Open Access
Add to BibSonomy
Abstract
We report on the first laser operation of ytterbium-doped potassium yttrium double molybdate crystal (Yb:KY(MoO4)2). Single-crystals containing 3 at.% Yb3+-ions were grown by the low temperature gradient Czochralski method. The crystal structure (orthorhombic, sp. gr. D142h – Pbna) was refined with the Rietveld method. Yb:KY(MoO4)2 exhibits a layered structure leading to a strong optical anisotropy and a perfect cleavage along the crystallographic (100) plane. The maximum stimulated-emission cross-section amounts to 3.70×10−20 cm2 at 1008.0 nm with an emission bandwidth of 37 nm (for light polarization E || b). The Stark splitting is determined at 6 K. Continuous-wave laser operation is achieved in a thin Yb:KY(MoO4)2 crystal plate (thickness: 286 µm) under diode pumping. The microchip laser generated a maximum output power of 0.81 W at 1021-1044 nm with a slope efficiency of 76.4% and polarized emission. Yb:KY(MoO4)2 crystal lamellae / plates are attractive for sub-ns passively Q-switched microchip lasers and thin-disk lasers.
© 2020 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Cleavage is the property of crystalline materials to split along certain crystallographic planes. The simplest case is basal or pinacoidal cleavage when this happens along only a single plane. It may result from a weaker bond strength or larger lattice spacing perpendicular to the plane, and it is frequently observed in crystals with a regular location of atoms forming “layers”. As a result, crystalline plates with continuous, smooth and flat faces, both strictly parallel to the so-called cleavage plane are easily obtained. An example of a cleaving mineral is mica showing perfect (or even “eminent”) cleavage. Crystals with perfect cleavage can cleave without leaving any rough surface (mirror-quality) [1]. Cleavage is used for identification of minerals. It is also useful for chipping of wafers of semiconductor crystals (e.g., silicon).
Laser crystals exhibiting perfect cleavage along a crystallographic plane are known [1–3]. On the one hand, this complicates their mechanical processing (cutting and polishing). On the other hand, it can turn out to be a useful feature for applications where thin laser elements with a large aperture and good parallelism / optical quality of the faces are required. Moreover, this simplifies fabrication of laser elements, especially for soft crystals. Two examples are the microchip and thin-disk lasers. In particular, a microchip laser consists of a thin (typically, few hundred µm thick) gain medium placed in a simple plano-plano cavity without air gaps, leading to a highly-compact and monolithic design (optionally, one or both of the cavity mirrors are coated on the crystal faces) [4]. Note a positive thermal lens of the gain material for stabilizing the laser mode of microchip lasers is an important requirement [5]. This robust design is almost insensitive to misalignment and leads to low intracavity losses, low laser threshold, good cavity stability and very short cavity roundtrip time beneficial for achieving sub-ns pulses in passively Q-switched lasers [6].
So far, only few crystals doped with rare-earth ions (RE3+) such as Nd3+, Yb3+ or Tm3+ showing perfect cleavage have been applied for lasers in the near-IR, cf. Table 1. These include borates (LaB3O6) [7], phosphates (YPO4 and LuPO4) [8,9] and molybdates (BaGd2(MoO4)4) [10]. Other cleaving crystals were proposed as potential laser materials (e.g., CsGd(MoO4)2) [11] but no laser operation has been reported so far. Yb3+-doped materials are known for efficient laser operation at ∼1 µm on the 2F5/2 → 2F7/2 electronic transition. The Yb3+ ion can be easily pumped by commercially available high-power InGaAs laser diodes emitting at ∼0.96-0.98 µm [12]. It also exhibits a simple energy level scheme eliminating parasitic energy-transfer processes. In-band pumping of Yb3+ leads to high pump Stokes efficiency, and, thus, reduced heat load and slope efficiencies up to ∼80% [13]. Moreover, as compared to Nd3+, Yb3+ shows larger Stark splitting of the ground-state (2F7/2) and, consequently, broader emission at ∼1 µm, beneficial for broadly tunable and mode-locked lasers. Liu et al. reported on a diode-pumped Yb:LuPO4 laser based on a (100)-oriented unprocessed as-grown crystal plate delivering 1.61 W at 1036-1040 nm with a slope efficiency of 75% [8]. Zhu et al. used cleaved plates of Yb:BaGd2(MoO4)4 and produced a similar output power (1.16 W at 1046-1054 nm) albeit with much lower slope efficiency (20%) [1].
Table 1. Laser Performancea of RE3+-doped Cleaving Crystals
Among the host crystals for Yb3+ doping, complex tungstates and molybdates are attracting a lot of attention. Their advantages include: (i) high attainable Yb3+ doping concentrations; (ii) weak luminescence quenching, (iii) low non-radiative relaxations and luminescence quantum yields approaching unity, (iv) polarized, intense and broad absorption and emission bands at 1 µm, and (v) Raman activity. A prominent example is the crystal family of monoclinic double tungstates with chemical formula KLn(WO4)2, where Ln = Gd, Y or Lu [15,16]. Efficient continuous-wave [12,17], passively Q-switched [6] and especially mode-locked [18,19] Yb:KLn(WO4)2 lasers are known. Thin-disk lasers based on these materials were also demonstrated using thin bulk or epitaxial samples [20,21].
Compared to the above mentioned potassium double tungstates, their double molybdate (DMo) counterparts with chemical formula KLn(MoO4)2 have been barely studied. One example of the DMo crystal family is potassium yttrium double molybdate, KY(MoO4)2 [22–24]. It is orthorhombic and exhibits an interesting layered structure together with a low-symmetry site for the RE ions (C2) leading to a strong anisotropy of the optical properties as well as to the natural cleavage habit [22]. The structure and vibronic properties of undoped KY(MoO4)2 have been reported [25,26]. The RE3+ site symmetry was revealed using Eu3+ as a structural probe [22]. Very recently, we achieved laser operation in cleaved single-crystalline plates and thin films of Tm:KY(MoO4)2 [3]. A crystal-plate laser generated 0.88 W at 1840-1905nm with a slope efficiency of 65.8%. The concept of the thin-film laser using cleaved Nd:KY(MoO4)2 was first proposed [27] by Kaminski et al.
