Review of Advances in Metal-Enhanced Fluorescence

Abstract

In this chapter, we review principles of metal-enhanced fluorescence (MEF), including critical experimental support for the Unified Plasmon-Fluorophore description for the mechanism of MEF. By this description, two routes of enhancement exist for a fluorophore coupled with a metal nanoparticle: namely enhanced absorption and enhanced emission. Literature reports included in this chapter describe the characteristics of a coupled system that influence the efficiency of MEF, including: nanoparticle morphology, distance dependence, the excitation volume effect (EVE) to name but just a few. Reported changes in the photophysical properties of free-space fluorophores, namely improved quantum yields and photostability, for these optimized systems establish MEF as a highly competitive technology for multifarious applications. Subsequently, various applications for MEF systems are highlighted, including MEF-based immunodiagnostics, bioluminescence assays, and the potential for MEF in photodynamic therapy. MEF can also find utility in the development of fluorescence-based electronics as a substitute for potentially toxic quantum dot technologies. Herein we include an effective overview of its principles and a glimpse into prospective advantages of MEF in application.

1 Principles of Metal-Enhanced Fluorescence (MEF)

Metal-enhanced fluorescence (MEF) is a photophysical process through which radiative technologies can gain distinct advantages in application. Luminescent agents are applied ubiquitously, finding incorporation in fields ranging from light emitting diode (LED) electronics to biomedical imaging, clinical therapies, and diagnostics [1,2,3,4]. Due to limitations in photostability, chemical reactivity, toxicity, quantum yields and blinking, however, continued research is necessary to refine current materials for superior technologies and assays [5,6,7]. Metal-enhanced fluorescence provides a route for such tuning, as MEF from traditional fluorophores is hallmarked by an amplified emission intensity, increased quantum yields, and improved photostability [8,9,10,11]. Because of this potential, the principles of MEF have received particular attention in the literature, with an emphasis on determining the mechanism by which this process occurs and, subsequently, how these systems may be optimized for practical use [12,13,14,15,16,17]. The key concepts outlined here for metal-enhanced fluorescence have also been explored for other emissive routes including phosphorescence, alpha fluorescence (delayed fluorescence), and for quenching pathways such as the generation of reactive oxygen species [18,19,20,21,22].

The photophysical underpinnings of MEF lie in the properties of the metals themselves, particularly in their response to the external application of electromagnetic radiation. Due to high free electron densities (plasmons) and nano-scale sizes of metal nanoparticles, a dipole is created as electrons within the metal outer layer oscillate at comparable frequencies to an applied wavelength [23]. This effect is particularly pronounced for wavelengths which overlap with the localized surface plasmon resonance (LSPR) band, which results in the generation of evanescent waves and a local electric field about the nanoparticle structure [9, 24, 25]. This is defined as the near-field, with a finite, or less than one wavelength of light, distance range where a fluorophore may feel its effect. As discussed in later sections, this fluorophore-nanoparticle coupling serves to increase the effective absorption cross-section of fluorophores in the metal near-field as compared to fluorophores in far-field conditions, thereby improving excitation parameters. The effect of enhanced absorption is paired with an alternate route of amplification: enhanced emission. For this effect to occur, excited surface plasmons couple fluorophore quanta and radiate as a coupled unit. As detailed in later sections, surface plasmon coupled fluorescence (SPCF), which is a related technology to MEF, results in a decreased lifetime of radiative emission and, subsequently, improved photostability and superior quantum yields as compared to free space emission [8,9,10,11]. Enhanced emission and absorption mechanisms combine to present the Unified Plasmon-Fluorophore description, depicted graphically in Fig. 1a. As a result, researchers can adjust parameters to optimize the MEF system for varied applications, as illustrated in Fig. 1b, c [26, 27].

Fig. 1

The unified plasmon-fluorophore description of metal-enhanced fluorescence (MEF) can be applied for a variety of metals, fluorophores, and substrates. a Graphical depiction of a coupled plasmon-fluorophore system for MEF. b MEF from chromophore deposited on a glass slide compared to a chromophore on silver island films (SiFs). Modified from Ref. [74]. c MEF achieved both in the ultraviolet (UV) and visible (vis) excitation ranges using different metals and substrates. Modified from Ref. [68]

1.1 Enhanced Absorption from Metal Nanoparticles Is Characteristic of MEF Systems

Fluorophore absorption cross section is a key factor in determining intensities of a luminescent system, and describes the probability that an absorption event will occur for a given fluorophore. In the far field, the likelihood of interaction between molecule and photon is significantly reduced relative to the probability of near-field absorption. Absorption cross sections are a function of molar absorptivity of the attenuating species, as described in Eq. 1, where σ is the absorption cross section, ε is the molar extinction coefficient, and N_A is Avogadro’s number [28].

$$\sigma = \frac{2.303\varepsilon }{{N_{A} }}$$

(1)

In a MEF system, however, the near-field area is extended through the generation of surface plasmons upon excitation. When the nanoparticles absorb incident light, local, high intensity electric fields are generated around the particle. Fluorophores within the near-field therefore experience a larger effective absorption cross section due to coupling with a much larger metal nanoparticle, which significantly exceeds the physical size of the fluorophore itself [29]. This results in enhanced absorption for the MEF system as compared to the fluorophore alone, as depicted for anthracene in Fig. 2 [10, 30, 31].

Fig. 2

The effective cross section of anthracene is increased for a system coupled to silver island films (SiFs). a Absorption spectra for anthracene on quartz versus anthracene coupled to SiFs on quartz substrate, demonstrating enhanced absorption. b Luminescence intensity of anthracene on quartz versus the coupled system, demonstrating metal-enhanced fluorescence (MEF). Modified from Ref. [30]

Extensive investigation has been conducted to characterize this effect for a variety of metals. The finite-difference time domain (FDTD) method is a preferred computational strategy for modeling the generation and distributions of electric fields in simple systems. Using this method, a distinct wavelength-dependence of e-field generation can be observed for a range of metals. Although silver and gold are commonly used in MEF technologies, these metals have been found to be primarily visible (vis) and near-infrared (NIR) absorbers and couplers, as shown in Fig. 3 [32,33,34,35]. This is helpful if the fluorophore probe of interest experiences an excitation event with vis-NIR incident light; however, many probes are excitable by higher energy, ultraviolet (UV) light. For these probes, alternative metals are necessary, and have been initially investigated using the FDTD method as shown in Fig. 3a [8,9,10, 31, 36]. Aluminum, zinc, and indium have emerged as possible metals for MEF systems in the UV, reaching maximum e-field values with UV exposure comparable to those for silver and gold in the vis-NIR range. It is important to note also that within these ranges, each metal will generate maximum e-fields at different wavelengths, highlighted for tin and zinc in Fig. 3c and shown by peak maximums for each metal in Fig. 3a, b. This permits further tuning of the system for maximum enhanced absorption effects depending on the fluorophore of interest.

