AFM Boosts Bottom-Up Nanofabrication

Sophia Hohlbauch1, F. Ted Limpoco1, Nathan Kirchhofer1, Ben Ohler1 and Donna Hurley2
1. Oxford Instruments Asylum Research Inc., Santa Barbara, California USA
2. Lark Scientific LLC, Boulder, Colorado USA
Donna Hurley is a consultant on AFM measurement methods and their application to materials science. Previously, she was a senior scientist at the National Institute of Standards & Technology (NIST) and led a team to develop and apply new AFM techniques for nanomechanics. She has a Ph.D. in condensed-matter physics from the University of Illinois–Urbana-Champaign.
State-of-the-art, ambient-pressure atomic force microscopes (AFMs) provide unique capabilities for characterizing nanomaterials fabricated with bottom-up techniques such as surface self-assembly and on-surface synthesis. For instance, morphology ranging from crystal grain structure to polymer network dispersion can be measured with superb resolution, and reaction dynamics can be monitored in real time. Flexibility in operating environments also enables in-situ experiments in realistic environments. Here, we discuss the capabilities of today’s AFMs and show examples that illustrate their valuable role in bottom-up nanofabrication research.
We thank Peter Beton, Jiajun Chen, James De Yoreo, Andras Kis, and Vladimir Korolkov for valuable assistance with figure preparation. The contributions of everyone at Oxford Instruments Asylum Research are also greatly appreciated.
Corresponding author
Dr. Ben Ohler
Oxford Instruments 
Asylum Research Inc.
6310 Hollister Ave.
Santa Barbara, CA 93117  USA 
Tel: +1 805 696-6466
Famously foreseen 60 years ago in Feynman’s lecture There’s Plenty of Room at the Bottom1, bottom-up nanofabrication has only fully blossomed in this century. This field seeks to construct nanoscale structures and devices using precise control of matter down to molecular or atomic levels. Bottom-up techniques are considered crucial for continued miniaturization of electronics and data storage but also have exciting implications for many other applications including carbon capture, catalysis, and purification and separation membranes.
Nanofabrication at surfaces can be broadly grouped into top-down and bottom-up approaches. Top-down methods such as lithography “impose” a structure through material deposition and removal. In contrast, bottom-up methods introduce atoms or molecules on a surface, which then “grow” into organized structures due to physical and chemical forces2. Surface confinement and the potential for unique nanoscale phenomena yield materials that are otherwise inaccessible. Bottom-up techniques promise to combine ease of fabrication with the ability to tailor material structure and properties.
Examples of bottom-up processes include self-assembly, on-surface synthesis, and atomic layer deposition. Self-assembly involves spontaneous or autonomous ordering through weak, noncovalent interactions (e.g., van der Waals, hydrogen bonds)2, for instance organization of supramolecular networks of organic molecules3. Such processes can incorporate biological principles: for example, formation of protein nanotriangles through native folding mechanisms4, and creating nanostructures using self-assembled DNA origami as templates (Figure 1). 
Figure 1. Topography images of DNA origami: (left) triangular constructs and (right) rectangular constructs with gold nanoparticles. The images illustrate DNA origami’s uniform geometry and its ability to be programmed into different 2D shapes. (left) Scan size 600 nm; imaged in liquid with the MFP-3D AFM in buffer solution in tapping mode. (right) Scan size 270 nm; imaged in liquid with the Cypher AFM in Fast Force Mapping mode. Triangular DNA sample courtesy of Paul W.K. Rothemund, California Institute of Technology. Rectangular DNA sample courtesy of Cheng Tian, Dmytro Nykypanchuk, and Oleg Gang, Brookhaven National Laboratory.
On the other hand, on-surface synthesis uses strong interactions such as covalent bonds for directed assembly5,6. Encompassing both solution-based reactions and ones that occur in ultra-high vacuum (UHV), it enables a wide range of novel nanoarchitectures. A prominent example is the new class of materials called metal-organic frameworks (MOFs) created with directional coordination bonds.
