A comparative analysis of the thermal shift of Sm3+ ion zero-phonon lines in YAG and sesquioxide (Y2O3 and Sc2O3) ceramics is presented. The Sm3+ lines in YAG show small red shifts, whereas in sesquioxides large blue shifts in absorption (up to ~9 cm−1 Y2O3 or ~6 cm−1 Sc2O3), and blue or red shifts in emission are observed for the C2 centers, while the C3i magnetic-dipole allowed lines exhibit small red shift. The data are analyzed in terms of competition between the dynamic (due to electron-phonon interaction) and the static shifts produced especially by the thermal local structural changes.
© 2013 OSA
The Sm3+ (4f5) ion has a complex electronic level structure and its 4G5/2 level (Fig. 1) is able for efficient, long lived visible emission that could be used for luminescent phosphors, high pressure sensors or for lasers. The applications of this Sm3+ emission were for long time restricted by the absence of suitable pump sources and this limited the fundamental interest for this ion too; however, the recent advent of efficient violet diode lasers and of new types of materials, the transparent ceramics of garnets, such as YAG (Y3Al5O12), or cubic sesquioxides R2O3 (R = Y, Lu, Sc) that can be doped with Sm and tailored to specific applications could change the situation. Whereas in garnets the Sm3+ ions occupy a unique structural site of D2 symmetry, the sesquioxides offer for substitution two sites, both with six-fold O2- coordination, of C2 and C3i symmetry. In single crystals, several basic spectroscopic data on Sm3+:YAG , including high pressure effects  and Sm3+ in C2 centers of Y2O3  have been reported. Sm-doped YAG transparent ceramic was investigated for suppression of amplified spontaneous 1 µm emission of the Nd:YAG lasers at room temperature [4–6]. The extension of applications requires a detailed knowledge of the spectroscopic properties of the Sm-doped materials and of their variation with external conditions, such as temperature (T). Due to thermal expansion of the lattice, the T influences the static interactions that determine the positions of the spectral lines, as well as the interaction with the lattice vibrations (electron-phonon interaction) that induces the dynamic shift and broadening of lines.
The present paper aims the investigation of the thermal shift of the optical lines of Sm3+ in YAG and in both structural centers of Y2O3 and Sc2O3. Such comparative analysis could be relevant for the elucidation of the mechanisms by which the structure of the doping centers and their interaction with the phonons determine the thermal behavior of the optical lines.
YAG, Y2O3 and Sc2O3 polycrystalline ceramics doped with 1 wt.% Sm3+ were produced by solid–state reaction method, using high-purity powders (99.99 mass %) of sesquioxides, with particles size of 3-7 µm; after sintering at 1550°C, translucent ceramic samples were obtained. XRD investigation confirmed the single-phase cubic structure of these materials, and the scanning electron microscopy evidenced well developed fairly uniform ceramic grains, without intergrain phase separation. The absorption spectra were recorded at different temperatures with a closed cycle He refrigerator, in a setup consisting of tungsten halogen lamp, a Jarell-Ash monochromator, photomultipliers, Ge photodiodes and a Lock-in SR830 amplifier on line with a computer, and emission spectra were excited with a xenon lamp.
3. Spectroscopic data on Sm3+ in YAG and sesquioxides
3.1 Thermal shift of Sm3+ zero-phonon lines in YAG
A partial energy level scheme of Sm3+, based on data of Sm3+ in YAG  is illustrated in Fig. 1(a); J manifolds are split in (J + 1/2) Stark levels in D2 symmetry. The IR lower sextets 6HJ, 6FJ are separated by ~6700cm−1 from the metastable level 4G5/2 (17601 cm−1) and the upper group contains many closely-spaced levels in visible and ultraviolet . The absorption into the levels from the superior group is weak, except for those around 405 nm, which can be used for excitation of visible emission. Sm3+ presents intense visible emission corresponding to 4G5/2→6H5/2,7/2,9/2 transitions (the most intense emission 4G5/2→6H7/2 is in orange-red) with lifetime at low concentrations close to the radiative lifetime, in the range of ms.
