Nonfluorescent optical probing of single molecules and nanoparticles

Non

fluorescent Optical Probing of Single Molecules and

Nanoparticles

Thomas Jollans, Martin D. Baaske, and Michel Orrit

*

Huygens−Kamerlingh Onnes Laboratory, Leiden University, Postbus 9504, 2300 RA Leiden, The Netherlands

ABSTRACT: Notwithstanding its unique power for imaging and investigation of transparent condensed and biological matter, fluorescence presents severe limitations: it requires specialfluorescent labels, which are prone to photobleaching, and the photon streams it provides are relatively weak. In the

past 10 to 20 years nonfluorescent optical methods have

appeared, which can also provide information on matter at the nanoscale, while presenting different limitations. In the present paper, we review some of these methods, with special emphasis on work from our group. We consider mostly the optical detection and study of single immobilized or transiently bound molecules and nanoparticles through their

scattering, the heat they dissipate in the environment upon light absorption, or their coupling to auxiliary optical resonators such as whispering-gallery modes.

■

INTRODUCTION

Optical microspectroscopy remains one of the most direct and powerful methods to investigate matter at nanometer scales: to explore and clarify the molecular structure, dynamics, and processes at work in physical chemistry, materials science, or bioscience. Much of the current knowledge of molecular materials at the nanoscale has been garnered through the fluorescence of suitable labels dispersed in, or attached to, the structures of interest. Fluorescence is the workhorse technique in cellular biology. Its exquisite sensitivity reaches down to

single-molecule detection1 and enables super-resolution

microscopy.2 Single-molecule methods enable exploration of molecular dynamics, free from time- and ensemble-averaging.3 For all their success, however, fluorescence techniques face limitations: they require labeling of the targeted molecules, which may alter their interactions and functions, and fluorescence signals are restricted in intensity and detection rate, as well as in spectral selectivity, which directly restricts the number of different labels observable in the same sample. The

urge to overcome the restrictions of fluorescence and to

directly observe unaltered molecules in action has led several groups to explorefluorescence-free optical methods, which we

briefly review in the present article. Although they face

limitations of their own, these fluorescence-free methods

cannot or do not have to rely on photon-counting detectors. Therefore, they potentially provide higher light intensities and larger detection bandwidths thanfluorescence.

Optical interaction of light with matter can give rise to nonoptical signals (e.g., a photovoltaic current) that we do not consider here or to optical signals. Because it changes the color of photons,fluorescence is easy to isolate from narrow-band (laser) excitation with suitable spectralfilters. However, other properties of light may also be modified by interaction with

matter, such as the polarization or the propagation wavefront. Processes affecting the latter are generally referred to as scattering. One usually distinguishes scattering processes according to the time dependence of the scattering potential. Truly static defects lead to truly elastic scattering, i.e., scattering without any change of frequency. Elastic scattering should thus be distinguished from Rayleigh scattering in fluids,4

where density fluctuations are never truly static, but relax on viscosity-dependent time scales. Acoustic waves in solid orfluid matter lead to Brillouin scattering lines, whereas higher-frequency vibrations lead to Raman scattering. As Brillouin and Raman scattering (the surface-enhanced variant5 excepted) are usually too weak to detect single small objects, we will focus on elastic and Rayleigh scattering here, i.e., scattering without a significant frequency (or color) change. The optical theorem6 states that interference of the scattered wave with the incident wave is responsible for the extinction signal, i.e., for a loss of intensity in the incident wave after it has interacted with the object of interest. For single nanoscale objects, the changes they induce, seen either in scattering or extinction, are extremely weak in relative values. Their observation thus requires very large numbers of detected photons to overcome shot noise. However, because the intensity that a nanoparticle or a molecule can withstand is limited, naı̈ve measurements of scattering or extinction are rendered impractical by the extremely long integration times they require; more subtle techniques are needed.

Essentially, two strategies have been developed to solve this problem, depending on whether the objects to be detected Received: January 28, 2019

Revised: April 1, 2019

Published: April 9, 2019

Feature Article

pubs.acs.org/JPCC

Cite This:J. Phys. Chem. C 2019, 123, 14107−14117

Derivative Works (CC-BY-NC-ND) Attribution License, which permits copying and redistribution of the article, and creation of adaptations, all for non-commercial purposes.

Downloaded via LEIDEN UNIV on January 22, 2020 at 14:30:59 (UTC).

absorb or are purely dielectric. In the first case, one uses a probe beam to optically detect a consequence of absorption that vanishes in the absence of absorption, for example, heat dissipated by the particle upon light absorption of a heating beam with another color. This method is called photothermal microscopy.7−15The change of intensity of the scattered probe beam is thus used as a proxy for absorption of the heating beam. The other strategy, applicable to nonabsorbing as well as absorbing objects, is to improve noise rejection in direct scattering (i.e., with a single probe beam), which is only possible with bright-field scattering on realistic samples in condensed matter. This technique has recently been popularized under the name iSCATbut has been implemented in different guises with various optical designs.16−19

Initial experiments at low temperatures16benefited from the resonance enhancement of the scattering cross section and have been improved significantly in recent years.20 However, the difficulty of measuring scattering in cryogenic environ-ments has limited their applications so far. Alternatively, optical resonators can be used as efficient detectors for small objects: whispering gallery modes21,22 or plasmonic struc-tures23can be used alone or in combination24−26to detect and study small objects. In the present article, we start with a short discussion of the signal-to-noise ratio in bright-field scattering and discuss some of the most usual methods: photothermal microscopy, interferometric methods in microscopy, and coupling to plasmonic or all-dielectric cavities. Although we focus on immobilized particles, these methods can also be applied to diffusing particles and to on-the-fly measurements.27

■

SIGNAL, BACKGROUND, AND NOISE IN

SCATTERING EXPERIMENTS

We consider the optical detection of a small object, which we call the scatterer and which we will often approximate as a point electric dipole. In the most general scheme, we illuminate this object and look at thefield Escit scatters. Together with the scattered field, we will also collect at least a part of the incidentfield, which we call Eref. We then measure an intensity I_mthat results from a superposition of these twofields

ϕ

∝ | + | = ̅ + ̅ + ̅ ̅

I_m E_ref E_sc2 E_ref2 E_sc2 2E E_{ref sc}cos (1)

where the bars indicate the (positive) amplitudes of thefields.

We have assumed both fields to be spatially matched (see

below), and ϕ is a phase angle between the reference and

scattered field, which depends on the geometry of the

experiment, on the position of the scatterer, and on the frequency of the excitation laser.

The simple expression in eq 1 allows us to discuss

fundamental issues of signal, background, and noise. Wefirst assume that detection noise is limited to photon or shot noise, i.e., the number fluctuation of detection events in an ideal photon-counting detection chain. In this case, we immediately see fromeq 1that, without scattering, the background intensity Bm= E̅ref2 gives rise to noise that scales as the square root of the detected intensity, i.e., as E̅_ref. The signal, which we identify as the change of detected intensity when we illuminate the scatterer, S_m= E̅_sc2 _{+ 2E̅}

refE̅sccosϕ, changes in a more complex way depending on the interference between reference and scatteredfields:

(i) For very weak scatterers, the scattered intensity E̅sc2 can often be neglected. Therefore, for an ideal detection, both the signal and the background noise scale as E̅ref: the

signal-to-noise ratio is independent of the reference field. In other words, the signal-to-noise ratio cannot be improved by increasing (or decreasing) the amplitude of the reference field. Of course, in a practical experiment, other noise sources must be considered, and the amplitude of the referencefield is a crucial parameter to optimize, as briefly outlined below.

