Low-loss surface channel waveguides with a cross-section of 30 × 30 μm2 are produced by diamond saw dicing of 6.2 at.% Tm3+, 3.5 at.% Gd3+:LiYF4 films grown by liquid phase epitaxy (LPE) on (001)-oriented bulk undoped LiYF4 substrates. Pumped by a Ti:Sapphire laser at 783 nm, a continuous-wave Tm:LiYF4 waveguide laser generated 1.30 W at 1880 nm (for π-polarization) with a slope efficiency of 80% with respect to the absorbed pump power. The laser threshold was at 80 mW. The waveguide morphology was studied revealing low roughness (3 ± 2 μm) as expressed by the propagation losses of <0.3 dB/cm. A combination of LPE and diamond saw dicing is a promising technology for multi-watt single-mode channel waveguide lasers and amplifiers.
© 2018 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
Thulium ions (Tm3+) are known for their laser emission at around 2 μm due to the 3F4 → 3H6 transition . This emission is eye-safe and it spectrally overlaps with absorption lines of certain bio- and atmospheric molecules (H2O, CO2), so that Tm lasers are used in remote sensing (LIDAR), range-finding, medicine and spectroscopy . They are also suitable for pumping of Holmium (Ho3+) based oscillators emitting at >2 μm .
Due to the typically large Stark splitting of the ground-state (3H6), the ~2 μm emission of Tm3+ is broadband, which is used in tunable and mode-locked lasers. Tm3+ ions feature high absorption cross-sections at ~0.8 μm (the 3H6 → 3H4 transition) so that they can be easily excited, e.g., by commercial AlGaAs laser diodes or Ti:Sapphire lasers. The cross-relaxation (CR) process for adjacent Tm3+ ions, 3H4 + 3H6→ 3F4 + 3F4, is efficient in many materials even for moderate doping concentrations, leading to efficient population of the upper laser level with a quantum efficiency approaching 2. This increases the laser slope efficiency to >80%  and suppresses the unwanted heat dissipation.
Tetragonal lithium yttrium fluoride crystals (LiYF4) are known as suitable hosts for Tm3+ doping [5,6]. They features long lifetime of the 3F4 state (~10 ms) and efficient CR for Tm3+ ions [5,7]. High Tm3+ doping concentrations are feasible  accompanied with minor lattice distortion, weak energy-transfer upconversion  and weak luminescence quenching. Moreover, the LiYF4 crystal possesses low phonon energy (560 cm−1), low refractive index (n ~1.44 at 1.9 μm) and high thermal conductivity (~6 W/mK). Efficient power-scalable bulk Tm:YLF lasers have been reported [5,6].
Waveguide (WG) laser geometry allows one to fully utilize the advantages of Tm3+-doped materials for efficient lasing at ~2 μm . Tm WG lasers are also promising for sensing and telecom applications. So far, planar and channel active Tm WGs were fabricated by different methods. Liquid phase epitaxy (LPE) was used to produce single-crystalline thin films of Tm3+-doped Y3Al5O12 , LiYF4  and KY(WO4)2 . Structuring of such films by Ar+-ion etching provided low-optical-loss (< 0.1 dB/cm) channel WGs . Other known fabrication methods are direct bonding (e.g., for Tm:Y3Al5O12 crystal) , ion diffusion (e.g., for lead germanate glass or Tm:LiNbO3 crystal) [14,15], femtosecond laserwriting (fs-DLW, e.g., for Tm:ZBLAN glass or Tm:KLu(WO4)2 crystal) [16,17] or reactive co-sputtering (e.g., for amorphous Tm:Al2O3) .
It has to be noticed that the highest slope efficiency from any Tm WG laser has been achieved using a LPE-grown and Ar+-ion-microstructured 8 at.% Tm:KY1-x-yGdxLuy(WO4)2 buried channel WG . Pumped by a Ti:Sapphire laser at 794 nm, it delivered 1.5 W of continuous-wave (CW) output at 1.84 μm with a slope efficiency η of ~81% (with respect to the absorbed pump power). Much higher power of 15 W at 2.02 μm has been reported from a direct-bonded diode-side-pumped Tm:Y3Al5O12 planar WG laser whilst at lower η = 43% and relatively poor beam quality (M2>300 in the horizontal direction) .