Fig. 3

Different metal nanoparticles generate electric fields in varying spectral ranges. a Electric field intensity as a function of wavelength for varying metal nanoparticles in the ultraviolet (UV) spectral region. Modified from Refs. [8,9,10, 31, 36]. b Electric field intensity as a function of wavelength for varying metal nanoparticles in the visible (vis) and near infrared (NIR) region. c Finite-difference time domain (FDTD) simulations for tin and zinc nanoparticles, modeled at excitation wavelength where the electric field reaches its maximum intensity. Modified from Refs. [8, 11]

Alternate metals have also been explored for more practical considerations. Silver, with a relatively simple nanoparticle synthesis, is advantageous for proof-of-concept studies; however, cost concerns will remain a limiting factor in pervasive incorporation into common technologies. This is also true for gold nanostructures. Copper, therefore, has been explored as a more cost-effective alternative and exhibits e-field generation in the vis-NIR region [25, 37]. In fact, copper exhibits e-field characteristics that are highly similar to those of gold, as shown in Fig. 3b. Palladium, platinum, germanium, and nickel have also been investigated by the FDTD method and are shown to generate electric fields with vis-NIR incident light, although to a lesser degree than copper, gold, and silver [11, 38, 39]. For UV-excited systems, rhodium was explored due to properties of high thermal and chemical stability, and subsequently the ability of the system to be autoclaved. Unlike other metals, it was observed that the MEF effects generated by rhodium nanoparticles held constant before and after the metal-deposited slide was autoclaved, proving the utility of these particles for re-usable assays, for instance those which are re-used after autoclaving [10]. The diversity of properties between these metals allows for creative development of MEF systems to meet a variety of application requirements.

Using the FDTD method, optimum spacer lengths for the fluorophore-metal system can also be predicted. It is well known that direct contact between a metal and a fluorophore will frequently result in luminescence quenching due to charge transfer, a principle frequently applied in Förster Resonance Energy Transfer (FRET) and inner filtering effects (IFE) [40, 41]; however, if a fluorophore is placed too distal from the nanoparticle, the metal near-field effect will no longer be a factor in fluorophore absorption. In many studies, this translates to enhancement from only a portion of a fluorophore solution over a metal film [8, 11, 25, 42]. By modeling the electric field in two dimensions as a function of distance from the nanoparticle, spacer thicknesses can be established for experimentation. In many FDTD simulations reported, 100 nm spherical nanoparticles are investigated, although additional studies have reported the e-field effects of smaller particles and particle arrays [10, 31]. Figure 4a displays the distance-dependence of electric field intensity, where signal decay is observed with loss of intensity by approximately 20 nm from the nanoparticle surface [43].

Fig. 4

Enhanced absorption from the generation of a nanoparticle electric field is distance dependent. a Electric field intensity plotted as a function of distance from the nanoparticle surface displaying a significant loss of field intensity at 20 nm. Inset displays the 2D finite-difference time domain (FDTD) model for a 100 nm particle. b Left: overlay of electric field intensities from FDTD simulation and normalized fluorescent DNA label intensity measured at incremental distances. Right: 2D FDTD simulation for the silver nanoparticle. c Experimental schematic displaying the strategy for obtaining the data reported in b using DNA scaffolds. Modified from Ref. [43]

The validity of these simulations has also been explored experimentally, as displayed in Fig. 4b, c. In one study by Dragan et al., the authors performed a FDTD simulation and plotted near-field intensities as a function of distance from the nanoparticles. They then created DNA scaffolds to place the fluorophore of interest at incremental distances from the silver nanoparticle surface. The resulting normalized fluorescence intensities were plotted as a function of these distances, revealing a tight correlation between the FDTD e-field predictions and the resulting experimental emission intensities [43]. Mishra et al. also explored the distance dependence of different emissive pathways from the same chromophore. As shown in Fig. 5, the group investigated the enhancement of fluorescence, alpha fluorescence, and phosphorescence from eosin using silicon dioxide as a spacer between chromophore and silver nanoparticles. For each of these processes, the enhancement factor for the system reached a peak value at the predicted maximum e-field intensity of the silicon-coated particle, verifying enhanced absorption as a likely mechanism for enhancement in these MEF systems. Interestingly, the emission enhancement factor profiles are all a similar shape, as shown in Fig. 5b. This is thought to be due to the respective emissions originating from an enhanced singlet state [18].

Fig. 5

Enhanced absorption generates amplified emission intensities as a function of distance from nanoparticle for different states within the same chromophore. a Emission intensities for chromophores at the surface of SiO₂-coated silver particles of varying shell thickness. Left: Fluorescence spectra. Right: alpha-fluorescence and phosphorescence spectra. b Plot of enhancement factor for each state as a function of shell thickness. c Electric field intensities from finite-difference time domain (FDTD) simulations for silver and silver-shell particles. Modified from Ref. [18]

Enhanced absorption can also be influenced by incident laser power, which is a phenomenon termed the MEF excitation volume effect (MEF-EVE) and is illustrated in Fig. 6a [16]. 2D FDTD simulations were conducted by Dragan et al. for silver particles to model the power dependence of the nanoparticle near-field, as shown in Fig. 6b. From these simulations, it seems clear that the electric field volume increases non-linearly with increasing laser power, placing more fluorophore in the near-field and thus enhancing overall absorption and subsequently emission intensities. In fact, the group found these effects to increase rapidly before reaching saturation at about 50 mW excitation power for each of seven silver island film samples collected at varying deposition times. From the model, it was predicted that enhancement factor for an experimental trial would also exhibit a non-linear correlation with excitation power. Using a whole blood assay and the IR-780 dye, this prediction was experimentally verified. As depicted in Fig. 6c, fluorescence intensity increased for the whole blood assay on silver island films as compared to the buffered glass control. This enhancement factor is plotted also and displays a non-linear increase that becomes asymptotic as the laser power approaches 50 mW [16]. Lifetime analysis by the time-correlated single photon counting (TCSPC) method has also confirmed the hypothesis of an enhanced near-field volume, as shown in Fig. 6d. MEF-EVE has also been observed for metals such as copper and gold, illustrated in Fig. 7. That this effect has been observed in multiple MEF systems is significant, as it provides a route for improved assay sensitivity with a simple adjustment of excitation power, with no other modifications to the system. Finally, the significance of the MEF-EVE effect cannot be overstated. In classical far-field fluorescence, altering the excitation power only changes the excitation probability with the excitation volume remaining constant. This is in stark contrast to what is observed in the near-field.

