DC-drift suppression of Ti: LiNbO3 waveguide chip by minimizing the contamination in oxide buffer layer

Analysis and Control of the DC Drift in LiNbO3Based Mach–Zehnder Modulators

“…In the short time range, the amplitude of is determined by the relative anisotropy of the dielectric permittivity ratio (see Appendix IV). This is in partial contradiction with the observations of Nagata et al [15] who have highlighted a correlation between the magnitude of the short dc drift and the quality of the SiO buffer layer altered by unwanted contaminants such as Na and K ions [17], [18] easily incorporated from the photoresist developers used during the patterning process, Li ions from LiNbO substrates, which may diffuse throughout the SiO layer during the heating process, and OH ions adsorbed from room atmosphere [19] in the silica buffer. Nevertheless, as shown below (see Section III-B), this correlation is more probably due to the modification of the LN substrate during the process (out-diffusion of Li ions) than the alteration of the silica buffer.…”

“…Therefore, among several different solutions, Kim et al [17] have proposed a technique to remove, using a chemical-cleaning process, all the alkali contaminants (especially Na and K ions) from the buffer layer, and to diminish drastically the OH ions contents with an adapted annealing process. To reduce the mobility of the Li ions, Jin et al [20] have used F ions implanted SiO layer.…”

Study of modulation phase drift in an interferometric fiber optic gyroscope

“…Many researches on LiNbO 3 optical intensity modulators and switches have shown that the SiO 2 buffer layer is subject to contaminants such as OH − groups and alkali-metal ions from the photoresist developers. [6][7][8] It is well known that these OH − groups and alkali-metal ions are incorporated into SiO 2 as network modifiers and act as mobile species. As shown in Fig.…”