Spectroscopic and optical properties of the rare earth-doped liquid PBr3/AlBr3/SbBr3
Introduction
Recently, much research has been directed at hosts which when doped with rare earths lead to efficient emission in the near infrared, up to wavelengths of about 5 μm. In general, such hosts have almost invariably been solid state. However, we have determined the spectroscopic properties of the trivalent lanthanide ions of Pr, Nd, Sm, Dy, Ho, Er, Tm and Yb as dopants in the liquid PBr3/AlBr3/SbBr3. In the past, extensive research was conducted into inorganic liquid laser systems, in particular trivalent neodymium salts in aprotic solvents such as phosphorous oxychloride (POCl3) or selenium oxychloride (SeOCl2) acidified with a Lewis acid such as tin tetrachloride (SnCl4) or zirconium tetrachloride (ZrCl4) 1, 2, 3, 4, 5. Previous research has demonstrated that systems such as Nd3+:POCl3/SnCl4 yield good 1.06 μm lasers. However, they do possess several unattractive features. The toxic and corrosive nature of the solvents, their high Raman cross sections and thermal expansion and convection problems all led to more manageable solid-state hosts being adopted. These problems are particularly acute in large flashlamp-pumped systems of the type originally envisaged. However, more modern techniques such as diode pumping and waveguide confinement, using only minute amounts of material offer new routes to lasers based around such systems.
The most recent studies have considered spectral intensities and transition probabilities given by Judd–Ofelt analysis of a POCl3/SnCl4 solvent and energy transfer between rare earth codopants in this system.
In 1976 neodymium dissolved as an anhydrous salt in PBr3/AlBr3/SbBr3 was investigated [6]. This solvent possesses a lower vibrational (phonon) energy spectrum than the more common, oxygen-containing solvents. This is expected to decrease the probability of non-radiative decay compared to that in the POCl3 or SeOCl2 systems. This immediately presents the possibility of laser action in the mid infrared region, an area not obtainable when using POCl3 as a host, due to quenching of such transitions by non-radiative decay processes. PBr3 is thus a reasonable choice as a host for such a laser as it has a highest fundamental frequency at 381 cm−1, compared to that in POCl3 which occurs at 1292 cm−1. A survey of potential host solvents showed it to be by far the most promising [7]. It is transparent over a wide range 800–26 000 cm−1, and can be prepared into optical quality solutions by filtering. However to realise such potential benefits purity is of paramount importance, and great care must be taken to ensure that the samples used are free from impurities with vibrational frequencies significantly in excess of those of PBr3. POBr3 was found to be a major contaminant of standard, reagent grade PBr3, and so was bought in a high purity form. This was shown spectroscopically to be free of any P–O stretch absorption prior to the preparation of the solutions. Fig. 1 shows the infrared absorption spectrum of the PBr3, taken prior to solution preparation. Overlayed is the spectrum of a sample containing POBr3, the absence of the P–O stretch can be noted. For POBr3, the vibrational bands of interest are the v1 (340 cm−1), v2 (488 cm−1), v4 (1261 cm−1) modes. The band at 760 cm−1 is an overtone of the fundamental P–Br stretch, explaining its appearance in both the pure and impure samples.
Absorption and fluorescence spectra were used in order to determine the energy level structure of the trivalent lanthanide-doped liquid. A comprehensive assignment of 4f levels has been given to the aforementioned rare earths in a PBr3/AlBr3/SbBr3 host. Absorption spectra of the visible, near infrared (NIR) and mid infrared (MIR) spectral regions and fluorescence spectra in the visible and NIR were taken. The 4f energy levels were determined from these spectra. The absorption spectra were also used to determine oscillator strengths between certain levels and ground state and to determine the linewidths, both of which were required in the Judd–Ofelt analysis.
Section snippets
Preparation of the solutions
Originally the base chemicals were purified from standard spectroscopic grade samples. However recently ultra pure PBr3 (99.99%), AlBr3 (99.99%) and SbBr3 (99.999%) have been obtained.
The doped liquids were prepared by mixing together the three powders of the rare earth, aluminium and antimony tribromides in the approximate molar ratio 1:30:30. The mix was stirred for 1 h. The PBr3 was then added with a molar ratio of about 20:1 with the Lewis acid and stirred at room temperature for several
Analysis
The oscillator strength of a transition between levels “aJ and bJ” is given bywhere the constants have their usual meanings and N and d are the rare earth concentration (ions per cm3) and the path length, respectively. A(E) is the absorbance as a function of energy, E being in cm−1. As first published by Judd and Ofelt 8, 9 the theoretical expression for an oscillator strength between levels aJ and bJ′ is shown in Eq. (2).
Results
Table 2Table 3Table 4Table 5Table 6Table 7Table 8 give a complete spectroscopic analysis of rare earth-doped PBr3/AlBr3/SbBr3. A comprehensive assignment of 4f energy levels is given as well as calculated and measured oscillator strengths and linewidths. Judd–Ofelt parameters for most of the lanthanides studied have been evaluated and are also given. ΔE is the FWHM linewidth, Irel are the relative peak intensities of the lines. The oscillator strengths were measured as described, by integrating
Discussion
The fluorescence spectra observed when the liquids were excited, are shown in Fig. 2, Fig. 3, Fig. 4. The fluorescences exhibit linewidths that are narrower than is typical in a solid state amorphous host. This is suggestive of there being less site broadening in the liquid host than in a host such as a glass. It is likely that the lanthanide ion is co-ordinated by a well-defined solvation shell, and that all ions experience a similar local environment, this would indeed be the case if some
Conclusion
Spectroscopic properties of trivalent rare earth ions in PBr3/AlBr3/SbBr3 and their 4f energy levels have been determined, oscillator strengths calculated and Judd–Ofelt parameters evaluated for all but one of the lanthanides studied. It has been shown that this solvent behaves as a low phonon energy host, with non-radiative relaxation rates that are extremely small compared to the radiative component. Although such hosts do present problems associated with their handling and instability, they
Acknowledgements
The authors wish to thank the DERA for initial support of this project, Dr. W.S. Brocklesby for initial assistance with fluorescence measurements and Dr. A.P. Coleman for early work associated with the preparation of the solutions.
References (12)
J. Molec. Spectrosc.
(1968)- et al.
Optical Materials
(1996) Appl. Phys. Letts.
(1966)- et al.
Characteristics of the Nd+3 SeOCl2 liquid laser
Appl. Phys. Letts.
(1966) - et al.
IEEE J. Quant. Electron.
(1972) - et al.
J. Opt. Soc. Am.
(1968)
Cited by (2)
Spectroscopy of Pr<sup>3+</sup> ions in lithium borate and lithium fluoroborate glasses
2001, Physica B: Condensed Matter