In the present work, we demonstrate laser operation of a Yb:KY(MoO4)2 crystal, for the first time to the best of our knowledge, by using its perfect cleavage feature.
2. Crystal growth
The Yb:KY(MoO4)2 compound melts at ∼1243 K. The single crystals were grown by the Low Temperature Gradient (LTG) Czochralski method [28]. As raw materials, we used Y2O3 (purity: 5N), Yb2O3 (4N), MoO3 (4N) and K2CO3 (5N) taken according to the composition 95-93mol% KY0.97Yb0.03(MoO4)2 solute – 5-7 mol% K2Mo3O10 solvent assuming substitution of Y3+ ions by the Yb3+ dopants (3 at.% Yb). The potassium trimolybdate (K2Mo3O10) was added to the melt to prevent its partial dissociation resulting in the formation of yttrium oxomolybdate (Y2MoO6) and, thus, to stabilize the growth process [22]. The raw materials were mixed and placed in a Pt crucible. It was heated up to ∼1320 K in air and kept at this temperature for 2–3 hours to homogenize the melt. Then, it was cooled to ∼1240 K (the temperature where the growth started). A [100]-oriented seed from an undoped KY(MoO4)2 was used. It was rotated at 20 rpm, the pulling rate was 1–2 mm/h and the cooling rate was ∼2 K/day. The temperature gradient in the melt was below 3 K/cm (in the vertical direction). After completing the growth, the crystal was removed from the melt and slowly cooled down to room temperature (RT, 293 K). No annealing was applied. More details can be found elsewhere [3,22].
The as-grown crystals were transparent and colorless, Fig. 1. They had a cylindrical shape with an elliptic cross-section with the semiaxes oriented along the [010] and [001] directions. Neither cracks nor inclusions were observed. The crystals showed an easy cleavage along the (100) plane (orthogonal to the growth direction). The orientation of the [010] and [001] axes was determined by single-crystal X-ray diffraction.
Fig. 1. A photograph of as-grown 3 at.% Yb:KY(MoO4)2 crystals. The growth direction is along the [100] crystallographic axis.
The actual Yb3+ doping concentration was determined by Energy Dispersive X-ray (EDX) spectroscopy to be NYb = 2.3 ± 0.5×1020 cm-3 (∼3.0 at.% Yb), so that the segregation coefficient for Yb3+ doping KYb = Ncrystal/Nmelt was close to unity.
3. Crystal structure
3.1 Rietveld refinement
The structure and phase purity of the 3 at.% Yb:KY(MoO4)2 was confirmed by X-ray powder diffraction (XRD), Fig. 2(a). The measurements were carried out in a θ-θ Bragg Brentano configuration using a Siemens D-5000 powder X-ray diffractometer with Cu Kα (1.5406 Å) radiation. The XRD pattern was recorded in a 2θ range from 10° to 65°, a step size of 0.02° and a step time of 16 s. The crystal belongs to the orthorhombic class (sp. gr. Pbna – D142h, centrosymmetric point group 2/m). Note that we use the non-conventional space group (the standard one is Pbcn, No. 60), following the early publications [25,29]. The unit cell parameters refined using the Le Bail method are: a = 18.212(2) Å, b = 7.9343(6) Å, c = 5.0705(5) Å (number of the formula units in the unit-cell Z = 4). The obtained R-factors were Rwp = 6.44% and Rexp = 5.19% (the reduced χ-squared value χ2 = (Rwp/Rexp)2 = 1.54). The calculated volume of the unit cell V is 732.7(1) Å3 and the crystal density ρcalc = 4.083(6) g/cm3. The structure from [22] was taken as the starting one for the refinement. The fractional atomic coordinates were refined by the Rietveld method using the TOPAS software, see the results in Table 2, with the fixed unit cell parameters obtained by the Le Bail’s method and considering the (100) preferred orientation. For this refinement, Rwp = 8.90% and Rexp = 5.19% (χ2 = 2.94). Figure 2 shows the Rietveld refinement plot with the observed, calculated and difference patterns. No other phases except the orthorhombic one are found in the pattern. The min / max residual electron densities are -0.6 e/Å3 and +0.1 e/Å3, respectively.
Fig. 2. (a) Rietveld analysis of the RT X-ray powder diffraction (XRD) pattern of a 3 at.% Yb:KY(MoO4)2 crystal, the numbers denote the Miller’s indices (hkl) for the sp. gr. Pbna; (b,c) Illustration of the layered structure of the Yb:KY(MoO4)2 crystal: (b) three-dimensional view, (c) a projection of the a-c plane. Black rectangle denotes the unit-cell.
Table 2. Fractional Atomic Coordinates (x, y, z), Occupancy Factors (O.F.) and Temperature Factors (Biso) for 3 at.% Yb:KY(MoO4)2 (Space Group Pbna)
The unit-cell parameters are slightly smaller than those for undoped KY(MoO4)2, a = 18.23 Å, b = 7.95 Å, c = 5.07 Å [25]. In KY(MoO4)2, the Yb3+ ions substitute for the Y3+ ones in a single type of crystallographic sites (Wyckoff symbol: 4c, site symmetry: C2, coordination number (C.N.): VIII). The corresponding ionic radii are RYb = 0.985 Å and RY = 1.019 Å for VIII-fold oxygen coordination [30], explaining the observed decrease of the lattice constants. The closeness of the ionic radii of Yb3+ and Y3+, the existence of the orthorhombic stoichiometric KYb(MoO4)2 phase [31] and the homovalent doping mechanism also explain the observed KYb ≈ 1.
