High-Tg TOPAS microstructured polymer optical fiber for fiber Bragg grating strain sensing at 110 degrees

: We present the fabrication and characterization of fiber Bragg gratings (FBGs) in an endlessly single-mode microstructured polymer optical fiber (mPOF) made of humidity-insensitive high-Tg TOPAS cyclic olefin copolymer. The mPOF is the first made from grade 5013 TOPAS with a glass transition temperature of Tg = 135°C and we experimentally demonstrate high strain operation (2.5%) of the FBG at 98°C and stable operation up to a record high temperature of 110°C. The Bragg wavelengths of the FBGs are around 860 nm, where the propagation loss is 5.1dB/m, close to the fiber loss minimum of 3.67dB/m at 787nm.

The environmental conditions can be compensated for with the dual-grating technology, where only one FBG is subjected to strain [20,21]. Although progress has been made towards reducing the influence of humidity on PMMA fibers by doping [22], the issue of humidity dependency was first solved with the fabrication of the first TOPAS mPOF in 2007 [4,5]. The polymer TOPAS belongs to the class of cyclic olefin copolymers (COCs), which is a class of optical thermoplastics that are chemically inert and have a very low moisture uptake, high water barrier, and good optical transmission [23]. This means that a TOPAS FBG is humidity insensitive [24]. However, the aim with the original fiber was to use the chemical inertness of TOPAS for localized fluorescence-based biosensing [4,5]. It was first in 2011 that a 1569 nm FBG was written into a TOPAS mPOF, using 325 nm UV writing [25]. Soon thereafter an 871 nm FBG in a TOPAS mPOF was reported and it was demonstrated that the TOPAS FBG was indeed humidity insensitive to within the accuracy of the climate chamber that was used [24]. Interestingly, it has been demonstrated that TOPAS is also a very good material for terahertz (THz) fiber fabrication due to its low material loss and material dispersion at THz frequencies [26][27][28].
In all published papers on TOPAS mPOFs, the fibers were made of the particular TOPAS grade 8007, which has a glass transition temperature of only T g = 80°C [29]. This is even lower than that of PMMA (typically 110°C), which means that the issue of a low operating temperature remains a problem for polymer FBGs. Here we demonstrate for the first time the fabrication of an mPOF made of high-T g TOPAS grade 5013 with T g = 135°C. We further inscribe FBGs into the fiber and demonstrate strain sensing of 2.5% strain at 98°C, where a PMMA fiber would malfunction (max. operation temperature of 92°C demonstrated in [17]). Further we also demonstrate strain sensing at a record high temperature of 110°C. Due to setup limitations the strain sensing at 110°C is limited to strains up to 0.3%. For applications in for example accelerometers, the strain rarely exceeds 0.3% [9].
Taking into consideration that the TOPAS fiber is humidity insensitive [24] and that the TOPAS FBG strain sensor may be temperature compensated using dual-FBG technology [20,21], this demonstration of high-temperature operation at 110 °C provides a significant step towards a practically applicable polymer FBG sensor platform. The fiber is shown to have a minimum loss of 3.67 dB/m at 787 nm and a loss of 5.1 dB/m at the FBG wavelength 853 nm.

