- Email: [email protected]
Properties of ytterbium and neodymium doped alkali metal yttrium double phosphates of the M3Y1−xLnx(PO4)2 type
Journal of Alloys and Compounds 380 (2004) 405–412
Properties of ytterbium and neodymium doped alkali metal yttrium double phosphates of the M3Y1−x Lnx (PO4 )2 type T. Aitasalo a,b , M. Guzik c , W. Szuszkiewicz d , J. Hölsä a , B. Keller c , J. Legendziewicz c,∗ a
d
Department of Chemistry, University of Turku, FI-20014 Turku, Finland b Graduate School of Materials Research, Turku, Finland c Faculty of Chemistry, University of Wrocław, 14 Joliot-Curie, PL-50-383 Wrocław, Poland Institute of Chemistry, Economy Academia, 118/120 Komandorska, PL-53-345 Wrocław, Poland
Abstract The spectroscopic behaviour of the Nd3+ and Yb3+ doped alkaline metal yttrium double phosphates, M3 Y1−x Lnx (PO4 )2 (M = Na, Rb; x = 0.01–0.3) were studied for both powder and single crystal samples. The high resolution absorption and emission spectra were measured in the visible and IR regions. Spectral changes with the Nd3+ and Yb3+ concentration were interpreted. The absorption strengths of the 4f–4f transitions were analysed and used to assess the structural modifications of the two double phosphates. Based on the 4 K absorption spectra the number of metal sites occupied by the dopants was investigated. Strong emission from Na3 Y1−x Ndx (PO4 )2 involving the 4 F3/2 → 4 I9/2 , 4 I11/2 , 4 I13/2 , 4 I15/2 transitions were observed whereas the corresponding emission from the rubidium phosphate was presumably quenched by multiphonon processes due to the water molecules absorbed in the channel-like structure. The IR spectra were used to assign the vibronic components of the electronic transitions. The Yb3+ emission bands were broadened depending on the Yb3+ concentration (1–10 mol%). The tentative energy level scheme of the ground and excited 2 FJ (J = 7/2, 5/2) levels was described. © 2004 Elsevier B.V. All rights reserved. Keywords: Neodymium; Ytterbium; Absorption; Emission; Yttrium double phosphates
1. Introduction The structural and spectroscopic properties of the alkali metal lanthanide double phosphates, M3 Y(PO4 )2 , doped with different lanthanide ions have been subject to considerable interest already for a long time [1–11]. These double phosphates have their host absorption edge at a rather short wavelength, and are thus suitable as host lattices for various luminescent materials including laser devices. The double phosphate lattice co-doped by different lanthanide ions can also exhibit up-conversion emission in a wide spectral range from blue to red. From this point of view the Yb3+ ion with co-doping by other Ln3+ ions is the most promising system [12]. The diode pumped solid state (tunable) laser (DPSSL) systems based on the Yb3+ doped materials are studied in∗ Corresponding author. Tel.: +48-71-375-7300; fax: +48-71-328-2348. E-mail address: [email protected] (J. Legendziewicz).
0925-8388/$ – see front matter © 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.jallcom.2004.03.066
tensively at the moment [13]. The potential host lattices studied for this promising pseudo three level lasing system include simple or complex oxide, garnet, borate, silicate and tungstate systems [14–16]. In addition to these lattices, the double phosphates may be of interest. The Yb3+ ion can create laser action in the IR range around 1000 nm and has several advantages when compared to the Nd3+ doped systems already established as several laser materials. No cross-relaxation processes exists in the Yb3+ system and the Yb3+ absorption broadened usually by extensive vibronic side bands is suitable for efficient IR diode pumping. The structures of the alkali metal lanthanide double phosphates, M3 Ln(PO4 )2 (M = Na, Rb), have been studied by several authors [3,8–11]. These double phosphates crystallise in three forms depending on the alkali metal and the ionic radius of the lanthanide ion. The sodium phosphates crystallise in the orthorhombic system (space group Pbc21 , #29, Z = 24) [8,9] and the rubidium phosphates as hexagonal (space group P3m, #164, Z = 1) from Dy to Lu and Y
406
T. Aitasalo et al. / Journal of Alloys and Compounds 380 (2004) 405–412
or monoclinic (space group P21 /m, #11, Z = 2) from La up to Tb [3,10,11]. The sodium double phosphates possess six lanthanide sites of low symmetry whereas both forms of the rubidium double phosphates have only one lanthanide site each, of low symmetry in the monoclinic form and of high symmetry in the hexagonal one [10]. The knowledge on the spectroscopic properties of the yttrium double phosphates, M3 Y(PO4 )2 , doped with lanthanide ions is rather limited but recently a study on mainly the Eu3+ doped M3 Y(PO4 )2 (M = Na, Rb) has been reported [17]. In the present paper the results of the spectroscopic studies on the Yb3+ doped yttrium double phosphates are given. Since the neodymium wide gap compounds are interesting as scintillators and laser materials both in single crystal and nanometric forms the optical studies were extended to Nd3+ doped materials, too.
2. Experimental 2.1. Sample preparation The Nd3+ and Yb3+ doped yttrium double phosphates, M3 Y(PO4 )2 :Ln3+ , (M = Na, Rb; Ln3+ = Nd3+ , Yb3+ ) were synthesised with a solid state reaction between the intimately ground stoichiometric amounts of the mixed lanthanide phosphate hydrate Y1−x Lnx (PO4 )2 ·zH2 O (x = 0.01–0.3) and M3 PO4 (M = Na, Rb) at 900 ◦ C in air for 12 h [17]. The post-reaction heat treatment was carried out at 1400 and 1250 ◦ C for the sodium and rubidium phosphates, respectively. The ICP determination of M+ , Ln3+ and P amounts in the products was performed using an ARL 3410 spectrometer. The X-ray diffraction of the powder samples measured using a HZG-4 diffractometer equipped with a Guinier camera and Cu K␣ radiation confirmed the formation of the appropriate double phosphate phases. The rubidium salt was found moisture sensitive and had to be handled in dry conditions. All operations were done in a dry box. 2.2. Spectroscopic measurements The absorption spectra were measured with a Cary–Varian 500 scan spectrophotometer equipped with an Oxford CF1204 helium flow cryostat between 4 and 293 K in the 200–1500 nm spectral region. Thin disks prepared under pressure (0.7 MPa) were used to detect the 4f–4f transitions of the Nd3+ and Yb3+ ions. The oscillator strengths of these transitions were calculated by integration of the Gauss–Lorentz shaped bands by using the ICH-10 program [18,19]. The high resolution emission spectra were measured at 4, 77 and 293 K with a Spectra-Pro 750 monochromator equipped with a Hamamatsu R 928 photomultiplier tube as a detector with a resolution of 0.1 cm−1 . A lock-in amplifier was used for the IR emission detection. Nd:YAG pulsed laser (λexc = 532 nm) and wide laser diode lines
were employed as the excitation sources. A liquid nitrogen cryostat was used to detect the spectra at 77 K. The IR spectra were recorded with a Bruker FS 88 FT-IR spectrometer with 0.1 cm−1 resolution. The Raman spectra were obtained using a Nicolet Magna 860 FT-IR spectrometer.
