17 Microsphere-Assisted Interference Microscopy
449
Tomographic diffractive microscopy With tomographic acquisitions, tomographic diffractive microscopy (TDM) is based on the principle of digital holographic
microscopy (DHM) in which interferometry enables the axial resolution to reach a
few tens of nanometres [74]. However, the lateral resolution in DHM is still limited by the diffraction of light, and in this case requires a coherent light source, i.e.
δ x,y ∼ λ [75]. TDM achieves an improvement in the lateral resolution by varying the
illumination angle of the plane wave [76–78]. By performing multiple incident illuminations, the numerical aperture of the interference system appears thus enhanced
in the Fourier domain and, further, the coherent noise is averaged, providing a higher
SNR [79]. TDM allows the reconstruction of the volume of the sample using an
adapted model of diffraction [80, 81] with a lateral resolution that is twice as high as
that in DHM. It was shown that a decrease in the wavelength of the light to 405 nm
allows TDM to visualize objects 90 nm of size [82].
Total internal reflection microscopy In 1981, D. Axelrod used the idea proposed
by E. J. Ambrose on total internal reflection [83] in order to increase the axial
resolution of biological imaging systems [84]. A transparent prism is placed above
the surface of the sample and only the evanescent waves generated at the glass–sample
interface allows the illumination of the sample. This method is known as total internal
reflection microscopy (TIRFM) and makes it possible to excite an interface to within
a few 100 nanometres depth, corresponding to the decay in the evanescent field. This
provides the advantage of removing the out-of-focus fluorescence signal, resulting in
a higher signal-to-noise ratio [85]. An axial resolution of 100 nm can thus be achieved
which is smaller than the diffraction limit. Nevertheless, the gain in resolving power
is only along the optical axis. In fact, the lateral resolution is still limited by the
diffraction although an immersion objective lens collects the fluorescence far-field
information. Exploiting the performance of SNOM, the emanating evanescent waves
were therefore recorded with a high-lateral resolution, leading to scanning tunnelling
optical microscopy (STOM) [32, 86].
Laser scanning tomography At the end of the 1980s, the group of J. P. Fillard
at the University of Montpellier (France) was one of the first groups to develop the
notion of optical nanoscopy [87]. In order to study defects inside silicon and the
key semiconductor alloys at the root of the communications and semiconductor light
source revolutions, several techniques were developed that allowed the observation
and study of nanometre-sized structures without resolving them. For example, laser
scanning tomography (LST) [88, 89] used the very high contrast of dark-field illumination with a focused laser beam orthogonal to the viewing direction, now known
as light sheet microscopy [90], to map micrometre-sized microprecipitates in 3D.
The technique was even able to show distributions of 10 nm-sized microprecipates
within annealed gallium arsenide [91]. By means of image processing, phase contrast
microscopy was able to show up the effect of pinning of looped dislocation paths in
gallium arsenide by individual micropreciptates [87]. The atomic sized defect was
indirectly visible due to it being extended in one direction and forming a stress field
that modified the surrounding refractive index sufficiently for the path to be discernible. Since then, many other optical nanoscopy techniques have been developed
for studying and measuring nanostructures without resolving them, such as surface-
449
Tomographic diffractive microscopy With tomographic acquisitions, tomographic diffractive microscopy (TDM) is based on the principle of digital holographic
microscopy (DHM) in which interferometry enables the axial resolution to reach a
few tens of nanometres [74]. However, the lateral resolution in DHM is still limited by the diffraction of light, and in this case requires a coherent light source, i.e.
δ x,y ∼ λ [75]. TDM achieves an improvement in the lateral resolution by varying the
illumination angle of the plane wave [76–78]. By performing multiple incident illuminations, the numerical aperture of the interference system appears thus enhanced
in the Fourier domain and, further, the coherent noise is averaged, providing a higher
SNR [79]. TDM allows the reconstruction of the volume of the sample using an
adapted model of diffraction [80, 81] with a lateral resolution that is twice as high as
that in DHM. It was shown that a decrease in the wavelength of the light to 405 nm
allows TDM to visualize objects 90 nm of size [82].
Total internal reflection microscopy In 1981, D. Axelrod used the idea proposed
by E. J. Ambrose on total internal reflection [83] in order to increase the axial
resolution of biological imaging systems [84]. A transparent prism is placed above
the surface of the sample and only the evanescent waves generated at the glass–sample
interface allows the illumination of the sample. This method is known as total internal
reflection microscopy (TIRFM) and makes it possible to excite an interface to within
a few 100 nanometres depth, corresponding to the decay in the evanescent field. This
provides the advantage of removing the out-of-focus fluorescence signal, resulting in
a higher signal-to-noise ratio [85]. An axial resolution of 100 nm can thus be achieved
which is smaller than the diffraction limit. Nevertheless, the gain in resolving power
is only along the optical axis. In fact, the lateral resolution is still limited by the
diffraction although an immersion objective lens collects the fluorescence far-field
information. Exploiting the performance of SNOM, the emanating evanescent waves
were therefore recorded with a high-lateral resolution, leading to scanning tunnelling
optical microscopy (STOM) [32, 86].
Laser scanning tomography At the end of the 1980s, the group of J. P. Fillard
at the University of Montpellier (France) was one of the first groups to develop the
notion of optical nanoscopy [87]. In order to study defects inside silicon and the
key semiconductor alloys at the root of the communications and semiconductor light
source revolutions, several techniques were developed that allowed the observation
and study of nanometre-sized structures without resolving them. For example, laser
scanning tomography (LST) [88, 89] used the very high contrast of dark-field illumination with a focused laser beam orthogonal to the viewing direction, now known
as light sheet microscopy [90], to map micrometre-sized microprecipitates in 3D.
The technique was even able to show distributions of 10 nm-sized microprecipates
within annealed gallium arsenide [91]. By means of image processing, phase contrast
microscopy was able to show up the effect of pinning of looped dislocation paths in
gallium arsenide by individual micropreciptates [87]. The atomic sized defect was
indirectly visible due to it being extended in one direction and forming a stress field
that modified the surrounding refractive index sufficiently for the path to be discernible. Since then, many other optical nanoscopy techniques have been developed
for studying and measuring nanostructures without resolving them, such as surface-