Hydrothermal Single-Crystal Growth of Lu2O3 and Lanthanide-Doped

ARTICLE pubs.acs.org/crystal

Hydrothermal Single-Crystal Growth of Lu2O3 and Lanthanide-Doped Lu2O3 Colin McMillen,† Daniel Thompson,‡ Terry Tritt,‡ and Joseph Kolis*,† †

Department of Chemistry and Center for Optical Materials Science and Engineering Technology and ‡Department of Physics and Astronomy,Clemson University, Clemson, South Carolina 29634, United States ABSTRACT: The hydrothermal growth of bulk single crystals of lutetium oxide (Lu2O3) is reported. Crystals were grown at 600 650 °C at 2 kbar in water with high concentrations (10 20 M) of KOH as a mineralizer. Under these conditions crystals formed spontaneously and could also be transported to suitable seeds to form large (3 8 mm/edge) single crystals. Experiments were done to optimize transport conditions and the quality of the crystals for the first time. The crystals were doped with trivalent laser active ions such as Er3+ and Yb3+. Absorption spectra of the doped materials were determined. Thermal conductivities of the pure single crystals as well as of the Yb-doped crystals were determined between room temperature and 77 K. The conductivity of the pure crystals increases significantly at lower temperatures as expected, and is significantly greater than that of YAG. The Yb-doped crystals have a thermal conductivity almost unchanged from the pure host, which is promising for use in high-energy laser applications.

1. INTRODUCTION The sesquioxides of scandium, yttrium, and lutetium, Ln2O3, have recently received a large amount of attention because of their potential application as host materials for 1.0 and 1.5 μm (eye-safe) solid-state lasers operating at high average powers.1 3 Thermal management is a key issue in high-power solid-state lasers, where the use of multiple diode pumps generates a significant amount of latent heat from the inherent quantum defects of the pumping process. This heat must be drawn away from the gain medium to prevent deleterious optical effects or even destruction of the coatings or gain medium itself. The Ln2O3 oxides are especially attractive candidates because of their robust thermal stability and high reported thermal conductivities.4,5 Scandia, yttria, and lutetia all melt at about 2450 °C and are all reported to have room-temperature thermal conductivities equaling or exceeding that of YAG. Improved thermal conductivity at cryogenic temperatures also makes these sesquioxides attractive for cryogenically cooled solid-state lasers which can further facilitate thermal management.6,7 Lutetia is of particular interest as a potential host because of its more favorable thermal properties when doped with lasing ions. Mass differences between dopant and host ions introduce phonon scattering in a material, typically lowering the thermal conductivity of a doped material compared to an undoped host. The magnitude of the drop in thermal conductivity depends on the mass mismatch between the host and dopant. In the cases of the typical dopants of interest, Yb3+ for 1 μm and Er3+ for 1.5 μm eye-safe emissions, the atomic mass of Lu is most similar to the dopants. Thus Yb:Lu2O3 and Er:Lu2O3 should exhibit higher thermal conductivities than similarly doped YAG, Y2O3, and Sc2O3. Griebner and co-workers have in fact reported that while r 2011 American Chemical Society

single crystals of YAG (11 W/mK) and Lu2O3 (∼12.5 W/mK) have somewhat similar room-temperature thermal conductivities, the thermal conductivity of Yb:YAG (6.8 W/mK) is only about 60% of that of similarly doped Yb:Lu2O3 (11 W/mK).4 Those authors also studied Yb:Y2O3 and Yb:Sc2O3 in the same work and found that although the thermal conductivities of undoped yttria (∼13.6 W/mK) and scandia (∼16.5 W/mK) exceeded that of lutetia, the thermal conductivities of Yb:Y2O3 (7.7 W/mK) and Yb:Sc2O3 (6.6 W/mK) were considerably lower than Yb:Lu2O3 at the same doping level. Later reports presented slightly adjusted thermal conductivity values but revealed the same overall trends in comparing the Yb-doped and undoped crystals.5 Like scandia and yttria, lutetia crystallizes in the cubic space group Ia3 and thus enjoys the advantages of broad absorption bands, large Stark splitting, and resistance to thermal lensing and anisotropic thermal expansion, making it a very attractive host targeted for high-power solid-state lasers.8,9 Despite these favorable thermal and optical properties, the growth of high optical quality single crystals of Lu2O3 is extremely limited due to its high melting point. Czochralski growth of the rare earth sesquioxides has proven particularly challenging, and more specialized growth techniques are thus needed to grow lutetia single crystals from a melt.10 In particular, the extreme temperatures create severe thermal strains and introduce a high level of defects from the growth crucible. Some initial improvements in crystal size were made by Petermann and co-workers at the University of Received: April 22, 2011 Revised: August 10, 2011 Published: August 11, 2011 4386