3. Results and discussion 3.1. Nd3+ doped sodium and rubidium double phosphates, M3 Y(PO4 )2 :Nd3+ It was reported earlier [17] that the syntheses of the yttrium double phosphates resulted in the formation of mainly the double phosphate phase only. This was confirmed presently both by the ICP-AES and the X-ray diffraction studies. In all spectroscopic studies the yttrium double phosphate (M3 Y(PO4 )2 ) was used as a matrix with Nd3+ and Yb3+ as the doping ions. At high temperatures used in the present work, sodium yttrium double phosphates with a complex orthorhombic structure [8,9] is formed, whereas the structure of the rubidium double phosphates should be of the high symmetry hexagonal type. With reference to the previous study [17], the Eu3+ ion may occupy two or even more sites in the sodium yttrium double phosphate lattice, whereas in the rubidium lattice the symmetry of the yttrium site seems to be much higher. However, the pure electronic 4f–4f transitions are then accompanied with transitions presumably of vibronic origin. The results for the praseodymium doped rubidium yttrium double phosphate [20] lead to similar conclusions. The absorption spectra of both the rubidium and sodium yttrium double phosphates at 4 and 293 K with 6 mol% Nd3+ doping level (Figs. 1 and 2) show 4f–4f transitions with rather complicated crystal field (CF) fine structure in the range of all transitions. Accordingly, it is concluded that the absorption from the 4 I9/2 level to the excited ones results from Nd3+ ions in at least two sites. Especially well this can be observed for the 4 I9/2 → 2 P1/2 transition where two well resolved lines at 23 264 and 23 231 cm−1 with a splitting of 33 cm−1 are recognised for this transition with a unsplit 2 P1/2 Kramers doublet. Previously [17] two components were also found but the line located at 432 nm belongs rather to a transition from a higher energy Stark component of the 4 I9/2 ground term of one Nd3+ site. Since the present spectra are recorded exactly at 4 K, the splitting determined for the energy separation for the 2 P1/2 levels of Nd3+ in the two sites is more certain and precise. Because of the low point symmetry of the lanthanide sites and the absence of transition selection rules between the CF components, the maximum number of lines for each transition from the 4 I9/2 level is given by the J multiplicity of the excited level, i.e. J + 1/2. One can clearly observe more than J + 1/2 lines for the 4 I9/2 → 2 H11/2 and 4 I9/2 → 4 G7/2 , 4 G5/2 transitions. For the latter group of lines, 14 well resolved lines are observed, which correspond well to two sites occupied by the
T. Aitasalo et al. / Journal of Alloys and Compounds 380 (2004) 405–412
407
Fig. 1. Absorption spectra of (a) Na3 Y0.94 Nd0.06 (PO4 )2 at 4 and 293 K and (b) Rb3 Y0.94 Nd0.06 (PO4 )2 at 4 K.
Nd3+ ions in the sodium double phosphate structure. The assignment of the transitions was made by using the Nd3+ energy level data for NdOF [21]. The character of the absorption spectra differs for the rubidium double phosphate for the same measurements conditions and with the same Nd3+ doping level. The 4I 2 9/2 → P1/2 transition presents a single line but other bands, mainly those corresponding to the hypersensitive 4I 4 9/2 → G5/2 transition are less clearly split into the CF components. The origin of the 4f–4f transition strengths is a little different as will be shown later. Table 1 presents the integrated strengths of the 4f–4f transitions for both the sodium and rubidium double phosphates together with the range of the Stark components for selected transitions. The areas of absorption bands were determined numerically by graphical integration and expressed in terms
of integrated strengths using the program described previously [18,19]. It was discovered that the strengths of the hypersensitive transitions change in an opposite way with decreasing temperature for the rubidium and sodium double phosphates. Moreover, with decreasing temperature, the strengths of all transitions were observed to vary in a different way for the sodium and rubidium double phosphates. Such untypical behaviour of the 4f–4f transition strengths in rubidium double phosphate seems to point out to the different character of these transitions. Note also that the total splitting of the majority of the bands in the absorption spectra of the rubidium double phosphate at 4 K is weaker. The larger total splitting of the bands for the sodium phosphate confirms that more than one lanthanide site in the structure is occupied by the dopants. The broadening of the bands is important in laser
Fig. 2. Details of the absorption spectra of (a) Na3 Y0.94 Nd0.06 (PO4 )2 at 4 and 293 K and (b) Rb3 Y0.94 Nd0.06 (PO4 )2 at 293 K (all values in cm−1 units unless otherwise noted).
408
T. Aitasalo et al. / Journal of Alloys and Compounds 380 (2004) 405–412
Table 1 Integrated strengths of the 4f–4f transitions in diluted neodymium double phosphates, M3 Y0.94 Nd0.06 (PO4 )2 4f–4f transition
Wavelength (nm)
Oscillator strengths (P × 108 )c Na3 Y0.94 Nd0.06 (PO4 )2
4F 4F
3/2
5/2 ,
4F
7/2 ,
4F
9/2
4G
5/2 ,
2K 2K 2P
2H 4S
9/2
3/2
2G
7/2
13/2 ,
4G 4 7/2 , G9/2 2G 2 4 , , 15/2 9/2 D3/2 , G11/2
1/2 4D 4 2 4 2 3/2 , D5/2 , I11/2 , D1/2 , L15/2 ,
920–840 840–772 772–715 710–656 610–560 550–493 492–448 445–425 370–346
application especially since the absorption cross-section for pumping is then increased. As for the emission spectra of the Nd3+ doped sodium and rubidium double phosphates, in general a strong concentration quenching of the neodymium emission is observed. Accordingly, in the emission experiments the yttrium double phosphates with low Nd3+ doping level (3 and 6 mol%) samples were used at 77 and 293 K (Figs. 3 and 4). Only weak emission was obtained from the rubidium double phosphate, presumably because of quenching by the multiphonon process via the O–H (stretching) vibrations of the water molecules absorbed in the open channel structure of the hexagonal rubidium double phosphates [10]. In contrast, recorded with the same conditions as for the rubidium double phosphates, the sodium double phos-
Rb3 Y0.94 Nd0.06 (PO4 )2
293 K
4K
293 K
4K
897 3278 3406 391 2299 (2%) 5298 9860 (10%) 3669 1162 259 5293
785 2941 2560 399
329 1589 1343 124
– 1598 1252 114
4707
3072
3150
3480 1439 227 7270
1315 351 107 1513
1190 444 82 1550
phates yielded excellent emission from the 4 F3/2 level to the four 4 I9/2 , 4 I11/2 , 4 I13/2 and 4 I15/2 multiplets (Fig. 3). The intensities of these lines correlate well with the values of the matrix elements for the respective transitions. In the high resolution emission spectra at 77 K the number of components correlates with J + 1/2 levels for one site, but the additional weak lines indicate the population of another site in the structure despite the low Nd3+ doping level (3 and 6 mol%). A drastic increase in the intensity of the 4 F3/2 → 4 I11/2 , 4I 9/2 emission lines was found at 77 K with decreasing Nd3+ ion concentration (Fig. 4). However, a reverse relation was observed at 293 K, because most probably the cross-relaxation is phonon assisted and thus temperature controlled. The present results prove those obtained for
Fig. 3. Emission spectra of Na3 Y0.94 Nd0.06 (PO4 )2 at 77 and 293 K and Rb3 Y0.94 Nd0.06 (PO4 )2 at 4 K with Nd:YAG laser excitation at 532 nm.
