Few-Atom Fluorescent Silver Clusters Assemble at Programmed Sites

Letter pubs.acs.org/NanoLett

Few-Atom Fluorescent Silver Clusters Assemble at Programmed Sites on DNA Nanotubes Patrick R. O’Neill,†,∥ Kevin Young,†,⊥ Daniel Schiffels,†,§ and Deborah K. Fygenson*,†,‡ †

Physics Department and ‡Biomolecular Science and Engineering Program, University of California, Santa Barbara, California 93106, United States § Department of Physics, Ludwig-Maximilians-Universität, München, Germany S Supporting Information *

ABSTRACT: We show that DNA hairpins template the site-specific assembly of fluorescent few-atom Ag clusters on DNA nanotubes. Fluorescent clusters form only at hairpin sites and not on the double-stranded DNA scaffold, allowing for spatially programmed self-assembly. Ag clusters synthesized on hairpins protruding from DNA nanotubes can have nearly identical fluorescence spectra to those synthesized on free hairpins of identical sequence. Analysis of the stepwise photobleaching of individual clusters suggests a chemical yield of ∼45%. Given the well-established sequence-specific optical properties of DNA stabilized Ag clusters, these results point the way toward high yield assembly of metal cluster fluorophores with control over spectra as well as spatial arrangement. KEYWORDS: Self-assembly, DNA nanotechnology, patterning

T

that could be the basis for qubits or for detectors of molecular events.17 It is well-known that DNA oligomers can template the formation of few-atom Ag clusters (AgDNAs) with sequencedependent fluorescence.18,19,15 We wondered whether DNA nanostructures could be used to spatially control cluster assembly. In particular, we asked: are tiled DNA nanostructures stable under the chemical conditions that lead to Ag cluster formation on ssDNA? Do fluorescent Ag clusters self-assemble at or around nicks or crossovers, or can their formation be directed to specific sites? If directed, will the spectral signatures of the Ag clusters be altered from those formed on freely diffusing strands? And, if directed, what fraction of programmed “cluster sites” are occupied with fluorescent clusters postsynthesis, that is, what is the chemical yield? We addressed these questions using nanotubes assembled from DNA tiles, with hairpin appendages on a subset of the tiles serving as the cluster growth sites. We chose DNA tiles that assemble into stiff tubes many micrometers in length, making them amenable to characterization by fluorescence microscopy.20,21 This approach was critical to assessing the formation of few-atom clusters, which are unlikely to alter hairpin size or electron density sufficiently for detection by atomic force or electron microscopy.22 Using fluorescence microscopy, we were able to identify gross perturbations to the DNA nanostructures caused by Ag-cluster synthesis and verify

oday the utility of DNA extends well beyond its role in biology. Applications ranging from the detection of cocaine in blood serum1 to the purification of single-walled carbon nanotubes2 take advantage of the nucleobases’ ability to bind molecules or nanomaterials with sequence-dependent specificity. Other applications exploit Watson−Crick base pairing for programmed self-assembly of DNA nanostructures, which have proven versatile templates for precise positioning of functional nanoparticles, proteins, and small molecules.3−7 The combination of programmed self-assembly and direct DNA binding would allow for device synthesis in fewer steps and with greater complexity than currently possible, not least because of the limited suite of available linker chemistries. Controlled positioning based solely on DNA has been demonstrated for carbon nanotubes,8 but the simplicity gained by the DNA-only assembly of such devices is largely offset by the difficulty of mechanically/electrically coupling to them. To date, DNA-based positioning schemes have primarily focused on making optically active arrays of noble-metal nanoparticles,9 with an eye toward self-assembled plasmonics10 and fieldenhanced Raman spectroscopy.11 All of these, however, have relied on non-DNA linkages between the active elements and the DNA scaffold. In this work, we advance toward DNA-only positioning of optically active elements based on noble-metal atom clusters. In contrast to nanoparticles, noble-metal clusters comprised of ∼10 atoms12−14 exhibit fluorescent transitions in the visible and near-infrared15 rather than plasmons.16 Their unique optical properties suggest that the precise positioning of noble metal clusters with the nanometer-scale resolution afforded by DNA nanotechnology would enable near-field optical interactions © 2012 American Chemical Society

