I was reading an AFM site and saw a contrived claim trying to make a comparison that was over 50% off the benchmark and thought it would be a good topic to open up for discussion. The comment was to the effect of: by scanning at a slower scan speed we are actually going faster because our poor scanner dynamics require so much rounding we have to make up for it by moving faster through the linear range - even though the image will take longer to acquire (because frame rate is determined by Lines/ScanFrequency, so the slower scan rate will always make a slower image. . .)
It got me thinking about how the community defines AFM speed. From the UIUC workshop last week, I hear Closed-Loop Transfer Function at constant force is emerging as the agreed figure of merit, and this is a tremendously positive step in providing transparency to the community. But the comment above, makes me think something is missing – namely Productivity.
How fast can you get your publishable data? That is the fundamental question facing the scientists that use our systems, and a critical question for defining fast AFM. In fact, it was the basis for one of the Grand Challenges we gave to ourselves: Image 3 demanding samples in under 300s. A normal AFM image takes 10-15minutes each, so everyone on this list will know the inherent difficulty in this challenge, and what it would mean to solve it.
Our samples were C60H122, SBS, and Celgard® and we did it in just over 200s. You can see the video here, we cut out some of the engage bits, but the clock at the bottom keeps true time. I can send the long form version to anyone interested:
Enjoy and comments welcome,
Steve
Celgard® is a registered trademark of Celgard, LLC and that Celgard, LLC is not associated with Bruker Nano.