Using the refined atomic coordinates, the structure of Yb:KY(MoO4)2 is illustrated in Fig. 2(b,c). The corresponding interatomic distances are listed in Table 3. The C.N. for Mo6+ cations is 4 + 1: a distorted tetrahedral coordination (with 4 closely located oxygens, with the Mo – O distances lying in the range 1.644(8) – 1.847(6) Å and 1 more distant oxygen at 2.5735 Å). For K+, the C.N. is 6 + 4, defined in a similar manner. The closest 6 oxygens are at 2.742(1) - 2.759(6) Å, while the other oxygens are at 3.335(2) – 4.223(0) Å. Such a coordination behavior is common for tungstates and molybdates with a C.N. of W (or Mo) of V or VI, where the cation – anion distances may vary in a broad range, so that a formal definition of the C.N. is not possible [25]. For the distorted [Y|YbO8] polyhedra (bicapped octahedra), the bond lengths are in the range 2.26(2) - 2.74(3) Å. The belts of edge-sharing [Y|YbO8] polyhedra run along the b-axis. Along the c-axis, they are separated by an empty polyhedron which is sharing 4 edges with four [MoO4] tetrahedra. Each Mo-tetrahedron connects two translationally identical belts of Y-polyhedra and, simultaneously, it connects the corners of two adjacent Y-polyhedra (within the belt). Thus, the multi-layer structure of Yb:KY(MoO4)2 is determined by [Y|Yb(MoO4)2] layers (lying in the b-c plane) formed by [MoO4] tetrahedra and [Y|YbO8] polyhedra and containing cavities, as well as zigzag K+-layers separating two neighboring [Y|Yb(MoO4)2] layers. The zigzag K+-layers are more loose than the [Y|Yb(MoO4)2] ones. Moreover, the latter are separated by a relatively large a/2 spacing of 9.106 Å. This determines the natural cleavage along the (100) plane.
Table 3. Selected Interatomic Distances in 3 at.% Yb:KY(MoO4)2
The shortest Y|Yb – Y|Yb distance is 3.973 Å (along the b-axis, in the layer plane). It is relatively long and similar to that (Y – Y) in the monoclinic KY(WO4)2 crystal (4.06 Å) [22]. This distance determines the weak cross-talk of Yb3+ ions (weak concentration quenching of luminescence) along the a-axis (orthogonal to the layer plane), 9.45 Å.
The coefficients of thermal expansion (CTE) of Yb:KY(MoO4)2 were calculated from the temperature dependence of the lattice constants determined by high-temperature powder XRD. For the XRD studies in the temperature range of 300–550 K, we used a temperature chamber (HTK10). The heating rate was 0.17 K/s with a delay of 300 s before each measurement. The 2θ angle varied from 10° to 70° with a step size of 0.03° and a step time of 5 s. The relative evolution of the unit cell parameters with temperature is shown in Fig. 3. The linear thermal expansion tensor (αij) in abc frame was obtained from the slopes of the linear fits shown in Fig. 3. Thus, (αij) is:
Fig. 3. Relative thermal evolution of the unit cell parameters of 3 at.% Yb:KY(MoO4)2 with the temperature up to 550 K.
The anisotropy of the thermal expansion in Yb:KY(MoO4)2 is relatively strong, as expressed by the ratios αa : αb = 1.64 and αa : αc = 7.15. The thermal expansion in KY(MoO4)2 is stronger than in its double tungstate counterpart. Indeed, for KY(WO4)2, αa = 8.4, αb = 2.0 and αc* = 19.8 [10−6 K-1] [32].
The thermal conductivity of KY(MoO4)2 is unknown but it is expected to be higher in the layer plane.
By mechanical cleavage orthogonal to the crystal growth direction (along the (100) plane), we produced single-crystalline plates of Yb:KY(MoO4)2 with a thickness (t) as thin as 100 µm. They were studied by Scanning Electron Microscopy (SEM) using a MERLIN microscope (Carl Zeiss). Under bending, the plates exhibited an elastic deformation along the [010] axis and broke with the fracture edge running parallel to the [001] axis, Fig. 4(a). This SEM image shows multiple “steps” parallel to the (100) cleavage plane. Both surfaces of the cleaved plates had mirror-quality. No polishing was applied to them. The SEM study in the µm-scale revealed dark flower-like defects, Fig. 4(b); however, the concentration of these defects was low. The clean aperture of the obtained crystal-plates was >1 cm2. They contained no macroscopic cracks.
Fig. 4. Scanning Electron Microscope (SEM) images of a cleaved single-crystal plate of 3 at.% Yb:KY(MoO4)2: (a) the fracture edge of the plate running along the c-axis, the a-axis is vertical; (b) the plate surface parallel to the (100) plane showing flower-like defects.
3.2 Raman spectra
Molybdate crystals are known as efficient Raman-active materials. Thus, Yb:KY(MoO4)2 can serve for self-frequency Raman conversion. The RT polarized Raman spectra were measured using a confocal Raman microscope (Renishaw inVia) equipped with an ×50 objective, an edge filter and an Ar+ ion laser (488 nm). A cleaved crystal plate (a-cut) was used and the excitation / collection geometries were a(mn)a, where m, n = b, c (according to Porto’s notations).
The Raman spectra are shown in Fig. 5. They are strongly polarized with the most intense Raman response in the a(bb)a geometry. The observed bands are classified into three groups of vibrations [22,26]. The low-frequency range, 80-272 cm-1, contains translational (T’) and rotational (R’) modes of the K, Y|Yb and Mo cations. Internal bending vibrations (δ) of the oxygen bridged [MoO4] tetrahedra are observed in the intermediate frequency range, 315–435 cm-1. The high-frequency range, 726-944 cm-1, contains intense stretching vibrations (ν) of these tetrahedra. The gap in the Raman spectra (500-700 cm-1) is due to the relatively weak oxygen linkage of the [MoO4] tetrahedra in the layer plane. This feature is similar to the one found in scheelite (CaWO4) type double tungstate and molybdate crystals with isolated [WO4] tetrahedra [33].
Fig. 5. RT Raman spectra of the 3 at.% Yb:KY(MoO4)2 crystal for the a(mn)a, m, n = b, c geometries (Porto’s notations), λexc = 488 nm, numbers indicate the peak frequencies in cm-1.