The inherent atomic and molecular nature of bottom-up nanofabrication requires characterization tools with exquisite resolution and sensitivity. This can give the impression that only UHV techniques such as scanning tunneling microscopy (UHV-STM) and atomic force microscopy (UHV-AFM) are suitable. But state-of-the-art AFM tools that work at ambient pressure, such as those from Oxford Instruments Asylum Research, can also characterize these structures, often while contributing new information and with greater ease of use. Recent innovations enable structural imaging of features down to picoscale lattices, capture of dynamic events at video rates, and operation in a range of environments. Here, we discuss AFM capabilities for evaluating bottom-up nanofabrication processes and give examples that illustrate their benefits.
Bottom-up processes can produce structures ranging from zero-dimensional quantum dots and one-dimensional wires to two- and three-dimensional (2D and 3D) layers and devices. In each case, understanding aspects of micro- to nanoscale morphology is vital. Sometimes, simply detecting the material’s presence may be sufficient, although not trivial. Other times, measuring surface roughness and uniformity – of the substrate, the synthesized material, or both – increases understanding. Height measurements (e.g., layer thickness, number of layers) and statistics on individual or ensemble size, shape, and dispersion may also be valuable.
With sub-nanometer spatial resolution, AFM excels at such tasks. Information on morphology is obtained by imaging topography (height) in either tapping or contact mode. Tapping mode, where the cantilever is oscillated near its fundamental resonant frequency, also yields a phase image that can supply useful contrast. Moreover, topography results can be spatially correlated with those from other techniques (e.g., photoluminescence, Raman scattering) for insight into how surface chemistry affects the resulting nanomaterial’s structure7.
Examples of topography images are shown in Figure 1 for DNA origami and Figure 2 for a 2D material grown by chemical vapor deposition (CVD). In addition, Figure 3 shows topographic characterization of a 3D textile created by layering MOF scaffolds. Here, topographic imaging was used to characterize both the overall textile structure and the finer structure of individual polymer strands9.

Figure 2.
 Commercial-scale CVD synthesis of transition metal dichalcogenides such as molybdenum disulfide (MoS2) requires improved techniques for controlling crystalline orientation during growth. These topography images contain a domain of MoS2 grown by CVD on annealed sapphire. The profile across the blue line reveals the domain is <1 nm thick, which is consistent with Raman observations on a MoS2 monolayer. Also seen are individual sapphire terraces with step heights of ~0.22 nm, which is sufficiently low for growth of continuous, single-crystal MoS2. Imaged with the Cypher AFM in contact mode. Adapted from Ref. 8.

Figure 3.
 Bottom-up fabrication of textiles remains an attractive but elusive goal of polymer chemistry. Here, a strategy was demonstrated with layer-by-layer assembly of crystalline coordination networks using liquid phase epitaxy. In a molecular weaving process, reaction precursors were systematically arranged into scaffold-like MOF structures. Initiating the reaction removed superfluous metal ions to obtain a layer of interwoven polymer threads. The topography images show textiles made with (A) three, (B) five, and (C) ten weaving cycles on a silicon substrate. Below each image is the height profile across the line indicated. (D) The graph of textile thickness measured by AFM versus number of weaving cycles reveals a linear dependence. Acquired with the MFP-3D AFM in tapping mode. Reproduced from Ref. 9 as licensed under CC BY 4.0.
Of course, a key driver of bottom-up research is the potential for materials with novel functional properties to enable nanoelectronics, spintronics, and other new applications. One of AFM’s greatest strengths is the ability to go “beyond topography” and characterize functional behavior of nanomaterials. For instance, modes such as conductive AFM (CAFM), electrostatic force microscopy (EFM), and Kelvin probe force microscopy (KPFM) evaluate local conductivity, photocurrents, surface potential, work function, and other electrical response. Piezoresponse force microscopy (PFM), meanwhile, probes nanoscale electromechanical activity. It is particularly useful to examine ferroelectric or piezoelectric domain growth and polarization switching in nanoarchitectures such as MOF nanocrystals10 and self-assembled arrays of polymer lamellae11.
Many bottom-up processes, particularly self-assembly, involve liquid-phase reactions. The ability to perform AFM experiments in a liquid environment can therefore be particularly beneficial. Equipment for fluid experiments ranges from simple droplet holders to more sophisticated cells that surround the entire cantilever and sample surface in liquid. Capabilities for operation under either static or perfusion conditions make these cells exceptionally versatile.
