Broadband Fluorescence Enhancement with Self-Assembled Silver

Figure 1. (a) Sketch of the dimer OA consisting of two 80 nm Ag particles attached to a DNA origami pillar together with a top-view (left inset). The ...
0 downloads 10 Views 2MB Size
Download PDF
Broadband Fluorescence Enhancement with Self-Assembled Silver Nanoparticle Optical Antennas Carolin Vietz,†,‡ Izabela Kaminska,†,⊥,‡ Maria Sanz Paz,†,∥ Philip Tinnefeld,*,† and Guillermo P. Acuna*,† †

Institute for Physical and Theoretical Chemistry, Braunschweig Integrated Centre of Systems Biology (BRICS), and Laboratory for Emerging Nanometrology (LENA), Braunschweig University of Technology, 38106 Braunschweig, Germany ⊥ Institute of Physics, Faculty of Physics, Astronomy, and Informatics, Nicolaus Copernicus University, Grudziadzka 5, 87-100 Torun, Poland S Supporting Information *

ABSTRACT: Plasmonic structures are known to affect the fluorescence properties of dyes placed in close proximity. This effect has been exploited in combination with singlemolecule techniques for several applications in the field of biosensing. Among these plasmonic structures, top-down zero-mode waveguides stand out due to their broadband capabilities. In contrast, optical antennas based on gold nanostructures exhibit fluorescence enhancement on a narrow fraction of the visible spectrum typically restricted to the red to near-infrared region. In this contribution, we exploit the DNA origami technique to self-assemble optical antennas based on large (80 nm) silver nanoparticles. We have studied the performance of these antennas with farand near-field simulations and characterized them experimentally with single-molecule fluorescence measurements. We demonstrate that silver-based optical antennas can yield a fluorescence enhancement of more than 2 orders of magnitude throughout the visible spectral range for high intrinsic quantum yield dyes. Additionally, a comparison between the performance of gold and silver-based antennas is included. The results indicate that silver-based antennas strongly outperform their gold counterparts in the blue and green ranges and exhibit marginal differences in the red range. These characteristics render silver-based optical antennas ready for applications involving several fluorescently labeled species across the visible spectrum. KEYWORDS: fluorescence enhancement, DNA origami, single-molecule detection, plasmonics, nanophotonics, silver nanoparticles

I

the micromolar range by reducing the observation volume and enhancing the fluorescence signal.6 There are two main available techniques: zero-mode waveguides (also known as nanoapertures) and optical antennas (OAs).7,8 Additionally, combinations of both techniques were also demonstrated.9 Zero-mode waveguides, in which the main effect exploited is a reduction of the observation volume, stand out as the only commercially available technique. One of the reasons of the success of this approach lies in the fact that a reduction of the observation volume can be achieved over a broad fraction of the visible spectrum. The diameter of the nanoaperture determines a cutoff wavelength above which light cannot propagate and is

n order to gain insight into the dynamics of complex and inhomogeneous biomolecular multicomponent structures, single-molecule techniques represent a major step forward from ensemble average measurements.1 In particular, singlemolecule fluorescence techniques are exploited, for example, in DNA sequencing,2 RNA expression3 and diagnostics.4 Singlemolecule detection requires the isolation of the fluorescence signal of a single molecule not only from background contributions but also from the signals of other equivalent molecules or impurities. In conventional microscopy techniques, the smallest observation volume is determined by the diffraction limit. This constrains single-molecule detection to the nanomolar range, while in many biological systems microto millimolar concentrations are reached.5 Over the last decades, nanophotonic approaches were developed to boost single-molecule fluorescence detection to © 2017 American Chemical Society

