Open Access
Add to BibSonomyAbstract
Transverse photodarkening loss variations in an LMA Yb-DCF are experimentally studied. Photodarkening rate depends on inversion, which is affected by the pump induced increase and signal induced depletion of the inversion. In double-clad fibers, intensity distributions of the pump and signal modes and their overlap with the core are significantly different, leading to transverse differences in inversion within the core. Moreover, practical fiber laser configurations aim at generating and preserving only the fundamental transverse-mode thus creating a high contrast in inversion within the core. Therefore, dramatically different rates of photodarkening across the core of the active fiber can be expected. We demonstrate the existence of transverse mode-induced photodarkening loss variations in an LMA Yb-DCF laser and discuss its implications. Composition-related transverse photodarkening loss variations are measured to be negligible in the studied Yb-DCF.
©2008 Optical Society of America
1. Introduction
Photodarkening in ytterbium-doped silica fibers (YDFs) has been recognized as a potential threat to the reliability of fiber lasers and amplifiers [1-5]. While the mechanism behind photodarkening is under intense research and debate [6-12, 15], inversion has been identified as a key parameter, i.e., the rate of photodarkening is greatly accelerated when an YDF is used in high inversion [3]. Photodarkening rate has been correlated with the number density of ytterbium ions in the excited state [8, 9, 13], with observed power-law rate dependencies ranging from 4 [9] to 7 [13] ions (required to create a color center). Fiber lasers and fiber amplifiers operating in different inversion regimes can therefore be expected to exhibit photodarkening on a wide range of timescales.
The proposed precursors and drivers behind photodarkening suggest that there are two causes inducing spatial variations in the photodarkening process: compositional inhomogeneities and spatial variations in inversion. Compositional inhomogeneities are largely determined by the chosen preform fabrication process. Rare-earth doped fiber fabrication methods are known to produce homogeneous doping along the preform length, but maintaining radial doping homogeneity is more challenging. Most fabrication processes rely on a layer-by-layer deposition of the core material. High process temperatures lead to in- and out-diffusion of volatile species and to compositional differences within and between the deposited layers. These inhomogeneities may lead to significant radial photodarkening rate variations through the various composition-related mechanisms.
The spatial inversion is determined by the local pump, signal, and amplified spontaneous emission (ASE) powers through their respective rates of absorption and emission. Pump depletion and signal/ASE amplification along the active fiber length result in longitudinal variations in the inversion, while transverse inversion variations are caused by distribution of optical powers to various modes having different intensity distributions. In fiber lasers and amplifiers efforts are typically made to minimize the longitudinal inversion variations i.e., active fiber length and pumping configuration are chosen not to result in poorly pumped regions potentially leading to unstable laser operation. As a result, photodarkening can be expected to have weak (but not insignificant) longitudinally varying dependence. In contrast to this, majority of fiber laser configurations aim at generating and preserving only the fundamental transverse LP01-mode, and therefore, the transverse inversion profile is strongly affected by the intensity profile of this mode [17]. Cladding pumping provides high and uniform excitation of the ytterbium ions in the core. Therefore, regions of the core where intensity of the inversion-depleting LP01-mode is low remain in high inversion and hence experience higher rate of photodarkening through the inversion dependence of the photodarkening process.
This work focuses on studying and demonstrating transverse photodarkening loss variations in a cladding-pumped, large-mode-area (LMA) ytterbium-doped active fiber (Yb-DCF). The used measurement method is described first. This is followed by a description of simulated inversion distributions in three experimental configurations. The experiments and results are then described. Finally, implications of this effect are discussed.
