In this work we report the spectroscopy and laser results of several Thulium doped BaY2F8 single crystals grown using the Czochralski technique. The doping concentration is between 2at.% and 18at.%. We performed room temperature laser experiments pumping the samples with a laser diode at 789 nm obtaining 61% as maximum optical-to-optical efficiency with a maximum output power of 290 mW and a minimum lasing threshold of 26 mW. The lasing wavelength changed with the dopant concentration from 1927 nm up to 2030 nm and the nature of the transition changed from purely electronic to vibronic, accordingly.
©2004 Optical Society of America
The interest in novel solid state laser emitting in the 2 μm region is continuously increasing because they can be used to develop eye-safe compact devices suitable for several applications such as DIAL (Differential Absorption LIDAR) measurements, telecommunication or medical equipment. The emission of Tm3+ at 2 μm has been studied in different hosts, but, to the best of our knowledge it is the first time that laser action from Tm3+ :BaY2F8 (Tm:BaYF) has been obtained in this wavelength region. A previous work  report laser action in the same crystal but at a different wavelength and with Yb3+ sensitizing.
Tm3+ is well-known as active ion to achieve population inversion for the 3F4 → 3H6 transition (between 1.9 and 2 μm) pumping the 3H4 manifold by means of diode laser operating around 800 nm via the cross-relaxation process (3H4,3H6)→(3F4,3F4).
The use of fluoride crystals as laser hosts is favoured because of their wide transparency window (ranging from UV up to far IR) and the low phonon energy; in particular BaYF is the one with the lowest width of phonon spectrum, only 400 cm-1.
Since no data are available on the optimal concentration of Thulium in BaYF to maximize the 3F4 → 3H6 emission, we grew several samples with various doping concentrations. Laser action was obtained with Tm concentration ranging from 2at.% to 18at.% in order to investigate the features of the laser emission as a function of the doping density.
The main feature of our laser is that the laser emission extends in a wavelength region where no electronic transitions exist: following the pioneer work of Ref.  we argue that our results are explained with a vibronic shift (observed also in Refs. [3, 4, 5]).
The samples under investigation were six BaYF single crystals doped with Thulium concentrations: 2at.%, 5at.%, 8at.%, 12at.% and 18at.%. BaYF has monoclinic structure and symmetry group C2/m. The unit-cell parameters are a=6.972Å, b=10.505 Å, and c=4.260 Å, with angle γ between the a-axis and the c-axis of 99.76° and the unit cell contains two molecules. The BaYF matrix is transparent from UV to IR (≈ 9μm), and its refractive index n is 1.5. The rare-earth dopant substitutionally enters the Y3+ sites.
2.1. Crystal growth
The crystal growth apparatus consists of a home-made Czochralski furnace with conventional resistive heating; special care has been devoted to the quality of the vacuum system, which has an ultimate pressure limit below 10-7 mbar.
All the crystals were grown using BaY2F8 powder as raw material for the crystal and the doping density was achieved by adding a proper amount of BaTm2F8 powder for the 2%, 5% and 8% doped samples. For the 12% and 18% samples we mixed suitable amounts of BaF2 and TmF3 powders, instead. To avoid OH- contamination, the powders were purified at AC Materials (Orlando,Fla.,USA) and the growth process was carried out in a high-purity (99.999%) argon atmosphere. During growth, the rotation rate of the sample was 5rpm, the pulling rate 0.5 mm/h, and the temperature of the melt was in the range 995-987°C. The furnace is also provided with an optical computer-controlled apparatus for diameter control. The average size of the BaYF crystals was about 15 mm in diameter and 55 mm in length. The single crystalline character of the samples was checked using a X-ray Laue technique that allows us to identify the crystallographic axis of the crystal and cut oriented samples. The dimensions of the crystal sample inside the laser were 4×5 mm2, with thicknesses depending on the dopant concentration. The wider faces were optically polished to laser tolerance.
Room temperature absorption measurements were performed by a CARY 500 spectrophotometer between 250 and 1900 nm with a resolution of 0.3 nm in the UV-VIS wavelength region and 1 nm in the NIR region.
