Open Access
Add to BibSonomyAbstract
We demonstrate the application of a Fluorescence Intensity Confocal Optical Microscopy technique to the determination of the relative erbium ion distribution in optical fibers. As well as being able to acquire two dimensional profiles of the relative erbium ion distribution, this method can also provide valuable information on a sub-micron level regarding physical and optogeometric parameters of the examined samples.
©2005 Optical Society of America
1. Introduction
Rare-earth doped optical fibers are now being used extensively for fiber lasers and fiber amplifiers [1]. The capabilities of these devices have found significant use in a range of scientific and industrial environments among which the field of optical communications has been influenced the most. Optimising the design and fabrication of such devices has become a continuing challenge to optical scientists and engineers. It has been demonstrated that the properties and performance of Er-doped fiber (EDF) lasers and amplifiers (EDFA) are related to a number of parameters such as the fiber glass material, the waveguide characteristics, and the distribution profile of the erbium ions. The accurate knowledge of the latter has been found to be vital for the optimal design and operation of these devices [2–4].
To date, two different approaches have been employed to measure the distribution of erbium ions in optical fibers. In one of these, chemical analytical techniques (i.e., Secondary Ion Mass Spectroscopy [5], Electron Probe Microanalysis [6], and Extended X-ray Absorption Fine Structure Spectroscopy [7]) have been applied to fiber preforms from which fiber is drawn, in order to investigate the erbium distribution profile. The information extracted about the preform is then scaled down to match the final fiber dimensions assuming that there are no changes during the drawing process that can alter the dopant profile of the drawn fiber. However, there is no guarantee that during fiber drawing the dopant profile of the resulting fiber will be the same as that of the original preform. It is expected that due to dopant diffusion some redistribution of the erbium ions will occur during the drawing process. In the second approach, measurements were made directly on the drawn fiber. Transmission electron microscopy (TEM) [8], Raman microscopy [9] and fluorescence intensity based confocal microscopy [10] are some of the techniques previously used for that purpose. In this Letter we present the results from the application of a Raman confocal imaging system to the determination of the erbium ion distribution in germano-alumino-silicate optical fibers by observing the backscattered fluorescence signal generated on the Er-doped fibers after appropriate optical excitation. The advantage of this technique is that it can provide information about the relative erbium ion distribution by exploiting the high wavelength resolution of a Raman spectrometer together with the reduced depth of field and the submicron spatial resolution of a confocal microscope. In addition there is no requirement for fiber sample preparation as in the case of the TEM scheme, and information about the erbium ion distribution can simply be gathered from the endface of a cleaved fiber sample.
2. Experimental method
In this work a micro-fluorescence scanning technique, based on the exploitation of a commercial Renishaw Raman Confocal Microscope, is employed for the investigation of the erbium ion distribution in optical fibers. An experimental EDF sample (code: Er35) acquired from the Laboratoire de Physique de la Martiére Condensée (LPMC), Université de Nice (Nice, France), one EDFA sample as well as a unique EDF sample, used in the production of a photonic crystal fiber (PCF) laser [11] supplied by the Optical Fiber Technology Centre (OFTC), University of Sydney (Australia), were examined. Images were obtained from the endface of freshly cleaved (~8 mm long) fiber samples.
The profiling scheme adopted for the imaging of the Er-ion distribution on the endface of cleaved fiber samples sees pump light from an argon ion laser (λ=514 nm, P~1mW) focused by a microscope objective (x100, NA=0.9) to a diffraction limited submicron spot incident on the endface of the fiber (Fig. 1). In the core region, where the erbium ions are located, the incident pump excites the ions from the ground state 4I15/2 to the 2H11/2 upper energy level. Radiative decay back to the ground state from this level, and non-radiative relaxation to the next lowest level, 4S3/2, followed by radiative emission to the ground state, yields fluorescence bands centered on 520 and 550 nm, respectively (Fig. 2(a)). The 550 nm fluorescence peak is then collected by the same microscope objective and is spectrally analyzed with the Raman system (Fig. 2(b)).
Fig. 2. (a) Partial energy level diagram for erbium in germanosilicate host. (b) The 550 nm fluorescence peak collected by the Raman system.
The existence of appropriate interference filters in the arrangement of the Raman system ensures that adequate discrimination between the excitation and backscattered fluorescence wavelengths can be achieved. The intensity of the backscattered fluorescence from each point is taken as an indication of the local dopant concentration. It has been assumed that the 550 nm backscattered fluorescence is proportional to the erbium ion concentration since the process of de-excitation of the 4S3/2 level is free from any possible cross-relaxation or upconversion effects. A two-dimensional fluorescence profile is then formed by scanning the core area in submicron steps across the endface of the fiber (X-Y plane). As a result, two-dimensional images of the relative ion distribution in the plane of the cleaved fiber face are obtained for each one of the examined samples.