Fig. 6

Enhanced absorption and subsequent metal-enhanced fluorescence (MEF) are dependent on excitation power, a phenomenon termed the MEF excitation volume effect (EVE). a Illustration of expanding electric field volume as a function of excitation power. b 2D finite-difference time domain (FDTD) simulation of a nanoparticle under increasing laser power. c Results of a whole blood assay experiment demonstrating MEF EVE. Left: Fluorescence intensities as a function of laser power. Right: Enhancement factors as a function of laser power. d Time correlated single photon counting (TCSPC) emission decay curves for a chromophore at increasing excitation powers. Modified from Refs. [16, 46]

Fig. 7

The metal-enhanced fluorescence excitation volume effect (MEF EVE) is observed for various metals. a MEF enhancement factor for a copper-based system as a function of copper thickness and laser power. b MEF enhancement factor for a gold-based system as a function of gold thickness and laser power

The results summarized above underscore how MEF systems can be a platform for highly sensitive assays. This concept is depicted by the graphic in Fig. 8, which demonstrates the possible utility of MEF systems in both DNA hybridization assays and immunoassays. By designing systems which place fluorescent-labeled species of interest near-to metal structures, detection limits can be significantly decreased.

Fig. 8

Graphical representation of possible diagnostic uses of metal-enhanced fluorescence (MEF) technologies

1.2 Enhanced Intensities from Surface Plasmon-Coupled Emission (SPCE) Is Characteristic of MEF Systems

In addition to the enhanced absorption effect, a phenomenon also occurs in MEF systems whereby chromophore quanta will radiate through the surface plasmons of the nearby metal. This results in radiation of the coupled quanta, which carries distinct characteristics. This phenomenon has been named surface plasmon coupled emission (SPCE) or fluorescence (SPCF) [12], the principles of which extend to MEF for dual-mechanism enhancement, although SPCF deals with thin continuous films. Detection of the coupled emission requires specific parameters, as a fluorophore in solution will also emit uncoupled, or free state, luminescence. To differentiate the coupled emission from free space emission, a reverse Kretschmann configuration is used, an example of which is shown in Fig. 9a [9, 24]. In this configuration, excitation occurs directly to the analyte, which is coated on the metal. Detection occurs on the reverse side of the metal layer such that only coupled quanta is analyzed. This results in a circular fluorescence pattern, as only emission light that angle matches with the surface plasmons will couple constructively into the system.

Fig. 9

Fluorescence emission can couple to surface plasmons for metal-enhanced fluorescence (MEF). a Schematic for detection of surface plasmon coupled emission (SPCE) using a reverse Kretschmann configuration of a 5-phase system. b Plot of Fresnel calculations for reflectivity from p- and s-polarized light versus fluorescence intensity of fluorescein isothiocyanate (FITC) in various dilutions of PVA. c Schematic displaying the evolving description of fluorophore-nanoparticle interactions. Modified from Ref. [24]

The generation of coupled quanta is maximized when there is a wavevector match between incident light ( ) and the surface plasmon ( ), which can be described by Eq. 2, where $n_{p}$ is prism refractive index and $\theta_{sp}$ is the surface plasmon angle [24].

(2)

To achieve the highest generation of surface plasmon modes, therefore, incident light must be applied at the surface plasmon angle that minimizes reflectivity of fluorescence by the metal. Fresnel calculations are employed in numerous studies to determine the optimal conditions for SPCE, as displayed in Fig. 9b [9, 24]. These data, collected by Aslan et al., demonstrate an inverse relationship between fluorescence intensity from fluorescein isothiocyanate (FITC) in polyvinyl alcohol (PVA) and normalized reflectivity for p-polarized light as the incident angle is changed [24]. A similar trend is not observed for s-polarized light, as a mirror dipole at the metal surface cancels the incident light, preventing the generation of a surface plasmon [24, 44]. This observation of angle-dependent emission intensity from a fluorophore-nanoparticle system supports the plasmon coupling description of MEF. Previously, it was thought that nanoparticles acted predominately as quenchers for fluorescence emission. This conclusion, however, was based in studies where fluorophores were placed in close range to very small nanoparticles, as depicted in Fig. 9c (left). Because MEF is dependent both on optimal fluorophore-nanoparticle distance and larger particle size, these earlier studies did not demonstrate MEF. The use of larger particles and distances in the range of 10 nm revealed the MEF effect [45]; however, early mechanisms of MEF discuss only resonance interaction between fluorophore and nanoparticle (Fig. 9c, middle). This description alone does not account for the observed coupled emission signals, suggesting that plasmons in fact radiate coupled quanta. This coupling is now the foundation for the mechanistic description of MEF (Fig. 9c, right). The coupled system experiences a decrease in luminescent lifetime that has been reported in numerous studies of MEF systems, which will be discussed in more detail in later sections [8, 9, 25, 46].

Reports have shown that the degree to which a luminescence signal is enhanced by coupling with surface plasmons is wavelength dependent. This due primarily to the propensity of a metal particle to either absorb or scatter light, which is highly dependent on particle size and can be modeled by Mie calculations [11, 31]. That the metal extinction $(C_{E} )$ is comprised of both a scattering $(C_{S} )$ and absorption $(C_{A} )$ portion is illustrated by the equation below [8].

$$C_{E} = C_{A} + C_{S}$$

(3)

Surface plasmon resonance is achieved when incident light induces migration of electrons within a particle, forming an oscillating dipole. This effect is highly dependent on the size of the particle, therefore anisotropic particles such as nanorods can have multiple LSPR bands within a single structure. Numerous studies have explored the resulting MEF effects from different anisotropic particles, including nanostars and nanorods as well as bimetallic structures, and have found the increased complexity conducive for strong MEF effects [32, 33, 47, 48]. The breadth of LSPR peaks within these anisotropic structures provides more opportunity for coupling and subsequent enhanced emission.

