Open Access
Add to BibSonomy
Abstract
We report the latest progress in fabrication and laser performance of the fully crystalline double-clad ‘Yb:YAG-core/undoped-YAG-clad’ fibers grown by the hybrid crystal growth method. The single-crystalline ytterbium (Yb) doped yttrium aluminum garnet (YAG) fiber cores were grown by the laser heated pedestal growth (LHPG) method, and the single-crystalline undoped YAG claddings were grown by the liquid phase epitaxy (LPE) technique, in which the single-crystalline Yb:YAG cores were used as the growth seeds. The key parameters of the hybrid-grown ‘crystalline core/crystalline clad’ (C4) fibers, including material composition, crystal structure, and fiber propagation loss, were characterized. The results confirmed that the grown C4 fibers, indeed, have both the single-crystalline fiber core and single-crystalline fiber clad. By utilizing a double-clad low-loss C4 fiber as a diode-cladding-pumped laser gain medium, we realized a fiber laser with the optical-to-optical conversion efficiency of 68.7% versus the incident pump power.
© 2018 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Recently, fiber lasers have been recognized to be the prominent laser architecture capable of meeting the constantly growing need in high power light sources with nearly diffraction limited beam quality. Despite the latest successes in fiber laser power scaling out of a single fiber aperture, where multi-kilowatt silica-glass based single-mode fiber lasers/amplifiers have been reported [1,2], further increase of fiber laser output power from the current multi-kilowatt level to the multi-ten-kilowatt level while maintaining near diffraction limited beam quality is required. This development is facing fundamental physical limitations, including the thermal - due to the low thermal conductivity of the silica-glass material (~1 W/m*K), the onset of the stimulated Brillouin scattering (SBS) - due to the relatively high SBS gain coefficient of the silica-glass (~5.5x10−11 m/W) [3], the onset of the transverse mode instability (TMI) [2], and the photo-darkening effects [4]. All of these limitations are strongly tied with laser host material properties. In search for paths towards further power scaling out of a single fiber aperture, and in order to overcome the above limitations, there have been some efforts in place, since early 2010, theoretically exploring the potential of fiber lasers based on non-silica crystalline materials [5,6]. They have shown that, since crystalline materials, such as yttrium aluminum garnet (YAG), possess much higher (about an order of magnitude) thermal conductivity and significantly reduced SBS gain (more than two orders of magnitude lower than that of silica glass), fiber lasers based on fibers made entirely from crystalline materials (fully-crystalline fibers) would have the potential for achieving multi-ten-kilowatt output power out of a single fiber aperture, even for the most challenging, narrowband, single-frequency fiber laser/amplifier case [5,6]. Fibers based entirely on single-crystalline materials are also likely not to be prone to photo-darkening. Indeed, though the photodarkening of bulk doped and undoped YAG materials under UV and short-wavelength visible light (‘solarization’) has been extensively studied and reported (e. g., in [7]), no photodarkening has been observed, and none is expected under the near-IR excitation (pump diodes) in such wide-bandgap crystalline laser hosts as YAG [5,6]. Also, as shown by a very recent TMI threshold estimates [8], a simplified TMI power threshold derivation indicates that this threshold is simply directly proportional to the thermal conductivity of the fiber material [8]. Since thermal conductivity of YAG is ~10 times higher than that of silica, the power threshold of TMI in fully crystalline YAG fibers is projected to be at least an order of magnitude higher than that in conventional silica glass fibers.
While fabrication techniques for double-clad glass fibers (‘glass-core/glass-cladding’ design, or GCGC) are well developed, fabrication of true double-clad, fully-crystalline fibers (‘crystalline-core/crystalline-cladding’ design - CCCC, or C4), was found to present significant technological challenge. Unlike glass fibers, crystalline fibers cannot be just pulled from a softened vitreous preform, because crystalline materials have very well defined melting point, and do not exhibit a convenient “soft” transition between the liquid and crystalline phases. Therefore, double-clad crystalline fibers simply cannot be fabricated in the same manner as their glass counterparts. Since recently, growing long, low-loss, rare-earth-doped (RE-doped) single-crystalline YAG fibers with diameters down to 30 µm via Laser Heated Pedestal Growth (LHPG) technique, to be used as fiber cores for C4-design fibers, has substantially matured. Indeed, a few successful reports were presented recently involving RE-doped YAG LHPG cores fabrication and their laser testing in a waveguided regime of propagation [9,10]. Laser efficiencies up to 67.5% were reported, which confirms that LHPG core fabrication is mature enough to support further fully-crystalline fiber development. In the meantime, fabrication of high quality (low-loss) undoped YAG claddings surrounding RE-doped YAG core, to form a C4 fiber structure, still presents a major technological challenge.