(ii) If the reference field can be eliminated altogether, the signal-to-noise ratio can in principle become arbitrarily large. This is the case for fluorescence, where a spectral detection filter completely suppresses the illumination light, while transmitting afluorescence signal proportional to the absorbed intensity. In that case, the background and the noise associated with it disappear completely, and the fluorescence signal is detected on (ideally) a dark background. The same argument would seem to apply to dark-field scattering. Unfortunately, as the intensity scattered by a single molecule or nanoparticle is exceedingly small, even a weak background will easily overcome it. Such a background is caused by Rayleigh and Brillouin scattering by the matrix surrounding the scatterer, by defects of the sample and of the optical components, or by stray reflections at the surfaces of the setup’s lenses and mirrors.

We now consider deviations from the ideal scheme above: (i) Noise sources: in a simple optical detection experiment, noise arises not only from quantumfluctuations of the optical signal itself, characterized by its noise equivalent power28

σ_ph= 2BPhν, B being the detection bandwidth, P the

detected power, and hν the photon energy. Important

additional noise sources are power fluctuations of the

excitation laser (laser noise σ_laser), detection noise arising either from the detection electronics (dark noise σdark) or excess noise from the electron avalanche often involved in the amplification process for photon-counting devices.29 This latter is given by σ_exc= F−1·σ_ph, F being an effective factor describing the total noise created by photon noise and excess avalanche noise of the detector,30so thatσph2 +σexc2 = F· 2BPhν). Excess noise is usually negligible for analog detectors which do not rely on avalanches to amplify the signal.

(ii) Mode matching: eq 1 supposes that the reference and scatteredfields interfere at a given point in space. However, in real experiments, thefields are integrated over a significant part of the wavefront in order to optimize the signal collected from a small object. Interference then takes place between extended wave fronts which in general do not coincide, e.g., the spherical wave scattered by a point dipole and an incident Gaussian reference beam.31 The overlap of the two interfering fields introduces a mode-matching factor which reduces the

amplitude of the interference term. Noting the field

distributions of each wave in the detector plane as Eref(x,y) = E̅refFref(x,y) and Esc(x,y) = E̅scFsc(x,y), where Fref and Fscare square-normalized functions bearing the spatial dependence of thefields, we find a complex mode overlap factor

∬

α = F*_ref( , ) ( , )d dx y F x y x y_sc

(2)

so that the intensity becomes

∬

α ϕ

= | + | = ̅ + ̅ + ̅ ̅ | |

I_m E_ref E_sc2d dx y E_ref2 E_sc2 2E E_{ref sc} cos

(3)

The amplitude|α| ≤ 1 reduces the interference term, whereas the argument of the complex overlap factor may modify the angleϕ.

In conclusion, to optimize the signal-to-noise ratio in scattering experiments, different strategies must be pursued simultaneously to reduce fundamental and experimental noise sources. To reduce statistical fluctuations of the number of detected photons, one should apply the highest intensities and integration times permitted by the photostability of the objects

under study. Laser power fluctuations can be removed to a

large extent by optical subtraction, either in an interference setup18,32 or by clever design of the optical path.19 The detected intensity resulting from the interference of reference and scatteredfields must be adapted to the best regime of the detector. This can be done in different ways, for example, in

reflection by careful index matching at the glass−sample

interface12or in transmission by interferometry.18Finally, the interference between reference and scatteredfields should be improved by matching the two modes as much as possible, for example by introducing a weakly transmitting mirror (Figure 3.1) or mask to reduce the referencefield.33,34

■

PHOTOTHERMAL TECHNIQUES

One of the main challenges in any scattering-based detection technique is the background created by stray scatterers such as dust, impurities in microscope oil, or the environment the measurement is being performed in. This problem of selectivity can be particularly daunting in an intrinsically scattering environment such as a cell.

Photothermal microscopy, a highly selective technique for detecting objects which absorb light of a chosen wavelength or set of wavelengths, relies on an indirect measurement. Here,

we focus on the all-optical scattering-based variant.7 Closely related to a large, well-established family of photothermal and photoacoustic techniques,15,35−39it uses a two-color approach to detect and localize absorbers: A resonant heating laser beam is absorbed by an object in the sample, which transforms a significant part of the absorbed energy into heat. The heat diffuses out of the particle and into the surrounding medium, where it establishes a localized, time-dependent temperature gradient and a corresponding refractive index gradient. This local refractive index gradient acts as a (thermal) lens and can be detected through the scattering of a second (probe) laser beam7,11,13,40(seeFigure 1). Therefore, the method is closely related to the scattering signal discussed above. The position of the probe focus with respect to the thermal lens influences the

angle ϕ and should be varied to optimize the photothermal

difference signal. Although the heating intensity is limited by photodamage and/or saturation of the absorber, the scattered probe signal depends only on a broadband refractive index change. Therefore, the probe wavelength can be judiciously chosen out of the spectral absorption range of the absorber to minimize damage and saturation by the probe, thereby allowing for high probe intensities and accordingly low photon noise.11

To discriminate the signal from background scattering, photothermal and related techniques modulate the heating beam at a certain frequency, generally in excess of 100 kHz. The corresponding oscillating component of the scattered probe beam is then extracted to recover the actual signal generally with a lock-in amplifier.7The technique is sensitive to absorbing objects only and extremely robust in the face of even substantial nonabsorbing scatterers. This makes noble-metal nanoparticles, which are only weakly luminescent but strongly absorbent, ideal contrast agents even in noisy biological

Figure 1.(a) Simplified scheme of a photothermal microscope. (b) Cartoon representation of thermal lens creation in a photothermal microscope’s focus.

Figure 2.(a) Photothermal signal-to-noise ratio as a function of the incident probe power measured for 20 nm gold NPs in water on glass (SNR∝ P1/2_{). (b) Photothermal SNR measured for 20 nm NPs in different fluids as a function of calculated photothermal strength for these fluids, Σ}

PT= n|∂n/∂T|Cp−1, scaled with respect to glycerol. (c) Normalized histograms of SNR for 20 nm gold NPs in glycerol: (dark gray) deposited on glass; (light gray) deposited on a 100 nm thick thermal isolation layer of PMMA (figure reproduced with permission from Gaiduk et al.11).

environments,10,41 although other contrast agents have been proposed42 and label-free live-cell photothermal imaging has been demonstrated.43In very carefully optimized experiments using judiciously chosen media and samples, absorption

measurements down to single molecules12,44 have been

demonstrated.

