Measurements at 351 nm of temporal dispersion in fibers

Abstract

1. Temporal dispersion at 351-nm was measured in the following: a 35-m bundle of 19 each 50-µm-core fibers, a companion 35-m single fiber, a 100-µm-core single fiber (at 4 lengths), and a 50-µm-core single fiber (two samples, 7 lengths). The 50-µm-core fiber was from preform #24; the 100-µm-core fiber was a prototype version having a thick cladding. All of the fibers were developed and manufactured at the Vavilov State Optical Institute, St. Petersburg, Russia. 2. Dispersion measurements were made by propagating a 20-ps 351-nm pulse through the fiber under test and recording the output on an S20 streak camera. The width of the pulse transmitted by the fiber was compared to that of a fraction of the pulse that had propagated over an air path. Values of dispersion were calculated as, D = {radical}(F² - A²) , where F and A are the full widths at half maximum (FWHM) for, respectively, the fiber-path and the air-path streaks. 3. In each of the experiments, the measured dispersion increased with counts in the streak record, which in principle, are proportional to intensity in the fiber. Measured values of dispersion ranged from about 0.6 to 1.0 ps/m for the single fibers. 4. The measured FWHMs of both the fiber-path pulse and the air-path pulse increased with increase in counts in the streak record. The rate of broadening was greatest for the fiber-path pulse, and the broadening of that pulse was the primary cause for the dependence of dispersion on counts in the streak record. Pulse broadening with increase in counts is symptomatic of camera saturation, but it is difficult to understand why saturation should have effected the fiber-path pulses more strongly. 5. There were spatial anomalies in the streak records of the output pulses from some of the fibers. Emission by the bundle of a "doubled" pulse is a primary example. In streaks recorded at about 800 counts, the total duration for the pair of pulses was about 100 ps. The maxima of the pulses occurred in different columns of the streak, so there was a relative spatial or angular offset between the two pulses. Pulses with extended tails were observed in each streak that was recorded at about 400 counts. 6. We frequently had difficulty obtaining adequate transmittance through 50-mm single fibers. Some of our problems probably were related to inexperience in cleaving this particular fiber.