Enhanced-Fluorescence of a Dye on DNA-Assembled Gold Nano

iBB – Institute for Bioengineering and Biosciences, Instituto Superior Técnico,. 8. Universidade de Lisboa, Av. Rovisco Pais 1, 1049-001 Lisboa, Po...
0 downloads 0 Views 2MB Size
Subscriber access provided by Kaohsiung Medical University

C: Plasmonics, Optical Materials, and Hard Matter

Enhanced-Fluorescence of a Dye on DNA-Assembled Gold NanoDimers Discriminated by Lifetime Correlation Spectroscopy Pedro M. R. Paulo, David Botequim, Agnieszka Jóskowiak, Sofia A.M. Martins, Duarte Miguel de França Teixeira dos Prazeres, Peter Zijlstra, and Silvia Brito Costa J. Phys. Chem. C, Just Accepted Manuscript • DOI: 10.1021/acs.jpcc.7b12622 • Publication Date (Web): 30 Apr 2018 Downloaded from http://pubs.acs.org on May 3, 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 39 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 Journal of Physical Chemistry

1

Enhanced-Fluorescence of a Dye on DNA-

2

Assembled Gold Nano-Dimers Discriminated

3

by Lifetime Correlation Spectroscopy

4

Pedro M. R. Paulo,1,* David Botequim,1 Agnieszka Jóskowiak,1 Sofia Martins,2 Duarte M. F. Prazeres,2

5

Peter Zijlstra 3 and Sílvia M. B. Costa 1 1

6 7 8 9 10 11

Centro de Química Estrutural, Instituto Superior Técnico,

Universidade de Lisboa, Av. Rovisco Pais 1, 1049-001 Lisboa, Portugal 2

iBB – Institute for Bioengineering and Biosciences, Instituto Superior Técnico, Universidade de Lisboa, Av. Rovisco Pais 1, 1049-001 Lisboa, Portugal 3

Molecular Biosensing for Medical Diagnostics,

Eindhoven University of Technology, P.O. Box 513, 5600 MB, Eindhoven, The Netherlands

12 13 14 15 16 17 18

ACS Paragon Plus Environment

1

The Journal of Physical Chemistry 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 39

1

Abstract

2

The surface plasmon modes of metal nanoparticles provide a way to efficiently enhance the

3

excitation and emission from a fluorescent dye. We have employed DNA-directed assembly to

4

prepare dimers of gold nanoparticles and used their longitudinally coupled plasmon mode to

5

enhance the fluorescence emission of an organic red-emitting dye, Atto-655. The plasmon-

6

enhanced fluorescence of this dye using dimers of 80 nm particles was measured at single

7

molecule detection level. The top enhancement factors were above 1000-fold in 71% of the

8

dimers within a total of 32 dimers measured, and, in some cases, they reached almost 4000-fold,

9

in good agreement with model simulations. Additionally, fluorescence lifetime correlation

10

analysis enabled the separation of enhanced from non-enhanced emission simultaneously

11

collected in our confocal detection volume. This approach allowed us to recover a short

12

relaxation component exclusive to enhanced emission that is attributed to the interaction of the

13

dye with DNA in the interparticle gaps. Indeed, the frequency of enhancement events is larger

14

than expected from the volume occupancy of the gap region, thus suggesting that interaction of

15

the dye with DNA linkers favors the observation of emission enhancement in our dimer particles.

16 17 18 19 20

ACS Paragon Plus Environment

2

Page 3 of 39 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 Journal of Physical Chemistry

1

Introduction

2

In recent years an increasing number of reports have illustrated the use of plasmonic

3

nanostructures as optical nanoantennas to modify the emission properties of fluorescent

4

molecules or other emitters.1-3 The phenomena described encompass emission enhancement and

5

quenching,4-7 tailoring of the emission spectrum or lifetimes,8-11 and controlling emission

6

directionality.12,13 The role of plasmonic antennas for modifying other photophysical or

7

photochemical processes has also been reported. For instance, plasmonic nanostructures have

8

been employed for the manipulation of optical selection rules,14 spatial confinement in

9

photopolymerization,15 enhancement of Förster energy transfer efficiency,16 reduction of dye

10

photobleaching17 or improvement of energy conversion in solar cells.18

11

The strong enhancement of fluorescence emission is nevertheless one of the most

12

promising features of plasmonic antennas. The enhancement of very weak emitters is potentially

13

interesting for studying intrinsically fluorescent systems with low quantum yields, e.g. proteins

14

or metal complexes, because it makes possible their detection by optical techniques without the

15

need of labelling. On the other hand, the emission enhancement of strongly fluorescent

16

molecules by their conjugation with metal nanoparticles can result in even brighter fluorescent

17

objects of interest for imaging and biosensing applications. Other examples that benefit from

18

plasmonic light confinement and emission enhancement are the detection of single molecules in

19

concentrated solutions19-21 and super-resolution localization fluorescence microscopy.22-24

20

The design principles to achieve emission enhancement with plasmonic antennas have

21

been extensively studied, both theoretical and experimentally, and are well established in the

22

literature.1-5,25-27 The emission enhancement that is possible with a single spherical gold

23

nanoparticle is limited to only of a few times the photon emission rate of the isolated fluorescent

ACS Paragon Plus Environment

3

The Journal of Physical Chemistry 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 39

1

molecule.4,5 In order to obtain large fluorescence enhancements, it is necessary to employ

2

plasmonic nanostructures that generate intense near fields. This is the case of metal nanoparticles

3

with sharp geometrical features and of nanostructure arrays or assemblies of particles with

4

nanometric gaps.6,28-35 For instance, dimers of spherical gold nanoparticles are capable of

5

enhancing fluorescence emission by two or more orders of magnitude.21,30-34 The longitudinal

6

coupling of the individual plasmons in dimer particles gives rise to an hybridized plasmon mode

7

that produces an intense near field concentrated in the gap region. The field enhancement is

8

particularly strong for dimers of gold particles of several tens of nanometer in diameter

9

combined with gap separations of a few nanometers. Such dimer particles afford the largest

10

emission enhancements reported up to now.21,33,34

11

The fabrication of dimer nanostructures can be accomplished by nanolithography, but

12

precise control of gap distances down to only a few nanometers is still challenging using these

13

methods. Alternatively, wet-chemistry synthesis and supramolecular assembly approaches give

14

access to dimers of gold nanoparticles with narrow interparticle distances. One successful

15

approach is the use of DNA-directed self-assembly to produce dimer nanoparticles that perform

16

efficiently as plasmonic antennas.10,30-33 Nevertheless, these methods generally require quite

17

elaborate molecular arrangements of DNA (e.g. singly DNA-functionalized particles or DNA

18

origami templates) that make the full process expensive and complex to replicate. On the other

19

hand, the spontaneous aggregation of gold nanoparticles upon surface immobilization has also

20

been shown to produce dimers that can be used as plasmonic antennas.21 Although this is a very

21

simple way to produce dimers, the process is intrinsically not selective for dimers, which then

22

appear on the surface mixed with single particles and large aggregates.

ACS Paragon Plus Environment

4

Page 5 of 39 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 Journal of Physical Chemistry

1

In this contribution, we report on fluorescence enhancement of a red-emitting dye using

2

gold nanoparticle dimers to achieve large enhancement factors. We have adapted a DNA-

3

directed approach to assemble gold nanoparticles into dimers,36 which is simple and does not

4

rely on exactly one ds-DNA in the gap, while being selective for dimer assembly contrasting to

5

spontaneous parrticle aggregation. The performance of these dimer particles as plasmonic

6

antennas for emission enhancement of Atto-655 fluorescent dye was investigated by single

7

molecule fluorescence microscopy. The dimers of gold particles having a diameter of 80 nm

8

afforded top emission enhancements of almost 4000-fold, which are comparable to the largest

9

enhancement factors reported so far.21,33 We have also introduced a fluorescence lifetime

10

correlation (FLCS) analysis in order to separate the dimer enhanced from non-enhanced emission

11

of dye molecules within the confocal detection volume. This analysis method can be readily

12

applied to fluorescence enhancement experiments performed with pulsed excitation and photon

13

detection by time-correlated single photon counting, which is a widespread technique of time-

14

resolved fluorescence. However, to the best of our knowledge, this approach has only been

15

reported once for the purpose of distinguishing enhanced and non-enhanced emission.37 The

16

accelerated decay time associated with plasmon-enhanced fluorescence allows to clearly define a

17

distinct fluorescence decay pattern that can be used in FLCS analysis to filter the correlation

18

funtion of enhanced (and non-enhanced) emission. Other approaches for discriminating between

19

enhanced and non-enhanced emission typically involve at least two polarization-resolved

20

measurements,21,28,32 while FLCS approach can be performed from a single measurement.

