434
A. V. Maslov and V. N. Astratov
SR microscope with 60× (NA = 1.42) oil immersion objective, FL laser excitation
at λ = 525 nm and detection at 604 − 644 nm.
The comparison of three imaging modalities, (i) without microspheres, (ii) through
microspheres without plasmonic arrays, and (iii) through microspheres using plasmonic arrays, is illustrated in Fig. 16.16. The analyses of these results show that in the
case (i) represented by Fig. 16.16a–c the resolution ∼λ/2.2 is, as expected, below the
diffraction limit determined by the microscope objective. In the case (ii) represented
by Fig. 16.16d–f the resolution of virtual imaging through microspheres ∼λ/3.7
approaches the SIL limit. Finally, in the case (iii) represented by Fig. 16.16g–i the
resolution is found to be ∼λ/6.8, similar to imaging of FL nanospheres in Fig. 16.15d.
The results of imaging FL labeled nanospheres and F-actin protein filaments presented in Figs. 16.15, 16.16 cannot be explained by the regular LPSIM mechanism
because the post-imaging processing was not used in these experiments [44]. In addition, the bandstructure folding into the escape cone, as it was already mentioned, is
not possible for arrays with periods as short as 80 and 100 nm in the range of detection wavelengths selected in these experiments. It has been hypothesized [44] that
the mechanism of the observed super-resolution is based on the coupling of the emission on nanoscale objects to the localized surface plasmon resonances (LSPRs) in the
adjacent Au nanodiscs. For sufficiently small gaps between the coupled nanodiscs,
the maxima of the plasmonic near-fields can be located between the nanodiscs [97].
The resonantly enhanced plasmonic near-field peaks with large in-plane k-vectors
are evanescently coupled to high-index microspheres, thus contributing to the formation of magnified virtual images of FL objects. While the regular LPSIM mechanism
leads to the moire fringes formation with the need to use post-imaging processing,
the proposed mechanism does not require post-imaging processing.
16.6 Conclusion
Microspherical nanoscopy, or imaging by dielectric microspheres placed in contact
with the objects, emerged as amazingly simple and powerful method of increasing
the resolution of standard microscopes. The method does not require any complicated hardware or software—just putting the microsphere on the top of the object and
finding its virtual image using a standard microscope. Scanning large areas require
translating the microsphere. However, the critical difference compared to such methods as NSOM or AFM is that instead of slow point-by-point scanning by a very narrow
nanoprobe, the microspherical nanoscopy offers much faster area-by-area scanning
followed by stitching the images obtained for each position of microsphere on the
sample surface. The translation of microspheres can be easily achieved by a variety
of techniques from using movable coverslips with embedded microspheres [35–38,
43, 44] to attaching microspheres to AFM probes [49]. Super-resolved imaging of a
broad range of objects such as metallic and semiconductor nanostructures, surfaces
of central processing units, biological cells, and proteins has been demonstrated.
Microspheres of various sizes and materials located in the air as well as in liquids
A. V. Maslov and V. N. Astratov
SR microscope with 60× (NA = 1.42) oil immersion objective, FL laser excitation
at λ = 525 nm and detection at 604 − 644 nm.
The comparison of three imaging modalities, (i) without microspheres, (ii) through
microspheres without plasmonic arrays, and (iii) through microspheres using plasmonic arrays, is illustrated in Fig. 16.16. The analyses of these results show that in the
case (i) represented by Fig. 16.16a–c the resolution ∼λ/2.2 is, as expected, below the
diffraction limit determined by the microscope objective. In the case (ii) represented
by Fig. 16.16d–f the resolution of virtual imaging through microspheres ∼λ/3.7
approaches the SIL limit. Finally, in the case (iii) represented by Fig. 16.16g–i the
resolution is found to be ∼λ/6.8, similar to imaging of FL nanospheres in Fig. 16.15d.
The results of imaging FL labeled nanospheres and F-actin protein filaments presented in Figs. 16.15, 16.16 cannot be explained by the regular LPSIM mechanism
because the post-imaging processing was not used in these experiments [44]. In addition, the bandstructure folding into the escape cone, as it was already mentioned, is
not possible for arrays with periods as short as 80 and 100 nm in the range of detection wavelengths selected in these experiments. It has been hypothesized [44] that
the mechanism of the observed super-resolution is based on the coupling of the emission on nanoscale objects to the localized surface plasmon resonances (LSPRs) in the
adjacent Au nanodiscs. For sufficiently small gaps between the coupled nanodiscs,
the maxima of the plasmonic near-fields can be located between the nanodiscs [97].
The resonantly enhanced plasmonic near-field peaks with large in-plane k-vectors
are evanescently coupled to high-index microspheres, thus contributing to the formation of magnified virtual images of FL objects. While the regular LPSIM mechanism
leads to the moire fringes formation with the need to use post-imaging processing,
the proposed mechanism does not require post-imaging processing.
16.6 Conclusion
Microspherical nanoscopy, or imaging by dielectric microspheres placed in contact
with the objects, emerged as amazingly simple and powerful method of increasing
the resolution of standard microscopes. The method does not require any complicated hardware or software—just putting the microsphere on the top of the object and
finding its virtual image using a standard microscope. Scanning large areas require
translating the microsphere. However, the critical difference compared to such methods as NSOM or AFM is that instead of slow point-by-point scanning by a very narrow
nanoprobe, the microspherical nanoscopy offers much faster area-by-area scanning
followed by stitching the images obtained for each position of microsphere on the
sample surface. The translation of microspheres can be easily achieved by a variety
of techniques from using movable coverslips with embedded microspheres [35–38,
43, 44] to attaching microspheres to AFM probes [49]. Super-resolved imaging of a
broad range of objects such as metallic and semiconductor nanostructures, surfaces
of central processing units, biological cells, and proteins has been demonstrated.
Microspheres of various sizes and materials located in the air as well as in liquids