dx.doi.org/10.1021/cg2005166 | Cryst. Growth Des. 2011, 11, 4386–4391

Crystal Growth & Design Hamburg using the Bridgman technique, but difficulty seeding the melt prompted the development of the heat exchanger method by the same research group.11 This technique has been used to grow multi-centimeter polycrystalline boules having regions of single crystalline Lu2O3 from a melt in rhenium crucibles under a reducing atmosphere. The boules require a postgrowth annealing treatment to remove color centers that result from oxygen deficiencies, and rhenium inclusions resulting from oxidation of the crucible are commonly observed in the new growth even under highly reducing conditions.12 Nevertheless, good progress has been made in this difficult system and some very promising optical and lasing studies have been performed on Yb:Lu2O3 crystals fabricated from the annealed boules.4,5,13 15 To alleviate some of the overall challenges of crystal growth at such high temperatures, we have recently begun to explore novel hydrothermal approaches to the growth of extremely refractory oxides. We encountered some success with a number of high melting oxides including Sc2O3,16 ThO2,17 cubic HfO2,18 and KNbO3.19 In particular we recently demonstrated that Sc2O3 single crystals can be readily grown hydrothermally around 550 650 °C from aqueous hydroxide solutions.16,20 This new approach expands the limited field of rare earth sesquioxide crystal growth and, since the growth is performed in aqueous solution at only a quarter of the melting point of the oxides, has the potential to produce high-purity single crystals with low thermal strain. In this paper we describe the extension of this technology to the first hydrothermal growth of single crystals of Lu2O3, Yb: Lu2O3, and Er:Lu2O3 and discuss the hydrothermal phase stability and some other properties of these crystals. The absorption spectra of the doped crystals are reported, the single-crystal thermal conductivity has been measured for several samples, and the effect of doping on thermal conductivity is discussed.

2. EXPERIMENTAL SECTION Spontaneous nucleation experiments were used to determine the phase stability of Lu2O3 in aqueous hydroxide mineralizer solutions and to synthesize an initial seed crystal required to commence seeded transport growth. Experiments were performed in 3/8 in. o.d. silver ampules that were weld-sealed using a tungsten inert gas welder. Lu2O3 powder (1 g, Alfa Aesar, 99.99%) was first weighed into an ampule along with solid KOH pellets (2.25 4.5 g, MV Laboratories, 99.99%), and an appropriate amount of deionized water (2 4 mL) was then added to form an aqueous KOH mineralizer of 10 20 M concentration. The ampule was weld-sealed and placed in an autoclave, and the remaining volume of the autoclave was filled with deionized water to serve as counter pressure to prevent the ampule from bursting during the heating cycle. The autoclave was then heated from 550 to 660 °C for 7 days and cooled to room temperature (over about 12 h), and the ampule was subsequently opened. Er:Lu2O3 and Yb:Lu2O3 crystals were synthesized in the same manner, with substitution of Er2O3 (Aldrich, 99.9%) or Yb2O3 (Alfa Aesar, 99.9%) for Lu2O3 in the desired quantities. The resultant small crystals were identified by powder and singlecrystal X-ray diffraction (XRD). Powder diffraction was performed using a Scintag XDS 2000-2 diffractometer with Cu Kα radiation (λ = 1.5415 Å) and a solid-state Ge detector. The well-ground reaction products were analyzed from 5 to 65° in 2θ using a scan speed of 0.5°/min. Er:Lu2O3 and Yb:Lu2O3 crystals were further analyzed by energy dispersive X-ray analysis (EDX) and single-crystal XRD to verify the presence and approximate quantity of the dopant ions. EDX was performed using an Oxford INCA EDX analyzer attached to a Hitachi 3400N scanning electron microscope. Average dopant concentrations for each crystal were obtained using an accelerating voltage of 20 kV at multiple analysis