21 22

ACS Paragon Plus Environment

5

The Journal of Physical Chemistry 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 6 of 39

1

Experimental

2

Materials

3

Gold nanoparticles with sizes of 80 nm and 40 nm, stabilized by a citrate coating, were acquired

4

from Nanopartz Inc. as aqueous suspensions with an optical density of 1 (product no. A11-80

5

and A11-40). DNA oligonucleotides purified by HPLC were purchased from STAB Vida (Monte

6

da Caparica, Portugal). DNA strands with (n + 10) nucleotides and having the general sequence

7

5’-SH-A10Nn-3’ were used. Two oligonucleotides were designed for “n” equal to 15, 30 or 60, in

8

such a way as to form hybrids by base pairing of the N15, N30 and N60 sequences (see Scheme S1

9

in the SI). Tris base (Eurobio, molecular biology grade), boric acid (Fisher Chemical, assay

10

100%) and EDTA (TCI Europe, 98%) were used to prepare TBE buffer. Sodium citrate tribasic

11

dihydrate (Sigma-Aldrich, ≥ 99.5%) and hydrochloric acid (Sigma-Aldrich, 37%) were used to

12

prepare citrate buffer. Ultrapure water was obtained with a Milli-Q purification system (Merck-

13

Millipore). Agarose with electrophoresis purity degree was purchased from NZYTech.

14

Polyethylenimine, branched polymer with an average Mw ~25,000 was purchased from Aldrich.

15

Equipment

16

Extinction spectra were measured in an UV/Vis spectrophotometer from PerkinElmer, model

17

Lambda 35. TEM characterization was performed on a HITACHI H-8100 electron microscope

18

operating at 200 kV. Glass surfaces were cleaned using an UV/Ozone chamber model PSD-UV3

19

from Novascan. Single-molecule fluorescence experiments were performed on a confocal

20

fluorescence lifetime microscope, MicroTime 200 (PicoQuant GmbH). The microscope setup

21

details were previously described.38 The single-particle spectra were obtained by means of a low

22

light level spectrometer, QEPro (Ocean Optics), that was fiber coupled to the confocal

ACS Paragon Plus Environment

6

Page 7 of 39 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 Journal of Physical Chemistry

1

microscope. Data acquisition and preliminary analysis were performed on SymPhoTime

2

software (PicoQuant GmbH). This program was also used to carry out data analysis by

3

Fluorescence Lifetime Correlation Spectroscopy (FLCS).39,40

4

Preparation of Gold Nanoparticle Dimers

5

The first step was the hybridization of the thiolated DNA oligonucleotides to form a DNA linker

6

with two thiol moieties, one at each end of the double strand (see Scheme S1 in the SI). Then, the

7

double-stranded DNA linkers were used to assemble gold nanoparticles into dimers and larger

8

aggregates in a similar approach to that described in Ref. 36. The concentration of DNA used

9

was always more than a 2-fold excess relatively to the theoretical maximum number of thiolated

10

oligos per particle, i.e. approx. 1400 and 430 chains for gold particles with 80 and 40 nm,

11

respectively.41 The solution of pre-hybridized DNA linkers was mixed with citrate-stabilized

12

gold nanoparticles following the low pH protocol of Zhang et al.42 The detailed description is

13

given in the SI. The large excess of DNA linkers and the low pH protocol were used to obtain an

14

extensive surface funcionalization of gold nanoparticles with oligonucleotides.

15

Gel electrophoresis was used to separate single particles from dimers and higher aggregates. The

16

concentrations of agarose gels used were 0.7% and 1.5% for particles sized 80 and 40 nm,

17

respectively. All gels were run at 120 V in 0.5× TBE buffer. Two red bands were typically

18

obtained after 30 min. These particle-containing agarose fractions were cut from the gel and

19

placed in a microtube filled with approx. 0.5 mL of 0.5× TBE buffer. The particles were allowed

20

to diffuse from the gel pieces for at least one week. Samples containing the extracted gold

21

nanoparticles were characterized by TEM, showing that the fraction extracted from the slowest

22

migrating band contained mostly dimers.

ACS Paragon Plus Environment

7

The Journal of Physical Chemistry 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 8 of 39

1

Surface Immobilization of Gold Nanoparticle Dimers

2

Round glass coverslips were cleaned by sonication in aqueous solution of RBS50 detergent 5%

3

(v/v) during 30 mins, followed by another sonication in absolute ethanol, and rinsing with water

4

after each step. The dried coverslips were then irradiated for 2 hrs in a UV/ozone chamber.

5

Before particle immobilization, the glass surface was coated with polyethyleneimine.43 First, the

6

coverslips were dipped in a cold piranha bath for 10 mins to improve the surface wettability, then

7

rinsed copiously with water and dried. The polymer coating was deposited by covering the

8

surface with an aqueous solution of polyethyleneimine 0.2% (w/v) during 20 mins, then rinsing

9

with water and nitrogen blow drying. The gold nanoparticle dimers were immobilized by spin

10

coating 10 µL from a dilute (sub-nM) aqueous solution of these particles.

11

Fluorescence Enhancement of Atto-655 Dye

12

The glass coverslips with immobilized dimers were first observed in the confocal microscope

13

under laser excitation at 482 nm using a power of approximately 50 kW/cm2 to scan an area of

14

20×20 µm2 (256 pixels per side) with an integration time of 0.6 ms/pixel. The surface

15

immobilization conditions were previously optimized to obtain microscope images showing well

16

dispersed difraction-limited spots. The surface-immobilized dimer particles were immersed in

17

ultrapure water during these measurements. The photoluminescence spectrum was collected for

18

each dimer particle within a total of 10 to 15 particles selected per image. The longitudinal

19

surface plasmon band was used to infer the gap separation using a plasmon ruler relation

20

determined from our simulation results. The laser excitation was changed to 639 nm and the

21

same area was scanned using an excitation power of approximately 4 kW/cm2. The diffraction-

22

limited spots in the new image were matched to the previously identified dimer particles. The

ACS Paragon Plus Environment

8

Page 9 of 39 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 Journal of Physical Chemistry

1

emission intensity time trace of each dimer particle was measured at 639 nm for an excitation

2

power of approximately 0.4 kW/cm2. This power was the same as later used for measuring

3

fluorescence enhancement of Atto-655 dye. At this stage the dimer particles were immersed in

4

ultrapure water, so the purpose of measuring intensity time traces was to characterize the

5

background signal from the dimer’s photoluminescence emission and to confirm the absence of

6

fluorescence bursts later attributed to enhanced emission from dye molecules interacting with

7

dimer particles. Then, the water drop covering the dimer particles was replaced with a diluted

8

aqueous solution of Atto-655 dye with a concentration of a few nanomolar. The exact

9

concentration of Atto-655 dye was determined for each fluorescence enhancement experiment by

10

performing a FCS measurement at a depth of 10 µm inside the dye solution. This afforded values

11

of dye concentration in the range of 0.7 to 4.8 nM. The emission intensity time trace of each

12

dimer particle was re-measured, at a power of 0.4 kW/cm2, now in the presence of Atto-655 dye

13

in aqueous solution. During the time acquisition of 60 s, it was possible to observe, in most

14

cases, from tens to hundreds of strong fluorescence emission events. The fluorescence

15

enhancement effect were evaluated from the comparison with the non-enhanced average

16

emission of Atto-655 dye, which was determined from a FCS measurement at the same

17

excitation power of 0.4 kW/cm2, similarly to other studies reported in the literature.19,28,32

18

DDA simulations

19

Simulations of discrete dipole approximation (DDA) were performed in order to theoretically

20

estimate the emission enhancement using dimers of gold nanoparticles. The free software

21

implementation A-DDA was used for running these simulations.Error! Reference source not found. The

22

near-field intensity maps around the gold nanoparticles were calculated using the subroutine

23

implemented in the A-DDA software package.45 The estimation of fluorescence enhancement

ACS Paragon Plus Environment

9

The Journal of Physical Chemistry 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 10 of 39

1

factors requires the calculation of the enhanced radiative and non-radiative decay rates. The

2

theoretical formalism described by D’Agostino et al.46 was used for this purpose (see SI).