The signal-to-noise ratio achievable in a photothermal measurement depends not only on optical considerations but also on the choice of modulation frequency and environment. Tuning the frequency allows one, ideally, to limit ubiquitous 1/ f noise while matching the thermal diffusion length to the optical probe volume. Adapting the environment, both medium and substrate, allows one to affect the strength of the refractive index gradient through∂n/∂T and limit invisible heat loss into the substrate (seeFigure 2). Moving away from “conventional” media, ∂n/∂T and the SNR can be increased dramatically by working near the critical point of, e.g., xenon or carbon dioxide, where most material properties (including the refractive index) vary violently with the thermodynamic variables of state. There, detection of dissipated powers∼64 pW has been demonstrated, nearly 2 orders of magnitude better than the∼3 nW seen in glycerol.45

■

INTERFEROMETRIC METHODS

Interference-based scattering microscopy (iSCAT, seeFigure 3) utilizes the interference between the field scattered by an analyte and the field reflected by the glass slide (eq 1, final term). The interference term scales linearly with the volume of

the analyte V and the magnitude of the reflected field. In

comparison to dark-field methods, which only detect the

scattered intensity Es2 ∝ V2, this technique yields improved contrast for small analytes. This method has enabled the spectroscopy of single gold nanoparticles with radii down to 10 nm17 and the extinction-based imaging of a single quantum dot.19 It was further successfully used in conjunction with single-moleculefluorescence microscopy to track translational and rotational motion of single virus particles (Simian Virus 40).46 Moreover, iSCAT has enabled the detection of single proteins as small as bovine serum albumin (BSA, molecular mass: 65 kDa) binding to a cover slide surface and their localization with 5 nm accuracy.47

The phase factor cosϕ of the interference term also depends on the distance between the scatterer and the interface on which the reference reflection occurs. This can be used for three-dimensional tracking of nanoparticles. The motion patterns obtained that way not only allow for determining the potential of electrostatic traps48 but also provide a new means to access and investigate dynamic biological processes like the movement of single proteins during diffusion in the

plasma membrane, transport along filopodia, and

endocyto-sis.49 Instead of backward scattering, coherent forward

scattering can also be used for tracking single virus particles in three dimensions.50

The imaging contrast that can be obtained via iSCAT is

technically limited by the dynamic range of the camera used to record the images. This, however, can be mitigated by artificially reducing the collection efficiency for the reflected field, while enabling high throughput of the scattered field by

placing transmission/reflection masks in the infinity

con-jugated path of the microscope to spatiallyfilter the respective

Figure 3.Interferometric scattering microscopy. (1) Adapted with permission from Cole et al.:33(1b) setup of a typical iSCATmicroscope, incident beam, focused into the back focal plane of the objective, and the reflection on the sample slide in darker blue; optical path of light scattered by the analytes in light blue. In this case the setup contains a partial reflector (PR) which couples the light into the microscope and attenuates the reflected beam while transmitting most of the scatteredfield in order to optimize the field overlap factor α and with it the imaging contrast. (1a, 1c) Working principle: spatial intensity distribution of the scattered light in the back focal plane. The position of the PR is represented by the circle/gray area. (2) Adapted with permission from Young et al.:55 iSCAT images that allow determination of the molecular mass of scattering analytes. (2A) Schematic of the experiment indicating the adsorption of analyte molecules with different mass. (2B) iSCATimage of BSA (bovine serum albumin) molecules adsorbing to a glass slide. (2C) Images showing the contrast increasing for BSA oligomers consisting of one to four molecules. The corresponding occurrence statistics are represented in the top panel of (2D) alongside the respective scatter plots (bottom).

fields (Figure 3.1).33,34Moreover, the method has enabled the observation of microtubule motion,51 the imaging of single proteins secreted from a single cell,52 and the monitoring of nanoparticle-labeled lipids diffusing inside lipid membranes spanned over pores.53Furthermore, it has been used to track

nanoparticles on the membrane of live cells54 and for

quantitative determination of single-molecule mass (seeFigure 3.2).55

While using the reflected field as a reference provides a common path for the interferingfields and thus makes for a more stable device, it is not intrinsically necessary. Using a Michelson-type configuration to provide for the interference allows one to control the intensity and polarization state of the referencefield, enabling the detection of nanoparticles moving through nanoscalefluidic channels.18,56,57

■

CAVITY-BASED TECHNIQUES

Another kind of optical tool that enables the label-free observation of nanoscopic entities is represented by optical microcavities. A specific family of microresonators, which relies on the confinement of light via total internal reflection (TIR) on the outer interface of an axially symmetric cavity, are the

so-called Whispering Gallery Mode (WGM) resonators.21,22

Named after a similar acoustic phenomenon observed in the whispering gallery at St Paul’s Cathedral in London, these devices are able to confine light in micrometer-sized mode volumes for extensive periods of time, ranging from nano- to microseconds (i.e., 106₋₁₀9 _{optical oscillations). An} electro-magnetic wave undergoing TIR exhibits an evanescent tail that extends into the lower refractive index medium beyond the interface, consequently making the resonators susceptible to refractive index perturbations in the surrounding medium. As

the light field probes the surrounding medium repeatedly

during successive cavity round trips, even minute perturbations by nanoscopic entities can be detected as changes in the cavity’s resonance spectrum. There, the excess polarizability of the entity with respect to the surrounding medium gives rise to a shift in the WGM’s resonance frequency,58

allowing for the detection of nanoparticles and viruses.59,60In addition, losses caused via scattering and absorption of light by the entity can

be observed as broadening of the WGM’s resonance.61−64 Precise quantification of single analyte properties, however, requires the accurate determination of its position, as line width broadening and frequency shift depend on the analyte’s location within the WGM’s field distribution. The necessary information can be obtained by scanning the cavity with a

probe beam and observation of the cavity’s photothermal

response24,65 or via simultaneous frequency tracking of

multiple nondegenerate WGMs with different polar mode

numbers.66,67 A WGM-based method that does not require

this information relies on analyte-induced mode splitting: Light scattered by nanoparticles is efficiently coupled into the counter-propagating cavity mode. This leads to the formation of two standing wave modes with either their intensity node or antinode at the scatterer’s location and thus to distinct frequencies and line widths.61,68−70The ratio of the differences in frequency and the differences in line width between the split

modes are independent of the local field strength and thus

allow for the direct determination of the analyte’s polar-izability.71,72

A recent study shows that further sensitivity enhancements, especially for perturbations by small scatterers, are possible using exceptional points.74 Mode-splitting-based sensing, however, goes hand in hand with the highest demands regarding cavity Q factors and suggests the choice of active/ lasing microcavities as sensors, to compensate for nanoparticle-induced losses via optical gain.74−77

The frequency shift a WGM undergoes when perturbed by an analyte scales with the electricfield intensity at the location of the analyte and with the analyte’s polarizability. Con-sequently, the near-field enhancement provided by plasmonic nanoparticles resonantly excited by WGMs has allowed boosting WGM sensitivity to a level at which single molecules (see Figure 4) can be resolved.63,73,78,79 This increase, however, comes at the cost of a strongly decreased sensing

volume, now limited to the volume where the field strength

around the plasmonic NP is highest (≈104_nm3_{). This imposes} severe restrictions on the total number of molecules detectable by assays of the one-way-binding type. Modification of the NPs with chemical agents that only allow the analyte of interest to

