Q&A - Made to measure – A life in metrology
Innovation in semiconductor technology is dependent on metrology. The work of Professor Wilfried Vandervorst, Senior Fellow and Department Director of Materials and Components Analysis at Imec, Belgium, has been instrumental to developments in processing. Ledetta Asfa-Wossen takes a look at his career in research.
What led you to enter the field of metrology?
I became interested in metrology for the semiconductor industry because physical analysis is such a mission-critical operation for the development of semiconductor technology. Once I was aware of the needs of this industry, I began to focus my research on metrology concepts for improved composition (surface, thin films) and impurity analysis (dopants, carriers). It’s such a rapidly evolving field that requires a high degree of creativity, innovation and a permanent insight. I guess I took it as an opportunity to differentiate myself and to demonstrate my competencies and visions in various disciplines, such as engineering, physics and materials science.
What excited you about the subject?
During my PhD project, I focused on the underlying physics of secondary ion mass spectrometry (SIMS) and electrical properties of point contacts. Through that I started to examine the metrological aspects, such as accuracy, quantification, and sensitivity and application value.
The excitement came when I was able to tell process and device engineers what they were making instead of what they thought they were making. Linking the results of observing metrology to the fundamentals of processing really excited me. My job ranges from fundamental physics to practical instrument engineering and technological applications.
I get to explore a wide range of metrology and scientific fields in many different disciplines, from ion-solid interactions, electrical measurements and laser-solid interactions to process technology and instrument design. No two days are ever the same.
Why is metrology so crucial to the fabrication of semiconductors?
Metrology is a great enabler. The fundamental insight created by metrology is key in accelerating R&D in process technology. Within the fabrication area, certifying that process tools behave the way they should and stay within their operational window represents an important tool for yield and production efficiency. Understanding the potential role of excursions is also crucial to the time window of yield ramping. There are vast economic benefits for conducting metrology too, specifically in the characterisation of nanoelectronics.
You pioneered device characterisation using scanning spreading resistance microscopy (SSRM). What was novel about this method?
SSRM is the sole method we have for providing carrier profiles with adequate spatial resolution (<1nm) and sensitivity and quantification accuracy. Its uniqueness is the combination of all these properties. At the heart of every transistor lies a complex, engineered, electrically active dopant distribution controlling its electrical performance. As technology progresses, being able to create and control this distribution on the sub-nanometre scale becomes the key to successful device development. SSRM probes the spatial extent of the carrier distribution with nanometre resolution, thereby providing the essential feedback on the dopant incorporation and activation processes used to engineer modern devices.
What direct application did the work have?
When developing SSRM, 2D-profiling was listed as a red brick wall in The International Technology Roadmap for Semiconductors. Our developments removed this deficiency and, essentially, the wall.
What would you say is your greatest achievement so far?
Bringing SSRM from an initial idea to the commercial product market by addressing the fundamental physics as well as the engineering problems, from tips instrument to standards. SSRM is accessible to any user and it is being embraced by all major semiconductor companies, such as Intel, IBM, TSMC and Toshiba.
How would you say metrology in this area has evolved over the last 10 years?
Scanning probe microscopy has become routine, transmission electron microscopy (TEM) a commodity, and secondary ion mass spectrometry now a trustworthy, standard approach. Equally, advanced physical and electrical metrology has moved from blanket film and lab analysis towards fab and device analysis. We are also seeing advanced metrology transitioning from a single experiment and expert-based approach in the R&D phase towards high-volume, easily accessible support in the fab. Obviously, this creates a stronger return on investment by improving productivity and yield ramp, but it also means a more time-critical operation with tighter tolerances on accuracy and reproducibility.
Are there any hindrances to metrology innovation?
The pure lack of dedicated funding opportunities for the science of metrology is an issue. The long development time to build and explore new instrumentation requires financial support that is sustained over many years prior to reaching the expected results. Nowadays, you also need very expensive instrumentation, which often exceeds standard budgets of funding agencies. A TEM or atom probe can cost £1.6–2.4m and it is hard for any researcher to get that kind of money solely for fundamental studies on how to improve metrology concepts.
Why is it difficult to secure funding for metrology compared to other material science fields?
Many reviewers do not appreciate the value of exploring the science and physics of metrology and its instrumentation. They just want to know what device is being developed and how your metrology tool would enable that. Appraisal for the generic work of metrology, including the development of standards, is low and very difficult to finance. Dedicated calls for research on metrology are almost non-existent. Given the small community focusing on metrology it is also difficult to get extensive response to publications and, therefore, high citation indices. The latter plays a negative role when faced with funding organisations using citation indices as selection criteria.
Can it be hard for metrologists to have their accomplishments recognised?
Metrologists operating in industrial or semiconductor metrology development cannot be measured with the standards of academia in, say, physics or chemistry, or with standard evaluation criteria used by many funding agencies and reviewers. Industrial metrology scientists can only publish a small part of their expertise and innovation in papers as, by nature, a multi-target and multimode operation using more than 10 different physical instrumentation systems is kept confidential as a unique resource for the research centre or company.
Moreover, due to the interdisciplinary nature of the subject, papers rarely fit well in highly rated pure physics or materials journals. Finally, the citation index (h-factor) might not be as high as in more physics/materials oriented fields, as the population of metrologists is substantially smaller (and the fraction that are allowed to publish is even smaller), making the target of a comparable h-index (>40–50) extremely difficult. The need to cover research with a substantial impact on industry and return on investment is also a publication strategy ruling journals across the industry.
In addition, as soon as a metrology concept reaches a useful state and is explored in materials science R&D, it moves to a supporting state, with the metrologist being added to the author list but no longer being the lead author.
What do you see as the focus areas for metrology in this field over the next 10 years?
Technologically, the largest challenge is in 3D-device structures where the most important information to be retrieved relates to interfacial properties. This means measuring in a very confined volume with 3D-spatial resolution.
What is currently the most significant development in metrology?
Faced with the evolution towards 3D-devices and confined volumes, true 3D metrology has become a must. Atom probe tomography (APT) is emerging as the most powerful solution in that respect as it provides 3D near-atomic resolution and is tuned towards the next generation of 3D-devices. For a metrology specialist, these are exciting times. APT has the potential to unravel atomic scale materials interactions and composition variations. Despite the many challenges it still faces, the reward in terms of unparalleled metrological power is astonishing.
What advice would you give to a student entering metrology?
Adopt a broad knowledge of many metrology concepts and explore these. A single method will rarely be a universal solution. To have a strong impact you need to understand the technology you are studying as well.
What attributes do you think a metrologist should have?
A broad technological insight and an analytical mind. And patience, persistence and willingness to network within the metrology community and participate in conferences, as they are a tremendous source of information.
Career top tip
A key lesson in my career has been to focus not only on the metrology aspects of semiconductor technology, but equally to build on fundamental insights related to process steps and material development. Metrology is not just about excellence in measurement, but being a valued partner in materials research and process development. I think it’s important for metrologists to have continuous interaction with process engineers and materials scientists to understand their demands and any deficiencies of available metrology.
Awarded the Imec prize for outstanding achievement for the creation of a Materials Characterisation Laboratory, acting as a Centre of Excellence in materials and process characterisation 2001 Nominated Imec Fellow
2013 Appointed Senior Fellow for ‘exceptional scientific contribution to the materials and component analysis for R&D in advanced semiconductor process technologies’. The Imec Fellowship has been granted to only 10 researchers.
630 papers published