The principle of wavelength-dependent coupled enhancement, however, is clearly modeled in simpler systems. In one study conducted by Zhang et al., the emission spectrum of Prodan was increasingly red shifted through suspension in solvents of varying polarity [49]. As shown by Fig. 10, MEF was observed most strongly for the bathochromic emission spectra. The authors attribute this observation to spectral overlap with different portions of the metal extinction spectrum. The scattering (C_s) and absorption (C_A) spectra for the silver nanoparticles used in this study are shown in Fig. 10c. As the emission spectrum of Prodan is red shifted, there is a greater exclusivity in spectral overlap with the scattering portion. At these wavelengths incident light is primarily scattered—or radiated in MEF systems—by the metal rather than being re-absorbed. By shifting the emission wavelengths away from the absorption component spectrum, the likelihood of inner filtering effects diminishes, and energy is primarily released as luminescent radiation [47, 49]. This is clarified further by comparing the MEF enhancement factor with the overall extinction spectrum in Fig. 10c. It seems from these data that there is no direct correlation between enhancement and extinction; however, after isolating each extinction component, a trend is observed between the scattering component and enhancement factor. In fact, wavelengths where the absorption spectrum is present demonstrate the lowest enhancement factors, lending validity to the assertion that wavelength dependence of enhancement from coupled emission may be in part reliant upon the scattering characteristics of the metal. In some cases of spectral overlap, however, small spectral distortions of the enhanced spectra have been observed [14, 50]. Although this effect has not been intensively studied to date, this observation may alter how the mechanism of MEF is interpreted in future studies [14].

Fig. 10

The scattering portion of the metal extinction spectrum facilitates metal-enhanced fluorescence (MEF). a Normalized emission spectra for Prodan in varying polarity solvents. b Photograph of results shown in a. c Demonstration of correlation between enhancement factor and scattering portion. Top: Extinction spectrum plotted against enhancement factor, displaying no clear correlation. Bottom: Absorption and scattering components of extinction spectrum plotted against enhancement factor. Modified from Ref. [49]

Wavelength dependence can also be ascertained through analysis of the synchronous scattering spectrum, or the spectrum generated when the wavelength of excitation is set equal to the emission wavelength that is being collected. This technique, investigated in relation to MEF by Dragan et al., also gives insight into the magnitude of MEF enhancement factors for the system studied. As shown in Fig. 11a, when excitation wavelength is held constant and MEF enhancement is measured at variable wavelengths, the enhancement spectrum closely mimics the structure of the synchronous spectrum. Enhancement factors can also be predicted when a constant emission is detected at variable excitation wavelengths, as shown in Fig. 11b. Thus, by looking only at the synchronous scattering spectrum, one could predict whether the MEF effect would be maximized for a desired emission value, or at the desired emission wavelength [51].

Fig. 11

Synchronous spectra are an accurate indicator of metal-enhanced fluorescence (MEF) effects for a coupled system. a Synchronous spectrum (blue) plotted against enhancement factor when excitation remained constant and the emission was collected at varying wavelengths (red). b Synchronous spectrum (blue) plotted against enhancement factor at a fixed emission wavelength after excitation at varying wavelengths (red). Modified from Ref. [51]

1.3 Principles of Enhanced Absorption and Emission Intersect to Form a Unified Description of MEF

Each previously described consideration is a key element in the mechanistic understanding of metal-enhanced fluorescence and can be summarized by the Unified Plasmon-Fluorophore Description. Under this model there are two modes of enhancement, as enumerated previously. Enhanced absorption is the result of increased near-field effects from the metal nanoparticle, which effectively increases the absorption cross section of fluorophores near-to the metal. Metal type, size, and spacing will dictate the degree to which the particles can act as an excitation “antennae” for fluorophores at varying wavelengths. Metals can be selected for spectral range in addition to more practical considerations such as cost, thermal stability, or ease of synthesis. Enhanced emission arises from the ability of fluorophore emission quanta to couple with surface plasmons of proximal nanoparticles. Enhancement from this strategy is largely dependent on metal particle size and subsequently the scattering portion of the extinction spectrum, as it is the scattering or radiation of the coupled quanta which imparts favorable lifetime, photostability, and quantum yields to the system. To understand these advantages, the equations for lifetime can be compared for free space emission $(\tau_{FS} )$ and MEF $(\tau_{MEF} )$.

$$\tau_{FS} = \frac{1}{{\Gamma + k_{nr} }}$$

(4)

$$\tau_{MEF} = \frac{1}{{\Gamma +\Gamma _{m} + k_{nr} }}$$

(5)

In these equations, $\Gamma$ is the radiative decay rate for system that remains uncoupled, $\Gamma _{m}$ is the radiative decay rate for the metal-coupled system, and k_nr is the cumulative non-radiative decay rate [11]. From these equations, it is mathematically predicted that a decreased lifetime should be observed for MEF systems, which has been experimentally verified on numerous occasions [8, 9, 25]. Understanding this principle, the equations for quantum yield in free space $(\Phi _{FS} )$ and for the coupled system $(\Phi _{MEF} )$ can also be examined.

$$\Phi _{FS} = \frac{\Gamma }{{\Gamma + k_{nr} }}$$

(6)

$$\Phi _{MEF} = \frac{{\Gamma +\Gamma _{m} }}{{\Gamma +\Gamma _{m} + k_{nr} }}$$

(7)

Since $\Gamma _{m}$ is a greater contributor in the numerator for the MEF system, an enhanced quantum yield is predicted. Although this effect has been shown experimentally, it has been noted that the enhancement factor decreases for fluorophores with high free space quantum yield. This can also be explained mathematically using Eqs. 6 and 7. For high efficiency yields, $\Phi _{FS}$ will approach 1, which indicates that the rate of radiative decay far exceeds the rate of alternate decay pathways, and Eq. 7 can be simplified to Eq. 8.

$$\Phi \approx \frac{{\Gamma +\Gamma _{m} }}{{\Gamma +\Gamma _{m} }}\rightarrow1$$

(8)

In this model, addition of the coupled decay rate has negligible impact on quantum yield, which translates to the reduced enhancement factors observed experimentally. The implications of these results are two-fold. Firstly, that a system can be verified as coupled if a reduced lifetime is measured, confirming the unified plasmon-fluorophore description. Secondly, chromophores can be selected for applications based on quantum yields for sensitive systems. Optimal MEF effects can also be achieved by manipulating the metal size and shape or fluorophore emission properties to achieve preferential overlap of emission and metal scattering component. This can be modeled mathematically, using Mie calculations, or experimentally by taking synchronous measurements of the metal system.

