1. Introduction

The characteristics of most radiative emissions are determined by the intrinsic nature of the active species and its immediate microscopic environment. Laser emission can, however, be controlled by macroscopic environment. Recently 2D photonic bandgap fibers have become a reality. This makes it possible that laser emission wavelength can be controlled by allowing only some wavelengths to be confined in a laser cavity. The impact of this is significant, as it can be used to provide laser architectures which enable many new wavelengths of efficient laser operations and consequently many new applications.

We report in this work an all-fiber monolithic cladding pumped Yb-doped all-solid photonic bandgap fiber laser. The Yb 3-level system at ~978nm can be operated just like the 4-level system in achieving single-mode powers in the hundreds of watts with comparable efficiency. The key is using a Yb-doped double-clad all-solid photonic bandgap fiber with its bandgap optimized to suppress the 4-level system near 1030nm. Owing to our specialty photonic bandgap fiber, we have achieved single-mode power of 151W at ~978nm with an efficiency of 63% with respect to the launched pump power in a practical all-fiber monolithic fiber laser configuration. This is a significant progress over our previous demonstration using free-space bulk optics laser scheme [1]. This breakthrough immediately provides new pumping options for diffraction-limited solid-state lasers and core-pumping options for fiber lasers and amplifiers, especially in nonlinearity-limited diffraction-limited pulsed lasers.

Previously, all-solid double-clad photonic bandgap fibers (AS-PBFs) have been used to suppress Yb ASE at shorter wavelengths, enabling high-power lasers above 1150nm [2–4]. We have also used AS-PBFs to enable an efficient Yb fiber laser at ~1018nm by suppressing Yb ASE at the longer wavelengths [5]. For suppressing the Yb 4-level system in an Yb 3-level system laser, the separation of the laser wavelength and ASE peak wavelength is ~50nm. This is much less than the ~150nm separation as reported in [2–4], and it presents a much more significant challenge. The fibers in [2–4] also had mode field diameter of ~10µm. This is also much smaller than what is desired in an Yb 3-level system as will be discussed later. The use of all-solid photonic bandgap fibers for 3-level Yb fiber lasers has been reported before [6]. However, in this case, it was a core-pumped fiber laser arrangement with an active fiber mode field diameter of just 3.4µm, and the maximum achieved output power was only ~140mW.

The 4-level Yb system has been a critical foundation for industrial high-power fiber lasers. The 3-level system has only attracted a little academic interest, having been limited by poor efficiencies and low powers in practical laser configurations. Conventional methods of mitigation are almost entirely based on large core-to-cladding ratio [7–13]. This lowers the intensity of the laser relative to that of the pump, therefore, allowing the required high inversion to be maintained at relatively lower pump powers. This conventional approach results in reduced unused pump (i.e. residual pump), but it produces only limited performance improvements in practical high-power fiber lasers.

In a recent state-of-the-art demonstration of a monolithic fiber laser configuration, a fiber with a 20μm core diameter and a square cladding of 80μm × 80μm was tapered to 50μm × 50μm in the middle, achieving 10.6W at 976nm and an efficiency of 18.4% [11]. The records in both power and efficiency for Yb 3-level fiber lasers were set in a rod-type photonic crystal fiber with a core diameter of 80μm and a pump waveguide diameter of 200μm. This large core cannot be bent, and the fiber must be kept straight. 94W at 977nm, an efficiency of 48% and an M² of 1.2 were achieved in a free-space bulk optics laser scheme [12]. The pump was double-passed in this case. Using the same fiber and the cavity structure, higher efficiencies of 53% and 63% were achieved for single and double pass pump respectively. However, M² was ~1.2 only for an output power below 25W [13]. At the highest output power of 95W, the M² was only 2.2.

In this paper, we present an all-fiber monolithic fiber laser operating at ~978nm which was built using in-house fabricated Yb-doped double-clad all-solid photonic bandgap fiber. We have achieved continuous wave output power of 151.4W with laser slope efficiency of 63%. Compared to our previous report using bulk optics for pump and output coupling [1], we have increased maximum output power by a factor of ~2 while maintaining the high laser efficiency in an all-fiber monolithic laser scheme. Furthermore, we have conducted long term power stability test. We have provided further experimental proof to our claim in the previous report [1] that output power is only limited by available pump power and the active fiber is almost photodarkening free.

2. Fiber design and characterization

The cross section of the all-solid photonic bandgap fiber used in this work is shown in Fig. 1 along with the measured bend loss showing the carefully engineered bandgap of the active fiber for the suppression of the Yb 4-level system. The fiber can be coiled down to below 20cm diameter without significant in-band bend loss and show strong loss above ~980nm due to a loss of core guidance. This fiber was designed and fabricated in-house by the optical fiber fabrication facility at Clemson University. It is the same active fiber as the one used in our previous work [1]. The fiber has a core with a corner-to-corner distance of 24μm and side-to-side distance of 21μm. The cladding has a corner-to-corner distance of 131μm and side-to-side distance of 124μm. A multiple-cladding-resonance design is used for enhanced higher-order-mode suppression [14–16]. The fiber is coated with low index acrylate to provide a pump NA of ~0.46, and the cladding pump absorption was measured to be ~1.76dB/m at 915nm.

