• No results found

On the origin of photodarkening resistance in Yb-doped silica fibers with high aluminum concentration

N/A
N/A
Protected

Academic year: 2022

Share "On the origin of photodarkening resistance in Yb-doped silica fibers with high aluminum concentration"

Copied!
12
0
0

Loading.... (view fulltext now)

Full text

(1)

On the origin of photodarkening resistance in Yb-doped silica fibers with high aluminum concentration

M AGNUS E NGHOLM ,

1,*

M ATTHEW T UGGLE ,

2

C OURTNEY

K UCERA ,

2

T HOMAS H AWKINS ,

2

P ETER D RAGIC ,

3

AND J OHN

B ALLATO

2

1Division of Electronics Design, Mid-Sweden University, SE-851 70 Sundsvall, Sweden

2Center for Optical Materials Science and Engineering Technologies (COMSET) and the Department of Materials Science and Engineering, Clemson University, Clemson, South Carolina 29634, USA

3Department of Electrical and Computer Engineering, University of Illinois at Urbana-Champaign, Urbana, Illinois 61822, USA

*magnus.engholm@miun.se

Abstract: Ytterbium (Yb) doped silica fibers are widely used in high power fiber lasers where co-doping the silica core material with other elements is pivotal for high efficiency, low detrimental effects, and reliable optical properties. Aluminum (Al) is one of the most preferred co-dopants, yet, purely Yb/Al-doped silica fibers suffer from relatively high levels of photodarkening (PD) when used in laser applications. A slightly improved PD-resistance has been reported for increasing Al-doping concentration. However, the source of this improved performance is still unknown. In this article, we present the origin of the improved PD-resistance observed in Yb-doped silica fibers with high Al-concentration. It is found that a high Al-co-doping concentration reduces the interaction strength between the Yb-ion and nearby oxygen ions, resulting in a significant PD-resistance with negligible induced loss in the entire visible spectral range. A negligible PD is observed even for significantly higher Yb-concentrations than commonly used in commercially available Yb-doped silica fibers.

© 2020 Optical Society of America under the terms of theOSA Open Access Publishing Agreement

1. Introduction

Optical fiber lasers have experienced a surge in power-scaling capabilities in recent years [1–3], most notably with those based on Yb-doped glasses. Maximum output power capabilities in these systems currently exceed multiple-kW levels for continuous wave operation [4] and approach MW peak power during pulsed operation [5]. These incredibly high power levels are made possible through continually expanding the boundaries for the gain material and fiber design [6]

and increasing the threshold for detrimental optical non-linearities, such as stimulated Brillouin scattering (SBS), [7] stimulated Raman scattering (SRS) [8], transverse mode instabilities (TMI) [9], and photodarkening (PD) [10]. While both SBS and SRS strongly depend on the effective mode area and fiber length, all of these detrimental effects can also be minimized by careful design of the core material. The material aspects of PD have been given considerable attention during the last two decades where several research groups have contributed to an increased knowledge related to the dynamics of the PD-process [10–22]. PD in Yb-doped fiber lasers is observed as a degradation in output laser power over time, limiting the operational lifetime [10]. The formation of color centers with influential absorption bands in the visible spectral range causes the power degradation due to an induced core-loss that extends to the pump- and lasing wavelengths near 1 µm. Some of the most successful mitigation techniques so far include co-doping the core glass with phosphorous (P) [17,18], a combination of aluminum and phosphorous (Al/P) [19], cerium (Ce) [20], or by post-processing methods such as hydrogen- or deuterium loading [21,22].

#413766 https://doi.org/10.1364/OME.413766

Journal © 2021 Received 2 Nov 2020; revised 24 Nov 2020; accepted 1 Dec 2020; published 10 Dec 2020

(2)

Of these approaches, co-doping with Al is of particular interest as the doping process is mature, offering high-yield silica preforms with accurate and controllable refractive index profiles.

Important for laser applications, the resulting fibers possess excellent mechanical- and thermal properties. Furthermore, fibers co-doped with Al are also known for their low background loss and, more importantly, allow for the fabrication of Yb-doped silica fibers with relatively high absorption- and emission cross-sections, enabling the use of reasonable active fiber length [23].

Although many advantages of using Al-codoping in Yb-doped silica fibers exist, they suffer from high PD-levels unless they are simultaneously co-doped with P and/or Ce. Interestingly, it has been reported that the PD-level in purely Yb/Al-doped silica fibers decreases (while still far from adequately) when increasing the Al concentration [24,25]. Jetschke, et al. [24], compared Yb/Al-doped silica fibers with different Yb and Al concentrations and observed a correlation between the PD-level and the absolute Al-concentration rather than the Al/Yb ratio.

Their investigation covered concentrations up to 6.5 mol% Al

2

O

3

and up to 1.7 mol% Yb

2

O

3

, which, in the case of Al-codoping, is close to the upper limit tenable using Modified Chemical Vapor Deposition (MCVD) [26]. However, the true origin of the observed correlation between increased Al concentration and improved PD resistivity is still unknown. Al-codoping increases the solubility of rare-earth (RE) ions in silica glass and helps to reduce clustering [27]. Hence, a common hypothesis (although still not proven) is that the clustering of Yb-ions has a major influence on the PD dynamics considering the observed correlation between PD and excitation density (C

Yb

* inversion) [11].