2 Applications of Metal-Enhanced Fluorescence

Luminescent agents have found ubiquitous use across fields, with applications ranging from electronic technologies to biomedical diagnostics [52,53,54,55]. The potential impact of traditional organic fluorophores in these areas is limited, however, due to low quantum yields and poor photostability. In diagnostics, low quantum yields reduce the sensitivity of luminescence-based detection; for technologies such as light emitting diode (LED) displays, such low-yield fluorophores require more excitation power to achieve sufficient brightness. In both cases, degradation of the fluorophore under light reduces the long-term utility of a fluorescent probe. Alternatives to organic fluorophores, such as inorganic quantum dots, have been developed to overcome these limitations [5, 6, 56]. Many of these options, however, contain toxic materials that raise long-term environmental and health concerns [57]. Metal-enhanced fluorescence is therefore a competitive alternate strategy for optimizing the characteristics of luminescent agents for future applications.

2.1 MEF from Silver-Coated Luminescent Nanostructures Diversifies Potential Applications

Fluorescent particles are highly desirable for use in biological sensing, and indeed much research has been done to achieve monodisperse, modifiable particles with high quantum yields and photostability to be used for this purpose [3, 4]. Many studies that explore the generation of these particles use silica nanoparticles as the core with fluorophores doped into these structures [53, 58]; however, replacing this system with a silver-silica nanocomposite has been shown to provide advantageous MEF characteristics to the nanoparticles. Aslan et al. investigated this strategy, creating a nanocomposite with a silver nanoparticle core coated in a silica shell, as shown in Fig. 12. This shell was then doped or covalently modified with fluorophores; in the case of this study, Tris(dibenzoylmethane) mono(5-amino phenanthroline) europium (Eu-TDPA) or Rhodamine 800 (Rh800) were used. An etched structure, or “nanobubble”, was used as a control to determine enhancement factors and featured the fluorophore functionalized shell without the enhancing silver core [53].

Fig. 12

Metal-enhanced fluorescence (MEF) can be achieved for silver-silica-fluorophore nanohybrids. a Top: Graphical depiction of the experimental strategy whereby a silver core is then etched for a fluorescent nanobubble control. Bottom: Transmission electron microscopy (TEM) images of the nanobubble (left) and silver core (right) structures. b Left: Absorption spectra for both structures demonstrating enhanced absorption. Right: Fluorescence spectra for both structures demonstrating enhanced emission. c Fluorescence spectra and graphical representation of silver core Rh800 structures demonstrating enhanced emission. d Lifetime decay curves collected by the time correlated single photon counting (TCSPC) method for the Rh800 structures. Modified from Ref. [53]

The utility of this strategy is rooted both in the ability of silver nanoparticles of certain sizes to produce a MEF effect and in the properties of the silica shell. As shown by the transmission electron microscopy (TEM) images in Fig. 12a, the silver particles used in this study were 130 nm in diameter with a 11 nm silica shell. The significance of this can also be visualized in Fig. 12b. Across the entire spectrum, the silver core structure displays a higher absorbance than the control nanobubble. In particular, an otherwise unobserved peak arises from the silver core structure at 400 nm, corresponding to the surface plasmon of the silver nanoparticles. This increased absorbance is mirrored by the 10-fold enhancement of the luminescence signal; a corresponding 10-fold drop in fluorescence lifetime confirms that the silver core provides a key MEF effect. This observation is highly dependent on the thickness of the silica shell, as MEF is known to be a near-field phenomenon. Thus, by optimizing the size parameters of these particles, a silver/silica nanocomposite can be made with an approximately 100-fold increased detectability over Eu-TDPA silica nanostructures. The versatility of this strategy is also reinforced through the incorporation of a second fluorophore, Rh800, as shown in Fig. 12c, d. Here once again there is an observable MEF effect in the fluorescence intensity which corresponds to a reduction of fluorophore lifetime [53].

Other dye-doped silicon oxide structures have been investigated, varying both fluorophore and metal for improved use in application [59,60,61]. Through the generation of silver/silica nanocomposites that can be either doped or covalently modified with traditional fluorophores, a broad library of fluorophores could be considered for applications where their quantum yields may have been previously limiting. As displayed by Fig. 13, this could provide an alternative to quantum dot technology, which is called to question for use in biological media due to possible toxicity [56]. Despite limitations, the MEF from quantum dots has also been explored. Luminescence from these inorganic MEF hybrid structures have seen improved brightness, reduced lifetimes, and superior stability comparable to those effects observed for organic chromophores [62]. These MEF-quantum dot hybrids also provide an intriguing route for improved application, including detection technologies [63].

Fig. 13

Metal-enhanced fluorescence (MEF) from silver core structures present an alternative to quantum dot technology. a Transmission electron microscopy (TEM) image of silver core/ultraviolet (UV) probe shell structures. b Photographs of structures from a in solution under UV excitation

Carbon nanodots have also been intensely investigated as organic quantum dot alternatives. These quasi-spherical structures are generally less than 10 nm in size and are frequently reported to emit blue to green fluorescence, which has restricted their application in biological media due to overlapping autofluorescence [64]. Quantum yields of these fluorophores are also typically low, limiting their utility [65]. Nonetheless, carbon nanodots have been investigated for use in sensing platforms, cellular imaging, and as light emitting diodes [64, 66, 67]. Utilization of MEF in these carbon nanodot applications can expand their potential uses. In a study conducted by Schmitz et al., MEF was observed from carbon nanodots on SiFs at multiple excitation wavelengths. This is due to broad absorption of carbon nanodots, which could prove advantageous for versatility in assay development [65]. More recent studies have attempted to achieve metal-enhanced fluorescence from carbon nanodots through the formation of nanohyrids [64, 66, 67]. In one study conducted by Tian et al., carbon nanodots were coated with silver to generate a hybrid structure on the order of 5 nm. Prior to coating, the carbon nanodots are reported to exhibit a quantum yield of ~1% at 500 nm excitation. While carbon nanodots can be excited at vis-NIR wavelengths for use in biological media, this quantum yield is not sufficient for sensitive assays or imaging. The hybrid, however, displays a broad LSPR range across more bathochromic wavelengths as compared to carbon nanodots. As such, the hybrid structure was able to achieve detectable fluorescence when excited up to 620 nm. The authors also report superior photostability, a hallmark of the MEF effect [64].

