Supplementary to:

Stability, resolution and ultra-low wear amplitude modulation atomic force microscopy ofDNA: small amplitude small set-point imaging

Sergio Santos1,2, Victor Barcons3, Hugo K Christenson1, Daniel J Billingsley1,4, William A Bonass4, Josep Font3 and Neil H Thomson1,4

1School of Physics and Astronomy, University of Leeds, LS2 9JT, UK. 2Laboratory for Energy and Nanosciences, Masdar Institute of Science and Technology, Abu Dhabi, UAE, 3Departament de Disseny i Programació de Sistemes Electrònics, UPC -Universitat Politècnica de CatalunyaAv. Bases, 61, 08242 Manresa, Spain

4Department of Oral Biology, Leeds Dental Institute, University of Leeds, LS2 9LU, UK.

Fig. 1S below details how to reach SASS as a function of amplitude set-point A. Fig. 2S shows the effects of relative humidity on the NC mode of operation and SASS and the apparent width and height of dsDNA molecules on mica. Fig. 3S compares the tip-mica force and energy at very low and at very high humidity. Fig. 4S shows high resolution images of dSDNA molecules in the SASS mode. The humidity has been controlled with a home-built humidity chamber with a flow of dry nitrogen to decrease the humidity and water vapor to increase it.

FIG. 1S Topography of dsDNA on mica for which Ais first a-b) 2.1nm, then d-e) 0.6 nm and finally to g-h) 0.1 nm. The images in the middle column are software zooms. The NC mode of operation prevails for A=2.1 and A=0.6 nm. Then the SASS mode is reached for A=0.1 nm. The step change in stability and resolution is observed in the SASS region where DNA helix periodicity is resolved as shownin the zoom in h). The normalised force (dashed blue) and energy (continuous black) is reconstructed in the left columnand the minimum distance of approach dm for each of the scans on the left is shown. Note how dm decreases with decreasing A as predicted in Fig. 1a in the main text (grey). In short, and in agreement with the interpretation in the main text, dm lies close to the zero value in Fts only in the SASS mode whereas in the NC mode dm is approximately 2-3 nm far from the steep slope in Fts where mechanical contact with the micasurface is established. We deduce that these 2-3 nm correspond with the water layer thickness for the tip-mica system. Such meniscus formation is due to confinement and therefore implies that even at water vapours below saturation the meniscus will be formed1. The normalizing value for Fts is ≈1.2 nN and for Edis is ≈50eV. The force and energy here have been computed with the use of equations (1) and (2) in the main text and the large values of energy dissipation have resulted by employing relatively large values of A0 ≈ 30 nm for reconstruction where smooth force transitions follow. In SASS the energy values should be computed with (2) with the A0 values employed to image in the SASS mode and these values are typically in the order of eV or fractions of eV as detailed in the main text. However, these larger values of A0employed to reconstruct Ftsin the above figures have been very recently2 reported to allow recovering the force for the whole relevant range of distance d. More importantly, the approximations for (1) are valid when A/A0 is close to 1 and this method for force reconstruction does not alter the validity of the interpretation in the main text (see below).In short, the smooth force transition is universal and has been previously reported in AM AFM3 for sufficiently large values of A04 allowing to sense the whole range of minimum distance of approach dm. Note that the interpretation of the region of force where SASS operates is based on dm (see argument below). Thus,since the value of the conservative force Fts depends on dm only, dm alone can be employed to confirm what the conservative force is in SASS once the force is reconstructed with any value of A0. In any case, for the given dm in SASS, the value of Fts is ≈ 0nN throughout. Comparing with the force in Fig. 1b in the main text, note that the first slope in Fts is positive rather than negative. While the dm values can be trusted in both Fig. 1 in the main text and in the figures above, the slope of the force, and its value at a given d, is more reliable in these (Figs. 1Sc, f, i) than in Fig. 1b. As stated this is because of the large A/A0 approximation assumed5, 6 in the theory employed to derive (1). The DNA apparent widths at half-height are ≈8 nm in the NC mode and ≈4.5 nm in SASS mode. The experimental parameters are f0=331 kHz, f= f0, k≈40 N/m, Q=550 and A0=2.5 nm. Scan sizes of a, d and g) are 300 nm by 300 nm and the relative humidity was ≈ 50%.

