Bryan Barnes (Fed)

Dr. Barnes is a Physicist in the Nanoscale Device Characterization Division (NDCD) of the Physical Measurement Laboratory (PML) at the National Institute of Standards and Technology (NIST). He develops new approaches to optical microscopy and electromagnetic modeling to enable improved metrology of nanoscale structures with dimensions more than an order of magnitude below traditional resolution limits.  New applications and standards produced from these methods include defect inspection and critical dimension metrology for nanoelectronics. His current research seeks solutions to anticipated limitations of optical measurements of nanometer scale features over large areas, which are critical for the effective manufacturing process control of products that incorporate billions of nanoscale features.  Methods currently studied to extend optical microscopy to near-atomic scales include atomistic simulation of dielectric constants, machine learning for defect inspection, and reduced wavelengths beyond the deep-ultraviolet.  He also provides key optical measurements of overlay offset between subsequent patterned layers in support of NDCD’s Atom Device Group.

Dr. Barnes has authored more than 20 peer-reviewed journal articles and 40 conference publications, presents invited talks at both academic and industrial conferences, and holds 1 patent.  He serves as the Chair of the North American Chapter of the SEMI Standards Microlithography Committee and is a Voting Member of the SEMI North American Regional Standards Committee and is a member of the IEEE and APS.  Most recently, he is a co-recipient of a 2016 Department of Commerce Silver Medal “for pioneering advances in optics, imaging structures 30 times smaller than the wavelength of light with near atomic accuracy.”  He is also a co-recipient of the 2013 R&D 100 Award for "Quantitative Hybrid Metrology," a new method that enhances multiple measuring instruments by tying them together statistically in novel combinations. Dr. Barnes leads the Quantitative Nanoscale Imaging Through Artificial Intelligence project and has recently successfully completed the Optical Methods for 3-D Nanostructure Metrology project.

Google Scholar Profile

Research Opportunities

I am seeking a post-doctoral researcher to augment our current research, which concentrates on new approaches to extending optical capabilities for the characterization of nanoscale devices as they increase in complexity, with challenging new materials properties, thicknesses, and length scales that defy simplistic applications of the fundamental equations of electromagnetism.  Post-doctoral opportunities are available presently in the following focus area: 

Quantitative Nanoscale Imaging through Artificial Intelligence  - We seek to expand optics-based measurement science not only to the end of the nanoscale device “roadmap” below 3 nm in critical dimension but to develop the underlying enabling characterization technologies over large areas for QIS-fostering elements such as two-dimensional materials and atom-based devices. Increasing intricacy coupled with the thickness dependence of the dielectric function (DF) below 5 nm will require a critical understanding of both the fundamental behavior of light reflecting and scattering off ultrathin films and structures as well as data-driven methods to fully characterize near-atomic sized features. Thus, as the electromagnetic modelling undergirding key industrial nanoscale measurements grows ever more complex with some adopting physics-free machine learning (ML), measurement strategies are to be developed to harness the potential of ML while both validating accuracy and uncertainty and enabling the integration of physics-based a priori information. We are developing quantitative approaches for interpreting optical scattering and imaging of nanoscale devices though the use of machine learning (ML). From the literature, nanoscale device characterization is already being performed in manufacturing using physics-free, algorithmic approaches with reports of improved uncertainty. None of these reports however make clear how accuracy is achieved nor what relationship exists between their uncertainty and accepted uncertainty quantification practices. This goal is also pursued through the development of advanced ML for the detection and potential identification of patterned and unpatterned defects.

Areas of expertise that would be of great benefit include the following:

• Successful coding and documentation of convolutional neural networks (CNNs) and/or Gaussian processes or “Kriging”
• Coding skills in python and MATLAB
• Familiarity with Bayesian statistics
• Electromagnetic scattering calculations and/or theory
• Optical properties determination of ultrathin films and features

If you have a subset of these skills and are seeking a post-doctoral position, please contact bryan.barnes [at] nist.gov to discuss in more detail the opportunities available to join our ongoing work.   US citizenship is not required for this opportunity. However, additional opportunities are available to U.S. Citizens through the NRC/NIST Postdoctoral Program (Accurately Establishing Uncertainties from Artificial Intelligence and Actinic Optical Dimensional Characterization of Deep-Subwavelength Nanostructures).