However, tapping-mode imaging in liquid has traditionally been plagued by the “forest of peaks”: spurious signals created by piezoacoustic excitation that make tapping in liquid difficult or even untenable. With new photothermal excitation approaches such as blueDrive by Oxford Instruments Asylum Research12, this phenomenon is eliminated. Driving the cantilever oscillation via laser light results in a near-ideal response on any cantilever, in any environment.
Temperature control during AFM experiments is another useful capability. For instance, it can ensure the viability of biological systems or help explore how to fine-tune liquid-phase reactions. Temperature control is usually achieved using specialized sample stages. On newer AFMs, these stages can provide extremely stable, precise control as high as 400°C.
The goal of bottom-up nanofabrication is creating surface structures with superb regularity on molecular and atomic scales. As a result, structural characterization with commensurate spatial resolution – to even smaller scales than discussed above – is virtually essential. For instance, information on lattice structure helps evaluate the synthesis process and verifies that the intended nanoarchitecture was produced.
Atomic and molecular lattices are often evaluated by high-resolution transmission electron microscopy (TEM), UHV-STM, or UHV-AFM6. In STM, scanning is performed with a sharp metal tip a few angstroms above the sample, and sub-nanoampere tunneling currents between the tip and sample are measured. In UHV-AFM in noncontact mode, a miniature tuning fork is scanned, and small changes in its resonant frequency are detected. These methods provide sufficient resolution to detect crystal structure and determine characteristics such as symmetry, orientation, and lattice constants. However, instruments require significant effort and expertise to maintain and operate. Such disadvantages are usually tolerated because these tools are often considered the only options.
Recently, however, significant resolution advancements have been made in surface analysis techniques that performed at ambient pressures and temperatures. As well as improved environmental TEM, ambient STM and AFM can now achieve spatial resolution comparable to that of UHV tools. Nanofabricated structures with extreme local flatness present optimal conditions for high-resolution imaging, since they allow fewer atoms (or even a single atom) in the tip to interact with the surface.
Examples of real-space lattice imaging with ambient STM are shown in Figure 4 for a self-assembled molecular network of nickel octaethylporphyrin (NiOEP). NiOEP and other metalloporphyrins are promising candidates for new catalysis technologies and in biomedical diagnostic applications. As Figure 4 demonstrates, ambient STM can now resolve molecular structures such as the porphyrin ring, at sufficiently low tunneling currents to avoid damaging the sample or tip.
Figure 4. Ambient STM tunneling current images of a 2D metallophorphyrin network spontaneously formed on the surface of highly oriented pyrolytic graphite (HOPG) from a solution of nickel octaethylporphyrin (NiOEP) in phenyloctane. (a) Survey scan revealing a grain boundary (arrows) and moiré patterns (wavy lines) that arise from lattice mismatch between the NiOEP layer and the HOPG substrate. Scans of successively smaller regions, indicated by the boxes, (b) reveal closer detail of the regular sub-nanometer lattice structure and (c) even resolve the porphyrin molecular rings. Imaged with the Cypher AFM in STM mode in air. Sample courtesy of Michael Hopkins, Univ. Chicago. For more information see Ref. 13.
As seen in Figure 4, many bottom-up results have been obtained with electrically conducting substrates such as highly oriented pyrolytic graphite (HOPG) or noble metals. However, today’s research frequently involves nonconducting surfaces, partly because the substrate can strongly influence the resulting nanomaterial’s properties and partly to decouple the electrical response of the substrate and the nanomaterial5. These systems preclude characterization with STM, which requires conducting samples.
Fortunately, AFM is amenable to insulating and semiconducting materials as well as conducting ones, and some models have sufficient resolution for lattice-scale imaging. Examples of ambient AFM images are shown in Figure 5 for supramolecular arrangements on insulating hexagonal boron nitride (hBN). As indicated in the figure, line section analysis enables quantitative measurements of lattice constants. Ambient AFM experiments can also be performed in a range of realistic environments (i.e., in liquid, at variable temperature), as discussed above.

Figure 5. 