The most intense Raman mode is at 865 cm-1 with a width at half maximum (FWHM) of 18.8 cm-1. The maximum phonon energy hνmax is 944 cm-1.
4. Optical spectroscopy
4.1 Optical absorption
Orthorhombic KY(MoO4)2 is an optically biaxial crystal. Only a mean value of its refractive index is known, ≈ 1.95. KY(MoO4)2 shows a strong birefringence, Δn = 0.018–0.087 [34]. The optical indicatrix axes coincide with the crystallographic axes. We will denote the principal light polarizations as E || a, b and c.
The absorption spectra were measured using a Varian CARY-5000 spectrophotometer (0.3-2 µm) and a FTIR spectrometer Bruker Tensor 27 (2-7 µm). A Glan-Taylor prism was used for polarization-resolved studies.
The transmission spectrum of a ∼1 cm-thick 3 at.% Yb:KY(MoO4)2 crystal sample is shown in Fig. 6(a) revealing a transparency range from 0.33 to 3.4 µm (extending up to 5 µm to some extent). The structured absorption at longer wavelengths is related to the ν vibrations of the MoO4 tetrahedra. The infrared cut-off is due to the 2ν1 overtone peak of the [MoO4]2- group fundamental vibration.
Fig. 6. RT absorption properties of a 3 at.% Yb:KY(MoO4)2 crystal: (a) unpolarized transmission spectrum of a ∼1 cm-thick (100)-oriented crystal plate; (b) absorption cross-section, σabs, spectra for light polarizations E || a, b, c. The arrow indicates the pump wavelength used in the laser experiments; (c) Tauc plot for the evaluation of the optical bandgap (Eg), Eph – photon energy.
The absorption at ∼1 µm originates from the 2F7/2 → 2F5/2 transition of Yb3+ ions. It is analyzed in Fig. 6(b) in terms of absorption cross-sections, σabs = αabs/NYb (αabs - absorption coefficient) for the principal light polarizations E || a, b, c. The absorption spectra exhibit strong polarization anisotropy. The maximum σabs is 1.77×10−20 cm2 at 977.1 nm and the FWHM of the absorption peak is 19.6 nm for light polarization E || b. This wavelength corresponds to the zero-phonon line (ZPL, see below) transition at RT. For the other two light polarizations, the absorption cross-sections are lower, as expressed by the ratios σabs(b) : σabs(c) = 1.26 and σabs(b) : σabs(a) = 4.2 at ∼980 nm. This anisotropy arises, in part, from the layered crystal structure (note the significant drop of the optical absorption for light polarized orthogonal to the b-c layer plane) and, in part, from the low symmetry of the Yb3+ site. The observed broad absorption spectra qualify Yb:KY(MoO4)2 for diode-pumping by InGaAs laser diodes emitting at ∼980 nm.
The absorption cross-sections in Yb:KY(MoO4)2 are lower than those for the monoclinic Yb:KY(WO4)2 crystal: σabs = 10.8×10−20 cm2 at 981.0 nm corresponding a narrower FWHM of 4.0 nm (for light polarization E || Nm) [35].
The optical bandgap of Yb:KY(MoO4)2 was evaluated with the Tauc plot, i.e, by plotting (αabs×Eph)2 vs. the photon energy Eph = h(c/λ). The intersection of the linear fit of the obtained plot with the horizontal axis yields Eg = 3.72 eV (the wavelength of the UV absorption edge is ∼330 nm).
4.2 Luminescence: spectra and lifetime
The polarized luminescence spectra were measured using an optical spectrum analyzer (OSA, Hamamatsu, AQ6373) and a Glan-Taylor polarizer. The stimulated-emission (SE) cross-sections, σSE, for i-th polarization (i = a, b, c) were calculated from the measured luminescence spectra calibrated for the spectral response of the set-up W'(λ) using the Füchtbauer–Ladenburg (F-L) formula [36]:
For Yb:KY(MoO4)2, the maximum σSE amounts to 3.70×10−20 cm2 at 1008.0 nm and the emission bandwidth (FWHM) Δλem is 37.0 nm for the high-gain light polarization E || b. The emission spectra are also strongly polarized, as expressed by the ratios σSE(b) : σSE(c) = 3.9 and σSE(b) : σSE(a) = 6.0 at ∼1.01 µm. These high ratios are a prerequisite for linearly polarized laser emission. Note that (100)-oriented crystal plates give access to the preferable light polarization E || b. Compared to its monoclinic double tungstate counterpart, Yb:KY(WO4)2, for which σSE = 3.2×10−20 cm2 at 1021.9 nm with Δλem = 30.2 nm for E || Nm [35], Yb:KY(MoO4)2 provides higher SE cross-sections and broader emission bandwidth. These are attractive features for broadly tunable and ultrashort-pulse lasers at ∼1 µm.
Fig. 7. Emission properties of Yb3+:KY(MoO4)2: (a) stimulated-emission (SE) cross-sections, σSE, for light polarizations E || a, b and c. The arrow indicates the observed laser wavelength; (b) gain cross-sections, σgain = σSE – (1 – β)σabs, β = N2(2F5/2)/NYb is the inversion ratio, the light polarization is E || b; (c) luminescence decay curve for a cleaved film, λexc = 930 nm, λlum = 1030 nm, symbols: experimental data, line: single-exponential fit.
For the Yb3+ ion, exhibiting reabsorption at the laser wavelength, the spectral behavior of the laser depends on the output-coupling losses via the rate of inversion in the gain medium, expressed by the inversion ratio β = N2(2F5/2)/NYb, where N2 is the population of the upper level (2F5/2) and N1 + N2 = NYb. Thus, the gain cross-sections, σgain = σSE – (1 – β)σabs, are calculated to predict the possible laser wavelengths for different β. The gain spectra for light polarization E || b are shown in Fig. 7(b). For small β < 0.03, the gain spectra are very flat and broad extending from ∼1040 nm to 1100 nm. For intermediate inversion ratios 0.05 < β < 0.10, a local peak in the spectra appears at ∼1039 nm. For even higher inversion ratios, another maximum at a shorter wavelength of ∼1019 nm is observed. For β = 0.15, the gain bandwidth Δλg is 33.0 nm in agreement with the luminescence studies.
