Dopant segregation in YAG single crystal fibers grown by the laser heated pedestal growth technique
Introduction
Despite successful application of glass optical fibers in development of devices such as fiber optic sensors, fiber lasers, etc.; applications requiring extreme material properties have proven challenging given silica stability limits. Conditions such as high temperatures, radioactivity, high energy fluences, corrosion, etc., limit applicability of glass fibers. For such applications, single crystal (SC) fibers have shown exceptional promise. Crystalline fibers grown from materials like sapphire or YAG are thermally, chemically, and mechanically more robust compared to silica fibers, making them ideal for extreme environment sensing applications such as in boilers, jet engines, and nuclear reactors [1]. Additionally, high thermal conductivity and low nonlinear gain coefficients, coupled with high surface-area-to-volume ratio of the fiber geometry make SC fibers suitable for applications involving high optical power density [2], [3].
Single crystal fibers, grown by processes like the LHPG technique, do not intrinsically have a functional cladding. The cladding is not only essential to confine light and reduce frustrated total internal reflection that may arise due to contact with the surface when the fiber is deployed in various applications, but it also helps reduce the modal volume required for distributed sensing applications [4], and cladding pumping schemes in high power laser applications [5]. Glasses have been explored, to limited success, as potential cladding materials for crystal fibers [6], [7]. For high temperature or high-power laser applications, the differences in thermal and mechanical properties of glasses and crystals are exacerbated and lead to significant losses and ultimately catastrophic device failure. To preserve homogeneity of material, and exploit the advantages of a crystalline phase, growing a crystalline cladding is essential.
While a few post-cladding techniques (where a crystalline layer is grown on an existing fiber) have been proposed [8], these procedures usually involve two separate growth processes, which not only introduce scattering losses at the core-clad interface but also increase the time and cost involved for production. Most importantly, current reports on these alternative growth methods reveal limitations in either maximum length of clad fiber, maximum cladding thickness, temperature stability, or finished fiber quality [5], [8], [9], [10], [11]. Hence, a single step process is more desirable. Unlike drawing of glass fibers, starting from a rod-in-tube type preform is not effective as low viscosity in the liquid molten pool causes intermixing and homogenizes the melt composition [12]. Instead, in-situ fabrication of a graded index fiber can be accomplished by leveraging crystal growth melt dynamics.
Previous researchers have shown that certain LHPG-grown crystal fibers exhibit non-uniform radial distribution of a dopant species [13]. Depending on dopant ions, a tendency may be observed for the dopant species to segregate towards the center of the fiber or the periphery, creating a distribution profile. In magnesium (Mg) doped sapphire, for example, the center axis of the fiber showed a much higher concentration of Mg2+ ions compared to fiber edges [13]. Such an effect has also been observed in YAG SC fibers as well. While neodymium (Nd3+) ions were observed to concentrate at the center axis [14], [15], chromium (Cr4+) ions were observed to have higher concentration near the fiber periphery [16].
The segregation of the ions in the YAG matrix were argued to correlate with their size, as well as the differential solubility of the dopant in the solid and liquid phases of the host crystal matrix [17]. The latter is quantitatively described by the segregation coefficient (k) which is the ratio of the solubility in the solid phase to solubility in the liquid phase. Hence, if the value of k is 1, it implies there is no depletion or accumulation of dopants in the melt when the crystal grows, which is indicative of a homogenous solid solution in the crystal. The segregation coefficient also seems to be related to the relative size difference between the dopant ion and the host ion it replaces. Ions such as Nd3+ and cerium (Ce3+), which are larger in size compared to the yttrium (Y3+) ions (see Table 1), have k < 1 (see Table 2) and tend to segregate towards the center of the fiber. On the other hand, ions like Cr4+, which are smaller compared to Y3+ ions, have k > 1 and tend to segregate towards the outer rim of the fiber. In addition, ions such as Yb3+, which is very similar in size to Y3+ ion (see Table 1), has a reported k value of about 1 in YAG. During crystal growth, Yb ions indicate no tendency to self-segregate. The same should be also true for ions such as Ho, Er, Tm, etc. In short, the value of k determines where a particular dopant species will move radially with respect to the center axis of the fiber during growth. Since the refractive index of YAG changes linearly with addition of dopant ions (see Table 3), auto-segregation of dopant ions in YAG can be explored as a means to create an effective refractive index gradient to achieve a monolithic core-clad structured crystalline fiber.
