15 Super-Resolution Imaging and Microscopy by Dielectric …
373
and neurobiology. It has potential to revolutionize the entire fields of biology and
medicine. In 2014, Nobel Prizes in Chemistry were awarded to the three leading scientists (Stefan W. Hell, Erik Betzig and W. E. Moerner) of the techniques. One practical
limit is that these techniques are not applicable to imaging non-fluorescence samples,
including for example the semiconductor chip devices and biological viruses and subcellular structures which cannot be labelled using existing fluorophores. Moreover,
using of fluorescence could also alter original function and dynamic processes of biological specimens. As a result, strong needs still exist to develop a super-resolution
lens which can offer high-resolution and label-free imaging of samples.
Recently, through manipulation of the diffraction of light with binary masks or
gradient met surfaces, several miniaturized and planar lenses have been reported with
intriguing functionalities such as subdiffraction-limit focusing in far-field, ultrahigh
numerical aperture and large depth of focus, which provides a viable solution for the
label-free super-resolution imaging. The most well-known example is possibly the
‘superoscillatory lens (SOL)’ by Zheludev’s group in Southampton [8]. The basic
idea, which has root connection with Toraldo di Francia’s proposal in 1956 [16], is
to use a carefully-designed amplitude or phase zone plate to modulate the beam to
achieve a super-resolution spot in the far-field, through constructively and destructive
interference without evanescence waves being involved. The key disadvantage in the
technique is the appearance of giant sidelobes near to the central spot, which affected
the practical adoption of the technology. The central spot resolution can, in theory,
range from infinite small to 0.38λ/NA (known as ‘super-oscillation criteria’ in [17],
NA: Numerical Aperture). More recently, Qin et al. reported the development of
supercritical lens (SCL), a planar diffraction component which has a focusing spot
smaller than 0.61λ/NA but slightly larger than SOL (0.38λ/NA), and a needle-like
focal region with its Depth of Focus (DOF) z = 2λ/NA
2 that differs from traditional
spherical lens, Fresnel Zone Plate, SOL, and others [18]. The main advantage of
SCL is its 3D imaging capability (DOF: 12λ) in axial direction with modest superresolution (0.41λ) in lateral direction. For more information on latest development
of planar diffractive lens, please refer to a latest review by Huang and co-workers
[17].
15.1.2 Microsphere Super-Resolution Imaging
Wang and co-workers published their pioneering work on microsphere nanoscopy
in 2011 [19]. The technique employs micro-sized spheres as super-resolution lenses
(superlens) to magnify underlying objects before projecting them into the objective lens of a conventional microscope (see Fig. 15.1). The spheres generate subdiffraction illumination on the underlying object, excite and collect the near-field
evanescent object information and form virtual images that are subsequently captured
by the conventional lens. This is a label-free and real-time imaging technique, which
can directly resolve ~50 nm features under white light illumination. Such resolving power corresponds to a calibrated resolution of ~λ/6–λ/8 based on convolution
373
and neurobiology. It has potential to revolutionize the entire fields of biology and
medicine. In 2014, Nobel Prizes in Chemistry were awarded to the three leading scientists (Stefan W. Hell, Erik Betzig and W. E. Moerner) of the techniques. One practical
limit is that these techniques are not applicable to imaging non-fluorescence samples,
including for example the semiconductor chip devices and biological viruses and subcellular structures which cannot be labelled using existing fluorophores. Moreover,
using of fluorescence could also alter original function and dynamic processes of biological specimens. As a result, strong needs still exist to develop a super-resolution
lens which can offer high-resolution and label-free imaging of samples.
Recently, through manipulation of the diffraction of light with binary masks or
gradient met surfaces, several miniaturized and planar lenses have been reported with
intriguing functionalities such as subdiffraction-limit focusing in far-field, ultrahigh
numerical aperture and large depth of focus, which provides a viable solution for the
label-free super-resolution imaging. The most well-known example is possibly the
‘superoscillatory lens (SOL)’ by Zheludev’s group in Southampton [8]. The basic
idea, which has root connection with Toraldo di Francia’s proposal in 1956 [16], is
to use a carefully-designed amplitude or phase zone plate to modulate the beam to
achieve a super-resolution spot in the far-field, through constructively and destructive
interference without evanescence waves being involved. The key disadvantage in the
technique is the appearance of giant sidelobes near to the central spot, which affected
the practical adoption of the technology. The central spot resolution can, in theory,
range from infinite small to 0.38λ/NA (known as ‘super-oscillation criteria’ in [17],
NA: Numerical Aperture). More recently, Qin et al. reported the development of
supercritical lens (SCL), a planar diffraction component which has a focusing spot
smaller than 0.61λ/NA but slightly larger than SOL (0.38λ/NA), and a needle-like
focal region with its Depth of Focus (DOF) z = 2λ/NA
2 that differs from traditional
spherical lens, Fresnel Zone Plate, SOL, and others [18]. The main advantage of
SCL is its 3D imaging capability (DOF: 12λ) in axial direction with modest superresolution (0.41λ) in lateral direction. For more information on latest development
of planar diffractive lens, please refer to a latest review by Huang and co-workers
[17].
15.1.2 Microsphere Super-Resolution Imaging
Wang and co-workers published their pioneering work on microsphere nanoscopy
in 2011 [19]. The technique employs micro-sized spheres as super-resolution lenses
(superlens) to magnify underlying objects before projecting them into the objective lens of a conventional microscope (see Fig. 15.1). The spheres generate subdiffraction illumination on the underlying object, excite and collect the near-field
evanescent object information and form virtual images that are subsequently captured
by the conventional lens. This is a label-free and real-time imaging technique, which
can directly resolve ~50 nm features under white light illumination. Such resolving power corresponds to a calibrated resolution of ~λ/6–λ/8 based on convolution