It was found that the wavelength range (536.5–563.5 nm) taken as an indication of the local relative erbium ion concentration is free from any possible Raman peaks arising in the core silica-dopants environment. That observation was achieved by scanning the cladding area, where no erbium ions exist, or the core area of fibers with similar composition to one of the active fibers with the exception that no erbium ions were present. Furthermore, after repeatedly examining the backscattered fluorescence signal at one point for an extended amount of time it was found that the measured fluorescence signal was fairly constant suggesting that for the duration of the measurements the relative populations and hence fluorescence signals of the thermally coupled 2H11/2 and 4S3/2 energy levels would remain independent of each other. From this it was concluded that the effect of local heating of the sample due to the probe beam was small and, in particular, did not disturb the accuracy of the measurement.
3. Results and discussion
3.1 Examination of the amplifier fiber
Employing the method described above and scanning the endface of the freshly cleaved EDFA sample in steps of 0.25±0.05 µm the fluorescence image displayed in Fig. 3 was acquired. A three dimensional projection of the relative erbium distribution is also shown in that figure where some form of depression in erbium concentration at the central core region is clearly visible. There have been a number of previous studies [5, 12, and 13] on the preforms of EDFAs that have revealed a similar depression in erbium concentration at the central core area, which is known to negatively impact on the performance of an EDFA.
Fig. 3. Two dimensional fluorescence profile detected from the endface of the freshly cleaved EDFA sample.
Fig. 4. A transverse profile through the centre of the fiber showing the relative erbium ion distribution. Also shown is the average radial refractive index profile determined using QPM
The measurements presented here are in direct contrast to reports that aluminum oxide improves the solubility of rare earth ions and as a result there is little [12] or no indication at all [13] of any depression in erbium concentration at the central core region, especially in the case where erbium concentration does not exceed the critical limit of 100 ppm. Consequently, knowing that the EDFA sample under investigation had an estimated erbium concentration that is not higher than this critical limit and that aluminum had been incorporated during the fabrication of this fiber, it was not expected that the erbium ions would be depleted in the central core area. On the contrary, a Gaussian like erbium distribution profile was anticipated. It becomes even more evident that the erbium is depleted at the central core area when the transverse profile through the centre of the fiber of the relative erbium ion concentration is studied (Fig. 4). In that figure, the extracted erbium profile is directly compared with the refractive index profile of the fiber that was measured using quantitative phase microscopy (QPM) [14]. Clearly evident is a resemblance of the erbium distribution profile to that of the refractive index including the depleted region in the core centre.
3.2 Examination of Er35
This fiber was fabricated using the modified chemical vapour deposition (MCVD) technique with the rare earth ions being incorporated via solution doping. An estimated erbium concentration of ~7600 ppm was incorporated in the silica core matrix together with a number of network modifiers such as germanium, aluminum and phosphorus. Measurement of the backscattered fluorescence was performed by scanning the endface of the freshly cleaved fiber in steps of 0.50±0.05 µm across the whole core area under investigation producing the two dimensional image shown in Fig. 5.
Fig. 5. Two dimensional fluorescence profile detected from the endface of the freshly cleaved Er35 fiber
It can be seen that the EDF sample Er35 contains an asymmetric configuration, which is believed to be the outcome of an unsuccessful collapse during the fabrication process possibly due to the very high erbium ion concentration.
3.3 Examination of PCF sample
Finally an EDF sample used in the production of a PCF laser was examined with the technique presented here. The erbium ion concentration in this sample was estimated to be around 1000 ppm and with aluminium incorporated in the core area. Measurements were acquired directly from the freshly cleaved endface of the sample in steps of 0.5±0.05 µm. The two dimensional fluorescence profile (Fig. 6) shows that this fiber is free from any fabrication artifacts or any depression in erbium concentration in the centre core region. In fact this sample exhibits a well-defined, almost Gaussian, erbium ion profile with most of the ions located in the centre core region.
Fig. 6. Two dimensional fluorescence profile detected from the endface of the freshly cleaved PCF sample.
