Joel Nunes1*, Mark Farries1, David Furniss1, Zhuoqi Tang1, Richard Crane1, Dinuka Jayasuriya1, Lukasz Sojka2, Slawomir Sujecki1,2, Trevor M. Benson1 and Angela B. Seddon1.
1Mid-Infrared Photonics Group, George Green Institute for Electromagnetics Research, University of Nottingham, UK.
2Telecommunications & Teleinformatics Department, Wroclaw University of Technology, Wroclaw, Poland.
Many chemicals and biological materials exhibit distinct spectroscopic interactions with mid-infrared (MIR) electromagnetic (EM) waves between 2 to 50 µm  especially in the ‘spectral fingerprint’ region from 6.66 to 20 µm (1500 to 500 cm-1). Advancing MIR light sources to efficiently probe these interactions can unlock answers to critical challenges across the fields of medical, chemical, and environmental sensing. Chalcogenide glasses are vitreous materials based on some of the elements from Group-16 of the Periodic Table. They are suitable host materials for trivalent lanthanide dopants, facilitating their photoexcitation by relatively low cost Near-Infrared (NIR) laser sources, while preferentially promoting the resultant MIR photoluminescent (PL) emissions due to their low phonon energies . Lanthanide doped, chalcogenide-glass fibres are promising candidates for realising bright, inexpensive, tuneable, and coherent MIR light sources beyond 4 µm , which is a spectral region that currently has a deficiency in MIR light sources.
Despite past research into these doped materials, challenges in their fabrication, characterisation and numerical modelling persist, and prevent MIR fibre lasers from being fully realized. The aim of this endeavour is toward the advancement of attaining MIR lasing in chalcogenide glass fibres, by developing bespoke optical characterisation methods. To achieve this the MIR absorption and PL emission behaviour of praseodymium ‘Pr3+’, and terbium ‘Tb3+’ ion dopants in Ge-As-Se-Ga host chalcogenide glass composition were investigated. This was done to obtain accurate empirical data to feed back into the glass fabrication process, and to gain insight into the spectroscopic behaviour of the lanthanide dopant and host glass for new photoexcitation schemes, and improved numerical models. The ground state absorptions (GSA) of doped bulk samples were measured in transmission using an FTIR spectrometer. The excited state absorptions (ESA) were then measured between 2.5 to 8.5 µm in doped fibre samples with a FTIR (Fourier transform infrared) spectrometer, while depopulating the ground state via photoexcitation with a NIR laser source. Additionally, the MIR PL emission spectra of fibre samples, excited with a NIR source, were then collected. The preliminary results from these measurements are shown in Tables 1 and 2. The GSA spectra showed the relative heights of the absorptions caused by the chalcogenide glass host, lanthanide ion dopant and impurities, and were used to obtain the corresponding energy level diagrams for the lanthanide ion dopant. The ESA spectra showed the changes in intensity of transmitted light, with an increasing intensity corresponding to a reduction in GSA due to depopulation of the ground state, while a decreasing intensity corresponded to an increasing ESA. These changes in transmitted intensity were then matched to potential upward transitions in their respective energy level diagrams.
The features seen in the ESA spectra for praseodymium (III) measured using 1.547 and 2.0 µm photoexcitation schemes show that both exhibit similar trends (see Table 1 (c)(d)). An increase in transmitted intensity at 4.74 µm (2109 cm-1) could be attributed to a reduction in GSA between the 6H4 →6H5 energy levels, while a decrease in transmitted intensity centred around 3.65 µm (2739 cm-1) could be attributed to an increasing ESA between the 6H5→3F2 energy levels. Additionally, the ESA results for praseodymium (III) seem unaffected by the underlying selenium-hydride impurity (in the glass host) absorption . The ESA seen at 3.65 µm overlaps with the GSA at 4.74 µm, both of which are within the MIR PL emission span of Pr3+ from 3.4 to 5.9 µm (see table 1 (e)) thus potentially contributing to increased reabsorption.
The ESA spectra for terbium (III) measured using a 2.0 µm photoexcitation scheme (see Table 2 (c)) show multiple distinct features. An increase in transmitted intensity around 4.72 µm (2118 cm-1) and 2.92 µm (3424 cm-1) could both be attributed to a reduction in GSA between the 7F6 →7F5, and 7F6 →7F4 energy levels respectively. Reductions in transmitted intensity centred around 3.45 µm (2898 cm-1), 4.23 µm (2364 cm-1), and 7.7 µm (1298 cm-1) could each be attributed to an increasing ESA between the 7F5→7F2, 7F5→7F3, and 7F5→7F4 energy levels respectively. Additionally, the ESA seen at 4.23 µm seems to overlap the GSA at 4.72 µm, and both of which are within the MIR PL emission span in Tb3+ from 4.2 to 5.54 µm (see Table 2 (d)). Additionally, the ESA measurements for both samples required heatsinking to avoid premature damage of the fibre at the laser launch face.
We have developed a novel method for characterising Lanthanide doped optical fibre in the MIR. This method enables measurement of excited state parameters including ESA, ground state depopulation and gain. These results show the importance of ESA in supressing gain in Lanthanide doped fibres and the requirement to select pumping schemes and lasing wavelengths that avoid ESA.