Figure 4.Single-molecule detection with plasmonically enhanced optical microcavities. (a) Typical layout of a prism-coupled WGM single-molecule sensor (adapted from Baaske et al.63). Light from a tunable laser is evanescently coupled into a microsphere cavity via total internal reflection on the prism’s surface. The microsphere cavity is modified by adsorption of gold nanorods on its surface. (b) Typical WGM transmission spectra exhibiting Lorentzian dips where the wavelength of the tunable excitation laser matches a WGM resonance (red: unperturbed state, blue: the resonance shifted upon perturbation). (c) Sketch illustrating transient (left) and the permanent interaction (right) of an analyte molecule with receptor molecules on the surface of gold nanorods inside a WGM’s evanescent field. The wavelength traces displayed in (d) and (e) show typical wavelength shift patterns (spikes and steps) associated with transient and permanent analyte−receptor interactions, respectively (adapted with permission from Kim et al.73).

bind is necessary in order to ensure detection specificity. Consequently, significant statistics of single-molecule events can only be obtained if transient, highly specific interactions between the target and receptor molecules (tethered to the

plasmonic NP) are monitored,63 which also enables strong

positive and negative controls. Nonetheless, plasmonically boosted WGM sensors can be used to monitor single-molecule

surface modification processes and enzyme conformational

changes associated with single-molecule interactions in real time and over a wide range of environmental conditions.73,79 WGM-based recognition of single zinc and mercury ions interacting with gold nanorods has been reported;80however, the precise physical mechanism underlying this observation has yet to be determined.

Recently, the capacity of WGM-based resonators to act as single-particle photothermal spectrometers has been demon-strated:24,65,81 Heat dissipated by single nanoparticles upon absorption of a probe beam locally heats a microcavity, altering its refractive index profile and thus the WGM’s frequency. At wavelengths where the probe wavelength coincides with a WGM resonance, the photothermal absorption spectrum

exhibits distinct Fano patterns on top of the particle’s

absorption line. The shape of each individual Fano pattern reflects the frequency detuning of the nanoparticle’s resonance with respect to the resonance of the corresponding WGM.

■

PLASMONICS

In a perfect conductor, charge carriers would respond to any electricfield instantly. In real metals, this clearly cannot be the

case, not least because a frequency-independent response would be inconsistent with, among others, energy thresholds in

the photoelectric effect. In the simple Drude−Sommerfeld

family of models, the frequency dependence of the electric and optical properties of a metal can be reduced to two parameters, the plasma frequency and the Drude time.82,83In the bulk, this primarily represents the frequency of a charge oscillation with a large wavelength;84at the interface with a dielectric, however, the electron plasma supports captive resonances in its

neighborhood,85 known as surface plasmon resonances

(SPRs).

While SPRs are significant in thin films, localized SPRs

(LSPRs) in metal nanoparticles and, more generally, three-dimensional nanostructures have a much stronger and more versatile optical response.86The strong plasmonic response is what gives solutions of gold nanoparticles their distinctive red appearancedescribed by Faraday87 and first explained fully by Gustav Mie88and what gives many medieval stained glass windows their color.89

The strong and stable optical response of metal nano-particles makes them suitable for use as optical labels with which to track biomolecules. In contrast tofluorophores, they do not suffer from photoblinking or -bleaching, at the cost of significantly larger sizes. As such, they have been used, e.g., to

track phospholipids and proteins diffusing in cell

mem-branes49,90 and to detect substeps of the rotation of F1 -ATPase.91

Since the SPR is a surface effect, it is highly sensitive to the refractive index of the dielectric medium near the surface.

Figure 5.Spectral signature of two nearby plasmonic nanoparticles coupling (plasmonic nanoruler). (a) Nanoparticles attached to a glass surface/ to each other with BSA−biotin/biotin−spreptavidin. Silver (b) and gold (c) have different colors depending on whether they are individual particles (b/c left) or pairs (b/c right) (inset: TEM.) (d) Representative scattering spectra of single particles and pairs of silver (top) and gold (bottom). Reproduced with permission from Sönnichsen et al.96

LSPRs couple primarily to a nearfield with a size scale of some nanometers, which is especially enhanced in regions of high curvature by the lightning-rod effect.92 This coupling of a

property measurable in the far field, the surface plasmon

resonance, to the optical properties in a zeptoliter volume is what gives single metal nanoparticles their power for sensing.93 While nanoscopic sensors are impractical for many

applica-tions,94 they allow for the detection of rare events and

potentially tiny concentrations of analytes. In particular, a single-gold-nanorod-based sensor detecting the arrival in, and departure from, the nearfield of individual molecules has been demonstrated.23,95

When two plasmonic nanoparticles approach each other to

distances comparable with the size of the near field, the

coupling between the two, which can be thought of as plasmon hybridization,86 is highly dependent on the distance (see

Figure 5). This opens the use of a pair of nearby nanoparticles as a plasmonic nanoruler.96Suitably attached to a biomolecule, such a ruler can be used to monitor conformational dynamics in real time.97 Recently, Ye et al. demonstrated continuous monitoring of the conformational dynamics of a single protein

for 24 h.98 If two nanoparticles are linked with

well-characterized elastic molecules, a nanoruler becomes a plasmonic nanospring which can be used as a force sensor with optical readout.99

■

CONCLUSIONS AND OUTLOOK

In this paper, we have reviewed current optical methods providing single-molecule or single-nanoparticle sensitivity, while not relying on fluorescence detection. In practice, all these methods rely on light scattering in one or the other guise. Scattering has three main advantages overfluorescence:

(i) Scattering signals are stable, and they do not blink nor bleach. Indeed, as scattering relies on the molecule’s optical polarizability, it essentially reflects the molecule’s integrity and chemical stability, rather than the complex electron-vibration coupling processes affecting fluorescence.

(ii) Bright-field scattering signals are often intense.

Scattering is not limited by nanosecondfluorescence lifetimes

and can often be detected with higher efficiency than the

fluorescence of single molecules. Nonabsorbing objects can be illuminated by strong laser beams without photodamage, provided the laser wavelength is chosen properly. Absorbing molecules or particles, in particular, metal nanoparticles, usually sustain much higher intensities and illumination doses thanfluorescent dyes.

(iii) Scattering makes it possible to avoid labels altogether. Nonlabeled species placed in a suitable medium, e.g., a protein in solution, will scatter light. Of course, the absence of labels comes at a cost: the lack of specificity of scattering. Two molecules with the same polarizability will give indistinguish-able scattering signals. One could restore specificity through

chemical functionalization or through specific chemical

recognition, which is the solution favored by life throughout

evolution. However, just as with fluorescent labels, these

additional chemical interactions may alter the properties and functions of the objects to be detected.

We have described three main nonfluorescent optical

methods.

(i) Photothermal microscopy is based on the absorption of specific objects, which can be considered as possible labels for molecules or nanoparticles. Just asfluorescent dyes, absorbing metal nanoparticles can be fabricated in different colors, thanks

to variations in metal composition, size, and shape of the nanoparticles. Absorbing labels can be imaged on a virtually dark background thanks to modulation techniques because the optical absorption of most media of interest, including living cells, is very low when the heating wavelength is chosen in a

proper transparency window. Compared to the fluorescent

labels so ubiquitous in cell biology, these absorbing labels are often much more stable but also much bulkier.