FIG. 2SBoth thez-piezo or topography and the phase contrast are shown for a, b) low 5% relative humidity (RH), c, d) intermediate RH of 50% and e, f) high RH of 95% in both the a, c and e) NC and b, d and f) SASS modes of operation. Corresponding force Fts and energy Ets versus distance dcurves are shown in g) to i) respectively in dashed blue and continuous black respectively. The asterisks in Fts and Ets imply that the values are normalised as in the main text. Here, as in Fig. 1S, the smooth force transition was reached by employing a relatively large value of A0≈ 30 nm to reconstruct Fts as in Fig. 1S. Very recent modelling of cantilever dynamics2 has further shown that a step in energy dissipation should be expected at the onset, i.e. don, of capillary interactions. Such step is shown in Fig. 2Sg (circle) implying that at distances smaller than that, i.e. d<don, the water layers on the tip and the sample overlap. Similar results have been very recently reported in frequency modulation AFM7. It is also observed that the apparent height in the NC mode gradually decreases with RH as the surface features disappear both in the j) z-piezo signal andk) the width (FWHM) of the molecule also decreases almost linearly. In the SASS mode the height and width are rather similar at 5% and 50% RH. At 95% RH, however, the height dramatically increases while the lateral contrast remains below 6nm on the average. By inspection it can be seen that the phase contrast also increases dramatically in the SASS mode at 95% RH. Note that loss of apparent height does not follow from large sample deformations only8. In fact, the water layers might cover the molecule in the NC mode and the height might be lost even when mechanical contact does not occur9. The images correspond to the same tip-sample system as that shown in Fig. 1S implying that in the SASS mode the repeat can be recovered as shown in Fig. 1Sh. The normalizing values for Fts and Etsare similar to those in Fig. 1S. In summary, from Figs. 2Sg-i it is observed that the step in energy, i.e. d≈don where the meniscus should form, occurs at all values of RH and for SASS d<don throughout implying that permanent tip-water contact occurs throughout in SASS mode. The formation of the meniscus at all values of RH should be due to confinement between the tip and the sample and is predicted to occur by standard theory at even small values of RH1.

FIG. 3S In this figure, the force (dashed) and energy (continuous) in Figs. 2Sg and 2Si corresponding to 5% (light colors) and 95% (dark colors) RH respectively have been plotted for comparison. It can be observed that the force and energy profiles are relatively similar in both range and magnitude. The forces are in fact only slightly, i.e. fractions of a nm, longer in range at the higher humidity. This explains why SASS can operate for the whole spectrum of relative humidity as shown in Figs. 2S1-f. The fact that there is also water present at very low humidity, and in particular, meniscus formation due to confinement1, is also in agreement with other reports in the literature10.

FIG. 4S a-e) Examples of topographic images of dsDNA molecules on mica obtained in the SASS mode of operation where the helical repeat is resolved. In our experience, there are two conditions that allow resolving the helical repeat in ambient AM AFM, namely 1) employing the SASS mode of operation and 2) employing super sharp tips, i.e. R<5 nm. We would like to note however that some experience by the user is required in order to obtain the highest resolution. On the other hand, the SASS mode is otherwise easily accessed by selecting operational parameters similar to those reported in this work, i.e. A0≈1-3 nm and A≈0.05-0.2 nm. It should also be expected that the technique will improve in the future by optimizing sample preparation. Sample preparation in this work is standard and as discussed in the literature11. Moreover, it is clear that future developments might exploit other advantages, i.e. other than the phenomenon exploited in SASS, of AFM operation12, 13.

1.J. Israelachvili, Intermolecular & Surface Forces, 2 ed. (Academic Press, 1991).

2.S. Santos, C. A. Amadei, A. Verdaguer and M. Chiesa, The Journal of Physical Chemistry C 117 (20), 10615–10622 (2013).

3.A. S. Paulo and R. Garcia, Physical Review B 66 (4), 041406-041409 (2002).

4.S. Santos, V. Barcons, J. Font and N. H. Thomson, J. Phys. D: Appl. Phys. 43, 275401-275407 (2010).

5.J. E. Sader, T. Uchihashi, M. J. Higgins, A. Farrell, Y. Nakayama and S. P. Jarvis, Nanotechnology 16, S94–S101 (2005).

6.A. J. Katan, M. H. van Es and T. H. Oosterkamp, Nanotechnology 20, 165703-165711 (2009).

7.D. S. Wastl, A. J. Weymouth and F. J. Giessibl, (2013).

8.S. Santos, V. Barcons, H. K. Christenson, J. Font and N. H. Thomson, PLoS ONE 6 (8), e23821 (2011).

9.S. Santos, M. Stefancich, H. Hernandez, M. Chiesa and N. H. Thomson, Journal of Physical Chemistry C 116 (4), 2807–2818 (2012).

10.S. Rozhok, R. Piner and C. A. Mirkin, J. Phys. Chem. B 107 (3), 751–757 (2003).

11.S. Santosand N.H. Thomson “High resolution imaging of Immunoglobulin G (IgG) antibodies and other biomolecules using amplitude modulation atomic force microscopy in air.” In Atomic Force Microscopy in Biomedical Research

Eds. Davide Ricci and Pier Carlo Braga (Humana Press, New York, 2011)

Methods Mol Biol, 736, Part 2, 61-79 (2011).

12.R. Garcia and R. Proksch, European Polymer Journal (2013).

13.A. Cerreta, D. Vobornik and G. Dietler, European Polymer Journal DOI:10.1016/j.eurpolymj.2013.03.026 (2013).

1