2. Measurement method
Photodarkening-induced changes are measured indirectly by comparing the near-field intensity distributions of transmitted light of a pristine and photodarkened fiber core. Under uniform and repeatable launch conditions and low mode-coupling in comparison to the measurable losses, various core modes having different spatial localization experience different transmission loss depending on their overlap with the spatial loss distribution. Measured near-field intensity distribution therefore reflects the induced radial losses (asymmetry is lost because of circularly symmetric modes). Justification of this method relies on the principles of differential mode attenuation measurements developed for characterization of multimode optical fibers [18]. Measurement is done in the 450–650 nm wavelength range which coincides with the peak of the photodarkening loss. A halogen lamp is used as the source, while a CCD camera is used to capture the intensity distributions. A matching passive fiber is used to provide low-loss coupling from the source into the active fiber and to remove any cladding modes originating from the source. Splice/coupling loss is repeatable within ±0.2 dB. Fibers are kept straight or loosely coiled for the near-field intensity profile measurement in order not to cause any bend-induced mode discrimination. A Gaussian fit is used to detect the focus point of the core intensity distribution in order to repeatedly produce a sharp image at the camera. The root-mean-square (RMS) deviation of the intensity change measurement is ±0.3 dB within the core (from -8.0 µm to 8.0 µm in a 10-µm radius core) and <±0.8 dB at the falling edges of the intensity change. The measurement uncertainty is based on nine samples and was measured by comparing the intensity profiles of the delivery fiber with and without attenuation (using neutral-density filters). The intensity change between the pristine and the photodarkened profile is attributed to photodarkening by comparing the intensity change to a spectrally resolved photodarkening-induced excess loss within the same wavelength span. Excess loss is measured by replacing the lens and the CCD camera by a standard multi-mode fiber (MMF) to collect the core light and by measuring and comparing the pristine and the photodarkened transmission spectra using an optical spectrum analyzer. Excess loss spectra measured in the different experiments are summarized in Fig. 1.
Fig. 1. Photodarkening-induced excess loss spectra measured in the different experiments (losses are scaled with sample length). Noise between 850-1000 nm (due to Yb-absorption) has been removed from experiment #3.
3. The modeled inversion distributions
Three configurations were used to study photodarkening under different inversion conditions.
- in experiment #1, cladding pumping of an LMA Yb-DCF is used to achieve longitudinally and transversally uniform excitation of the Yb-ions while the ASE-induced depletion of the inversion is minimized. Therefore, any transverse changes in the core intensity distribution can be attributed to composition-induced photodarkening.
- experiment #2 uses core pumping of an LMA Yb-DCF with resulting core intensity changes attributed to transverse pump-mode-induced photodarkening.
- experiment #3 presents the core intensity distribution of an LMA Yb-DCF used in a cladding-pumped laser configuration with the resulting transverse changes attributed lasing-mode-induced photodarkening.
Figure 2 presents the simulated inversion profiles for the above configurations in a 20-µm-diameter LMA core. Steady-state rate equations were used in calculations [15], and inversion was assumed to be solely defined by the LP01-mode, cladding pumping and spontaneous emission. Pump and signal powers and fiber parameters are described in detail in the next section. Coiling of the Yb-DCF in exp. #2 and #3 is required to reduce the impact of the higher-order-modes (HOMs) on the inversion. Therefore, coiling-induced shift in the LP01-mode intensity and the consequent effect on the inversion was also simulated. For comparison, inversion profile in a straight fiber is also presented (other parameters as in exp. #3). Absorption and emission cross-sections used in the model were σpa=9.4*10-26 m2, σpe=3.3*10-27 m2, σsa=3.0*10-27 m2, σse=3.8*10-26 m2 for the pump and signal, respectively.
Fig. 2. Simulated inversion profiles for three different experimental configurations for a 20-µm-diameter LMA core. LP01-mode profile is calculated for a 60-mm-diameter fiber bend radius and was applied in calculation of inversion profiles of exp. #2 and #3.
4. Experiments and results
The experiments were done using a fiber from the same batch of 20-µm-core-diameter, Yb-DCF with core NA of 0.08 and Yb-ion concentration of roughly 8*1025 m-3. Fiber diameter is 125 µm. Sample lengths and pump deactivation times were varied between the experiments to stay within the sensitivity of the CCD camera.
4.1 Experiment #1 –composition-induced photodarkening
The measurement setup is presented in Fig. 3. Based on the simulation results, ASE power has negligible effect on the inversion, if sample length is kept shorter than 50 mm. Therefore, Yb-DCF sample length of 46 mm was used. Roughly 6.6 W of pump at 915 nm was coupled in the Yb-DCF through the combiner, with a resulting inversion level of nearly 63 %. Pumping time was 5 minutes. The output of the Yb-DCF was either focused onto a power meter for the purpose of determining the transmitted pump, or imaged onto a CCD camera for measurement of the core-intensity distribution. A band-pass filtered (Schott BG40) halogen lamp was used as the source. Pump is turned off while the intensity distribution is measured.