Received: March 7, 2017 Accepted: April 26, 2017 Published: April 26, 2017 4969

DOI: 10.1021/acsnano.7b01621 ACS Nano 2017, 11, 4969−4975

Article

www.acsnano.org

Article

ACS Nano confined to a reduced volume in the zeptoliter range.5 Therefore, experiments involving several species labeled with different fluorophores can be performed on the same platform.10 The most prominent example of the application of this technique is single-molecule real-time DNA sequencing in which the incorporation of single fluorescently labeled nucleotides by a polymerase immobilized in a zero-mode waveguide can be monitored at a concentration range of nucleotides between 0.1 and 10 μM.2 Among other approaches,11,12 dimer OAs consist typically of two metallic nanostructures placed in close vicinity forming a nanometric gap.13,14 The optical response of these OAs is determined by the size, shape, and material composition of the elements, gap geometry, and optical properties of the surrounding medium. Upon suitable illumination, localized surface plasmons of the coupled elements produce a highly intensified light field in the gap leading to both a reduction of the observation volume and an enhancement of the fluorescence signal of dyes placed within this region. The fluorescence enhancement (FE) produced by an OA is given by the product of two contributions: the electric-field intensity enhancement (at the dye’s excitation wavelength) and the relative change in the dye’s quantum yield induced by the OA (at the dye’s emission wavelength range).15 Additionally, a higher or lower fluorescence signal may be detected in experiments due to alterations of the angular emission directionality produced by the OA.16 Although the FE is a near-field effect,17,18 it can be generally linked to far-field properties of the OA.19,20 Thus, the absorption cross section of the OA is related to quantum yield reduction or quenching, whereas the scattering cross section is related to electric-field enhancement.15 Examples of top-down dimer OAs for FE include triangular and semispherical shaped gold elements.9,21 These OAs, exhibit a FE of more than 3 orders of magnitude, enabling singlemolecule detection at 20 μM in the red and near-infrared range. Recently, we have shown that bottom-up OAs can outperform their lithographic top-down counterparts in terms of FE and single-molecule detection at high concentrations.22 To this end, we employed the DNA-origami technique23 to self-assemble dimer OAs based on 100 nm gold colloidal nanoparticles (NPs). This approach constitutes a major improvement as it enables parallel fabrication and the positioning of the fluorescent dye inside the OA gap with nanometer precision. Nevertheless, as in previous top-down OAs based on gold elements, FE is restricted to the red-infrared spectral range. This shortcoming strongly limits the implementation of OAs for multicolor applications involving the detection of different species. In order to circumvent these limitations, it is desirable to replace gold by other materials that exhibit at the same time both low absorption and high scattering cross section over a broad spectral band in the visible range. Silver NPs, with a surface plasmon resonance in the violet-blue spectral range, appear as ideal candidates.24 However, dimer OAs based on silver NPs were not extensively studied. Most experiments involve rather small- or medium-sized particles,25−29 as colloidal silver NPs are less stable and more challenging to functionalize than gold.30 In this contribution, we self-assemble dimer OAs based on silver NPs onto a DNA origami structure using an improved protocol for functionalizing large metallic NPs.31 We show that strong FE can be obtained in the blue, green, and red spectral range using a single Ag OA in contrast to OAs based on gold NPs.

RESULTS AND DISCUSSION Figure 1a depicts a sketch of the dimer OA employed for broadband FE. It is based on a 3D pillar-shaped DNA origami structure where two 80 nm silver NPs are incorporated with an interparticle gap of approximately 12 nm.31 For synthesis, self-

Figure 1. (a) Sketch of the dimer OA consisting of two 80 nm Ag particles (spherical structures in gray) attached to a DNA origami pillar (tubular-shaped structures in gray) immobilized on a functionalized coverglass (light-blue). The left inset depicts a topview of the dimer OA in which the expected position of the fluorophores is indicated as a red spot. The right inset describes the single dyes positioned at the OA hotspot used in this experiment: Alexa488 (blue), Atto542 (green), and Atto647N (red). Numerical simulations of the absorption and scattering cross sections for 100 nm Au (b) and 80 nm Ag (c) dimer OAs. 4970

DOI: 10.1021/acsnano.7b01621 ACS Nano 2017, 11, 4969−4975

Article

ACS Nano assembled DNA origami nanostructures were immobilized on neutravidin-modified glass coverslips using biotin modification in the base of the DNA nanostructures. From the DNA nanostructures, three single-stranded DNA polyA sequences protrude per NP binding site. The coverslips with the immobilized DNA origami nanostructures were subsequently incubated with NPs, previously functionalized with thiolated T25 sequences. The stable functionalization of large 80 nm Ag NPs was achieved by a combination of improvements: long DNA sequences and a revised protocol.31 In order to minimize the interparticle gap, the zipper hybridization geometry was employed.31 A detailed description of the DNA origami structure22 and the OA preparation protocol can be found in the Supporting Information (SI). Approximately at the center of the gap space, i.e., the OA hotspot, we placed a single fluorescent dye. To test the performance of the OAs throughout the visible range, we employed three different dyes: Alexa488, Atto542, and Atto647N corresponding approximately to the blue, green, and red spectral range, respectively. These dyes exhibit a strong photostability and high quantum yield; however, dyes that operate in a comparable spectral range with a lower quantum yield might lead to a higher FE.22 Furthermore, for a comparison we repeated the measurements with OAs based on 100 nm Au NPs in the same dimer geometry. Both Ag- and Au-based OAs were stable for at least 3 days. This configuration of NP size and gap of Au OAs led to the highest FE reported to date in the red spectral range at the single-molecule level,22 even higher than the one obtained with larger NPs of 150 nm.31 In order to predict the OAs’ spectral response, we performed numerical simulations of the absorption and scattering cross sections. The results for the Au and Ag OAs with an excitation parallel to the dimer orientation are included in Figure 1b,c, respectively. Au OAs show significant absorption from the ultraviolet to the green range, whereas scattering dominates in the yellow to red range. The optical properties for Ag OAs differ significantly from their Au counterparts, showing an absorption peak in the violet range (approximately at 370 nm) and significant scattering from the blue to the red range with a peak in the green range (approximately at 540 nm). These results further confirm that Au OAs can lead to FE in the red-infrared spectral range, as has been extensively reported in the literature.9,21,22,32−34 In addition, these results also predict that dimer Ag OAs could lead to FE spanning from the blue to the red spectral range due to the high scattering cross section and negligible absorption, confirming our strategy for broadband FE. We characterized the OAs using a wide-field epifluorescence microscope using circularly polarized incident light (further details are included in the Materials and Methods section). OAs were immobilized on the surface of a chambered coverglass for imaging. Three samples were prepared containing dimer OAs with the different dyes, Alexa488 (blue), Atto542 (green), and Atto647N (red) at the hotspot. Figure 2a−c depicts exemplary fluorescence images of the three samples from a set of approximately 5000 frames in the red, green, and blue range, respectively. For all samples, spots of different intensities were imaged. For each spot, the corresponding fluorescent transient was extracted from the set of frames. This procedure enabled the identification of single OAs through the detection of blinking events and single-step photobleaching, which constitute characteristic features of single molecules. The fluorescence intensity of each spot was derived from the corresponding fluorescence transients following background