In the present work, we aimed to demonstrate the first Tm fluoride channel WG laser featuring watt-level laser output with a slope efficiency exceeding 80% and based on LPE technology for the thin film growth of Tm:LiYF4 and diamond saw dicing for its micro-structuring. The previously reported Tm:LiYF4 planar WG laser generated 0.56 W at 1.88 μm with η = 76% . Recently, optical grade diamond blade dicing was implemented for fabrication of LiNbO3 ridge waveguides [19,20].
2.1 LPE growth
The Tm,Gd:LiYF4 active layers of high optical quality were grown on undoped LiYF4 substrates by LPE . The bulk LiYF4undoped crystals were grown by the Czochralski method using a -oriented seeds. Double-side laser-grade-quality polished substrates (dimensions: 7 × 20 × 2.5 mm3) were oriented with their surfaces being parallel to the (001) crystallographic plane. The films were grown using a batch with a composition of 73 mol% LiF – 27 mol% YF3 with 7 at.% Tm3+ and 5 at.% Gd3+ replacing the Y3+ ions. LiF was used both as solvent and as solute. The optically passive Gd3+ ions were added to increase the refractive index difference between the substrate and the active layer Δn. The Tm,Gd:LiYF4 layers are isostructural to LiYF4 (tetragonal space group C64h – I41/a).
The growth was performed at 2°C below the saturation temperature (Ts = 761 °C) for 25 min resulting in a transparent and crack-free layer with an area of 7 × 10 mm2 and a thickness of ~50 μm, Fig. 1(a).
The composition of the films was studied by Energy Dispersive X-ray spectroscopy (EDX). The results on the segregation coefficients of Tm3+ (KTm = 0.88) and Gd3+ (KGd = 0.7) ions are shown in Fig. 1(b) and compared with our previous data for rare-earth ions (RE3+) in LPE-grown LiYF4 films. In LiYF4, these ions replace the Y3+ ones in a single type of sites (S4) with an VIII-fold O2- coordination. The KRE value is strongly dependent on the ratio of ionic radii RRE/RY (RY = 1.019 Å, RGd = 1.053 Å and RTm = 0.994 Å). The stoichiometric layer composition was LiY0.903Gd0.035Tm0.062F4. The actual concentration of Tm3+ was 8.6 × 1020 cm−3.
2.2 Waveguide fabrication
LiYF4 is an optically uniaxial (positive) crystal. Its optical axis is parallel to the c-axis, so that there are two principal light polarizations, π (E || c) and σ (E⊥c). The Tm,Gd:LiYF4 / LiYF4 epitaxy was oriented for light propagation along the a-axis giving access to both polarizations. The top surface of the active layer was polished with a laser-grade quality to a thickness of 30 ± 1 μm. The length of the sample l was 8.0 mm.
For micro-structuring of the Tm,Gd:LiYF4 active layer, we used a DISCO DAD3350 high-precision circular saw. The individual WGs were fabricated by repetitive passing a circular resinoid blade along the crystal a-axis through the whole length of the WG. This simultaneously provided cutting of the layer and polishing of the WG side walls . The latter is because the blade made of resin and grains of fine and calibrated diamonds, mills LiYF4 with a precision of about 1 μm and 1” (angular second).The rotation speed was 10000 revolutions per minute (rpm) and the cutting speed - 0.2 mm/s. The cuts were cooled with a clean flowing water (18 °C). The produced channel WGs were directly operational.