Fabrication of high-Tg TOPAS mPOF and FBG inscription
In our experiments we employed a high-Tg TOPAS mPOF fabricated in-house at DTU Fotonik using the drill and draw technique. The material used for the mPOFs was TOPAS COC of the particular grade 5013, obtained from TOPAS Advanced Polymers, Inc. The TOPAS was cast into 6 cm diameter rod and the desired hexagonal holes structure was drilled into the rod. The preforms were then drawn to fiber by first drawing a 5 mm cane, which was then sleeved and drawn again. The two final high-T g TOPAS mPOFs we use here have a diameter of 130 μm and 180 μm, respectively, and a solid core surrounded by three rings of air holes arranged in a hexagonal lattice. For the fiber with a diameter of 130 μm the air-hole diameter is d = 2.2 ± 0.2 μm and the distance between the air-holes (pitch) is on average Λ = 6 μm, which means that the fiber is endlessly single moded (d/Λ < 0.43). The end facet of the high-Tg fiber with a diameter of 130 μm is shown in the inset of Fig. 1(a).
The grating inscription was performed using a 30 mW CW HeCd laser operating at 325 nm (IK5751I-G, Kimmon). A circular Gaussian laser beam was expanded by a cylindrical lens in one direction, the one along the fiber, from 1.2 mm to 1.2 cm. The laser beam was then focused downwards using another cylindrical lens to sideways expose the core of the fiber through a phase-mask. The phase mask has been custom made for 325 nm writing with a uniform period of 572.4 nm (Ibsen Photonics). This would result in a Bragg grating wavelength λ B of 872 nm if no annealing or pre-strain was applied. However, depending on the applied pre-strain and thermal treatment of the fiber the grating resonance will be blueshifted [20,21]. As an example the final grating wavelength of the fiber with a diameter of 130 μm was 853.4 nm and the strength of the reflected peak was around 20 dB, as shown in Fig. 1(a). The about 19 nm blue-shift is because we both used pre-strain and afterwards annealed the FBG.
Since the TOPAS high-T g optical grade 5013 mPOF was not reported before, a cut-back measurement with 5 cuts was performed in order to determine the transmission loss of the fiber. Cleaving was done with a homemade electronically controlled hot blade cleaver with flat side blade at a temperature of 75°C for both the blade and fiber, giving a high quality cleaves [30]. Figure 1(b) shows the loss profile of the high-Tg TOPAS mPOF for the wavelength span of 500-1000 nm. The loss at the Bragg wavelength 853.4 nm is found to be 5.12 dB/m, whereas the minimum loss is 3.67 dB/m at 787 nm. The loss can be reduced by optimizing the drawing process [16], but in our experiments the loss is not a limiting factor.

Thermal response of high-T g mPOF FBGs
The temperature response of the grating of the 130 μm fiber shown in Fig. 1(a) was studied by monitoring the reflection spectrum using a fiber circulator, a supercontinuum source (SuperK Versa by NKT Photonics A/S) and an optical spectrum analyzer (Ando AQ6317B). Light was launched into the mPOF using a silica step index fiber, which was butt-coupled to the mPOF. In order to reduce the background noise, a small amount of refractive index matching gel was applied between the silica fiber and the mPOF to minimize Fresnel reflections. The grating was then placed on a resistive hot stage to control the temperature. A thermocouple was placed as close to the grating as possible with a temperature uncertainty of 2%, while several layers of lens paper were covering the grating to isolate it and create a more uniform temperature. For stable strain sensing it is crucial to anneal the fiber before operation [17,19]. We therefore annealed the FBGs for ~3 hours at 80 degrees before operation.  Figure 2(a) shows how the Bragg wavelength blue-shifts by 1.2 nm as the temperature is increased to 107°C, waiting 20 minutes at each temperature for the fiber to stabilize before making a recording. Only a very small hysteresis is observed, while the wavelength shift is returning to zero as the temperature is ramped down again at 20 minute intervals, which indicates that the fiber was properly annealed before operation. Figure 2(b) shows the reflected normalized power at the (temperature dependent) Bragg wavelength. The power returns also to the initial level after a full temperature cycle, but fluctuations are apparent and a maximum hysteresis of 0.75dB occurs at 50°C. The small power variation between the forward and reverse temperature cycle is probably due to experimental errors, such as laser fluctuation, coupling instabilities, etc. The total drop of the reflected power is 3.2 dB at the maximum temperature of 107°C.

Small strain response of high-T g mPOF FBGs under different temperatures
The small strain response of the FBG in the 130 μm in diameter high-T g TOPAS mPOF was investigated by mechanical stretching. The two ends of the fiber were clamped and glued to two micro-translation stages (Thorlabs) with UV-cured glue, with one of them used to buttcouple the mPOF to the single-mode silica fiber. The glue is mechanically much stiffer than the high-Tg TOPAS mPOF, so that it does not unduly influence the strain. One stage can move longitudinally to apply axial strain to the grating manually with a low loading speed. The axial strain values were determined by dividing the fiber longitudinal elongation by the length of fiber between the two gluing points. The gratings were left to stabilize for about ten minutes at each value of the strain, before a recording was made.  Fig. 2(b) we see that the 2.5 dB drop corresponds well with the drop due to temperature alone. Again the small power variation between the forward and reverse strain loading cycle is probably due to experimental errors, such as laser fluctuation, coupling instabilities, etc.
PMMA FBGs have been taken to 92°C without applied strain [17] and here the high-Tg TOPAS mPOF survived 100°C at 0.3% strain, which is clearly a record. A golden rule for long term stability is to keep the temperature at least 20°C below the glass-transition temperature. PMMA has a T g around 110°C, so a PMMA FBG could therefore in principle survive 100°C for a short time. Our high-T g TOPAS has a T g of 135°C, so we therefore finally test our 130 µm FBG for small strain operation at 110°C. The Bragg wavelength shift for a strain loading-unloading cycle is shown in Fig. 4(a) giving a slope of 1.0 pm/µε. In order for the Bragg wavelength to be recoverable and avoid destroying the grating the operational strain regime of the FBG was kept below 0.16%, where the peak power has decreased by 6 dB as shown in Fig. 4(b). We note that 0.16% is still significantly more than the 0.02% used in the recent demonstration of highly sensitive mPOF FBG accelerometers taken to 15 g [9].

