We report and correct a few typographical and scaling errors in our manuscript
© 2013 OSA
In the first paragraph of section 2 of our manuscript , ε 0 should be replaced by εh which was correctly used throughout the procedure. In Eq. (1), k is the wave number in the host medium. As a result C 1, C 2, and C 3 should be multiplied by , and L 1 should be divided by . The procedure and subsequently the equivalent circuit configuration remain unchanged. From Eq. (12), ωCPR does not change. n, the order of the filter, in Eq. (14) should be replaced by i, with no effect on the design procedure. Considering the equation for L′ in the last paragraph of section 2, equations (16) and (17) should be divided by . Thus, in Table 1 all dimensions should be divided by . The recalculated Fig. (8) for F = 0.376 and a 1,2 = 9.5nm and a 3 = 19nm shows that the bandwidth is narrowed but still the same behavior as predicted in Fig. (7) is seen. The angle dependance is the same as described previously in .