Hybrid structures of carbon nanodots were also examined in one study conducted by Lin et al. In this report, silver nanoparticles and carbon nanodots were mixed in solution for a MEF effect that could be applied to LED technology. Although the authors report no detectable covalent or electrostatic interaction between the carbon nanodots and the silver structures, they nonetheless report fluorescence enhancement for these solutions [66]. This is possibly due to diffusion of carbon nanodots into the nanoparticle near-field at the time of excitation, although further studies would be required to confirm this theory. The reported enhancement is concentration dependent. Similar to previously discussed studies [47], the authors see an initial increase followed by a sharp decrease in enhancement as concentration is steadily increased [66]. This could be due either to an increase in dynamic quenching events as concentrations are levied or due to inner filtering effects of the nanoparticles, to name two possibilities. Bound hybrid structures have also been investigated for fluorescence tuning by this same method, as reported by Zhang et al. For these studies, carbon nanodots were coupled with gold particles for yellow luminescence, although in this case characterization data confirmed adsorption of the carbon nanodots onto the metal structures. Enhancement was similarly observed for this system as a function of concentration. Lifetimes of the coupled systems were also reduced, implicating coupled emission as a mechanism of enhancement consistent with MEF [67]. Although the creation of metal-carbon nanodot hybrids for metal-enhanced fluorescence is a relatively new area of study, it provides a promising platform for future incorporation of carbon nanodots into mainstream fluorescence applications.

2.2 Metal-Enhanced Systems on Plastic Substrates Yield Sensitive Assays for Biomedical Applications

Investigations into the mechanisms and properties of metal-enhanced fluorescence have largely been conducted on silica-based substrates such as glass or quartz. These materials are advantageous for reproducibility and experimental design due to well-known chemical characteristics and predictable photophysical properties; however, these substrates can be expensive and difficult to make, and are therefore not practical for ubiquitous use in biomedical applications. To address this limitation, Aslan et al. investigated the generation of MEF using plastic substrates. Although there are a wide variety of polymeric substrates available for the design of plastic scaffolds, polycarbonate is commonly used in high throughput assays and was therefore a focus of this study. The authors also report investigation of polypropylene films; however, due to lack of surface functionalization this substrate was not suitable for silver deposition [68].

To optimize surface functionalization for silver deposition, the authors first treat the polycarbonate (PC) films through base-catalyzed hydrolysis and subsequent amino-coating. This places amine groups at the plastic surface, which have a high affinity for silver. Silver island films (SiFs) were then deposited on the treated plastic, and a slightly red-shifted plasmon absorbance band was reported. FITC-labeled human serum albumin (HSA) was used as the fluorescent probe to ensure a mean distance of 4 nm from the SiF surface, which is in the optimal range for MEF. As shown in Fig. 14a, FITC-HSA demonstrated enhanced fluorescence intensity on the silver-coated polycarbonate film. The authors also report decreased lifetime for the enhanced signal, consistent with previous reports of MEF on silica substrates [68]. The results of this study provide the foundation for MEF from plastic substrates in future applications, opening the diverse properties of different plastic materials to be investigated. Techniques for plastics development also allow for facile creation of customizable devices to suit application needs, making this technology even more powerful for widespread use.

Fig. 14

Metal-enhanced fluorescence (MEF) can be implemented in 96-well plates for biomedical applications. a Absorption spectra for FITC-HSA MEF experiment demonstrating enhanced absorption on plastic substrates. b Graphical representation of a typical MEF experiment on 96-well plates. Inset: Photographs of fluorescein in wells. c Photograph of silver island films (SiFs) in disposable wells, now commercially available as “Quanta Plates.” D Fluorescence as a function of ctDNA concentrations for SiFs compared to glass wells. Arrows indicate limits of detection. Modified from Refs. [68,69,70]

Studies have been conducted to further establish MEF from plastic substrates as a useful tool in biomedical applications ranging from drug discovery to diagnostics. The 96-well plate, depicted in Fig. 14b, is a common tool employed in high throughput screening (HTS) assays. In research also performed by Aslan et al., a proof-of-concept study was conducted to support the use of silver-coated well plates in HTS. In this study, the microwave assisted metal-enhanced fluorescence (MAMEF) technique was employed [68]. This technique combines the advantages of increased sensitivity from MEF with the improved agent diffusion kinetics generated by microwave-induced ohmic heating, which results in a highly-sensitive, ultra-fast assay [69]. The authors functionalize the SiFs with biotinylated bovine serum albumin (BSA) and use FITC-labeled Avidin to model agent affinities in HTS analysis. They report an observed MEF effect, which is photographically represented in the inset of Fig. 14b [68]. This technology could also be easily applied to disposable well plates, as shown in Fig. 14c.

Enhancement for HTS in biochemical assays results in a decreased detection limit and therefore improved assay sensitivity, which can be applied also to DNA detectability. In a study by Dragan et al., this principle was investigated also using silver-coated 96-well plates, which have now been commercialized through Ursa BioScience under the tradename, “Quanta Plates.” Prior to analysis, PicoGreen^® was bound to polymeric DNA strands [70]. A similar strategy was employed previously with DNA chains of 16 bp, and a resulting 5-fold decrease in fluorescence intensity was reported, which can be difficult to detect at low concentrations of DNA. This stands in contrast to the 7-fold enhancement that was reported for the metal-coupled system, demonstrating how in simple systems MEF can greatly improve the ratio of fluorescence from bound fluorophore to free fluorophore [71]. Polymeric DNA strands have more degrees of freedom and are therefore less predictable in conformation than shorter DNA chains, so it was unclear if the MEF effect would still occur given variable metal-fluorophore distances; however, MEF was observed for the systems studied. This greatly improved the limit of detection for ctDNA as compared to glass wells in the study by Aslan et al., as demonstrated in Fig. 14d [70]. This principle could therefore be extended for research requiring DNA quantification.

While MEF is clearly advantageous in fundamental biomedical research, it can also be powerful if applied in diagnostic assays for improved medical care. Infection with Bacillus anthracis, the etiologic agent of anthrax, is a highly lethal condition but is also asymptomatic. This results in late diagnosis and high mortality rates, as the disease will progress quickly if untreated and significant accrual of lethal toxins will occur. For this condition and others which share similarities, faster and more sensitive assays are required for early diagnosis and a rapid clinical response. Current technologies such as FRET assays or enzyme-linked immunoabsorbent assays (ELISAs) often take hours to complete, and are therefore clinically limiting. In contrast, Dragan et al. employ the MAMEF technology for a rapid detection strategy of protective antigen (PA), a biomarker for anthrax. The superior detection limit of the MAMEF assay is demonstrated in Fig. 15a, where pg/ml concentrations of PA83 can be detected with SiFs on plastic as compared to the plastic plate alone. This stands in contrast to the ng/ml detection limit inherent to many other assay techniques. Figure 15b demonstrates the utility of this assay using whole blood samples as compared to the phosphate buffered milk diluent/blocking solution (PBS-Milk). Overall these MAMEF assays were conducted in their entirety in 40 min, significantly improving time to diagnosis compared to commonly utilized techniques [52].

