Subscriber access provided by Kaohsiung Medical University
Biological and Environmental Phenomena at the Interface
The Sapphire (0001) Surface: A Transparent and Ultra-flat Substrate for DNA Nanostructure Imaging Masudur Rahman, Zachary Boggs, David Neff, and Michael L. Norton Langmuir, Just Accepted Manuscript • DOI: 10.1021/acs.langmuir.8b01851 • Publication Date (Web): 15 Aug 2018 Downloaded from http://pubs.acs.org on August 21, 2018
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 9 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Langmuir
The Sapphire (0001) Surface: A Transparent and Ultra-flat Substrate for DNA Nanostructure Imaging Masudur Rahman, Zachary Boggs, David Neff and Michael Norton Department of Chemistry, Marshall University, Huntington, West Virginia, USA
Supporting Information Placeholder ABSTRACT: Mica is the current substrate of choice for
DNA nanostructure imaging; mainly due to its atomically flat surface. However, these mica substrates are often not optically clear. In this work, sapphire has been evaluated as an alternative substrate, with potential to enable parallel optical and AFM studies. Well known for its thermal and chemical, sapphire is a hard ionic material with excellent optical properties. Because sapphire lacks the excellent basal cleavage properties of the sheet silicate mica, a process to anneal it at high temperature in water vapor was developed to achieve near atomically smooth (average roughness = 0.141nm) terraces. AFM imaging was used to determine the dimensions of these terraces and to characterize the morphology of the DNA nanostructures, revealing that their structures were preserved, indicating that annealed c-plane cut (0001) sapphire is a promising substitute for mica as a flat and transparent substrate for DNA nanostructure studies.
INTRODUCTION: Mica has served as the standard substrate for DNA nanostructure research from its very beginning. In addition to its atomically flat surface it also displays moderately strong adhesion to DNA. While based on its nominal chemical composition mica should be intrinsically transparent, the mica substrates often employed are natural mineral samples, with attendant high compositional variability from sample to sample. Most samples have a combination of light scattering gas inclusions (“bubbles”) and light absorbing chemical contaminants. Therefore mica, at present, is not a realistic candidate as a commodity optically transparent material. When alternative ultra-flat substrates, for example graphite1-2 and graphene3-4 have been tested, interactions with these surfaces have been found to induce unanticipated distortions of the DNA based test structure. For example, when cross-shaped DNA origami5 are applied to graphite, significant rearrangement of the DNA structure is observed, as demonstrated in Figure 1.
Figure 1 Comparison of DNA nanostructures on mica and graphite. A is on mica. B is on graphite. Significant rearrangement can be seen on graphite.1
Similar distortions were observed when MoS2 was evaluated as a substrate.6 Although several other air stable layered materials remain to be studied, this trend of distortions being induced by the surfaces of two dimensional van der Waals solids motivated a consideration of 3D materials. It is possible that the same van der Waals interactions which lead to layer to layer binding in these 2D materials may also lead to competitive binding of their surfaces to DNA, causing partial disruption of the hydrogen bonding responsible for maintaining the structure of DNA origami constructs. Sapphire, Al2O3 or α-alumina, is an ionic 3D compound with excellent mechanical, thermal, chemical and optical properties. There are several aspects of sapphire which would lead one to anticipate that DNA structures will adhere to sapphire without the disruption observed with the van der Waals materials. First, in common with mica layers, the major forces stabilizing the structure of sapphire are ionic. The structure of the unit cell for sapphire is quite complex, consisting of 6 Al-O layers, as depicted in Figure 2.7 The layers in muscovite mica, the form of mica most commonly used for AFM imaging, with formula KAl2(AlSi3O10)(OH)2, are also bound together by ions, usually potassium ions, rather than by van der Waals forces. Secondly, the isoelectric point of the 0001 face of single crystal sapphire has been determined to be between 48 and 4.59. The surface charge
ACS Paragon Plus Environment
Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 2 of 9
