|
Kenneth A. Goldberg, Ph.D., Optical Physicist Center for X-Ray Optics • Lawrence Berkeley National Laboratory Extreme Ultraviolet Optics • Ultra-High-Accuracy Interferometry • EUV Lithography EUV 'Actinic' Mask Inspection • Optical System Modeling • Synchrotron Light • Coherence |
![]() |
| INTERFEROMETRY |

The interference of two coherent waves.
Interferometry is the science, and the art, of making measurements with coherent light. Light moves in waves, and light from a single source can travel different paths and then recombine in ways we call interference. When we measure that interference, we learn about the different paths the light has taken; and the yard-stick, that is, the length-scale of the measurement, is the wavelength of light itself.
One of the primary applications of interferometry is in testing, polishing, and aligning very high quality lenses, like those in cameras, telescopes, and photo-lithography tools called steppers, which are used in the fabrication of the intricate circuit patterns on computer chips.
In principle, if you can measure with a smaller wavelength of light, you can measure with finer steps. But there are many reasons why you'd want to use a smaller wavelength of light in your measurement. Some lenses and mirrors are designed specifically to focus ultraviolet light, extreme ultraviolet light, so-called soft x-rays, and even x-rays. Those special lenses often focus different colors of light (i.e. different wavelengths) differently. So to know how a lens or mirror will perform for a given wavelength, it is essential to test the lens using that same wavelength.
Since 1993, I have been working to new ways to measure lenses using extreme ultraviolet light. These special lenses may be the highest quality light-focusing optics ever made, and the measurements I perform are the most accurate wavefront measuring tests of their kind. By adapting techniques from the earliest types of interferometry, and inventing many new methods, our measurement accuracy reaches levels smaller than the radius of a hydrogen atom, below 0.5-Å, or 50 picometers, 50 trillionths of a meter.
In close collaboration with semiconductor industry sponsors, including SEMATECH and the EUV LLC, and in cooperation with scientists at LBNL, and other national laboratories, including Lawrence Livermore and Sandia National Laboratories, my team has measured nine prototype EUV lithography lenses, capable of lithography down to 12-nm line width. These lenses reached diffraction-limited wavefront quality with aberrations as small as 0.5-nm RMS.
We use several different interferometer configurations to perform measurements. Unlike most visible-light interferometers, we measure the full pupil without null-compensating, or re-imaging lenses, because such EUV lenses of the required high quality are simply unavailable. For that reason, we are extremely sensitive to the geometry of the measurement, the propagation of the interfering waves to the detector, and we require very careful calibration to achieve high accuracy. Furthermore, since EUV light sources typically have a short coherence length, the so-called common-path interferometer designs are the most suitable.
Each test begins with Foucault testing, also known as the knife-edge test. We also use lateral shearing interferometry using a cross-grating to perform a variation of the Ronchi test with two simultaneous measurement directions. Our most accurate measurement technique is a significant improvement over the point-diffraction interferometer (also known as the Linnik interferometer or the Smartt Interferometer) and was invented by Hector Medecki in the early days of EUV interferometry at LBNL. It is called the phase-shifting point-diffraction interferometer (PS/PDI). It uses pinhole diffraction to create spherical reference waves of extremely high accuracy, and a transmission-grating beam-splitter to separate the test and reference beams. In many ways, the PS/PDI has been our most successful interferometer design. However, while the PS/PDI offers high accuracy, the shearing techniques offer high efficiency and ease of alignment.
Besides myself, a number of other researchers have made essential contributions to this work. They include Hector Medecki, Jeffrey Bokor, David Attwood, Gary Sommargren, Patrick Naulleau, Senajith Rekawa, Paul Denham, Erik Anderson, J. Alex Liddle, Keith Jackson, Phil Batson, Matt Bjork, and Edita Tejnil. Other researchers who have played important roles include John Taylor, Henry Chapman, SangHun Lee, Raul Beguiristain, and Michael Shumway. Essential engineering and technical contributions have been made by Robert Gunion, C. Drew Kemp, Rene Delano, Brian Hoef, Gideon Jones, Deirdre Olynick, Bruce Harteneck, Farhad Salmassi, Ron Tackaberry, Jeff Gamsby, and others. Expert guidance and support has come from our SEMATECH project leader, Kim Dean, and formerly Pat Gabella as well.
"Testing EUV optics with EUV light: If you can measure it, you can make it," Kenneth Goldberg, SPIE Newsroom: Optical Design & Engineering (2006). [LINK]
"At-Wavelength Alignment and Testing of the 0.3 NA MET Optic," K. A. Goldberg, P. Naulleau, et al., JVST B 22 (6), 2956-61 (2004). [LINK]
"Extreme-ultraviolet phase-shifting point-diffraction interferometer: a wave-front metrology tool with subangstrom reference-wave accuracy," P. P. Naulleau, K. A. Goldberg, et al., Applied Optics 38 (35), 7252-63 (1999).
"A Phase-Shifting Point Diffraction Interferometer," H. Medecki, E. Tejnil, K. A. Goldberg, and J. Bokor, Optics Letters 21 (19), 1526-8 (1996). [LINK]
To learn more about this research, please browse the publication list where a number papers are available, view slides from recent public presentations, or contact me. Wikipedia.org has entries for Photolithography and Interferometry.