Fig. 3. Setup used in experiment #1. CF is a color filter (Schott BG40), 20/125 is a 20-µm-core, low-NA passive fiber (high index coating), 20/125DC is a double cladding (low index coating) matched passive, L is a microscope objective (60x), and CCD is a beam-imaging digital camera (Spiricon LSB-230).
The measured near-field intensity profiles are presented in Fig. 4(a) with the calculated intensity change profile presented in Fig. 4(b). Inset graph shows the symmetric intensity change distribution, featuring low amount of speckle. The intensity change profile is flat (variations are within the measurement error) with an average intensity change of 5.5 dB between -8.0 µm and 8.0 µm. This value compares very well with the spectrally resolved excess losses of 5.4 dB calculated from data presented in Fig. 3, confirming photodarkening as the origin of the intensity change. In the absence of modes having enough power to deplete the inversion (as seen in simulated inversion of “exp #1” in Fig. 2), radially flat intensity change distribution reflects good compositional homogeneity with regards to photodarkening precursors. Therefore, composition-induced photodarkening variations can be assumed negligible and any radial intensity variations in experiments #2 and #3 can be attributed to mode-induced changes.
Fig. 4. (a) Near-field intensity profiles of pristine and photodarkened (5 minutes) fiber. (b) Intensity change profile with the inset graph showing the intensity change distribution.
4.2 Experiment #2 –pump-mode-induced photodarkening
The setup to core pump the coiled LMA fiber is presented in Fig. 5. Roughly 250 mW of pump power at 976 nm was available from the output of the passive 20/125 fiber after mode conversion from the single-mode pigtail. A 100-mm long piece of the Yb-DCF was spliced to the delivery fiber and coiled on a 60-mm diameter spool to produce loss for the HOMs. Pump power was almost completely absorbed in the active fiber and the total ASE was estimated to be low (~35 µW). Fiber was pumped for 15 minutes in this configuration. For the intensity profile measurements, the passive 20/125 delivery fiber was coupled directly to the filtered halogen light source.
Fig. 5. Setup used in experiment #2. The fiber pigtail of a single-mode pump laser diode is spliced onto a matching 20-µm-core, low-NA passive fiber (high index coating) through a mode-converter. L is a microscope objective (60x).
Again, the measured near-field intensity profiles are presented first in Fig. 6(a) with the calculated intensity change profile presented in Fig. 6(b). The core pumped intensity change profile reveals a central higher-loss (~5.5 dB) region surrounded by a lower-loss (~5 dB) shoulder. This loss distribution can be understood in light of simulated inversion profile presented in Fig. 2 (labeled “exp. #2). The inversion and therefore rate of photodarkening is highest at the peak of the LP01-mode and lowest at the edge of the core where the mode intensity is low. However, the simulated inversion asymmetry does not show in the measurement because of mode symmetry. The averaged intensity change of 5.3 dB (across - 8.0 µm to 8.0 µm) again compares well with the spectrally resolved excess losses of 5.2 dB (calculated from data presented in Fig 1.).
Fig. 6. (a) Near-field intensity profiles of pristine and photodarkened (15 minutes) fiber. (b) Intensity change profile with the inset graph showing the intensity change distribution.
4.3 Experiment #3 – lasing-mode-induced photodarkening
The fiber laser setup is presented in Fig. 7. A straight-cleaved, 2-meter-long Yb-DCF was pumped through a free-space optics configuration having two dichroic filters to separate the lasing signal and pump. The Yb-DCF was coiled on a 60-mm diameter spool to produce loss for the HOMs. During the first five minutes, the pump power was step-wise increased up to 17 W (of coupled power) to determine slope efficiency. The pump power was then reduced to roughly 9.2 W for the rest of the experiment. Total output power was 5.2 W at around 1030 nm, with measured slope efficiency of about 75% (with respect to absorbed pump power). There were no signs of output power degradation during the 90 minutes long experiment. The Yb-DCF was spliced directly to the passive 20/125 fiber for the intensity distribution measurements.
Fig. 7. Setup used in experiment #3. DF is a dichroic filter (HR @ 1030-1100 nm, HT @ 900-980 nm), L1 and L2 are aspheric lenses (f: 11 mm and 4.5 mm, respectively).
