Choosing the optimum optical fiber for a specific use may be an obvious task: the common requirement is that the fiber's lifetime should be longer than the expected system-operational time. In this respect, reliability issues have been studied and much has been achieved.1,2 Our previous studies concerned optical fibers aging in water for long periods (from several months up to two years), followed by shorter periods (from several days up to three months), and then adding supplementary energy (a controlled applied stress) to the water-corrosion factor.3
The reliability and expected lifetimes of optical fibers are closely related to environmental chemical action on silica, which in turn influences fiber strength. Previous work has emphasized two major mechanisms of strength degradation, including aging and stress corrosion. Silica is sensitive to exposure to water but also to methanol and other reagents through the classical stress-corrosion phenomenon. The latter is estimated through the stress-corrosion factor, a parameter characterizing a material's capacity to resist stress. Considering the numerous application fields for optical fibers, one may wonder to what extent more aggressive chemical reagents may influence their mechanical reliability.
We exposed multimode optical fibers to acetylene (C2H2), ammonia (NH3), and dimethyl sulfoxide (DMSO: C2H6SO) for different durations after prior vacuum exposure. Dynamic-tensile testing results were treated using Weibull statistics. A comparison of mean failure strength of as-aged fibers to the influence of water for similar exposure duration revealed the highest sensitivity to DMSO. We noticed differing influences of the chemical reagents on optical-fiber strength. For acetylene, an interesting, unexpected effect followed short-term gaseous exposure. The fibers exhibited a decrease in mean strength and a broader distribution with increasingly evident extrinsic defect populations, i.e., the Weibull slope for aged fibers was smaller than that of nonaged fibers. Humidity influence subsequent to acetylene exposure seems to favor extrinsic defects and a broader distribution. After prolonging the exposure to seven days, however, we noticed an unexpected, less severe effect. We hypothesize that this may be a mutual compensating effect, i.e., up to a certain exposure time, the acetylene molecule appears less reactive. This may be caused by partial polymerization at the silica-polymer interface, which leads to a less severe decrease in mean strength.
The behavior of fibers soaked in DMSO reagent revealed that, for short immersion times, the mean strength decreases (in comparison with nonaged fibers) but maintains a monomodal, steeper plot. Increasing the exposure duration produced an increasingly significant mean-strength decrease, and extrinsic-defect populations lead to an excessively broad multimodal distribution, in particular for the 18h exposure duration. Compared to the effect of water for similar immersion times, DMSO appears more aggressive than deionized water. Longer immersion in water (18h compared to 2h) leads to a curing effect at crack tips, slightly increasing the mean strength, as we saw in our previously reported observations. That was not the case for DMSO, where prolonged exposure led to irreversible fiber failure.
Finally, the ammonia effect for an exposure of seven days appears more aggressive than that of acetylene for the same exposure, but less so than of DMSO aging for 18h. For aging in ammonia, the mean strength decreases again in comparison with the nonaged fiber strength, but the Weibull plot presents a monomodal distribution that is quite similar to that of acetylene after seven days. We note that the ammonia effect is less disastrous than that of the DMSO solution. Water is a known factor in the propagation of cracks at glass-fiber surfaces, because it makes it much easier to break the silicon-oxygen bonds that build the vitreous network.
Certain reagents, such as DMSO, acetylene, and ammonia, lead to more severe damage of fiber strength than that generated by water. Silica optical fiber was most severely damaged by a DMSO solution. After 18h of exposure, its polymer-acrylate coating appeared to provide ineffective protection from severe chemical damage. We conclude that using coated optical fibers does not ensure mechanical stability in extreme conditions. Scanning-electron-microscope observations of DMSO-exposed fibers (see Figure 1) were consistent with mechanical measurements.
Figure 1. Polymer damage from exposure to a dimethyl sulfoxide (DMSO) reagent.
Acetylene action on optical fibers is more complex, because a prolonged aging duration of seven days can lead to a less severe effect than a short aging duration. And after a certain aging time, acetylene molecules appear less reactive, perhaps because of partial polymerization at the silica-polymer interface. We observed damage and swelling to the polymer coatings (see Figure 2). We advise use of hermetically sealed optical fibers in acetylene and DMSO environments. Such fibers are designed to improve aging behavior by preventing acetylene and DMSO diffusion through the glass surface and coating.
Figure 2. Polymer-coating damage after exposure to acetylene.
Our research is focused on optical-fiber characterization when subjected to different, harsh environments to study their optical properties. We foresee continued cooperation between the Romanian and French teams.
This work was performed at the University of Rennes 1 (France), together with Marcel Poulain and Rochdi El Abdi. The authors express their gratitude to Verrillon Inc. (North Grafton, MA) for technical assistance and material support.
Irina V. Severin
Polytechnic University of Bucharest
Irina Severin is a mechanical engineer with a PhD in composite materials. A professor at the Polytechnic University of Bucharest, she has been a visiting professor at the University of Rennes (France) since 2001 for teaching and research on optical-fiber characterization. She is an expert European Foundation for Quality Management (European Commission) assessor and has published more than 40 journal articles, 60 proceedings papers, and six books.
1. F. Berghmans, Reliability of optical fibers and components, Proc. SPIE
5855, pp. 20, 2005. doi:10.1117/12.623286
2. E. Austin, A. van Brakel, M. N. Petrovich, D. J. Richardson, Fibre optical sensor for C2H2gas using gas-filled photonic bandgap fibre reference cell, Sens. Act. B Chem
. 139, pp. 30, 2009.
3. I. Severin, M. Poulain, R. El Abdi, Progress in reliability of silica optical fibres, Proc. SPIE
7003, pp. 700322, 2008. doi:10.1117/12.781662