To determine the true origin of the reduced PD when increasing Al-codoping, we have investigated Yb-doped YAG-derived (Yb:YAG-d) all-glass silica fibers drawn using the molten core method (MCM) [28–31]. The MCM offers the opportunity to manufacture highly-doped silica-based optical fibers with core compositions challenging to fabricate using more common preform manufacturing methods, e.g., MCVD. One unique feature of MCM-derived fibers is the inherent incorporation of SiO

2

into the core due to a diffusion-controlled dissolution process from the silica cladding during fiber drawing [32]. The incorporation of silica in the molten YAG core results in the formation of a glassy yttrium aluminosilicate core, but with a higher Al

2

O

3

concentration than silica fibers made using MCVD. High co-doping levels of Al in silica fibers have also gained attention as intrinsically low optical non-linearity gain media for fiber lasers and amplifiers. In particular, the sesquioxide components (Y

2

O

3

, Al

2

O

3

, and Yb

2

O

3

) are known to increase the thresholds for SBS and SRS [7,8]. However, high co-doping levels do often yield an undesirable increase in the refractive index and core numerical aperture (NA), making single-mode operation challenging. A more in-depth spectroscopic investigation on this unique core composition is still highly relevant. It offers valuable insights on the interplay of the atomic processes accountable for the induced optical losses, most notably PD in high power silica fiber lasers, that could push the technology forward for advanced fiber laser gain materials.

This article presents an investigation on Yb:YAG-d optical fibers with high Al

2

O

3

concentration

exhibiting a significantly increased PD resistance compared to MCVD-made Yb/Al-doped silica

fibers. This increased PD resistance is found to originate from the lower probability for interactions

between the Yb

3+

-ions and the host lattice, as determined by a blue-shifted charge-transfer (CT)

absorption band in the ultraviolet (UV) spectral range, similar to that observed in Yb-doped silica

fibers co-doped with phosphorous (P) [18]. However, in contrast to Yb/P-doped silica fibers, the

absorption cross-section for the 4f-4f transitions near 1 µm has similar intensity to Yb/Al-doped

silica fibers. Hence, Yb-doped silica fibers with high Al-concentration appear to be an attractive

choice as laser gain material, provided one can limit the undesirable increase in refractive index

in order to support single-mode operation.

(3)

2. Experimental

Four different Yb-doped yttrium aluminosilicate glass core optical fibers were drawn using the MCM, utilizing Yb:YAG crystals as the precursor material [33]. Two of the fibers were formed by inserting Yb:YAG rods (Roditi International, London, UK) into a 3 mm inner by 30 mm outer diameter telecommunications-grade silica capillary preform (Heraeus-Tenevo, Buford, GA, USA) that served as the cladding glass. Two additional fibers were drawn using Yb:YAG rods (1 mm outer diameter), sleeved inside a pure sapphire (Al

2

O

3

) tube (1.1 mm inner by 1.5 mm outer diameter, Saint-Gobain, France), inserted into respective 3 mm inner by 30 mm outer diameter preforms. This concentric YAG/sapphire precursor assembly was then sleeved inside the pure silica cladding tube. The main purpose of the sapphire sleeve is to further increase the concentration of Al

2

O

3

in the resultant core composition [34,35]. All preforms were heated to approximately 2000°C, above the melting point of the core YAG phase and where the cladding glass is sufficiently softened to draw into fiber. Given the high quenching rates (on order of 2000 K/s) and dissolution of SiO

2

from the cladding into the core melt during the draw, the Yb-doped yttrium aluminosilicate molten core is quenched into a glassy state upon cooling of the fiber as it exits the draw furnace. In addition, a Yb/Al-doped reference fiber was manufactured by using standard MCVD preform manufacturing techniques as described elsewhere [18]. All fibers were drawn to a 125 µm outer diameter and a conventional (higher-index) acrylate coating was applied during the draw yielding a total outer diameter of approximately 250 µm.

The core composition of the fibers was determined using a Tescan MAYA3 Scanning Electron Microscope (SEM) equipped with an Oxford instruments X-MAX 50 Energy Dispersive X-ray Spectroscopy (EDS). UV/visible transmission measurements on the fiber samples were conducted using a combined deuterium and halogen light source coupled to an Ocean Optics 2000+

(200–1000 nm) or Ocean Optics MAYA (170–330 nm) spectrometer. Short fiber samples with different lengths were spliced in-between two solarization-resistant fibers, where each fiber sample was measured several times to ensure high accuracy. More specifically, the sample length was decreased for measurements in wavelength regions with stronger absorption. The resulting spectra were merged, thereby enabling a wide dynamic range spanning the UV to visible. Measurements in the NIR spectral range were made using a similar procedure but with an ANDO AQ-6315E optical spectrum analyzer. PD experiments were performed by using a wavelength division multiplexer (WDM) (Thorlabs WD202E-1) where short fiber samples (10–30 mm) were spliced to the WDM and core pumped using a 450 mW, 976 nm pump laser (Oclaro LC95A76ULR-20RC2). All fiber samples were pumped close to the pump saturation level, reaching 45–50% level of inversion across the entire sample length, as verified by simulations using the RP Fiber Power simulation software [36]. The transmission loss over time was measured (with the pump diode off) in the range 450–800 nm using the Ocean Optics 2000 + spectrometer and a fiber-coupled white light source (halogen lamp).

3. Results

3.1. Intrinsic properties of Yb:YAG derived fibers

Due to the inherent dissolution process of silica into the YAG core during the fiber drawing, the amount of silica in the core and the core diameter highly depends on the fiber drawing time (for a set temperature as performed here). In general, the concentration of silica increases, and the core diameter decreases as the fiber drawing progresses; see details in Cavillon, et al. [32].

Figure 1 shows a typical compositional profile for a Yb:YAG-d fiber (Fiber 2; see Table 1) and

a Yb:YAG-d fiber with the additional sapphire sleeve (Fiber 3; see Table 1). The use of an

additional sapphire sleeve resulted in an increased Al-concentration and reduced effective Y- and

Yb-concentrations due to the dilution by Al

2

O

3

in the core.

