Tarja T. Volotinen
Tarja Volotinen continues research of coloured glasses in collaboration with Sheffield Univ. and Uppsala Univ.. Her expertise contains reliability, durability, fracture analysis, manufacturing methods, optical properties and test methods of optical fibres for communications networks. She has a PhD in physics (optics, optical fibres, 1991) from Helsinki Univ. and the other one in engineering materials from Sheffield Univ. (glasses, 2008). She worked 8 years for Nokia Cables, 11 years for Ericsson Cables, 2 years at OFS Fitel LTD, and over 10 years as consultant in optical fibres. She also worked one year at Bellcore in USA, learning the durability and mechanical reliability issues of fibres. She has also the associate professor title in solid state physics (2011) from Uppsala Univ.
Long term durability and corrosion of optical fiber silica glass surface
Tarja T. Volotinen
Tarja’s Consulting AB, Hudiksvall, Sweden
Optical fibres, consisting of almost pure silica glass cladding and Ge-doped silica glass core, have been used for communications networks worldwide, since beginning of 1980. Very high durability of the glass surface and low enough stress are required in order to maintain the mechanical reliability with a failure rate less than one failure in a 100000 km fibre during service in 40 years.
During 1990 – 2002 a lot of intensive research and development was done worldwide to reach these properties. In EU three COST research projects (COST218, 246(1-3) and 270) were needed to find out the deep understanding in the silica glass durability behaviour in the chemical and physical conditions applied for optical fibres in installed cables and networks. Reliable and repeatable chemical and physical tests methods for the fibre aging and corrosion were developed (4-18). In addition, the proper mechanical tests methods and life-time calculation theories were obtained (1, 4-18). The results of these studies are used for the international ITU and IEC standards applied for communication fibres.
Fused silica (melted from pure natural silica sand) was tested for fibre cladding material. But it was found not suitable enough to reach the long-term reliability (Fig. 1a-d). One reason is the way the contamination is located in fused silica glass. Contamination appears as few dark particles in the sand of which the fused silica is melted. In the melted glass, the contamination ions and molecules are segregated to certain particles/locations, which get solved much faster to water and other chemicals causing local deep enough roughness at corrosive conditions, and also cause a fracture of a fibre at much lower stress than pure, homogenous synthetic silica glass would do. If the contamination is located one-by-one ion or molecule, it cannot decrease strength much or increase the overall corrosivity. The other property that was found to mater for the corrosion behaviour and mechanical properties was the material structure of fused silica vs synthetic silica. A glass made from fused silica could have some part of crystalline structures remained, while the synthetic silica glass was found smooth and homogenous (Fig, 1c-d). Fibres are therefore manufactured from homogeneous, pure synthetic silica (e.g. Heraeus, F300) as the outer cladding material.
The other material issue, found in long term aging tests at elevated temperatures and humid and wet conditions, was the coating materials used. Typically, the UV-cured polyurethane-acrylate materials used in early years, could contain one corroding component that when the fibre was aged, the colour of the coating changed quickly to yellowish or brown and at the same time also provided out an aging product that was found very corrosive for the silica glass surface. Thus, the whole fibre world changed to the coating materials that are non-corrosive.
Because the mechanical strength of a fibre varies a lot along fibre length, it was found obligative to proof-test every produced fibre length. It was found that the drawing conditions may affect very much the mechanical quality of the fibres (4, 13), and that the standardised proof-test can not be done directly at the drawing process (19). There must be at least 20 min waiting time in room conditions between the drawing process of a primary-coated 250 µm fibre to get the humidity molecules to diffuse on the glass surface through the primary coating layers of 62,5 µm (20, 1). The fracture at the typical 1 % strain proof-test during the tension time of 1 s, requires a normal room environment humidity to get diffused on the glass surface. It was also found that the chemical conditions in the fibre aging tests are also dependent on the aging container material and wet surface areas of the fibre cladding and the container (9). The fibres should therefore be finally aged in the cables, not outside of them (5-6). It is the local chemical conditions inside the cable that matter most.
Furthermore, there are also fusion splices in the fibres in a cable network. It takes only a few seconds and warms up the joint position of the fibre glass up to temperature above 2000 °C. But the fibre heats up even around the splice point. At a distance 5 – 10 mm from the splice point, the glass surface is found to be significantly weaker than for the un-spliced fibre (7,13,15). Furthermore, it was found in the aging tests at hot water and humid conditions that the surface corrosion at the splice point could be so fast that it kept the surface smooth, and at the 5 – 10 mm distance the corrosion speed was lower leaving a much rougher surface (Figs 2 – 3) (7). Thus, the fusion splices are normally protected with a special protection construction that supports the strength of the fibre. Other fibre components, that require any kind of touching of the glass surface (e.g. stripping) or a treatment of the glass, need to be specially protected to keep the reliability required.