(ii) Bright-field scattering exploits the tiny changes of a focused wavefront by a small scatterer. Interferometric scattering (iSCAT) is the most advanced modality of this method. It provides unique sensitivity and accuracy for in vitro conditions, in the absence of other scatterers or at least of any time-dependent scattering processes. In that case, background is removed through highly sensitive image subtraction

techniques. The iSCAT method has been compared to mass

spectrometry in the liquid phase, but it is also particularly useful to study interactions and assembly of single molecules in vitro. Such studies would be much more difficult, if possible at all, in a complex system such as a live cell because of the lack of molecular specificity.

(iii) It is possible to engineer light waves to enhance light− matter interactions with respect to those at the focus of a microscope. Those methods use optical cavities in a broad sense. A plasmonic hot spot can act as a cavity with low quality factor (Q) and high confinement, whereas optical cavities such as whispering gallery modes present a high Q and a (relatively) large mode volume. Whereas those two types of cavities have their own specific advantages, they can also be combined to

benefit from both a high Q and a small mode volume,

mitigating the original compromises.25,26The main advantage of cavity-enhanced detection is its extreme sensitivity, but its disadvantages are its complexity of operation compared to the previous methods and the difficulty involved in interfacing the optical cavity or cavities with soft matter or biological systems of interest, such as live cells. Specificity also remains an issue for these methods, as they still require either specific binding or labeling with specific molecular groups, which may alter the biomolecule’s properties and functions. A further issue is the quantification of scattering signals, which is possible only if the precise structure of the electricfield is controlled or known. Unlike in photothermal or iSCATmicroscopy, the magnitude of the optical signal in cavity-enhanced scattering cannot be obviously correlated to the detected object, unless its position is controlled very precisely. The same type of difficulty arises in surface-enhanced Raman scattering and excitation and/or emission enhancement schemes, two methods we did not discuss here.

made possible by the much higher scattering intensities than

those achieved in fluorescence. Of course, a scattering

correlation scheme would still leave two problems open: (i) the inability to correlate single-molecule dynamics on time scales longer than the diffusion time and (ii) that of the specificity of scattering signals. The latter point could be addressed by measuring additional dimensions of the scattering bursts. This could be done in real time during each burst, as in the multiparameter analysis offluorescence bursts proposed by Seidel’s group.100 _{Alternatively, different quantities could be}

measured at different times on the same samples. First steps in this direction have been proposed in scattering media by Hiroi

and Shibayama101 and with plasmonic gold nanorods by

Sönnichsen’s group.27The search is on for convenient optical signals to provide additional dimensions during the scattering bursts.

■

AUTHOR INFORMATION Corresponding Author *E-mail:orrit@physics.leidenuniv.nl. ORCID Martin D. Baaske:0000-0003-2384-7557 Michel Orrit:0000-0002-3607-3426 Notes

The authors declare no competingfinancial interest. Biographies

Thomas Jollans received his M.Phys. from the University of Oxford in 2015, where he completed a project on single-virus particle tracking using near-TIRF microscopy. He is currently pursuing a Ph.D. in the single-molecule optics group at the Leiden Institute of Physics, using photothermal techniques to study heat transfer and boiling at the nanoscale, focusing in particular on the dynamics of plasmonic vapor nanobubbles.

Martin D. Baaske is currently a postdoctoral researcher in the single-molecule optics group at the Leiden University’s Institute of Physics, where he holds a Marie Skłodowska-Curie fellowship. Previously he was a Ph.D. student at the Max Planck Institute for the Science of Light in Erlangen and received his Ph.D. in physics from the Friedrich-Alexander-Universität Erlangen-Nürnberg in 2017 for his studies on microcavity-based single-molecule detection.

Michel Orrit studied at E. N. S. in Paris and obtained his Ph.D. in Bordeaux. He observed the first fluorescence signal from a single molecule in 1990. Since then, single-molecule fluorescence has revolutionized cell biology and material science. In Leiden since 2001, Orrit uses single molecules to remove ensemble averaging and study dynamics free from synchronization. He received the Edison-Volta Prize (2016) and the Spinoza Prize (2017).

■

ACKNOWLEDGMENTS

The authors acknowledge funding by the Netherlands

Organisation for Scientific Research (NWO) and the

Zwaartekracht program NanoFront (TJ, MO). The project of transient scattering by dielectric particles has received

funding from the European Union’s Horizon 2020 research

and innovation programme under the Marie Skłodowska-Curie Grant Agreement no. 792595 (MDB).

■

REFERENCES

(1) Selvin, P.; Ha, T. Single-molecule Techniques: A Laboratory Manual; Cold Spring Harbor Laboratory Press, 2008.

(2) Hell, S. W. Nanoscopy with Focused Light (Nobel Lecture). Angew. Chem., Int. Ed. 2015, 54, 8054−8066.

(3) Moerner, W. E.; Orrit, M. Illuminating Single Molecules in Condensed Matter. Science 1999, 283, 1670−1676.

(4) Cummins, H. Z.; Gammon, R. W. Rayleigh and Brillouin Scattering in Liquids: The Landau−Placzek Ratio. J. Chem. Phys. 1966, 44, 2785−2796.

(5) Schlücker, S. Surface-Enhanced Raman Spectroscopy: Concepts and Chemical Applications. Angew. Chem., Int. Ed. 2014, 53, 4756− 4795.

(6) van de Hulst, H. C. Light scattering by small particles; Dover Classics Sci. Math.; Dover: Originally published by Wiley, 1981; New York, 1957.

(7) Boyer, D.; Tamarat, P.; Maali, A.; Lounis, B.; Orrit, M. Photothermal Imaging of Nanometer-Sized Metal Particles Among Scatterers. Science 2002, 297, 1160−1163.

(8) Berciaud, S.; Cognet, L.; Blab, G. A.; Lounis, B. Photothermal Heterodyne Imaging of Individual Nonfluorescent Nanoclusters and Nanocrystals. Phys. Rev. Lett. 2004, 93, 257402.

(9) Berciaud, S.; Cognet, L.; Lounis, B. Photothermal Absorption Spectroscopy of Individual Semiconductor Nanocrystals. Nano Lett. 2005, 5, 2160−2163.

(10) Lasne, D.; Blab, G. A.; Berciaud, S.; Heine, M.; Groc, L.; Choquet, D.; Cognet, L.; Lounis, B. Single Nanoparticle Photo-thermal Tracking (SNaPT) of 5-nm Gold Beads in Live Cells. Biophys. J. 2006, 91, 4598−4604.

(11) Gaiduk, A.; Ruijgrok, P. V.; Yorulmaz, M.; Orrit, M. Detection Limits in Photothermal Microscopy. Chem. Sci. 2010, 1, 343−350.

(12) Gaiduk, A.; Yorulmaz, M.; Ruijgrok, P. V.; Orrit, M. Room-Temperature Detection of a Single Molecule’s Absorption by Photothermal Contrast. Science 2010, 330, 353−356.

(13) Selmke, M.; Braun, M.; Cichos, F. Photothermal Single-Particle Microscopy: Detection of a Nanolens. ACS Nano 2012, 6, 2741− 2749.

(14) Selmke, M.; Braun, M.; Cichos, F. Nano-Lens Diffraction Around a Single Heated Nano Particle. Opt. Express 2012, 20, 8055− 8070.

(15) Bialkowski, S. E. Photothermal Spectroscopy Methods for Chemical Analysis; Wiley: New York, 1996.