3 4

Results and Discussion

5

The dimers of gold nanoparticles used in fluorescence enhancement experiments were produced

6

by bridging two nanoparticles with a DNA-based linker formed by hybridizing two thiolated

7

oligonucleotides. These short linkers contain a central, double-stranded region of less than 60

8

base pairs and overhanging single strand segments with 10 nucleotides. These short linkers

9

yielded dimers with interparticle gaps of a few nanometers that can be observed in the TEM

10

images (Fig. 1A, B). The gap distances are shorter than the length of the double-stranded region

11

because of the single strand segments on both ends of the DNA linker. The single strand

12

segments confer flexibility to the DNA linker and allow for the rigid double strand segment to be

13

positioned at a tilted angle within the gap, which would explain why the interparticle separation

14

is shorter than the length of the double-stranded region. Another fundamental difference

15

regarding our dimerization protocol relatively to other described in the literature, is that the DNA

16

linkers are previously hybridized and only then added to the citrate-stabilized nanoparticles. In

17

this way, it is possible that initially one (or few) DNA linkers connect two particles at a tilted

18

angle, before the particles get extensively coated, thereby, locking them at a separation distance

19

smaller than the fully extended size of the linker.

20

The fraction of dimer particles obtained after purification by gel electrophoresis, as

21

observed from TEM images, was on average 68%, while single particles represent 23% of

22

analysed samples. Single particles are present as smear in the gel and end up as a contamination

ACS Paragon Plus Environment

10

Page 11 of 39 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 Journal of Physical Chemistry

1

in the dimer sample when extracted from the respective gel band (inset of Fig. 1C). Gold

2

nanoparticles that are only citrate-stabilized show a gel without any bands, only a smear is

3

visible, which emphasizes the role of thiolated DNA linkers in particle assembly (Fig. S1 of the

4

SI).

5

The optical extinction spectrum of dimer particles features two surface plasmon bands

6

instead of the single band that is characteristic of single gold nanoparticles (Fig. 1C). This occurs

7

because of plasmon coupling between the individual modes in neighbouring particles of the

8

dimer gives rise to two optically active hybridized plasmon modes.47 The surface plasmon band

9

appearing at longer wavelengths corresponds to the bonding mode, and it generates strong local

10

fields across the interparticle gap that are interesting for fluorescence enhancement. The

11

resonance wavelength of the longitudinal plasmon mode depends strongly on the gap separation.

12

For instance, at shorter gap distances the plasmon coupling is stronger and this energy stabilizing

13

interaction induces a shift of the longitudinal surface plasmon band toward longer wavelengths.

14

This effect is illustrated in Figure 1D that shows the calculated extinction spectra for dimers of

15

gold particles of 80 nm simulated with gap distances of 2, 3 and 4 nm – red, orange and green

16

curves, respectively. Using this plasmon ruler obtained from simulated spectra (inset of Fig. 1D),

17

we estimate from the experimental spectrum of our dimer sample that the average interparticle

18

distance is around 2 to 3 nm, which matches well with our results from TEM (Table S1 of the

19

SI).

20

ACS Paragon Plus Environment

11

The Journal of Physical Chemistry 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 12 of 39

1 2

Figure 1. A) Electron microscopy images of dimers of gold nanoparticles with a diameter of 80

3

nm assembled by a DNA linker with a 60-bp double-stranded region. B) Magnification of TEM

4

images of four selected examples of dimer nanoparticles. C) Extinction spectrum of dimers of

5

gold nanoparticles (red curve) extracted from the gel band no. 2; and that of gold nanoparticles

6

functionalized with DNA, but not-dimerized (blue curve), separated from gel band no. 1– the

7

inset shows a photograph of the gel with bands indicated. D) Calculated extinction spectra from

8

DDA simulations of a single gold nanoparticle of 80 nm (blue curve) and that of dimer

9

nanoparticles with a gap separation of 2, 3 and 4 nm (red, orange and green curves, respectively)

10

– the inset shows the peak wavelength of the longitudinal surface plasmon band (LSP), as a

11

function of the gap separation considered in DDA simulations.

12 13

The dimer particles were further characterized by their optical spectrum measured at the

14

single-particle level. For this purpose, dimer samples were immobilized onto glass coverslips at

ACS Paragon Plus Environment

12

Page 13 of 39 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 Journal of Physical Chemistry

1

low surface density. The optical microscopy images show diffraction-limited spots with a narrow

2

dispersion of emission intensity (Fig. 2A and B). The photoluminescence spectrum was collected

3

from several of these spots on each image. The spectrum of dimer particles typically features two

4

bands: one more intense at around 550 nm accompanied by another band or shoulder at longer

5

wavelengths (Fig. 2C). It is clearly different from the spectrum of single particles that show only

6

one band at 550 nm with a Lorentzian lineshape – as exemplified in spectrum no. 6 of Fig. 2C.

7

Also, the emission intensity from dimer particles is approximately twice that of single particles

8

and, in addition, the emission from single particles is non-polarized, while the dimers emission

9

displays polarization in the sample plane. These features allow to easily distinguish the emission

10

of individual dimers from that of a minor fraction of single particles that is always present even

11

after purification by gel electrophoresis.

12

The emission spectrum of dimer particles changes slightly from particle to particle within

13

each sample. In fact, the flexible ends of the DNA oligonucleotide linkers allow for some

14

variation in the gap distance between paired particles, which strongly affects the longitudinal

15

surface plasmon band of the dimer spectrum. This variation in the gap distance was firstly

16

noticed in the TEM images of dimer samples (Fig. S3 of the SI). Another contribution to

17

heterogeneity in the emission spectrum of individual dimers arises from deviations from the

18

spherical shape of the gold nanoparticles. Nevertheless, we have used the plasmon ruler shown in

19

the inset of Figure 1D to infer about the approximate interparticle gap distance in individual

20

dimer particles from their single particle spectra.

21

ACS Paragon Plus Environment

13

The Journal of Physical Chemistry 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 14 of 39

1 2

Figure 2. A) Optical microscopy image of dimer nanoparticles obtained with excitation at 482

3

nm (∼50 kW/cm2) and detection with a longpass filter above 510 nm. Each spot on the image is a

4

single dimer nanoparticle or an individual particle of size 80 nm – the scale bar is 5 µm. B) The

5

same area scan as shown in image A obtained with excitation at 639 nm (∼4 kW/cm2) and

6

detection with a bandpass filter centred at 695 nm with a transmission window of 55 nm. C)

7

Photoluminescence spectrum of single dimer particles identified as spots 1 to 5 in image A, and

8

one example of an individual gold nanoparticle labelled no. 6.

9 10

The measurement of fluorescence enhancement in the emission of Atto-655 dye was

11

performed by collecting light from a spot where an individual dimer was found. For this purpose,

12

the dimer sample was immersed in a nanomolar solution of Atto-655 molecules, which during

ACS Paragon Plus Environment

14

Page 15 of 39 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 Journal of Physical Chemistry

1

the measurement diffuse and explore the space surrounding the individual dimer (Fig. 3A). The

2

interparticle gap is the hot-spot for fluorescence enhancement because of the large local field

3

created there by the bonding plasmon mode. The broad spectral distribution of this plasmon

4

mode, ranging from 600 to 800 nm, is suitable for fluorescence enhancement of a wide range of

5

red-emitting dyes. The most favorable situation occurs when the dye’s emission is slightly red-

6

shifted relatively to the plasmon resonance frequency, in order to minimize energy dissipation by

7

the particle’s absorption. The emission of Atto-655 is well matched with the longitudinal mode

8

in dimers’ spectra (Fig. 3B) and, for this reason, this dye was chosen to illustrate fluorescence

9

enhancement.