Figure 2 Sapphire unit cell. a=4.758 Å, c=12.991 Å Only the Al (black) atoms are shown in the above diagram. The rest (white) are empty sites. Due to slight positional differences in each layer, the unit cell only repeats every six layers. (Drawing adapted from reference 7)
state is necessarily made more complex by the six different types of surface hydroxyls attendant with the suggested reconstruction of the sapphire surface to a Gibbsite-like structure in aqueous solution.10 AFM studies support hydroxylation of the surface.11 While their isoelectric points do differ significantly, with mica having a reported value ranging from 3 – 3.5, this does mean that at pH above 7, where most DNA imaging is performed, the hydrated surface of sapphire will be, like mica, negatively charged.12 This allows ionic bridging to be a strong contributor to the adhesion of nanostructures to the surface of sapphire. In fact, double stranded DNA bound to sapphire was imaged using AFM well before the advent of DNA origami nanostructures.13 This paper details the route to achieving the twin objectives of generating near atomically smooth terraces with lateral dimensions useful for DNA nanostructure imaging while retaining high optical quality and determining the compatibility of these surfaces with origami, enabling them to maintain their designed structures. The surface of sapphire, as received from optical window manufacturers, is optically, but not atomically flat. Many articles have been written pertaining to the development of flat terraces on the surface of 0001 sapphire. The consensus appears to be that higher temperatures, between 1300 and 1500oC, temperatures not obtainable with the laboratory furnaces available in most laboratories, are required to achieve extended ter-
Figure 3 AFM image and associated line profile analysis of terrace heights from a sample annealed for 18 hours in air (image above graph). The miscut angle of the single crystal sapphire was calculated from the above image, where tangent of the miscut angle = 0.024 µm/ 5 µm, for a miscut angle of 0.3 degrees.
races.14-15 These high temperatures are required in order to increase surface diffusion but more importantly, to enable “bunching” of elementary steps, the formation of higher steps through the coalescence of smaller steps. The minimum average width of a terrace can be predicted based on the miscut angle, the angle between the normal to the surface and the crystallographic c axis.16 For example, a miscut angle (M) of 0.3 degrees, quite acceptable in optical window materials, would lead to a terrace width (λ) of 43nm according to equation 1, assuming all steps are O-Al-O elementary steps of 2.164 Å, the distance between planes shown in Figure 2. λ = (c/6) / (sin M)
Equation 117
Fortunately terraces much wider than this prediction form because of step bunching. As an example, an AFM image of a high temperature annealed sample with non-elementary steps, enabling the formation of terraces an order of magnitude greater in width than this prediction, is provided in Figure 3. To enhance the surface mobility and increase step bunching while maintaining a relatively low (1200oC) processing temperature, this work annealed sapphire
ACS Paragon Plus Environment
2
Page 3 of 9 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Langmuir
samples under an atmosphere of pure water vapor. Under these conditions, surface species volatilization through reaction of alumina with water to form Al(OH)3 is expected to cause etching. Etching at the higher temperature of 1450oC in 68% water vapor/ 32% oxygen has been reported to yield terraces on the order of 20 microns in extent.18
MATERIALS AND METHODS Sapphire Substrate Preparation and Annealing: Round Sapphire windows (part #MSPW100/010M 1inch diameter, 1mm thick) used for producing images for Figures 3 and 4 and Table 1 were purchased from Meller Optics of Providence, RI. All windows purchased were from the same lot. In order to allow these windows to fit into the quartz tube furnace seen in Figure S1 the windows had to be mechanically cleaved. Cleaving was performed by a rapid impact on the center of the window with a sharp SiC tipped awl. Windows employed for Figures 5 and 6, also from Meller Optics, were 2 inch diameter, 0.2 mm thick (part #SCD2470-02A). These thin plates were cleaved by scribing along an edge. Sapphire is rinsed with MeOH on each side. The edge is then allowed to touch a chemwipe to wick off excess methanol, then air dried. An alumina (~ 7.6cm long, 5 mm wide) boat is used to maintain the sapphire in a near vertical orientation. The sapphire only contacts the boat at two points, at the