Letter pubs.acs.org/NanoLett
Painting with Biomolecules at the Nanoscale: Biofunctionalization with Tunable Surface Densities Robert Schlapak,†,⊥,# Jürgen Danzberger,‡,⊥ Thomas Haselgrübler,† Peter Hinterdorfer,†,§ Friedrich Schaf̈ fler,‡ and Stefan Howorka*,†,∥ †
Center for Advanced Bioanalysis GmbH, 4020 Linz, Austria Institute for Semiconductor and Solid State Physics, §Institute for Biophysics, Johannes Kepler University, 4040 Linz, Austria ∥ Department of Chemistry, Institute of Structural and Molecular Biology, University College London, London, England ‡
S Supporting Information *
ABSTRACT: We present a generic and flexible method to nanopattern biomolecules on surfaces. Carbon-containing nanofeatures are written at variable diameter and spacing by a focused electron beam on a poly(ethylene glycol) (PEG)-coated glass substrate. Proteins physisorb to the nanofeatures with remarkably high contrast factors of more than 1000 compared to the surrounding PEG surfaces. The biological activity of model proteins can be retained as shown by decorating avidin spots with biotinylated DNA, thereby underscoring the universality of the nano-biofunctionalized platform for the binding of other biotinylated ligands. In addition, biomolecule densities can be tuned over several orders of magnitude within the same array, as demonstrated by painting a microscale image with nanoscale pixels. We expect that these unique advantages open up entirely new ways to design biophysical experiments, for instance, on cells that respond to the nanoscale densities of activating molecules. KEYWORDS: Nanopattern, surface, protein, DNA, SEM, AFM, fluorescence microscopy
N
throughput and reproducibility, but a costly master-mold has to be fabricated for any new pattern geometry.22 By contrast, dippen nanolithography is much more flexible in terms of geometry and pattern modifications, but substantially slower. Nanoarrays of biomolecules have been generated on a variety of substrates including Au,13,14 Si,13,22 SiO2,13 and glass.22,23 Despite these advancements, it is currently not possible to obtain patterns for cell-biological applications on optically transparent substrates that offer both sufficiently high contrast between target and non-target areas, and the ability to tune the surface density of biomolecules within a nanoarray. Here we adapt a biofunctionalization approach9,24 that is based on resist-less electron-beam-induced deposition (EBID)25,26 of carbon-containing nanofeatures and develop it to a universal biofunctionalization platform with high contrast and wide-range control of the local density of the biomolecules. As a result, we are able to achieve contrasts between target and nontarget areas of more than 1000, while being able to simultaneously control the density of the selectively bound biomolecules over three orders of magnitude within one and the same array. The size of the EBID target areas can in principle be controlled with nanometer-scale precision,27 while their lateral distances can be freely chosen to allow, for example, optical readout near the resolution limit of the microscopic technique.
anostructured and biofunctionalized surfaces are of increasing importance in the bio-sciences. DNA-and protein nanoarrays can help to determine the genome and proteome content in a highly parallel fashion, leading to improved diagnostics and therapeutics.1 Benefits of nanoarrays include the higher density of reaction sites and a much smaller sample volume compared to microarrays that can translate into potentially higher sensitivity, improved kinetics, and faster reaction times.2 In addition, nanopatterns of proteinaceous cellsurface receptors are essential tools for studying cell adhesion1,3−5 and signaling processes. Well-defined arrays are advantageous because they offer control over the biologically relevant spatial distribution of activating biomolecules on the nanoscale. Given the high sensitivity of cells, it is of paramount importance to control the surface density of biomolecules in the target areas of the array and to avoid their spurious presence in non-target areas, where they would result in undesired cellular activation. Current approaches for nanopatterning comprise top-down methods, such as nanocontact printing, 6,7 nanoimprint lithography,8 e-beam and UV lithography,9−12 dip-pen nanolithography,13−15 AFM-based nanografting,16 and direct depositing with a nanopipette.17 Alternatively, bottom-up strategies, such as micelle,4,18 particle,2 or transfer nanolithography19 as well as nanopositioning on DNA20 or protein21 scaffolds are successfully employed. Each approach features a set of advantages with regard to feature size, pattern fidelity, costs, throughput, and ease of fabrication. Nanoimprint lithography, for instance, offers features as small as 10 nm with high © 2012 American Chemical Society