T. Aitasalo et al. / Journal of Alloys and Compounds 380 (2004) 405–412
Fig. 4. Emission spectra of Na3 Y0.94 Nd0.06 (PO4 )2 and Na3 Y0.97 Nd0.03 (PO4 )2 at 77 and 293 K with Nd:YAG laser excitation at 532 nm.
409
Fig. 5. Absorption spectra of Na3 Y0.8 Yb0.2 (PO4 )2 and Rb3 Y0.85 Yb0.15 (PO4 )2 at 4 and 293 K.
K3 Nd(PO4 )2 [5] where the decay rate of the emission was strongly concentration dependent. 3.2. Yb3+ doped sodium and rubidium double phosphates, M3 Y(PO4 )2 :Yb3+ As stated above, considerable research activity has been directed toward the Yb3+ doped materials aimed to new laser applications. The ytterbium doped sodium and rubidium yttrium double phosphates may offer efficient hosts for such laser materials, too. For diode pumped laser applications strong absorption of the pumping radiation is required and thus a high absorption cross-section should be beneficial. The low temperature absorption spectra for the Rb3 Y1−x Ybx (PO4 )2 and Na3 Y1−x Ybx (PO4 )2 disks with 15 and 20 mol% concentration of the Yb3+ ion (Figs. 5 and 6) show the typical 2 F7/2 → 2 F5/2 transition of the Yb3+ ion at around 1000 nm. The double phosphates with lower Yb3+ concentrations could not be used in the absorption spectra measurements since the disks became non-transparent when thick and spectra could not be recorded. The bands in the absorption spectra are shifted to shorter wavelengths (from 930 to 905 nm) for phosphates in comparison to ytterbium fluoride, LiLuF4 :Yb3+ [12], despite the strength of the spin-orbit coupling varies only very little from one compound to another. The blue shift is due to the larger splitting of the 2 F7/2 ground level because of the strong
Fig. 6. Absorption spectra of Na3 Y0.8 Yb0.2 (PO4 )2 at 4, 150 and 293 K.
410
T. Aitasalo et al. / Journal of Alloys and Compounds 380 (2004) 405–412
crystal field effect. The crystal field is usually stronger in the oxide materials when compared to the fluoride hosts [22]. This is confirmed by the much larger splitting of the excited 2 F5/2 level, 793 against 420 cm−1 in LiLuF4 :Yb3+ . An important fact worth noticing is that the absorption spectra for both the rubidium and sodium double phosphates are almost identical at 293 K whereas more complex spectra were observed at 4 K. The 293 K spectra suggest occupation of only one site of very similar character in both rubidium and sodium double phosphates, but this conclusion is in contradiction to both the situation observed for the Nd3+ , Pr3+ and Eu3+ doped double phosphates [17,20] and the rather different structures, lanthanide site symmetries and lanthanide-oxygen coordination in these two double phosphates [3,9,10]. It seems reasonable to suppose that due to the high Yb3+ doping levels, the main part of the ytterbium has segregated to its own phase which, moreover, is the same for both the rubidium and sodium double phosphates. This excludes the compositions such as the simple Na3 Yb(PO4 )2 and Rb3 Yb(PO4 )2 phases which should have different structures though no specific structural data is available for the pure ytterbium compounds. In any case, from the absorption spectra at 4 and 293 K one can conclude that only one site is present in the compound yielding the main bands in the absorption spectra since three electronic components of the 2F 5/2 excited level can be easily observed at 10 245, 10 525 and 11 038 cm−1 with a total CF splitting of 793 cm−1 . As for Na3 Y1−x Ybx (PO4 )2 , the absorption spectra at 4 K show that, most probably, up to 20 mol% one site is preferred by the Yb3+ ion in the yttrium double phosphate lattice. However, the close similarity between the complex structure of the weak components at 4 K for both types of the double phosphates imposes some uncertainty to this interpretation, too. Additional components observed in the absorption spectra can also be a result of strong electron–phonon coupling which is very typical for the Yb3+ doped materials and/or by cooperative ion pair components. The superposition of the absorption spectra at 4 K with the 293 K IR spectra (Fig. 7) indicates that some vibronic components are indeed present corresponding to the IR active vibrations at around 150, 350 and maybe also at around 700 cm−1 . However, the majority of weaker components seems to be of other origin. At the present stage of the studies it is rather difficult to observe the electron–phonon coupling in the absorption spectra with the Raman active modes but this coupling seems to prefer the IR active vibrations. As for the CF splitting of the 2 F7/2 ground level, the temperature dependence of absorption spectra measured from 4 to 293 K (Fig. 6) allows to elucidate three weak components at about 987 (10 131); 1005 (9950) and 1030 nm (9708 cm−1 ) for the 2 F7/2 → 2 F5/2 transition in Na3 Y0.8 Yb0.2 (PO4 )2 . Assuming that these lines correspond to the hot band absorption to the lowest CF component of the 2 F5/2 excited level at 10 245 cm−1 the partial CF splitting of the 2 F7/2 ground level could be determined to be
Fig. 7. Emission spectra (diode laser excitation at 990.1 nm) at 293 K and absorption spectra of Na3 Y0.8 Yb0.2 (PO4 )2 at 4 and 293 K with superposition of IR spectra.
as follows: 0, 114, 295 and 537 cm−1 . The total splitting of 537 cm−1 seems to be too low in comparison with that observed in LiLuF4 :Yb3+ (486 cm−1 ). However, in order to estimate the 2 F7/2 ground level splitting a method presented earlier [23] can be used. This method basically relates the effects of the spin-orbit coupling and the crystal field with the position of the 2 F5/2 excited level and the total crystal field splitting of the 2 F7/2 ground level. The relationship between the total CF splitting of the 2 F7/2 and 2 F5/2 levels predicts for the splitting of the former level to be about 950 cm−1 when the splitting of the latter is 793 cm−1 . It is thus very improbable that hot band absorption from the highest CF component of the 2 F7/2 level at 950 cm−1 would be observed even at 293 K. Moreover, the weak lines observed in the hot band absorption spectra might well result from transitions to the excited CF components of the excited 2F 5/2 level. From the relationship between the barycenters of the 2 F7/2 and 2 F5/2 levels, the barycenter for the 2 F7/2 ground level is predicted to be at about 450 cm−1 when that of the 2 F5/2 level is at 10 603 cm−1 . Thus, the partial energy level scheme for the 2 F7/2 level can be deduced to be as follows: 0, 537 and 950 cm−1 . With the barycenter at 450 cm−1 , the position of the missing level is estimated
T. Aitasalo et al. / Journal of Alloys and Compounds 380 (2004) 405–412
411
a function of temperature for Na3 Y0.8 Yb0.2 (PO4 )2 and Na3 Y0.85 Yb0.15 (PO4 )2 . For the sodium double phosphate a decrease in temperature leads to a decrease in the intensities from 4366 at 293 K to 2039 at 4 K, whereas for rubidium phosphate almost no change was observed in accordance with the results discussed for Nd3+ in Section 3.1. and the earlier data reported for the appropriate europium double phosphate [11].