Received: May 10, 2012 Revised: September 29, 2012 Published: October 1, 2012 5464

dx.doi.org/10.1021/nl3017797 | Nano Lett. 2012, 12, 5464−5469

Nano Letters

Letter

AgDNAs. The mixtures were submerged in a 90 °C water bath and left to cool to room temperature over 2 days inside a Styrofoam box. Ag cluster synthesis was performed in solutions of hairpin tubes and bare tubes following the most common approach for AgDNA synthesis18 whereby AgNO3 is added to the DNA solution, followed by reduction of Ag+ to Ag with NaBH4. Ag+ ions bind to DNA bases (not the phosphate backbone),24 and reduction of bound Ag+ ions by BH4− catalyzes the formation of Ag clusters. The DNA guides the formation of Ag-atom clusters in a sequence-dependent manner, limits their growth, and stabilizes against aggregation. A portion of 50 μL of DNA tube solution (with or without hairpins) was mixed with 2.8 μL of 192 μM AgNO3. After 30 min at 4 °C, the mixture was reduced with 2.8 μL of 1 mM NaBH4 and stored at 4 °C. Final concentrations were 1.6 μM DNA tiles, 35.6 μM ammonium acetate, 8.9 μM magnesium acetate, 9.7 μM AgNO3, and 50.4 μM NaBH4. Both AgNO3 and NaBH4 were prepared fresh in water just minutes prior to use. For imaging, 10 μL of the Ag/ DNA tubes solution was mixed with 90 μL of a solution containing 5 mg/mL polyvinylalcohol (PVA, average molecular weight 16 kD), 40 mM ammonium acetate, and 10 mM magnesium acetate. This mixture was spin-cast onto a bare glass coverslip at 1700 rpm for 1 min to create a thin PVA film. The samples were imaged on an inverted fluorescence microscope (Olympus IX70) equipped with a CCD camera (DVC 1310), a 100×, 1.4 NA oil immersion objective (Olympus PlanApo, 100×, 1.4NA), and the following filter sets (from Omega Optical, except as noted): Green emitterEX: 475AF40, DC: 505DRLP, EM: 510ALP. Red emitterEX: 535RDF45, DC: 560DRLP, EM: HQ620/75 (Chroma Technologies). To examine the effect of Ag+ on DNA tubes, prior to reduction of the silver to form fluorescent clusters, hairpin tubes and bare tubes were imaged before and after the addition of AgNO3 (Figure 2). The 3′-end of strand no. 3 in these tubes was covalently attached to a fluorescein derivative (FAM) by

that fluorescent clusters form specifically at hairpins programmed into the nanotube lattice. Figure 1 shows a schematic of the tube design, adapted from ref 20. Each tube consists of a cylindrical lattice of identical tiles

Figure 1. Schematic of DNA nanotube with hairpin protrusions. Top left: tube, tiles, with hairpins represented by red asterisks. Top right: axial view showing outward orientation of the hairpins. Bottom: Tile, arrows point to the 3′ end. The hairpin region of strand no. 1 was omitted to create bare tubes.

