Reference Metrology=Red Bricks?

Traditionally, the International Technology Roadmap for Semiconductors (ITRS) has marked future key requirements and challenges to semiconductor technology by printing them in bright crimson — a feature that has earned them the "Red Brick Wall" nickname. Reference metrology is now rapidly becoming the next "Great Wall of Silicon Technology."

At a panel discussion about this subject that took place during February's SPIE Advanced Lithography Conference, reference metrology was identified as a major challenge facing the measurement capability analysis area. The event's description put it best, "The rapid introduction of new instruments, and the increased complexity of such instruments, makes it difficult to evaluate their capability without having a reference instrument with sensitivity over the same range. In areas such as film metrology, the number of measurands and instruments with fundamentally different detection mechanisms increases the need for a reference. Applications such as optical proximity correction verification and the introduction of design-based metrology methods highlight the need for having a baseline reference for a range of applications."

Ronald Dixson, one of the panel's moderators and senior scientist at the dimensional metrology effort at the Precision Engineering Division (PED; one of five divisions of the Manufacturing Engineering Laboratory) of the National Institute of Standards and Technology (NIST, Gaithersburg, Md,), discussed the problem's source, observing that because of the industry's rapid progress and unique metrology requirements, relevant standards are often unavailable. Consequently, there often is no traceable linkage between the realization of the SI (Systeme International d'Unites, or International System of Units) unit of length — the meter — and measurements in the fab line.

Reference metrology is crucial for advanced technologies such as nanotech, where quantum dimensional effects are becoming problematical. The industry is increasingly requiring some form of reference metrology and, although nanotech will have grown to a revenue level comparable with that of mainstream semiconductors over the next five years or so, there is insufficient activity in reference metrology, particularly as it refers to nanotech, although it is even more crucial for its continued development and continuity.

Reference metrology is difficult and costly because it is extremely complicated and must be 10× better than baseline metrology. To add to this metrological Gordian knot, the reference metrology market is relatively small and returns are very low, making finding investors or venture capitalists as easy as getting a good home loan.

Presently, the need for reference metrology is beginning to make itself felt, with criticality at a near horizon. The consensus between metrologists is that a group of experts from industry, academia and government is urgently needed to address the problem. For some time now, Sematech's Advanced Metrology Advisory Group (AMAG) has been concerned over the status of reference metrology, as have been NIST and IMEC; meanwhile, IBM, Texas Instruments and other companies are also seeing the writing on the wall.

A larger, more concerted effort is needed now to define what the needs will be over the next 5-10 years, how long-term research is to be done, and how assistance is to be afforded to start-ups or small companies with good technology.

A good start is scheduled to take place during Scanning 2008, a conference to be held at NIST this month, sponsored by SPIE. Michael Postek, a supervisory physical scientist at the PED, suggested that discussions take place at the conference's scanning probe microscopy session. Those invited include metrology leaders worldwide and representatives of NIST's German and Japanese counterparts, as well as IMEC and LETI. The idea is to form some sort of a working group to coordinate different activities in this complex field.

It behooves all of us who can attend this conference to participate in the scheduled roundtable discussion — everybody is welcome. Viable solutions are needed for the future, and we will not get them through a one company or organization effort — it must be industry-wide because it involves the creation of a vast infrastructure and understanding at the public, government and industry levels that, without reference metrology, the semiconductor industry as a whole — and nanotech in particular — will be unable to attain its future goals.

This will not be simple or quickly done; the time to start is now.