Figure 2. Wide-field fluorescence images obtained for Ag OAs with a single dye, Alexa488 (a), Atto542 (b), and Atto647N (c) positioned at the antenna hotspot. In each image, an exemplary spot is highlighted. Fluorescence transients of the highlighted spots corresponding to a single OA are included. For comparison, exemplary fluorescence transients of DNA origami structures with single dyes and no NPs are included in black. For better visualization, the fluorescence intensity of the single dye without AgNP is plotted with 20-fold magnification.

subtraction. Only spots exhibiting transients with a single photobleaching step were considered. In order to estimate the FE induced by the OAs, we repeated the measurements with DNA origami structures with single dyes and no NPs. From these measurements, reference fluorescence transients were obtained. Exemplary reference transients are included in Figure 2 (in black) for comparison. The mean reference intensity in each spectral range (blue, green, and red) was calculated based on more than 100 transients (further details are included in the SI). Thus, the FE was estimated by normalizing with the mean reference intensity. In addition to dimer OAs, due to our fabrication procedure, samples might also contain a distribution of DNA origami structures with a single NP (monomer OAs) and with no NPs. To disentangle the monomer and the dimer populations, we repeated the experiments with samples containing only one NP binding side, i.e., monomer OAs. The results for the FE obtained from Au and Ag dimer and monomer OAs together with DNA origami structures with no NPs for each spectral range are histogrammed in Figure 3 (see also Figure S1 in the SI). The FE distributions for the monomer and dimer samples show an overall strong enhancement in all cases aside from Au OAs in the blue and green range. For these samples with strong FE, a reduced excitation power was employed (further details are available in the Materials and Methods section) to avoid fast photobleaching of OAs. As a result, this approach hindered the detection of structures with FE below approximately 2-fold, such as, for example, DNA origami structures without NPs. Due to this cutoff and the appearance of extremely high FE factors, the results obtained follow log-normal distributions. Excluding the presence of structures without NPs, the monomer and dimer OAs distributions were fitted with a single and double log-normal distribution, respectively. The separation of the populations is best visible for the case of Atto647N in Figure 3. The first peak of the dimer OAs lies in good agreement with the monomer OAs peak, and therefore it 4971

DOI: 10.1021/acsnano.7b01621 ACS Nano 2017, 11, 4969−4975

Article

ACS Nano

Figure 3. (a) FE histogram plot for 80 nm Ag dimer (black) and monomer (red) OAs with a single dye, Alexa488 (blue range, left), Atto542 (green range, middle), and Atto647N (red range, right). Histograms are plotted logarithmically and fitted with a log-normal distribution (black and red, respectively). The dimer OA is fitted with a double log-normal distribution to account for the underlying monomer population (green curve) in addition to the dimer population (purple). (b) Analogous FE histogram plots and fits for Au OAs. In the blue spectral range both the monomer and dimer samples are fitted with a single log-normal distribution due to the negligible enhancement.

Figure 4. (a) Mean experimental FE with standard error for Ag and Au OAs in the blue, green, and red spectral ranges (squares). The experimentally obtained maximal FE values are also included (crosses). (b) Numerical simulations of the electric-field intensity at the OA’s hotspot for different excitation wavelengths, 487 nm (blue), 532 nm (green), and 644 nm (red). (c) Relative change of the quantum yield for dyes positioned at the OA’s hotspot. For this calculation, the intrinsic quantum yield and peak emission wavelength of Alexa488, Atto542, and Atto647N were employed. The error bars in (b) and (c) take into account the dispersion in NP size and interparticle gap distance.