comprised of five different DNA strands, numbered 1 to 5. Each tile consists of a pair of parallel double helices joined at two points where one strand from each helix crosses over to the adjacent helix. Single-stranded overhangs of five unpaired bases at each corner of the tile direct tile−tile interactions. Extensive characterization by atomic force and fluorescence microscopy has shown that such “double-crossover” or “DX” tiles assemble into hollow tubes roughly 10 nm in diameter,20 with exponential length distributions yielding many tubes 10 μm or longer.23 As originally designed, the nanotube lattice is completely double-stranded, save for unpaired overhangs at the tube ends. Fluorescent Ag clusters, on the other hand, preferentially form on single-stranded DNA.19 We therefore used a modified version of strand no. 1 to incorporate a DNA hairpin into the nanotube lattice, as previously described.20 The modification we used orients hairpins radially outward from the tube surface. We changed the sequence in the single-stranded loop region from four adenines to nine cytosines, because earlier work indicates that poly dC are better than poly dA at templating the formation of fluorescent silver clusters.19,12 DNA tubes were assembled with or without hairpins, referred to herein as hairpin tubes and bare tubes, respectively. Strands no. 2−5 were identical for both types of tubes. Hairpin tubes used the modified version of strand no. 1 described above, and bare tubes used the original version of strand no. 1 without the hairpin. The five DNA strands comprising the tile were mixed at 1.8 μM (each strand, see Supporting Information for sequences) in 40 mM ammonium acetate/10 mM magnesium acetate. The Tris-based buffer most commonly used for DNA self-assembly was avoided because AgDNA syntheses performed in Tris-acetate resulted in significantly less fluorescence compared to those in ammonium acetate, reflecting either reduced chemical- or quantum-yields for the

Figure 2. Fluorescent images of dye-labeled DNA tubes before and after addition of Ag+ ions. Bare tubes (top images) remain welldispersed on Ag+ addition, while Ag+ mediated interactions between hairpins causes tangling of the hairpin tubes. Scale bars: 10 μm. 5465

dx.doi.org/10.1021/nl3017797 | Nano Lett. 2012, 12, 5464−5469

Nano Letters

Letter

the manufacturer (IDT DNA). Initially well-dispersed hairpin tubes aggregated into large tangles in the presence of Ag+. Bare tubes, on the other hand, remained well-dispersed. Ag+ ions can mediate non-Watson−Crick base pairing25 and are reported to link cytosine homopolymers through cytosine− Ag+−cytosine interactions.26 Since the hairpin tubes present a high linear density of polycytosine loops, it is likely that C− Ag+-C bridges between hairpins on different tubes mediate the observed tangling. Ag+ ions can bind DNA bases in doublestranded DNA too.25 We therefore expect that Ag+ ions also bind to the bare DNA tubes. However, because all the bases in the bare tubes are already paired in double helices, there are no sites for Ag-mediated interactions between different tubes. Figure 3 shows the formation of red and green emitters on hairpin tubes after reduction of the Ag+ ions with NaBH4.

Figure 4. Ag clusters assembled on hairpin tubes exhibit nearly identical fluorescence spectra to those assembled on free hairpins. Hairpin tubes (solid curves) and free hairpins (dashed curves) stabilize the same green emitter (ex/em: 460/545 nm), while the red emitter (ex: 560 nm) that forms on hairpin tubes shows slightly blue-shifted emission relative to that on free hairpins (620 vs 628 nm).

was slightly blue-shifted compared to fluorescent species that form on free hairpins (620 vs 628 nm), but the three other peak wavelengths were the same to within the resolution of our measurement (±2 nm). Additional experiments were performed to differentiate between Ag cluster formation directly on the nanotubes and the possibility that nonspecific adsorption of free hairpinstabilized Ag clusters accounted for the nanotube fluorescence (Figure S1 of the Supporting Information). Ag clusters were synthesized in a solution of free hairpins, mixed with a solution of untreated hairpin tubes 1 day after Ag reduction, and imaged 1 day after mixing. Some weak green fluorescence was observed along the tubes, but the signal was small compared to the fluorescence observed when Ag was added and reduced directly in the solution of hairpin tubes. Dialysis of the free hairpinstabilized Ag clusters to remove free AgNO3 and NaBH4 prior to mixing with the untreated hairpin tubes did not substantially reduce the bulk fluorescence of the solution, but no fluorescence was observed along the tubes after mixing. These results indicate that Ag cluster fluorescence that forms along nanotubes following Ag reduction in solutions of hairpin tubes results primarily from the formation of clusters directly on the nanotubes at the programmed hairpin sites. The ability of hairpin protrusions to template the same Ag cluster fluorophores as free hairpins is consistent with the expectation, based on the high predicted melting temperature of the 9 base pair hairpin stem (57.5 °C), that free hairpins and hairpin protrusions should be similar in their conformational flexibility and in their ability to template fluorescent Ag cluster growth. The inability of bare tubes to support the assembly of fluorescent Ag clusters is consistent with previous observations that, while Ag+ ions can bind to DNA bases in both single and double-stranded DNA, only single-stranded DNA offers the flexibility and accessibility of the bases required for fluorescent Ag cluster formation.19 Our results show that this preferential formation on single-stranded DNA allows fluorescent cluster assembly to be confined on a DNA nanostructure to sites that present single-stranded regions, like DNA hairpins. Recently, Pal et al. demonstrated fluorescence resulting from Ag reduction on sugar labeled oligomers via the Tollens reaction.9 In this case there was no ssDNA on the DNA nanostructure during Ag-reduction. Labels were localized to one leg of a DNA origami triangle by hybridization to fully