Ambient AFM images of 2D supramolecular networks: (A) hBN substrate, (B) porous perylene tetracarboxylic diimide (PTCDI) and melamine network, and (C) 5,10,15,20-tetrakis(4-carboxylphenyl)porphyrin (TCPP) lattice. Sections correspond to the orange lines in the topography images. Analysis of multiple line sections yielded the following values of lattice constants: 0.25 ± 0.01 nm (hBN), 3.54 ± 0.04 nm (PTDCI), and 2.24 ± 0.05 nm (TCCP). Images acquired with the Cypher AFM in (a, b) contact mode and (c) tapping mode in air.  Adapted from Refs. 14 and 15.
The impressive spatial resolution seen in Figures 4 and 5 is the result of instrumentation advances in newer AFMs. In the lateral direction, closed-loop control provides more accurate XY scanning than the open-loop operation of older AFMs. In the vertical direction, improving the design of the mechanical loop between the tip and sample mean vertical noise floors can be as low as 5-10 pm in many laboratories. Higher spatial resolution is also attained using smaller cantilevers, which have intrinsically lower thermal noise for the same spring constant. Another critical factor for achieving ambient STM is the introduction of ultra-low-noise current amplifiers that routinely detect currents below 1 pA.
Evaluating a reaction’s intermediate steps as well as its end products gives deeper insight into reaction pathways. Some such experiments have been accomplished using careful, time-consuming methods to stabilize intermediate states at cryogenic temperatures for investigation with UHV tools6.
However, modern ambient AFMs also provide many capabilities for investigating dynamic behavior and offer greater ease of use. For instance, “fast-scanning” AFMs developed over the last decade can acquire a single image in seconds. Events on even shorter timescales can be captured with new high-speed, video-rate AFMs such as the Cypher VRS AFM. Capable of rates exceeding 10 frames per second (10 fps), the VRS and other video-rate AFMs are especially well suited to study self-assembly energetics, which typically occur on millisecond to second timescales.
Figure 6 demonstrates the power of video-rate AFMs for this purpose. In this work, the two-dimensional nucleation and growth of arrays of short protein fragments called peptides were examined in detail. Molecular-resolution imaging and high-speed movies revealed that crystal formation initiated from highly ordered nuclei and assembled one row at a time. Combined with molecular dynamics simulations, the results confirmed predictions of Gibbs’ classical nucleation theories developed over a century ago. Individual frames are shown in Figure 6, but the data were acquired directly in movie format for ease of use and better event recognition17.

Figure 6.
 In-situ monitoring of self-assembly dynamics of MoSBP1 peptides, which contain the seven amino acids Tyr-Ser-Ala-Thr-Phe-Thr-Tyr (YSATFTY), on MoS2. (top row) High-resolution topography images reveal the structure and ordered arrangement of peptides as they attach to the surface from solution. (middle row) Close-up images from a topography movie acquired at 0.39 fps show nucleation and growth of an island. (bottom row) Frames from the same movie show formation of new rows beside older ones, such as those indicated by the circles (dashed, before creation; solid, after). Height (color) scale is 1.5 nm (top and middle rows) and 1.1 nm (bottom row). Acquired on the Cypher VRS AFM in tapping mode. Adapted from Ref. 16; for a related movie, see Ref. 17.
Figure 7 shows a further illustration of video-rate AFM imaging to directly observe bottom-up dynamics. Cetyl trimethylammonium bromide (CTAB) is a surfactant that spontaneously self-assembles into rows of hemicylindrical structures, or hemimicelles, upon adsorption at the solid-liquid interface with HOPG. Occurring on timescales of microseconds, the dynamics of surfactant processes such as this have important implications for dispersion stabilization, corrosion inhibition, and boundary lubrication. As Figure 7 demonstrates, high-speed imaging can help elucidate the kinetics of adsorption, assembly, and motion of CTAB aggregates at the solid-liquid interface.

Figure 7.
 Tapping mode phase images of the dynamic rearrangement of CTAB hemimicelles on HOPG during perfusion of a dilute solution of isopropyl alcohol. The images represent individual frames from a movie acquired at ~2 s per image (0.48 fps). Two CTAB grains with opposite row orientation are bounded on the left and right by two grains with parallel orientation. The grain boundaries significantly shift over time, but the micelles within each grain maintain their orientation. Acquired on the Cypher VRS AFM.