The 10 K near IR absorption spectra of Sm3+ (1wt.%) in YAG ceramics show well separated and intense zero-phonon 6H5/2→6F7/2,9/2 absorption lines, that provide accurate data for T dependence; by heating to 300K these lines present very weak, transition-dependent (~0 to 2 cm−1) red shift, as illustrated in Fig. 1(b) and 1(c). All lines broaden strongly, from 2 to 4 cm−1 at 10 K to 8-10 cm−1 at 300K, and the peak intensities decrease due to this broadening and to the reduction of the Boltzmann population coefficient of the 6H5/2(1) level. Small (~1.5 cm−1) red shift in this T range is observed for the 4G5/2→6H7/2 visible emission (Fig. 1(d)), with the most intense lines Y1(16185 cm−1) and Y2(16230 cm−1) corresponding to the 4G5/2(1)→6H7/2(1,2) transitions, accompanied by broadening, from 20 to 43 cm−1 for Y1 and from 17 to 41 cm−1 for Y2.
3.2 Temperature dependences of Sm3+ lines in Y2O3 and Sc2O3
Although earlier studies on Sm:Y2O3  assign the low temperature absorption spectra to the dominant (75%) center of C2 symmetry, it was recently shown  that some of the transitions with ΔJ = 0,1, that are magnetic-dipole allowed (with relevant oscillator strengths ), contain lines that belong to both C2 and C3i centers. The 6H5/2→6F7/2,9/2 absorption spectra, associated to C2 centers, show pronounced blue shifts of 7-9 cm−1 as the T increases from 10 to 300K, as shown in Fig. 2(a) or Table 1. The T shift relative to 10 K, ΔE(T) = E(T)-E(10K) for several lines is presented in Fig. 2(b): up to ~50K it is negligible, then the dependence bends and for T>100 K it increases almost linearly with T, with similar rates, + 0.04 cm−1/°K for 6H5/2(1)→6F9/2(1) and + 0.04 cm−1/°K, + 0.03 cm−1/°K for 6H5/2(1)→6F7/2(1,4) transitions. The FWHM for different 6H5/2→6F7/2,9/2 lines varies from 3 to 4 to 15-18 cm−1 from 10 to 300K.
The three most intense lines in the visible 6H5/2(1)→4G5/2 absorption at 10K (Fig. 2(c)) have been earlier attributed to the C2 centers . It was subsequently demonstrated , by a time-resolved quasi-selective technique, that the line denoted by C3i (17698 cm−1) corresponds to 6H5/2(1)→4G5/2(1) transition of C3i centers. The 300 K emission on this line at low concentrations has lifetime of ~8.4ms, larger than for lines attributed to C2 centers, 1.48 ms. The T effects on the visible absorption lines exhibit different behavior, the line C2 in 6H5/2(1)→4G5/2(1) absorption (17560 cm−1) (Fig. 2(c)) shifts to blue with ~6 cm−1 from 10 to 300 K, while the line C3i presents red shift of ~3.5 cm−1; the 6H5/2→4F3/2 absorption lines present blue shift of ~9 cm−1. Again for T>100 K the shifts of C2 lines exhibit almost linear T dependence with slope + 0.024 cm−1/°K for 6H5/2(1)→4G5/2(1), and + 0.035 cm−1/°K for 6H5/2(1)→4F3/2(2) lines; the red shift of C3i 6H5/2→4G5/2 line presents nonlinear T dependence.
The T shift of the 4G5/2→6HJ (J = 5/2-9/2) of Sm:Y2O3 C2 emission lines, excited with a cw lamp filtered around ~405 nm (in a dominant electric-dipole transition 6H5/2→6P3/2), is also transition dependent (Fig. 2(d)): only the 4G5/2(1)→6H5/2(1) line shows blue T shift (~6 cm−1), while all the 4G5/2(1)→6H5/2(2,3) and 4G5/2(1)→6H7/2,9/2) lines show small (0-2.5 cm−1) red shifts. This indicates that the involved Stark levels of the 6H7/2, 6H9/2 manifolds as well as 6H5/2(2,3) levels are shifted to blue with similar (or slightly larger) quantities than the emitting 4G5/2(1) level (Table 1). The blue shift of C2 4G5/2(1)→6H5/2(1) emission could be connected to the fact that we took the ground level H5/2(1) as the zero reference point at all T. The emission lines broaden, from 6 to 22 cm−1 between 10 and 300 K for 4G5/2(1)→6H7/2(1).