In this work, we present a more careful look at growth of graded index single crystal fibers via adding dopants to pure fibers and re-growing them with LHPG. The selective segregation of dopants across the fiber cross-section demonstrates a concentration distribution profile, which is expected to also approximate a graded refractive-index profile. Control of the concentration distribution profile is engineered not only by varying dopant species and the fiber matrix material, but also growth conditions like fiber draw speed.
Section snippets
Experimental details
Doped YAG SC fibers presented in this study have been grown in a custom-designed LHPG apparatus, details of which can be found elsewhere [24]. A schematic of the LHPG apparatus is shown in Fig. 1. The Gaussian beam from a stabilized CO2 laser is transformed into a collimated symmetric ring by a pair of ZnSe axicons. The CO2 laser ring is then focused onto the tip of a feedstock-rod to form a mini-melt pool from which the crystal fiber is grown. Commercially available single crystal feedstock
Results
The occurrences of cracks along the fiber axis during the growth of highly doped Nd:YAG SC fibers are not uncommon, as the strong segregation of the dopant at the solid-melt interface leads to constitutional supercooling and subsequent structure loss [28]. However, the Nd:YAG and Nd-Ho:YAG SC fibers presented in this study were of good crystalline uniformity. No visible defects or cracking are seen in the fiber cross-section SEM micrographs recorded by secondary electron detection, as shown in
Discussion
Radial segregation of dopants in a crystal result arise due a non-planar growth interface and the incomplete mixing of the melt in contact to it. While there have been some extensive studies analyzing melt flows in float zone crystal growths [16], [30], [31], in practice, detailed numerical investigation is quite challenging as the shape of the molten zone and the crystal growth interface are highly dependent on growth conditions. In the following section, a phenomenological model is discussed
Conclusions
In this work, we presented experimental results demonstrating self-segregation of RE ions in LHPG grown YAG SC fibers. We believe that this technique can eventually be used to provide high-quality cladded single crystal fibers suitable for harsh-environment applications, owing to the intrinsic thermal and mechanical stability of bulk crystals. Preferential segregation of Nd3+ ions in Nd-Ho co doped YAG SC fibers have been demonstrated, furthering credibility of the dopant segregation theory
CRediT authorship contribution statement
Subhabrata Bera: Conceptualization, Methodology, Investigation, Formal analysis. Paul Ohodnicki: Project administration, Funding acquisition, Supervision. Keith Collins: Investigation, Formal analysis. Matthew Fortner: Resources. Yoosuf N. Picard: Investigation. Bo Liu: Software, Investigation. Michael Buric: Project administration, Funding acquisition.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgements
We acknowledge Marc De Graef at Carnegie Mellon University for the EBSD dynamical simulations.
The work was performed in support of the US Department of Energy’s Fossil Energy Crosscutting Technology Research Program. It was executed through the NETL Research and Innovation Center’s Advanced Sensors and Controls FWP. Research performed by Leidos Research Support Team staff was conducted under the RSS contract 89243318CFE000003. Part of the crystal growth work was supported by the HEL-JTO under
Disclosure
This work was funded by the Department of Energy, National Energy Technology Laboratory, an agency of the United States Government, through a support contract with Leidos Research Support Team (LRST). Neither the United States Government nor any agency thereof, nor any of their employees, nor LRST, nor any of their employees, makes any warranty, expressed or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus,
References (35)
Design and implementation of distributed ultra-high temperature sensing system with a single crystal fiber
J. Lightwave Technol.
(2018)Coilable single crystals fibers of doped-YAG for high power laser applications
J. Cryst. Growth
(2014)Simulation and experiment on laser-heated pedestal growth of chromium-doped yttrium aluminum garnet single-crystal fiber
J. Cryst. Growth
(2011)- et al.