(16) Plakhotnik, T.; Palm, V. Interferometric Signatures of Single Molecules. Phys. Rev. Lett. 2001, 87, 183602.

(17) Lindfors, K.; Kalkbrenner, T.; Stoller, P.; Sandoghdar, V. Detection and Spectroscopy of Gold Nanoparticles Using Super-continuum White Light Confocal Microscopy. Phys. Rev. Lett. 2004, 93, 037401.

(18) Ignatovich, F. V.; Novotny, L. Real-Time and Background-Free Detection of Nanoscale Particles. Phys. Rev. Lett. 2006, 96, 1−4.

(19) Kukura, P.; Celebrano, M.; Sandoghdar, V. Imaging a Single Quantum Dot when it is Dark. Nano Lett. 2009, 9, 926−929.

(20) Faez, S.; Türschmann, P.; Haakh, H. R.; Götzinger, S.; Sandoghdar, V. Coherent Interaction of Light and Single Molecules in a Dielectric Nanoguide. Phys. Rev. Lett. 2014, 113, 213601.

(21) Foreman, M.; Swaim, J.; Vollmer, F. Whispering Gallery Mode Sensors. Adv. Opt. Photonics 2015, 7, 168−168.

(22) Kim, E.; Baaske, M. D.; Vollmer, F. Towards Next-Generation Label-Free Biosensors: Recent Advances in Whispering Gallery Mode Sensors. Lab Chip 2017, 17, 1190−1205.

(23) Zijlstra, P.; Paulo, P. M. R.; Orrit, M. Optical Detection of Single Non-Absorbing Molecules using the Surface Plasmon Resonance of a Gold Nanorod. Nat. Nanotechnol. 2012, 7, 379−382. (24) Heylman, K. D.; Thakkar, N.; Horak, E. H.; Quillin, S. C.; Cherqui, C.; Knapper, K. A.; Masiello, D. J.; Goldsmith, R. H. Optical Microresonators as Single-Particle Absorption Spectrometers. Nat. Photonics 2016, 10, 788−795.

(25) Ruesink, F.; Doeleman, H. M.; Hendrikx, R.; Koenderink, A. F.; Verhagen, E. Perturbing Open Cavities: Anomalous Resonance Frequency Shifts in a Hybrid Cavity-Nanoantenna System. Phys. Rev. Lett. 2015, 115, 203904.

(26) Doeleman, H. M.; Verhagen, E.; Koenderink, A. F. Antenna− Cavity Hybrids: Matching Polar Opposites for Purcell Enhancements at Any Linewidth. ACS Photonics 2016, 3, 1943−1951.

(27) Wulf, V.; Knoch, F.; Speck, T.; Sönnichsen, C. Gold Nanorods as Plasmonic Sensors for Particle Diffusion. J. Phys. Chem. Lett. 2016, 7, 4951−4955.

(28) Bachor, H.-A. A guide to experiments in quantum optics; Wiley-VCH Weinheim: New York, 1988.

(29) van Dijk, M. A.; Lippitz, M.; Stolwijk, D.; Orrit, M. A Common-Path Interferometer for Time-Resolved and Shot-Noise-Limited Detection of Single Nanoparticles. Opt. Express 2007, 15, 2273−2287.

(30) McIntyre, R. J. Multiplication Noise in Uniform Avalanche Diodes. IEEE Trans. Electron Devices 1966, ED-13, 164−168.

(31) Zumofen, G.; Mojarad, N. M.; Sandoghdar, V.; Agio, M. Perfect Reflection of Light by an Oscillating Dipole. Phys. Rev. Lett. 2008, 101, 180404.

(32) van Dijk, M. A.; Lippitz, M.; Orrit, M. Detection of Acoustic Oscillations of Single Gold Nanospheres by Time-Resolved Interferometry. Phys. Rev. Lett. 2005, 95, 267406.

(33) Cole, D.; Young, G.; Weigel, A.; Sebesta, A.; Kukura, P. Label-Free Single-Molecule Imaging with Numerical-Aperture-Shaped Interferometric Scattering Microscopy. ACS Photonics 2017, 4, 211−216.

(34) Liebel, M.; Hugall, J. T.; van Hulst, N. F. Ultrasensitive Label-Free Nanosensing and High-Speed Tracking of Single Proteins. Nano Lett. 2017, 17, 1277−1281.

(35) McDonald, F. A. Photoacoustic, Photothermal, and Related Techniques: A Review. Can. J. Phys. 1986, 64, 1023−1029.

(36) Strohm, E. M.; Moore, M. J.; Kolios, M. C. Single Cell Photoacoustic Microscopy: A Review. IEEE J. Sel. Top. Quantum Electron. 2016, 22, 137−151.

(37) Terazima, M.; Hirota, N.; Braslavsky, S. E.; Mandelis, A.; Bialkowski, S. E.; Diebold, G. J.; Miller, R. J. D.; Fournier, D.; Palmer, R. A.; Tam, A. Quantities, Terminology, and Aymbols in Photo-thermal and Related Spectroscopies (IUPAC Recommendations 2004). Pure Appl. Chem. 2004, 76, 1083−1118.

(38) Tokeshi, M.; Uchida, M.; Hibara, A.; Sawada, T.; Kitamori, T. Determination of Subyoctomole Amounts of Nonfluorescent Molecules Using a Thermal Lens Microscope: Subsingle-Molecule Determination. Anal. Chem. 2001, 73, 2112−2116.

(39) Wang, L. V. Ultrasound-Mediated Biophotonic Imaging: A Review of Acousto-Optical Tomography and Photo-Acoustic Tomography. Dis. Markers 2004, 19, 123−138.

(40) van Dijk, M. A.; Tchebotareva, A. L.; Orrit, M.; Lippitz, M.; Berciaud, S.; Lasne, D.; Cognet, L.; Lounis, B. Absorption and Scattering Microscopy of Single Metal Nanoparticles. Phys. Chem. Chem. Phys. 2006, 8, 3486−3495.

(41) Cognet, L.; Tardin, C.; Boyer, D.; Choquet, D.; Tamarat, P.; Lounis, B. Single Metallic Nanoparticle Imaging for Protein Detection in Cells. Proc. Natl. Acad. Sci. U. S. A. 2003, 100, 11350−11355.

(42) Kim, J.-W.; Galanzha, E. I.; Shashkov, E. V.; Moon, H.-M.; Zharov, V. P. Golden Carbon Nanotubes as Multimodal Photo-acoustic and Photothermal High-Contrast Molecular Agents. Nat. Nanotechnol. 2009, 4, 688−694.

(43) Lasne, D.; Blab, G. A.; Giorgi, F. D.; Ichas, F.; Lounis, B.; Cognet, L. Label-Free Optical Imaging of Mitochondria in Live Cells. Opt. Express 2007, 15, 14184−14193.

(44) Hou, L.; Adhikari, S.; Tian, Y.; Scheblykin, I. G.; Orrit, M. Absorption and Quantum Yield of Single Conjugated Polymer Poly[2-methoxy-5-(2-ethylhexyloxy)-1,4-phenylenevinylene] (MEH-PPV) Molecules. Nano Lett. 2017, 17, 1575−1581.