The luminescence decay curve of the sample was measured using an InGaAs photodetector under ns pulse excitation from an optical parametric oscillator (Horizon, Continuum) tuned to 930 nm, a 1/4 m monochromator (Oriel 77200) and an 8 GHz oscilloscope (DSA70804B, Tektronix). A thin (∼30 µm) cleaved film was used to avoid the effect of radiation trapping. The decay curve plotted in a semi-log scale, Fig. 7(c), is single-exponential in agreement with a single type of Yb3+ sites. The luminescence decay time τlum = 458 µs is longer than that of monoclinic Yb:KY(WO4)2 crystals (τlum = 231 µs [35]).
4.3 Crystal-field splitting
For Yb3+ ions in C2 sites, each 2S+1LJ multiplet is split into J + 1/2 Stark sub-levels which are numbered as 0…3 for the ground-state (2F7/2) and 0’…2’ for the excited-state (2F5/2). To resolve their energies, we measured the unpolarized absorption and luminescence spectra at low temperature (LT, 6 K) using an Oxford Instruments Ltd. cryostat (model SU 12) with helium-gas close-cycle flow. The results are shown in Fig. 8. The LT spectra were interpreted accounting for the Raman spectra in order to assign the possible vibronic sidebands.
Fig. 8. Absorption and luminescence spectra of Yb3+:KY(MoO4)2 plotted vs. the energy difference (Eph – EZPL), where Eph = h(c/λ) is the photon energy and EZPL = 10246 cm-1 is the zero-phonon line (ZPL) energy, measured at LT (6 K) and RT (293 K) with unpolarized light. For LT emission spectra, λexc = 976 nm. Electronic transitions are indicated as i ↔ j’, phonon sidebands – as hν.
The zero-phonon line (ZPL) is the transition between the lowest Stark sub-levels of both multiplets, designated as 0 ↔ 0’. It is observed at 10246 cm-1 corresponding to the wavelength of 976.0 nm (at 6 K). The full set of energy-levels of Yb3+:KY(MoO4)2 is 2F7/2 = (0, 315, 411, 601) cm-1 and 2F5/2 = (10246, 10440, 10749) cm-1, see Fig. 9(a), showing the energy-level scheme and the wavelengths of the electronic transitions in absorption and emission. In particular, the emission peak at ∼1007 nm corresponding to the maximum SE cross-sections is assigned to the 0’ → 1 transition. The partition functions for the lower (m = 1) and upper (m = 2) manifolds are Z1 = 1.397 and Z2 = 1.469 (the ratio Z1/Z2 = 0.951). The determined crystal-field splitting is different from that in monoclinic Yb:KY(WO4)2 for which 2F7/2 = (0, 169, 407, 568) cm-1 and 2F5/2 = (10187, 10476, 10695) cm-1 [16]. In particular, the total Stark splitting of the ground-state, ΔE(2F7/2), is larger for Yb:KY(MoO4)2 (601 cm-1), which explains the broadband emission properties of this material.
Fig. 9. (a) Crystal-field splitting of the Yb3+ ion in the KY(MoO4)2 crystal (site symmetry: C2), Z1(2) are the partition functions for the ground- and excited-state, respectively, arrows indicate the transitions in absorption / emission at 6 K; (b) barycenter plot [37] for the Yb3+ ion showing the position of Yb:KY(WO4)2 and Yb:KY(MoO4)2 crystals (blue circles).
Previously, for the isostructural KYb(MoO4)2, it was pointed out that even at LT, the spectra of the Yb3+ ion revealed a strong electron-phonon coupling [31]. Thus, the assignment of the Stark levels was complicated. The following splitting for the ground-state (2F7/2) was proposed: (0, 240, 460, 460) cm-1. This poorly agrees with our data.
For all RE ions, the barycenter energy of any isolated 2S+1LJ 4fn multiplet depends linearly on the barycenter energy of any other isolated multiplet. This relation is expressed by the barycenter plot [37], Fig. 9(b). The determined barycenter energies < E(2F5/2)> and < E(2F7/2)> agree well with the linear fit to this plot, expressed by the equation E(2F5/2) = 10166.6 + 0.997×E(2F7/2) cm-1, where E0 = 10166.6 cm-1 denotes the energy of the Yb3+ excited-state for a free-ion. This confirms the correctness of the constructed energy-level scheme.
5. Laser operation
5.1 Laser set-up
The active elements for the laser experiments were fabricated by mechanical cleavage of the 3 at.% bulk Yb:KY(MoO4)2 crystal along the (100) plane. In this way, we produced a thin cleaved crystal-plate with a thickness t = 286 µm and a high uniformity of ±2 µm over the clear aperture. No post-cleavage treatment (e.g., polishing) was applied to both surfaces which remained uncoated. The crystal orientation corresponded to light propagation along the a-axis (a-cut) which is beneficial for the pump absorption providing access to light polarizations E || b and E || c with higher σabs values, Fig. 6(b).
For laser experiments, we selected a plano-plano (microchip) laser cavity, Fig. 10, formed by a flat pump mirror (PM) coated for high transmission (HT) at ∼0.97 µm and for high reflection (HR) at 1.02–1.2 µm, and a set of flat output couplers (OCs) with a transmission at the laser wavelength TOC of 0.5%-10%. Both cavity mirrors were gently pressed towards the crystal-plate resulting in a nearly monolithic design. The geometrical cavity length Lcav nearly equaled the crystal-plate thickness t. The whole stack (cavity mirrors and crystal-plate) was passively cooled.
Fig. 10. Scheme of the diode-pumped 3 at.% Yb:KY(MoO4)2 cleaved crystal-plate microchip laser: LD – laser diode, PM – pump mirror, OC – output coupler.