(4)

Fig. 1. Typical compositional profiles for the Yb:YAG-d fibers. Fiber 2 (solid lines) and Fiber 3 with an additional sapphire sleeve (dashed lines). A relative increase in Al and decrease in Y are observed for the sapphire sleeved Yb:YAG-d Fiber 3.

Table 1. Measured physical parameters of the Yb:YAG-d fibers (all conc. in mole % oxide)

Fiber 1 2 3 4 Ref (MCVD)

Yb conc. in YAG rod 3.0 5.6 3.0 5.6 –

Sapphire sleeve – – Yes Yes –

Avg. Si conc. (SiO2) 71.9 79.1 78.7 86.9 97.4

Avg. Al conc. (Al2O3) 16.5 12.3 17.6 11.1 2.0

Avg. Y conc. (Y2O3) 11.0 7.8 3.5 1.8 –

Avg. Yb conc. (Yb2O3) 0.60 0.87 0.12 0.25 0.55

Core diameter (2a) (µm) 25.0 10.2 11.0 4.3 20

Al/Yb ratio 27 14 146 44 3.7

Al/Y ratio 1.5 1.6 5.0 6.2 –

Table 1 provides a summary of the physical properties for the four investigated Yb:YAG-d fibers. The radial profile of the core differs between the fibers due to the dissolution of silica.

Therefore, the core diameter of each fiber is defined as the distance between the points of maximum slope (differential) for the compositional profile of silica. This core diameter is also used for determining the average elemental concentration. The additional MCVD-made Yb/Al-doped silica fiber is included and was used for comparison in this investigation.

Figure 2 shows the absorption spectra for Fibers 1 and 4 in both the ultraviolet (UV) and near-infrared (NIR) spectral range (inset). Several interesting observations are unveiled compared to the Yb/Al-doped MCVD-made reference fiber. The NIR spectra show strong similarities, both in shape and cross-section, exposing a similar local, disordered symmetry around the Yb-ions. Concomittantly, some rather unique features are observed in the UV spectral range.

In regards to Yb-doped materials, spectroscopy in the UV spectral range is strongly associated with local interactions between the Yb-ion and the host lattice, as well as inter-configurational 4f

14

– 4f

13

5d

1

transitions of divalent Yb-ions (Yb

2+

) [37]. A first observation is the absorption band appearing around 330 nm in Fibers 1 and 4, which is not present in the MCVD-made fiber.

Although not shown in the Figure, this 330 nm absorption is observed for all Yb:YAG-d fibers

(5)

listed in Table 1 and indicates the presence of Yb

2+

-ions in all as-drawn Yb:YAG-d fibers [37]. A second observation is the different spectral shape at shorter wavelengths for the sapphire sleeved fiber (Fiber 4) as compared to the non-sleeved counterpart (Fiber 1), indicating that a higher Al/Yb ratio influences the local environment around the Yb-ions. A third observation is the comparable absolute value of the absorption band at 330 nm, although Fiber 1 has nearly 2.4 times higher Yb-concentration. The fraction of Yb

2+

-ions is much higher in the sapphire-sleeved Fiber 4, which is likely related to the higher Al-concentration.

Fig. 2. Ultraviolet (UV) absorption spectrum of Yb:YAG-d Fiber 1 (blue solid curve) and sapphire sleeved Fiber 4 (black solid curve) in comparison with charge-transfer (CT)-band of the Yb/Al-doped MCVD made fiber (red solid curve). The CT-band in a Yb:YAG crystal is also shown for comparison (yellow dashed curve). The band appearing near 330 nm for Fibers 1 and 4 is due to inter-configurational transitions of Yb

2+

. Comparison of the absorption cross-section for Yb:YAG-d Fiber 4 and the Yb/Al-doped reference fiber (inset).

The origin of Yb

2+

ions in the YAG-d fibers is believed to originate from the MCM fiber drawing process as no Yb

2+

could be observed in the Yb:YAG precursor crystal. The hypothesis is that the formation of divalent Yb is due to the release of oxygen in the high-temperature melt (2000°C) of mixed YAG material with dissolved silica. Still, the observation of Yb

2+

is important, especially considering it can have profound implications for YAG-d fibers developed for use in laser applications. Yb-doped silica glass with Yb

2+

ions are known to reduce the efficiency of Yb-doped high power fiber lasers [38]. Our findings here imply it can be challenging to manufacture Yb:YAG-d fibers using the MCM without forming some fraction of Yb

2+

-ions.

The valence stability of the Yb

3+

-ion also plays an essential role in the induced optical losses (photodarkening) in Yb-doped fiber lasers [39]. In general, light absorbing color centers are formed in the Yb-doped material via a pair-creation process upon excitation to the charge transfer (CT) absorption band in the UV spectral range [40]. More specific, the spectral position of the CT-band can be seen as an indication of the interaction strength between the Yb

3+

-ion and the host lattice, i.e., the energy required to transfer an electron from a nearby oxygen to the Yb

3+

-ion.

However, the true spectral position of the CT-band is difficult to determine in the case of the Yb:YAG-d fibers as it is obscured by the superimposed Yb

2+

related absorption bands.