The low temperature spectra of Sm3+ in Sc2O3 ceramics  resemble those in Y2O3, but the energy levels are shifted and the J global splitting generally increases for C2 centers (Table 1), from Y2O3 to Sc2O3 and for C3i centers too. However, in the 10-300 K range, the 6H5/2→6F7/2,9/2 absorption lines shifts to blue (up to ~6 cm−1 in Sc2O3) are smaller than those in Y2O3. The T behavior of 6H5/2→6G5/2 absorption in Sc2O3 is similar to that of Sm3+ in Y2O3, but the line assigned to C2 centers (17454 cm−1) shift to blue with only ~4.5 cm−1, while the C3i centers line (17621 cm−1) shows a red shift of only ~2 cm−1. If the 4G5/2(1)→6H5/2(1) emission is shifted to blue with ~4.5 cm−1, other emission lines present very small red shifts.
The thermal shift of the optical lines is determined by the individual shifts of the energy levels involved in the transition, caused by the simultaneous effect of temperature on the static (due to thermal expansion) and dynamic (electron-phonon) interactions of the doping ion, . Most of the works attributed the observed T shift exclusively to the electron-phonon interaction and it was argued that although the direct one-phonon and Raman two-phonon processes could be of significance, the last process involving acoustic phonons dominates. For Debye distribution of acoustic phonons one has :Eq. (1) can fit the experimental red T shift; however, the inferred TD was lower (~500-600 K for YAG and ~460 K for Y2O3) than that determined by other methods or than calculated with the cut-off energy of the global phonon spectra, confirming thus the dominant role of the acoustic phonons (specific heat for example). The EP coupling can also describe the thermal broadening of the lines with two parameters and TD, but in many cases TD values differ from those estimated from T shifts. A major shortcoming of the Raman two-phonon T shift theory is the impossibility to account for the observed blue [11–13] or no-shift of several lines and this imposed reconsideration of the static T shift [12–14].
The thermal lattice expansion could change the geometry (bond lengths and angles) of the doping center (especially for low symmetry), which could alter the static interactions of the ion with the host. In a thermodynamic model [12–14] the T dependence of the static shift can be characterized by the differential shift and the global shift is . Physically, the thermal expansion of the lattice is the opposite of its contraction under isostatic compression and the differential T shifts could be then linked to the isothermal differential pressure shift, , where is the isothermal volume compressibility of the center and is the bulk thermal expansion coefficient, proportional to the heat capacity CV. Since can be measured independently, this relation opens the way to separate the static and dynamic effects in the thermal shift. By using static shifts evaluated from the isostatic compression data , the complex T shift behavior of Nd:YAG , with red-, blue- or no-shifted lines could be explained consistently  by the specific competition between the static and dynamic effects for each line.
In the Debye model the T dependence of CV and thus that of the static T shift is similar to that for the two-phonon Raman process. The T shift under the joint action of the static and dynamic effects can be thus described  by an equation similar to Eq. (1), with replaced by , where the parameter A is related to the static contribution and is the true dynamic shift parameter, different from the parameter entering in Eq. (1). For T close or larger than TD this equation would predict quasi-linear T dependence of the thermal shift, , as observed for Sm3+ in sesquioxides. The compensation of the opposite-sign blue static shift (A) and red dynamic shift (-) in could explain consistently the observed red (A<), blue (A>), or no (A) shift; obviously>. The static effects contribution to red shifted lines of Cr3+ in YAG is not negligible too .