Received: December 24, 2011 Revised: February 8, 2012 Published: February 29, 2012 1983
dx.doi.org/10.1021/nl2045414 | Nano Lett. 2012, 12, 1983−1989
Nano Letters
Letter
Results and Discussion. For EBID patterning, we used a high-resolution scanning electron microscope (SEM) with a beam blanker and an attached control unit for e-beam lithography. More details of the lithography system and the other experimental techniques employed in this work are given in the Supporting Information. The complete process sequence from e-beam writing to the incubation with biomolecules is schematically summarized in Figure 1. First, a dense, self-
focused on biofunctionalization, we employed here standard SEM techniques in a residual gas atmosphere, keeping in mind that smaller EBID features are possible if needed. We produced well-defined squares with nominal base widths of 100 and 300 nm, or circular dots with diameters ≤50 nm, which are shown as SEM images in Figure 2A−C. In this
Figure 2. Structural characterization of carbon-nanodeposits via (A− C) scanning electron microscopy on ITO-glass substrates, and (D−F) atomic force microscopy of 100 × 100 nm2 features on silicon substrates. The features were written using the following conditions: (A) nominal dimensions, 300 × 300 nm2; electron dose, 18 pC; (B) 100 × 100 nm2, 2 pC; (C) dots created by a stationary, focused electron beam, 2 pC; (D) 0.15 pC; (E) 0.72 pC; (F) 1.44 pC. A working distance of 8 mm and an acceleration voltage of 10 kV were used for all islands.
Figure 1. EBID-process sequence employed for the generation of biomolecular nanopatterns. Following (1) PEG-silanization of the ITO-glass substrate, (2) e-beam-induced deposition generates a pattern of nanosized features. (3) Incubation with biomolecules and (4) rinsing leads to the selective decoration of the nanofeatures with biomolecules.
imaging mode, e-beam-exposed areas appear dark, because the EBID areas reduce the local emission of secondary electrons and thus the intensity of the SEM signal. To quantify the height of the nanodeposits, atomic force microscopy (AFM) analyses were performed in ultrapure water in contact mode. Atomically flat Si substrates were employed for these reference experiments to avoid potential artifacts introduced by the surface roughness of the polycrystalline ITO film. The results are still representative, because Si and ITO-coated glass substrates behave similarly with regard to e-beam writing. As expected,29 the AFM measurements confirm the presence of mesa-like carbon nanofeatures that protrude from the surrounding PEG layer (Figure 2D−F). To study in detail the size evolution of the written features as a function of the e-beam parameters, we systematically varied the exposure time at fixed beam currents and beam energies. Figure 3A shows the evolution of the lateral dimensions of written squares with side lengths of 100 and 300 nm and of dots as a function of the integral electron dose deposited per feature at an electron energy of 10 keV. In agreement with earlier reports,26,29 an increase in the electron dose has only a minor influence on the lateral dimension. For example, the measured diameter of both types of squares increased by less than 20% with higher electron doses (Figure 3A), presumably due to secondary electron emission from the sidewalls of the growing deposits.26 By comparison, the height of the nanodeposits was strongly dependent on electron dose and increased six-fold within the same dose range, which is in line with a carbon deposition process29 (Figure 3B). In additional experiments, we found that lower beam energies of 3 and 5 keV lead to somewhat more efficient deposition, which is likely due
assembled monolayer (SAM) of poly(ethylene glycol) (PEG) is generated by PEG-silanization on an indium tin oxide (ITO) coated glass substrate. The 17 nm thick ITO film was used to prevent electrostatic charging of the substrate during e-beam exposure, while preserving the transparency of the substrate for subsequent optical assessment (Figure 1-1).28 Exposing the PEG film to a focused e-beam causes organic residues in the SEM chamber atmosphere to decompose and deposit near the irradiated substrate surface.26 The deposition process is mainly induced by secondary electrons with energies
