Past exam of the MPhil in micro- and nanotechnology enterprise course of the University of Cambridge / 2018 / Paper 1 / Section A(iii) / 5

In contact-mode atomic force microscopy, a sharp tip on a compliant cantilever remains in contact with the surface while a feedback loop controls its normal deflection. A laser reflected from the cantilever onto a position-sensitive detector converts bending into an electrical error signal. With calibrated deflection $d$ and stiffness $k_c$, the normal force is approximately $F=k_{c}d$ under the stated contact convention.

A piezoelectric scanner raster-scans the sample or tip. The controller adjusts vertical position to maintain the chosen deflection setpoint; that vertical displacement is recorded as topography. The right panel of the microscopy diagram identifies tip, cantilever, optical deflection readout, controller and vertical actuator.

The result is a probe-accessible surface, affected by finite tip shape, sample deformation, adhesion and feedback bandwidth. Lateral friction can distort or damage a soft surface, so constant deflection is not a guarantee of unchanged specimen structure.

Four useful interactions are dispersion attraction, electrostatic interaction, capillary attraction and short-range repulsion. Their ranges are relative to atomic contact and depend on geometry and environment.

Van der Waals forces provide an attractive contribution extending beyond direct atomic overlap; for an ideal sphere and plane it scales as $F=-AR/(6z^2)$ in a simple nonretarded model. Electrostatic forces can also be long-ranged: biased conducting surfaces are commonly attractive, whereas charged insulating surfaces can attract or repel depending on their charge and screening.

A capillary meniscus in humid air can pull tip and sample together. It acts over the formation and rupture range of the liquid bridge and can greatly exceed atomically localized forces; it is not a universal interaction in a dry vacuum. At the shortest separations, overlap-related Pauli repulsion provides a steep repulsive contact contribution.

Atomic-scale contact contrast relies on strongly localized short-range tip–surface forces, especially the rapidly varying repulsive interaction in the simple contact-mode picture. Long-range attraction mainly adds a background. Real atomic-resolution imaging also depends on apex structure, stability and operating mode; a generic long-range force alone does not resolve individual surface atoms.

For a polymer curve parameterized by contour distance $s$, let $\mathbf{t}(s)$ be its unit tangent. The persistence length is the characteristic contour distance over which tangent orientations lose correlation. In the usual three-dimensional worm-like chain convention,where ${\kappa }_b$ is the bending stiffness. Segments much shorter than $L_p$ look approximately straight; a much longer chain can contain many statistically distinct orientations.

A surface image can be traced to estimate tangent correlations as a function of contour separation, after accounting for tracing resolution and substrate effects. A strictly equilibrated two-dimensional chain has a different relation between bending stiffness and correlation decay length from the three-dimensional convention. The definition must therefore state whether the surface preserves a projected conformation or allows in-plane equilibration. This is a statement about statistical filament mechanics; biological interpretation is deferred.

Interpret the stipulated model as a freely jointed chain with $N$ independent, isotropically oriented segments of length $b$. With end-to-end vector $\mathbf{R}={\sum }_{j=1}^N{\mathbf{b}}_j$, cross terms vanish on averaging:Using only the model's supplied segment size and count givesThe mean vector is zero, not the mean magnitude. In the three-dimensional large-$N$ Gaussian chain limit,Thus the common root-mean-square interpretation of an average chain size is $10.1\mathrm{nm}$, while the mean scalar distance is about $9.32\mathrm{nm}$. The contour length $Nb=320\mathrm{nm}$ is a different quantity. A count of $999$ links between $1000$ positions changes these estimates by less than $0.1\%$.

These are results of the explicitly idealized statistical chain, not a prediction about the conformation or persistence length of an actual biological molecule. TODO (user-deferred biology): material-specific biological interpretation of that model.