The room temperature fluorescence spectra of the 3F4 → 3H6 transition were performed with the aim to measure the emission cross-section and were taken only for the 8% doped sample. As a pumping source we used a home-made cw tunable Ti:Al2O3 (pumped by an Ar+ laser). The fluorescence signal of the steady-state emission measurements was detected perpendicularly to the pump laser direction to avoid pump spurious scattering. The luminescence was chopped and focused by a monochromator with 25 cm focal length and equipped with a 300 gr/mm grating (blaze at 2 μm); the resolution of the measurement was 1.2 nm. To record the spectra as a function of the orientation of the samples we used a Glan-Thomson polarizer in front of the input slit of the monochromator. The signal was filtered by a silicon filter and detected by a liquid nitrogen cooled InSb detector, fed into pre-amplifiers, processed by a lock-in amplifier and subsequently stored on a PC.
We measured the room temperature fluorescence decay-time for the 3F4 manifold for all the samples under investigation. The crystals were excited by a pulsed tunable Ti:Al2O3 laser (pumped by a doubled Nd:YAG pulsed laser) with 10 Hz repetition rate, 30 ns pulse width. We collected the signal from a short portion of the sample (≈ 1 mm) to have the energy density as constant as possible in the observed region and to reduce any undesired effect such as radiation trapping, that can cause an incorrect evaluation of the decay time. The dynamic fluorescence was detected by the same experimental apparatus described above, but the signal from the InSb detector was sent, by fast amplifiers, to a digital oscilloscope connected to a computer.
Room temperature laser experiments were aimed to determine the cw performance of the active crystals as a function of the orientation and of the dopant concentration. The laser resonator was a nearly hemispherical optical cavity. The active medium was pumped by a AlGaAs laser diode emitting at around 790 nm with a maximum output power of 1 W with an emission bandwidth of approximately 2 nm. Astigmatism compensation was accomplished by means of an aspheric lens and a couple of anamorphic prisms. The pumping beam was coupled through a input coupler (IC) plane mirror with high reflectivity (HR) between 1850 and 2150 nm (R > 99.9%) with anti-reflection (AR) coating at 790 nm (T > 96%). The pump laser was focussed into the active medium by a 50-mm lens. The output coupler (OC) was a curved mirror (radius of curvature=100 mm) transmitting 0.5% between 1850 and 2150 nm. The active material was placed close to the IC plane mirror. The active material was placed close to the IC plane mirror. Using the standard theory for optical propagation in resonators  for the calculation of the laser mode waist, we estimated ω0 ≈ 60 μm. The samples without any anti-reflection coating or cooling, were mounted with the b-axis parallel to the fast-axis of the diode laser. The OC was placed on a horizontal stage translator in order to optimize the cavity length at the maximum output power; the total length of the laser cavity was about 10 cm. To avoid the residual light coming from the diode, the output laser emission was filtered by an AR-coated Ge filter and the power was measured with a power meter. The emission spectra of the laser were measured with a set-up similar to the one used to collect the fluorescence spectra.
3. Spectroscopic analysis
3.1. Absorption spectra
The first optical measurement performed on the samples consisted in the acquisition of the absorption spectrum. Those data can be helpful when it is necessary to verify the absence of impurities that could be introduced inside the crystal during the growth and that seriously compromise the optical performances. We checked that the UV absorption spectrum showed no evidence of incorporation of optically active impurities, within the sensitivity of the spectrophotometer, as a hint of the high purity and quality of the crystals.
Furthermore we recorded the IR absorption spectra for the samples under investigation down to 1900 nm for different orientations of the crystals. Fig. 1 shows the polarized absorption spectra for the 12% Tm:BaYF for the pumping and emission regions. The main peak position of the 3H4 level is 789.2 nm in E ∥ b polarization with a FWHM of 2 nm for all the samples and the values of the absorption coefficient depend on the Tm concentration: therefore this peak is suitable for diode-pumping.
3.2. Static and dynamic fluorescence measurements
The aim of our spectroscopic analysis was not to fully characterize the Tm3+ ion inside BaYF, but to evaluate the possibility to use such a system as a laser active material. For this purpose we limited our measurements only to the 3F4 manifold. We recorded the room temperature fluorescence spectra for the 3F4 → 3H6 transition with respect to the polarization of the 2% and 8% doped samples.