ARTICLE

points measured against a copper standard. Concentration is reported in atomic percent relative to the host metal site (for example, 3% Er:Lu2O3 represents a composition of Er0.06Lu1.94O3) . Unit cell parameters were obtained by single-crystal X-ray diffraction using 480 diffraction images collected with a Rigaku AFC8 diffractometer with Mo Kα (λ = 0.710 73 Å) radiation and a Mercury CCD area detector. Hydrothermal transport growth of larger Lu2O3 crystals was also performed in 3/8 in. o.d. welded silver ampules as well as directly in autoclaves containing fixed silver liners. The initial seed crystal was a relatively large crystal obtained from a spontaneous nucleation experiment as described above, and additional seeds were cultivated from the resulting bulk growth on that first crystal followed by repeated subsequent seeded growth experiments. Initially, Er- or Yb-doped lutetia crystals were grown on undoped seed crystals, and the first doped seeds were accordingly cut from that resulting growth. The seeds were drilled, measured, and weighed prior to being tied onto a silver ladder, and the ladder was then placed in the liner along with feedstock consisting of commercial powder or smaller crystals harvested from previous hydrothermal runs. Solid KOH and deionized water were also added to reach a target mineralizer concentration of 20 M KOH and a percent fill of 50 70% in the liner. The liner and autoclave were sealed and heated using two ceramic band heaters strapped around the dissolution and growth zones. The temperature of the dissolution zone ranged from 630 to 660 °C. The temperature of the growth zone was 20 50 °C cooler, establishing a positive solubility gradient between the dissolution and growth zones. These growth experiments lasted about 2 6 weeks and were terminated by cooling to room temperature over a 12 h period. The as-grown single crystals were fabricated both in-house and by Ceramare Corp. (Piscataway, NJ, USA) for subsequent characterization. Absorption spectra of optically polished crystals having parallel flats were obtained using a Perkin-Elmer Lambda 900 UV/vis/near-IR spectrometer. Spectra ranging from 350 to 800 nm were measured using a photomultiplier tube (PMT) detector and 850 1700 nm using a PbS detector at a scanning rate of 4 nm/s. Thermal conductivity (k) measurements were performed on a custom-designed system from 10 to 300 K using a steady-state technique. The system and measurement technique are detailed by Pope et al.21 Crystals approximately 2  2  5 mm3 in size having three sets of mutually perpendicular, optically polished parallel flats, were mounted in the removable sample apparatus using Dupont silver paste 4929N. Joule heating establishes a temperature gradient, ΔT, across the sample from power, P, supplied to a 120 Ω strain gauge resistor mounted on top of the sample. The slope of the P vs ΔT curve gives the thermal conductance, K (K = kA/L), where A is the sample area and L is the distance between the thermocouples from whence k is calculated. An undoped YAG crystal was used as an independent standard for comparison.