In case of Sm:YAG the weak thermal shifts of the absorption and emission lines described above contrast with the quite large pressure red shifts reported for the Y1 (−8.03 cm−1/GPa) and Y2 (−6.67 cm−1/GPa) emission lines , making this material suitable for pressure sensing. By using these pressure shift coefficients, the calculated for these lines are 35 × 10−3 cm−1K−1 and respectively 29.5 × 10−3 cm−1K−1; then, by using the observed red shift the calculated are 42.5 × 10−3 cm−1K−1 and 37 × 10−3 cm−1K−1, i.e. A = 0.82 and A = 0.8, respectively. Compensation of the strong static T effects indicates for Sm:YAG the electron-phonon interaction could be quite strong too; this is supported by the thermal line broadening.
Unfortunately, the absence of pressure shift data of any RE3+ ion in sesquioxides precludes a similar analysis to that for Sm3+ in YAG. However, in the discussion of the complex host, center and transition dependent thermal shift of Sm3+ in sesquioxides, several particular characteristics of these materials could be of relevance. In sesquioxides the phonon energies are determined by the vibrations of the RO6 groups, the maximum phonon energy is largest for Sc2O3 (~672 cm−1), while for Lu2O3 (~618 cm−1) it is only slightly higher than in Y2O3 (~597 cm−1), despite of the large differences in atomic mass of R3+ in these materials (44.96 for Sc, 174.47 for Lu and 88.91 for Y); this suggests that the phonon energies in sesquioxides are controlled by the cation-anion distance rather than by the mass of the cation. Similar tendency could be expected to hold for the acoustic phonons involved in the T line shift too; this could determine stronger dynamic thermal shift in Sc2O3 than in Lu2O3 or Y2O3. The static effects determined by the thermal expansion could be expected to follow the trends manifested in the electronic structure of Sm3+ in Sc2O3 compared with Y2O3 at low temperatures (Table 1): lower positions of the emitting manifolds (particularly 4G5/2) for both centers and stronger crystal field effects on the positions of lines, determining transition-dependent blue shift of the optical lines. In both materials the crystal field effects are larger for the C3i  than for the C2 centers, although their T variation is considerably weaker.
The observed blue thermal shift of the absorption lines of the C2 center of Sm3+ in both Y2O3 and Sc2O3 could then result from the clear dominance of the blue static shift over the red dynamic shift. The difference from YAG could be linked particularly with the weaker dynamic effects and stronger low-symmetry crystal field in sesquioxides, while the weaker shift in Sc2O3 relative to Y2O3 could be determined by the increased relative contribution of the dynamic effects due to the higher phonon energies and to stronger EP coupling. On the same line of reasoning, the observed red shift of the lines of the C3i center could indicate the dominance of the dynamic effects over the weaker static thermal shift effects compared with the C2 centers as suggested by comparison of the crystal field effects in Sc2O3 and Y2O3.
A very complex system (host, structural center, electronic transition) dependent red-, blue- or no-thermal shift of the optical absorption and emission lines of Sm3+ in the D2 sites of YAG and in the C2 or C3i symmetry sites in cubic sesquioxide (Y2O3, Sc2O3) ceramics from 10 K to 300 K was observed. This behavior was explained consistently by the balance specific to each case between the static effects determined by the modification of static interactions (nephelauxetic effect, crystal field interaction) due to thermal expansion of the lattice (structural center modifications) and the dynamic effects due to electron-phonon interaction.
This work was supported by CNCSIS – UEFISCSU, project PNII –PCE-IDEI 35/2011, Romania.