As a pump source, we employed an InGaAs fiber coupled laser diode (fiber core diameter: 105 µm; numerical aperture (N.A.): 0.22) emitting unpolarized output at a central wavelength of ∼968 nm. The pump was collimated and focused into the crystal through the PM using a lens assembly (reimaging ratio: 1:1, focal length: f = 30 mm). The pump spot diameter 2wP in the focus was 100 ± 10 µm. The crystal-plate was pumped in a double-pass, as all the used OCs provided reflection at the pump wavelength, R ≈ 90%. The total pump absorption in two passes was estimated from the small-signal value accounting for the Fresnel losses at the uncoated crystal surfaces and amounted to 9.4%. In one pass, ηabs,0 = 1 – exp(–<σabs>NYbt) = 4.8%, where <σabs> = 0.75×10−20 cm2 is the polarization-averaged absorption cross-section. This value does not account for the ground-state bleaching but at the same time it does not overestimate the laser slope efficiency. Due to the small leakage of the residual pump through the OC, it was difficult to estimate the pump absorption under lasing conditions.
A long-pass filter (FEL1000, Thorlabs) separated the laser output and the residual pump. The laser emission spectra were measured with an accuracy of ±0.2 nm.
5.2 Laser performance
Laser operation was achieved for all the employed OCs. The CW input-output dependences of the diode-pumped microchip cleaved crystal-plate Yb:KY(MoO4)2 laser are shown in Fig. 11(a). For the 1% OC, the laser generated a maximum output power of 0.81 W at 1021-1044 nm with a high slope efficiency (η) of 76.4% vs. the absorbed pump power Pabs. The laser threshold Pth was as low as 56 mW whilst the optical-to-optical efficiency (vs. the pump power incident on the crystal) ηopt was rather low, 7.0%, owing to the low pump absorption in the thin crystal. With increasing the output coupling, the laser threshold gradually increased, from 55 mW for TOC = 0.5% to 143 mW for TOC = 10%. The input-output dependences were linear for all the OCs. The absorbed pump power in the crystal-plate Pabs was limited to below 1.1 W in order to avoid thermal fracture of the passively cooled crystal.
Fig. 11. Diode-pumped Yb:KY(MoO4)2 cleaved crystal-plate microchip laser: (a) input-output dependences, η – slope efficiency; (b) typical laser emission spectra measured at Pabs = 1.0 W. The laser polarization is E || b.
Typical laser emission spectra are shown in Fig. 11(b). The spectra experienced a blue-shift with the increase of the output coupling: from 1020-1051 nm for TOC = 0.5% to 1012-1020 nm for TOC = 10%. This is due to the quasi-three level nature of the Yb3+ laser transition 2F5/2
→ 2F7/2 exhibiting reabsorption. The observed spectral behavior agrees with the gain spectra for E || b, Fig. 7(b). The laser output was linearly polarized (E || b) naturally selected by the anisotropy of the gain.
The laser spectra were modulated by etalon (Fabry-Perot) effects showing two wavelength separations, Δλ1 = 1.0 ± 0.1 nm and Δλ2 = 6.2 ± 0.5 nm. The smaller one is attributed to the etalon effect of the crystal itself and the larger one – to the residual small airgap between the crystal and one of the cavity mirrors. The free spectral range (FSR) of the Fabry-Perot etalon with a thickness t and a refractive index n at normal incidence is ΔλFSR ≈ λ2/(2nt) which yields 0.95 nm for the crystal (t = 286 µm and <λ> = 1030 nm). The thickness of the equivalent airgap is ∼90 µm. The roughly 20 nm broad laser spectra originated from the large gain bandwidth of Yb3+ in KY(MoO4)2.
The laser operation in the plano-plano cavity indicated a positive thermal lens for a-cut Yb:KY(MoO4)2. Positive thermal lens was observed also for monoclinic Yb:KY(WO4)2 crystal [38] and ascribed to the counteraction of negative thermo-optic coefficients dn/dT [39] and the positive thermal expansion. Negative dn/dT values were also measured for tetragonal double molybdates [40]. Note that for Yb:KY(MoO4)2, the largest coefficient of linear thermal expansion (38.3×10−6 K-1) is along the a-axis which may explain the positive thermal lens for the present crystal cut.
6. Conclusion
To conclude, we report on the growth, structure refinement by the Rietveld method, polarized room-temperature spectroscopy and crystal-field splitting (at 6 K), and first laser operation of Yb3+-doped potassium yttrium double molybdate, Yb:KY(MoO4)2. Its layered crystal structure dictates a strong polarization-anisotropy of the transition cross-sections for the Yb3+ dopant and a perfect natural cleavage habit along the (100) plane. Mechanically cleaved thin crystal films and plates with a thickness as thin as tens of µm can be directly applied as gain medium for laser operation. We report on a watt-level output from a diode-pumped microchip 3 at.% Yb:KY(MoO4)2 crystal-plate laser operating with a high slope efficiency (76.4%) almost approaching the Stokes limit. Highly Yb3+-doped KY(MoO4)2 crystal plates and films (e.g., up to 10 at.%, as obtained in our preliminary growth experiments) are attractive for short-pulse (sub-ns) passively Q-switched microchip lasers, as well as thin-disk lasers.
Funding
Russian Foundation for Basic Research (19-32-90199); Ministerio de Economía y Competitividad (MAT2016-75716-C2-1-R (AEI/FEDER,UE)); Agència de Gestió d'Ajuts Universitaris i de Recerca (2017 SGR 755); National Natural Science Foundation of China (51761135115, 61850410533, 61875199, 61975208); Deutsche Forschungsgemeinschaft (PE 607/14-1).
Acknowledgments
P. L. thanks Ms. Liza Basyrova for the help with the SEM studies.
Disclosures
The authors declare no conflicts of interest.