Figure 3. Fluorescent Ag clusters on DNA tubes with hairpin appendages. Unlike the tubes in Figure 2, these tubes are not labeled with a fluorescent dye, so all fluorescence is attributable to Ag clusters. The green (ex: 440 nm, em: 540 nm) and red (ex: 560 nm, em: 620 nm) fluorescence observed by microscopy are consistent with the solution spectra in Figure 4. Both large nanotube tangles (A,B) and individual nanotubes (C,D) were decorated with red and green emitters. (A,B): 17 h after Ag reduction. (C,D): 1 week after reduction. Prior to imaging, the nanotubes were spin-cast in PVA. Scale bars: 10 μm.

These tubes were not dye-labeled, so Ag clusters provided the only fluorescence. Shortly after reduction (30 min), only red emitters were visible, but at longer times (1 h to 1 week), both emitters were present. While the majority of tubes, initially tangled by the addition of Ag+, remained tangled (Figure 3A,B), some individual tubes were found and were also labeled with red and green emitters (Figure 3C,D), illustrating that hairpin protrusions from individual tubes are sufficient to stabilize the fluorescent clusters. For the specific hairpin sequence studied here, hairpin tubes and free hairpins (in solutions without tubes) supported nearly identical Ag cluster fluorescence spectra (Figure 4): a green species with peak excitation/ emission wavelengths at 460/545 nm and a red species with peak excitation/emission at 560/620 nm. The latter emission 5466

dx.doi.org/10.1021/nl3017797 | Nano Lett. 2012, 12, 5464−5469

Nano Letters

Letter

Figure 5. Yield estimates of red Ag clusters synthesized on DNA nanotubes. DNA tubes assembled with 0.5% hairpin tiles were deposited on a supported lipid bilayer, followed by addition of AgNO3 and NaBH4. (A) A single DNA nanotube imaged two hours after Ag reduction, with individual fluorescent clusters visible along the tube contour. (B) All of the fluorescent clusters on the tube photobleached irreversibly within one minute of exposure to excitation light (535/45 nm). Intense nonphotobleaching spots away from the nanotube are presumably strongly scattering larger Ag aggregates. (C) Time sequence of a zoomed region of the nanotube, showing irreversible photobleaching of individual Ag clusters. The nanotube is also labeled uniformly with a fluorescent dye, and weak spectral bleed through makes the tube contour dimly visible. Scale bar: 5 μm.