A wide variety of approaches have been pursued in order to address that challenge. The oldest approach of overcladding the LHPG-grown single-crystalline cores was designed to create a high-quality glass coat in a function of cladding on a crystalline fiber core [11,12]. Although very successfully realizing a high quality core/cladding fiber structure, this approach only supports fabrication of a ‘crystalline-core/glass-cladding’ fiber designs (or CCGC). Unfortunately, CCGC fibers simply do not have the required power scaling potential projected for C4 fibers - due to the low thermal conductivity of the thick glass cladding layer (thermal insulator around the crystalline core). An interesting approach proposed for obtaining crystalline fiber cladding was to create an in situ refractive index profile during the LHPG process by controlling the diffusion of dopant ions [13,14]. Although this approach may, perhaps, work for some dopants in a limited way, it would not work for many others. For example, dopants like erbium (Er) have high mobility in their liquid state that can quickly uniformize the dopant concentration and make it hard to realize the required gradient refractive index profile.
Since recently, in order to realize fully-crystalline, ‘crystalline-core/crystalline-clad’ (C4) fibers, a variety of techniques, such as hydrothermal growth, molten salt growth, and magnetron sputtering, were implemented to grow crystalline claddings around cylindrical circumference of the LHPG-grown crystalline fiber cores [15,16]. Despite the reported successes in cladding deposition derived from the above-mentioned techniques, no laser results have been demonstrated from these efforts so far. An alternative liquid solution growth method, which can be used for crystalline cladding deposition over the LHPG-grown fiber (used as a seed) - a liquid phase epitaxy (LPE), has been recently explored and further developed [17,18], and very encouraging first laser results have been reported based on the (LHPG + LPE)-derived C4 fiber [18]. It was the first cladding-pumped laser operation ever reported from a true double-clad, fully-crystalline ‘Yb:YAG-core/undoped-YAG cladding’ C4 fiber, and it was delivered with the slope efficiency of ~37%.
Reported here are the advanced techniques used for further development of high quality growth of the fully-crystalline C4 fibers as well as characterization results of the key properties of the grown C4 fibers, including the crystalline structure, material composition, and propagation loss, which were partially described in [19]. We also report recent significant improvements in optical-to-optical conversion efficiency of the diode-cladding-pumped laser operation based on an ‘Yb:YAG-core/undoped-YAG-clad’ low-loss C4 fiber.
2. Fabrication of the fully-crystalline C4 fiber
As mentioned above, a hybrid, two-step, crystal growth approach has been recently developed for fabrication of fully-crystalline double-clad C4 fibers [17,18]. The first step in this approach is to grow the single-crystalline fiber core by the well-known LHPG method [20]. Figure 1 illustrates the LHPG system implemented for drawing the single-crystalline fiber. The systemincludes the following steps: (1) a donut-shaped CO2 laser beam is formed by passing a regular round CO2 laser beam through a reflexicon and a parabolic reflection mirror; (2) the donut-shaped laser beam is then focused on the tip of a preform so that a molten zone is formed; (3) a seed rod is brought into a contact with the molten zone to initiate the seeding; and (4) the fiber is drawn by coordinated pulling (by the computerized procedure with the fiber diameter control loop) of the seed rod and the preform.
Fig. 1 A simplified schematic of an LHPG system used to grow the single-crystalline core of a fully-crystalline C4 fiber.
The second step in our approach was to grow the crystalline cladding by the LPE method, in which the single-crystalline fiber core, grown by the LHPG, is used as a seed for growing the single-crystalline cladding. Figure 2 illustrates the configuration of a recently developedadvanced LPE crystal growth system, which is mainly composed of the following components: (1) heating furnace, (2) Pt crucible, (3) proper holding and pulling units. During the cladding growth process, the single-crystalline seed (fiber core) was immersed in the molten flux. The composition of the flux and the growth temperature depend on the materials to be grown. For growing the undoped YAG crystalline cladding, the yttria (Y2O3) and alumina (Al2O3) powders were added in the flux, and the cladding growth temperature was kept between the 900 and 1150 °C.