The temporal decay-curve of the 3F4 manifold has been recorded as a function of the doping density. The recorded decay-time curves were single exponential with a value nearly constant for the different doping levels: 17.0±1.5 ms, giving no hint of concentration quenching; the result is comparable to those found in the literature, i.e., 18.3 ms  and 16 ms[8, 9].
In Fig. 2 we show the emission cross sections of the 3F4 → 3H6 transition, calculated by means of the β - τ method described in Ref. , for the a and b polarizations and for the emission perpendicular to a and b: we need to use three different polarizations since the BaYF is a monoclinic crystal. The peak value is 2.5 · 10-21 cm2 at the wavelength of 1918.8 nm with E ∥ a; this value is not in agreement with those of ref.  neither for the peak position (1858 nm), nor for the value (1.1·10-21 cm2), but the discrepancy can be probably be ascribed to the fact that in that paper the emission-cross section was calculated for a non-oriented sample.
4. Laser experiments
The first attempt was performed on the 2% doped crystal using two different orientations:
We used the notation b–c and b–a to indicate the way the samples were cut and oriented. So b–a means that the crystal was cut as a parallelepiped in which the a and b axis were parallel to two sides (i.e., they lie on the input face and the pump and emitted laser beam can be polarized with E parallel to either the two axes). In both cases the polarization of the pump laser was parallel to the b-axis. In the first case the laser emission was polarized parallel to the b-axis with a maximum output power of 40 mW at a lasing wavelength of 1938 nm, while in the second case we obtained 62 mW polarized along the a-axis at a lasing wavelength of 1932 nm. Once observed that the best results were achieved with b–a cut crystals, we chose this orientation for all the other crystals and for every doping level we studied the output power as a function of the thickness of the sample in order to optimize the performance of the laser. A summary of the results obtained with the different samples is shown in Table 1: for the best samples at different doping densities we report the value of the slope efficiency calculated with respect to the incident power (ηinc) and absorbed power (ηabs); in this case the efficiency of the 12% crystal is as high as 61%, a value even higher than 59% obtained in Tm:YAG of Ref.  using a 5% as output coupler. It may be worth noticing the values of the threshold power, that in the best condition is as low as 26 mW: such a low value, obtained with a laser diode as pumping source, demonstrates the quality and the low losses of our crystals. The output characteristics for the higher doped samples are plotted in Fig. 3.
In Fig. 4 we show the lasing emission as a function of the wavelength for the 5%, 8%, 12% and 18% crystals (the 2% is not showed because it overlaps to the 5% spectrum). According to the energy level scheme of Tm:BaYF  the longest wavelength electronic emission is 1956 nm, even if it is not clear if it is active in the polarizations under investigation. Our fluorescence spectra indicate a maximum emission wavelength at 1927 nm in E ∥ b polarization, therefore we divided Fig. 4 into two parts: electronic and vibronic. As you can see in Table 1, we observed, two different conditions for the polarization states of the laser emission. In fact the laser emission in the samples with Tm concentration of 2% and 5% was polarized and showed a single peak emission, but for the other samples the emission was not polarized and above all the emission was spread over a wide wavelength range. We interpreted these results with the fact that the laser emission of the 2% and 5% samples could be ascribed to a purely electronic transition, while the laser emission of the 8%, 12% and 18% crystals, being extended to wavelength region were no electronic transition can be found, could be explained by the vibronic shift mechanism described in Ref. .
Another hint of the vibronic character of the emission is the fact that the emission is not polarized. In fact as pure electronic transitions are subject to the selection rules of their dipole operator, they give rise to polarized emission according to the point group symmetry of the rare earth site. On the other hand if the transition is vibronic the tensor operator is no more the “pure” dipole one, but takes into account the phononic contribution that is free from the symmetry rules of the point group symmetry. Therefore in this case it can result in unpolarized emission, as in our experimental results.
In our experimental conditions the vibronic shift could be estimated around 100 nm: this value was measured considering the last peak in the emission cross-section (1927 nm in E ∥ b spectrum of Fig. 2) and the longest wavelength tail of the laser spectrum of Fig. 4. This value is consistent with the results of the literature [2, 3, 11].