Figures 8(a) and 8(b) present the measured near-field intensity profiles and the calculated intensity change profile, respectively. The intensity change profile features a symmetric higher-loss (~10.5 dB) ring around a central pedestal at about 8.5 dB. Therefore, contrary to core pumping, photodarkening has occurred faster at the highly-inverted edge region of the core, as is expected based on the simulated inversion profile presented Fig. 2. Averaged intensity change of 9.0 dB correlates quite well with the magnitude of measured excess losses presented in Fig. 1. However, with this excess loss level the source power falls below OSA sensitivity at wavelengths shorter than 600 nm and therefore accurate comparison of the intensity change and excess loss was not possible.
Fig. 8. (a) Near-field intensity distributions of pristine and photodarkened (90 minutes) fiber. (b) Intensity change profile with the inset graph showing the intensity change distribution..
5. Discussion
Demonstrated photodarkening loss dependence on the transverse signal/pump mode intensity distribution may lead to different results for measurements with cladding or core pumping, assuming same sample length and longitudinally uniform inversion (i.e. no signal, low ASE power). Where cladding pumping provides essentially uniform excitation of the core, core pumped sample inversion and therefore photodarkening cross-section is influenced by the pump intensity profile (as shown in Fig. 6(b)). Because the change of sample transmission is typically measured in the visible spectrum wavelength range where a single-mode (at 1.06 µm) YDF is multimode, a transmission loss measurement result becomes dependent on the overlap of the excited measurement modes with the transversal loss distribution. Therefore, a core pumped photodarkening measurement with tightly confined pump mode has a tendency to underestimate the induced losses in a multimode transmission loss measurement.
A non-uniform photodarkening distribution can also be expected to have an influence on the modal behavior of an LMA fiber. Progressive photodarkening may in fact result in differential mode loss in favor of the LP01-mode, thus improving the beam quality. Using results from experiment #1 and published 635 nm-to-NIR loss ratios [2, 16], the excess loss at the edge of the core at 1.06 µm region is estimated to be in the range 0.6–1.2 dB/m. Loss of only 0.05 dB/m (at 1.06 µm) can be extrapolated from exp. #3, but this measurement result is believed to be affected by presence of HOMs (i.e. coiling was not efficient enough in leaking out the HOMs). With typical application lengths of several meters, the accumulated losses can therefore be significant. However, preliminary simulations indicate that the differential mode loss between the LP01-mode and the HOMs is not high enough to provide significant improvement in beam quality before the LP01-mode power begins decay due to the progressive photodarkening. Further theoretical and experimental work is needed to verify this assumption.
6. Conclusions
Transverse photodarkening loss variations in a cladding-pumped LMA Yb-DCF laser were observed for the first time. Measured near-field core intensity change distribution reveals higher photodarkening loss at the edge of the core. This result is consistent with an inversion distribution defined by a lasing LP01-mode (depleting the inversion) and cladding pumping (increasing the inversion) in a coiled fiber laser. The LP01-mode intensity is low at the edge of the core and therefore this region resides in higher inversion, and through the photodarkening rate dependence on inversion, experiences faster photodarkening. The presented results imply that different modes experience different rates of photodarkening through the imprint they and other similarly localized modes have on the inversion. Therefore, in a uniformly doped fiber, the lasing mode having the highest intensity experiences the lowest rate of photodarkening.
Acknowledgments
The authors would like to acknowledge Dr. Jeffrey Koplow and Dr. Dahv Kliner from Sandia National Laboratories (Livermore, CA) for invaluable discussions. Support from TEKES, the Finnish Funding Agency for Technology and Innovation, is acknowledged.
References and links
1. R. Paschotta, J. Nilsson, P. R. Barber, J. E. Caplen, A. C. Tropper, and D. C. Hanna, “Lifetime quenching in Yb doped fibres,” Opt. Commun. 136, 375 (1997). [CrossRef]
2. J. Koponen, M. J. Söderlund, H. J. Hoffmann, and S. K. T. Tammela, “Measuring photodarkening from single-mode ytterbium doped silica fibers,” Opt. Express 14, 11539–11544 (2006). [CrossRef] [PubMed]
3. J. Koponen, M. Söderlund, H. J. Hoffman, D. Kliner, and J. Koplow, “Photodarkening measurements in large-mode-area fibers,” in Fiber Lasers IV: Technology, Systems, and Applications, D. J. Harter, A. Tünnermann, J. Broeng, and C. Headley, Proc. SPIE6453, 64531E-1-11 (2007).
