Comparison of high resolution negative electron beam resists
SEM image of lines with a pitch of written in TEBN-1 and transferred into silicon by the RIE process listed on Table II. Any remaining TEBN-1 after pattern transfer has been removed by oxygen plasma. The inset shows a zoom-out of the etched lines.
Normalized dose curves for each of the four investigated resists: SU-8 2000, mr-L 6000.1XP, ma-N 2401XP, and TEBN-1. The data points represent the average resist height measured by atomic force microscopy along a structure. The standard deviation on the data points are less than the size of the legend.
SEM images of patterns written in mr-L 6000.1XP and transferred into silicon. Remaining mr-L 6000.1XP has been removed by oxygen plasma. Average etch depth of . (A) lines with a pitch of . (B) lines with a pitch of . (C) Minimum obtained linewidth is with a pitch down to . (D) Minimum obtained pitch is with a linewidth down to .
SEM images of patterns written in TEBN-1 and transferred into silicon by the RIE process outlined in Table II. Remaining TEBN-1 has been removed by oxygen plasma (A) (top left part), (bottom left part), (bottom right part), and (top right part) half-pitch lines written in TEBN-1 and etched into silicon. (B) grid with a period of written in TEBN-1 and etched into silicon. (C) Minimum obtained linewidth with TEBN-1 is with a pitch down to written in TEBN-1. Average etch depth of . (D) Minimum obtained pitch is with a linewidth down to ( half-pitch lines). Average etch depth of .
Processing and results summary. The four resists were spin coated onto silicon substrate and prebaked to remove the solvent. The film properties (average thickness and standard deviation) were measured with a Dektak 8 profilometer across a wafer. The electron beam exposure was performed in a Jeol JBX9300FS tool, and the sensitivity and contrast of each resist were determined from the dose curves in Fig. 2. After exposure the two chemically amplified resists, SU-8 2000 series and mr-L 6000.1XP, were postexposure baked to cross-link the exposed areas. Each resist was developed by conventional dip developed in suitable developers. After development the written patterns were transferred into a silicon substrate by the reactive ion etch process shown in Table II, and the minimum obtained resolution was characterized by scanning electron microscopy (see Figs. 3 and 4).
Reactive ion etch process used to transfer the written patterns into the silicon substrate. The process is optimized with regard to vertical sidewalls and smooth surfaces (see Fig. 1).
Article metrics loading...
Full text loading...