Fig. 15

Metal-enhanced fluorescence (MEF) can be employed in sensitive diagnostic assays. a Fluorescence intensities as a function of protective antigen concentration. b Fluorescence intensities as a function of protective antigen concentration in different assay solutions. Modified from Ref. [56]

Luminescence is also employed heavily in biological research through the use of bioluminescence imaging. Using this technique, researchers can monitor the presence or absence of certain agents in a system as a function of a biological response to these agents. One commonly used protein for this strategy is Firefly Luciferase, depicted graphically in Fig. 16, which emits a fluorescent signal when active. The utility of bioluminescence as a detection strategy is limited, however, due to the relatively low intensities of bioluminescent signals as compared to fluorescent probes. Systems requiring bioluminescence detection, therefore, could benefit from employing a MEF analysis strategy. Eltzov et al. investigated the phenomenon of metal-enhanced bioluminescence using a bacterial model with luciferase responses to metabolic changes. Although only a small portion of luminescence can couple with surface plasmons under this assay design, the authors nevertheless report luminescence enhancement for this assay that is clear from the photographical inset of Fig. 16 [72].

Fig. 16

Graphical representation of a metal-enhanced bioluminescence (MEB) system. Inset: Photographs of luciferase emission for metabolic changes in bacteria on silver island films (SiFs) and on plates. Modified from Ref. [72]

Additional MEF studies for in vitro applications have been explored, including the enhancement of the Fluo-3 fluorescent indicator, which permits detection of calcium for intracellular systems. The ability to monitor calcium concentration is a powerful tool, as calcium operates ubiquitously in many cellular processes and is an indicator of various diseases, including hypoparathyroidism and renal failure. Bondre et al. provide foundational data for this application, reported in Fig. 17, which demonstrates enhanced emission for Fluo-3 on SiFs compared to glass at varying calcium ion concentrations. The overall result is increased signal to noise ratio, which ultimately supports the use of MEF for in vitro ion detection. This strategy can help to overcome low quantum yields of fluorescent indicators and the prevalence of background autofluorescence from biological samples. By surmounting these limitations through MEF, fluorescence ion detection in vitro becomes a more practical technique in biomedical research [73].

Fig. 17

Metal-enhanced fluorescence (MEF) can be applied to calcium ion detection for future use in cellular ion detection assays. a Photographs of Fluo-3 emission as a function of calcium concentration on silver island films (SiFs) and on Glass. b Fluorescence intensity plotted as a function of calcium concentration displaying improved detection power. Modified from Ref. [73]

2.3 Metal-Enhanced Systems Can Be Engineered for the Generation of Reactive Oxygen Species

As previously mentioned, metal nanostructures can also enhance luminescence signals from different states. One radiative pathway of particular interest is phosphorescence. Unlike fluorescence, phosphorescence occurs when there is relaxation between two unequal spin states; for many molecules, this constitutes a transition between a triplet excited and singlet ground state. For this process to happen, an excitation event must first occur to excite the chromophore into its singlet excited state; from here, intersystem crossing will proceed whereby the excited state now has triplet character. Subsequent relaxation yields phosphorescence, although back intersystem crossing can also occur for a long-lived singlet excited to singlet ground emission (alpha fluorescence). Due to the energetic limitations of transitions between unequal states, phosphorescence signals typically display longer lifetimes than their accompanying fluorescent emission.

Researchers have investigated the effect of placing phosphors near-to metal nanostructures. In one study by Zhang et al., Rose Bengal was examined in conjunction with SiFs. The researchers performed MEF analysis at room temperature to ascertain any differences between the observed effect in these conditions, as shown in Fig. 18. Since phosphorescence is a long-lived luminescence process, dissolved oxygen will readily quench its excited state. As such, it was necessary to conduct metal-enhanced phosphorescence (MEP) analysis at low temperatures, where quenching effects are negligible, to observe the phosphorescent signal. The authors observed an increased MEF enhancement factor at lower temperatures, which they explain by general decreased quantum efficiency at higher temperatures. Enhancement was also observed for the phosphorescence signal at low temperatures, indicating that the metal system has an overarching impact on the system as a whole. Reduced lifetimes reported by the group support that both signals couple to the surface plasmons for enhanced emission. These same data reinforce the prediction that MEF is likely due to enhanced absorption at low temperatures rather than back intersystem crossing, as this reverse process would lead to a lengthening of fluorescence lifetimes despite plasmon coupling [19].

Fig. 18

Metal-enhanced phosphorescence (MEP) is observed for Rose Bengal at low temperatures. Left: Enhanced absorption (top) and emission (bottom) for Rose Bengal fluorescence at room temperature. Middle: Enhanced absorption (top) and emission (bottom) for Rose Bengal fluorescence at 77 K. Right: Enhanced absorption (top) and emission (bottom) for Rose Bengal phosphorescence at 77 K. Modified from Ref. [19]

MEP is of particular interest due to its clinical relevance. As previously mentioned, triplet excited states are readily quenched by diffused oxygen. This process occurs when energy is transferred from the triplet excited state to ground state molecular oxygen through a collision event. The chromophore is returned to its singlet ground state and oxygen enters its highly reactive state, namely: singlet oxygen (¹O₂) [19]. This pathway for the generation of reactive oxygen species (ROS) has been employed extensively in photodynamic therapies (PDT), which are used for oncological tumor treatment among other clinical applications [19, 22]. These therapies are generally comprised of three elements: the photosensitizer, ambient oxygen, and an excitation source. In this system, the phosphorescence-emitting chromophore is the photosensitizer. Current strategies for controlling singlet oxygen yields frequently emphasize the excitation source, striking a delicate balance between long irradiation times and possible subsequent hypoxia and high intensities, which can cause photobleaching and therefore inactivation of the photosensitizer. MEP technologies could therefore find utility in this application, as SPCE has been shown to reduce lifetimes and improve photostability.