3. RESULTS AND DISCUSSION 3.1. Spontaneous Nucleation and Phase Stability. The hydrothermal phase stability of some related sesquioxide powders such as Sc2O3 and Y2O3 has been studied by Roy and Shafer in deionized water and served as a starting point for the present spontaneous nucleation study of Lu2O3 in 20 M KOH.22,23 Given the chemical similarity of Y2O3 and Lu2O3, we anticipated that hydrothermal solutions above 600 °C would be required to stabilize lutetia relative to hydrated phases such as LuO(OH) and Lu(OH)3. In fact at all temperatures from 550 to 600 °C we observed only the formation of large (3 8 mm) crystals of LuO(OH). From 600 to 620 °C a mixture of LuO(OH) and Lu2O3 (appearing as very small cubes) was obtained. Above 620 °C the cubic Lu2O3 crystals were the lone product, generally occurring in sizes of 0.5 1 mm3 with one crystal of ∼2 mm3 4387

dx.doi.org/10.1021/cg2005166 |Cryst. Growth Des. 2011, 11, 4386–4391

Crystal Growth & Design

ARTICLE

occurring out of numerous attempted spontaneous nucleation reactions. Some representative crystals are shown in Figure 1. An interesting feature of these cubic crystals is the opaque and irregular material deposited on the outer 50 100 μm of the crystals, physically resembling polycrystalline LuO(OH). However no bulk LuO(OH) crystals were observed in the reaction products, and the observed cubic morphology was that expected for Lu2O3. It appears that a layer of LuO(OH) is only deposited on the nucleated Lu2O3 crystals during the cool-down process as the system cools through the temperature regime where LuO(OH) is stable. Thus only the outermost surfaces of the crystals contain LuO(OH). This was verified by powder XRD in Figure 2 by comparing the powder pattern of the as-synthesized crystals and crystals from the same reaction that have had their outermost surfaces polished away to reveal a clear, colorless interior. The primary peaks in the pattern of the as-synthesized cubic crystals are accounted for by Lu2O3 (PDF 12-0728), while the series of additional lower intensity peaks can be

Figure 1. Lu2O3 crystals from hydrothermal spontaneous nucleation.

Figure 2. Powder XRD analysis of Lu2O3 spontaneous nucleation. (hkl) assignments for Lu2O3 are made on the top pattern, while (hkl) assignments for LuO(OH) are made on the bottom pattern. Positions of the most intense standard reflections for each phase are denoted below the experimental patterns.

attributed to the strongest indexed peaks of LuO(OH) (PDF 720982). The powder pattern of the clear, interior material can be accounted for solely by Lu2O3 (PDF 12-0728). 3.2. Seeded Transport Growth. Using the largest crystal pictured in Figure 1 (2.3  2.1  1.8 mm3), an initial proof-ofconcept mass transport experiment was performed duplicating the conditions producing the largest spontaneous nucleation (using 20 M KOH at 650 °C with a 45 °C temperature gradient) for 13 days. After transport, the seed crystal grew to a size of 4.5  4.4  3.2 mm3. The new growth was mostly transparent throughout, with the exception of ∼100 μm of polycrystalline LuO(OH) on the outermost surfaces. This crystal was cut into multiple pieces to be used as seed crystals in subsequent growth experiments designed to optimize growth rate and optical clarity as a function of temperature and temperature gradient and as seeds to grow bulk Er3+ and Yb3+-doped crystals (thus all lutetia grown in our research laboratory so far is ultimately traceable to that first 2 mm3 seed). The growth optimization studies were performed at 640 and 650 °C with thermal gradients of 20 55 °C to ensure that the growth zone of the autoclave was above 600 °C to maximize growth of the oxide over nucleation of LuO(OH) crystals. This study is summarized in Table 1, including an average growth rate (in all three dimensions) for all experiments performed under a given set of conditions and a qualitative assessment of optical clarity. In all cases the as-grown Lu2O3 crystals were colorless (or pink in the case of Er:Lu2O3) and required no postgrowth annealing treatment to remove color centers as is necessary for crystals grown using melt techniques in oxygen-poor atmospheres.8 Average growth rates of 0.55 1.1 mm/week were regularly observed in this study. For all thermal gradients, faster growth always occurred at 650 °C compared to 640 °C, presumably because of a greater overall solubility at higher temperatures (which is required for any growth to occur on the seed in this experimental design). Moderate thermal gradients of 30 45 °C produced the fastest observed growth for both dissolution zone temperatures studied. These conditions afforded a larger solubility gradient than narrower temperature gradients but still minimized competing spontaneous nucleation which quenched growth when wider gradients were used. In terms of optical clarity, a decided dependence on temperature gradient was observed. In all cases lutetia growth resulting from 20 to 30 °C gradients was fully transparent. Growth using 30 45 °C gradients had mixtures of opaque and transparent regions with opacity generally increasing with wider gradients. The wider gradients of these experiments also correspond to growth zone temperatures that are very close to the border of lutetia stability relative to