 

Fig. 1 The 2D all-solid photonic bandgap fiber cross section and measured bend loss at 20cm, 30cm and 40 cm coil diameters in a passive fiber which has the same bandgap structure as the active fiber.

Download Full Size | PDF

3. Experimental setup

The basic laser arrangement is a counter-pumped monolithic configuration shown in Fig. 2 with two pump diodes at ~915nm (200W, 0.22NA 105/125µm) spliced to a 2 + 1 pump combiner. Three meters of 20/105µm Er-doped fiber with a highly Er doped core (7wt% Er) coiled at 5cm in diameter was used in each pump path to absorb any backward propagating leakage light at the lasing wavelength at ~978nm. Pump loss at ~915nm was measured to be ~0.1dB for the 3m Er fiber and ~0.46dB for the pump combiner. The high-reflectivity fiber Bragg grating (FBG) was written in-house using a frequency-quadrupled YAG laser at 266nm in a 24/125µm photosensitive fiber which was also made in-house. The FBG has a reflectivity of >99% and a bandwidth of 2nm. The output is angle cleaved. This configuration was previously found to minimize output at the FBG end, i.e. laser 2 in Fig. 2, without compromising the laser efficiency [1].

 

Fig. 2 Configuration of the Yb 3-level-system monolithic fiber laser.

Download Full Size | PDF

The length of the bandgap fiber was optimized first by progressively cutting back the bandgap fiber while fully characterizing the laser performance. The bandgap fiber was coiled to 15cm diameter, and the residual pump light as well as the light at the lasing wavelength (laser 2) were monitored at the far end.

4. Results and discussions

Laser efficiency, total efficiency, residual pump as percentage of the launched pump power, and laser 2 as percentage of the launched pump power are shown versus the photonic bandgap fiber length in Fig. 3. The optimized fiber length is ~13m, and 12.3m fiber was used in the last experiment. The reason we ended up using only 12.3m long active fiber in the final test (as in Fig. 5) is that we had been using the same piece of active fiber for several measurements. Eventually, the active fiber got shortened during this process due to cut-back measurements and fiber cleaving. But, we did not replace the active fiber with a new piece because of limited amount of active fiber left in the lab.

 

Fig. 3 Laser performance versus active fiber length. Laser efficiency: accounting for just output power from the pump combiner with respect to the launched pump, total efficiency: accounting for both laser output powers with respect to the launched pump, residual pump: with respect to the launched pump power, and laser 2: with respect to the launched pump.

Download Full Size | PDF

Output powers with a single pump and double pump are shown in Fig. 4, with the spectra at various powers. This output power does not include the output power at laser 2. A maximum output power of 90.9W at the output was achieved with a single pump and 151.4W with double pumping, limited by available pump powers in each case. The corresponding output for the double pumping at laser 2 is 7W, i.e. 4.4% of total power of 158.4W. With double pumping, the efficiency was ~63% and ~75.4% with regards to the launched pump power after the combiner and absorbed pump power respectively. The efficiency of ~63% is at the same level as the efficiency of 62.7% that we achieved in a free-space bulk optics laser configuration [1]. However, the novelty in this work is that we were able to achieve much higher output power using a practical monolithic configuration. The ASE from the 4-level system was well suppressed to below 40dB at highest laser output power.

 

Fig. 4 Left: measured output versus pump power for both with respect to the launched pump power and absorbed pump power; Right: measured output spectrum for double pump case under various output power.

Download Full Size | PDF

The M² at ~150W was 1.25/1.24 (Fig. 5), which was also found to be almost constant across the whole power range. The M² at ~3W was 1.20/1.21.

 

Fig. 5 M² measurement at output of 3W and 150W.

Download Full Size | PDF

Photo-darkening increases significantly with inversion levels in Yb fiber lasers and is expected to be a severe problem for the Yb 3-level system due to its high inversion. The Yb phosphosilicate core glass used in our fiber is well known for its high resistance to photo-darkening [17], and it exhibited negligible degradation of laser performance over a period of several months and numerous tests. We have conducted a long-term power stability experiment over ~60 hours with a single-pump configuration under the output power at ~75W (Fig. 6). Apart from some power fluctuations in the few percent levels, most likely due to temperature changes in the lab, there was very little sign of photo-darkening. It is worth mentioning that the long term stability test was performed with only a single pump. This is because we wanted to be safe while running the laser continuously especially over several nights.

 

Fig. 6 Laser stability test at 75W. Laser was turned off at t = 60.8hrs.

Download Full Size | PDF

5. Conclusion

To conclude, we have demonstrated that the Yb 3-level system at ~978nm can be operated just like the well-established Yb 4-level system in achieving single-mode high-power laser output with comparable efficiency using a Yb-doped double-clad all-solid photonic bandgap fiber with its bandgap optimized to suppress the 4-level system in a practical monolithic setup. We have achieved record output power of 151.4W, M² of ~1.2 at full power, and laser efficiency of 63% with respect to the launched pump power in an all-fiber monolithic Yb-doped double-clad all-solid photonic bandgap fiber laser.