The morphology of the WGs was studied with a confocal laser microscopy using a blue (405 nm) light source. Figure 2(a) shows a top-view of the sample with a bottom part representing the initial (planar) Tm,Gd:LiYF4 / LiYF4 epitaxial structure and the formed channel WGs having a width d of 30 ± 1μm and separated from each other by a diced area of 200 ± 5μm. In Fig. 2(b), we show a closer look on a single channel WG. The roughness of the cut is 3 ± 2 μm. In Figs. 2(c) and 2(d), a side view on the structure is shown. The depth of the dicing is 35 ± 3 μm, so no active layer is preserved in the gap between two adjacent WGs. The cross-section of the channel WG is almost a square (d ≈t). A total of 8 WGs were written.
2.3 Mode analysis
At first, we studied the refractive index of the substrate and the active layers using the M-line method , see Fig. 3(a). The measurements were performed for TM modes corresponding to π polarization (ne refractive index). A similar 9 at.% Tm, 5 at.% Gd:LiYF4 layer (batch composition) was studied. The measured ne values for the substrate are in good agreement with the Sellmeier curve reported by Barnes et al. . The measured ne values for the active layer were thus fitted using an equation of a similar form, Fig. 3(a). Considering a linear variation of the refractive index with Tm3+doping, the calculated refractive indices for the 7 at.% Tm, 5 at.% Gd:YLF layer at 1.880 μm (laser wavelength) are ne(substrate) = 1.4652 and ne(layer) = 1.4675. The refractive index contrast Δn is 2.3 ± 0.5 × 10−3.
The mode analysis for the channel WG was performed using software from . At the pump wavelength (0.78 μm), the WG is multimode. The plot of effective refractive indices nefffor the five first TM WG modes vs. the size of a square channel WG (d × d) is shown in Fig. 3(b) for the laser wavelength (1.880 μm). For the WG lateral size of 30 × 30 μm2, three modes (TM00, TM10 and TM01) are supported.
2.4 Laser set-up
The scheme of the Tm:LiYF4 channel WG laser is shown in Fig. 4(a). As a pump source, we used a CW Ti:Sapphire laser (model 3900S, Spectra Physics) tuned to 783 nm (out of the local maximum in the 3H6 → 3F4 absorption band of Tm3+ for π-polarization, to ensure a uniform pump distribution over the whole WG length) and generating ~3.2 W of output power. The pump polarization was defined by a Glan-Taylor polarizer. The pump power incident onto the WG was varied by a rotatory λ/2 plate. The pump beam was focused by a spherical lens (AC254-050-B, Thorlabs, focal length f = 50 mm, transmission T = 88%) providing a pump spot diameter at the input face of the WG 2wP of 32 μm. The pump coupling efficiency ηcoupl was 87 ± 1% (the value including the Fresnel loss). It was determined from pump-transmission experiments at 0.84 μm, out of Tm3+ absorption. The pump absorption under lasing conditions ηabs was 74.0 ± 0.5% also determined from pump-transmission measurements but at 0.78 μm, and at the threshold pump power.
The laser cavity was composed by a flat pump mirror (PM) coated for high reflection (HR) at 1.87-2.30 μm and for high transmission (HT, T = 96%) at 0.78 μm, and a set of flat output couplers (OCs) with transmissions TOC ranging from 2% to 50% at 1.88 μm. Both PM and OC were placed close to the WG end-facets with minimum air gaps, <100 μm. No index-matching liquid was used to avoid optical damage. The WG was mounted on a Cu-holder using a silver paste for better heat removal from the substrate side. It was passively cooled. The laser output was filtered from the residual (non-absorbed) pump by a cut-off filter (FEL 900, Thorlabs, T = 83.2% at 1.88 μm and T < 0.01% at 0.78 μm). The laser emission spectra were measured using an optical spectrum analyzer (OSA, model AQ6375B, Yokogawa).
The channel WG laser exhibited a weak blue (0.47 μm) emission related to the 1G4 → 3H6 Tm3+ transition after upconversion, Fig. 4(b).