bottom edge and near the middle of the sapphire. Annealing is performed in a long unsealed quartz tube envelope centered in a tube furnace, as presented in Figure S1. The system is maintained at 100 oC, prior to annealing. The furnace is ramped up to full temperature over several separate temperature changes with an average heating rate of 650oC/hour. The cooldown is a similar process with multiple temperature changes, yielding an average cooling rate of 220oC/hour, until the 100 oC idle temperature is reached. An example Time-Temperature annealing profile is included in the supplementary material as Figure S2. For samples annealed in pure water vapor, this semi-hydrothermal condition was maintained by feeding vapor from a round bottom flask containing boiling distilled water into the furnace tube throughout the annealing process. Safety Note: Superheated steam can cause significant burns and can ignite flammable materials. Care should be taken to prevent contact with hot vapor including unpredictable rises in pressure which can accompany “bumping” in the water evaporator. The measured liquid water transport rate
of 0.2g/min translates to a water vapor transport rate of 1.3L/min at 1200oC. AFM Imaging The AFM imaging studies were performed using a Bruker Multimode 8 and Nanoscope VI controller (Billerica, MA, USA) in the SCANASYST-AIR mode, which uses PEAKFORCE tapping, a non-resonant, 2000 Hz tapping-like imaging mode. Tips on this device consist of silicon nitride with a spring constant of 0.4 N/m. These tips also have a nominal radius of 2 nm. Production of Cross shaped DNA Origami A schematic outline of the process for the production of cross shaped DNA origami is shown in Figure S3. Briefly, a mixture of staple strand DNA oligomers (30 nM) and M13mp18 DNA genome (10 nM) was brought to a volume of 60 uL using 1x TAE buffer solution containing 40 mM Tris-HCl at a pH 8.0, 20 mM acetic acid, 1 mM EDTA, and 12.5 mM magnesium chloride (origami buffer). The constructs were annealed in a slow cooling process, which drops the temperature linearly from 90 to 20 oC over a 13 hour period. Excess staples were removed from solution by the purification process described below. This preparation is based on the work published by Liu et al.5 The one dimensional arrays (1D) of origami shown in Figures 5, 6 and Figure S9 were produced by utilizing staples complementary to two opposite arms of the cross shaped structures.19 The origami shown in S9 has the same dimensions as the Liu origami but has all unique staple sequences designed in this lab. 1D origami constructs were generated with each origami containing one (Figure 6) or three (Figure S9) staple(s) modified with Cy3 (IDT) in place of biotin for use as fluorescent test objects. Purification of DNA Origami The origami solutions are filtered using a 30 kDa MW centrifuge filter (Amicon Ultra from Millipore/Sigma). The retained solutions are then washed three times, for a final dilution factor of ~8000:1, with 400 uL of 1x DNA origami buffer using a benchtop microcentrifuge. After filtration, the retrieved solution was stored at 4 o C.
ACS Paragon Plus Environment
3
Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 4 of 9
Table 1: Analysis of Substrate Surface Roughness
Table 1 Roughness analysis of terraces, representative AFM images and histograms of terrace widths for well cut sapphire samples after various annealing conditions, including as received (for expanded views, see supplementary material Figure S4). Ra = average of absolute values of Z deviation from the mean plane. All images 1.7x1.7 microns in x and y and 7 nm in the z dimension. Histograms of the terrace widths of all 5 samples (conditions above each histogram). Longer annealing times generate much larger terraces and the addition of water vapor also aids in terrace development. Mica value is provided for reference.
Origami Deposition onto Sapphire To evaluate the viability of sapphire as an alternative substrate, surfaces were dosed with DNA origami in origami buffer for 5 minutes, rinsed with 50-100uL of doubly distilled H2O, then quickly blown dry with a stream of dry nitrogen gas. This same treatment was used for the 0D and the 1D constructs. Correlative Microscopy (AFM/Fluorescence) A homebuilt system consisting of a FastScan (Bruker) AFM coupled to a Nikon Eclipse TE300 stand, based on the design reported by Fantner et al.20, was used to produce the correlated image presented in Figure 6 and Fig. S9. Cy3 was excited using a DPSS (LASOS) 532nm laser. Images were captured using a Pro EM512 (Princeton Instruments) CCD Camera.