4. Conclusions
We have demonstrated the applicability of a Fluorescence Intensity Confocal Optical Microscopy technique utilizing a commercial Renishaw Raman microscope to the characterization of erbium doped optical fibers. The advantage of this technique is that it can provide two-dimensional information about the relative erbium ion distribution with resolution that has been estimated to be around 0.4 µm. As a result, in contrast to the Raman Microscopy technique [9] that was applicable only for the characterization of multimode fibers, this method can be employed for the examination of both multimode and single mode fibers. In addition, by exploiting the high wavelength resolution of a Raman spectrometer adequate wavelength discrimination between the pump and backscattered fluorescence signals is guaranteed, while the reduced depth of field of a confocal microscope ensures that measurements are free from any possible effects due to guiding fluorescence within the fiber. In addition there is no requirement for fiber sample preparation as in the case of the TEM scheme, and information about the erbium ion distribution can simply be gathered from the endface of a cleaved fiber sample. Although this method provides information about the relative erbium distribution rather than the absolute concentration, it still has the capacity to offer new insights regarding the physical and optical parameters of fiber lasers and amplifiers. Investigation into the determination of absolute concentration is the subject of ongoing work.
Acknowledgments
The authors would like to thank John Canning of the Optical Fiber Technology Centre (OFTC), Australia, as well as Thinh Nguyen of Victoria University for the provision of the fiber specimens. This work is sponsored in part by the Australian Research Council. Fotios Sidiroglou would like to thank the Victorian state government (ICT scheme) and the Science Faculty at the University of Melbourne for partly funding his PhD candidature.
References
1. Michel J.F. Digonnet, Rare-earth-doped fiber lasers and amplifiers (2nd ed, Marcel Dekker, New York, 2001). [CrossRef]
2. E. Desurvire, J. L. Zyskind, and C. R. Giles, “Design optimization for efficient erbium-doped fiber amplifiers,” IEEE J. Lightwave Technol. 8, 1730–1741(1990). [CrossRef]
3. W. L. Barnes, R. I. Laming, E. J. Tarbox, and P. R Morkel, “Absorption and emission cross section of Er3+ doped silica fibers,” IEEE J. Quantum. Electron. 27, 1004–1010 (1991). [CrossRef]
4. J. C. Martin, “Erbium transversal distribution influence on the effectiveness of a doped fiber: optimization of its performance,” Opt. Commun. 194, 331–339 (2001). [CrossRef]
5. N. Kagi, A. Oyobe, and K. Nakamura, “Gain characteristics of Er3+ doped fiber with a quasi-confined structure,” IEEE J. Lightwave Technol. 8, 1319–1322 (1990). [CrossRef]
6. T. Kashiwada, M. Shigematsu, T. Kougo, H. Kanamori, and M. Nishimura, “Erbium-doped fiber amplifier pumped at 1.48 µm with extremely high efficiency,” IEEE Photonics Technol. Lett. 3, 721–723 (1991). [CrossRef]
7. D. T. Bowron, R. J. Newport, J. S. Rigden, E. J. Tarbox, and M. Oversluizen, “An X-ray absorption study of doped silicate glass, fiber optic performs,” J. Mat. Sci. 38, 485–490 (1996). [CrossRef]
8. A. C. Pugh, R. P. Stratton, and D. B. Lewis, “Investigation of elemental diffusion during the drawing and heat treatment of glass optical fibers,” J. Mat. Sci. 29, 1036–1040 (1994). [CrossRef]
9. S. Rogard, A. Bulou, H. Poignant, and M. Rousseau, “Non-destructive concentration measurements of erbium-doped ZBLAN fluorozirconate glass,” Opt. Mat. 4, 557–563 (1995). [CrossRef]
10. A. Othonos, J. Wheeldon, and M. Hubert, “Determining erbium distribution in optical fibers using phase-sensitive confocal microscopy,” Opt. Eng. 34, 3451–3455 (1995). [CrossRef]
11. N. Groothoff, J. Canning, T. Ryan, K. Lyytikainen, and H. Inglis, “Distributed photonic crystal fiber (DFB-PCF) laser,” Opt. Express , 13, 2924–2930 (2005) [CrossRef] [PubMed]
12. B. J. Ainslie, “A review of the fabrication and properties of erbium-doped fibers for optical amplifiers,” IEEE J. Lightwave Technol. 9, 220–227 (1991) [CrossRef]
13. B. J. Ainslie, J. R. Armitage, S. P. Craig, and B. Wakefield, “Fabrication and optimisation of the erbium distribution in silica based doped fibers,” in Proceedings of IEE Fourteenth European Conference on Optical Communication (Brighton, UK, 1988) pp.62–65.
14. A. Roberts, E. Ampem-Lassen, A. Barty, K.A. Nugent, G.W. Baxter, N.M. Dragomir, and S.T. Huntington, “Refractive-index profiling of optical fibers with axial symmetry by use of quantitative phase microscopy,” Opt. Lett. 27, 2061–2063 (2002). [CrossRef]