References
1. H. Zhu, Y. Chen, Y. Lin, X. Gong, Q. Tan, Z. Luo, and Y. Huang, “Growth, spectral properties, and laser demonstration of Yb3+:BaGd2(MoO4)4 cleavage crystal,” J. Appl. Phys. 101(6), 063109 (2007). [CrossRef]
2. Y. J. Chen, Y. D. Huang, X. Q. Lin, Q. G. Tan, Z. D. Luo, and Y. F. Lin, “Laser emission from unprocessed cleavage microchip,” Appl. Phys. Lett. 86(2), 021115 (2005). [CrossRef]
3. A. Volokitina, P. Loiko, A. Pavlyuk, S. Slimi, R. M. Solé, E. B. Salem, E. Kifle, J. M. Serres, U. Griebner, V. Petrov, M. Aguiló, F. Díaz, and X. Mateos, “Laser operation of cleaved single-crystal plates and films of Tm:KY(MoO4)2,” Opt. Express 28(7), 9039–9048 (2020). [CrossRef]
4. J. J. Zayhowski, “Microchip lasers,” Opt. Mater. 11(2-3), 255–267 (1999). [CrossRef]
5. J. M. Serres, X. Mateos, P. Loiko, K. Yumashev, N. Kuleshov, V. Petrov, U. Griebner, M. Aguiló, and F. Díaz, “Diode-pumped microchip Tm:KLu(WO4)2 laser with more than 3 W of output power,” Opt. Lett. 39(14), 4247–4250 (2014). [CrossRef]
6. P. Loiko, J. M. Serres, X. Mateos, K. Yumashev, A. Yasukevich, V. Petrov, U. Griebner, M. Aguiló, and F. Díaz, “Sub-nanosecond Yb:KLu(WO4)2 microchip laser,” Opt. Lett. 41(11), 2620–2623 (2016). [CrossRef]
7. Y. Chen, X. Gong, Y. Lin, Q. Tan, Z. Luo, and Y. Huang, “Continuous-wave laser characteristics of a Nd3+:LaB3O6 cleavage microchip and the influence of thermal effects,” Appl. Opt. 45(32), 8338–8345 (2006). [CrossRef]
8. J. Liu, W. Han, X. Chen, D. Zhong, B. Teng, C. Wang, and Y. Li, “Spectroscopic properties and continuous-wave laser operation of Yb:LuPO4 crystal,” Opt. Lett. 39(20), 5881–5884 (2014). [CrossRef]
9. X. Zhang, J. He, T. Tang, B. Teng, D. Zhong, X. Xu, and Z. Wang, “Efficient laser operations of unprocessed thin plate of Nd:YPO4 crystal,” Opt. Express 26(20), 26179–26187 (2018). [CrossRef]
10. Y. J. Chen, Y. F. Lin, X. H. Gong, H. M. Zhu, Z. D. Luo, and Y. D. Huang, “805-nm diode-pumped continuous-wave 2-µm laser performance of Tm3+:BaGd2(MoO4)4 cleavage plate,” Appl. Phys. B 98(1), 55–60 (2010). [CrossRef]
11. W. Zhao, Y. S. Huang, Z. B. Lin, B. Wei, F. W. Wang, M. Xu, X. Zhao, Q. H. Zheng, and W. W. Zhou, “Spectra and energy levels of a layered Yb3+:CsGd(MoO4)2 crystal with perfect cleavage: a candidate for microchip lasers,” RSC Adv. 5(44), 34730–34736 (2015). [CrossRef]
12. J. Liu, V. Petrov, H. Zhang, and J. Wang, “Power scaling of a continuous-wave and passively Q-switched Yb:KLu(WO4)2 laser end-pumped by a high-power diode,” Appl. Phys. B 88(4), 527–530 (2007). [CrossRef]
13. P. Loiko, J. M. Serres, X. Mateos, X. Xu, J. Xu, V. Jambunathan, P. Navratil, A. Lucianetti, T. Mocek, X. Zhang, U. Griebner, V. Petrov, M. Aguiló, F. Díaz, and A. Major, “Microchip Yb:CaLnAlO4 lasers with up to 91% slope efficiency,” Opt. Lett. 42(13), 2431–2434 (2017). [CrossRef]
14. H. Zhu, Y. Chen, Y. Lin, X. Gong, Z. Luo, and Y. Huang, “Polarized spectral properties and laser demonstration of Nd3+:BaGd2(MoO4)4 cleavage crystal,” J. Opt. Soc. Am. B 24(10), 2659–2665 (2007). [CrossRef]
15. V. Petrov, M. C. Pujol, X. Mateos, O. Silvestre, S. Rivier, M. Aguiló, R. M. Solé, J. Liu, U. Griebner, and F. Díaz, “Growth and properties of KLu(WO4)2, and novel ytterbium and thulium lasers based on this monoclinic crystalline host,” Laser Photonics Rev. 1(2), 179–212 (2007). [CrossRef]
16. A. A. Lagatsky, N. V. Kuleshov, and V. P. Mikhailov, “Diode-pumped CW lasing of Yb:KYW and Yb:KGW,” Opt. Commun. 165(1-3), 71–75 (1999). [CrossRef]
17. J. M. Serres, P. Loiko, X. Mateos, K. Yumashev, N. Kuleshov, V. Petrov, U. Griebner, M. Aguiló, and F. Díaz, “Prospects of monoclinic Yb:KLu(WO4)2 crystal for multi-watt microchip lasers,” Opt. Mater. Express 5(3), 661–667 (2015). [CrossRef]
18. U. Griebner, S. Rivier, V. Petrov, M. Zorn, G. Erbert, M. Weyers, X. Mateos, M. Aguiló, J. Massons, and F. Díaz, “Passively mode-locked Yb:KLu(WO4)2 oscillators,” Opt. Express 13(9), 3465–3470 (2005). [CrossRef]
19. S. Pekarek, C. Fiebig, M. C. Stumpf, A. E. H. Oehler, K. Paschke, G. Erbert, T. Südmeyer, and U. Keller, “Diode-pumped gigahertz femtosecond Yb:KGW laser with a peak power of 3.9 kW,” Opt. Express 18(16), 16320–16326 (2010). [CrossRef]
20. F. Brunner, T. Südmeyer, E. Innerhofer, F. Morier-Genoud, R. Paschotta, V. E. Kisel, V. G. Shcherbitsky, N. V. Kuleshov, J. Gao, K. Contag, A. Giesen, and U. Keller, “240-fs pulses with 22-W average power from a mode-locked thin-disk Yb:KY(WO4)2 laser,” Opt. Lett. 27(13), 1162–1164 (2002). [CrossRef]
21. S. Rivier, X. Mateos, Ò Silvestre, V. Petrov, U. Griebner, M. C. Pujol, M. Aguiló, F. Díaz, S. Vernay, and D. Rytz, “Thin-disk Yb:KLu(WO4)2 laser with single-pass pumping,” Opt. Lett. 33(7), 735–737 (2008). [CrossRef]
22. A. Volokitina, P. Loiko, E. Vilejshikova, X. Mateos, E. Dunina, A. Kornienko, N. Kuleshov, and A. Pavlyuk, “Eu3+:KY(MoO4)2: A novel anisotropic red-emitting material with a layered structure,” J. Alloys Compd. 762, 786–796 (2018). [CrossRef]
23. Y. Chen, Y. Lin, X. Gong, Q. Tan, Z. Luo, and Y. Huang, “Polarized spectral characteristics of Nd3+: KY(MoO4)2 crystal with perfect cleavage planes: a promising microchip gain medium,” J. Opt. Soc. Am. B 24(3), 496–503 (2007). [CrossRef]
24. P. Loiko, E. Kifle, J. M. Serres, X. Mateos, M. Aguiló, F. Díaz, E. Vilejshikova, N. Kuleshov, and A. Pavlyuk, “Efficient continuous-wave in-band pumped Nd:KY(MoO4)2 laser,” Laser Phys. Lett. 15(6), 065002 (2018). [CrossRef]
25. R. F. Klevtsova and S. V. Borisov, “X-ray structural study of the double molybdate KY(MoO4)2,” Sov. Phys. Dokl. 12, 1095 (1968) [Transl. from Dokl. Akad. Nauk SSSR 177(6), 1333-1336 (1967)].