complementary strands. TEM revealed the presence of crystalline nanoparticles, roughly two nanometers in diameter, suggesting each Ag particle consisted of hundreds of Ag atoms. In contrast, our fluorophore-decorated tubes show fluorescence spectra that are consistent with the formation of few-atom Ag clusters. In particular, the red species has excitation and emission spectra nearly identical to a DNA stabilized Ag cluster previously identified as Ag13.12 A high chemical yield of Ag fluorophores will be essential for elaborate patterning. Chemical yields have rarely been reported for fluorescent AgDNAs, and their measurement typically requires determination of the emitter concentration by fluorescence correlation spectroscopy.25 The ability to limit Ag cluster growth to hairpin sites on DNA nanotubes provides a unique means of estimating the chemical yield by counting individual emitters along tubes presenting a known linear density of hairpins. Under the conditions used in Figure 3, the high linear density of emitters and the tangling of the tubes prevented such yield measurements. To enable yield estimates, the fraction of tiles containing hairpins was lowered from 100% to 0.5%, and tubes were immobilized by adsorption on a supported lipid bilayer prior to Ag cluster synthesis. Bilayer formation, DNA nanotube adsorption, and AgDNA synthesis were performed in a flow cell.26 Briefly, glass coverslips were cleaned, exposed to UV-ozone, and kept under vacuum before construction of the flow cell. A high concentration of vesicle solution (DOPC, from Avanti Polar Lipids) was introduced into the flow cell to form the bilayer. After a few minutes, the flow cell was rinsed with buffer (40 mM ammonium acetate with 10 mM magnesium acetate), followed by addition of 3 μL of DNA nanotubes at a total tile concentration of 2 μM. After

waiting 20 min to allow adsorption of tubes on the bilayer, small volumes of AgNO3 and NaBH4 were added to the flow cell to create final concentrations of 10 μM each. Tubes were imaged using the microscope setup described above, with the following filter combinations: Cy5 (EX: 585/40 nm, EM: 660/ 32 nm), red AgDNA (EX: 535/45 nm, EM: 610/70 nm). Figure 5 shows a single tube, adsorbed on a bilayer coated glass surface, two hours after the addition of AgNO3 and reduction with NaBH4. Under these conditions, red emitters formed at a high enough linear density to observe the contour of the tube, but at a low enough density that individual emitters could be spatially resolved. The tubes were also uniformly labeled with Cy5 to allow for visualization of tubes on the bilayer prior to Ag cluster synthesis. Cy5 excites and fluoresces at longer wavelengths than the red Ag clusters, which enabled Ag clusters and Cy5 to be imaged in different channels. Nevertheless, Cy5 fluorescence was weakly excited and collected in the Ag cluster channel, making the entire tube contour faintly visible (Figure 5C). Irreversible photobleaching of the red emitters allowed us to count the individual emitters. We counted 154 emitters over 141 μm of DNA nanotubes (3 tubes), corresponding to an average of 1 emitter per 916 nm of tube. Previous AFM work put bounds on the number of tiles in the circumference of 4− 10 tiles, and a model of the tube stiffness for our guess of 7 tiles is consistent with previous stiffness measurements for nanotubes annealed according to the protocol used here.20 Using 15 nm for the length of a tile, 7 tiles in circumference, and 0.5% hairpin tiles gives a mean linear density of hairpins corresponding to 1 hairpin over about 430 nm of tube. The fraction of hairpins housing emitters is then 430/916 = 0.469, 5467