RESULTS AND DISCUSSION: Initial studies were performed to study the evolution of the sapphire surface with different annealing protocols and are presented next. Sapphire Surface Morphology Development with Annealing The maximal annealing temperature employed in this study was limited to 1200 C, the maximum of most commonly available laboratory tube furnaces. As delivered, optically polished sapphire is mechanically amorphized at the surface. Snapshots of the evolution of the surface with annealing are provided in Table 1.
ACS Paragon Plus Environment
4
Page 5 of 9 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Langmuir
Figure 4 0D DNA origami applied to sapphire surfaces annealed at 1200 oC in a) air for 18 hours and b) water vapor for 14 hours. Significant differences in origami quality are apparent. The 14 hour water annealed surface appears to be significantly more compatible with the origami structure. Images are 1.7 X 1.7µm2 with a z value range of 7nm.
While a short (1 hr) anneal at 1100C in air is not sufficient to obliterate the scratches on polished, as delivered sapphire, terraces do become observable. The higher roughness of the 1100 C annealed sample compared with the as received sample is attributed to surface coarsening, including erosion of amorphous material and terrace emergence. Individual terrace roughness was not measured for this sample due to the minimal size of these terraces. As anticipated for a thermally activated, diffusion related process, increased annealing temperature, longer annealing time and high-water vapor content leads to larger terrace widths, with an unavoidable increase in step height. A roughness analysis was performed on all the sapphire samples prepared and unless otherwise noted, average single terrace values are included in Table 1. From this data it appears that the smoothest samples result from an 18 hour air, rather than water, annealing protocol. However, the effects of sub-angstrom differences in roughness on image quality appear minimal, as discussed below. Noting that the samples are quite uniform, with the Ra values being quite consistent between multiple terraces of a single annealing run, the water annealed samples’ Ra values are slightly higher, possibly reflecting a roughness contribution from Al(OH)3 cluster precipitation/decomposition during annealing.11 It is noted that these samples never achieve the baseline Ra of 0.0719 nm observed in this study for mica, a value which convolves intrinsic noise
of our AFM system and true roughness of the mica surface. Temperatures above 1200C may produce terraces with smother surfaces, however these temperatures are not readily available to most researchers. Also, longer annealing times than the arbitrarily selected periods studied here will provide larger terraces. Histograms representing the terrace widths for all five treatments of the sapphire samples are included in Table 1. Clearly longer annealing times give larger terraces, with some displaying widths of 2 µm, sufficient to support many origami on a single terrace. The effect of the hydrated environment in enhancing terrace growth when compared to dry air conditions is clear. The transport of alumina, in the form of Al(OH)3 is the most probable mechanism for water aiding in the growth rate of terraces at these temperatures.18 Although the Ra values increase under water annealing conditions, the larger terraces made available for origami placement recommends the use of water vapor during substrate annealing. Compatibility with Origami Structure Imaging
ACS Paragon Plus Environment
5
Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
The cross-shaped origami designed by Liu et al.5 has been used in this work as a test object. Two parameters of this “test” origami form may be used to evaluate the utility of sapphire as a substrate for origami. The first parameter, edge to edge width of the central, raised portion of the origami, was adopted in view of the observation that several incompatible substrates “expand” origami1-4, 6 presumably through partial melting of the structure. The second parameter is the measured depth of the corrugation, the “trough” that runs through the center of the origami. While certainly related to surface roughness as moderated by the molecular layer of DNA, this parameter reflects the functional image contrast, important for visualization and arguably one of the reasons for the popularity of mica, as opposed to glass, as a substrate. Figure 4 presents AFM images of well separated single (0D) cross shaped DNA origami applied to a well-cut substrate annealed at 1200oC for A) 14 hours in water vapor and B) 18 hours in air. Example images reflecting origami deposition on three surfaces resulting from other annealing protocols are included in Figure S5. It is apparent, both from inspection and from analysis that the image quality of the origami on the water annealed sample rivals that obtained using mica substrates. The corrugation depth is identical (Dmica = 1.36nm, Dsapphire = 1.35nm) within tip width induced error, and the width of the origami is essentially not