26. J. Hanuza and L. Łabuda, “Polarized Raman and infrared spectra of a multilayer KY(MoO4)2 crystal,” J. Raman Spectrosc. 11(4), 231–237 (1981). [CrossRef]
27. A. A. Kaminskii and S. N. Bagayev, “Ribbon and sheet miniature crystal laser,” Quantum Electron. 24(12), 1029–1030 (1994). [CrossRef]
28. Y. A. Borovlev, N. V. Ivannikova, V. N. Shlegel, Y. V. Vasiliev, and V. A. Gusev, “Progress in growth of large sized BGO crystals by the low-thermal-gradient Czochralski technique,” J. Cryst. Growth 229(1-4), 305–311 (2001). [CrossRef]
29. A. A. Kaminskii, P. V. Klevtsov, and A. A. Pavlyuk, “Stimulated emission from KY(MoO4)2-Nd3+ crystal laser,” Phys. Status Solidi A 1(3), K91–K94 (1970). [CrossRef]
30. R. D. Shannon, “Revised effective ionic radii and systematic studies of interatomic distances in halides and chalcogenides,” Acta Crystallogr., Sect. A: Cryst. Phys., Diffr., Theor. Gen. Crystallogr. 32(5), 751–767 (1976). [CrossRef]
31. G. I. Frolova, L. E. Reznik, and I. E. Paukov, “Crystal-field effects on the thermodynamic properties of the KLn(MoO4)2 isostructural compounds,” J. Chem. Thermodyn. 21(1), 25–36 (1989). [CrossRef]
32. P. A. Loiko, K. V. Yumashev, N. V. Kuleshov, G. E. Rachkovskaya, and A. A. Pavlyuk, “Detailed characterization of thermal expansion tensor in monoclinic KRe(WO4)2 (where Re = Gd, Y, Lu, Yb),” Opt. Mater. 34(1), 23–26 (2011). [CrossRef]
33. L. Macalik, J. Hanuza, and A. A. Kaminskii, “Polarized Raman spectra of the oriented NaY(WO4)2 and KY(WO4)2 single crystals,” J. Mol. Struct 555(1-3), 289–297 (2000). [CrossRef]
34. L. A. Kulakova, “Acoustooptical and elastic properties of laminated KY(MoO4)2 crystals,” Phys. Solid State 42(1), 55–58 (2000) [transl. from Fizika Tverdogo Tela 42(1), 53–56 (2000)]. [CrossRef]
35. X. Mateos, J. M. Serres, P. Loiko, U. Griebner, V. Petrov, K. Yumashev, M. Aguiló, and F. Díaz, “Indium-modified Yb:KLu(WO4)2 crystal: Growth, spectroscopy and laser operation,” J. Lumin. 183, 391–400 (2017). [CrossRef]
36. B. Aull and H. Jenssen, “Vibronic interactions in Nd:YAG resulting in nonreciprocity of absorption and stimulated emission cross sections,” IEEE J. Quantum Electron. 18(5), 925–930 (1982). [CrossRef]
37. P. H. Haumesser, R. Gaumé, B. Viana, E. Antic-Fidancev, and D. Vivien, “Spectroscopic and crystal-field analysis of new Yb-doped laser materials,” J. Phys.: Condens. Matter 13(23), 5427–5447 (2001). [CrossRef]
38. P. A. Loiko, S. Manjooran, K. V. Yumashev, and A. Major, “Polarization-anisotropy of thermal lens in Yb:KY(WO4)2 laser crystal under high-power diode pumping,” Appl. Opt. 56(10), 2937–2945 (2017). [CrossRef]
39. P. A. Loiko, K. V. Yumashev, N. V. Kuleshov, and A. A. Pavlyuk, “Thermo-optical properties of pure and Yb-doped monoclinic KY(WO4)2 crystals,” Appl. Phys. B 106(3), 663–668 (2012). [CrossRef]
40. P. A. Loiko, X. Han, K. V. Yumashev, N. V. Kuleshov, M. D. Serrano, C. Cascales, and C. Zaldo, “Thermo-optical properties of uniaxial NaT(XO4)2 laser host crystals, (where T = Y, La, Gd or Bi and X = W or Mo),” Appl. Phys. B 111(2), 279–287 (2013). [CrossRef]