10

The enhancement effect was measured by using laser excitation at 639 nm, which

11

addresses both the dye’s absorption and the longitudinal plasmon band of the dimer

12

nanoparticles. At this excitation wavelength, the photoluminescence images show only spots

13

from dimer particles (Fig. 2B), and their relative intensity is determined by the overlap between

14

the dimer’s spectrum and the emission filter range (Fig. 3B). The signal collected from the dimer

15

nanoparticles in the absence of Atto-655 is steady over long observation times (orange trace in

16

Fig. 3C). When the dye is added to the solution, the time traces begin to show strong emission

17

bursts (red trace in Fig. 3C) that are several orders of magnitude more intense than the dimers’

18

background signal. The emission of Atto-655 molecules in solution was also collected from a

19

surface region free of dimer nanoparticles (grey trace in Fig. 3C). In this case, the background

20

emission from Atto-655 within the confocal detection volume of the microscope is detected, but

21

without any events of intense fluorescence emission.

22

Therefore, strong fluorescence bursts are only detected from spots where dimers are

23

immobilized and when Atto-655 molecules are present in solution. These bursts are attributed to

ACS Paragon Plus Environment

15

The Journal of Physical Chemistry 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 16 of 39

1

the exploration of hot-spot regions on the surface of dimer nanoparticles by Atto-655 molecules

2

diffusing in solution and transiently interacting with the DNA. The gap region is a very small

3

volume, in the order of zeptoliter, when compared to the confocal detection volume which is

4

around four femtoliters. For the typical concentrations of Atto-655 used in our measurements,

5

the dye’s occupation number in the confocal detection volume varies between 1 and 10, and thus,

6

the occupation probability of the gap region is several orders of magnitude lower, as further

7

discussed below. For this reason, the probability of multiple occupancy of the gap region is

8

negligible, and each burst is ascribed to the exploration of the interparticle gap by only one dye

9

molecule. The strong fluorescence bursts experimentally observed then result from single Atto-

10

655 molecules exploring the interparticle gap region.

11 12

Figure 3. A) Scheme of the fluorescence enhancement experiment showing a single Atto-655

13

dye molecule (red star) diffusing through the gap (hot-spot) of a dimer nanoparticle. B)

ACS Paragon Plus Environment

16

Page 17 of 39 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 Journal of Physical Chemistry

1

Photoluminescence spectrum of an individual dimer nanoparticle (orange curve) overlapped with

2

the absorption and emission spectra of Atto-655 dye (green and red curves, respectively), and the

3

detection range of the bandpass emission filter (light pink rectangle). C) Emission intensity time

4

traces measured from the spot of an individual dimer nanoparticle in the absence of Atto-655 dye

5

in solution (top, orange curve); emission time trace from the same spot as before, but in the

6

presence of Atto-655 dye in nanomolar concentration (middle, red curve); emission time trace

7

from a region of the surface without any particle, but in the presence of Atto-655 dye in the same

8

concentration as previous (bottom, grey curve).

9 10

The duration and intensity of each fluorescence burst is closely related to the trajectory of

11

the Atto-655 molecule as it diffuses across the interparticle gap. In this region, the plasmon’s

12

near field changes rapidly. Likewise, the enhancement of the dye’s excitation and emission

13

varies as the dye molecule changes its position in the gap. The spatial dependence of the

14

emission enhancement effect together with the randomness of the dye’s trajectory explains the

15

intensity variations observed from burst to burst within each time trace collected for several

16

dimers. In order to characterize the enhancement effect, we will focus our discussion on the most

17

intense event of each time trace. Some examples of intense bursts observed for dimers of 80 nm

18

gold particles are presented in Figure 4A-D. We assume that if the observation time is long

19

enough, then the most intense event should correspond to the molecular trajectory through the

20

region of largest fluorescence enhancement. The strongest burst in each trace was used to

21

calculate the top enhancement factor, which is the ratio between enhanced and non-enhanced

22

emission for that event. The top enhancement factor is then calculated from the maximum burst

23

intensity corrected for the background signal and normalized to the average intensity of a non-

ACS Paragon Plus Environment

17

The Journal of Physical Chemistry 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 18 of 39

1

enhanced Atto-655 dye. The latter was determined to be 0.35 counts/ms for the same conditions

2

of the enhancement experiment (Fig. S4 of the SI).

3

We have found that dimers of gold particles with 40 nm give maximum enhancement

4

factors between 500 and 1000-fold (open symbols in Fig. 4E). On the other hand, dimers of 80

5

nm particles are able to enhance the emission of Atto-655 by more than a 1000-fold, and in

6

several examples, the top enhancements reach almost 4000-fold (closed symbols in Fig. 4E). As

7

expected, the emission enhancement increases with the dimer particle size, because the radiation

8

efficiency of the plasmon also increases. The range of enhancement factors is comparable for

9

different linker sizes, which are shown together in Fig. 4E, because our dimerization protocol

10

resulted in similar gap distances among samples. Surprisingly, a fraction of 16% of the dimers of

11

80 nm particles do not show any enhancement events in the sampled time interval of 60 s. This is

12

tentatively attributed to an obstruction of dimer gaps with an excess of DNA linkers that prevents

13

the dye to access the hot-spot region. Nevertheless, about 71% out from a total of 32 dimers

14

analysed show events with top enhancements above 1000-fold, and out of these about 16% are

15

above 3000-fold.

16

The top enhancement factors reported here match up the largest known values from the

17

literature for plasmonic emission enhancement of single-molecule fluorescence. Other examples

18

of enhancements larger than 1000-fold have been obtained with gold nano-bowtie or nanorod

19

antennas.6,7,28 Recently, enhancement factors of 5000-fold on the emission of Atto-655 were

20

reported with a system that uses complex nanostructures of DNA origami to assemble dimers of

21

gold nanoparticles.33 Another work reported on fluorescence enhancement using dimers

22

spontaneously assembled by particle aggregation upon surface immobilization, which resulted in

23

enhancement factors of more than 1000-fold.21 Dimer nanoantennas produced by

ACS Paragon Plus Environment

18

Page 19 of 39 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 Journal of Physical Chemistry

1

nanolithography showed enhancements of 15000-fold for the weakly fluorescent crystal violet,

2

while the enhancement reported for the fluorescent dye Alexa Fluor 647 was around 1500-fold.34

3

In comparison, the dimer antennas used in this work deliver comparable or better enhancement

4

factors, but using a DNA-directed assembly approach that is considerably more simple and

5

affordable than those based on DNA origami, while being selective for dimer assembly which is

6

not possible with spontaneous aggregation.

7 8

Figure 4. A-D) Examples of intense fluorescence bursts from Atto-655 emission enhanced by

9

dimers of gold nanoparticles – the emission enhancement factor corresponding to each burst is

10

indicated inside the figure. E) Top emission enhancement factors for single-molecule

ACS Paragon Plus Environment

19

The Journal of Physical Chemistry 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 20 of 39

1

fluorescence of Atto-655 dye plotted against the LSP peak wavelength of the individual dimer

2

nanoparticles used for fluorescence enhancement. Each symbol represents a measurement on a

3

different dimer nanoparticle within a total of 32 dimers analyzed. The symbols are color coded

4

according to the enhancement factor: blue, < 1000; green, 1000 - 2000; orange, 2000 - 3000; red,

5

> 3000. The open and closed symbols show enhancement factors obtained for dimers of 40 nm

6

and 80 nm particles, respectively.