Table 1 summarizes the main results obtained including the mean FE (μ) with its corresponding standard error (SE) and

is ascribed to the DNA origami population that due to the synthesis procedure contains a single NP instead of forming a dimer structure as intended. The second peak of the dimer OAs distribution, with a higher mean FE, is therefore ascribed to the dimer population. Based on the comparison between the extracted monomer and dimer distributions (peaks in green and violet in Figure 3) for samples with comparable FE, i.e., OAs with a single Atto647N dye, we conclude that both Au and Ag NPs exhibit a similar binding yield. It is worth mentioning that within each population, additional factors are responsible for the heterogeneous FE distribution. These include NP dispersion in size and shape and the orientation of the OA immobilized on the previously functionalized coverglass surface.14 In the case of the FE distributions arising from the Au OAs in the blue spectral range, single log-normal distributions were employed to fit both the monomer and the dimer sample due to the negligible enhancement obtained. Based on this analysis, the mean FE together with its corresponding standard error (SE, for further details see the Supporting Information section statistical analysis) were calculated. These values are included in Figure 4a together with the maximum FE obtained.

Table 1. Mean FE (μ) with Corresponding Standard Error (SE), Maximum FE (max) Together with FE Figure of Merit (FoM) Extracted from the Maximum FE

AgNP

AuNP

μ SE max FoM μ SE max FoM

Alexa488

Atto542

Atto647N

139 17 183 168 1.53 0.01 3 2.4

149 14 207 192 3.07 0.08 17 16

162 5 400 260 176 8 430 280

the maximum FE (max). In the case of Ag-based dimer OAs, FE mean values of 139 ± 17, 149 ± 14, 162 ± 5 together with FE maximum values of 183, 207, and 400 were measured in the blue, green, and red spectral range, respectively. For Au-based OAs mean FE values of 1.53 ± 0.01, 3.07 ± 0.08, 176 ± 8 together with FE maximum values of 3, 17, 430 were measured 4972

DOI: 10.1021/acsnano.7b01621 ACS Nano 2017, 11, 4969−4975

Article

ACS Nano

circularly polarized light in order to minimize the dispersion of the FE distributions. Thus, a factor of approximately two could be achieved by employing incident light linearly polarized along the dimer OA orientation.22 Furthermore, these previous works utilized dyes with a low quantum yield (under 10%) or quenching agents to reduce the dye’s quantum yield. In this way, through the coupling with the OA, the quantum yield was improved, and as a result, an extra enhancement factor of approximately an order of magnitude was obtained. To avoid the strong dependence of the FE with the intrinsic quantum yield of the dye employed, we calculated additionally the FE figure of merit (FoM) as the product of the FE and the dye’s intrinsic quantum yield.37 The results are included in Table 1 and show that single Ag OAs achieve the highest reported FoM in the blue and green range as well as only slightly lower (10%) FoM compared to Au OAs in the red range.

in the blue, green, and red spectral range, respectively. The results clearly demonstrate that Ag-based OAs can yield a broadband FE in the spectral range from the blue to the red as suggested by the cross section simulations in Figure 1c. Furthermore, Au-based OAs only yield significant FE in the red spectral range as also predicted by the cross section numerical simulations in Figure 1b. In addition to the optical response of OAs, we simulated the near-field interaction between OAs and dyes.20,35 These simulations were performed at the excitation wavelengths of 487 nm (blue), 532 nm (green), and 644 nm (red) with an incident polarization along the dimer orientation. Figure 4b includes the electric-field intensity enhancement at the Ag and Au OA hotspot for the blue, green, and red spectral range. The error bars take into account the dispersion in NP size (±5 nm as provided by the manufacturer) and interparticle gap distance (±2 nm).31 The results are consistent with our observations based on the optical properties of OAs. Ag OAs with a significant scattering cross section from the blue to the red spectral range yield a strong electric-field intensity enhancement with factors reaching above 250 along that fraction of the visible spectrum. In contrast, Au OAs lead to strong electricfield intensity enhancement in the red range where they also exhibit strong scattering. Besides the change of the electric-field (excitation), the simulations also yield the relative change of the quantum yield for dyes placed at the OA hotspot and with the transition dipoles aligned along the dimer orientation (Figure 4c). For these calculations the intrinsic quantum yield and the peak emission wavelength of the dyes employed were considered, Alexa488 (0.92 and 519 nm), Atto542 (0.93 and 562 nm), and Atto647N (0.65 and 664 nm). The results are also in good agreement with the far-field calculations in Figure 1b,c. For Ag OAs, the quantum yield at the antenna hotspot remains essentially unaffected for the high quantum yield dyes employed in the blue and green range, whereas for the Atto647N dye, there is even an improvement of approximately 40%. These simulations are consistent with the negligible absorption cross section exhibited by Ag OAs in the blue to red spectral range (Figure 1c). Au OAs display a different effect on the dye’s quantum yield, with both Alexa488 and Atto542 dyes severely quenched and only Atto647N enhanced by 20% in agreement with the considerable absorption cross section of Au OAs in the blue to yellow region. A direct comparison between the experimental FE results (Figure 4a) and the FE simulation arising as the product between the electric-field intensity enhancement (Figure 4b) and the relative change in the quantum yield (Figure 4c) remains challenging. This is mostly due to the fact that circularly polarized light was employed in our measurements and dyes are incorporated to the DNA origami structure through one anchoring group and therefore are free to rotate in a time scale orders of magnitude faster than the integration time.14 Some of the dyes employed, such as Atto647N, carry a net positive charge and can stick without a preferred orientation36 to the DNA origami structure. Nevertheless, numerical simulations are in good agreement with the main features extracted from measurements: Au OAs lead to strong FE in the red range, whereas Ag OAs can strongly enhance fluorescence also in the green and blue ranges. Recent works have shown higher FE values over 3 orders of magnitude.9,21,22,33 It is worth mentioning that in our experiments, OAs are randomly oriented on the coverglass. Therefore, the excitation of the OAs was performed with

