Integration of Fabry-Perot and Bragg grating captures best of each.
Bruce H. Walker, Walker Associates
Researchers at the University of Central Florida and at OptiGrate, both in Orlando, have demonstrated how a Fabry-Perot etalon and a volume Bragg grating can be arranged sequentially to provide a filter with the narrow spectral bandwidth of the etalon and the broad rejection band of the grating (Figure 1). And because the transmission of the etalon matched the channels of the International Telecommunications Union grid, the filter could be tuned to isolate any single channel of the grid.
Figure 1. The etalon transmits a comb of frequencies, and the grating reflects only one of the frequencies to the detector.
For their etalon, the investigators used a 1.036-mm-thick fused silica plate with an ∼43 percent reflecting dielectric coating on each face. The Bragg grating, manufactured by OptiGrate, was recorded in photothermal-refractive glass with a process consisting of ultraviolet exposure followed by thermal development. By rotating the grating on a precision rotation stage, they could tune its transmission from 1530 to 1552 nm (Figure 2).
Figure 2. The researchers tuned the reflection of the Bragg grating from 1530 to 1552 nm by rotating it (θ = 3.06°, 4.13°, 10.79° and 14.70° curves 1 through 4, respectively). The transmission of the Fabry-Perot also is shown (curve 5).
They adjusted the grating so that its transmission matched one of the transmission peaks of the etalon. The spectral selectivity of the Bragg grating in this experiment was 320 pm, and that of the etalon, 220 pm, so the selectivity of the combined filter was the etalon’s 220 pm.
Figure 3. By combining the Bragg grating and the etalon as shown in Figure 1, the researchers could select a single transmission peak of the etalon as the pass frequency of the filter. The four numbered peaks here correspond to the same grating angles indicated in Figure 2.
With a different etalon, the scientists believe that their filter could have much higher selectivity. The limit for an etalon, they calculate, is ∼10 pm because of imperfect flatness and imperfect parallelism of the mirrors and because of the divergence of the incoming beam.
Optics Letters, Aug., 15, 2006, pp.2417-2419.