4. Summary
Fig. 8. Emission spectra of Na3 Y0.9 Yb0.1 (PO4 )2 and Rb3 Y0.9 Yb0.1 (PO4 )2 at 293 K with laser diode excitation (990.1 nm).
at 313 cm−1 which is in a rather good agreement with the observed value of 295 cm−1 . The complete determination of the CF splitting of the 2F 7/2 ground level will be possible only by measuring the Yb3+ emission spectra at 4 K by using a titanium sapphire laser as an excitation source. At the moment, only the emission spectra detected at 293 K by using wide band laser diode excitation (Fig. 8) and a spectrum excited by a dye laser at 293 K are available. Based on these emission spectra one can find strong reabsorption at 10 256 cm−1 for a rubidium double phosphate with 10 mol% Yb3+ concentration. A broadening of the bands was observed with increasing Yb3+ concentration, confirming a disorder in the structure mainly in rubidium double phosphate, which is in accordance with the structure possessing open channels leading to a strong tendency to absorb water into these channels. The disorder is less evident in the emission of sodium double phosphate with 10 mol% Yb3+ . The similarity between emission spectra irrespective of the Yb3+ doping level and the alkali metal confirms the hypothesis that Yb3+ occupies one center in the yttrium lattice. The unusual behaviour of the rubidium open channel lattice is manifested again in the changes of intensities with decreasing temperature. Table 2 collects the integrated strengths of the 2 F7/2 → 2 F5/2 transition as Table 2 The temperature dependence of the oscillator strength of the 2F 2F 3+ in diluted double phosphates, 7/2 → 5/2 transition for Yb M3 Y1−x Ybx (PO4 )2 Temperature (K)
293 150 75 35 4
Oscillator strength (P × 108 )c Na3 Y0.80 Yb0.20 (PO4 )2
Rb3 Y0.85 Yb0.15 (PO4 )2
4366 4017 3559 2660 2039
644 615 618 608 623
The spectroscopic properties of the Nd3+ and Yb3+ doped alkaline metal yttrium double phosphates of the M3 Y1−x Lnx (PO4 )2 (M = Rb, Na; Ln = Nd3+ , Yb3+ ) type were investigated. Structural differences were found between the Nd3+ doped sodium and rubidium double phosphates. They are manifested in the presence of two 4I 2 9/2 → P1/2 transitions at 4 K and by the intensities of 4I 4 4 9/2 → G5/2 , G7/2 transitions. The similarities in the low temperature absorption spectra of the sodium and rubidium double phosphates doped by Yb3+ indicate complex composition of the samples with possible phase segregation. The tendency of the rubidium double phosphate structure to adsorb water molecules into the open channels is manifested in a long range disorder in the structure and in the nature of the 4f–4f transition strengths. The splitting and strength of the 2 F7/2 → 2 F5/2 transition of Yb3+ in both types of double phosphates vary only slightly. The studies of the spectra as a function of the concentration of the active ions and the analysis of the vibronic components based on the IR spectra help in the elucidation of the spectral lines. The electronic transition probabilities were calculated. The temperature dependence of the strengths of the 4f–4f transitions in both Nd3+ and Yb3+ doped double phosphates points out to the different nature of the transitions in the spectra of rubidium and sodium salts and/or a transformation of the structure of the rubidium double phosphate. The number of lines detected in the Nd3+ absorption spectra (mainly for the 4 I9/2 → 2 P1/2 transition) at 4 K indicates the occupation of two lanthanide sites in the structure of sodium double phosphates. These materials would be interesting for different luminescence applications if samples of good quality could be synthesised.
Acknowledgements Financial support from the Academy of Finland and KBN (J.H. and J.L.), and the European Union Marie Curie program (T.A. and J.L.) is gratefully acknowledged. The authors would like to thank professor W. Ryba-Romanowski (Institute of Low Temperature and Structure Research, Polish Academy of Science, Wrocław) for measurement of the Yb3+ emission spectra.
412
T. Aitasalo et al. / Journal of Alloys and Compounds 380 (2004) 405–412
References [1] B. Finke, L. Schwarz, P. Gürtler, M. Kraas, M. Joppien, S. Becker, J. Lumin. 60–61 (1994) 975. [2] L. Schwarz, B. Finke, M. Kloss, A. Rohmann, U. Sasum, D. Haberland, J. Lumin. 72–74 (1997) 257. [3] P.P. Melnikov, V.B. Kalinin, V.A. Efremov, L.N. Komisarowa, Iz. Akad. Nauk SSR Neorg. Mater. 17 (1981) 1452. [4] C.E. Bamberger, P.R. Robinson, R.L. Sherman, Inorg. Chim. Acta 34 (1979) L203. [5] H.Y.-P. Hong, S.R. Chinn, Mater. Res. Bull. 11 (1976) 421 (and references cited therein). [6] J.Z. Letho, P. Houenou, R. Eholie, C. R. Acad. Sci. (Paris) Sér. II 307 (1988) 1177. [7] M. Kloss, L. Schwarz, J. Hölsä, Acta Phys. Pol. A 95 (1999) 343. [8] Z.J. Chao, C. Parent, G. Le Flem, P. Hagenmuller, J. Solid State Chem. 82 (1989) 255. [9] R. Salmon, C. Parent, M. Vlasse, G. Le Flem, Mater. Res. Bull. 13 (1978) 439. [10] S. Areva, J. Hölsä, M. Kloss, M. Lahtinen, R.-J. Lamminmäki, M. Lastusaari, L. Schwarz, J. Valkonen, Mater. Sci. Forum 378–381 (2001) 644. [11] K. Okada, J. Ossaka, Acta Cryst. B 36 (1980) 919.
[12] A. Bensalah, Y. Guyot, H. Sato, T. Fukuda, G. Boulon, J. Alloys Comp., in press. [13] W.F. Krupke, IEEE J. Sel. Topics Quant. Electr. 6 (2000) 1287. [14] F. Druon, S. Chenais, P. Raybaut, F. Balembois, P. Georges, R. Gaumé, G. Aka, B. Viana, S. Mohr, D. Kopf, Opt. Lett. 27 (2002) 197. [15] P.-H. Haumesser, R. Gaumé, B. Viana, G. Aka, D. Vivien, B. Ferrand, Trends Opt. Photon. 50 (2001) 425. [16] P.-H. Haumesser, R. Gaumé, B. Viana, E. Antic-Fidancev, D. Vivien, J. Phys.: Condens. Mat. 13 (2001) 5427. [17] W. Szuszkiewicz, B. Keller, M. Guzik, T. Aitasalo, J. Niittykoski, J. Hölsä, J. Legendziewicz, J. Alloys Comp. 341 (2002) 297. [18] J. Legendziewicz, T. Głowiak, G. Oczko, D.C. Ngoan, J. Less-Common Met. 125 (1986) 45. [19] J. Legendziwicz, Wiad. Chem. 42 (1988) 605. [20] M. Guzik, T. Aitasalo, W. Szuszkiewicz, J. Hölsä, B. Keller, J. Legendziewicz, J. Alloys Comp. (2004) in press. [21] J. Hölsä, E. Säilynoja, P. Ylhä, E. Antic-Fidancev, M. Lemaˆıtre-Blaise, P. Porcher, J. Chem. Soc., Faraday Trans. 94 (1998) 481. [22] E. Antic-Fidancev, J. Hölsä, M. Lastusaari, A. Lupei, Phys. Rev. B 64 (2001) 195108. [23] E. Antic-Fidancev, J. Hölsä, M. Lastusaari, J. Phys. Condens. Mat. 15 (2003) 863.