Subsequent investigation has occurred to confirm that MEP can induce enhanced singlet oxygen yields. Singlet Oxygen Sensor Green^® (Invitrogen, USA) is commonly used in these studies to detect singlet oxygen generation [20, 22]. Invitrogen has reported that Sensor Green is bichromophoric, with anthracene and fluorescein components. This probe, which is highly selective for singlet oxygen, operates through a FRET system whereby fluorescein emission is quenched by anthracene. Following singlet oxygen exposure, an endoperoxide is formed and fluorescein can relax radiatively; this process is irreversible and permits quantification of singlet oxygen production [20]. Although Sensor Green is a fluorophore that will experience MEF thus skewing data, control studies can be conducted to account for this prior to calculating enhanced singlet oxygen yields [22]. In a study performed by Karolin and Geddes, MEP-generated singlet oxygen is measured using Rose Bengal as the photosensitizer. The authors investigate the distance dependence of metal-enhanced singlet oxygen generation through subsequent coatings of silicon oxide over SiFs, as depicted graphically in Fig. 19d. MEF is well-known to be a distance-dependent phenomenon; as such, the authors conducted an FDTD simulation for the experimental nanoparticles, as shown by Fig. 19a, b. The electric field modeled decreases in intensity by 10 nm distance, correlating closely with previous MEF studies. These data were then plotted against singlet oxygen enhancement factors from luminescence intensities, as shown in Fig. 19c. Although there is not an exact correlation, enhancement factor decreases with increased distance comparable to the electric field, indicating that MEP-induced singlet oxygen generation also follows a mechanism of enhanced absorption. Karolin and Geddes also report a quantum yield dependence similar to that of MEF. As shown in Fig. 20, enhancement factors were highest for photosensitizers of lower quantum efficiency [22].

Fig. 19

Metal-enhanced singlet oxygen yields are distance dependent. a Close up of finite-difference time domain (FDTD) simulation from B displaying e-field at experimental distances. b 2D FDTD simulation of a 100 nm silver nanoparticle. c Plot of field enhancement and singlet oxygen enhancement as a function of Rose Bengal distance from metal. d Graphical representation of experimental design. Modified from Ref. [22]

Fig. 20

Plot of corrected metal-enhanced singlet oxygen enhancement factors as a function of far-field quantum yields, displaying quantum yield dependence of enhancement. Modified from Ref. [22]

Photosensitizer selection and nanoparticle-fluorophore distances offer two ways to adjust singlet oxygen generation in PDT by MEP technologies; however, tuning of these systems is not limited to these components. In clinical settings, singlet oxygen yields may be controlled by increasing irradiation time. As shown in Fig. 21a, the effect of this strategy can be demonstrated using Sensor Green. Figure 21b displays how this method is only so effective, and in fact yield intensities becomes asymptotic at longer illumination times. It is known, however, that MEF effects can be heightened with greater excitation power, known as MEF-EVE. This effect has also been observed for MEP-induced singlet oxygen generation, as shown in Fig. 21c [20]. In this study performed by Karolin and Geddes, Rose Bengal was combined with Sensor Green and deposited over silver-coated wells. Following excitation of Rose Bengal, enhanced singlet oxygen yields were observed at each laser power setting; however, yields also increased for MEP samples as a function of increased laser fluency rate [20]. This strategy provides yet another avenue to tune singlet oxygen yields, establishing MEP as a potential technology for photodynamic therapy applications.

Fig. 21

Singlet oxygen generation can be tuned with incident laser power in metal-enhanced phosphorescence (MEP) systems. a Fluorescence emission spectra of Sensor Green as a function of illumination time. b Integrated intensity of Sensor Green as a function of illumination time. c Real-color photograph of Sensor Green detection of singlet oxygen from Rose Bengal phosphorescence on silver island films (SiFs) at different excitation powers. Modified from Ref. [20]

Superoxide generation has similarly been investigated given its utility as a reactive oxygen species, using dihydroethidium (DHE) probe. This probe, when in the presence of superoxide, forms a luminescent cation in an irreversible reaction. As such, DHE can be used to monitor superoxide generation much like Sensor Green is used for singlet oxygen [20]. In a study by Zhang et al., metal-enhanced superoxide generation was studied in this manner, where acridine was used as the photosensitizer. As displayed in Fig. 22, emission from the DHE probe was greatly enhanced when excitation occurred on the silvered substrate [21]. The EVE effect has also been observed for superoxide generation, although this effect is less pronounced than that for singlet oxygen [20]. Overall metal-enhanced systems provide a promising technology for expanded application in biomedical sciences.

Fig. 22

Superoxide yields can be enhanced in coupled metal-photosensitizer systems. a Real-color photographs of dihydroethidium (DHE) probe with acridine on silver island films (SiFs) for superoxide detection. b Absorption spectra for DHE probe demonstrating enhanced superoxide generation following irradiation on SiFs. Modified from Ref. [21]

3 Conclusion

Metal-enhanced fluorescence is an intriguing area of research with continued relevance in the development of new luminescent technologies. Enhancement can be observed from two possible mechanistic pathways. The first, enhanced absorption, is the result of surface plasmon bands observed characteristically for metal nanoparticles. For excitation wavelengths overlapping with these bands, there is generation of evanescent waves which can excite a fluorophore in the near-field. Since this process is extremely rapid, it can be thought of as a simultaneous occurrence, drastically amplifying the absorption cross section for a fluorophore in the near-field and subsequently enhancing absorption. Signal intensity increase can also be the result of enhanced emission, whereby the radiative emission of the fluorophore couples with metal surface plasmons to radiate as a collective unit. Given the additional route of decay, quantum yields are typically reported to increase for this mechanistic pathway while lifetimes decrease. A secondary result of this effect is that fluorophores spend less time in the excited state, improving resistance to photobleaching. MEF effects can also be amplified as a function of excitation power, a phenomenon known as the MEF excitation volume effect. The implication of these MEF characteristics are such that this technology can be applied across a multitude of luminescence-based applications. Some discussed here include improved diagnostics, high throughput screening, LED technologies, and clinical use for example in photodynamic therapies. This, however, is not a comprehensive description of possible MEF applications, a larger scope of which is illustrated by Fig. 23. Overall the characteristics of MEF intersect for a unified description, which will continue to expand as the phenomenon is further characterized and explored in research and application.

Fig. 23

Schematic for the unified description of metal-enhanced fluorescence (MEF), as published by Geddes et al., demonstrating general principles and investigative routes for application. Modified from Ref. [74]