Since we report vibronic laser emission with a very low absorbed power, we can state that the samples were carefully grown resulting in high purity crystals with negligible scattering centers inside the material (whose presence would have been seriously detrimental to the vi-bronic effect). It is important to point out that our results are obtained with only 0.5% as output coupler: we expect to improve the laser performances using different output couplers in order to significantly increase the maximum output power and the slope efficiencies.
5. Conclusion and further developments
In conclusion, we have grown and spectroscopically analyzed several Tm:BaYF single crystals; we obtained cw laser operation demonstrating a maximum slope efficiency of 33% (61% with respect of the absorbed power) and a maximum output power of 290 mW. Moreover we obtained a very low threshold power (26 mW) using a laser diode as pumping source that has a very poor M2 in comparison to other systems like Ti:Al2O3. The vibronic laser emission indicates the high quality of our crystals and the negligible amount of scattering losses in the samples.
The other important result of our work can open the possibility to achieve emission in a wide wavelength range (1927-2030 nm) that nominate Tm:BaYF laser as a good pump for other solid-state laser (i.e., Ho-doped crystals) as well as for applications in surgery or LIDAR systems. Since the output characteristics shows no roll-off behavior we expect to obtain higher output power with a more powerful pumping source. Furthermore a cooling system and anti-reflection coatings could considerably improve the emission performances of our laser. All these improvements will be useful in order to obtain a significantly larger tunability range and higher output power (in Ref.  the authors obtained 162 nm in Tm:YLF pumping with two laser diodes 40 W each). Moreover a new design of the cavity could open the possibility to achieve TEM00 emission.
The authors wish to thank Ilaria Grassini for the preparation of the samples, H.P. Jenssen and A. Cassanho for helpful discussions.
References and links
1. R.J. Thrash and L.F. Johnson, “Uconversion laser emission from Yb3+-sensitized Tm3+ in BaY2F8,” J. Opt. Soc. Am. B 11, 881–885 (1994). [CrossRef]
2. L.F. Johnson and H.J. Guggenheim, “Electronic- and phonon-terminated laser emission from Ho3+ in BaY2F8,” IEEE J. Quantum Electron. 10, 442–449 (1974). [CrossRef]
4. A. Dergachev, K. Wall, and P.F. Moulton, “A CW side-pumped Tm:YLF laser,” in Advanced Solid State Lasers, M.E. Fermann and L.R. Marshall, eds., Vol. 68 of OSA Proceeding Series (Optical Society of America, Washington , D.C., 2002) pp. 343–346.
5. L.E. Batay, A.A. Demidovich, A.N. Kuzmin, A.N. Titov, M. Mond, and S. Kück, “Efficient tunable laser operation of diode-pumped Yb,Tm:KY(WO4)2 around 1.9 μm,” Appl. Phys. B 75, 457–461 (2002). [CrossRef]
6. A.E. Siegman, Lasers, (University Science, Mill Valley, Calif., 1982).
7. S.A. Payne, L.L. Chase, L.K. Smith, W.L. Kway, and W.F. Krupke, “Infrared cross-section measurements for crystals doped with Er3+, Tm3+ and Ho3+,” IEEE J. Quantum Electron. 28, 2619–2630 (1992). [CrossRef]
8. K. M. Dinndorf, “Energy transfer between Thulium and Holmium in laser hosts”, Ph.D. Dissertation MIT (1993).
9. M.A. Noginov, M. Curley, P. Venkateswarlu, A. Williams, and H.P. Jenssen, “Excitation scheme for the upper energy levels in a Tm:Yb:BaY2F8 laser crystal,” J. Opt. Soc. Am. B 14, 2126–2136 (1997). [CrossRef]
10. B.F. Aull and H.P. Jenssen, “Vibronic interactionsin Nd:YAG resulting in nonreciprocity of absorption and stimulated emission cross section,” IEEE J. Quantum Electron. 18, 925–930 (1982). [CrossRef]
11. I.F. Elder and M.J.P. Payne, “Lasing in dode-pumped Tm:YAP, Tm,Ho:YAP and Tm,Ho:YLF,” Opt. Commun. 145, 329–339 (1998). [CrossRef]