Properties of ytterbium and neodymium doped alkali metal yttrium double phosphates of the M3Y1−x Lnx (PO4 )2 type T. Aitasalo a,b , M. Guzik c , W. Szuszkiewicz d , J. Hölsä a , B. Keller c , J. Legendziewicz c,∗ a
d
Department of Chemistry, University of Turku, FI-20014 Turku, Finland b Graduate School of Materials Research, Turku, Finland c Faculty of Chemistry, University of Wrocław, 14 Joliot-Curie, PL-50-383 Wrocław, Poland Institute of Chemistry, Economy Academia, 118/120 Komandorska, PL-53-345 Wrocław, Poland
Abstract The spectroscopic behaviour of the Nd3+ and Yb3+ doped alkaline metal yttrium double phosphates, M3 Y1−x Lnx (PO4 )2 (M = Na, Rb; x = 0.01–0.3) were studied for both powder and single crystal samples. The high resolution absorption and emission spectra were measured in the visible and IR regions. Spectral changes with the Nd3+ and Yb3+ concentration were interpreted. The absorption strengths of the 4f–4f transitions were analysed and used to assess the structural modifications of the two double phosphates. Based on the 4 K absorption spectra the number of metal sites occupied by the dopants was investigated. Strong emission from Na3 Y1−x Ndx (PO4 )2 involving the 4 F3/2 → 4 I9/2 , 4 I11/2 , 4 I13/2 , 4 I15/2 transitions were observed whereas the corresponding emission from the rubidium phosphate was presumably quenched by multiphonon processes due to the water molecules absorbed in the channel-like structure. The IR spectra were used to assign the vibronic components of the electronic transitions. The Yb3+ emission bands were broadened depending on the Yb3+ concentration (1–10 mol%). The tentative energy level scheme of the ground and excited 2 FJ (J = 7/2, 5/2) levels was described. © 2004 Elsevier B.V. All rights reserved. Keywords: Neodymium; Ytterbium; Absorption; Emission; Yttrium double phosphates
1. Introduction The structural and spectroscopic properties of the alkali metal lanthanide double phosphates, M3 Y(PO4 )2 , doped with different lanthanide ions have been subject to considerable interest already for a long time [1–11]. These double phosphates have their host absorption edge at a rather short wavelength, and are thus suitable as host lattices for various luminescent materials including laser devices. The double phosphate lattice co-doped by different lanthanide ions can also exhibit up-conversion emission in a wide spectral range from blue to red. From this point of view the Yb3+ ion with co-doping by other Ln3+ ions is the most promising system [12]. The diode pumped solid state (tunable) laser (DPSSL) systems based on the Yb3+ doped materials are studied in∗ Corresponding author. Tel.: +48-71-375-7300; fax: +48-71-328-2348. E-mail address: [email protected] (J. Legendziewicz).
0925-8388/$ – see front matter © 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.jallcom.2004.03.066
tensively at the moment [13]. The potential host lattices studied for this promising pseudo three level lasing system include simple or complex oxide, garnet, borate, silicate and tungstate systems [14–16]. In addition to these lattices, the double phosphates may be of interest. The Yb3+ ion can create laser action in the IR range around 1000 nm and has several advantages when compared to the Nd3+ doped systems already established as several laser materials. No cross-relaxation processes exists in the Yb3+ system and the Yb3+ absorption broadened usually by extensive vibronic side bands is suitable for efficient IR diode pumping. The structures of the alkali metal lanthanide double phosphates, M3 Ln(PO4 )2 (M = Na, Rb), have been studied by several authors [3,8–11]. These double phosphates crystallise in three forms depending on the alkali metal and the ionic radius of the lanthanide ion. The sodium phosphates crystallise in the orthorhombic system (space group Pbc21 , #29, Z = 24) [8,9] and the rubidium phosphates as hexagonal (space group P3m, #164, Z = 1) from Dy to Lu and Y
406
T. Aitasalo et al. / Journal of Alloys and Compounds 380 (2004) 405–412
or monoclinic (space group P21 /m, #11, Z = 2) from La up to Tb [3,10,11]. The sodium double phosphates possess six lanthanide sites of low symmetry whereas both forms of the rubidium double phosphates have only one lanthanide site each, of low symmetry in the monoclinic form and of high symmetry in the hexagonal one [10]. The knowledge on the spectroscopic properties of the yttrium double phosphates, M3 Y(PO4 )2 , doped with lanthanide ions is rather limited but recently a study on mainly the Eu3+ doped M3 Y(PO4 )2 (M = Na, Rb) has been reported [17]. In the present paper the results of the spectroscopic studies on the Yb3+ doped yttrium double phosphates are given. Since the neodymium wide gap compounds are interesting as scintillators and laser materials both in single crystal and nanometric forms the optical studies were extended to Nd3+ doped materials, too.
2. Experimental 2.1. Sample preparation The Nd3+ and Yb3+ doped yttrium double phosphates, M3 Y(PO4 )2 :Ln3+ , (M = Na, Rb; Ln3+ = Nd3+ , Yb3+ ) were synthesised with a solid state reaction between the intimately ground stoichiometric amounts of the mixed lanthanide phosphate hydrate Y1−x Lnx (PO4 )2 ·zH2 O (x = 0.01–0.3) and M3 PO4 (M = Na, Rb) at 900 ◦ C in air for 12 h [17]. The post-reaction heat treatment was carried out at 1400 and 1250 ◦ C for the sodium and rubidium phosphates, respectively. The ICP determination of M+ , Ln3+ and P amounts in the products was performed using an ARL 3410 spectrometer. The X-ray diffraction of the powder samples measured using a HZG-4 diffractometer equipped with a Guinier camera and Cu K␣ radiation confirmed the formation of the appropriate double phosphate phases. The rubidium salt was found moisture sensitive and had to be handled in dry conditions. All operations were done in a dry box. 2.2. Spectroscopic measurements The absorption spectra were measured with a Cary–Varian 500 scan spectrophotometer equipped with an Oxford CF1204 helium flow cryostat between 4 and 293 K in the 200–1500 nm spectral region. Thin disks prepared under pressure (0.7 MPa) were used to detect the 4f–4f transitions of the Nd3+ and Yb3+ ions. The oscillator strengths of these transitions were calculated by integration of the Gauss–Lorentz shaped bands by using the ICH-10 program [18,19]. The high resolution emission spectra were measured at 4, 77 and 293 K with a Spectra-Pro 750 monochromator equipped with a Hamamatsu R 928 photomultiplier tube as a detector with a resolution of 0.1 cm−1 . A lock-in amplifier was used for the IR emission detection. Nd:YAG pulsed laser (λexc = 532 nm) and wide laser diode lines
were employed as the excitation sources. A liquid nitrogen cryostat was used to detect the spectra at 77 K. The IR spectra were recorded with a Bruker FS 88 FT-IR spectrometer with 0.1 cm−1 resolution. The Raman spectra were obtained using a Nicolet Magna 860 FT-IR spectrometer.