7 8

The single-particle spectrum is known for each dimer used for fluorescence enhancement,

9

which allows us to plot the top enhancement factor relative to the peak wavelength of the

10

longitudinal surface plasmon band (Fig. 4E). This plot does not show any particular trend

11

between these two experimental observables, in contrast with other plasmonic nanoparticles used

12

for fluorescence enhancement, such as single gold nanorods.28 A plausible explanation for the

13

lack of correlation between top enhancements and the LSP peak wavelength of dimer particles is

14

the molecular congestion of gaps with DNA linkers that, by randomly occupying the interparticle

15

volume, may limit or obstruct the dye’s access to the central region of the hot-spot. In ref. 21, it

16

was demonstrated that, upon attaching bulky thiolated-PEG chains onto the surface of dimer

17

particles, the respective enhancement factors decrease, in agreement with an exclusion effect of

18

the dye from the dimer hot-spot due to molecular crowding of the interparticle gap. Even though,

19

in our case, the dimers are extensively coated with DNA, these linkers do not seem to have such

20

a detrimental effect, and large enhancement factors above 1000-fold can be measured in 71% of

21

the dimers analysed. Actually, we hypothesize that the DNA linkers may increase the effective

22

concentration of Atto-655 in the hot spot, as supported by a simple argument based on the hot-

23

spot volume and its expected occupancy. For this purpose, we have first determined from DDA

ACS Paragon Plus Environment

20

Page 21 of 39 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 Journal of Physical Chemistry

1

near field simulations that the decay length of the plasmon field is around 10 nm in a direction

2

perpendicular to the interparticle axis (for a dimer of 80 nm particles with a gap of 10 nm).

3

Assuming that the hot-spot is approximately cylindrical with radius and height of 10 nm × 10 nm

4

gives a volume of 3.1 zL, which for an average dye concentration around 4 nM gives an

5

occupancy of 7.6 × 10-6 molecules in the hot-spot. Another way to analyze this volume

6

occupancy would result from multiplying it by the duration of time traces and calculate the

7

estimated occupation time of one molecule in the hot-spot. For time traces with a duration of 60

8

s, this would correspond to one molecule being present in the hot-spot for about 450 µs during

9

the time trace interval. The frequency of burst events observed ranges from 20 events up to

10

several hundred per trace with an average burst duration around 120 µs (Fig. 5). The total

11

duration of enhanced emission events obtained from multiplying the number of bursts per trace

12

by its average duration is much larger than the estimated occupation time of 450 µs. Thus, we

13

hypothesize that the dye’s interaction with the DNA somehow favors the occupancy of the hot-

14

spot and the observation of fluorescence enhancement from our dimer particles.

15

ACS Paragon Plus Environment

21

The Journal of Physical Chemistry 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 22 of 39

1 2

Figure 5. A) Emission intensity time trace showing enhanced fluorescence bursts – the red

3

dashed line shows the threshold used to discriminate intense events of enhanced single-molecule

4

fluorescence. We have used an intensity threshold of µ + 6σ, as proposed in ref. 32. where µ is

5

the average background signal and σ is the standard deviation. B) Photon count rate histogram

6

from the emission trace shown in “A” and the normal distribution fitted (magenta curve) fitted to

7

the background emission. This fitting was used to find the average background signal and its

8

standard deviation (µ and σ). The number and duration of events with intensity above µ + 6σ

9

were counted on each time trace for the 32 dimers analyzed. C) Top emission enhancement

10

factors for single-molecule fluorescence of Atto-655 dye plotted against the number of

11

fluorescence bursts during the time trace. The symbols and respective colors represent the same

12

enhancement factors shown in Fig. 4E. D) Similar plot using the average burst duration to

13

represent the top enhancement factors. The weak correlation found in the plots (C) and (D)

ACS Paragon Plus Environment

22

Page 23 of 39 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 Journal of Physical Chemistry

1

suggests that dimers with more available gaps, i.e. showing a higher number of bursts or of

2

longer duration, can reach higher enhancement factors.

3 4

We have performed model simulations of gold nanoparticle dimers to compare

5

theoretically predicted fluorescence enhancements with our experimental results. The model

6

dimers were defined for a particle size of 80 nm and the interparticle gap was varied between 2

7

and 10 nm, i.e. in the range of those experimentally inferred from optical spectra. The gap region

8

is where the near field enhancement is mostly concentrated (Fig. 6A). Even for a gap separation

9

of 10 nm, the near field enhancement corresponds to an intensity increase of almost 1000-fold

10

along the interparticle axis, but it decays within approx. 10 nm. We have assumed three fixed

11

positions of the dye in the gap to evaluate how the fluorescence enhancement changes with the

12

position (right side of Fig. 6A). It also makes it possible to discuss how the maximum

13

enhancement effect would be limited if access to the central regions of the hot-spot is restricted

14

by molecular congestion of DNA linkers. The enhancement of fluorescence emission results

15

from the combined changes in the excitation, radiative and non-radiative decay rates of the dye

16

induced by interaction with the nanoparticle antenna. The enhancement factor of excitation rate

17

(exc ) and fluorescence quantum yield (ϕ⁄ϕ ) were calculated separately, as presented in the SI.

18

Here, we will focus the discussion on the overall emission enhancement, i.e.  ⁄  calc = exc ×

19

ϕ⁄ϕ , and on the reduced fluorescence lifetime, i.e. = 1⁄ r + nr .

20

The overall emission enhancements calculated for the central position of the gap yielded

21

values ranging from 1000 to 10000-fold increase in the dye’s emission (Fig. 6B). This result

22

compares well with our experimental top enhancement factors which range from 1000 to almost

23

4000-fold. The comparison is made on the basis of the top enhancement values experimentally

ACS Paragon Plus Environment

23

The Journal of Physical Chemistry 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 24 of 39

1

measured assuming that these correspond to the dye being close to gap’s centre, where the

2

largest enhancements are expected. Actually, the absolute maximum is probably at some position

3

along the interparticle axis shifted away from the centre, where there is an optimal radiative over

4

non-radiative decay rate enhancements.32 On the other hand, the intensity distribution of each

5

fluorescence burst experimentally measured is affected by the trajectory and residence time of

6

the dye in the gap. The position dependence of the enhancement effect can be evaluated from the

7

values of  ⁄  calc for positions away from the gap centre (semi-filled and open symbols in

8

Fig. 6B). The enhancement effect decays by orders of magnitude over a few tens of nanometers,

9

as it closely follows a similar distance dependence of the excitation rate enhancement. As

10

previously hypothesized, the lack of correlation between top enhancements and the LSP peak

11

wavelength of dimer particles may be explained by molecular hindrance of DNA linkers that

12

limits the dye’s access to the central region of the hot-spot. The strong spatial dependence of the

13

overall emission enhancement, as seen in our calculations, suggests that this is a plausible

14

explanation.

ACS Paragon Plus Environment

24

Page 25 of 39 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 Journal of Physical Chemistry

1 2

Figure 6. A) Near field maps calculated from DDA simulations of a dimer of gold nanoparticles

3

with a size of 80 nm and a gap separation of 10 nm for an incident wavelength of 639 nm

4

polarized across the interparticle axis – the scale bar is 20 nm. The near field map on the right

5

shows a magnification of the interparticle gap where the positions considered for the point-like

6

dipole of the emitting dye molecule were fixed – the number shown is the displacement in nm

7

away from the interparticle axis. B) Enhancement factor calculated for the overall fluorescence

8

emission  ⁄  calc for the interaction of Atto-655 dye with a dimer of gold nanoparticles of

9

size 80 nm at different gap separations – closed symbols correspond to a dye molecule

ACS Paragon Plus Environment

25

The Journal of Physical Chemistry 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 26 of 39

1

positioned at the interparticle joining axis, semi-filled and open symbols are for a position 10 and

2

20 nm away from the axis, respectively. C) Fluorescence lifetime decrease ( ⁄ ) calculated for

3

the same conditions described before, where  is the decay time of Atto-655 in water (1.8 ns).