Table 1. Lutetia Growth Optimization Summary temp (°C)

temp gradient (°C)

av growth rate (mm/week)

apparent clarity of new growth, other comments

650

20

0.74

clear; some small Lu2O3 spontaneous nucleation (sn)

650

25

0.75

clear; some small Lu2O3 sn

650

30

1.03

most areas clear; some medium-size Lu2O3 sn

650

45

1.07

mostly clear, cloudy near seed; much Lu2O3 + LuO(OH) sn

650

50

0.60

cloudy; significant small Lu2O3 + LuO(OH) sn

650

55

0.65

mostly cloudy with small clear areas; much LuO(OH) sn

640

25

0.59

clear; some small Lu2O3 + LuO(OH) sn

640 640

30 35

0.75 0.70

mostly clear; some small Lu2O3 +LuO(OH) sn clear and cloudy areas; much Lu2O3 + LuO(OH) sn

640

40

0.74

mostly cloudy; much Lu2O3 + LuO(OH) sn 4388

dx.doi.org/10.1021/cg2005166 |Cryst. Growth Des. 2011, 11, 4386–4391

Crystal Growth & Design

ARTICLE

Figure 3. Growth progression of Er:Lu2O3 (top) and Yb:Lu2O3 (bottom) crystals: seed (left), as-grown (center), and polished (right).

LuO(OH), so higher temperature experiments are needed to determine whether the observed opacity truly results from a wider thermal gradient or if it is rather a function of proximity to the stability limit of lutetia. Regardless it appears in this system the growth zone temperature should be kept well above 600 °C. Thus we identified a dissolution zone temperature of 650 °C with a temperature gradient of 20 30 °C as optimum conditions for long-term bulk growth of optical quality Lu2O3, Er:Lu2O3, and Yb:Lu2O3. Bulk crystals from longer growth experiments generally grew toward a three-dimensional habit, although it was not always obviously cubic due to variations in the initial shape of the seed crystal and softening of the physical features by the LuO(OH) (or Er:LuO(OH) or Yb:LuO(OH)) deposited on the outside of the crystals during cool-down. Crystals with dimensions up to 7 10 mm were grown over periods of 4 6 weeks, depending on the initial size of the seed crystal. After polishing off the oxide hydroxide surface, the new oxide growth was cut away from the seed crystal and silver wire and then optically polished to yield clear regions 2 5 mm in size. Typical progressions from seed crystals to fabricated, polished pieces resulting from feedstock compositions of 3% Er:Lu2O3 and 11% Yb:Lu2O3 are shown in Figure 3. High doping levels were selected in these particular growth experiments to ensure a reasonable detection resolution by EDX and basic optical absorption experiments. Future growth experiments will be designed to produce measurable crystals having application-specific dopant concentrations. The pictured crystals were each grown at 650 °C with a temperature gradient of 20 °C at an average rate of 0.65 mm/week for Er:Lu2O3 and 0.73 mm/week for Yb:Lu2O3. Growth on the Er:Lu2O3 seed was also accompanied by some relatively large spontaneous nucleation crystals on the side of the autoclave liner that were also transparent once polished. 3.3. Optical and Thermal Properties Characterization. The average concentration of Er3+ in the resulting Er:Lu2O3 crystals was 3.2 ( 0.4 at. % measured by EDX. This doping resulted in a measured increase in unit cell parameter from 10.3980(12) Å in undoped hydrothermal lutetia to 10.4038(12) Å in the doped crystal, as determined by single-crystal XRD using a small spontaneously nucleated crystal that accompanied the growth on the seed. EDX of the resulting Yb:Lu2O3 crystals indicated an average Yb3+ concentration of 11 ( 1 at. %, but a suitable single crystal small enough for lattice parameter measurements by single-crystal XRD was not produced in this case. On the basis of the similarities of the dopant concentrations in the measured crystals to those in the feedstock used to grow them, there