3. Results and discussion
The input-output dependences of the Tm:LiYF4 WG laser are shown in Fig. 5(a). For all OCs, the laser output was linearly polarized (π). The polarization was selected by the gain anisotropy. The laser slope efficiency gradually increased with TOC. For TOC = 50%, the WG laser generated 1.30 W at 1880 nm with a maximum η of 79.7% (with respect to the absorbed pump power Pabs). The laser threshold was at Pabs = 70 mW and the optical-to-optical conversion efficiency ηconv amounted to 47.2% (with respect to the incident pump power). No thermal roll-over and no WG damage (no end-face damage by the pump beam nor thermal fracture of the WG) were observed up to at least Pabs = 1.8 W. The laser performance was similar for all 8 prepared WGs; the variation of the output power was less than 10%. At the maximum Pabs, stable operation without degradation of the output power was observed for at least 10 min. Further improvement of the long-term laser stability is expected with active cooling.
The typical laser emission spectra are shown in Fig. 5(b). They were weakly dependent on the output coupling and the pump level, the emission occurred between 1874 and 1888 nm. The multi-band spectral behavior is due to the etalon effects at the crystal-mirror interfaces partially dependent on the laser alignment.The observed emission spectra are well in line with the gain spectra of Tm3+ ions in LiYF4 for π-polarization, see Fig. 6. For inversion ratiosβ = N2(3F4)/NTm ranging from 0.16 to 0.26 (as expected for TOC = 2% - 50%), the local peak centered at around 1.88 μm dominates in the spectra.
Laser operation with the Tm:LiYF4 WG was also studied for the case of very high output couping (without any OC, so that TOC = 96.4% is determined by Fresnel losses at the polished WG end-facet), Fig. 5. The pump absorption was only 27 ± 1% due to the strong bleaching of the WG. The laser generated 0.35 W at 1878 nm with η = 67.4%. The laser threshold was at Pabs = 115 mW. The laser emission was π-polarized. The decrease of the slope efficiency with respect to TOC = 50% is most probably due to the strong energy-transfer upconversion (ETU). Indeed, the intensity of blue emission from the WG increased by few orders of magnitude. Thus, the optimum OC for the studied laser is expected to be near or slightly above 50%.
The performance of the Tm:LiYF4 channel WG laser for TOC ranging from 2% to 50% was analyzed using Caird plot  modified for the case of high output coupling , Fig. 7. Thedependence of η on the output coupling loss γOC is represented as 1/η = 1/η0(1 + 2γ/γOC), where γ = –ln(1 – L), L is the internal loss per pass, γOC = –ln(1 – TOC) and η0 is an intrinsic slope efficiency. The best-fit of the experimental data in Fig. 7 yields η0 = 77.0% and a coefficient of passive losses δ = 4.34L/l = 0.27 ± 0.1 dB/cm. Note that this value includes the losses at the mirror / crystal interfaces and is averaged over all used OCs. Even lower passive losses are expected for the WG itself (they are revealed for the high-quality bulk OCs with TOC = 35% and 50%), estimated to be about 0.1 dB/cm in agreement with the value reported in  for a planar Tm:LiYF4 layer. The Caird analysis is valid if the laser emission wavelength λL and, consequently, the stimulated-emission (SE) cross-section, σSE(λL), are constant. In our case, see Fig. 5(b), λL experiences only a slight blue-shift with increasing TOC. Within this spectral range, σSE of Tm3+ ions in LiYF4 is almost constant and equal to 0.4 × 10−20 cm2 .
The far-field profile of the output laser beam from the Tm:LiYF4 channel WG laser, see Fig. 8(a), was measured using a spherical lens (f = 40 mm) and a thermal imaging screen. The profile is nearly circular. In Fig. 8(b), we show the intensity profile of the TM00 mode calculated using software from  and the Δn value determined above. The mode is expected to be well-confined within the WG (the profile overlap integral Г is 0.987).
The laser performance of the Tm:LiYF4 channel WG was simulated using a rate-equation model accounting for (i) depopulation of the ground-state (3H6), (ii) CR, (iii) ETU from the 3F4 state, 3F4 + 3F4 → 3H6 + 3H4, and (iv) energy migration from the 3F4 state. The scheme of the Tm3+ energy levels in LiYF4 and the relevant spectroscopic processes is shown in Fig. 9(a).