Instrumentation advances that have made video-rate scanning possible include the use of small cantilevers (~10 μm long). Accommodating these cantilevers, which have resonance frequencies greater than 1 MHz in air, has required development of faster control electronics and higher instrument resonances than those of older AFMs. Operating small cantilevers in tapping mode using conventional piezoacoustic excitation presents difficulties, however. Response spectra (“tunes”) often exhibit severe distortions and greater variability in time, complicating setup and stable operation. In contrast, photothermal excitation provides a clean, stable response at high frequencies, even during liquid perfusion. This makes it invaluable for capturing dynamic events that occur at liquid-solid interfaces16-18 and is why blueDrive photothermal excitation is standard on Cypher VRS AFMs.
Combining chemistry, physics, and materials science, the world of bottom-up nanofabrication at surfaces has progressed immensely since Feynman’s first vision. Yet myriad opportunities still remain, such as development of other surface-confined chemical reactions, achieving finer control of material properties, and further deployment in advanced technology. Today’s AFMs provide powerful capabilities for characterizing bottom-up nanomaterials including superb spatial resolution, imaging speeds up to video rates, and versatile environmental control. These and many other benefits mean that AFMs are rapidly becoming an essential tool for bottom-up research.
1. R. P. Feynman, Eng. Sci. 23, 22 (1960).
2. J. V. Barth, G. Costantini, and K. Kern, Nature 437, 671 (2005).
3. T. Kudernac, S. Lei, J. A. A. W. Elemans, and S. De Feyter, Chem. Soc. Rev. 38, 402 (2009).
4. W. M. Park, M. Bedewy, K. K. Berggren, and A. E. Keating, Sci. Rep. 7, 10577 (2017).
5. R. Lindner and A. Kühnle, Chem. Phys. Chem. 16, 1582 (2015).
6. Q. Shen, H.-Y. Gao, and H. Fuchs, Nano Today 13, 77 (2017).
7. S. Zhao, G. B. Barin, L. Rondin, C. Raynaud, A. Fairbrother, T. Dumslaff, S. Campidelli, K. Müllen, A. Narita, C. Voisin, P. Ruffieux, R. Fasel, and J. S. Lauret, Phys. Status Solidi B 254, 1700223 (2017).
8. D. Dumcenco, D. Ovchinnikov, K. Marinov, P. Lazić, M. Gibertini, N. Marzari, O. Lopez Sanchez, Y.-C. Kung, D. Krasnozhon, M.-W. Chen, S. Bertolazzi, P. Gillet, A. Fontcuberta i Morral, A. Radenovic, and A. Kis, ACS Nano 9, 4611 (2015).
9. Z. Wang, A. Błaszczyk, O. Fuhr, S. Heissler, C. Wöll, and M. Mayor, Nature Commun. 8, 14442 (2017).
10. Y. Sun, Z. Hu, D. Zhao, and K. Zeng, Nanoscale 9, 12163 (2017).
11. M. F. Guo, J. Jiang, J. F. Qian, C. Liu, J. Ma, C.-W. Nan, and Y. Shen, Adv. Sci. 6, 1801931 (2019).
12. A. Labuda, S. Hohlbauch, M. Kocun, F. T. Limpoco, N. Kirchhofer, B. Ohler, and D. Hurley, Micros. Today 26, 12 (2018).
14. V. V. Korolkov, S. A. Svatek, S. Allen, C. J. Roberts, S. J. B. Tendler, T. Taniguchi, K. Watanabe, N. R. Champness, and P. H. Beton, Chem. Commun. 50, 8882 (2014).
15. V. V. Korolkov, S. A. Svatek, A. Summerfield, J. Kerfoot, L. Yang, T. Taniguchi, K. Watanabe, N. R. Champness, N. A. Besley, and P. H. Beton, ACS Nano 10, 10347 (2015).
16. J. Chen, E. Zhu, J. Liu, S. Zhang, Z. Lin, X. Duan, H. Heinz, Y. Huang, and J. J. De Yoreo, Science 362, 1135 (2018).
18. L. Costa, G. Li-Destri, N. H. Thomson, O. Konovalov, and D. Pontini, Nano Lett. 16, 5463 (2016).
Website developed by S8080 Digital Media