3. Results and discussion 3.1. Nd3+ doped sodium and rubidium double phosphates, M3 Y(PO4 )2 :Nd3+ It was reported earlier [17] that the syntheses of the yttrium double phosphates resulted in the formation of mainly the double phosphate phase only. This was confirmed presently both by the ICP-AES and the X-ray diffraction studies. In all spectroscopic studies the yttrium double phosphate (M3 Y(PO4 )2 ) was used as a matrix with Nd3+ and Yb3+ as the doping ions. At high temperatures used in the present work, sodium yttrium double phosphates with a complex orthorhombic structure [8,9] is formed, whereas the structure of the rubidium double phosphates should be of the high symmetry hexagonal type. With reference to the previous study [17], the Eu3+ ion may occupy two or even more sites in the sodium yttrium double phosphate lattice, whereas in the rubidium lattice the symmetry of the yttrium site seems to be much higher. However, the pure electronic 4f–4f transitions are then accompanied with transitions presumably of vibronic origin. The results for the praseodymium doped rubidium yttrium double phosphate [20] lead to similar conclusions. The absorption spectra of both the rubidium and sodium yttrium double phosphates at 4 and 293 K with 6 mol% Nd3+ doping level (Figs. 1 and 2) show 4f–4f transitions with rather complicated crystal field (CF) fine structure in the range of all transitions. Accordingly, it is concluded that the absorption from the 4 I9/2 level to the excited ones results from Nd3+ ions in at least two sites. Especially well this can be observed for the 4 I9/2 → 2 P1/2 transition where two well resolved lines at 23 264 and 23 231 cm−1 with a splitting of 33 cm−1 are recognised for this transition with a unsplit 2 P1/2 Kramers doublet. Previously [17] two components were also found but the line located at 432 nm belongs rather to a transition from a higher energy Stark component of the 4 I9/2 ground term of one Nd3+ site. Since the present spectra are recorded exactly at 4 K, the splitting determined for the energy separation for the 2 P1/2 levels of Nd3+ in the two sites is more certain and precise. Because of the low point symmetry of the lanthanide sites and the absence of transition selection rules between the CF components, the maximum number of lines for each transition from the 4 I9/2 level is given by the J multiplicity of the excited level, i.e. J + 1/2. One can clearly observe more than J + 1/2 lines for the 4 I9/2 → 2 H11/2 and 4 I9/2 → 4 G7/2 , 4 G5/2 transitions. For the latter group of lines, 14 well resolved lines are observed, which correspond well to two sites occupied by the
T. Aitasalo et al. / Journal of Alloys and Compounds 380 (2004) 405–412
407
Fig. 1. Absorption spectra of (a) Na3 Y0.94 Nd0.06 (PO4 )2 at 4 and 293 K and (b) Rb3 Y0.94 Nd0.06 (PO4 )2 at 4 K.
Nd3+ ions in the sodium double phosphate structure. The assignment of the transitions was made by using the Nd3+ energy level data for NdOF [21]. The character of the absorption spectra differs for the rubidium double phosphate for the same measurements conditions and with the same Nd3+ doping level. The 4I 2 9/2 → P1/2 transition presents a single line but other bands, mainly those corresponding to the hypersensitive 4I 4 9/2 → G5/2 transition are less clearly split into the CF components. The origin of the 4f–4f transition strengths is a little different as will be shown later. Table 1 presents the integrated strengths of the 4f–4f transitions for both the sodium and rubidium double phosphates together with the range of the Stark components for selected transitions. The areas of absorption bands were determined numerically by graphical integration and expressed in terms
of integrated strengths using the program described previously [18,19]. It was discovered that the strengths of the hypersensitive transitions change in an opposite way with decreasing temperature for the rubidium and sodium double phosphates. Moreover, with decreasing temperature, the strengths of all transitions were observed to vary in a different way for the sodium and rubidium double phosphates. Such untypical behaviour of the 4f–4f transition strengths in rubidium double phosphate seems to point out to the different character of these transitions. Note also that the total splitting of the majority of the bands in the absorption spectra of the rubidium double phosphate at 4 K is weaker. The larger total splitting of the bands for the sodium phosphate confirms that more than one lanthanide site in the structure is occupied by the dopants. The broadening of the bands is important in laser
Fig. 2. Details of the absorption spectra of (a) Na3 Y0.94 Nd0.06 (PO4 )2 at 4 and 293 K and (b) Rb3 Y0.94 Nd0.06 (PO4 )2 at 293 K (all values in cm−1 units unless otherwise noted).
408
T. Aitasalo et al. / Journal of Alloys and Compounds 380 (2004) 405–412
Table 1 Integrated strengths of the 4f–4f transitions in diluted neodymium double phosphates, M3 Y0.94 Nd0.06 (PO4 )2 4f–4f transition
Wavelength (nm)
Oscillator strengths (P × 108 )c Na3 Y0.94 Nd0.06 (PO4 )2
4F 4F
3/2
5/2 ,
4F
7/2 ,
4F
9/2
4G
5/2 ,
2K 2K 2P
2H 4S
9/2
3/2
2G
7/2
13/2 ,
4G 4 7/2 , G9/2 2G 2 4 , , 15/2 9/2 D3/2 , G11/2
1/2 4D 4 2 4 2 3/2 , D5/2 , I11/2 , D1/2 , L15/2 ,
920–840 840–772 772–715 710–656 610–560 550–493 492–448 445–425 370–346
application especially since the absorption cross-section for pumping is then increased. As for the emission spectra of the Nd3+ doped sodium and rubidium double phosphates, in general a strong concentration quenching of the neodymium emission is observed. Accordingly, in the emission experiments the yttrium double phosphates with low Nd3+ doping level (3 and 6 mol%) samples were used at 77 and 293 K (Figs. 3 and 4). Only weak emission was obtained from the rubidium double phosphate, presumably because of quenching by the multiphonon process via the O–H (stretching) vibrations of the water molecules absorbed in the open channel structure of the hexagonal rubidium double phosphates [10]. In contrast, recorded with the same conditions as for the rubidium double phosphates, the sodium double phos-
Rb3 Y0.94 Nd0.06 (PO4 )2
293 K
4K
293 K
4K
897 3278 3406 391 2299 (2%) 5298 9860 (10%) 3669 1162 259 5293
785 2941 2560 399
329 1589 1343 124
– 1598 1252 114
4707
3072
3150
3480 1439 227 7270
1315 351 107 1513
1190 444 82 1550
phates yielded excellent emission from the 4 F3/2 level to the four 4 I9/2 , 4 I11/2 , 4 I13/2 and 4 I15/2 multiplets (Fig. 3). The intensities of these lines correlate well with the values of the matrix elements for the respective transitions. In the high resolution emission spectra at 77 K the number of components correlates with J + 1/2 levels for one site, but the additional weak lines indicate the population of another site in the structure despite the low Nd3+ doping level (3 and 6 mol%). A drastic increase in the intensity of the 4 F3/2 → 4 I11/2 , 4I 9/2 emission lines was found at 77 K with decreasing Nd3+ ion concentration (Fig. 4). However, a reverse relation was observed at 293 K, because most probably the cross-relaxation is phonon assisted and thus temperature controlled. The present results prove those obtained for
Fig. 3. Emission spectra of Na3 Y0.94 Nd0.06 (PO4 )2 at 77 and 293 K and Rb3 Y0.94 Nd0.06 (PO4 )2 at 4 K with Nd:YAG laser excitation at 532 nm.