The populations N2 and N3 (N1 + N2 + N3 = NTm) are expressed using the following set of rate equations:26] and our unpublished results). Thus, for the condition of negligible ground-state bleaching, the CR rate constant WCR = C'CR∙(NTm)2 is about 1.8 ± 0.2 × 104 s−1 and the theoretical quantum efficiency of excitation to the 3F4 state ηq is about 1.95 ± 0.02 (for 6.2 at.% Tm:LiYF4). The ETU parameter CETU for 6 at.% Tm doping is estimated from  as 5.5 × 10−19 cm3s−1. τ2 is the experimental luminescence lifetime of the 3F4 level, for which the energy migration (expressed by a rate Wd) was taken into account, 1/τ2 = 1/τ20 + Wd = 6.07 ms  (τ20 is the unquenched lifetime).
All variables (N2, N3, IP and IL) were considered as functions of the axial (z) and radial (r) coordinates. The pump and laser beams were assumed to have Gaussian intensity profiles. The output power was calculated from the photon flux for the straight-forward propagating wave I+L (h is the Planck constant):
The results of the modeling of the output laser performance are shown in Fig. 8(b). The measured input-output dependence for TOC = 50% is plotted vs. the incident pump power because it was used as an initial condition for the modeling. The model predicts generation of 1.21 W at 1.88 μm with ηopt = 44.7% and it agrees with the experimental data. When assuming a constant WCR for the whole range of the incident pump power, the calculated input-output dependence deviates from the linear one. This is due to population of the 3H4 state when the pump rate becomes comparable to the rates of competing processes (CR and radiative decay). As a result, depopulation of the ground-state (3H6) and, thus, decrease of ηabs are expected. However, the experimental curve in Fig. 8(b) is linear well above the threshold. A good agreement was achieved when considering an increase of WCR with the pump power.
Physically, the increase of WCR with temperature (T) is attributed to heating of the WG due to the absorbed pump power. CR is a multi-phonon assisted energy transfer process . For Tm3+ ions in LiYF4, the 3H4 + 3H6→ 3F4 + 3F4 CR mechanism requires emission of Ne = 3 phonons with an energy hνph = 560 cm−1 for each of them. The temperature dependence of this process is expressed as [27,28]:Eq. (3) assuming a temperature rise below 100 K.
In Table 1, we compared the output characteristics of Tm channel WG lasers reported so far (the best results were selected from multiple publications). Channel WGs produced by Ti or Zn ion diffusion in Tm:LiNbO3 [14,32] generated low (<1 mW) output power while they can have specific applications due to multi-functionality of lithium niobate. Ion exchange was used to fabricate channel WGs in Tm:germanate glass  showing certain prospects over ion implantation . However, these methods are suitable mostly for glasses. Co-sputtering of amorphous Tm:Al2O3  has an advantage of silicon-based technology and it is suitable for the fabrication of distributed feedback (DFB) and distributed Bragg reflector (DBR) structures. Recently, a lot of attention has been paid for fs-DLW of channel waveguides in different Tm3+-doped materials, including glasses , ceramics [25,31] and single-crystals . This method features fast fabrication time, a variety of possible materials and multiple WG shapes. However, it is typically associated with higher optical losses (0.5-1 dB/cm) [17,30], relatively low refractive index contrast (Δn ~6 × 10−4)  and low potential for power scaling due to increased thermal fracture probability. Using fs-DLW, the highest CW output power to date has been extracted from a multimode Tm:ZBLAN glass WG, namely 205 mW at 1890 nm with η = 67% .
