Polymer Optical Fiber Tip Mass Production Etch Mechanism to Achieve CPC Shape for Improved Biosensor Performance

Abstract

We report on a simple chemical etching method that enables nonlinear tapering of Polymer Optical Fiber (POF) tips to manufacture Compound Parabolic Concentrator (CPC) fiber tips. We show that, counter-intuitively, nonlinear tapering can be achieved by first etching the core and not the cladding. The etching mechanism is modelled and etched tips are characterized both geometrically and optically in a fluorescence glucose sensor chemistry. A Zemax model of the CPC tipped sensor predicts an optimal improvement in light capturing efficiency of a factor of 3.96 compared to the conventional sensor with a plane-cut fiber tip. A batch of eight CPC fiber tips has been manufactured by the chemical etching method. The batch average showed an increase of a factor of 3.16, which is only 20% less than the predicted value. The method is reproducible and can be up-scaled for mass production.

Keywords: blood or tissue constituent monitoring; etching; fiber optic sensors; nonimaging optics; polymers.

Figure 1

Setup for chemical etching of the fibers for compound parabolic concentrator (CPC) tip formation. The tips are first immersed in dibromomethane (DBM) for etch of the core, then cleaned with distilled water and finally immersed in triethylphosphate (TEP) for etch removal of the cladding.

Figure 2

Microscope picture of etched CPC fiber tip with its specific dimensions. The bright zone from the right side of the picture to the center is a reflection from the microscope light source.

Figure 3

Sketch illustrating the proposed etch mechanism. Only the dipped section of the fiber is shown. Note that the etched fiber tip shape is not following a DBM iso-concentration profile. This is due to PMMA molecular orientation along the fiber axis originating from the fiber drawing process determining local etch speed and combined cladding and dissolved polymer buoyancy restriction of polymer transport away from the fiber.

Figure 4

COMSOL Multiphysics 5.3a simulation of PMMA total displacement (surface) and displacement field (arrow surface) after 100 s for the POF dipped 700 µm into DBM. Involved physics in the simulation: Transport of diluted Species, Solid Mechanics and Hygroscopic Swelling. Swelling is assumed isotropic. Dissolution (free PMMA molecules) is not included in the simulation. We assume no movement of the cladding. Real values for DBM diffusion constants, hygroscopic swelling coefficient, solubility in Poly(vinylidene difluoride) (PVDF), and PMMA are not known, but also not important for illustration of the combined mechanism. However, in the shown simulation, diffusion constants for DBM was assumed to be $5·{10}^{-12}$ m²/s in PMMA perpendicular to the fiber axis and in the PVDF cladding. Along the fiber in PMMA it was assumed to be $5·{10}^{-14}$ m²/s. DBM concentration at all surfaces was at all times set to be equal to some arbitrary solubility of DBM in PVDF. From the simulated displacement field, it is seen, that PMMA was locally moved by swelling in directions and extents, which could very well lead to the CPC shape. This picture is unchanged if e.g., the diffusion is assumed to be isotropic and/or the cladding is thicker.

Figure 5

Formation of CPC fiber tip. Five fibers were used. Each fiber was etched with cladding using different etching times. To make the shape visible, the cladding was removed in each case before taking the picture. The picture at 9 min is marked with red tangent lines to indicate the angle ${\Theta }_{\mathrm{C}}$ with the dominant PMMA molecular direction along the fiber axis. The vertical bright zones on the fibers are reflections from the microscope light source.

Figure 6

Comparison of ideal and chemically etched CPC profile. The CPC length values are from the fiber tip of one of the eight CPCs. The microscope measurement precision was ±1 µm.

Figure 7

Spectrum for plane-cut and CPC fiber tips for dummy sensor.