By comparing the absolute values for the absorption band at 330 nm, the relative fraction

of Yb

2+

-ions appear to be higher in the sapphire sleeved Fiber 4, considering the difference in

Yb-concentration. Using the absorption cross-section value for the 330 nm band as estimated

(6)

by Kirchhof, et al. [41], the calculated fraction of Yb

2+

ions in Fibers 1 and 4 are 3.3% and 6.6% respectively. The calculated fraction of Yb

2+

is found to be even higher in sapphire sleeved fiber with a lower Yb-concentration, Fiber 3, nearly 13% Yb

2+

. Fiber 3 also has the highest Al/Yb ratio as well as the highest Al-concentration of the four YAG-d fibers. It appears the total Al-concentration has a large influence on the valence state of the Yb-ion as also recently observed by others [42].

3.2. Oxidation of the YAG-d fibers by thermal treatment in air

To better investigate the true spectral position of the CT-band, short, 8–10 cm sections of the YAG-d fibers were oxidized by thermal treatment in air at 800°C for up to 12 hours using a tube furnace. The UV spectra resulting from this oxidizing experiment are shown in Fig. 3 for Fibers 1 and 4. It is clearly seen that the characteristic absorption band of Yb

2+

at 330 nm as well as the shoulder at 230 nm vanishes, suggesting all Yb-ions are now in their trivalent state. Interestingly, and somewhat surprisingly, no clear CT-band structure is present as normally observed in Yb/Al MCVD-made fibers (compare with Yb/Al reference fiber in Fig. 2). Only the onset of the CT-band is unveiled, which has a clear blue-shift compared to the Yb/Al-doped reference fiber. Another observation is the nearly 6 times lower intensity of the CT-band for the sapphire-sleeved Fiber 4, which is significantly lower than expected considering the difference in Yb-concentration is only ∼2.4 times. The origin for this discrepancy is likely related to the difference in Al/Yb ratio.

Fig. 3. UV absorption spectra of Fiber 1 (blue) and Fiber 4 (black) before (solid) and after (dashed) oxidation by thermal treatment at 800°C for 12 hours.

There is a clear difference in the spectral shape at shorter wavelengths between Fiber 1 and

Fiber 4. The superimposed absorption bands related to Yb

2+

-ions in the 200–260 nm range

has more structure and higher intensity for the sapphire sleeved Fiber 4 compared to Fiber

1, which has only a minor shoulder near 230 nm. Likely this difference is due to the higher

yttria concentration in Fiber 1 and/or difference in Al/Y ratio (∼1.5 vs. ∼6.2). A higher yttria

concentration could influence the onset of the bandgap in the YAG-d fibers as Y

2

O

3

, which is

more covalent, has a smaller bandgap than Al

2

O

3

[43]

(7)

3.3. Photodarkening characteristics

The fibers in Table 1 were continuously core pumped at 976 nm to evaluate the induced losses over time until saturation. The Yb:YAG-d fibers (Fibers 1–4) exhibited large spectroscopic differences compared to the MCVD-made reference fiber. Figure 4 depicts the saturated induced loss for Fiber 1 and Fiber 4 along with the Yb/Al doped reference fiber in the range 460–800 nm.

A surprisingly low, almost negligible loss is observed for the Yb:YAG-d fibers for wavelengths longer than 600 nm. For shorter wavelengths only the onset of an induced absorption band is seen for Fiber 1. The exact origin of this absorption band is difficult to determine due to a lower dynamic sensitivity at these wavelengths, but believed to be related to the formation of additional Yb

2+

-ions during the PD process. A significantly reduced loss at shorter wavelengths is observed for the sapphire sleeved Fiber 4. Considering the above hypothesis, the fraction of Yb

2+

-ions is higher in the sapphire sleeved fibers (see Section 3.1) and therefore closer to the finite fraction of Yb

2+

-ions that can form in silica fibers, which could explain the lower loss at shorter wavelengths observed in Fiber 4.

Fig. 4. Saturated induced core loss for the as-drawn Fiber 1 and Fiber 4 (solid blue- and black curves), the thermally treated, oxidized, Fiber 1 and Fiber 4 (dashed blue- and black curve). The saturated loss for the MVCD-made Yb/Al silica reference fiber (red solid line, right y-axis scale).

Considering that the as-drawn Yb:YAG-d fibers reported in this manuscript comprise a relatively large fraction of Yb

2+

ions, one would expect that the observed PD resistivity and the presence of Yb

2+

are related [44]. Jetschke, et al. reported a substantial improvement of PD resistance in Yb-doped silica fibers prepared under reducing conditions [44]. However, the PD characteristics of the oxidized (thermally treated) Yb:YAG-d fibers (Fiber 1 and Fiber 4), possessing only trivalent Yb-ions, show an even further improved PD resistance; see blue- and black dashed curves in Fig. 4. In fact, the saturated induced loss is negligible in the entire visible spectral range. This negligible loss for the oxidized fibers is a remarkable finding, suggesting more complex dynamics in the PD process.

Figure 5 shows the induced loss measured at 600 nm as a function of time for Fiber 1, Fiber

4 and the Yb/Al-doped reference fiber. Fiber 2 (not shown) demonstrates very similar PD

characteristics as Fiber 1, even though Fiber 2 has a higher Yb-concentration. All Yb:YAG-d

fibers exhibit negligible loss over time measured at 600 nm, whereas the induced loss for the

Yb/Al-doped reference fiber exceeded 400 dB/m before saturation. Although not directly seen

(8)

in Fig. 5, the induced loss at shorter wavelengths for the as-drawn Fiber 1 and Fiber 4 occur immediately upon core pumping at 976 nm with a rapid saturation, which suggests a termination of the PD process.

Fig. 5. Induced loss measured at 600 nm as a function of time for the as-drawn Fiber 1 and Fiber 4 (blue- and black solid- curves), the thermally treated, oxidized, Fiber 1 (blue dashed curve) (left y-axis scale). The induced loss over time for the MVCD-made Yb/Al silica reference fiber (red solid line, right y-axis scale).