CONCLUSIONS We have fabricated OAs by self-assembling large colloidal silver NPs onto three-dimensional DNA origami structures. Single fluorescent dyes were incorporated at the OA hotspot to characterize the antenna performance in terms of FE. Three different dyes, Alexa488, Atto542, and Atto647N, were employed to cover a significant fraction of the visible spectrum from the blue to the red range. The results obtained show that large silver NPs can be functionalized with a protocol developed for large gold NPs with equivalent binding efficiency to DNA origami structures. Furthermore, silver-based OAs lead to a strong broadband FE. In particular, in the blue and green regions, a maximum factor of approximately 200-fold was measured for dyes with an intrinsic quantum yield close to one. In the red region, the maximum enhancement factor is doubled to 400× for a dye with an intrinsic quantum yield of 0.65. For comparison, the broadband performance of gold-based OAs was also studied in the same spectral range. Though Au OAs showed a slightly higher maximum FE in the red range (430×), they lead to significantly lower enhancement in the green (17×) and negligible in the blue (3×). These results were further rationalized through numerical simulations of the farand near-field properties of gold and silver OAs. In addition, the calculation of the FE FoM, a parameter which is independent of the quantum yield of the dye employed, reveals that silver-based OAs yield the highest reported values in the blue (168×) and green (192×) ranges, with a slightly lower factor in the red (260×) as compared to Au OAs (280×). These characteristics turn self-assembled DNA origami-based Ag OAs into promising alternatives to other plasmonic structures for broadband applications such as zero-mode waveguides. In combination with parallel fabrication and ultraprecise placement of single moieties, Ag OAs hold promise for future applications like DNA sequencing or molecular diagnostics where multiplexing is highly required. MATERIALS AND METHODS Silver and gold NPs, 80 and 100 nm in diameter, respectively, were purchased from BBI Solutions and functionalized with single-stranded DNA oligonucleotides 25T containing a thiol modification at the 3′ end (Ella Biotech GmbH). A protocol for the DNA functionalization of NPs is included in the SI. DNA origami pillars were prepared like described elsewhere.22 For the details on DNA origami design, DNA sequences, and folding program see SI. The DNA origami pillars were immobilized on the glass surface of a Lab-Tek chamber (Thermo Fisher Scientific) coated with BSA-biotin/neutravidin (Sigma-Aldrich). 4973