T. Aitasalo et al. / Journal of Alloys and Compounds 380 (2004) 405–412
Fig. 4. Emission spectra of Na3 Y0.94 Nd0.06 (PO4 )2 and Na3 Y0.97 Nd0.03 (PO4 )2 at 77 and 293 K with Nd:YAG laser excitation at 532 nm.
409
Fig. 5. Absorption spectra of Na3 Y0.8 Yb0.2 (PO4 )2 and Rb3 Y0.85 Yb0.15 (PO4 )2 at 4 and 293 K.
K3 Nd(PO4 )2 [5] where the decay rate of the emission was strongly concentration dependent. 3.2. Yb3+ doped sodium and rubidium double phosphates, M3 Y(PO4 )2 :Yb3+ As stated above, considerable research activity has been directed toward the Yb3+ doped materials aimed to new laser applications. The ytterbium doped sodium and rubidium yttrium double phosphates may offer efficient hosts for such laser materials, too. For diode pumped laser applications strong absorption of the pumping radiation is required and thus a high absorption cross-section should be beneficial. The low temperature absorption spectra for the Rb3 Y1−x Ybx (PO4 )2 and Na3 Y1−x Ybx (PO4 )2 disks with 15 and 20 mol% concentration of the Yb3+ ion (Figs. 5 and 6) show the typical 2 F7/2 → 2 F5/2 transition of the Yb3+ ion at around 1000 nm. The double phosphates with lower Yb3+ concentrations could not be used in the absorption spectra measurements since the disks became non-transparent when thick and spectra could not be recorded. The bands in the absorption spectra are shifted to shorter wavelengths (from 930 to 905 nm) for phosphates in comparison to ytterbium fluoride, LiLuF4 :Yb3+ [12], despite the strength of the spin-orbit coupling varies only very little from one compound to another. The blue shift is due to the larger splitting of the 2 F7/2 ground level because of the strong
Fig. 6. Absorption spectra of Na3 Y0.8 Yb0.2 (PO4 )2 at 4, 150 and 293 K.
410
T. Aitasalo et al. / Journal of Alloys and Compounds 380 (2004) 405–412
crystal field effect. The crystal field is usually stronger in the oxide materials when compared to the fluoride hosts [22]. This is confirmed by the much larger splitting of the excited 2 F5/2 level, 793 against 420 cm−1 in LiLuF4 :Yb3+ . An important fact worth noticing is that the absorption spectra for both the rubidium and sodium double phosphates are almost identical at 293 K whereas more complex spectra were observed at 4 K. The 293 K spectra suggest occupation of only one site of very similar character in both rubidium and sodium double phosphates, but this conclusion is in contradiction to both the situation observed for the Nd3+ , Pr3+ and Eu3+ doped double phosphates [17,20] and the rather different structures, lanthanide site symmetries and lanthanide-oxygen coordination in these two double phosphates [3,9,10]. It seems reasonable to suppose that due to the high Yb3+ doping levels, the main part of the ytterbium has segregated to its own phase which, moreover, is the same for both the rubidium and sodium double phosphates. This excludes the compositions such as the simple Na3 Yb(PO4 )2 and Rb3 Yb(PO4 )2 phases which should have different structures though no specific structural data is available for the pure ytterbium compounds. In any case, from the absorption spectra at 4 and 293 K one can conclude that only one site is present in the compound yielding the main bands in the absorption spectra since three electronic components of the 2F 5/2 excited level can be easily observed at 10 245, 10 525 and 11 038 cm−1 with a total CF splitting of 793 cm−1 . As for Na3 Y1−x Ybx (PO4 )2 , the absorption spectra at 4 K show that, most probably, up to 20 mol% one site is preferred by the Yb3+ ion in the yttrium double phosphate lattice. However, the close similarity between the complex structure of the weak components at 4 K for both types of the double phosphates imposes some uncertainty to this interpretation, too. Additional components observed in the absorption spectra can also be a result of strong electron–phonon coupling which is very typical for the Yb3+ doped materials and/or by cooperative ion pair components. The superposition of the absorption spectra at 4 K with the 293 K IR spectra (Fig. 7) indicates that some vibronic components are indeed present corresponding to the IR active vibrations at around 150, 350 and maybe also at around 700 cm−1 . However, the majority of weaker components seems to be of other origin. At the present stage of the studies it is rather difficult to observe the electron–phonon coupling in the absorption spectra with the Raman active modes but this coupling seems to prefer the IR active vibrations. As for the CF splitting of the 2 F7/2 ground level, the temperature dependence of absorption spectra measured from 4 to 293 K (Fig. 6) allows to elucidate three weak components at about 987 (10 131); 1005 (9950) and 1030 nm (9708 cm−1 ) for the 2 F7/2 → 2 F5/2 transition in Na3 Y0.8 Yb0.2 (PO4 )2 . Assuming that these lines correspond to the hot band absorption to the lowest CF component of the 2 F5/2 excited level at 10 245 cm−1 the partial CF splitting of the 2 F7/2 ground level could be determined to be
Fig. 7. Emission spectra (diode laser excitation at 990.1 nm) at 293 K and absorption spectra of Na3 Y0.8 Yb0.2 (PO4 )2 at 4 and 293 K with superposition of IR spectra.
as follows: 0, 114, 295 and 537 cm−1 . The total splitting of 537 cm−1 seems to be too low in comparison with that observed in LiLuF4 :Yb3+ (486 cm−1 ). However, in order to estimate the 2 F7/2 ground level splitting a method presented earlier [23] can be used. This method basically relates the effects of the spin-orbit coupling and the crystal field with the position of the 2 F5/2 excited level and the total crystal field splitting of the 2 F7/2 ground level. The relationship between the total CF splitting of the 2 F7/2 and 2 F5/2 levels predicts for the splitting of the former level to be about 950 cm−1 when the splitting of the latter is 793 cm−1 . It is thus very improbable that hot band absorption from the highest CF component of the 2 F7/2 level at 950 cm−1 would be observed even at 293 K. Moreover, the weak lines observed in the hot band absorption spectra might well result from transitions to the excited CF components of the excited 2F 5/2 level. From the relationship between the barycenters of the 2 F7/2 and 2 F5/2 levels, the barycenter for the 2 F7/2 ground level is predicted to be at about 450 cm−1 when that of the 2 F5/2 level is at 10 603 cm−1 . Thus, the partial energy level scheme for the 2 F7/2 level can be deduced to be as follows: 0, 537 and 950 cm−1 . With the barycenter at 450 cm−1 , the position of the missing level is estimated
T. Aitasalo et al. / Journal of Alloys and Compounds 380 (2004) 405–412
411
a function of temperature for Na3 Y0.8 Yb0.2 (PO4 )2 and Na3 Y0.85 Yb0.15 (PO4 )2 . For the sodium double phosphate a decrease in temperature leads to a decrease in the intensities from 4366 at 293 K to 2039 at 4 K, whereas for rubidium phosphate almost no change was observed in accordance with the results discussed for Nd3+ in Section 3.1. and the earlier data reported for the appropriate europium double phosphate [11].
