Diamond Colloidal Probe Force Spectroscopy

Diamond is a highly attractive coating material as it is characterized by a wide optical transparency window, a high thermal conductivity, and an extraordinary robustness due to its mechanical properties and its chemical inertness. In particular, the latter has aroused a great deal of interest for scanning probe microscopy applications in recent years. In this study, we present a novel method for the fabrication of atomic force microscopy (AFM) probes for force spectroscopy using robust diamond-coated spheres, i.e., colloidal particles. The so-called colloidal probe technique is commonly used to study interactions of single colloidal particles, e.g., on biological samples like living cells, or to measure mechanical properties like the Young’s modulus. Under physiological measurement conditions, contamination of the particle often strongly limits the measurement time and often impedes reusability of the probe. Diamond as a chemically inert material allows treatment with harsh chemicals without degradation to refurbish the probe. Apart from that, the large surface area of spherical probes makes sensitive studies on surface interactions possible. This provides detailed insight into the interface of diamond with other materials and/or solvents. To fabricate such probes, silica microspheres were coated with a nanocrystalline diamond film and attached to tipless cantilevers. Measurements on soft polydimethylsiloxane (PDMS) show that the manufactured diamond spheres, even though possessing a rough surface, can be used to determine the Young’s modulus from a Derjaguin-Muller-Toporov (DMT) fit. By means of force spectroscopy, they can readily probe force interactions of diamond with different substrate materials under varying conditions. The influence of the surface termination of the diamond was investigated concerning the interaction with flat diamond substrates in air. Additionally, measurements in solution, using varying salt concentrations, were carried out, which provide information on double-layer and van-der-Waals forces at the interface. The developed technique offers detailed insight into surface chemistry and physics of diamond with other materials concerning long and short-range force interactions and may provide a valuable probe for investigations under harsh conditions but also on biological samples, e.g., living cells, due to the robustness, chemical inertness, and biocompatibility of diamond.


AFM and FIB characterization of the diamond colloidal probe
Further characterization of the diamond probes is shown in Figure S1. In addition to scanning electron microscopy (SEM), the surface morphology after oxidation was determined by atomic force microscopy (AFM) reverse imaging on a standard silicon AFM tip (Nanoworld, PPP-NCHR, tip radius: < 10 nm) (reverse imaging: a high aspect ratio object is imaged which provides the topography of the scanning probe). The roughness was determined in a region of the AFM image, which does not show artifacts resulting from the strong curvature of the sphere as shown in Figure  1d). A root-mean square (RMS) roughness of 18 nm was obtained, which is close to the RMS roughness of 16 nm obtained for a nanocrystalline diamond (NCD) film grown on a flat substrate using the same growth conditions (see Figure S2). Variations of up to 2 nm are commonly observed for NCD layers with such a thickness. The measured region is close to the hole in the diamond film which is resulting from the placement of the templates onto a substrate. As this region is not facing in the growth direction, this indicates that the diamond growth is homogeneous over the silica sphere. An oxygenated diamond sphere is shown in Figure S1, which shows the expected RMS roughness. This shows that the oxygen plasma treatment did not significantly alter the surface of the diamond film as expected. For hydrogen plasma treatments etch rates are lower and hence there will also be no change in roughness. 1 The thickness of the diamond film of approx. 200 nm was confirmed using a focused ion beam (FIB) cross-section as shown in Figure S3. Here, due to the strong ion bombardment and redepositions, a surface damage is observed, which results in smoothening of the diamond film.

Young's modulus determination by Hertz and DMT fitting
For determining the elasticity of a substrate with diamond colloidal probes, a polydimethylsiloxane (PDMS) standard sample was used that has a Young's modulus of 2.5 ± 0.7 MPa as defined by the manufacturer. As measurements were conducted in air, the hydrogenated diamond colloid was used to suppress strong capillary forces (Note: measurements with the uncoated silica colloid are not possible using the same conditions). The used probe with a spring constant of 0.14 N/m is shown in Figure S5 a). Fitting was done with the JPK Data Processing software employing eq. (1) and a radius of 2.7 µm for the sphere (2.5 µm silica particle + 200 nm diamond film). It should be noted that due to the large size of the colloidal particle, varying the radius by 200 nm will also lead to a result that is in the range of the manufacturer specification. The Derjaguin-Muller-Toporov (DMT) model (Eq. (1)) takes the adhesion into account, which facilitates the fitting procedure for the shown measurements as in this case the whole force curve may be fitted. In order to keep noise out of the fit, however, only the contact region including the cantilever snap-in (approx. 8% of the curve) were used. 2 F is the applied force, F(adh) is the adhesion, ν the Poisson's ratio, E the Young's modulus, R the radius of the sphere and δ the indentation depth.
The results and additional information for the fit is shown in Table S1 S-5

Hydrophobic interaction between H-Colloid and HPCD
The measurements of the H-colloid on HPCD in air is different to the three other combinations, as here, meniscus forces should be rather low due to the hydrophobicity of both surfaces. Nevertheless, a higher adhesion force is obtained (5.72 nN) in comparison to the O-Colloid on HPCD (2.74 nN). It is known that on hydrophobic surfaces, higher amounts of water can be present than on hydrophilic surfaces. This is explained by the formation of nanodroplets preferably on defect sites, which in total result in a higher volume than a homogeneous water layer. 3 In our study polished PCD has been used, which does not only contain defects but also grain boundaries. Apart from that, the diamond sphere is coated with NCD that has even more grain boundaries as well as sp 2 -carbon impurities. Both, the substrate and the sphere, are far from perfect and should easily allow nanodroplet formation. Hence, we assume that the increased adhesion force is caused by such nanodroplets and induced by imperfection of the substrate and the colloid surface. Figure S4 shows the adhesion force in dependence of the measurement cycle. Here, a trend towards higher adhesion is observed, which may be assigned to restructuring of the nanodroplets after contact (Note: all measurements were taken on the same spot). In future studies, single crystalline diamond substrates may be used to get a better understanding of this interaction as they provide a more homogeneous surface.

DLVO fitting of the force curves obtained in solution
Electrostatic double layer and van der Waals forces should govern the measurements in solution. These forces are described by the DLVO theory (Derjaguin, Landau, Verwey, Overbeek), which is commonly used to explain the stability of colloidal suspensions. With Igor Pro and a fitting procedure by McKee based on an algorithm by Chan et al., 3,4 we fitted our approach curves to obtain the surface potential and the surface charge density. The fits and the corresponding measured approach curve for the different concentrations are shown in Figure S6. Table S2 summarizes the fit results.
The fitting procedure assumes that for both surfaces, the potential is identical. As both surfaces are oxygen terminated, this approximation is plausible. For the different KCl concentrations the varying Debye length is calculated according to following equation (2) c is the bulk concentration of the ion, NA Avogadro's constant, is the valence of the ion, is the elementary charge, 0 is the permittivity in vacuum, is the relative permittivity of the solvent, is the Boltzmann constant, and is the absolute temperature.
Surface charge density and surface potential were adjusted to fit the measured data. It is observed that the surface potential is decreasing with increasing KCl concentration. For the surface charge density, the opposite trend is apparent. This may be explained by the ion sensitivity of diamond surfaces as used in ion-sensitive field effect transistors (ISFETS), 6 and is consistent with previous results in our group on detonation nanodiamond, where the KCl concentration strongly influences the surface assembly on silicon substrates. 7 Figure S6: Approach curves of an oxygen-terminated colloid to an oxygen terminated polycrystalline diamond in different KCl concentrations and corresponding DLVO fits (constant potential and constant charge).