LPE is a well-known method to produce high-optical-quality thin crystalline films . A combination of LPE and Ar+-ion etching produced highly-efficient channel WGs based on “mixed” Tm:KY1-x-yGdxLuy(WO4)2 monoclinic double tungstate (MDT) crystals [4,12] featuring low passive losses and watt-level output. The main drawbacks of MDTs are moderate thermal conductivity (3 W/mK) and strongly anisotropic thermal expansion . This increases the risk of thermal fracture and makes the growth more complicated. However, both these disadvantages are not present for LiYF4 and diamond saw dicing is the easier way to fabricate active μm-size channel WGs as compared to Ar+-ion etching. It allows one to achieve highly-efficient multi-watt Tm3+ lasing, cf. Table 1.
Further improvement of the output characteristics of Tm:LiYF4 channel WG lasers is possible in several ways. To decrease the passive losses, one need to optimize the dicing parameters. Overgrowth of the WGs by undoped layer of LiYF4 can also serve this aim. In this way, passive losses well below 0.1 dB/cm are expected. The power scaling of the WG laser and much higher amplification performances can be reached by increase of Tm3+ doping concentration to 10-15 at.%. Such active layers of high optical quality has been already achieved in our preliminary growth experiments. Diode-pumping of channel WGs is also achievable in double clad (overgrown) WG structures. Concerning a single transverse mode operation, reduction of WG cross-section to about 20 × 20 μm2 is needed, see Fig. 3(b). It seems feasible due to the low WG roughness achieved under diamond saw dicing.
We report on the first highly-efficient watt-level 2 μm channel waveguide laser produced by a combination of liquid phase epitaxy and diamond saw dicing. This laser is based on Tm,Gd:LiYF4 active layers grown on bulk LiYF4 substrates. It generates 1.30 W of CW output at 1880 nm with a slope efficiency of ~80% exploiting a quantum efficiency of the Tm3+ excitation of about 1.96 due to the efficient CR. The WGs feature low propagation losses of about 0.1-0.3 dB/cm without overgrown cladding. Thermal effects are negligible in passively-cooled cm-long devices under multi-watt laser pumping. The WGs are promising for highly-efficient lasers and amplifiers at ~2 μm. Functionalization of their top surface by nonlinear optical materials (e.g., graphene) or bio-molecules is of interest for generation of pulsed output and for sensing applications.
AgenceNationale de la Recherche (ANR) through the LabEx EMC3 project FAST-MIR; European Community funds FEDER and the Normandy region.
2. K. Scholle, S. Lamrini, P. Koopmann, and P. Fuhrberg, “2 μm laser sources and their possible applications,” in Frontiers in Guided Wave Optics and Optoelectronics, B. Pal, Ed. (Intech, 2010, pp. 471–500).
3. P. A. Budni, M. L. Lemons, J. R. Mosto, and E. P. Chicklis, “High-power/high-brightness diode-pumped 1.9-μm thulium and resonantly pumped 2.1-μm holmium lasers,” IEEE J. Sel. Top. Quantum Electron. 6(4), 629–635 (2000). [CrossRef]
4. K. van Dalfsen, S. Aravazhi, C. Grivas, S. M. García-Blanco, and M. Pollnau, “Thulium channel waveguide laser with 1.6 W of output power and ∼80% slope efficiency,” Opt. Lett. 39(15), 4380–4383 (2014). [CrossRef] [PubMed]
