The quality of any AFM image is dependent on a number of factors, but most importantly the condition of the nanoscale AFM probe tip.
We all know the frustration of having to pause work whilst we change a probe in the lab; put that same issue into AFM being used for the large scale production of semiconductor devices, and the ramifications are magnified.
During semiconductor fabrication, AFM is used to confirm critical parameters at key points in the production process – be that depth measurements after an etch process, surface roughness of a thin film, lateral measurements of a transistor gate or otherwise. These measurements often need a quick turnaround to ensure no delay in production flow. Consistent measurements are crucial to ensure processes are monitored accurately without inducing any incorrect process adjustments which could lead to wafers being scrapped.
Having to change probes slows down the production team – not only for the 10 minutes it takes to make the change but also an additional time spent calibrating the tool with the new probe, in order to confirm it is in line with the historical data for the given probe model. If this action has to be repeated multiple times a day in a fab, production is slowed and valuable labour and equipment resources are wasted.
The key reason probes need changing is because they deteriorate over time. AFM probes are rarely indestructible!
Since the inception of the AFM, work has been undertaken looking at the lifetime expectancy of the probe tips. The material of the tip and of the sample being imaged are both factors affecting the outcome, together with the specific imaging mode used. This paper by Associate Professor Koo-Hyun Chung, University of Uslan, South Korea, Wear characteristics of atomic force microscopy tips: A review [1] provides a comprehensive review of the research into this subject and conclusions reached.
Reliability is NuNano’s No. 1 priority, so we were keen to capture data on the lifetime performance of our own AFM probes by imaging semiconductor-relevant samples.
By partnering with scientists at the UK’s National Physical Laboratory (NPL) we’ve developed and begun executing these tests for both surface roughness characterization and known topography measurements. Some really interesting and useful data is emerging from our initial investigations.
These tests are still ongoing, to ensure robust experimental data and statistical relevance, but we thought you might like to see some of the images of our AFM probe tips captured before and after some of these tests.
SEM image of a NuNano SCOUT 70 AFM probe tip before use. The tip apex is measured to have a radius of curvature of only 4 nm.
SEM image of a NuNano SCOUT 70 AFM probe tip after use in AC (tapping) mode. The tip apex has broadened during use, with the radius now measured to be 18 nm.
In this image a ball of material has formed on the tip apex, this could be contamination accumulated during scanning or silicon from the tip reforming as a result of the forces incurred during imaging.
This example shows the tip becoming chamfered to an angle of 12° – the same angle as the probe is mounted in the instrument. This demonstrates the impact of prolonged tapping on a sample surface, since the flat section of the tip would be parallel to the sample surface. A build-up of tip material or contamination can be seen on the righthand side of the tip.
As with all good science, every question asked creates yet more questions to be answered in the pursuit of robust, reliable and most of all meaningful data. But even these initial images perfectly showcase how the nanoscale tip of an AFM probe can change over time.
Having a good grasp of how the tip evolves during use is crucial to ensure that the data and images captured by the AFM are accurate and consistent, and not impacted by deformation of the tip apex over time. Clearly this is especially important in a production environment, where the AFM output directly impacts production flows, material use and labour efficiencies.
Watch this space for more updates on our findings over the coming months.
[1] Int. J. Precis. Eng. Manuf. 15, 2219–2230 (2014). https://doi.org/10.1007/s12541-014-0584-6