Figure 4. Single-crystal absorption spectra of hydrothermally grown samples (a) Yb:Lu2O3, (b) Er:Lu2O3 (900 1700 nm), and (c) Er: Lu2O3 (350 850 nm).

appears to be minimal partition coefficient difference between the host and dopant ions in these hydrothermal systems and essentially any reasonable dopant concentration can be achieved. Doping was also verified by absorption spectroscopy shown in Figure 4. The spectra in Figure 4a,b are characteristic of the typical near-infrared transitions for Yb3+ and Er3+. The 2F7/2 to 2 F5/2 transition in Yb:Lu2O3 spans 900 1040 nm with the most intense absorption into the various Stark levels occurring at 976 nm. On the basis of this spectrum from the hydrothermal crystal, it appears that pumping at 976 nm will result in the subsequent emission at 1033 nm, which is consistent with meltgrown material (1034 nm).2 Er:Lu2O3 exhibits an absorption peak centered at 975 nm attributed to the 4I15/2 to 4I11/2 transition often used to pump the eye-safe 1.5 μm emission that subsequently results from the 4I13/2 to 4I15/2 transition. The corresponding absorption from the ground state to the 4I13/2 manifold is accounted for in this region of Figure 4b with the most intense absorption occurring at 1536 nm. The remaining transitions of Er3+ occurring in the visible region are resolved in Figure 4c. Detailed spectroscopic studies of these lanthanidedoped lutetia crystals are underway and will be reported elsewhere. The thermal conductivity of pure lutetia crystals was determined along with the thermal conductivities of the Er-doped lutetia and Yb-doped lutetia. As a reference point single crystals of YAG were also measured using the exact same technique, so all discussion can be made relative to YAG data. It is anticipated that thermal conductivity will always decrease upon doping relative to a pure host since any dopant will act as a lattice defect in the phonon transport mechanism. We postulate that the decrease for 4389

dx.doi.org/10.1021/cg2005166 |Cryst. Growth Des. 2011, 11, 4386–4391

Crystal Growth & Design

Figure 5. Thermal conductivity of undoped oxide single crystals.

Figure 6. Thermal conductivity of lanthanide-doped sesquioxide single crystals.

Lu2O3 will be significantly less upon doping than it would be for Sc2O3 or Y2O3. For example the similarity in mass between Lu and the dopants Er or Yb would decrease the phonon defects in the doped material. A comparison of undoped host crystals is made in Figure 5. In the case of undoped Sc2O3 and Lu2O3, both materials show a greater thermal conductivity than YAG at room temperature (ca. 10 W/mK for YAG versus about 12 and 15 W/mK for Lu2O3 and Sc2O3, respectively). All of the undoped compounds show a substantial increase in thermal conductivity as the temperature is lowered due to the loss of phonon phonon coupling as expected. We note that the shapes of the curves are similar to those of YAG reported previously by Aggarwal et al., but the absolute values of our YAG conductivities at low temperatures are somewhat lower than their values.24 We attribute this differential to the experimental difficulty in obtaining absolute values of thermal conductivity of our small samples at cryogenic temperatures. For the present we will use the curves as relative values to study the comparative behavior of the samples. We observe that the thermal conductivity of Sc2O3 and Lu2O3 increases dramatically to 77 K. It can be noted that the values for undoped Sc2O3 and Lu2O3 are respectively about 40 and 30% greater than undoped YAG at 77 K. The thermal conductivities of the doped single crystals are shown in Figure 6. The Er-doped materials show a significant drop off in thermal conductivity as expected, but 3% Er:Lu2O3 shows a significantly smaller drop off than identically doped Sc2O3 relative to their respective undoped crystals (Figure 5). Thus 3% Er-doped Lu2O3 shows a 25% decrease to a thermal conductivity of ∼9 W/mK at room