4. Discussion

A significant PD resistance is observed for all as-drawn Yb:YAG-d fibers at wavelengths longer than 600 nm; more interestingly, the thermally treated (oxidized) Yb:YAG-d fibers demonstrated a negligible induced loss in the entire visible spectral range. The thermal history of the molten core-derived fibers has been shown to be equivalent to conventional fibers, i.e., those drawn from MCVD-fabricated preforms [45]. This is expected since both types of fibers experience equivalent fiber draw conditions. Accordingly, the observed PD properties of the fibers treated herein are considered to result from their novel compositions and not the fabrication process. It becomes surprising that this negligible induced loss is obtained for silica fibers with relatively high Yb-concentration up to 0.87 mol% Yb

2

O

3

. This is typically 4 times higher than normally used in commercially available Yb-doped silica fibers and approaching concentration levels demonstrated in Yb-doped phosphate fibers [46,47]. We attribute these remarkable findings to the blue-shifted CT-band resulting from the high alumina concentration that significantly decreases the interactions between the Yb

3+

-ions and host lattice. A correlation between the spectral position of the CT-band and the level of induced losses has been recognized earlier in, e.g., Yb-doped phosphosilicate fibers [18]. The high PD resistivity observed for the oxidized Yb:YAG-d fibers is therefore expected (to some degree). However, the small induced loss at shorter wavelengths for the as-drawn Yb:YAG-d fibers, which appears to result from additionally formed Yb

2+

-ions, is still surprising. The true dynamics for this process as well as the energy transfer route to the CT-state with NIR-photons are under further investigations.

Determining the origin of the blue-shifted CT-band in the Yb:YAG-d fibers (compared to

MCVD-made Yb/Al-doped silica fibers) is still pivotal for the interpretation of the observed results

and deserves further discussion. The CT-band in Yb-doped silica glass is a complex superposition

of CT sub-bands introduced by different local glass environments, ion covalency, coordination

(9)

symmetry and coordination number [18,48,49]. Based on the Author’s experience, covalency is the single most important parameter influencing the spectral position and intensity of the CT-band, which in turn depends on the choice and concentration of co-dopants. Interestingly, the results presented here differ from those recently presented by Okazaki, et al., [48] of Yb/Al-doped silica glass with different Al-concentrations. They determined the spectral position of the CT sub-band associated with Yb-O-Al linkages at an energy of ∼5.05 eV (∼245 nm) and could also correlate an intensity increase of this band with increasing Al-concentration. A simultaneous decrease in the CT sub-band associated with Yb-O-Si linkages at ∼5.62 eV (∼221 nm) was also observed for increasing Al/Yb ratio. Unger, et al. [42] recently presented an extensive investigation of the optical properties in Yb/Al-doped silica glasses. In contrast to Okazaki, et al., a different spectral shape of the CT-band was observed, indicating the intensity of the Yb-O-Al related CT sub-band decreased rather than increased for constant Yb-concentration and increasing Al/Yb ratio. Most likely, the difference is explained by the higher Yb-concentration used by Unger, et al., ∼ 0.4 mol%, while the investigation by Okazaki, et al. did not include samples with Yb-concentrations exceeding 0.18 mol%. Apparently, the Yb-concentration also has a considerable influence on the spectral position of the CT-band and highlights the complex nature of CT-transitions in Yb-doped glass materials.

With increasing Al-concentration it is reasonable to assume that the amount of Yb-O-Al linkages will increase while the amount of Yb-O-Si linkages should decrease. The electronegativity of the cation in the second coordination shell has earlier been shown to influence the spectral position of the CT-band due to the high polarizability of the oxygen ion [18]. An increased amount of Al in the second coordination shell around Yb should therefore red-shift, rather than blue-shift the spectral position of the CT-band (increased Al/Si ratio). However, the electronegativity of the cation in the second coordination shell is not the only factor determining the degree of covalency between Yb and O. Also the Yb – O bond length and coordination number can have a significant influence on covalency. Okazaki, et al., [48] reported a nearly constant Yb – O distance for increasing Al/Yb ratio and a general trend of an increasing mean-bond distance for increasing coordination number when comparing different Yb/Al-doped crystalline materials. The Yb:YAG crystal has an 8-fold coordination and a longer Yb – O mean-bond distance than those found in Yb/Al-doped silica glass, which likely explain the reduced covalency and observed blue-shifted CT-band (see yellow dashed curve in Fig. 2). Based on the above arguments, we attribute the reduced covalency between Yb and O a consequence of an increased coordination number of oxygen around the Yb

3+

-ions also in the Yb:YAG-d fibers indicated by the blue-shifted CT-band.

It is not unlikely that the high Al concentration in the YAG/silica glass mixture promotes a local chemical environment around the Yb-ions with a resemblance of the 8-fold coordinated YAG-crystalline structure.

