430
A. V. Maslov and V. N. Astratov
sentative SEM (Fig. 16.13a), PCLR (Fig. 16.13b), and white light (WL) (Fig. 16.13c)
images of a Cr/Au double-line sample.
Although PCLR has not been integrated with microsphere-based imaging yet, we
believe that this method is highly promising in this regard. The limited resolution
advancement in PCLR is related to the absence of the magnification of the image in
a near-field zone where the higher spatial frequencies of the object are contained. A
combination of MN with PCLR can allow additional image magnification required
for further resolution improvement.
16.5.3 Microspherical Nanoscopy Based on Short-Period
Arrays with Small Gaps: Super-Resolution Without
Post-imaging Processing
In previous Sects. 16.5.1 and 16.5.2 we considered different situations when the
nanoplasmonic layer adjacent to the objects was designed to collect the objects’s
near-fields that resulted in super-resolution imaging of these objects. In PSIM
(Sect. 16.5.1), the illumination of the objects was provided with short-period standing plasmonic waves. In LPSIM (Sect. 16.5.1), the individual LSPRs in nanodisks
were arranged with the wavelength-scale periodicity to provide illumination with
collective SLRs due to out-of-plane diffraction. In PCLR method (Sect. 16.5.2), the
surface plasmon polaritons with large in-plane k-vectors were used to offset the
parallel momentum gained by the diffraction.
The question emerges regarding a principle possibility of using arrays with much
shorter periods where the coupling to SLRs becomes inefficient. In a simple 1D
model, the cutoff condition for coupling with SLRs is a > λ/(1 + n eff ), where n eff
is the index of the surrounding medium. If the gaps are sufficiently small (<20 nm),
the coupling between the neighboring LSPRs will create regions of very strong electromagnetic field enhancement between the metallic stripes in 1D model or between
the nanocylinders in 2D case. Such plasmonic hot spots would not be discernable
by the far-field microscopy. If the period of such nanoplasmonic arrays, which can
be termed plasmonic metasurfaces, is below the cutoff condition, they would produce illumination of the nanoscale objects located in the vicinity of such arrays with
near fields containing extremely high in-plane k-vectors, however, these illumination
peaks would not be resolvable in the far-field region.
In fact, such formation of strong maxima of the plasmonic near fields between the
Au nanoparticles was observed almost two decades ago [97] using photon scanning
tunneling microscope, see Fig. 16.14. Although this experiment was performed for
1D chains of nanoparticles, similar plasmonic hot spots can be expected in the case of
2D arrays. Theoretically, it is important to know the photon dispersions in such shortperiod arrays where strong hybridization of LSPRs can take place. In principle, the
basic properties can be understood based on a tight-binding approximation in 1D [98]
and 2D [99] cases. However, the modeling of LSPR as point dipoles in tight-binding
A. V. Maslov and V. N. Astratov
sentative SEM (Fig. 16.13a), PCLR (Fig. 16.13b), and white light (WL) (Fig. 16.13c)
images of a Cr/Au double-line sample.
Although PCLR has not been integrated with microsphere-based imaging yet, we
believe that this method is highly promising in this regard. The limited resolution
advancement in PCLR is related to the absence of the magnification of the image in
a near-field zone where the higher spatial frequencies of the object are contained. A
combination of MN with PCLR can allow additional image magnification required
for further resolution improvement.
16.5.3 Microspherical Nanoscopy Based on Short-Period
Arrays with Small Gaps: Super-Resolution Without
Post-imaging Processing
In previous Sects. 16.5.1 and 16.5.2 we considered different situations when the
nanoplasmonic layer adjacent to the objects was designed to collect the objects’s
near-fields that resulted in super-resolution imaging of these objects. In PSIM
(Sect. 16.5.1), the illumination of the objects was provided with short-period standing plasmonic waves. In LPSIM (Sect. 16.5.1), the individual LSPRs in nanodisks
were arranged with the wavelength-scale periodicity to provide illumination with
collective SLRs due to out-of-plane diffraction. In PCLR method (Sect. 16.5.2), the
surface plasmon polaritons with large in-plane k-vectors were used to offset the
parallel momentum gained by the diffraction.
The question emerges regarding a principle possibility of using arrays with much
shorter periods where the coupling to SLRs becomes inefficient. In a simple 1D
model, the cutoff condition for coupling with SLRs is a > λ/(1 + n eff ), where n eff
is the index of the surrounding medium. If the gaps are sufficiently small (<20 nm),
the coupling between the neighboring LSPRs will create regions of very strong electromagnetic field enhancement between the metallic stripes in 1D model or between
the nanocylinders in 2D case. Such plasmonic hot spots would not be discernable
by the far-field microscopy. If the period of such nanoplasmonic arrays, which can
be termed plasmonic metasurfaces, is below the cutoff condition, they would produce illumination of the nanoscale objects located in the vicinity of such arrays with
near fields containing extremely high in-plane k-vectors, however, these illumination
peaks would not be resolvable in the far-field region.
In fact, such formation of strong maxima of the plasmonic near fields between the
Au nanoparticles was observed almost two decades ago [97] using photon scanning
tunneling microscope, see Fig. 16.14. Although this experiment was performed for
1D chains of nanoparticles, similar plasmonic hot spots can be expected in the case of
2D arrays. Theoretically, it is important to know the photon dispersions in such shortperiod arrays where strong hybridization of LSPRs can take place. In principle, the
basic properties can be understood based on a tight-binding approximation in 1D [98]
and 2D [99] cases. However, the modeling of LSPR as point dipoles in tight-binding