Fig. 2 A simplified sketch of an LPE growth system for growing crystalline claddings around the crystalline core.
By implementing the above hybrid, two-step, double-clad fiber growth approach, the fully-crystalline C4 fibers with rare-earth-doped YAG core diameters between 30 and 100 µm and undoped YAG cladding thicknesses between 1 and 150 µm were successfully grown. Figures 3(a) and 3(b) below show a cross-section scanning electron microscope (SEM) image and an optical transmission image, respectively, of the fully-crystalline C4 fiber, which was composed of a 1% ytterbium (Yb3+) doped YAG crystalline core and an undoped YAG crystalline cladding. As can be seen from Fig. 3, both the fiber core and the cladding exhibit clear cross-sectional hexagonality (less obvious on the core, but much more prominent on the cladding) – the indication of the natural habit of the YAG single crystal occurring due to the seeded core growth in a <111> orientation. Furthermore, since the refractive index of the 1%Yb:YAG core is ~1.6x10−4 higher than that of undoped YAG cladding [21], one could clearly see the light guiding effect (i.e., the brightness of the core area is higher than that of the cladding), as shown in Fig. 3(b).
Fig. 3 (a) A cross-section SEM image of an Yb:YAG /YAG C4 fiber and (b) the corresponding cross-section optical transmission image.
As another example, Fig. 4 shows an optical reflection image of an Yb:YAG/YAG C4 fiber that has a much thicker, 120 µm, undoped YAG cladding.
3. Characterization of the key properties of the hybrid-grown C4 fiber
To ensure that the hybrid-grown fiber has the required quality to be used as a C4 fiber in fiber lasers, we performed quantitative characterization of the key properties of the as-grown C4 fibers. First, the material compositions of the fibers were characterized by the energy-dispersive X-ray spectroscopy (EDS). Elemental compositions in both the core region (marked as Spectrum 1 in Fig. 5), and the cladding regions (marked as Spectra 2 & 3 in Fig. 5), were quantitatively measured by the EDS. The measurement results are presented in Table 1.
Fig. 5 A cross-section SEM image of a grown fiber with the marked core and cladding regions probed by EDS.
Table 1. The material compositions of core and cladding regions measured by the EDS
One can see that the major elements of the core and cladding regions are oxygen (O), Aluminum (Al), and yttrium (Y), in proportions consistent with the YAG composition. One can also see that the measured concentration level of Yb is at the negligible (noise) level in the cladding region, whereas the concentration of Yb in the core region is consistent with the target concentration of Yb. This confirms that there is almost no diffusion of the Yb ions contained in the fiber core into the cladding region during the LPE deposition process of the undoped YAG cladding.
Second, in order to ensure that the hybrid-grown fiber has, indeed, a single-crystalline cladding, we conducted an x-ray diffraction (XRD) analysis on the fragments of LPE-grown cladding. Figure 6 shows the XRD pattern for the cladding. As can be seen from Fig. 6, there is only one peak of Laue reflection indexed (220) identifying the YAG single crystal, and no other phases are present. Notice, that the (440), (660) and (880) represent the same substance. Thus, the grown cladding was proven to be, indeed, a single-crystalline YAG cladding.
Finally, the propagation attenuation coefficient of the C4 fiber was determined by measuring the propagation loss of the hybrid-grown fiber. The following standard formulas were used to calculate the attenuation coefficients on both linear and log scale:
wheredenote the fiber length, the output laser power, and the input laser power, respectively (that is, if the laser is used as a probe source). As an example, in one experiment, a laser beam from HeNe laser with an output wavelength of 632.8 nm (which is not absorbed by the Yb3+ ions of the Yb:YAG core) was coupled into a double-clad C4 fiber by an objective. The fiber under testing had a length of L = 5 cm and a core numerical aperture of NA ≈0.06. The numerical aperture of the objective was 0.13, which is twice the numerical aperture of the fiber core, thus the measured propagation attenuation coefficient is representative of the combined propagation loss of both fiber core and the cladding. Both fiber ends were polished flat to laser quality, and all of the measurement results were corrected for Fresnel loss on both fiber ends. The obtained measurement results are presented in Table 2.Table 2. Measured attenuation coefficient
So, the current level of propagation loss of ~0.0024/cm is certainly low enough to enable high efficiency cladding-pumped laser demonstrations.
