Experiments have shown that a substrate can be patterned with submicron features by the following procedure:

  1. Coat the substrate with a layer of poly(methyl methacrylate) [PMMA], which will serve as an electron-beam resist to be patterned into an etch mask.
  2. Pattern the PMMA by electron-beam lithography.
  3. Ion-etch through the PMMA to transfer the pattern to the substrate.

The procedure has been used to fabricate a 600-nm-pitch diffraction grating; it could also be used to form submicron features on semiconductor and other devices.

Prior to the development of this procedure, it was known that submicron features can be produced in a PMMA resist by exposing it to an electron beam, then developing it. However, the transfer of submicron features from a PMMA mask to a substrate by either dry or wet etching had not been demonstrated. As a result of the development of the present procedure, it is now known that success in the transfer of submicron features depends on the proper choice of the substrate material(s), the thicknesses of the substrate and PMMA, the etching method, and the etching conditions.

The substrate for the 600-nm-pitch grating was a 100-nm-thick gold film on a 5-nm-thick chromium layer on a quartz disk; the Cr and Au films were deposited on the disk in an electron-beam evaporation system. A PMMA resist 120 nm thick was spun onto the Au surface.

The 600-nm-pitch grating pattern was imparted to the PMMA by use of an electron-beam lithography machine. The electron-beam dose in the exposed regions was about 200 µC/cm2, and the beam current was 5 nA. Following the electron-beam exposure, the pattern in the resist was developed by use of a 1:3 mixture of methyl isobutyl ketone and isopropyl alcohol.

The ion etching technique used to transfer the pattern to the substrate could be characterized, more precisely, as milling by a beam of argon ions. The etching was performed in an argon-ion-milling system that included an ion-beam source in a vacuum chamber that was capable of evacuation to a base pressure of 4 × 10 -8 torr (≈5 µPa). During operation of the ion-beam source, the pressure of argon in the system was raised to 2 ×10 -4 torr (≈0.03 Pa) by addition of argon gas to the chamber at a controlled rate. This operating pressure was chosen to be high enough to generate a sufficient population of ions to obtain the desired beam current density, but not so high that the beam would be excessively dispersed through collisions with gas molecules.

The ion-beam-generating potential was 1.5 kV. The total beam current was 10 mA. The beam was spread over a cross section approximately 5 cm in diameter. The etching time was about 2 minutes.

This work was done by Thomas George, Richard Muller, Paul Maker, and David Barsic of Caltech for NASA's Jet Propulsion Laboratory.