DOI: 10.1021/acsnano.7b01621 ACS Nano 2017, 11, 4969−4975

Article

ACS Nano Present Address

Afterward, immobilized DNA origami pillars were incubated for 48 h at 4 °C with metallic NPs diluted with 1xTE containing 12 mM MgCl2 and 100 mM NaCl to the absorption of 0.1−0.15 on the UV−vis spectrometer (Nanodrop 2000, Thermo Scientific). Finally, the sample was washed with 1xTE containing 12 mM MgCl2 and 100 mM NaCl to remove unbound NPs, and single-molecule fluorescence measurements were performed. Imaging and Analysis. Fluorescence imaging was performed on a custom-built wide-field epifluorescence microscope, based on an Olympus IX-71 inverted microscope. For excitation a 644 nm diode laser (iBeam smart, Toptica photonics, cleanup filter Brightline HC 650/13, Semrock), a 532 nm fiber laser (MPB Communications, cleanup filter Z532/647x, Chroma) and a 487 nm diode laser (iBeam smart, Toptica photonics, cleanup filter 488/1.9, AHF Analysentechnik) were utilized, respectively. Circular polarization for all three colors was achieved using a linear polarizer followed by a quarter-wave plate. Red and green excitation beams were coupled into the microscope with a dual-color-beamsplitter (Dual Line zt532/640 rpc, AHF Analysentechnik), while blue excitation was coupled in with a dichroic beamsplitter (Di01-R488, Semrock). The excitation beam was focused on the backfocal plane of an oil-immersion objective (100×, NA 1.4, UPlanSApo, Olympus). For significant reduction of sample and setup drift, an actively stabilized optical table (TS-300, JRS Scientific Instruments) and nosepiece stage (IX2-NPS, Olympus) were used. The fluorescence signal was spectrally filtered by emission filters (ET 700/75, Chroma; BrightLine 582/75, AHF Analysentechnik or BrightLine 531/40, Semrock) and imaged on an EMCCD camera (Ixon X3 DU-897/Andor) with a pixel size of 100 nm. The filters and beamsplitters were changed between sequential acquisitions for the different excitations. The fluorescence signals were acquired for 5000 frames (10,000 frames for green excitation) at 100 ms integration time and an electron multiplying gain of 5 (10 for blue excitation). The excitation intensity was set that the fluorophores were bleached in this period (not before or after). Therefore, excitation intensities were reduced for samples showing strong enhancement (20× for red, 3× for green, and 5× for blue). Only for Au green and blue, the same excitation intensities were applied for all three cases: reference, monomer, and dimer. The analysis was carried out with customwritten LabVIEW software. Molecules were selected by single-step photobleaching transients to the background intensity (surrounding pixel). Fluorescence intensities were background corrected by the bleached intensity level. Enhancement values were calculated due to reference measurements. See SI for further data processing. Simulations. All simulations were realized using a commercial FDFD software (CST). A medium with a permittivity of 1.77 was employed as background to match the experimental buffer conditions. The optical properties of silver and gold were extracted from ref 38. For the quantum yield simulations, dyes were modeled by a current source oscillating at a frequency corresponding to the wavelength of maximum emission as previously reported.36

∥

The Institute of Photonic Sciences (ICFO), The Barcelona Institute of Science and Technology, 08860 Castelldefels (Barcelona), Spain Author Contributions ‡

These authors contributed equally.

Notes

The authors declare no competing financial interest.

ACKNOWLEDGMENTS This work was supported by a starting grant (SiMBA, EU 261162) of the European Research Council (ERC) and the Deutsche Forschungsgesellschaft (AC 279/2-1 and TI 329/91). I.K. acknowledges support by the Mobility Plus grant 1269/ MOB/IV/2015/0 from the Polish Ministry of Science and Higher Education (MNiSW). C.V. is grateful for a scholarship of the Studienstiftung des deutschen Volkes. We acknowledge funding of the state ministry for research of lower saxony in the frame of the “Quantum- and Nanometrology” (QUANOMET) strategic research area. Quanomet is part of the LUH-TUBS research alliance. REFERENCES (1) Weiss, S. Fluorescence Spectroscopy of Single Biomolecules. Science 1999, 283, 1676−1683. (2) Eid, J.; Fehr, A.; Gray, J.; Luong, K.; Lyle, J.; Otto, G.; Peluso, P.; Rank, D.; Baybayan, P.; Bettman, B.; Bibillo, A.; Bjornson, K.; Chaudhuri, B.; Christians, F.; Cicero, R.; Clark, S.; Dalal, R.; deWinter, A.; Dixon, J.; Foquet, M.; et al. Real-Time DNA Sequencing from Single Polymerase Molecules. Science 2009, 323, 133−138. (3) Hesse, J.; Jacak, J.; Kasper, M.; Regl, G.; Eichberger, T.; Winklmayr, M.; Aberger, F.; Sonnleitner, M.; Schlapak, R.; Howorka, S.; Muresan, L.; Frischauf, A.-M.; Schütz, G. J. RNA Expression Profiling at the Single Molecule Level. Genome Res. 2006, 16, 1041− 1045. (4) Mayr, R.; Haider, M.; Thünauer, R.; Haselgrübler, T.; Schütz, G. J.; Sonnleitner, A.; Hesse, J. A Microfluidic Platform for Transcriptionand Amplification-Free Detection of Zepto-Mole Amounts of Nucleic Acid Molecules. Biosens. Bioelectron. 2016, 78, 1−6. (5) Moran-Mirabal, J. M.; Craighead, H. G. Zero-Mode Waveguides: Sub-Wavelength Nanostructures for Single Molecule Studies at High Concentrations. Methods 2008, 46, 11−17. (6) Holzmeister, P.; Acuna, G. P.; Grohmann, D.; Tinnefeld, P. Breaking the Concentration Limit of Optical Single-Molecule Detection. Chem. Soc. Rev. 2014, 43, 1014−1028. (7) Acuna, G.; Grohmann, D.; Tinnefeld, P. Enhancing SingleMolecule Fluorescence with Nanophotonics. FEBS Lett. 2014, 588, 3547−3552. (8) Di Fabrizio, E.; Schlücker, S.; Wenger, J.; Regmi, R.; Rigneault, H.; Calafiore, G.; et al. Roadmap on Biosensing and Photonics with Advanced Nano-Optical Methods. J. Opt. 2016, 18, 063003. (9) Punj, D.; Mivelle, M.; Moparthi, S. B.; van Zanten, T. S.; Rigneault, H.; van Hulst, N. F.; Garcia-Parajo, M. F.; Wenger, J. A Plasmonic A ̀ ntenna-in-Box’ Platform for Enhanced Single-Molecule Analysis at Micromolar Concentrations. Nat. Nanotechnol. 2013, 8, 512−516. (10) Uemura, S.; Aitken, C. E.; Korlach, J.; Flusberg, B. A.; Turner, S. W.; Puglisi, J. D. Real-Time tRNA Transit on Single Translating Ribosomes at Codon Resolution. Nature 2010, 464, 1012−1017. (11) Khatua, S.; Paulo, P. M. R.; Yuan, H.; Gupta, A.; Zijlstra, P.; Orrit, M. Resonant Plasmonic Enhancement of Single-Molecule Fluorescence by Individual Gold Nanorods. ACS Nano 2014, 8, 4440−4449. (12) Hoang, T. B.; Akselrod, G. M.; Argyropoulos, C.; Huang, J.; Smith, D. R.; Mikkelsen, M. H. Ultrafast Spontaneous Emission Source Using Plasmonic Nanoantennas. Nat. Commun. 2015, 6, 7788.