4. High efficiency laser operation of the low-loss C4 fiber
To demonstrate the high efficiency diode-cladding-pumped laser capability of the most recently hybrid-grown advanced C4 fibers, we carried out the laser experiment described below. The C4 fiber used in our experiments had a 1%Yb doped YAG core with the core diameter of ~100 µm, a 10 µm thick undoped YAG (u-YAG) cladding and no additional coating around the u-YAG cladding. The composite fiber propagation loss in this fiber was measured to be around 0.003 cm−1. The fiber length between the two laser quality polished ends was ~70 mm. Note that the 7 cm long fiber was used for our laser experiments because (as calculated and then confirmed experimentally) it already absorbs most (but not all, which would be inappropriate for the 3-level laser operation) of the pump with the given dopant concentration and the fiber clad-to-core ratio. Further increase in fiber length, with the given dopant concentration and fiber clad-to-core ratio, would not result in laser efficiency increase. We have used no antireflection (AR) coatings, and one of the uncoated polished fiber ends was used as a Fresnel reflection-based fiber laser output coupler (reflectivity of ~8.5%). Because this C4 fiber had no other second cladding but air, we had to try and minimize possible total internal reflection frustration at the ‘u-YAG cladding/air’ interface, which, in turn, would minimize pump power losses in the u-YAG cladding. Therefore, in order to hold the fiber with sufficient mechanical stability and protection during laser experiments, the fiber was placed into a U-shaped copper channel (U-channel) with the cross-section of 2x2 mm, where it was held, in only two spots, with small drops of clear epoxy resin. In fact, the fiber was slightly bent inside the U-channel in the resin-mounting process, so that one end of the fiber was showing in the corner at the bottom of the U-channel (as can be seen from a magnified end view of the ‘Yb:YAG-core/undoped-YAG-clad’ C4 fiber mounted in a 2x2 mm copper U-channel, presented in Fig. 7), while the other end was showing in the middle at the bottom of the U-channel cross section.
Fig. 7 End view of the ‘Yb:YAG-core/undoped-YAG-clad’ C4 fiber mounted in a 2x2 mm copper U-channel for laser experiments
Since this mounting does not provide any cooling to the C4 fiber, in order to avoid thermal effects during the laser experiments, they were performed in the quasi-continuous-wave (Q-CW) regime. A pulsed fiber coupled laser diode module (FCLDM) with an output wavelengthof 969 nm, pulse width of 1 ms, pulse repetition rate of 10 Hz (duty cycle of 1%), and a spectral bandwidth of 4 nm, was used as the Q-CW pumping source. The LCLDM output was coupled into a standard multimode delivery fiber, dia. 105/125 µm and NA of 0.22. With the pump pulse duration of 1 ms and the upper laser level lifetime of Yb3+:YAG of ~1 ms, the operating regime of our Yb:YAG-core C4 fiber laser is quite fairly representative of the true CW regime for the purposes of assessing laser CW power scaling potential, possible non-thermal saturation effects, and, most importantly, pump-to-output conversion efficiency, which could otherwise be obtained in true CW regime if the adequate fiber cooling is present.
Figure 8 depicts the experimental setup for the cladding-pumped laser testing of the above fully-crystalline ‘1%Yb:YAG-core/undoped-YAG-clad’ C4 fiber. The diverging pumping diode laser beam is collimated by the Lens 1 with the focal length of. The collimated beam then goes through the dichroic mirror (high transmission at 969 nm - pump wavelength, and high reflection at ~1030 nm – Yb-laser wavelength). After passing through the dichroic mirror, the collimated pump beam is focused into the C4 fiber by the Lens 2 withthe focal length of and numerical aperture of in order to most fully utilize the cladding NA. Yb:YAG laser cavity is formed between the dichroic pump mirror and the Fresnel reflection at the output end of the C4 fiber, which enables the low-threshold cladding-pumped lasing.