4 5

The interparticle distance has a pronounced effect in the longitudinal LSP peak

6

wavelength, as well as in the fluorescence enhancement although in two opposite ways. Short

7

interparticle distances induce larger near fields in the gap, which favor the dye’s excitation and

8

radiative decay rate enhancements, but the close proximity of the metal surface also favors the

9

non-radiative decay through energy transfer and ohmic losses. The interplay between these

10

opposing factors reduces the variation of emission enhancement with the LSP peak wavelength

11

(or conversely, the gap distance) in dimer particles. On the other hand, the effect on the

12

fluorescence lifetime from the acceleration of radiative and non-radiative decay rates is always

13

characterized by a pronounced lifetime decrease (Fig. 6C). The predicted decay rates in the hot-

14

spot region are in the ps and sub-ps timescales, which is below the time resolution of our setup

15

(grey dashed line). Indeed, the fluorescence decays retrieved from strong burst events show a

16

decay profile that approximately coincides with the instrument response function (Fig. S5 of the

17

SI).

18

The fluorescence emission time traces contain detected photons from both enhanced and

19

non-enhanced dye molecules within the confocal detection volume (Fig. 7). In order to separate

20

these two contributions, we have used the information about fluorescence lifetime to analyse our

21

results by fluorescence lifetime correlation spectroscopy (FLCS).39,40 Fluorescence enhancement

22

by plasmonic antennas is characterized by the acceleration of excited-state decay rates, and thus

ACS Paragon Plus Environment

26

Page 27 of 39 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 Journal of Physical Chemistry

1

the emission from enhanced dye molecules can be distinguished from that of non-enhanced

2

molecules by its short decay time. The fluorescence decays recorded simultaneously with each

3

emission trace shows multi-exponential decay profiles, which can be decomposed in short and

4

long decay components, as shown in Figure 7B (red and blue curves, respectively). The short

5

decay components are attributed to the enhanced decay rates of the dye together with some

6

background signal from the dimer particle. It can be described by two exponential components

7

with average decay times of 70 ps (84.5%) and 1.08 ns (3.5%). The long lifetime of 2.44 ns is

8

similar to the lifetime of Atto-655 in aqueous solution (1.8 ns), although it is slightly longer

9

probably due to adsorption onto the polymer-glass substrate. The FLCS analysis uses the decay

10

components previously defined from decay analysis to build mathematical filters (Fig. 7C),

11

which are then used to weigh the detected photons in order to retrieve separate correlation

12

functions for each decay component (Fig. 7D). In this work, the FLCS analysis was used to

13

distinguish between the correlation function of photons from enhanced dye molecules, which are

14

associated with short decay times, from that of non-enhanced molecules (red and blue curves in

15

Fig. 7D, respectively).

16

ACS Paragon Plus Environment

27

The Journal of Physical Chemistry 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 28 of 39

1 2

Figure 7. A) Emission intensity time trace showing intense fluorescence bursts from Atto-655

3

enhanced emission (dark grey) and background signal from the dimer particle alone (light grey).

4

B) Fluorescence decay of the same emission trace (dark grey) decomposed into short and long

5

decay components shown as red and blue curves, respectively (IRF is depicted in light grey, and

6

the inset shows the weighed-residuals from a tri-exponential fit). C) Filter weights obtained by

7

FLCS analysis based on the short and long decay components defined to separate the enhanced

8

and non-enhanced emission (red and blue curves, respectively). D) Correlation curve of

9

enhanced and non-enhanced emission (red and blue curves, respectively) obtained by FLCS

ACS Paragon Plus Environment

28

Page 29 of 39 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 Journal of Physical Chemistry

1

analysis, and non-filtered correlation curve of the emission trace (dark grey) and of the

2

background signal (light grey), as shown in (A).

3 4

The main difference between the correlation function of enhanced and non-enhanced

5

emission is that the first one shows an additional decay at short times that is not present in the

6

latter. We have confirmed this trend by performing the FLCS analysis of other emission traces

7

showing substantial enhancement effects (Fig. S6 of the SI). We have fitted the correlation

8

functions with the commonly used free-diffusion model, which afforded reasonable fits using

9

two diffusion terms. The free-diffusion model is used here just to extract relaxation times, which

10

are then compared to calculated diffusion times, in order to discuss the possible origin of the

11

several decay components observed in the correlation function. As it will later become obvious,

12

the free-diffusion picture is not adequate to explain our results. The fast decay component that is

13

only present in the correlation function of enhanced emission has an average relaxation time of

14

20 µs, and it can be associated with the short and very intense fluorescence bursts observed in

15

emission traces. Assuming that these events correspond to dye molecules crossing the hot-spot in

16

the gap region, the expected free-diffusion time would be in the sub-microsecond time scale,

17

because of the small volume of the hot-spot. The estimated diffusion time would be 0.22 µs for

18

an Atto-655 molecule diffusing in water across the transverse dimension of 10 nm calculated for

19

the near field in the gap region. The retrieved relaxation time of 20 µs could be tentatively

20

explained by transient sticking of the dye to DNA linkers, or simply, by the limited time

21

resolution of our correlation curves that allows us only to measure a relaxation tail of the fast

22

decay component. On the other hand, the long decay component occurs on a similar time scale

23

for the correlation function of enhanced and non-enhanced emission (inset of Fig. 7D). The

ACS Paragon Plus Environment

29

The Journal of Physical Chemistry 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 30 of 39

1

average relaxation time of this long component is around 2.7 ms, and is attributed to the dye’s

2

adsorption/desorption on the surface surrounding the dimer particle. A similar behavior has been

3

previously reported for emission enhancement of single-molecule fluorescence using surface

4

immobilized gold nanorods.19,28 The slower motion of dye molecules due to sticking onto the

5

surface, or due to interaction with DNA linkers, seems to be corroborated by intense emission

6

events that last longer than a few ms, which are seldom observed in emission traces (Fig. S7 of

7

the SI). Alternatively, it is possible to retrieve separate correlation functions for enhanced and

8

non-enhanced emission by using the polarization properties of the longitudinal surface plasmon

9

mode.21 Here, we have used the emission polarization of dimer particles to select those that are

10

aligned in such a way that emission along the LSP mode, or perpendicular to it, was divided by

11

the beam-splitter analyzer into two detection channels (Fig. S8 of the SI). The correlation

12

function of the enhanced emission shows at short times an additional relaxation component that

13

is not present in correlation of non-enhanced emission, similarly to the FLCS analysis. The

14

comparable results between these approaches validate the use of FLCS to separate enhanced

15

from non-enhanced emission in plasmon-coupled fluorescence, as shown here.

16 17

Conclusions

18

We have confirmed that gold nanoparticle dimers produced by DNA self-assembly make

19

efficient plasmonic antennas for fluorescence enhancement of red-emitting dyes. The

20

combination of large-sized nanoparticles with nanometric gaps generate hot-spots with intense

21

near fields in the interparticle region that are able to enhance the fluorescence of Atto-655 dye by

22

more than 1000-fold, while in some cases, top emission enhancements reach almost 4000-fold.

ACS Paragon Plus Environment

30

Page 31 of 39 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 Journal of Physical Chemistry

1

Fluorescence lifetime correlation lead to the recovery of a short relaxation component exclusive

2

to the enhanced emission likely due to the dye interaction with DNA linkers. The DNA-directed

3

assembly approach used here is selective for dimer assembly, which is not possible with

4

spontaneous particle aggregation, and is considerably more simple and affordable than

5

approaches based on DNA origami templates. These two features of dimer particles, i.e. large

6

fluorescence enhancements and preparation by affordable self-assembly colloidal methods,

7

renders these plasmonic antennas promising for application in optical imaging or sensing. We

8

have also shown that fluorescence lifetime correlation spectroscopy can be used to separate

9

plasmon-enhanced from non-enhanced emission. Further studies using this approach will enable

10

to explore its potential in extracting more information from plasmon-enhanced fluorescence.

