10nm Issue? Metrology problems at the Supply side.

[Francister]

I am interested in comments. Problem statement is metrology for materials, chemicals, Gases, Precursors, etc required for 10nm device technology. I am aware of the Metrology advancements that we are using for looking at patterned wafers, mask etc. in terms of defects, variability, film characteristics, etc. FCMN 2015, SPIE, IEEE conference and the advancements. I am still studying the different metrology we are using on bare wafers, wafer cleans, etc. down to 10e-9 levels The problem is when I work on correlating this to the anaylical techniques needs at the material level that ensure process performance at 10nm, I do not see the level of advancement required. Tomography, Particles size, Grain size control, Precursor and material purity control metrology are not up to the requirements. Problems with impurities or minor species in a chemicals, materials and precursors that are not specified seem to be causing more problems for the device manufactures. What are your thoughts?

 

[benb]

Here's a little known fact: You can't win a Nobel Prize for analytical chemistry. It isn't considered a science, but rather an art. That neatly characterizes part of the problem you describe.

Chemistry has long operated at 10E-9 scale, in research. Translating that into reliable, online 24-7 operation required in manufacturing is an art. It takes time and money; researchers and engineers working together. There just aren't many organizations willing to do this. When it is done, it involves fiercely limited scope and time and budget. And the budgets get blown away inevitably, with little hope of ROI.

So the problem is really economics, not capability.

 

[chris.sparks]

Here's a little known fact: You can't win a Nobel Prize for analytical chemistry. It isn't considered a science, but rather an art. That neatly characterizes part of the problem you describe.

While I partially agree with the sentiment, the analytical chemist in me has to quibble with this broad statement.
http://www.nobelprize.org/<wbr>nobel_prizes/chemistry/fields.<wbr>html
Analytical chemistry has a few prizes noted. I would also consider the 2014 prize for super resolved fluorescence microscopy as analytical. Don't forget all the physicists who won for their work with x-rays leading to spectroscopic methods.

I think there's hope though for production, at least in some areas of analysis. For example, Elemental Scientific (no association on my part, just interested) has the Scout for automated process monitoring with ICP-MS which looks promising.

 

[Francister]

President

I am interested in comments. Problem statement is metrology for materials, chemicals, Gases, Precursors, etc required for 10nm device technology. I am aware of the Metrology advancements that we are using for looking at patterned wafers, mask etc. in terms of defects, variability, film characteristics, etc. FCMN 2015, SPIE, IEEE conference and the advancements. I am still studying the different metrology we are using on bare wafers, wafer cleans, etc. down to 10e-9 levels The problem is when I work on correlating this to the anaylical techniques needs at the material level that ensure process performance at 10nm, I do not see the level of advancement required. Tomography, Particles size, Grain size control, Precursor and material purity control metrology are not up to the requirements. Problems with impurities or minor species in a chemicals, materials and precursors that are not specified seem to be causing more problems for the device manufactures. What are your thoughts?

I should add that specialized techniques of SPM can give you some great result but you need to be very precise on technique and know what you are looking for.

 

[mhagmann]

Since you mention SPM; at the 10-nm node we reach a point where scanning spreading resistance microscopy (SSRM) can no longer satisfy the rule-of-thumb that the resolution in carrier profiling be finer than 10 percent of the node dimension. Only under ideal conditions can a resolution of 1-nm be achieved with SSRM and I question the validity of resistance measurements over a distance smaller than the mean-free-path for the carriers and the morphology is changed by the probe. Furthermore, random dopant fluctuations (RDF) potentiate the need for finer resolution below the 22-nm node. For comparison, scanning capacitance microscopy (SCM) is not appropriate below the 45-nm node. I believe that there is considerable need for improved methods of metrology below the 22-nm node.