4. Summary
Fig. 8. Emission spectra of Na3 Y0.9 Yb0.1 (PO4 )2 and Rb3 Y0.9 Yb0.1 (PO4 )2 at 293 K with laser diode excitation (990.1 nm).
at 313 cm−1 which is in a rather good agreement with the observed value of 295 cm−1 . The complete determination of the CF splitting of the 2F 7/2 ground level will be possible only by measuring the Yb3+ emission spectra at 4 K by using a titanium sapphire laser as an excitation source. At the moment, only the emission spectra detected at 293 K by using wide band laser diode excitation (Fig. 8) and a spectrum excited by a dye laser at 293 K are available. Based on these emission spectra one can find strong reabsorption at 10 256 cm−1 for a rubidium double phosphate with 10 mol% Yb3+ concentration. A broadening of the bands was observed with increasing Yb3+ concentration, confirming a disorder in the structure mainly in rubidium double phosphate, which is in accordance with the structure possessing open channels leading to a strong tendency to absorb water into these channels. The disorder is less evident in the emission of sodium double phosphate with 10 mol% Yb3+ . The similarity between emission spectra irrespective of the Yb3+ doping level and the alkali metal confirms the hypothesis that Yb3+ occupies one center in the yttrium lattice. The unusual behaviour of the rubidium open channel lattice is manifested again in the changes of intensities with decreasing temperature. Table 2 collects the integrated strengths of the 2 F7/2 → 2 F5/2 transition as Table 2 The temperature dependence of the oscillator strength of the 2F 2F 3+ in diluted double phosphates, 7/2 → 5/2 transition for Yb M3 Y1−x Ybx (PO4 )2 Temperature (K)
293 150 75 35 4
Oscillator strength (P × 108 )c Na3 Y0.80 Yb0.20 (PO4 )2
Rb3 Y0.85 Yb0.15 (PO4 )2
4366 4017 3559 2660 2039
644 615 618 608 623
The spectroscopic properties of the Nd3+ and Yb3+ doped alkaline metal yttrium double phosphates of the M3 Y1−x Lnx (PO4 )2 (M = Rb, Na; Ln = Nd3+ , Yb3+ ) type were investigated. Structural differences were found between the Nd3+ doped sodium and rubidium double phosphates. They are manifested in the presence of two 4I 2 9/2 → P1/2 transitions at 4 K and by the intensities of 4I 4 4 9/2 → G5/2 , G7/2 transitions. The similarities in the low temperature absorption spectra of the sodium and rubidium double phosphates doped by Yb3+ indicate complex composition of the samples with possible phase segregation. The tendency of the rubidium double phosphate structure to adsorb water molecules into the open channels is manifested in a long range disorder in the structure and in the nature of the 4f–4f transition strengths. The splitting and strength of the 2 F7/2 → 2 F5/2 transition of Yb3+ in both types of double phosphates vary only slightly. The studies of the spectra as a function of the concentration of the active ions and the analysis of the vibronic components based on the IR spectra help in the elucidation of the spectral lines. The electronic transition probabilities were calculated. The temperature dependence of the strengths of the 4f–4f transitions in both Nd3+ and Yb3+ doped double phosphates points out to the different nature of the transitions in the spectra of rubidium and sodium salts and/or a transformation of the structure of the rubidium double phosphate. The number of lines detected in the Nd3+ absorption spectra (mainly for the 4 I9/2 → 2 P1/2 transition) at 4 K indicates the occupation of two lanthanide sites in the structure of sodium double phosphates. These materials would be interesting for different luminescence applications if samples of good quality could be synthesised.
Acknowledgements Financial support from the Academy of Finland and KBN (J.H. and J.L.), and the European Union Marie Curie program (T.A. and J.L.) is gratefully acknowledged. The authors would like to thank professor W. Ryba-Romanowski (Institute of Low Temperature and Structure Research, Polish Academy of Science, Wrocław) for measurement of the Yb3+ emission spectra.
412
T. Aitasalo et al. / Journal of Alloys and Compounds 380 (2004) 405–412
References [1] B. Finke, L. Schwarz, P. Gürtler, M. Kraas, M. Joppien, S. Becker, J. Lumin. 60–61 (1994) 975. [2] L. Schwarz, B. Finke, M. Kloss, A. Rohmann, U. Sasum, D. Haberland, J. Lumin. 72–74 (1997) 257. [3] P.P. Melnikov, V.B. Kalinin, V.A. Efremov, L.N. Komisarowa, Iz. Akad. Nauk SSR Neorg. Mater. 17 (1981) 1452. [4] C.E. Bamberger, P.R. Robinson, R.L. Sherman, Inorg. Chim. Acta 34 (1979) L203. [5] H.Y.-P. Hong, S.R. Chinn, Mater. Res. Bull. 11 (1976) 421 (and references cited therein). [6] J.Z. Letho, P. Houenou, R. Eholie, C. R. Acad. Sci. (Paris) Sér. II 307 (1988) 1177. [7] M. Kloss, L. Schwarz, J. Hölsä, Acta Phys. Pol. A 95 (1999) 343. [8] Z.J. Chao, C. Parent, G. Le Flem, P. Hagenmuller, J. Solid State Chem. 82 (1989) 255. [9] R. Salmon, C. Parent, M. Vlasse, G. Le Flem, Mater. Res. Bull. 13 (1978) 439. [10] S. Areva, J. Hölsä, M. Kloss, M. Lahtinen, R.-J. Lamminmäki, M. Lastusaari, L. Schwarz, J. Valkonen, Mater. Sci. Forum 378–381 (2001) 644. [11] K. Okada, J. Ossaka, Acta Cryst. B 36 (1980) 919.
[12] A. Bensalah, Y. Guyot, H. Sato, T. Fukuda, G. Boulon, J. Alloys Comp., in press. [13] W.F. Krupke, IEEE J. Sel. Topics Quant. Electr. 6 (2000) 1287. [14] F. Druon, S. Chenais, P. Raybaut, F. Balembois, P. Georges, R. Gaumé, G. Aka, B. Viana, S. Mohr, D. Kopf, Opt. Lett. 27 (2002) 197. [15] P.-H. Haumesser, R. Gaumé, B. Viana, G. Aka, D. Vivien, B. Ferrand, Trends Opt. Photon. 50 (2001) 425. [16] P.-H. Haumesser, R. Gaumé, B. Viana, E. Antic-Fidancev, D. Vivien, J. Phys.: Condens. Mat. 13 (2001) 5427. [17] W. Szuszkiewicz, B. Keller, M. Guzik, T. Aitasalo, J. Niittykoski, J. Hölsä, J. Legendziewicz, J. Alloys Comp. 341 (2002) 297. [18] J. Legendziewicz, T. Głowiak, G. Oczko, D.C. Ngoan, J. Less-Common Met. 125 (1986) 45. [19] J. Legendziwicz, Wiad. Chem. 42 (1988) 605. [20] M. Guzik, T. Aitasalo, W. Szuszkiewicz, J. Hölsä, B. Keller, J. Legendziewicz, J. Alloys Comp. (2004) in press. [21] J. Hölsä, E. Säilynoja, P. Ylhä, E. Antic-Fidancev, M. Lemaˆıtre-Blaise, P. Porcher, J. Chem. Soc., Faraday Trans. 94 (1998) 481. [22] E. Antic-Fidancev, J. Hölsä, M. Lastusaari, A. Lupei, Phys. Rev. B 64 (2001) 195108. [23] E. Antic-Fidancev, J. Hölsä, M. Lastusaari, J. Phys. Condens. Mat. 15 (2003) 863.