11 12

ASSOCIATED CONTENT

13

Supporting Information. Oligonucleotide sequences of DNA linkers; Details on DNA-

14

directed assembly of dimer particles; Additional images of purification gels, and TEM of dimer

15

particle samples; Average emission intensity of non-enhanced Atto-655 dye; Fluorescence

16

decays of enhanced and non-enhanced Atto-655 emission; Details on the analysis of correlation

17

curves and examples of time traces showing long events of enhanced emission; FCS analysis of

18

polarization resolved emissions; Details on the estimation of fluorescence enhancement from

19

DDA calculations.

20

The following files are available free of charge.

21

brief description (file type, i.e., PDF)

ACS Paragon Plus Environment

31

The Journal of Physical Chemistry 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 32 of 39

1 2

AUTHOR INFORMATION

3

Corresponding Author

4

* E-mail: [email protected]

5

Author Contributions

6

The manuscript was written through contributions of all authors. All authors have given approval

7

to the final version of the manuscript.

8

Funding Sources

9

Fundação para a Ciência e a Tecnologia, FCT.

10 11

ACKNOWLEDGMENT

12

Authors gratefully acknowledge financial support from Fundação para a Ciência e a Tecnologia,

13

FCT

14

PTDC/CTM-NAN/2700/2012).

15

(SFRH/BPD/111906/2015). DB acknowledges a PhD grant from BIOTECnico Program

16

(PD/BD/113630/2015).

(REEQ/115/QUI/2005,

Pest-OE/QUI/UI0100/2013/2014, PMRP

acknowledges

a

UID/BIO/04565/2013

Post-Doc

grant

from

and FCT

17 18 19

REFERENCES 1

Giannini, V.; Fernández-Domínguez, A. I.; Heck, S. C.; Maier, S. A. Plasmonic

20

Nanoantennas: Fundamentals and Their Use in Controlling the Radiative Properties of

21

Nanoemitters. Chem. Rev. 2011, 111, 3888-3912.

22

2

Novotny, L.; van Hulst, N. Antennas for Light. Nat. Photon. 2011, 5, 83-90.

ACS Paragon Plus Environment

32

Page 33 of 39 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

1

The Journal of Physical Chemistry

3

2 3

Rep. Prog. Phys. 2012, 75, 024402. 4

4 5

Biagioni, P.; Huang, J.-S.; Hecht, B. Nanoantennas for Visible and Infrared Radiation.

Anger, P.; Bharadwaj, P.; Novotny, L. Enhancement and Quenching of Single-Molecule Fluorescence. Phys. Rev. Lett. 2006, 96, 113002.

5

Kühn, S.; Håkanson, U.; Rogobete, L.; Sandoghdar, V. Enhancement of Single-Molecule

6

Fluorescence Using a Gold Nanoparticle as an Optical Nanoantenna. Phys. Rev. Lett.

7

2006, 97, 017402.

8

6

9

Single-Molecule Fluorescence Enhancements Produced by a Bowtie Nanoantenna. Nat.

10 11

Kinkhabwala, A.; Yu, Z.; Fan, S.; Avlasevich, Y.; Müllen, K.; Moerner, W. E. Large

Photon. 2009, 3, 654−657. 7

Yuan, H.; Khatua, S.; Zijlstra, P.; Yorulmaz, M.; Orrit, M. Thousand-Fold Enhancement

12

of Single-Molecule Fluorescence Near a Single Gold Nanorod. Angew. Chem., Int. Ed.

13

2013, 52, 1217−1221.

14

8

Muskens, O. L.; Giannini, V.; Sánchez-Gil, J. A.; Rivas, J. G. Strong Enhancement of the

15

Radiative Decay Rate of Emitters by Single Plasmonic Nanoantennas. Nano Lett. 2007,

16

7, 2871-2875.

17

9

Ringler, M.; Schwemer, A.; Wunderlich, M.; Nichtl, A.; Kürzinger, K.; Klar, T. A.;

18

Feldmann, J. Shaping Emission Spectra of Fluorescent Molecules with Single Plasmonic

19

Nanoresonators. Phys. Rev. Lett. 2008, 100, 203002.

20

10

Busson, M. P.; Rolly, B.; Stout, B.; Bonod, N.; Bidault, S. Accelerated Single Photon

21

Emission from Dye Molecule-Driven Nanoantennas Assembled on DNA. Nat. Commun.

22

2012, 3, 962−966.

ACS Paragon Plus Environment

33

The Journal of Physical Chemistry 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

1

11

Page 34 of 39

Li, J.; Krasavin, A. V.; Webster, L.; Segovia, P.; Zayats, A. V.; Richards, D. Spectral

2

Variation of Fluorescence Lifetime Near Single Metal Nanoparticles. Sci Rep. 2016, 6,

3

21349.

4

12

5 6

Kosako, T.; Kadoya, Y.; Hofmann, H. F. Directional Control of Light by a Nano-Optical Yagi–Uda Antenna. Nat. Photon. 2010, 4, 312-315.

13

Curto, A. G.; Volpe, G.; Taminiau, T. H.; Kreuzer, M. P.; Quidant, R.; van Hulst, N. F.

7

Unidirectional Emission of a Quantum Dot Coupled to a Nanoantenna. Science 2010,

8

329, 930-933.

9

14

Jain, P. K.; Ghosh, D.; Baer, R.; Rabani, E.; Alivisatos, A. P. Near-Field Manipulation of

10

Spectroscopic Selection Rules on the Nanoscale. Proc. Natl. Acad. Sci. 2012, 109, 8016-

11

8019.

12

15

Ueno, K.; Juodkazis, S.; Shibuya, T.; Yokota, Y.; Mizeikis, V.; Sasaki, K.; Misawa, H.

13

Nanoparticle Plasmon-Assisted Two-Photon Polymerization Induced by Incoherent

14

Excitation Source. J. Am. Chem. Soc. 2008, 130, 6928-6929.

15

16

de Torres, J.; Mivelle, M.; Moparthi, S. B.; Rigneault, H.; Van Hulst, N. F.; García-

16

Parajó, M. F.; Margeat, E.; Wenger, J. Plasmonic Nanoantennas Enable Forbidden

17

Förster Dipole−Dipole Energy Transfer and Enhance the FRET Efficiency. Nano Lett.

18

2016, 16, 6222-6230.

19

17

Pellegrotti, J. V.; Acuna, G. P.; Puchkova, A.; Holzmeister, P.; Gietl, A.; Lalkens, B.;

20

Stefani, F. D.; Tinnefeld, P. Controlled Reduction of Photobleaching in DNA Origami-

21

Gold Nanoparticle Hybrids. Nano Lett. 2014, 14, 2831-2836.

22 23

18

Warren, S. C.; Thimsen, E. Plasmonic Solar Water Splitting. Energy Environ. Sci. 2012, 5, 5133-5146.

ACS Paragon Plus Environment

34

Page 35 of 39 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

1

The Journal of Physical Chemistry

19

Khatua, S.; Yuan, H.; Orrit, M. Enhanced-Fluorescence Correlation Spectroscopy at

2

Micro-Molar Dye Concentration Around a Single Gold Nanorod. Phys. Chem. Chem.

3

Phys. 2015, 17, 21127-21132.

4

20

Pradhan, B.; Khatua, S.; Gupta, A.; Aartsma, T.; Canters, G.; Orrit, M. Gold-Nanorod-

5

Enhanced Fluorescence Correlation Spectroscopy of Fluorophores with High Quantum

6

Yield in Lipid Bilayers. J. Phys. Chem. C 2016, 120, 25996-26003.

7

21

Punj, D.; Regmi, R.; Devilez, A.; Plauchu, R.; Moparthi, S. B.; Stout, B.; Bonod, N.;

8

Rigneault, H.; Wenger, J. Self-Assembled Nanoparticle Dimer Antennas for Plasmonic-

9

Enhanced Single-Molecule Fluorescence Detection at Micromolar Concentrations. ACS

10 11

Photonics 2015, 2, 1099-1107. 22

Cang, H.; Labno, A.; Lu, C.; Yin, X.; Liu, M.; Gladden, C.; Liu, Y.; Zhang, X. Probing

12

the Electromagnetic Field of a 15-Nanometre Hotspot by Single Molecule Imaging.

13

Nature 2011, 469, 385-388.

