The efficiency and methodology of coupling light into microcavities has improved exponentially in the last decade. One such advancement is coupling light onto silicon microspheres. The material, size, and shape of a silicon microsphere are ideal for optical devices. Silicon microspheres are not the primary material used to fabricate microspheres for optical coupling because current methods used for microsphere fabrication cannot produce single-crystal silicon in the 16-μm scale, which is best for current optical technology.
The pulsed laser ablation method can now produce silicon microspheres that are ideal for optical use. This is a process in which the surface of a silicon substrate is super-heated by a high-power laser until molten, and a second laser pulse hits the molten silicon, ejecting micronsized spherical particles. Furthermore, silicon is naturally abundant, which further enables the large-scale production of optically compatible microspheres.
Processes for forming spherical structures exist in nature. The most commonly known example of a microsphere that nature provides is a rain droplet. Rain droplets form a spherical structure while falling in air because of the surface tension in the water molecules taking advantage of this shape, which has the smallest surface-area-to-volume ratio. Thus, a raindrop has a nearly perfect spherical shape as it travels in space. The approach used in this work is to fabricate a single-crystal silicon microsphere as inspired by this natural process of raindrop formation.
As with the microspherical liquid water droplets, ablation of a silicon wafer momentarily makes liquid silicon droplets in space, allowing these droplets to form spheres and cool to a solid state before settling onto the silicon wafer surface. This process allows a reproducible large-scale production of silicon microspheres in the 100-μm size scale.
The procedure for fabrication first requires stripping off the jacket and buffer from a short section at the end of an optical fiber, followed by the slow heating and stretching of this exposed area until the taper is the desired diameter. A fiberoptic tapering setup was used to create tapered fiber. This tapering setup consisted of two motorized clamps and a torch.
Motorized clamps pull the fiber apart as the gas torch, which a separate motor controls, moves slowly vertically while the two clamp stages move laterally. A small fiber about 5 to 6" was placed between the two clamp stages. First, with the single mode fiber (SMF) clamped in place, researchers manually moved the torch such that the SMF was directly above it. The distance between the preset location and the set torch z-axis position was recorded for automation.
In the pre-stretching phase, the two clamp stages were moved in phase to place the stripped fiber into the pre-heat phase for further cleaning. The pre-heating removes the excess debris left on the cladding strip by burning off the buffer and jacket residue. After the pre-heating phase, the program returned the torch and the fiber to the alignment position. The program has a 2-second wait period at this stage to soften the cladding and core further so the fiber would be less likely to fracture when stretched.
The right and left stages move individually to stretch the fiber one side at a time. This step controls the taper diameter by augmenting the number of stretches in the program. At the end of the stretching, an additional post-heating process begins. The post-heating phase added extra gravitational sag to the tapered fiber. If the fiber was not post-heated, the tapered fiber would have cooled at a higher tension because the stretching procedure would make removal of the tapered fiber far more difficult as it would have a high tendency to break.
This work was done by B.N.L. Pascoguin, R.P. Lu, J.M. Kvavle, and A.D. Ramirez of SPAWAR Systems Center Pacific. SPAWAR-004