expanded, W mica = 35.3nm, Wsapphire = 35.8nm) within experimental (tip convolution) error (depth and width analysis provided in supplementary information Figure S6). It may be noted that when one of the 0D origami bridges across the joining edge of two terraces, there is no apparent change in the structure for small steps. In contrast, images of the air annealed surface reflect significant restructuring of the origami, including apparent folding over and the appearance of holes in a fraction of the structures. While the air annealed sapphire surface appears indistinguishable from the water annealed surface as observed by AFM, the DNA origami, serving as probes, may reveal that chemically different environments are presented by the air annealed surfaces. Al2O3 is known to present different acidic sites when presented in different crystallographic forms.21 The surface reorganization produced under the specific annealing and application conditions used here is not well understood. It has been suggested that the hydrated alumina surface is not fully dehydrated until temperatures above 700oC are reached.10 The folding-over behavior observed for the air annealed samples is never observed on mica, and may be ascribed to a high density of strong binding
Page 6 of 9
sites. A recent study seeking to mimic the optimal surface binding properties of mica observed such folding over behavior for rectangular origami on a high density amine modified oxide on silicon surface.22 Because the originating event leading to the “folding over” of these paper like structures was not observed, one may hypothesize that they either result from “freezing in” of randomly folded solution phase structures, which for DNA plasmids has been associated with strong binding,23 or surface adhesion of one arm of the origami, followed by folding over by the flow of rinse solution over the substrate. Either genesis of this folding event can be ascribed to strong surface sites binding the DNA structure. The observation of the development of holes in the origami structures, which may be ascribed to the loss of staples from the structures, is also consistent with the presence of strong pinning or binding sites capable of removing staples from an intact origami. The nature of the pinning site responsible for these observed defects is not certain at this time, and will require further study of the surface chemistry of sapphire and its interactions with DNA. As has been the case with mica,24-26 it is possible that different solution phase surface treatments can titrate the binding to the surface.
Figure 5 1D origami applied to 15 hour water vapor annealed sample. No significant structural variability is apparent over a distance of ~ 2 microns as the structure crosses several terraces.Imaged area: 1.4 X 1.1 µm2 and z range is 8.4nm.
ACS Paragon Plus Environment
6
Page 7 of 9 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Langmuir
Because we did observe these indications of non-ideal origami-substrate interactions for sapphire annealed at high temperature in air, further investigations were limited to water vapor annealed samples, 1D DNA Origami Arrays (15 Hour Water Vapor) Stacking effects are known to influence the surface electronic states in materials. While the surface atomic arrangements were not determined in these studies, an attempt to probe different stacking sequences using extended origami arrays (1D arrays) was performed. Such 1D arrays can be produced in lengths sufficient to span several terraces.19 If the step height is sufficiently small, then these arrays would necessarily land on different unit layers within the unit cell shown in Figure 1. We would anticipate that significant differences in surface binding, or variability among surface micro- regions, may be revealed by differences in the otherwise quite homogeneous appearance of monomers of the extended DNA array. Figure 5 provides an AFM image of a 1D origami array applied to a 200um thick 15 hour, 1200oC, water annealed window with apparent 0.14 degree miscut angle and slightly higher step density. No significant variability is observed across the array, indicating that for origami imaging, placement location will not significantly impact origami structure. This is certainly an important assumption which can be made with certainty when using mica as a substrate. Correlative Microscopy (AFM/Fluorescence Microscopy) of 1D DNA Origami Arrays 1D arrays composed of cross shaped origami, each decorated with one Cy3 fluorophore modified staple, were imaged via AFM and wide field fluorescence microscopy. Supplemental Figure S7 provides AFM characterization data for the sapphire window employed in these optical imaging studies. Figure 6 and Figure S9 provide examples of correlative AFM/Optical microscopy enabled by these sapphire
Figure 6: A) AFM image, B) Fluorescence Image (129 nm/pixel) and C) overlay for curved array of singly Cy3 labeled origami. Images are 1.2µm X 1.0µm.
substrates. The acquired image scales differ significantly with our camera chip capturing 66um x 66um regions and the FastScan AFM scanner being limited to