14

23

Wertz, E.; Isaacoff, B. P.; Flynn, J. D.; Biteen, J. S. Single-Molecule Super-Resolution

15

Microscopy Reveals How Light Couples to a Plasmonic Nanoantenna on the Nanometer

16

Scale. Nano Lett. 2015, 15, 2662-2670.

17

24

Su, L.; Yuan, H.; Lu, G.; Rocha, S.; Orrit, M.; Hofkens, J.; Uji-i, H. Super-resolution

18

Localization and Defocused Fluorescence Microscopy on Resonantly Coupled Single-

19

Molecule, Single-Nanorod Hybrids. ACS Nano 2016, 10, 2455-2466.

20

25

Taminiau, T. H.; Stefani, F. D.; van Hulst, N. F. Single Emitters Coupled to Plasmonic

21

Nano-Antennas: Angular Emission and Collection Efficiency. New J. Phys. 2008, 10,

22

105005.

ACS Paragon Plus Environment

35

The Journal of Physical Chemistry 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

1

26

2 3

Rogobete, L.; Kaminski, F.; Agio, M.; Sandoghdar, V. Design of Plasmonic Nanoantennae for Enhancing Spontaneous Emission. Opt. Lett. 2007, 32, 1623-1625.

27

4 5

Page 36 of 39

Bharadwaj, P.; Novotny, L. Spectral Dependence of Single Molecule Fluorescence Enhancement. Opt. Express 2007, 15, 14266-14274.

28

Khatua, S.; Paulo, P. M.; Yuan, H.; Gupta, A.; Zijlstra, P.; Orrit, M. Resonant Plasmonic

6

Enhancement of Single-Molecule Fluorescence by Individual Gold Nanorods. ACS Nano

7

2014, 8, 4440-4449.

8

29

9

Nanoantennas Increase Maximal Photon Stream and Total Photon Number. J. Phys.

10 11

Wientjes, E.; Renger, J.; Cogdell, R.; van Hulst, N. F. Pushing the Photon Limit:

Chem. Lett. 2016, 7, 1604-1609. 30

Acuna, G. P.; Möller, F. M.; Holzmeister, P.; Beater, S.; Lalkens, B.; Tinnefeld, P.

12

Fluorescence Enhancement at Docking Sites of DNA Directed Self-Assembled

13

Nanoantennas. Science 2012, 338, 506-510.

14

31

Bidault, S.; Devilez, A.; Maillard, V.; Lermusiaux, L.; Guigner, J.-M.; Bonod, N.;

15

Wenger J. Picosecond Lifetimes with High Quantum Yields from Single-Photon Emitting

16

Colloidal Nanostructures at Room Temperature. ACS Nano 2016, 10, 4806-4815.

17

32

Zhang, T.; Gao, N.; Li, S.; Lang, M. J.; Xu Q.-H. Single-Particle Spectroscopic Study on

18

Fluorescence Enhancement by Plasmon Coupled Gold Nanorod Dimers Assembled on

19

DNA Origami. J. Phys. Chem. Lett. 2015, 6, 2043-2049.

20

33

Puchkova, A.; Vietz, C.; Pibiri, E.; Wünsch, B.; Paz, M. S.; Acuna, G. P.; Tinnefeld, P.

21

DNA Origami Nanoantennas with over 5000-fold Fluorescence Enhancement and Single-

22

Molecule Detection at 25 µM. Nano Lett. 2015, 15, 8354-8359.

ACS Paragon Plus Environment

36

Page 37 of 39 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

1

The Journal of Physical Chemistry

34

Flauraud, V.; Regmi, R.; Winkler, P. M.; Alexander, D. T. L.; Rigneault, H.; van Hulst,

2

N. F.; Garcia-Parajo, M. F.; Wenger, J.; Brugger, J. In-Plane Plasmonic Antenna Arrays

3

with Surface Nanogaps for Giant Fluorescence Enhancement. Nano Lett. 2017, 17, 1703-

4

1710.

5

35

Teixeira, R.; Paulo, P. M. R.; Costa, S. M. B. Gold Nanoparticles in

6

Core−Polyelectrolyte−Shell Assemblies Promote Large Enhancements of Phthalocyanine

7

Fluorescence. J. Phys. Chem. C 2015, 119, 21612-21619.

8

36

9

Nanodimers with Tunable Sizes and Interparticle Distances and Their Surface Plasmonic

10 11

Properties. Small 2013, 9, 2308-2315. 37

12 13

Lan, X.; Chen, Z.; Liu, B.-J.; Ren, B.; Henzie, J.; Wang, Q. DNA-Directed Gold

Ray, K.; Zhang, J.; Lakowicz, J. R. Fluorescence Lifetime Correlation Spectroscopic Study of Fluorophore-Labeled Silver Nanoparticles. Anal. Chem. 2008, 80, 7313-7318.

38

Martin-Garcia, B.; Paulo, P. M. R.; Costa, S. M. B.; Velazquez, M. M.

14

Photoluminescence Dynamics of CdSe QD/Polymer Langmuir-Blodgett Thin Films:

15

Morphology Effects. J. Phys. Chem. C 2013, 117, 14787-14795.

16

39

17 18

Gregor, I.; Enderlein, J. Time-resolved Methods in Biophysics. 3. Fluorescence Lifetime Correlation Spectroscopy. Photochem. Photobiol. Sci. 2007, 6, 13-18.

40

Kapusta, P.; Macháň, R.; Benda, A.; Hof, M. Fluorescence Lifetime Correlation

19

Spectroscopy (FLCS): Concepts, Applications and Outlook. Int. J. Mol. Sci. 2012, 13,

20

12890-12910.

21

41

Hill, H. D.; Millstone, J. E.; Banholzer, M. J.; Mirkin, C. A. The Role Radius of

22

Curvature Plays in Thiolated Oligonucleotide Loading on Gold Nanoparticles. ACS Nano

23

2009, 3, 418-424.

ACS Paragon Plus Environment

37

The Journal of Physical Chemistry 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

1

42

Page 38 of 39

Zhang, X.; Gouriye, T.; Göeken, K.; Servos, M. R.; Gill, R.; Liu, J. Toward Fast and

2

Quantitative Modification of Large Gold Nanoparticles by Thiolated DNA: Scaling of

3

Nanoscale Forces, Kinetics, and the Need for Thiol Reduction. J. Phys. Chem. C 2013,

4

117, 15677−15684.

5

43

Teixeira, R.; Paulo, P. M. R.; Viana, A. S.; Costa, S. M. B. Plasmon-Enhanced Emission

6

of a Phthalocyanine in Polyelectrolyte Films Induced by Gold Nanoparticles. J. Phys.

7

Chem. C 2011, 115, 24674–24680.

8

44

9

Capabilities and Known Limitations. J. Quant. Spectrosc. Radiat. Transfer 2011, 112,

10 11

Yurkin, M. A.; Hoekstra, A. G. The Discrete-Dipole-Approximation Code ADDA:

2234-2247. 45

D’Agostino, S.; Pompa, P. P.; Chiuri, R.; Phaneuf, R.; Britti, D. G.; Rinaldi, R.;

12

Cingolani, R.; Della Sala, F. Enhanced Fluorescence by Metal Nanospheres on Metal

13

Substrates. Opt. Lett. 2009, 34, 2381-2383.

14

46

D’Agostino, S.; Della Sala, F.; Andreani, L. C. Dipole-Excited Surface Plasmons in

15

MetallicNanoparticles:

16

DipoleApproximation. Phys. Rev. B: Condens. Matter Mater. Phys. 2013, 87,

17

205413−13.

18 19

47

Engineering

Decay

Dynamics

within

the

Discrete-

Nordlander, P.; Oubre, C.; Prodan, E.; Li, K.; Stockman, M. I. Plasmon Hybridization in Nanoparticle Dimers. Nano Lett. 2004, 4, 899-903.

20 21

ACS Paragon Plus Environment

38

Page 39 of 39 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 Journal of Physical Chemistry

1 2

TOC graphic

ACS Paragon Plus Environment

39