Considering there is a rather large difference in the local chemical environment around the

different Yb-ions in the YAG-d fibers, as evidenced by the shifted CT-band, it is noteworthy to

mention that there is just a minor spectroscopic difference observed in the NIR spectral range

(inset of Fig. 2). There is a striking similarity in spectral shape and absorption cross-section

with the MCVD made Yb/Al-doped silica fiber, which has a significantly lower Al-concentration

and Al/Yb ratio. This fact suggests that the first coordination shell (oxygen) has a slightly lower

influence on the spectroscopic properties of the

2

F

7/2

2

F

5/2

transitions in the 4f

13

manifold

compared to the more distant chemical environment around the Yb-ion. Further support is

given by comparing the spectroscopic properties observed for the Yb:YAG-d fibers with e.g.,

Yb/P-doped silica glass for an increasing P-concentration. Okazaki, et al. [49] performed an

extensive investigation on the local structure and optical properties of Yb/P-doped silica glass for

increasing P-concentration and observed considerable spectral changes for an increasing P/Yb

ratio. The observed spectral changes, i.e., shape and lower (∼0.5x) integrated cross-section, were

mainly attributed to an increased coordination number (from 6 to 8) and a higher symmetry

(10)

(more ordered structure) around the Yb

3+

-ion. The higher symmetry was determined to originate from the formation of microcrystalline phases of YbPO

4

, Yb(PO

3

)

3

, and YbP

5

O

14

-like structures depending on the P/Yb ratio. In contrast to the results observed for Yb/P-doped silica, our findings presented here suggests the chemical environment of the high Al/Yb-doped YAG-d fibers remains disordered even for high Al-doping concentrations and high Al/Yb ratio, with no sign of formed microcrystalline phases of Yb and Al. Hence, we propose that the chemical environment around the Yb

3+

-ions has a low symmetry preserving the relatively high absorption- and emission cross-sections of the

2

F

7/2

2

F

5/2

transitions.

Purely Yb/Al-doped silica fibers are perhaps the most attractive glass composition for use in high power fiber lasers offering high absorption- and emission cross-sections, excellent mechanical- and thermal properties and the use of a mature fabrication technology. The only major issue with this composition is the relatively high PD-levels observed even for moderate Yb-doping concentrations. Adding Ce to the core glass composition is a well known mitigation technique [20], but usually requires the addition of yet another element like Fluorine or Boron to reduce the refractive index and core numerical aperture. As presented in this article, using high Al-co-doping concentrations can reduce PD to negligible levels, even for Yb-concentrations significantly higher than currently used in commercial Yb-doped fibers. That said, using Yb:YAG-d fibers manufactured by the MCM may be a feasible route to high power fiber lasers with high PD resistance, although this fabrication technique is relatively young and several challenges remain before a core glass with ideal optical properties can be manufactured. However, there are other doping techniques compatible with MCVD that potentially could demonstrate Yb/Al-doped silica fibers with high Al-content locally around the Yb-ions. This would provide a Yb/Al-doped core glass with similar optical properties as presented here, but with a much lower average Al-concentration, reducing the need for additional elements to lower the refractive index.

Using synthesized nanoparticles (NPs) is one such method with demonstrated success in spectral engineering [50,51].

5. Conclusions

We have shown that Yb/Al-doped silica fibers with high Al-concentrations (in the case of this work,

> 11 mol%) exhibit negligible PD-levels even for significantly higher Yb-concentrations than those typically used in commercial Yb-doped silica fibers. Yb:YAG-d fibers with significantly higher Al

2

O

3

concentrations than normally reached with MCVD were manufactured using the MCM. The high PD-resistivity is correlated to a blue-shifted CT-band (relative to Yb/Al-doped MCVD made fibers) that decrease the covalency (and interaction strength) between the Yb

3+

-ions and oxygen ions in the first coordination shell. The blue-shifted CT-band is proposed to originate from an increased coordination number of oxygen ions around the Yb

3+

-ion lowering the covalency between Yb – O similar to that observed in the Yb:YAG single crystal. In addition, we propose that the local environment around the Yb-ions in highly Al-codoped silica glass has a low symmetry (more disordered) preserving relatively high absorption- and emission cross-sections.

This is opposite to that observed for, e.g., Yb/P-doped silica fibers with high P-concentration levels exhibiting significantly lower cross-sections due to a higher symmetry (less disorder) of the local environment. Our findings suggests that Yb/Al-doped silica fibers with a high, local, Al-content around the Yb-ions, manufactured by using, e.g., NP-doping technology, could be a very promising gain material for future high power fiber lasers.

Disclosures

The authors declare that there are no conflicts of interest related to this article.

(11)

References

1. M. N. Zervas and C. A. Codemard, “High power fiber lasers: a review,”IEEE J. Sel. Top. Quantum Electron.20(5), 219–241 (2014).

2. L. Dong and B. Samson, Fiber Lasers: Basics, Technology, and Applications (CRC Press, 2017).

3. Z. Liu, X. Jin, R. Su, P. Ma, and P. Zhou, “Development status of high power fiber lasers and their coherent beam combination,”Sci. China Inf. Sci.62(4), 41301 (2019).

4. K. Shima, S. Ikoma, K. Uchiyama, Y. Takubo, M. Kashiwagi, and D. Tanaka, “5-kw single stage all-fiber Yb-doped single-mode fiber laser for materials processing,”Proc. SPIE10512, 11 (2018).

5. G. Shi, S. Fu, Q. Sheng, J. Li, Q. Fang, H. Liu, A. Chavez-Pirson, N. Peyghambarian, W. Shi, and J. Yao,

“Megawatt-peak-power picosecond all-fiber based laser in MOPA using highly Yb3+doped LMA phosphate fiber,”

Opt. Commun.411, 133–137 (2018).

6. F. Kong, T. W. Hawkins, M. Jones, J. Parsons, M. Kalichemvsky-Dong, C. Dunn, and L. Dong, “Ytterbium-doped 30/400 LMA fibers with Record-Low∼NA of 0.028,” in Conference on Lasers and Electro-Optics (CLEO) (2016), paper SM2Q.2.

7. P. Dragic, P.-C. Law, J. Ballato, T. Hawkins, and P. Foy, “Brillouin spectroscopy of YAG-derived optical fibers,”Opt.

Express18(10), 10055–10067 (2010).

8. P. D. Dragic and J. Ballato, “Characterisation of Raman gain spectra in Yb:YAG-derived optical fibers,”Electron.

Lett.49(14), 895–897 (2013).