Time response of high-T g mPOF FBGs at 110°C
In this section, we demonstrate the thermal stability of the 130 µm in diameter mPOF FBG made of high-T g TOPAS mPOF at 110°C. In this measurement one end of the fiber was glued to a stage and used for incoupling, while the other end of the fiber was not mounted to avoid unwanted strain. Figure 5(a) shows the blue-shift of the Bragg grating with time while the variation of the Bragg wavelength shift is shown in Fig. 5(b). The total shift of the resonant wavelength was 5 nm after 7 hours at 110°C. After 7 hours the temperature was decreased to room temperature and after 17 hours at room temperature the resonant wavelength did not recover, but remained at the same wavelength. This permanent blue-shift of the resonant wavelength can be explained by fiber shrinking. The polymer optical fibers are pulled under tension, which results in an axial orientation of the polymer chains and a residual strain in the fiber [8,31]. At a certain temperature the chains relax from the axial orientation, which causes shrinkage of the fiber. This effect is directly linked with the drawing tension, thermal history and the annealing process of the fiber and has been extensively studied [8,31]. The blue-shift indicates that possibly the high-T g TOPAS FBGs would benefit from being annealed for a longer time than 3 hours at 80°C or at a higher temperature. However, an investigation into the optimum annealing time and temperature for the high-T g TOPAS FBG is outside the scope of this initial demonstration of fiber drawing, FBG writing, and high-temperature operation.

Large strain response of high-T g mPOF FBGs under different temperatures
Although in many applications only small strains are needed [9], polymer fibers are superbly well suited in applications where large strains are needed, such as in geo-textiles monitoring land-slides and in respiratory monitoring, due to their high failure strain [8]. In this section the large strain response of the high-T g TOPAS FBG is investigated. Using the same setup as in the low strain section and, a similar fiber, 180 μm in diameter with a hole diameter around 3 µm and a pitch around 8 µm, is subjected to strains up to 3% at room temperature, 2.2% at 50°C and 2.5% at 98°C. The Bragg grating wavelength of the unstrained fiber was 869.6 nm. The gratings were left to stabilize for about two minutes at each level of applied strain. The result of this strain investigation is shown in Fig. 6. The gratings were strained until the glue holding the fiber failed. For this reason there is no data on the releasing of the strain. The large-strain response is also found to be approximately linear, with the grating response at 98°C having a fitted linear slope of 0.75 pm/µε.

Conclusions
In this paper, we have demonstrated for the first time the response of FBGs inscribed in a new high-T g TOPAS mPOF. The host material used for the fabrication of fiber was grade 5013 TOPAS cyclic olefin copolymer, which has a glass transition temperature as high as Tg = 135°C. We demonstrate that FBGs in the high-T g TOPAS mPOF are able to operate at 110°C under applied tensile strain up to 0.16% and at 98°C under applied tensile strain up to 2.5%. It should be emphasized that there is not any report so far demonstrating the operation of PMMA POFs or mPOFs at higher temperature than 92°C [17]. The fiber was characterized under applied tensile strain at 50°C, 100°C and 110°C demonstrating consistent repeatability for all cases. We also investigated the thermal stability of the FBGs over time. The fiber was heated for around 7 hours at 110°C with a total resonant wavelength blue-shift of ~5 nm. It should be mentioned that after 24 hours at room temperature, the resonant Bragg wavelength remained at the same position. Overall, we believe that the current work is a significant step towards fabrication and commercialization of polymer FBG sensors based on TOPAS, which further enhances the ability of polymer fibers to operate at temperatures up to 110°C.