4. I. Manek-Hönninger, J. Boullet, T. Cardinal, F. Guillen, M. Podgorski, R. Bello Doua, and F. Salin, “Photodarkening and photobleaching of an ytterbium-doped silica double-clad LMA fiber,” Opt. Express 15, 1606–1611 (2007). [CrossRef] [PubMed]
5. B. Morasse, S. Chatigny, E. Gagnon, C. Hovington, J.-P. Martin, and J.-P. De Sandro, “Low photodarkening single cladding ytterbium fibre amplifier,” in Fiber Lasers IV: Technology, Systems, and Applications, D. J. Harter, A. Tünnermann, J. Broeng, and C. Headley, Proc. SPIE6453, 64530H-1-9 (2007). [CrossRef]
6. T. Kitabayashi, M. Ikeda, M. Nakai, T. Sakai, K Himeno, and K. Ohashi, “Population inversion factor dependence of photodarkening of Yb-doped fibers and its suppression by highly aluminum doping,” OFC 2005, paper OThC5.
7. J. Jasapara, M. Andrejco, D. DiGiovanni, and R. Windeler, “Effect of heat and H2 gas on the photodarkening of Yb3+ fibers,” in Conference on Lasers and Electro-Optics/Quantum Electronics and Laser Science Conference and Photonic Applications Systems Technologies 2006, Technical Digest (Optical Society of America, Washington, DC,2006), paper CTuQ5.
8. A. V. Shubin, M. V. Yashkov, M. A. Melkumov, S. A. Smirnow, I. A. Bufetov, and E. M. Dianov, “Photodarkening of aluminosilicate and phosphosilicate Yb-doped fibers,” in Conf. Digest of CLEO Europe-EQEC 2007, CJ3-1-THU.
9. S. Jetschke, S. Unger, U. Röpke, and J. Kirchof, “Photodarkening in Yb doped fibers: experimental evidence of equilibrium states depending on the pump power,” Opt. Express , 15, 14838–14843 (2007). [CrossRef] [PubMed]
10. S. Yoo, C. Basu, A. J. Boyland, C. Sones, J. Nilsson, J. K. Sahu, and D. Payne, “Photodarkening in Yb-doped aluminosilicate fibers induced by 488nm irradiation,” Opt. Lett. 32, 1626–1628 (2007). [CrossRef] [PubMed]
11. J. Kirchhof, S. Unger, A. Schwuchow, S. Grimm, and V. Reichel, “Materials for high-power fiber lasers,” J. Non-Cryst. Solids 352, 2399–2403 (2006). [CrossRef]
12. M. Engholn, L. Norin, and D. Åberg, “Strong UV absorption and visible luminescence in ytterbium-doped aluminosilicate glass under UV excitation,” Opt. Lett. 32, 3352–3354 (2007). [CrossRef]
13. J Koponen, M. Söderlund, H. J. Hoffman, D. Kliner, J. Koplow, and M. Hotoleanu, “Photodarkening Rate in Yb-doped Silica Fibers,” Appl. Opt. 47, 1247–1256 (2008). [CrossRef] [PubMed]
14. J. Koplow, D. Kliner, and L. Goldberg, “Single-mode operation of a coiled multimode fiber amplifier,” Optics Lett. 25, 442 (2000).
15. R. Paschotta, J. Nilsson, A. C. Tropper, and D. C. Hanna, “Ytterbium-Doped Fiber Amplifiers,” IEEE J. Quantum Electron. 33, 7 (1997).
16. K. E. Mattson, S. N. Knudsen, B. Cadier, and T. Robin, “Photo darkening in ytterbium co-doped silica material,” Fiber Lasers V: Technology, Systems, and Applications, J. Broeng, C. Headley, D. V. Gapontsev, and D. Kliner, Proc. SPIE6873, 6873-48 (2008).
17. M. Hotoleanu, J. Koponen, and T. Kokki, “Spatial Distribution of Photodarkening in Large Mode Area Ytterbium Doped Fibers,” in Advanced Solid-State Photonics Conference 2008, Technical digest (Optical Society of America, 2008), paper WE23.
18. R. Olshansky and S. M. Oaks, “Differential mode attenuation measurements in graded-index fibers,” Appl. Opt. 17, 1830–1835 (1978). [CrossRef] [PubMed]
