mentioned, the active fiber allowed only one polarization to propagate without significant attenuation. Measurement of polarization extinction ratio for the 1.2 m long fiber is presented in Fig. 5. A non-standard measurement based on rotating high quality polarizer was used. The polarization extinction ratio greater than 21 dB was achieved. Together with the spectral width given as the full width at half maximum, which is less than 0.13 nm, the presented laser output offers an adequate source for second harmonic generation.

 figure: Fig. 5

Fig. 5 Transmitted amplitude in dependence of the rotation angle of the analyzer. The polarization contrast was greater than 21dB in 1.2 m long fiber.

Download Full Size | PDF

5. Conclusion

We have successfully demonstrated the design and operation of gain switched fiber lasers based on micro-structured fiber operating in the near infrared with high peak power and short pulse duration. The wavelength dependence of the gain switched pulse duration in ytterbium fibers is treated theoretically and experimentally and good agreement is found. The theory points out that the shorter pulses can be more readily achieved when the gain switched laser is working at wavelength close to the ytterbium emission peak.

To achieve the goal of high power single polarized gain switched laser pulses produced by a device based on oscillator stage only, the 1.2 m long active micro structured fiber was used. The special design of the fiber allowed a polarization contrast of 21 dB without any additional measures to achieve linearly polarized pulses. These have duration under 60 ns and peak power above 2.3 kW which is one of the highest reported values using single stage gain switched ytterbium doped fiber lasers. Combining the laser pulse characteristics, the polarization contrast and spectral width of the output of the presented gain switched laser offers a compact source for second harmonic generation.

Acknowledgment

This work was supported by Slovenian research agency – ARRS (project L2-4174).

References and links

1. M. Laurila, J. Saby, T. T. Alkeskjold, L. Scolari, B. Cocquelin, F. Salin, J. Broeng, and J. Lægsgaard, “Q-switching and efficient harmonic generation from a single-mode LMA photonic bandgap rod fiber laser,” Opt. Express 19(11), 10824–10833 (2011). [CrossRef]   [PubMed]  

2. R. Petkovšek, J. Saby, F. Salin, T. Schumi, and F. Bammer, “SCPEM-Q-switching of a fiber-rod-laser,” Opt. Express 20(7), 7415–7421 (2012). [CrossRef]   [PubMed]  

3. C. Larsen, D. Noordegraaf, P. M. W. Skovgaard, K. P. Hansen, K. E. Mattsson, and O. Bang, “Gain-switched CW fiber laser for improved supercontinuum generation in a PCF,” Opt. Express 19(16), 14883–14891 (2011). [CrossRef]   [PubMed]  

4. S. Maryashin, A. Unt, and V. P. Gapontsev, “10-mJ pulse energy and 200 W average power Yb-doped fiber laser,” in Fiber Lasers III (SPIE, 2006), 61020O.

5. Y. Sintov, M. Katz, P. Blau, Y. Glick, E. Lebiush, and Y. Nafcha, “A frequency doubled gain switched Yb3+-doped fiber laser,”Proc. SPIE 7195, 719529 (2009).

6. V. Agrež and R. Petkovšek, “Gain-switched Yb-doped fiber laser for microprocessing,” Appl. Opt. 52(13), 3066–3072 (2013). [CrossRef]   [PubMed]  

7. C. Larsen, M. Giesberts, S. Nyga, O. Fitzau, B. Jungbluth, H. D. Hoffmann, and O. Bang, “Gain-switched all-fiber laser with narrow bandwidth,” Opt. Express 21(10), 12302–12308 (2013). [CrossRef]   [PubMed]  

8. R. Petkovšek and V. Agrež, “Single stage Yb-doped fiber laser based on gain switching with short pulse duration,” Opt. Express 22(2), 1366–1371 (2014). [CrossRef]   [PubMed]  

9. A. Yariv and P. Yeh, “Amplified spontaneous emission,” in Photonics: Optical Electronics in Modern Communications (Oxford University, 2007), pp. 755–759.

10. W. Koechner, “Relaxation oscillations,” in Solid-State Laser Engineering (Springer, 2006), pp. 128–134.

11. L. A. Zenteno, E. Snitzer, H. Po, R. Tumminelli, and F. Hakimi, “Gain switching of a Nd+3-doped fiber laser,” Opt. Lett. 14(13), 671–673 (1989). [CrossRef]   [PubMed]  

12. M. Giesberts, J. Geiger, M. Traub, and H.-D. Hoffmann, “Novel design of a gain-switched diode-pumped fiber laser,” Proc. SPIE 7195, 71952P (2009).

13. R. Paschotta, J. Nilsson, A. C. Tropper, and D. C. Hanna, “Ytterbium-doped fiber amplifiers,” IEEE J. Quantum Electron. 33(7), 1049–1056 (1997). [CrossRef]  

Cited By

Optica participates in Crossref's Cited-By Linking service. Citing articles from Optica Publishing Group journals and other participating publishers are listed here.

Alert me when this article is cited.


Figures (5)

Fig. 1
Fig. 1 A single polarization micro structured ytterbium doped fiber is used. It is pumped with high power 976 m pump diodes, temperature stabilized and controlled through the feedback loop.
Fig. 2
Fig. 2 Measured a) laser pulse peak power and b) pulse duration at different wavelengths in strongly pumped fiber. Solid line is the theoretical prediction in this wavelength range.
Fig. 3
Fig. 3 Comparison of three laser pulses at different operation wavelengths of 1015 nm, 1030 nm and 1040 nm. The corresponding pump pulses with pump pulse power of 330 W are also shown.
Fig. 4
Fig. 4 Figure a) shows the measured laser pulse duration and peak power in dependence of pump pulse power. The data (circles) are connected with solid line corresponding to linear fit and inverse square root fit as obtained from theoretical predictions. In b) the laser pulses and pump pulses energies are shown at corresponding pump pulse powers.
Fig. 5
Fig. 5 Transmitted amplitude in dependence of the rotation angle of the analyzer. The polarization contrast was greater than 21dB in 1.2 m long fiber.

Equations (5)

Equations on this page are rendered with MathJax. Learn more.

N 2 t = Γ σ c 0 V n s ( N 2 N 1 ) ϕ N 2 τ 21 + w ,
ϕ t = Γ σ c 0 V n s ( N 2 N 1 ) ϕ ϕ τ L + β N 2 τ 21 .
w = λ p P abs h c 0 = λ p h c 0 ( 1 10 α L / 10 ) P P .
t L p 2 π ( n s V h Γ λ p 1 σ P a b s ) 1 2 .
t P p P a b s 1 / 2 .
Select as filters


Select Topics Cancel
© Copyright 2024 | Optica Publishing Group. All rights reserved, including rights for text and data mining and training of artificial technologies or similar technologies.