ASSOCIATED CONTENT S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsnano.7b01621. Details on raw fluorescence enhancement data, a detailed description of nanoparticle functionalization, statistical analysis including a table with all values from fits and measurements and detailed information on DNA origami design (PDF)

AUTHOR INFORMATION Corresponding Authors

*E-mail: [email protected]. *E-mail: [email protected]. ORCID

Guillermo P. Acuna: 0000-0001-8066-2677 4974

DOI: 10.1021/acsnano.7b01621 ACS Nano 2017, 11, 4969−4975

Article

ACS Nano

(34) Flauraud, V.; Regmi, R.; Winkler, P. M.; Alexander, D. T. L.; Rigneault, H.; van Hulst, N. F.; García-Parajo, M. F.; Wenger, J.; Brugger, J. In-Plane Plasmonic Antenna Arrays with Surface Nanogaps for Giant Fluorescence Enhancement. Nano Lett. 2017, 17, 1703− 1710. (35) Taminiau, T. H.; Stefani, F. D.; van Hulst, N. F. Single Emitters Coupled to Plasmonic Nano-Antennas: Angular Emission and Collection Efficiency. New J. Phys. 2008, 10, 105005. (36) Acuna, G. P.; Bucher, M.; Stein, I. H.; Steinhauer, C.; Kuzyk, A.; Holzmeister, P.; Schreiber, R.; Moroz, A.; Stefani, F. D.; Liedl, T.; Simmel, F. C.; Tinnefeld, P. Distance Dependence of SingleFluorophore Quenching by Gold Nanoparticles Studied on DNA Origami. ACS Nano 2012, 6, 3189−3195. (37) Gill, R.; Tian, L.; Somerville, W. R. C.; Le Ru, E. C.; van Amerongen, H.; Subramaniam, V. Silver Nanoparticle Aggregates as Highly Efficient Plasmonic Antennas for Fluorescence Enhancement. J. Phys. Chem. C 2012, 116, 16687−16693. (38) Palik, E. D. Handbook of Optical Constants of Solids; Academic Press: Orlando, 1985.