5. S. So, J. I. Mackenzie, D. P. Sheperd, W. A. Clarkson, J. G. Betterton, and E. K. Gorton, “A power-scaling strategy for longitudinally diode-pumped Tm:YLF lasers,” Appl. Phys. B 84(3), 389–393 (2006). [CrossRef]
6. M. Schellhorn, “High-power diode-pumped Tm:YLF laser,” Appl. Phys. B 91(1), 71–74 (2008). [CrossRef]
7. B. M. Walsh, N. P. Barnes, and B. Di Bartolo, “Branching ratios, cross sections, and radiative lifetimes of rare earth ions in solids: Application to Tm3+ and Ho3+ ions in LiYF4,” J. Appl. Phys. 3(5), 2772–2787 (1998). [CrossRef]
8. P. Loiko, J. M. Serres, X. Mateos, S. Tacchini, M. Tonelli, S. Veronesi, D. Parisi, A. Di Lieto, K. Yumashev, U. Griebner, and V. Petrov, “Comparative spectroscopic and thermo-optic study of Tm:LiLnF4 (Ln = Y, Gd, and Lu) crystals for highly-efficient microchip lasers at ~2 μm,” Opt. Mater. Express 7(3), 844–854 (2017). [CrossRef]
9. A. Rameix, C. Borel, B. Chambaz, B. Ferrand, D. P. Shepherd, T. J. Warburton, D. C. Hanna, and A. C. Tropper, “An efficient, diode-pumped, 2 μmTm:YAG waveguide laser,” Opt. Commun. 142(4–6), 239–243 (1997). [CrossRef]
10. W. Bolanos, F. Starecki, A. Benayad, G. Brasse, V. Ménard, J.-L. Doualan, A. Braud, R. Moncorgé, and P. Camy, “Tm:LiYF4 planar waveguide laser at 1.9 μm,” Opt. Lett. 37(19), 4032–4034 (2012). [CrossRef] [PubMed]
11. W. Bolaños, J. J. Carvajal, X. Mateos, E. Cantelar, G. Lifante, U. Griebner, V. Petrov, V. L. Panyutin, G. S. Murugan, J. S. Wilkinson, M. Aguiló, and F. Díaz, “Continuous-wave and Q-switched Tm-doped KY(WO4)2 planar waveguide laser at 1.84 µm,” Opt. Express 19(2), 1449–1454 (2011). [CrossRef] [PubMed]
13. J. I. Mackenzie, S. C. Mitchell, R. J. Beach, H. E. Meissner, and D. P. Shepherd, “15 W diode-side-pumped Tm:YAG waveguide laser at 2 μm,” Electron. Lett. 37(14), 898–899 (2001). [CrossRef]
14. J. P. De Sandro, J. K. Jones, D. P. Shepherd, M. Hempstead, J. Wang, and A. C. Tropper, “Non-photorefractive CW Tm-indiffused Ti:LiNbO3 waveguide laser operating at room temperature,” IEEE Photonics Technol. Lett. 8(2), 209–211 (1996). [CrossRef]
15. D. P. Shepherd, D. J. B. Brinck, J. Wang, A. C. Tropper, D. C. Hanna, G. Kakarantzas, and P. D. Townsend, “1.9-microm operation of a Tm:lead germanate glass waveguide laser,” Opt. Lett. 19(13), 954–956 (1994). [CrossRef] [PubMed]
16. D. G. Lancaster, S. Gross, H. Ebendorff-Heidepriem, K. Kuan, T. M. Monro, M. Ams, A. Fuerbach, and M. J. Withford, “Fifty percent internal slope efficiency femtosecond direct-written Tm3+:ZBLAN waveguide laser,” Opt. Lett. 36(9), 1587–1589 (2011). [CrossRef] [PubMed]
17. E. Kifle, P. Loiko, X. Mateos, J. R. V. de Aldana, A. Ródenas, U. Griebner, V. Petrov, M. Aguiló, and F. Díaz, “Femtosecond-laser-written hexagonal cladding waveguide in Tm:KLu(WO4)2: µ-Raman study and laser operation,” Opt. Mater. Express 7(12), 4258–4268 (2017). [CrossRef]
18. N. Li, P. Purnawirman, Z. Su, E. Salih Magden, P. T. Callahan, K. Shtyrkova, M. Xin, A. Ruocco, C. Baiocco, E. P. Ippen, F. X. Kärtner, J. D. B. Bradley, D. Vermeulen, and M. R. Watts, “High-power thulium lasers on a silicon photonics platform,” Opt. Lett. 42(6), 1181–1184 (2017). [CrossRef] [PubMed]