9. T. Eidam, C. Wirth, C. Jauregui, F. Stutzki, F. Jansen, H.-J. Otto, O. Schmidt, T. Schreiber, J. Limpert, and A.

Tünnermann, “Experimental observations of the threshold-like onset of mode instabilities in high power fiber amplifiers,”Opt. Express19(14), 13218–13224 (2011).

10. 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,” OSA Technical Digest Series (OFC, 2006), Paper No. OThC5.

11. J. Koponen, M. J. Soderlund, H. J. Hoffman, and S. K. T. Tammela, “Measuring photodarkening from single-mode ytterbium doped silica fibers,”Opt. Express14(24), 11539–11544 (2006).

12. M. J. Söderlund, J.J. Montiel i Ponsoda, J. P. Koplow, and S. Honkanen, “Thermal bleaching of photodarkening-induced loss in ytterbium-doped fibers,”Opt. Lett.34(17), 2637–2639 (2009).

13. R. Peretti, A.-M. Jurdyc, B. Jacquier, C. Gonnet, A. Pastouret, E. Burov, and O. Cavani, “How do traces of thulium explain photodarkening in Yb doped fibers?”Opt. Express18(19), 20455–20460 (2010).

14. F. Mady, M. Benabdesselam, and W. Blanc, “Thermoluminescence characterization of traps involved in the photodarkening of ytterbium-doped silica fibers,”Opt. Lett.35(21), 3541–3543 (2010).

15. H. Gebavi, S. Taccheo, D. Tregoat, A. Monteville, and T. Robin, “Photobleaching of photodarkening in ytterbium doped aluminosilicate fibers with 633 nm irradiation,”Opt. Mater. Express2(9), 1286–1291 (2012).

16. T. Deschamps, H. Vezin, C. Gonnet, and N. Ollier, “Evidence of AlOHC responsible for the radiation-induced darkening in Yb doped fiber,”Opt. Express21(7), 8382–8392 (2013).

17. A. V. Shubin, M. V. Yashkov, M. A. Melkumov, S. A. Smirnov, I. A. Bufetov, and E. M. Dianov, “Photodarkening of aluminosilicate and phosphosilicate Yb-doped fibers,” OSA Technical Digest Series (CLEO, Europe, 2007), Paper No. CJ3-1-THU

18. M. Engholm and L. Norin, “Preventing photodarkening in ytterbium-doped high power fiber lasers; correlation to the UV-transparency of the core glass,”Opt. Express16(2), 1260–1268 (2008).

19. S. Jetschke, S. Unger, A. Schwuchow, M. Leich, and J. Kirchof, “Efficient Yb laser fibers with low photodarkening by optimization of the core composition,”Opt. Express16(20), 15540–15545 (2008).

20. M. Engholm, P. Jelger, F. Laurell, and L. Norin, “Improved photodarkening resistivity in ytterbium-doped fiber lasers by cerium co-doping,”Opt. Lett.34(8), 1285–1287 (2009).

21. M. Engholm and L. Norin, “Reduction of photodarkening in Yb/Al doped fiber lasers,”Proc. SPIE6873, 68731E (2008).

22. N. Zhao, W. Li, J. Li, G. Zhou, and J. Li, “Elimination of the photodarkening effect in an yb-doped fiber laser with deuterium,”J. Lightwave Technol.37(13), 3021–3026 (2019).

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

24. S. Jetschke, S. Unger, M. Leich, and J. Kirchof, “Photodarkening kinetics as a function of Yb concentration and the role of Al-codoping,”Appl. Opt.51(32), 7758–7764 (2012).

25. B. Morasse, S. Chatigny, E. Gagnon, C. Hovington, J-P. Martin, and J-P de Sandro, “Low photodarkening single cladding ytterbium fibre amplifier,”Proc. SPIE6453, 64530H (2007).

26. K. A. Mat Sharif, N. Y. M. Omar, M. I. Zulkifli, S. Z. Muhamad Yassin, and H. A. Abdul-Rashid, “Fabrication of alumina-doped optical fiber preforms by an MCVD-metal chelate doping method,”Appl. Sci.10(20), 7231 (2020).

27. J. Ballato and P. Dragic, “On the clustering of rare earth-dopants in fiber lasers,” J. Dir. Energy 6(2), 175–181 (2017).

28. J. Ballato, T. Hawkins, P. Foy, B. Kokuoz, R. Stolen, C. McMillen, M. Daw, Z. Su, T. M. Tritt, M. Dubinskii, J.

Zhang, T. Sanamyan, and M. J. Matthewson, “On the fabrication of all-glass optical fibers from crystals,”J. Appl.

Phys.105(5), 053110 (2009).

29. P. D. Dragic, J. Ballato, T. Hawkins, and P. Foy, “Feasibility study of Yb:YAG-derived silicate fibers with large Yb content as gain media,”Opt. Mater.34(8), 1294–1298 (2012).

(12)

30. S. Zheng, J. Li, C. Yu, Q. Zhou, and D. Chen, “Preparation and characterization of Nd:YAG ceramic derived silica fibers drawn by post-feeding molten core approach,”Opt. Express24(21), 24248–24254 (2016).

31. M. Cavillon, C. Kucera, T. Hawkins, J. Dawson, P. Dragic, and J. Ballato, “A unified materials approach to mitigating optical nonlinearities in optical fiber. III. Canonical examples and materials roadmap,”Int. J. Appl. Glass Sci.9(4), 447–470 (2018).

32. M. Cavillon, P. Dragic, B. Faugas, T. W. Hawkins, and J. Ballato, “Insights and aspects to the modeling of the molten core method for optical fiber fabrication,”Materials12(18), 2898 (2019).

