LBL Center for X-Ray Optics Researches X-Ray Lithography

January 10, 1992

By Lynn Yarris,

Housed in the Advanced Materials Laboratory, adjacent to LBL's Advanced Light Source, is the Center for X-ray Optics (CXRO), the world's first research center devoted exclusively to the study and utilization of x-ray light. During the past year, scientists at this center won a contract from the Advanced Research Projects Agency (ARPA) to develop a program in one of the prime new critical technologies--x-ray lithography.

Lithography is the process by which a beam of light is used to transfer intricate patterns from a mask onto the surface of a material in order to make a device, such as an integrated circuit. Today's microchips are fashioned with visible light. As tiny and powerful as these devices are, they could be made even tinier and more powerful if fashioned from the much shorter wavelengths of x-rays. It is like carving with a scalpel rather than a sword.

The ARPA program at the CXRO is being supported by nearly $60 million, spread over a five-year period. A large portion of this money will go to the design and construction of a pair of beamlines at the ALS--an undulator beamline for x-ray interferometry and a bending magnet beamline for metrology.

According to CXRO leader David Attwood, the light generated by an ALS undulator with its high spectral brightness and its partial coherence at wavelengths ranging from 45 to 300 angstroms, is a "perfect match" for studying and testing the materials used in x-ray lithography.

Initial plans for the metrology project, which is led by CXRO physicist James Underwood, call for the beamline to be divided into two branches--one for extreme ultraviolet light (25 to 250 angstroms) and the other for soft x-rays (6 to 30 angstroms). Together, these two branches will enable researchers to do a number of metrological activities, such as measure the reflectivity of x-ray mirror coatings or test the sensitivity of various lithography pattern recording materials.

Most of the remainder of the ARPA grant will go towards building a nanostructure fabrication facility. The main component of this facility will be an electron beam writing tool capable of generating a beam 25 angstroms in diameter that will accurately carve patterns over an area of 65 microns. These 65-micron areas can then be stitched together to form patterned areas that approach one square centimeter.

Says Attwood, "We will have full in-house capabilities for nanostructure patterning, synthesis, and processing. We will also be able to provide a very powerful tool-set for scientists who wish to pursue the novel properties of new materials in structures of near-atomic dimensions."

Managing LBL's x-ray lithography program is Keith Jackson, a physicist who came to the CXRO this year from the Rocketdyne Corporation, where he had been developing diamond films and optical elements for synchrotrons and free electron lasers. In addition to his program manager duties, Jackson is also overseeing the design and construction of the project's undulator beamline.

One of Jackson's first objectives as program manager will be to demonstrate the feasibility of projection lithography for making circuit patterns. Current x-ray lithography programs use proximity or shadow printing, where light is beamed through a mask that is essentially in direct contact with the target material. Projection printing offers a number of advantages over this technique, Jackson says, but none more important than the size of the patterns that can be made.

"In projection printing, the patterns that are transferred can be as much as five times smaller than the features on the mask," he says. "This is because imaging optics demagnify or reduce the mask pattern when they project it onto the surface of a semiconductor substrate." In proximity printing, the mask features and the printed patterns are the same size.

Jackson is also developing a proposal for a new program in what is called deep x-ray projection lithography, using a technique known as LIGA (a German acronym for Lithographic Galvanoformung Abformung, or lithography electroplating molding).

Unlike the lithography processes discussed earlier where the action takes place on the surface of a material, deep x-ray lithography, as its name implies, is used to sculpt out material several hundred microns below the surface. As a result, instead of producing devices such as computer chips, which are essentially two-dimensional, truly three-dimensional devices, such as the microscopic machines known as MEMS (for micro- electromechanical systems), can be made.

"Using x-rays at about 5 kilovolts of energy (the kind produced by ALS bending magnets), we get deep exposures into a target material," Jackson says. "If we then tilt the target through a series of such exposures, we can create highly complex structures."

In the LIGA technique, high-energy x-rays penetrate the transparent portion of a lithography mask, the exposed material is removed, and a metal is electroplated into the open space. This produces either a working device or a master mold for use in injection molding.

Currently, there are only two LIGA programs under development in the United States, one at the University of Wisconsin, and one that is just getting stated at Louisiana Tech. The LIGA program that Jackson and his CXRO colleagues are proposing will have the advantage of the ALS as its source of x- rays.

"Light from an ALS bending magnet, with its high flux at 2- to 3-angstrom wavelengths, will enable us to get the exposures needed for a LIGA operation in a matter of minutes, rather than the hours it now takes using x-rays from other synchroton facilities," says Jackson. "Also, unlike other facilities, at the ALS we can operate under atmospheric pressure and with fewer optical elements."

The combination of beams from an ALS bending magnet with the LIGA technique can provide mechanical devices that, according to Jackson, "will enjoy all the manufacturing advantages associated with integrated circuits."

On the LIGA proposal, Jackson and his LBL colleagues are working in close collaboration with Richard Muller, Dick White, and Roger Howe, researchers with the University of California at Berkeley, who are considered pioneers in the field.