The radial distribution of dopant (Cr, Nd, Yb, or Ce) in yttrium aluminum garnet (Y3Al5O12) single crystals grown by the micro-pulling-down method
J. Cryst. Growth
(2009) Stress-birefringence associated with facets of rare-earth garnets grown from the melt; a model and measurement of stress-birefringence observed in thin sections
J. Cryst. Growth
(1983)Distribution of ytterbium in Yb: YAG crystals and lattice parameters of the crystals
J. Cryst. Growth
(2003)- et al.
Crystal growth and perfection of large Nd: YAG single crystals
J. Cryst. Growth
(1972) Growth and lasing of single crystal YAG fibers with different Ho3+ concentrations
Opt. Mater.
(2018)- et al.
Radial dopant segregation in zero-gravity floating-zone crystal growth
J. Cryst. Growth
(1993) Three-dimensional simulation of floating-zone crystal growth of oxide crystals
J. Cryst. Growth
(2003)
Physical properties of a Y3Al5012 melt
J. Cryst. Growth
Marangoni convection in a floating zone under reduced gravity
J. Cryst. Growth
Some evidence for the existence and magnitude of a critical Marangoni number for the onset of oscillatory flow in crystal growth melts
J. Cryst. Growth
Single-crystal sapphire-based optical high-temperature sensor for harsh environments
Opt. Eng.
Predicted performance limits of yttrium aluminum garnet fiber lasers
Opt. Eng.
Low-loss ‘crystalline-core/crystalline-clad’ (C4) fibers for highly power scalable high efficiency fiber lasers
Opt. Express
Cited by (15)
Robust YAG:Ce single crystal for ultra-high efficiency laser lighting
2022, Journal of Rare EarthsCitation Excerpt :Despite this, the previous method of compounding organic resin and phosphor can no longer meet the requirements of high-power lasers because organic resins with poor thermal conductivity cannot handle the heat generated by high excitation power intensity. Therefore, YAG:Ce phosphor was carefully designed and combined with various materials to prepare phosphor transparent ceramics,15–18 phosphor in glasses (PiGs)19 and single-crystals20 to improve thermal stability. For example, Liu et al. prepared YAG:Ce transparent ceramics modified by MgO and SiO2, and the luminous efficiency was 223 lm/W under the laser power density of 3.5 W/mm2.21
Fabrication and evaluation of sol-gel derived magnesium aluminate spinel-clad sapphire fiber
2022, Optical Fiber TechnologyCitation Excerpt :Hence, one of the most important challenges in the realization of sapphire-based optical fiber sensors is the development of an appropriate cladding for sapphire fibers. Several strategies have been suggested for cladding single-crystal fibers, of which the wet chemical based post-cladding technique provides a valuable platform for experimentation [13–16]. Under this scheme, a cladding layer is deposited on the crystal fiber via sol–gel dip-coating and subjected to subsequent thermal treatment.
Monocrystalline fiber growth technique: New critical radius considerations
2021, Journal of Crystal GrowthCitation Excerpt :It’s well known that the LHPG technique produces high axial thermal gradients at the as-growth crystal fiber in comparison to other methods for crystal growth [1–10,13].
A novel route to fabricate Yb:YAG ceramic fiber and its optical performance
2021, Journal of the European Ceramic SocietyCitation Excerpt :Therefore, the kW level of output power has achieved the ceiling power of glass fiber laser and hardly increases furtherly due to the limitation of its physical properties of material [8,9]. Single crystal fibers combine the excellent properties of single crystal with the advantage of thermal management of fiber [3,10], having the potential of meeting the application requirements of high-power laser. YAG single crystal fiber is one of most representative single crystal fibers and has attracted much attentions.
Silver nanoparticles as a new sintering additive for the fabrication of YAG:Ce optical luminescent ceramics
2024, Journal of the American Ceramic Society