Fig. 8 Optical layout of the co-pumped fiber laser based on the ~70 mm long, 100/120 µm ‘Yb:YAG-core/u-YAG-clad’ C4 fiber. Laser cavity is formed by the dichroic pump mirror and the Fresnel reflection of the output C4 fiber end.
The results of the ‘1%Yb:YAG-core/u-YAG-clad’ C4 fiber laser testing are presented in Fig. 9, where the Q-CW fiber laser output at 1030 nm is plotted as a function of the 969-nm Q-CW laser diode pump power launched into the C4 fiber end after the Lens 2.
Fig. 9 The Yb:YAG/u-YAG C4 fiber Q-CW output power at 1030 nm versus the Q-CW incident laser diode pump power at 969 nm. Black line – experimental result; red line – simulation result.
The black solid circles (and the line) in Fig. 9 represent the experimental data with the modeling results (simulation of the laser performance based on the same fiber parameters: and) presented by the red solid squares (and the line). As seen from Fig. 9, experimental data set nicely fits our numerical simulation results. In the Q-CW regime we achieved ~50 W of Q-CW laser power with the optical-to-optical laser efficiency approaching the 70% mark. We also believe that the fiber loss figure can be further reduced.
5. Conclusion
In conclusion, we reported our latest progress in fabrication and laser performance of the fully-crystalline ‘Yb:YAG-core/undoped-YAG-clad’ C4 fibers, which were grown by the hybrid crystal growth method. The single-crystalline Yb:YAG fiber core was grown by the LHPG technique, and the single-crystalline undoped YAG cladding was grown by the LPE method, in which the single-crystalline fiber core played a role of the growth seed. The material composition and the single-crystalline structure were confirmed by the EDS and XRD analyses. We have demonstrated that high efficiency diode-cladding-pumped laser operation can be realized with the low-loss fully-crystalline C4 fiber as the gain medium. In the Q-CW regime ~50 W of Q-CW Yb-laser power at 1030 nm has been achieved with the optical-to-optical laser efficiency (versus the incident pump power) approaching the 70% mark. Experimental results nearly perfectly agreed with the numerical simulation of the laser performance.
Funding
US Army Contract #W911QX-17-C0023.
References and links
1. J. Zheng, W. Zhao, B. Zhao, C. Hou, Z. Li, G. Li, Q. Gao, P. Ju, W. Gao, S. She, P. Wu, and W. Li, “4.62 kW excellent beam quality laser output with a low-loss Yb/Ce co-doped fiber fabricated by chelate gas phase deposition technique,” Opt. Mater. Express 7(4), 1259–1266 (2017). [CrossRef]
2. F. Beier, C. Hupel, S. Kuhn, S. Hein, J. Nold, F. Proske, B. Sattler, A. Liem, C. Jauregui, J. Limpert, N. Haarlammert, T. Schreiber, R. Eberhardt, and A. Tünnermann, “Single mode 4.3 kW output power from a diode-pumped Yb-doped fiber amplifier,” Opt. Express 25(13), 14892–14899 (2017). [CrossRef] [PubMed]
3. J. W. Dawson, M. J. Messerly, R. J. Beach, M. Y. Shverdin, E. A. Stappaerts, A. K. Sridharan, P. H. Pax, J. E. Heebner, C. W. Siders, and C. P. Barty, “Analysis of the scalability of diffraction-limited fiber lasers and amplifiers to high average power,” Opt. Express 16(17), 13240–13266 (2008). [CrossRef] [PubMed]
4. S. Jetschke, S. Unger, M. Leich, and J. Kirchhof, “Photodarkening kinetics as a function of Yb concentration and the role of Al codoping,” Appl. Opt. 51(32), 7758–7764 (2012). [CrossRef] [PubMed]
5. J. Dawson, M. Messerly, J. Heebner, P. Pax, A. Sridharan, A. Bullington, R. Beach, C. Siders, C. Barty, and M. Dubinskii, “Power scaling analysis of fiber lasers and amplifiers based on non-silica materials,” Proc. SPIE 7686, 768611 (2010). [CrossRef]
6. T. Parthasarathy, R. Hay, G. Fair, and F. Hopkins, “Predicted performance limits of yttrium aluminum garnet fiber lasers,” Opt. Eng. 49(9), 094302 (2010). [CrossRef]