(45) Ding, T. X.; Hou, L.; van der Meer, H.; Alivisatos, A. P.; Orrit, M. Hundreds-Fold Sensitivity Enhancement of Photothermal Microscopy in Near-Critical Xenon. J. Phys. Chem. Lett. 2016, 7, 2524−2529.

(46) Kukura, P.; Ewers, H.; Muller, C.; Renn, A.; Helenius, A.; Sandoghdar, V. High-Speed Nanoscopic Tracking of the Position and Orientation of a Single Virus. Nat. Methods 2009, 6, 923−927.

(47) Piliarik, M.; Sandoghdar, V. Direct Optical Sensing of Single Unlabelled Proteins and Super-Resolution Imaging of their Binding Sites. Nat. Commun. 2014, 5, 4495.

(48) Krishnan, M.; Mojarad, N.; Kukura, P.; Sandoghdar, V. Geometry-Induced Electrostatic Trapping of Nanometric Objects in a Fluid. Nature 2010, 467, 692−695.

(49) Taylor, R. W.; Gholami Mahmoodabadi, R.; Rauschenberger, V.; Giessl, A.; Schambony, A.; Sandoghdar, V. Interferometric

Scattering Microscopy Reveals Microsecond Nanoscopic Protein Motion on a Live Cell Membrane. bioRxiv 2018.

(50) Huang, Y.-F.; Zhuo, G.-Y.; Chou, C.-Y.; Lin, C.-H.; Chang, W.; Hsieh, C.-L. Coherent Brightfield Microscopy Provides the Spatiotemporal Resolution To Study Early Stage Viral Infection in Live Cells. ACS Nano 2017, 11, 2575−2585.

(51) Andrecka, J.; Ortega Arroyo, J.; Lewis, K.; Cross, R. A.; Kukura, P. Label-Free Imaging of Microtubules with sub-nm Precision Using Interferometric Scattering Microscopy. Biophys. J. 2016, 110, 214− 217.

(52) McDonald, M. P.; Gemeinhardt, A.; König, K.; Piliarik, M.; Schaffer, S.; Völkl, S.; Aigner, M.; Mackensen, A.; Sandoghdar, V. Visualizing Single-Cell Secretion Dynamics with Single-Protein Sensitivity. Nano Lett. 2018, 18, 513−519.

(53) Spindler, S.; Sibold, J.; Gholami Mahmoodabadi, R.; Steinem, C.; Sandoghdar, V. High-Speed Microscopy of Diffusion in Pore-Spanning Lipid Membranes. Nano Lett. 2018, 18, 5262−5271.

(54) de Wit, G.; Albrecht, D.; Ewers, H.; Kukura, P. Revealing Compartmentalized Diffusion in Living Cells with Interferometric Scattering Microscopy. Biophys. J. 2018, 114, 2945−2950.

(55) Young, G.; Hundt, N.; Cole, D.; Fineberg, A.; Andrecka, J.; Tyler, A.; Olerinyova, A.; Ansari, A.; Marklund, E. G.; Collier, M. P.; et al. Quantitative Mass Imaging of Single Biological Macromolecules. Science 2018, 360, 423−427.

(56) Ignatovich, F. V.; Topham, D.; Novotny, L. Optical Detection of Single Nanoparticles and Viruses. IEEE J. Sel. Top. Quantum Electron. 2006, 12, 1292−1299.

(57) Deutsch, B.; Beams, R.; Novotny, L. Nanoparticle Detection using Dual-Phase Interferometry. Appl. Opt. 2010, 49, 4921−4925.

(58) Arnold, S.; Khoshsima, M.; Teraoka, I.; Holler, S.; Vollmer, F. Shift of Whispering-Gallery Modes in Microspheres by Protein Adsorption. Opt. Lett. 2003, 28, 272−274.

(59) Vollmer, F.; Arnold, S.; Keng, D. Single Virus Detection from the Reactive Shift of a Whispering-Gallery Mode. Proc. Natl. Acad. Sci. U. S. A. 2008, 105, 20701−4.

(60) Lu, T.; Lee, H.; Chen, T.; Herchak, S.; Kim, J.-H.; Fraser, S. E.; Flagan, R. C.; Vahala, K. High Sensitivity Nanoparticle Detection using Optical Microcavities. Proc. Natl. Acad. Sci. U. S. A. 2011, 108, 5976−5979.

(61) Gorodetsky, M. L.; Pryamikov, A. D.; Ilchenko, V. S. Rayleigh Scattering in High-Q Microspheres. J. Opt. Soc. Am. B 2000, 17, 1051−1057.

(62) Shao, L.; Jiang, X.-F.; Yu, X.-C.; Li, B.-B.; Clements, W. R.; Vollmer, F.; Wang, W.; Xiao, Y.-F.; Gong, Q. Detection of Single Nanoparticles and Lentiviruses Using Microcavity Resonance Broad-ening. Adv. Mater. 2013, 25, 5616−5620.

(63) Baaske, M. D.; Foreman, M. R.; Vollmer, F. Single-Molecule Nucleic Acid Interactions Monitored on a Label-Free Microcavity Biosensor Platform. Nat. Nanotechnol. 2014, 9, 933−939.

(64) Shen, B.-Q.; Yu, X.-C.; Zhi, Y.; Wang, L.; Kim, D.; Gong, Q.; Xiao, Y.-F. Detection of Single Nanoparticles using the Dissipative Interaction in a High-Q Microcavity. Phys. Rev. Appl. 2016, 5, 024011. (65) Knapper, K. A.; Pan, F.; Rea, M. T.; Horak, E. H.; Rogers, J. D.; Goldsmith, R. H. Single-Particle Photothermal Imaging via Inverted Excitation Through High-Q All-Glass Toroidal Microresonators. Opt. Express 2018, 26, 25020−25030.

(66) Keng, D.; Tan, X.; Arnold, S. Whispering Gallery Micro-Global Positioning System for Nanoparticle Sizing in Real Time. Appl. Phys. Lett. 2014, 105, 071105.

(67) Foreman, M. R.; Keng, D.; Treasurer, E.; Lopez, J. R.; Arnold, S. Whispering Gallery Mode Single Nanoparticle Detection and Sizing: The Validity of the Dipole Approximation. Opt. Lett. 2017, 42, 963−966.

(68) Weiss, D. S.; Sandoghdar, V.; Hare, J.; Lefèvre-Seguin, V.; Raimond, J.-M.; Haroche, S. Splitting of High-Q Mie Modes Induced by Light Backscattering in Silica Microspheres. Opt. Lett. 1995, 20, 1835−1837.

(69) Kippenberg, T. J.; Spillane, S. M.; Vahala, K. J. Modal Coupling in Traveling-Wave Resonators. Opt. Lett. 2002, 27, 1669−1671.

(70) Kippenberg, T. J.; Tchebotareva, A. L.; Kalkman, J.; Polman, A.; Vahala, K. J. Purcell-Factor-Enhanced Scattering from Si Nanocrystals in an Optical Microcavity. Phys. Rev. Lett. 2009, 103, 027406.

(71) Mazzei, A.; Götzinger, S.; de S. Menezes, L.; Zumofen, G.; Benson, O.; Sandoghdar, V. Controlled Coupling of Counter-propagating Whispering-Gallery Modes by a Single Rayleigh Scatterer: A Classical Problem in a Quantum Optical Light. Phys. Rev. Lett. 2007, 99, 173603.

