Surface precipitates on single crystal LiNbO 3 after dry-etching by CHF 3 plasma
- PDF / 152,064 Bytes
- 3 Pages / 612 x 792 pts (letter) Page_size
- 39 Downloads / 231 Views
MATERIALS RESEARCH
Welcome
Comments
Help
Surface precipitates on single crystal LiNbO3 after dry-etching by CHF3 plasma Kaori Shima Advanced Materials Research Division, New Technology Research Laboratories, Sumitomo Osaka Cement Co., Ltd., 585 Toyotomi-cho, Funabashi-shi, Chiba 274, Japan
Naoki Mitsugi and Hirotoshi Nagata Optoelectronics Research Division, New Technology Research Laboratories, Sumitomo Osaka Cement Co., Ltd., 585 Toyotomi-cho, Funabashi-shi, Chiba 274, Japan (Received 19 February 1997; accepted 5 August 1997)
The CHF3 electron cyclotron resonance (ECR) plasma etched LiNbO3 (LN) surface was analyzed chemically and crystallographically to investigate the dry-etch machining process for LN crystal, which was recently needed to obtain broader-band optical modulators. The etched surface was entirely covered with amorphous-like precipitates having ,70 nm diameter. These precipitates (or a part of them) were thought to be LiF from Auger electron and x-ray photoelectron spectroscopy. The results indicated that the LiF was formed and remained on the etched surface while the Nb was almost completely removed.
In order to expand the optical bandwidth of LiNbO3 (LN) based modulators, the plasma dry-etching technique is applied to fabricate a trench structure on the LN substrate surface, in which the LN is removed to a depth of about 3 mm along the optical waveguide.1,2 Such a structure effectively reduces the permittivity of the substrate, broadening the bandwidth while reducing the driving voltage. Dry-etching of the LN is performed fluorocarbons such as CF4 and CHF3 rather than Ar, in order to prevent surface damage by high energy ions.1,2 Jackel et al. reported originally the reactive ion etching of the LN and suggested the possibility of the formation of a thermally stable compound of Li with halogen during the chemical etching process.3 They recommended mixing Ar into the chemical etching gases to physically remove the Li-halides. However, to our knowledge, there has been no report to show any evidence for the existence of the Li-halide layer due to the dry-etching. Here, we investigated the LN surface etched by an electron cyclotron resonance (ECR) plasma of CHF3 and confirmed the formation of LiF on the surface. The CHF3 was chosen in this experiment because the coexistence of F and H was commonly known to be suitable for the etching of oxide materials.4 Commercial X-cut (crystallographic a-face) and Z-cut (c-face) LN substrates with an optically flat surface were treated by an ANELVA L-310R ECR plasma etcher. After the substrate was inserted, the etching chamber was evacuated to about 3 3 1025 Pa by a cryopump. Then, CHF3 was introduced into the chamber at a rate of 3 sccm, and the pressure was kept at about 0.01 Pa. Microwave power and high voltages applied to the magnet and bias electrode were chosen to J. Mater. Res., Vol. 13, No. 3, Mar 1998
http://journals.cambridge.org
Downloaded: 14 Mar 2015
generate a stable CHF3 plasma and to obtain a higher etching rate. The substrate holder, made o
Data Loading...