19. N. Courjal, B. Guichardaz, G. Ulliac, J.-Y. Rauch, B. Sadani, H.-H. Lu, and M.-P. Bernal, “High aspect ratio lithium niobate ridge waveguides fabricated by optical grade dicing,” J. Phys. D Appl. Phys. 44(30), 305101 (2011). [CrossRef]
20. M. F. Volk, S. Suntsov, C. E. Rüter, and D. Kip, “Low loss ridge waveguides in lithium niobate thin films by optical grade diamond blade dicing,” Opt. Express 24(2), 1386–1391 (2016). [CrossRef] [PubMed]
21. F. Starecki, W. Bolaños, G. Brasse, A. Benayad, M. Morales, J. L. Doualan, A. Braud, R. Moncorgé, and P. Camy, “Rare earth doped LiYF4 single crystalline films grown by liquid phase epitaxy for the fabrication of planar waveguide lasers,” J. Cryst. Growth 401, 537–541 (2014). [CrossRef]
22. N. P. Barnes and D. J. Gettemy, “Temperature variation of the refractive indices of yttrium lithium fluoride,” J. Opt. Soc. Am. 70(10), 1244–1247 (1980). [CrossRef]
23. O. V. Ivanova, R. Stoffer, and M. Hammer, “A variational mode solver for optical waveguides based on quasi-analytical vectorial slab mode expansion,” University of Twente, technical report (2009, pp. 1–19).
24. J. A. Caird, S. A. Payne, P. R. Staber, A. J. Ramponi, L. L. Chase, and W. F. Krupke, “Quantum electronic properties of the Na3Ga2Li3F12:Cr3+ laser,” IEEE J. Quantum Electron. 24(6), 1077–1099 (1988). [CrossRef]
25. J. Morris, N. K. Stevenson, H. T. Bookey, A. K. Kar, C. T. A. Brown, J.-M. Hopkins, M. D. Dawson, and A. A. Lagatsky, “1.9 µm waveguide laser fabricated by ultrafast laser inscription in Tm:Lu2O3 ceramic,” Opt. Express 25(13), 14910–14917 (2017). [CrossRef] [PubMed]
26. I. Razumova, A. Tkachuk, A. Nikitichev, and D. Mironov, “Spectral-luminescent properties of Tm:YLF crystal,” J. Alloys Compd. 225(1–2), 129–132 (1995). [CrossRef]
27. J. Ganem, J. Crawford, P. Schmidt, N. W. Jenkins, and S. R. Bowman, “Thulium cross-relaxation in a low phonon energy crystalline host,” Phys. Rev. B 66(24), 245101 (2002). [CrossRef]
28. F. Auzel, “Multiphonon-assisted anti-Stokes and Stokes fluorescence of triply ionized rare-earth ions,” Phys. Rev. B 13(7), 2809–2817 (1976). [CrossRef]
30. D. G. Lancaster, S. Gross, A. Fuerbach, H. E. Heidepriem, T. M. Monro, and M. J. Withford, “Versatile large-mode-area femtosecond laser-written Tm:ZBLAN glass chip lasers,” Opt. Express 20(25), 27503–27509 (2012). [CrossRef] [PubMed]
32. E. Cantelar, J. A. Sanz-García, G. Lifante, F. Cussó, and P. L. Pernas, “Single polarized Tm3+ laser in Zn-diffused LiNbO3 channel waveguides,” Appl. Phys. Lett. 86(16), 161119 (2005). [CrossRef]
33. B. Ferrand, B. Chambaz, and M. Couchaud, “Liquid phase epitaxy: A versatile technique for the development of miniature optical components in single crystal dielectric media,” Opt. Mater. 11(2–3), 101–114 (1999). [CrossRef]
34. O. Silvestre, J. Grau, M. C. Pujol, J. Massons, M. Aguiló, F. Díaz, M. T. Borowiec, A. Szewczyk, M. U. Gutowska, M. Massot, A. Salazar, and V. Petrov, “Thermal properties of monoclinic KLu(WO4)2 as a promising solid state laser host,” Opt. Express 16(7), 5022–5034 (2008). [CrossRef] [PubMed]