7. W. Koechner, Solid-State Laser Engineering, Sixth Edition (Springer, 2006).
8. M. N. Zervas, “Transverse mode instability analysis in fiber amplifiers,” Proc. SPIE 10083, 100830M (2017). [CrossRef]
9. Y. Li, K. Miller, E. G. Johnson, C. D. Nie, S. Bera, J. A. Harrington, and R. Shori, “Lasing characteristics of Ho:YAG single crystal fiber,” Opt. Express 24(9), 9751–9756 (2016). [CrossRef] [PubMed]
10. J. Zhang, Y. Chen, B. Ponting, E. Gebremichael, R. Magana, G. Maxwell, and M. Dubinskii, “Highly efficient waveguided laser performance of diode pumped unclad Yb:YAG crystalline fibre,” Laser Phys. Lett. 13(7), 075101 (2016). [CrossRef]
11. R. Byer, A. Cordova, M. Digonnet, M. Fejer, C. Gaeta, H. Shaw, and S. Sudo, “Method of cladding single crystal optical fiber,” US Patent No. 5,077,087 (1991).
12. K. Y. Huang, K. Y. Hsu, D. Y. Jheng, W. J. Zhuo, P. Y. Chen, P. S. Yeh, and S. L. Huang, “Low-loss propagation in Cr4+:YAG double-clad crystal fiber fabricated by sapphire tube assisted CDLHPG technique,” Opt. Express 16(16), 12264–12271 (2008). [CrossRef] [PubMed]
13. S. Rusanov, A. Yakovlev, and A. Zagumennyi, “Graded index single crystal optical fibers,” US Patent No. 5,579,427 (1996).
14. S. Bera, C. Nie, J. Harrington, T. Chick, A. Chakrabarty, S. Reichert, J. Champman, and S. Rand, “Cladding single crystal YAG fibers grown by laser heated pedestal growth,” Proc. SPIE 9726, 97260C (2016).
15. W. Kim, B. Shaw, S. Bayya, C. Askins, J. Peele, D. Rhonehouse, J. Meyers, R. Thapa, D. Gibson, and J. Sanghera, “Cladded single crystal fibers for high power fiber lasers,” Proc. SPIE 9958, 99580O (2016). [CrossRef]
16. W. Kim, S. Bayya, L. B. Shaw, J. Myers, S. N. Qadri, R. Thapa, C. Askins, J. Peele, D. Rhonehouse, S. Bowman, and D. Gibson, “Joseph Kolis, Brad Stadleman, and J. Sanghera, “Crystal fiber lasers,” Proc. SPIE 10382, 103820Q (2017).
17. S. Yin and C. Luo, “Method and apparatus for producing crystalline cladding and crystalline core optical fibers,” US Patent Application No. US 2017/0031091, Jul. 27 (2016).
18. J. Zhang, Y. Chen, S. Yin, C. Lou, and M. Dubinskii, “High-power diode-cladding-pumped Yb:YAG laser based on true double-clad fully crystalline fiber,“ in Advanced Solid State Lasers Conference, 2016 OSA Technical Digest Series (Optical Society of America, 2016), Postdeadline paper, ATu6A.4. [CrossRef]
19. M. Dubinskii, J. Zhang, V. Fromzel, Y. Chen, S. Yin, and C. Luo, “Fiber lasers with ‘crystalline-core/crystalline-clad’ (C4) architecture fibers for highly power scalable, high efficiency, diode-cladding-pumped operation,” in Advanced Solid State Lasers Conference, 2017 OSA Technical Digest Series (Optical Society of America, 2017), AW3A.2. [CrossRef]
20. J. Haggerty, W. Menashi, and J. Wenekkus, “Method of forming refractory fibers by laser energy,” US Patent No. 3,944,649 (1976).
21. M. Malinowski, J. Sarnecki, R. Piramidowicz, P. Szczepanski, and W. Wolinski, “Epitaxial RE3+:YAG planar waveguide lasers,” Opto-Electron. Rev. 9(1), 67–74 (2001).