dx.doi.org/10.1021/nl3017797 | Nano Lett. 2012, 12, 5464−5469

Nano Letters

Letter

indicating a yield of ∼45%. This yield estimate assumes random assembly of tiles. It is conceivable that tiles with hairpins may preferentially bind to one another, in which case we may have overestimated the yield. However, we identified tubes for analysis by their Cy5 fluorescence prior to imaging individual silver clusters along the contour of the tube, so it is therefore unlikely that our measurements are biased toward nanotubes with higher hairpin densities. Ag cluster synthesis on DNA commonly produces red emitters at short times after Ag reduction, followed by their oxidation and the emergence of green emitters on the time scale of hours.12,27 Consistent with this trend, the yield of red emitters appears to be much higher in Figure 5A (2 h after reduction, 0.5% hairpin tiles) than in Figure 3D (1 week after reduction, 100% hairpin tiles). Conversely, we did not observe any green Ag clusters on DNA nanotubes in the flow cell, where the slow generation of silver precipitate prevented imaging more than 12 h after reduction. Also contributing to the lack of observed individual green emitters could be their generally poorer photostability in aqueous solution compared to PVA and their lower quantum yield compared to red emitters.15 In summary, we have shown that DNA hairpins can template the site-specific assembly of fluorescent few-atom silver clusters on DNA nanotubes. Fluorescent Ag clusters form only at the hairpins: bare tubes do not support fluorescent cluster formation, and Ag clusters that form on hairpin tubes can have nearly identical fluorescence spectra to those synthesized on free hairpins consisting of the same loop and stem sequence. These results suggest that it may be possible to use DNA nanostructures to assemble fluorescent Ag clusters with high yield and nanometer control over cluster position. Given the well-documented sequence-dependent fluorescence of DNAtemplated silver clusters,19,28,15,28 these results point the way toward self-assembly of Ag clusters with control over fluorescence spectra as well as spatial position. With a better structural understanding of DNA stabilized Ag clusters, control over transition dipole orientations might also be achieved. Spectrally, spatially, and orientationally controlled assembly of Ag cluster arrays on DNA nanostructures would ultimately enable new experiments in energy transfer and quantum optics.

■

manuscript. D.S. is grateful to the Elite Network of Bavaria (IDK-NBT) for a doctoral fellowship and to BaCaTec for travel support. This work was supported by NSF CHE-0848375 and by the NSF MRSEC program under Award No. DMR-1121053 through a RISE fellowship to K.Y.

■

ASSOCIATED CONTENT

S Supporting Information *

DNA sequences, evidence against non-specific adsorption. This material is available free of charge via the Internet at http:// pubs.acs.org.