(72) Zhu, J.; Ozdemir, S. K.; Xiao, Y.-F.; Li, L.; He, L.; Chen, D.-R.; Yang, L. On-Chip Single Nanoparticle Detection and Sizing by Mode Splitting in an Ultrahigh-Q Microresonator. Nat. Photonics 2010, 4, 46−49.

(73) Kim, E.; Baaske, M. D.; Vollmer, F. In Situ Observation of Single-Molecule Surface Reactions from Low to High Affinities. Adv. Mater. 2016, 28, 9941−9948.

(74) Chen, W.; Özdemir, S. K.; Zhao, G.; Wiersig, J.; Yang, L. Exceptional Points Enhance Sensing in an Optical Microcavity. Nature 2017, 548, 192−196.

(75) Li, B.-B.; Clements, W. R.; Yu, X.-C.; Shi, K.; Gong, Q.; Xiao, Y.-F. Single Nanoparticle Detection using Split-Mode Microcavity Raman Lasers. Proc. Natl. Acad. Sci. U. S. A. 2014, 111, 14657−14662. (76) He, L.; Özdemir, S. K.; Zhu, J.; Kim, W.; Yang, L. Detecting Single Viruses and Nanoparticles using Whispering Gallery Micro-lasers. Nat. Nanotechnol. 2011, 6, 428−432.

(77) Özdemir, Ş. K.; Zhu, J.; Yang, X.; Peng, B.; Yilmaz, H.; He, L.; Monifi, F.; Huang, S. H.; Long, G. L.; Yang, L. Highly Sensitive Detection of Nanoparticles with a Referenced and Self-Heterodyned Whispering-Gallery Raman Microlaser. Proc. Natl. Acad. Sci. U. S. A. 2014, 111, E3836−E3844.

(78) Dantham, V. R.; Holler, S.; Barbre, C.; Keng, D.; Kolchenko, V.; Arnold, S. Label-Free Detection of Single Protein Using a Nanoplasmonic-Photonic Hybrid Microcavity. Nano Lett. 2013, 13, 3347−3351.

(79) Kim, E.; Baaske, M. D.; Schuldes, I.; Wilsch, P. S.; Vollmer, F. Label-Free Optical Detection of Single Enzyme-Reactant Reactions and Associated Conformational Changes. Sci. Adv. 2017, 3, No. e1603044.

(80) Baaske, M. D.; Vollmer, F. Optical Observation of Single Atomic Ions Interacting with Plasmonic Nanorods in Aqueous Solution. Nat. Photonics 2016, 10, 733−739.

(81) Horak, E. H.; Rea, M. T.; Heylman, K. D.; Gelbwaser-Klimovsky, D.; Saikin, S. K.; Thompson, B. J.; Kohler, D. D.; Knapper, K. A.; Wei, W.; Pan, F.; et al. Exploring Electronic Structure and Order in Polymers via Single-Particle Microresonator Spectroscopy. Nano Lett. 2018, 18, 1600−1607.

(82) Drude, P. Zur Ionentheorie der Metalle. Phys. Z. 1900, 1, 161− 165.

(83) Sommerfeld, A.; Bethe, H. Aufbau der Zusammenhängenden Materie; Handbuch der Physik 24/2; Springer Berlin Heidelberg, 1933; pp 333−622.

(84) Dressel, M.; Scheffler, M. Verifying the Drude Response. Ann. Phys. 2006, 15, 535−544.

(85) Ritchie, R. H. Plasma Losses by Fast Electrons in Thin Films. Phys. Rev. 1957, 106, 874−881.

(86) Zijlstra, P.; Orrit, M.; Koenderink, A. F. In Nanoparticles; de Mello Donegá, C., Ed.; Springer Berlin Heidelberg, 2014; pp 53−98. (87) Faraday, M. The Bakerian Lecture.Experimental Relations of Gold (and other metals) to Light. Philos. Trans. R. Soc. 1857, 147, 145−181.

(88) Mie, G. Beiträge zur Optik trüber Medien, speziell kolloidaler Metallösungen. Ann. Phys. 1908, 330, 377−445.

(89) Colomban, P.; Tournié, A.; Ricciardi, P. Raman Spectroscopy of Copper Nanoparticle-Containing Glass Matrices: Ancient Red Stained-Glass Windows. J. Raman Spectrosc. 2009, 40, 1949−1955.

(90) Fujiwara, T.; Ritchie, K.; Murakoshi, H.; Jacobson, K.; Kusumi, A. Phospholipids Undergo Hop Diffusion in Compartmentalized Cell Membrane. J. Cell Biol. 2002, 157, 1071−1082.

(91) Yasuda, R.; Noji, H.; Yoshida, M.; Kinosita, K.; Itoh, H. Resolution of Distinct Rotational Substeps by Submillisecond Kinetic Analysis of F1-ATPase. Nature 2001, 410, 898−904.

(92) Kelly, K. L.; Coronado, E.; Zhao, L. L.; Schatz, G. C. The Optical Properties of Metal Nanoparticles: The Influence of Size, Shape, and Dielectric Environment. J. Phys. Chem. B 2003, 107, 668− 677.

(93) Hill, R. T. Plasmonic Biosensors. Wiley Interdisciplinary Rev.: Nanomed. Nanobiotechnol. 2015, 7, 152−168.

(94) Dahlin, A. B. Size Matters: Problems and Advantages Associated with Highly Miniaturized Sensors. Sensors 2012, 12, 3018−3036.

(95) Beuwer, M. A.; Prins, M. W. J.; Zijlstra, P. Stochastic Protein Interactions Monitored by Hundreds of Single-Molecule Plasmonic Biosensors. Nano Lett. 2015, 15, 3507−3511.

(96) Sönnichsen, 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.

(97) Visser, E. W.; Horáček, M.; Zijlstra, P. Plasmon Rulers as a Probe for Real-Time Microsecond Conformational Dynamics of Single Molecules. Nano Lett. 2018, 18, 7927.

(98) Ye, W.; Götz, M.; Celiksoy, S.; Tüting, L.; Ratzke, C.; Prasad, J.; Ricken, J.; Wegner, S. V.; Ahijado-Guzmán, R.; Hugel, T.; et al. Conformational Dynamics of a Single Protein Monitored for 24 h at Video Rate. Nano Lett. 2018, 18, 6633−6637.

(99) Xiong, B.; Huang, Z.; Zou, H.; Qiao, C.; He, Y.; Yeung, E. S. Single Plasmonic Nanosprings for Visualizing Reactive-Oxygen-Species-Activated Localized Mechanical Force Transduction in Live Cells. ACS Nano 2017, 11, 541−548.

(100) Sisamakis, E.; Valeri, A.; Kalinin, S.; Rothwell, P. J.; Seidel, C. A. In Single Molecule Tools, Part B:Super-Resolution, Particle Tracking, Multiparameter, and Force Based Methods; Walter, N. G., Ed.; Methods in Enzymology; Academic Press, 2010; Vol. 475, pp 455− 514.

(101) Hiroi, T.; Shibayama, M. Dynamic Light Scattering Micro-scope: Accessing Opaque Samples with High Spatial Resolution. Opt. Express 2013, 21, 20260−20267.