References and links
1. J. B. Gruber, Z. Bahram, and M. F. Reid, “Spectra, energy levels, and transition line strengths for Sm3+:Y3Al5O12,” Phys. Rev. 60(23), 15643–15653 (1999). [CrossRef]
2. Y. Zhao, W. Barvosa-Carter, S. D. Theiss, S. Mitha, M. J. Aziz, and D. Schiferl, “Pressure measurement at high temperature using ten Sm:YAG fluorescence peaks,” J. Appl. Phys. 84(8), 4049–4059 (1998). [CrossRef]
3. N. C. Chang, J. B. Gruber, R. P. Leavitt, and C. A. Morrison, “Optical spectra, energy levels, and crystal field analysis of tripositive rare earth ions in Y2O3. I. Kramers ions in C2 sites,” J. Chem. Phys. 76(8), 3877–3889 (1982). [CrossRef]
4. H. Yagi, J. F. Bisson, K. Ueda, and T. Yanagitani, “Y3Al5O12 ceramic absorbers for the suppression of parasitic oscillation in high-power Nd:YAG lasers,” J. Lumin. 121(1), 88–94 (2006). [CrossRef]
5. A. Lupei, V. Lupei, C. Gheorghe, and A. Ikesue, “Spectroscopic investigation of Sm3+ in YAG ceramic,” Rom. Rep. Physics 6(3), 817–822 (2011).
6. R. Huß, R. Wilhelm, C. Kolleck, J. Neumann, and D. Kracht, “Suppression of parasitic oscillations in a core-doped ceramic Nd:YAG laser by Sm:YAG cladding,” Opt. Express 18(12), 13094–13101 (2010). [CrossRef]
7. A. Lupei, C. Tiseanu, C. Gheorghe, and F. Voicu, “Optical spectroscopy of Sm3+in C2 and C3i sites of Y2O3 ceramics,” Appl. Phys. B 108(4), 909–918 (2012). [CrossRef]
8. C. M. Dobson and R. Zia, “Magnetic dipole and electric quadrupole transitions in the trivalent lanthanide series:Calculated emission rates and oscillator strengths,” Phys. Rev. B 86(12), 125102 (2012). [CrossRef]
9. C. Gheorghe, A. Lupei, F. Voicu, and M. Enculescu, “Sm3+-doped Sc2O3 polycrystalline ceramics: Spectroscopic investigation,” J. Alloy. Comp. 535, 78–82 (2012). [CrossRef]
10. D. E. McCumber and M. D. Sturge, “Linewidth and Temperature Shift of the R Lines in Ruby,” J. Appl. Phys. 34(6), 1682–1684 (1963). [CrossRef]
11. T. Kushida, “Linewidths and thermal shifts of spectral lines in neodimium-doped Yttium Aluminium Garnet and Calcium Fluorophosphate,” Phys. Rev. 185(2), 500–508 (1969). [CrossRef]
12. Th. Sesselmann, W. Richter, D. Haarer, and H. Morawitz, “Spectroscopic studies of impurity-host interactions in dye-doped po]ymers:Hydrostatic-pressure effects versus temperature effects,” Phys. Rev. B 36(14), 7601–7611 (1987). [CrossRef]
13. A. Kuznetsov, A. Laisaar, and J. Kikas, “Temperature dependence of spectral positions and widths of 5DJ-7FJ fluorescence lines originating from Sm2+ ions in SrFCl crystals,” Opt. Mater. 32(12), 1671–1675 (2010). [CrossRef]
14. W. C. Zheng, P. Su, H. G. Liu, and G. Y. Feng, “Relative importance of static contribution to the thermal shifts of spectral lines in Nd3+-doped Y3Al5O12 laser crystals,” J. Phys. D Appl. Phys. 45(34), 345305 (2012). [CrossRef]
15. S. Kobyakov, A. Kaminska, A. Suchocki, D. Galanciak, and M. Malinowski, “Nd3+-doped yttrium aluminum garnet crystal as a near-infrared pressure sensor for diamond anvil cells,” Appl. Phys. Lett. 88(23), 234102 (2006). [CrossRef]
16. W. C. Zheng, B. X. Li, and G. Y. Feng, “Thermal shifts and electron–phonon coupling parameters of the R-lines for Cr3+ion in Y3Al5O12 crystal,” Opt. Mater. 35(3), 626–628 (2013). [CrossRef]
17. A. Lupei, C. Tiseanu, C. Gheorghe, ”Electronic structure and energy transfer processes of Sm3+ in sesquioxides,” presented at ICOM 2012, Belgrad, Serbia, 3–6 Sept, 2012.