33. J. Ballato and A. Peacock, “Perspective: Molten core optical fiber fabrication – A route to new materials and applications,”APL Photonics3(12), 120903 (2018).

34. M. Tuggle, C. Kucera, T. Hawkins, D. Sligh, A. F. J. Runge, A. C. Peacock, P. Dragic, and J. Ballato, “Highly nonlinear yttrium-aluminosilicate optical fiber with a high intrinsic Brillouin scattering threshold,”Opt. Lett.42(23), 4849–4852 (2017).

35. P. Dragic, T. Hawkins, S. Morris, and J. Ballato, “Sapphire-derived all-glass optical fibers,”Nat. Photonics6(9), 627–633 (2012).

36.https://www.rp-photonics.com/fiberpower.html

37. M. Engholm, L. Norin, and D. Aberg, “Strong UV absorption and visible luminescence in ytterbium-doped aluminosilicate glass under UV excitation,”Opt. Lett.32(22), 3352–3354 (2007).

38. J. Kirchhof, S. Unger, A. Schwuchow, S. Grimm, and V. Reichel, “Materials for high-power fiber lasers,”J. Non-Cryst.

Solids352(23-25), 2399–2403 (2006).

39. S. Rydberg and M. Engholm, “Charge transfer processes and ultraviolet induced absorption in Yb:YAG single crystal laser materials,”J. Appl. Phys.113, 223510 (2013).

40. S. Rydberg and M. Engholm, “Experimental evidence for the formation of divalent ytterbium in the photodarkening process of Yb-doped fiber lasers,”Opt. Express21(6), 6681–6688 (2013).

41. J. Kirchhof, S. Unger, A. Schwuchow, S. Jetschke, V. Reichel, M. Leich, and A. Scheffel, “The influence of Yb2+ ions on optical properties and power stability of ytterbium doped laser fibers,”Proc. SPIE7598, 75980B (2009).

42. S. Unger, A. Schwuchow, J. Dellith, and J. Kirchhof, “Optical properties of ytterbium/alimunium doped silica glass,”

Opt. Mater. Express10(4), 907–925 (2020).

43. J. Robertson, “High dielectric constant oxides,”Eur. Phys. J.: Appl. Phys.28(3), 265–291 (2004).

44. S. Jetschke, S. Unger, A. Schwuchow, M. Leich, V. Reichel, and J. Kirchhof, “Photodarkening in Yb-doped silica fibers: influence of the atmosphere during preform collapsing,”Proc. SPIE6873, 68731G (2008).

45. P. D. Dragic, M. Cavillon, A. Ballato, and J. Ballato, “A unified materials approach to mitigating optical nonlinearities in optical fiber. II. B. The optical fiber, material additivity and the nonlinear coefficients,”Int. J. Appl. Glass Sci.

9(3), 307–318 (2018).

46. N. G. Boetti, D. Pugliese, E. Ceci-Ginistrelli, J. Lousteau, D. Janner, and D. Milanese, “Highly doped phosphate glass fibers for compact lasers and amplifiers: a review,”Appl. Sci.7(12), 1295 (2017).

47. E. Balliu, N. G. Boetti, D. Pugliese, J. Lousteau, M. Engholm, D. Milanese, and H. E. Nilsson, “Single-frequency, pulsed Yb3+-doped multicomponent phosphate power fiber amplifier,”J. Opt.22(11), 115606 (2020).

48. T. Okazaki, E. H. Sekiya, and K. Saito, “The difference in the coordination environment around Yb3+ions between Yb–Al-doped and Yb–P-doped silica glasses,”Jpn. J. Appl. Phys.59(10), 102003 (2020).

49. T. Okazaki, E. H. Sekiya, and K. Saito, “P concentration dependence of local structure around Yb3+ions and optical properties in Yb–P-doped silica glasses,”Jpn. J. Appl. Phys.58(6), 062001 (2019).

50. C. Kucera, B. Kokuoz, D. Edmonson, D. Griese, M. Miller, A. James, W. Baker, and J. Ballato, “Designer emission spectra through tailored energy transfer,”Opt. Lett.34(15), 2339–2341 (2009).

51. T. Lindstrom, E. Garber, D. Edmonson, T. Hawkins, Y. Chen, G. Turri, M. Bass, and J. Ballato, “Spectral engineering of optical preforms through active nanoparticle doping,”Opt. Mat. Exp.2(11), 1520–1528 (2012).

References

Related documents

A few copies of the complete dissertation are kept at major Swedish research libraries, while the summary alone is distributed internationally through the series Digital

Purpose The critical shoulder angle (CSA) and the acromion index (AI) are measurements of acromial shape reported as predictors of degenerative rotator cuff tears (RCT)

Situationen runt de båda här aktuella barnen kan inte rimligen ha varit den att alla övriga barn runt om skulle vara befriade från normal sexuell utveckling och från beteenden som

In conclusion, individuals with PsA are metabolically distinct with greater VAT and ectopic liver fat and lower thigh muscle volume than age-, sex- and BMI-matched healthy

ALP från njure och humant serum band inte in alls till kollagentäckta plattor men band in ganska starkt till både bovint kollagen typ I och II samt humant kollagen typ I vid

An effective approach to handle the solving procedure of the ODE (6.8) is that, first, discretize the space domain to yield a nonlinear system of algebraic equations, and then

The definition of in vivo amyloid (Nomenclature Committee of the International Society of Amyloidosis) declares that there has to be tissue deposits, with a

Deposition of Al-doped ZnO films by high power impulse magnetron sputtering. Linköping Studies in Science and Technology