ARTICLE

temperature and 50% drop off to about 20 W/mK at 77 K, while 3% Er:Sc2O3 shows about a 60% drop off to ∼6 W/mK at room temperature and nearly a 75% drop off to ∼13 W/mK at 77 K. The effects of Er doping on the thermal conductivities of oxide hosts such as the sesquioxides or garnets are not as commonly discussed in the literature as Yb doping, but the magnitudes of the decreases in thermal conductivity observed in the present study seem reasonable compared to those observed in Yb-doped materials. As expected, Yb doping in Lu2O3 does not result in such a significant decrease in thermal conductivity as Er doping. We observed only about a 13% decrease in room-temperature thermal conductivity (to about 10.5 W/mK) for 11% Yb:Lu2O3. Other workers have similarly observed only small drops in the room-temperature thermal conductivity from Yb doping of lutetium-based oxide hosts, namely, about a 10% drop in 3% Yb:Lu2O3,11 and about a 4% drop in 10% Yb:LuAG compared to about a 35% drop in 10% Yb:YAG.25 Thus from a thermal management standpoint Lu2O3 single crystals appear to be very optimal hosts for Er- or Yb-doped lasers.

4. CONCLUSIONS Single crystals of Lu2O3, Er:Lu2O3, and Yb:Lu2O3 have been grown using hydrothermal methods for the first time. A seeded growth process using an aqueous KOH mineralizer was used to perform successive growth on one initial seed crystal in order to harvest new seed crystals and optimize transport conditions. Growth was performed at 600 650 °C, about 1800 °C below the melting point of lutetia, and growth rates of about 0.5 1 mm/ week were typically observed for both undoped and doped crystals. The resulting oxide crystals were always coated with a surface layer of LuO(OH) in their as-grown form. This layer was likely deposited on the bulk crystal as the autoclave was cooled through the oxide hydroxide stability region, and the layer could be easily polished away to reveal clear, bulk oxide material on the interior of the as-grown crystals. Characteristic absorption spectra of Er- and Yb-doped lutetia were obtained from these optical quality crystals. Improved thermal conductivities in the doped lutetia crystals compared to similarly doped scandia and YAG crystals make Lu2O3 a particularly attractive host for high-power laser applications. Lasing and other advanced spectroscopic studies of these promising hydrothermal Er:Lu2O3 and Yb: Lu2O3 single crystals are currently underway. ’ AUTHOR INFORMATION Corresponding Author

*E-mail: [email protected]. Tel.: (864) 656-4739. Fax: (864) 656-6613, ext. 485.

’ ACKNOWLEDGMENT The authors acknowledge financial support from the Joint Technology Office High Energy Laser Multidisciplinary Research Initiative (JTO HEL MRI) program through AFOSR Contract No. FA9550-07-1-0566, as well as Grant No. NSFDMR-0907395. ’ REFERENCES (1) Petermann, K.; Fagundes-Peters, D.; Johansen, J.; Mond, M.; Peters, V.; Romero, J. J.; Hutovoi, S.; Speiser, J.; Giesen, A. J. Cryst. Growth 2005, 275, 135–140. 4390