■

REFERENCES

(1) Swensen, J. S.; Xiao, Y.; Ferguson, B. S.; Lubin, A. A.; Lai, R. Y.; Heeger, A. J.; Plaxco, K. W.; Soh, H. T. J. Am. Chem. Soc. 2009, 131, 4262−4266. (2) Tu, X. M.; Manohar, S.; Jagota, A.; Zheng, M. Nature 2009, 460, 250−253. (3) Sharma, J.; Ke, Y. G.; Lin, C. X.; Chhabra, R.; Wang, Q. B.; Nangreave, J.; Liu, Y.; Yan, H. Angew. Chem., Int. Ed. 2008, 47, 5157− 5159. (4) Chhabra, R.; Sharma, J.; Ke, Y. G.; Liu, Y.; Rinker, S.; Lindsay, S.; Yan, H. J. Am. Chem. Soc. 2007, 129, 10304−10305. (5) Stephanopoulos, N.; Liu, M. H.; Tong, G. J.; Li, Z.; Liu, Y.; Yan, H.; Francis, M. B. Nano Lett. 2010, 10, 2714−2720. (6) Voigt, N. V.; Torring, T.; Rotaru, A.; Jacobsen, M. F.; Ravnsbaek, J. B.; Subramani, R.; Mamdouh, W.; Kjems, J.; Mokhir, A.; Besenbacher, F.; Gothelf, K. V. Nat. Nanotechnol. 2010, 5, 200−203. (7) Bui, H.; Onodera, C.; Kidwell, C.; Tan, Y.; Graugnard, E.; Kuang, W.; Lee, J.; Knowlton, W. B.; Yurke, B.; Hughes, W. L. Nano Lett. 2010, 10, 3367−3372. (8) Maune, H. T.; Han, S. P.; Barish, R. D.; Bockrath, M.; Goddard, W. A.; Rothemund, P. W. K.; Winfree, E. Nat. Nanotechnol. 2010, 5, 61−66. (9) Pal, S.; Deng, Z. T.; Ding, B. Q.; Yan, H.; Liu, Y. Angew. Chem., Int. Ed. 2010, 49, 2700−2704. (10) Tan, S. J.; Campolongo, M. J.; Luo, D.; Cheng, W. Nat. Nano 2011, 6, 268−276. (11) Ding, B. Q.; Deng, Z. T.; Yan, H.; Cabrini, S.; Zuckermann, R. N.; Bokor, J. J. Am. Chem. Soc. 2010, 132, 3248−3249. (12) O’Neill, P. R.; Velazquez, L. R.; Dunn, D. G.; Gwinn, E. G.; Fygenson, D. K. J. Phys. Chem. C 2009, 113, 4229−4233. (13) Petty, J. T.; Fan, C. Y.; Story, S. P.; Sengupta, B.; Iyer, A. S.; Prudowsky, Z.; Dickson, R. M. J. Phys. Chem. Lett. 2010, 1, 2524− 2529. (14) Schultz, D. E.; Gwinn, E. Chem. Commun. 2012, 48, 5748−5750. (15) Richards, C. I.; Choi, S.; Hsiang, J.-C.; Antoku, Y.; Vosch, T.; Bongiorno, A.; Tzeng, Y.-L.; Dickson, R. M. J. Am. Chem. Soc. 2008, 130, 5038−5039. (16) Zheng, J.; Nicovich, P. R.; Dickson, R. M. Annu. Rev. Phys. Chem. 2007, 58, 409−431. (17) Yeh, H.-C.; Sharma, J.; Han, J. J.; Martinez, J. S.; Werner, J. H. Nano Lett. 2010, 10, 3106−3110. (18) Petty, J. T.; Zheng, J.; Hud, N. V.; Dickson, R. M. J. Am. Chem. Soc. 2004, 126, 5207−5212. (19) Gwinn, E. G.; O’Neill, P.; Guerrero, A. J.; Bouwmeester, D.; Fygenson, D. K. Adv. Mater. 2008, 20, 279−283. (20) Rothemund, P. W. K.; Ekani-Nkodo, A.; Papadakis, N.; Kumar, A.; Fygenson, D. K.; Winfree, E. J. Am. Chem. Soc. 2004, 126, 16344− 16352. (21) O’Neill, P.; Rothemund, P. W. K.; Kumar, A.; Fygenson, D. K. Nano Lett. 2006, 6, 1379−1383. (22) Driehorst, T.; O’Neill, P.; Goodwin, P. M.; Pennathur, S.; Fygenson, D. K. Langmuir 2011, 27, 8923−8933. (23) Ekani-Nkodo, A.; Kumar, A.; Fygenson, D. K. Phys. Rev. Lett. 2004, 93, 268301. (24) Arakawa, H.; Neault, J. F.; Tajmir-Riahi, H. A. Biophys. J. 2001, 81, 1580−1587. (25) Sengupta, B.; Ritchie, C. M.; Buckman, J. G.; Johnsen, K. R.; Goodwin, P. M.; Petty, J. T. J. Phys. Chem. C 2008, 112, 18776−18782. (26) Weirich, K. L.; Israelachvili, J. N.; Fygenson, D. K. Biophys. J. 2010, 98, 85−92. (27) Ritchie, C. M.; Johnsen, K. R.; Kiser, J. R.; Antoku, Y.; Dickson, R. M.; Petty, J. T. J. Phys. Chem. C 2007, 111, 175−181.

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Present Addresses ∥

Department of Anesthesiology, Washington University School of Medicine, St. Louis, MO 63110. ⊥ Bioengineering Department, University of California, San Diego, La Jolla, CA 92093. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The authors thank Kim Weirich for assistance with flow cell experiments and P.W.K. Rothemund for a critical reading of the 5468

dx.doi.org/10.1021/nl3017797 | Nano Lett. 2012, 12, 5464−5469

Nano Letters

Letter

(28) Sharma, J.; Yeh, H. C.; Yoo, H.; Werner, J. H.; Martinez, J. S. Chem. Commun. 2010, 46, 3280−3282.

5469

dx.doi.org/10.1021/nl3017797 | Nano Lett. 2012, 12, 5464−5469