(13) Zhang, T.; Gao, N.; Li, S.; Lang, M. J.; Xu, Q.-H. Single-Particle Spectroscopic Study on Fluorescence Enhancement by Plasmon Coupled Gold Nanorod Dimers Assembled on DNA Origami. J. Phys. Chem. Lett. 2015, 6, 2043−2049. (14) Acuna, G. P.; Moller, F. M.; Holzmeister, P.; Beater, S.; Lalkens, B.; Tinnefeld, P. Fluorescence Enhancement at Docking Sites of DNADirected Self-Assembled Nanoantennas. Science 2012, 338, 506−510. (15) Anger, P.; Bharadwaj, P.; Novotny, L. Enhancement and Quenching of Single-Molecule Fluorescence. Phys. Rev. Lett. 2006, 96, 113002. (16) Taminiau, T. H.; Stefani, F. D.; Segerink, F. B.; van Hulst, N. F. Optical Antennas Direct Single-Molecule Emission. Nat. Photonics 2008, 2, 234−237. (17) Darvill, D.; Centeno, A.; Xie, F. Plasmonic Fluorescence Enhancement by Metal Nanostructures: Shaping the Future of Bionanotechnology. Phys. Chem. Chem. Phys. 2013, 15, 15709−15726. (18) Centeno, A.; Xie, F.; Alford, N. Predicting the Fluorescent Enhancement Rate by Gold and Silver Nanospheres Using FiniteDifference Time-Domain Analysis. IET Nanobiotechnol. 2013, 7, 50− 58. (19) Kuhn, S.; Hakanson, U.; Rogobete, L.; Sandoghdar, V. Enhancement of Single-Molecule Fluorescence Using a Gold Nanoparticle as an Optical Nanoantenna. Phys. Rev. Lett. 2006, 97, 017402. (20) Bharadwaj, P.; Novotny, L. Spectral Dependence of Single Molecule Fluorescence Enhancement. Opt. Express 2007, 15, 14266− 14274. (21) Kinkhabwala, A.; Yu, Z.; Fan, S.; Avlasevich, Y.; Mullen, K.; Moerner, W. E. Large Single-Molecule Fluorescence Enhancements Produced by a Bowtie Nanoantenna. Nat. Photonics 2009, 3, 654−657. (22) Puchkova, A.; Vietz, C.; Pibiri, E.; Wunsch, B.; Sanz Paz, M.; Acuna, G. P.; Tinnefeld, P. DNA Origami Nanoantennas with over 5000-Fold Fluorescence Enhancement and Single-Molecule Detection at 25 μM. Nano Lett. 2015, 15, 8354−8359. (23) Rothemund, P. W. Folding DNA to Create Nanoscale Shapes and Patterns. Nature 2006, 440, 297−302. (24) Coronado, E. A.; Encina, E. R.; Stefani, F. D. Optical Properties of Metallic Nanoparticles: Manipulating Light, Heat and Forces at the Nanoscale. Nanoscale 2011, 3, 4042−4059. (25) Zhang, J.; Fu, Y.; Chowdhury, M. H.; Lakowicz, J. R. MetalEnhanced Single-Molecule Fluorescence on Silver Particle Monomer and Dimer: Coupling Effect between Metal Particles. Nano Lett. 2007, 7, 2101−2107. (26) Weller, L.; Thacker, V. V.; Herrmann, L. O.; Hemmig, E. A.; Lombardi, A.; Keyser, U. F.; Baumberg, J. J. Gap-Dependent Coupling of Ag−Au Nanoparticle Heterodimers Using DNA Origami-Based Self-Assembly. ACS Photonics 2016, 3, 1589−1595. (27) Pal, S.; Deng, Z.; Ding, B.; Yan, H.; Liu, Y. DNA-OrigamiDirected Self-Assembly of Discrete Silver-Nanoparticle Architectures. Angew. Chem., Int. Ed. 2010, 49, 2700−2704. (28) Lee, J.-S.; Lytton-Jean, A. K. R.; Hurst, S. J.; Mirkin, C. A. Silver Nanoparticle−Oligonucleotide Conjugates Based on DNA with Triple Cyclic Disulfide Moieties. Nano Lett. 2007, 7, 2112−2115. (29) Sonnichsen, C.; Reinhard, B. M.; Liphardt, J.; Alivisatos, A. P. A Molecular Ruler Based on Plasmon Coupling of Single Gold and Silver Nanoparticles. Nat. Biotechnol. 2005, 23, 741−745. (30) Zhang, X.; Servos, M. R.; Liu, J. Fast pH-Assisted Functionalization of Silver Nanoparticles with Monothiolated DNA. Chem. Commun. (Cambridge, U. K.) 2012, 48, 10114−10116. (31) Vietz, C.; Lalkens, B.; Acuna, G. P.; Tinnefeld, P. Functionalizing Large Nanoparticles for Small Gaps in Dimer Nanoantennas. New J. Phys. 2016, 18, 045012. (32) Lakowicz, J. R. Radiative Decay Engineering 5: Metal-Enhanced Fluorescence and Plasmon Emission. Anal. Biochem. 2005, 337, 171− 194. (33) Punj, D.; Regmi, R.; Devilez, A.; Plauchu, R.; Moparthi, S. B.; Stout, B.; Bonod, N.; Rigneault, H.; Wenger, J. Self-Assembled Nanoparticle Dimer Antennas for Plasmonic-Enhanced SingleMolecule Fluorescence Detection at Micromolar Concentrations. ACS Photonics 2015, 2, 1099−1107. 4975

DOI: 10.1021/acsnano.7b01621 ACS Nano 2017, 11, 4969−4975