dx.doi.org/10.1021/cg2005166 |Cryst. Growth Des. 2011, 11, 4386–4391

Crystal Growth & Design

ARTICLE

(2) Petermann, K.; Huber, G.; Fornasiero, L.; Kuch, S.; Mix, E.; Peters, V.; Basun, S. A. J. Lumin. 2000, 87 89, 973–975. (3) Fornasiero, L.; Mix, E.; Peters, V.; Petermann, K.; Huber, G. Cryst. Res. Technol. 1999, 34, 255–260. (4) Griebner, U.; Petrov, V.; Petermann, K.; Peters, V. Opt. Express 2004, 12, 3125–3130. (5) Peters, R.; Kr€ankel, C.; Petermann, K.; Huber, G. Opt. Express 2007, 15, 7075–7082. (6) Brown, D. C. IEEE J. Sel. Top. Quant. Electron. 2005, 11, 587–599. (7) Fan, T. Y.; Ripin, D. J.; Aggarwal, R. L.; Ochoa, J. R.; Chann, B.; Tilleman, M.; Spitzberg, J. IEEE J. Sel. Top. Quant. Electron. 2007, 13, 448–459. (8) Petermann, K.; Fornasiero, L.; Mix, E.; Peters, V. Opt. Mater. 2002, 19, 67–71. (9) Laversenne, L.; Guyot, Y.; Goutaudier, C.; Cohen-Adad, M.Th.; Boulon, G. Opt. Mater. 2001, 16, 475–483. (10) Fornasiero, L.; Mix, E.; Peters, V.; Petermann, K.; Huber, G. Ceram. Int. 2000, 26, 589–592. (11) Peters, V.; Bolz, A.; Petermann, K.; Huber, G. J. Cryst. Growth 2002, 237 239, 879–883. (12) Peters, R.; Kr€ankel, C.; Petermann, K.; Huber, G. J. Cryst. Growth 2008, 310, 1934–1938. (13) Marchese, S. V.; Baer, C. R. E.; Peters, R.; Kr€ankel, C.; Engqvist, A. G.; Golling, M.; Maas, D. J. H. C.; Petermann, K.; S€udmeyer, T.; Huber, G.; Keller, U. Opt. Express 2007, 15, 16966–16971. (14) Baer, C. R. E.; Kr€ankel, C.; Saraceno, C. J.; Heckl, O. H.; Golling, M.; S€udmeyer, T.; Peters, R.; Petermann, K.; Huber, G.; Keller, U. Opt. Lett. 2009, 34, 2823–2825. (15) Baer, C. R. E.; Kr€ankel, C.; Saraceno, C. J.; Heckl, O. H.; Golling, M.; Peters, R.; Petermann, K.; S€udmeyer, T.; Huber, G.; Keller, U. Opt. Lett. 2010, 35, 2302–2304. (16) McMillen, C. D.; Kolis, J. W. J. Cryst. Growth 2008, 310, 1939–1942. (17) Mann, M.; Thompson, D.; Serivalsatit, K.; Tritt, T. M.; Ballato, J.; Kolis, J. W. Cryst. Growth Des. 2010, 10, 2146–2151. (18) Mann, M.; Kolis, J. W. J. Cryst. Growth 2010, 312, 461–465. (19) Mann, M.; Jackson, S.; Kolis, J. W. J. Solid State Chem. 2010, 183, 2675–2680. (20) Ballato, J.; McMillen, C.; Kokuoz, B.; Serivalsatit, K.; Kokuoz, B.; Kolis, J. Proc. SPIE 2008, 6871, 68711G/1–68711G/6. (21) Pope, A. L.; Zawilski, B.; Tritt, T. M. Cryogenics 2001, 41, 725–731. (22) Shafer, M. W.; Roy, R. Z. Anorg. Allg. Chem. 1954, 276, 275–288. (23) Shafer, M. W.; Roy, R. J. Am. Ceram. Soc. 1959, 42, 563–570. (24) Aggarwal, R. L.; Ripin, D. J.; Ochoa, R. J.; Fan, T. Y. J. Appl. Phys. 2005, 98, 103514-1–103514-14. (25) Beil, K.; Fredrich-Thornton, S. T.; Tellkamp, F.; Peters, R.; Kr€ankel, C.; Petermann, K.; Huber, G. Opt. Express 2010, 18, 20712–20722.

4391

dx.doi.org/10.1021/cg2005166 |Cryst. Growth Des. 2011, 11, 4386–4391