Questions about resolution and noise rejection come up a lot in AFM. At Asylum Research, scientists and engineers have explored this topic in depth and developed their Cypher range of atomic force microscopes to achieve even atomic resolution in labs challenged by vibrational noise. In this article, we will explore why Cypher performs so well and give one amazing example of its stability.
Considerations for Design
AFMs are mechanical instruments and are made of materials that distort their shape in response to external forces; in particular, sound and vibration. These distortions have the ability to cause noise, drift and damage of the tip and/or the sample surface. During the early stages of designing the Cypher AFM, considerations were given to the development of a microscope that would be as immune to these external vibrations as possible.
To achieve this, several prototypes were built in order to explore different ways of improving the resolution and stability of the instrument. Developers particularly tried to identify the main weak point(s) to external vibrations.
As shown in Figure 1, the AFM was divided into three separate components. The first was the optical detection system, the second the cantilever and the third, the sample. The question to be answered was simple: ‘Which of the three boundaries—between the detector and cantilever, the detector and sample or the cantilever and sample—was more vulnerable to external vibration noise?’
Figure 1. The three basic components of the AFM: the detector (typically an optical beam deflection sensor), the cantilever and the sample.
The cantilever detection system is similar in most AFMs and is an optical beam deflection system that contains a mirror-like surface on the back of the cantilever, a laser that reflects from this surface, and a position-sensitive detector (PSD) that receives the reflected laser beam.
When the cantilever distorts in any way, the light on the PSD is changed, and this motion is converted into a voltage. This is then used to infer the motion of the tip at the end of the cantilever. The cantilever acts as a carrier for a sharp tip that interacts with the sample surface. The sample is scanned in three dimensions relative to the detector and the cantilever.
Figure 2 shows the conventional arrangement of AFM systems with the cantilever and the optical detection system closely coupled. They are usually built into the same structure (typically called the “head”). At first glance, this is a natural and functional arrangement. The head is a simple unit that turns cantilever deflection into a voltage signal which is used to run the microscope and to measure tip-sample interactions. Initially, it was discovered that the weak link in the AFM mechanics was at the boundary between the cantilever and the sample.
To minimize the effects of this weak link, Asylum Research designed a system with two separate translation stages. The stages were placed between the cantilever and sample, and between the optical detector and the sample, as seen in Figure 2b below.
The mechanical loop between the cantilever and sample, and the predominant source of vibrational sensitivity, is then able to be as short and rigid as possible, thus minimizing susceptibility of vibrations at the tip-sample junction. The second stage transports the relatively high mass of the optical detector and isolates it from the tip-sample junction.
Figure 2. a) Arrangement for conventional AFMs, including most commercially available AFMs, where the optical detector and the cantilever are rigidly coupled together on a single z-translation stage, as seen in red. b) Dual-stage arrangement for the Cypher family SPMs, with a short, stiff cantilever stage, as seen in blue, and a separate optical detector stage, as shown in red.
Figure 3 qualitatively demonstrates the ray optics diagrams. By making the mechanical loop between the cantilever and sample as short and rigid as possible, there is potential to reduce the sensitivity of the AFM to external vibrations. In a normal AFM, the mass of the optical detection system is pointedly larger than that of the cantilever. As a result, vibrations in the optical detector are more common. Any vibrational motion of the relatively massive optical detector causes fluctuations in the tip-sample force.
As shown in Figure 3a, the optical detector and cantilever are coupled together. The cantilever is in contact with the sample, vibrational motion δz between the sample and the detector-cantilever assembly are amplified by the distortion of the cantilever. This results in a moderately large signal at the photodetector δVPSD. This noise is known as “real”; the motion of the optical detector directly affects how hard the cantilever is pressed into the surface.
In comparison to this, Figure 3b shows the cantilever-sample mechanical loop in the Cypher and it is small and tremendously stiff. Subsequently, the vibrational immunity of the cantilever is improved, and the cantilever-sample force is somewhat stable.
Simultaneously, the relatively massive optical detector will still be affected by vibrational motion δz. Movement of the optical detector, due to vibrational noise, changes the distance between the detector and the cantilever. Nevertheless, this movement only couples very weakly into the measured cantilever deflection.
Depicted in Figure 2b, the dual-stage approach used in the Cypher does come at some extra cost. Rather than one translation stage for engaging and for coarse vertical control of the tip-sample distance, this approach requires two translation stages. Additionally, the two translation stages need to be tightly integrated and move simultaneously. Although this increases the challenges in engineering, thankfully it has resulted in an improvement in stability giving real benefits for Cypher users.
Figure 3. Simple cartoons to explain the observed improvement in noise. a) For a conventional AFM, where the cantilever and optical detector are coupled together (see Figure 2a), the vibrational motion δz is amplified by the optical lever and results in a large voltage at the position sensitive detector δVPSD. b) In the Cypher approach, the vibrational susceptibility of the cantilever is minimized with a very stiff, small mechanical loop.
The Experiment: Real-World Performance
There are several ways in which better performance can be characterized based around noise rejection, on-surface noise and many more. Cypher performance was explored in relation to real-world environments: inside an operating glovebox. This unfriendly, but real-life environment is subjected to rather noisy pumps.
To get a measure of the vibration immunity of the Cypher, a demanding sample was chosen: point defect resolution in calcite. Changes in the imaging conditions were investigated as the glovebox pumps were flipped on and off. Figure 4 shows the setup of the experiment, with the Cypher installed inside the glovebox and the placement of the accelerometer on the table top.
Figure 4. A Cypher placed in an unfriendly, real-life environment. An accelerometer was placed on the table and its signal was measured with the auxiliary input on the ARC controller while the AFM was simultaneously being used to image a calcite surface in fluid.
Figure 5 below similarly shows some representative data with this setup. In this case, the pumps were switched on halfway through a scan. During the first part of the image, the calcite surface was measured with the pumps off. A pair of isolated point defects can be seen in the image, as depicted by the top two red circles. Halfway through the image, the pumps were switched on, meanwhile the scan continued. Clearly visible on the accelerometer signal in Figure 5b, a dramatic increase of the vibrational noise level can be identified. In comparison, this is not seen in the AFM image in Figure 5a, where there was no discernible change in image quality, and an additional atomic point defect was imaged, as depicted by the red circle on the lower left of the image. This experiment demonstrates the extraordinary vibration immunity of the Cypher platform.
Figure 5. a) Calcite topography measured simultaneously with b) accelerometer signal in a glovebox. The pumps were turned on half way through the scan, clearly noticeable in the accelerometer data. Isolated defects are evident, as circled in red, in the simultaneously acquired AFM topography.
This information has been sourced, reviewed and adapted from materials provided by Asylum Research - An Oxford Instruments Company.
For more information on this source, please visit Asylum Research - An Oxford Instruments Company.