Analysis of Thermally Diffused Single Mode Optical Fiber Couplers

The phenomenon of dopant diffusion as a viable means of coupler fabrication is investigated. It is well known that the diffusion of dopants can improve the uniformity of multimode star couplers manufactured by the fused biconical taper technique. The theoretical basis for the same phenomenon in a single mode coupler is developed, on the basis of the theory of diffusion and the Gaussian approximation for circular fibers.

A novel technique to manufacture and design single mode optical fiber couplers with a minimization of the manufacturing complexity is demonstrated. Traditionally fused biconical tapered couplers have been manufactured by twisting, fusing and elongating optical fibers at elevated temperatures. Usually, high temperature oxy-hydrogen flames are used for such purposes and some degree of skill is needed for a human operator. The complexity of control procedures for automation of the process is greatly increased by the fact that the tapering process is an integral part of the feedback loop. This can be eliminated if a constant tension is maintained on the fibers in the heating process while heat is applied uniformly from a source such as a platinum wire furnace. Since the refractive index differentials responsible for the guiding phenomenon at optical frequencies are directly dependent on concentration of dopants like fluorine and germania, radial diffusion of such dopants causes the fiber cores that are heated in a platinum wire furnace to come closer together. Such proximity leads to the phenomenon of evanescent field interaction or coupling of optical power from one arm of the coupler to the other.

The time evolution of the coupling process can be predicted in theory. While initial results are promising, the ability to automate the manufacture of couplers will be successful only after greater control over the variables is established. It is the intention of this work to understand the physics behind the mechanism as well as to prove the feasibility of modeling real world phenomena under controlled conditions.