As a conclusion, optical communications fibres in properly manufactured and designed cables are very durable i.e. chemically and physically protected from the chemical corrosion effects of the surrounding conditions. The fibre cladding glass and coating materials matter most. The mechanical stresses are kept below a standardised limit, and mainly taken care by cable construction.
Figure 1 a- d, Surface corrosion of a fused silica cladding of a fibre (a-c) and of a fibre with homogeneous pure synthetic silica glass surface. (14, 13)
Figure 2, Surface of the spliced aged fibre, showing the corrosion at a location a few mm from the splice point. These graphs were presented in the conference, not included in the publication. (7,2)
Figure 3 An AFM-graph of a aged spliced and corroded fibre cladding surface at a distance of 1mm from the splice point. (7,2) The corrosion speed at the splice point could be so high that the fibre became significantly thinner in this ageing test.
- Tarja Volotinen, Willem Griffioen, Michel Gadonna and Hans Limberger, “Reliability of Optical Fibres and Components, Final report of COST 246”, Springer-Verlag, London. 1 – 410 (1999).
- The course notes for the Short course FO209: Reliability of optical fibers and passive fiber components in long and short communications networks, The 63rd IWCS conference (2014) and also the 65th IWCS conference (2016).
- Tarja Volotinen, Kariofilis Konstadinidis, Victor Cusanello, Ed Tretheway, Ralph Lago and David Mazzarese, ”Mechanical reliability of short optical fiber links in data centers”, The Proc. of The 63rd IWCS conference, 47-54 (2014).
- Tarja T. Volotinen, Anu E. Konkarikoski, C. Bertil Arvidsson, and Thomas K Ericsson, “Impact of silica glass structure on transmission properties of Ge-doped single-mode fibers”, SPIE (The International Society of Optical Engineering) Vol. 4940, 1 – 13 (2003).
- T. Volotinen, “Reliability of optical fibres and components: achievements and conclusions of COST 246”, SPIE Vol. 3848, 88 – 94, (1999).
- T. Volotinen, “Water tests of optical fibres”, SPIE Vol. 3848, 134 – 143 (1999).
- Volotinen, M. Zimnol, M. Tomozawa, Y. – K. Lee, K. Raine, 1998, “Effect of mechanical stripping and arc-fusion on the strength and ageing of a spliced recoated optical fibre”, MRS (Materials, research Society Symposium Proceedings) Vol. 531, 163 – 168 (1998).
- Volotinen, A. Breuls, N. Evanno, K. Kemeter, C. Kurkjian, P. Regio, S. Semjonov, T. Svensson and S. Glaesemann, Mechanical behavior and B-value of an abraded optical fibre, Proc. of 47th IWCS, 881 – 890 (1998).
- Overgaard, P. Haslov, H. Knuuttila, A. Mazzotti, P. Regio, T. Svensson, T. Volotinen and S. Dodd, Effect of water quality and quantity on strength degradation of fused silica fibre in water tests, Proc. of the 45th IWCS, 928 – 938 (1996).
- Griffioen, T. Volotinen, P. Wilson, A. Gouronnec and T. Svensson, Handleability of Aged Optical Fibres, Proc. of the 44th IWCS, 857 – 864 (1995).
- T. Volotinen and O. S. Gebizlioglu, Mechanical behavior of coated fused silica optical fibres aged at elevated temperature in air and filling compound, SPIE Vol. 2611, 72 – 87 (1995).
- H. Yuce, R. A. Frantz, O. S. Gebizlioglu, I. M. Plitz, T. T. Volotinen, The mechanical performance of aged dual-coated fibers with varying extents of coating cure, Proc. of the 42nd IWCS, 875-863 (1993).
- T. Volotinen, H. H. Yuce and R. A. Frantz, Effects of glass preparation on the surface corrosion of fused silica optical fibres, SPIE Vol. 2074, 83 – 94 (1993).
- T. Volotinen, H. H. Yuce and R. A. Frantz, Ageing Behaviour of Fibres, SPIE Vol. 1973, 161 – 174 (1993).
- Volotinen, H. Yuce, N. Bonanno, R. Frantz and S. Duffy, Splicing of Aged Fibres, SPIE Vol. 1973, 186 – 192 (1993).
- T. Volotinen, Achieving long lifetimes and extremely low failure rates for silica optical fibres in communications networks, Invited presentation at SPIE5465 conference, Strasbourg, France, (2004).
- T. Volotinen, Service environment of optical fibres in telecommunications networks, Summary of the replies to COST 246 WG1 questionnaires I and II (93 -95), Proc. of the 1st COST 246 Workshop, (1995).
- J. Bonanno, H.C. Hartman, R. W. Contreras, H. H. Yuce, T .T. Volotinen and J .P. Varachi Jr., Handling Behaviour of Aged and Unaged Fibres during Splicing Operation, Proc. of 9th NFOEC (National Fibre Optic Engineering Conference), June 93, San Antonio, Texas, (1993).
- W Griffioen, Optical Fiber Mechanical Reliability, PhD- Thesis, Eindhoven Univ. of Technology, The Netherlands, 1994.