11 Label-Free Super-Resolution Microscopy by Nonlinear …
277
respectively. This represents reduction of ~16% in the PSF pp width, which is in close
agreement with the theoretical prediction.
11.6 Using Nonlinear Response in Other Label-Free
Microscopy Modalities
11.6.1 Temperature-Dependent Raman Scattering for SR
Raman microscopy, in which specific spectral features discriminate between distinct materials, provides a label-free counterpart to fluorescence microscopy. Tip
enhancement methods [42], combining scanning probe microscopy, such as STM
[43] or AFM [44] and Raman spectroscopy, provide rich information with nanometric resolution on the studied system. However, the use of tip-enhanced Raman
microscopy is limited to the surface. Katsumasa Fujita introduces in this book the
combination of structured illumination and micro-Raman microscopy to achieve farfield Raman-based SR. We present the use of nonlinear response to achieve SR by
monitoring photo-modulated Raman scattering [45].
As in NPMR, an ultra-short laser pulse (pump) is focused on the sample, and
induces (within ~1 ps) a diffraction-limited temperature profile. The integral intensity
of the Raman peaks, as well as the peak frequency (energy) and shape, changes with
the local temperature. An overlapping probe laser records specific spectral regions
of interest (ROI) in the Raman spectrum. Our SR approach relies on measuring the
changes in the Raman spectra, induced by the pump beam.
Simulations showed improvement of two to three times in resolution over the
diffraction limit. Experimental validation of the concept by monitoring the changes
in the total intensity of the Raman peaks shows mild resolution enhancement (
√
2).
The experimental setup is illustrated in Fig. 11.11. As in NPMR, the pump
(392 nm) and probe (785 nm) beams, with a controlled delay, are merged using
a dichroic mirror and enter the Raman microscope. Scattered light is removed by a
notch filter and the Raman spectra are detected by a spectrograph coupled to a CCD
camera. The sample is scanned by an x–y stage with 10 nm resolution. The difference
in Raman signal is taken at 0.1 Hz square wave modulation of the delay in the probe
beam.
Calibration measurements of the Raman spectra of silicon in the temperature
range of 300–900 K were performed using a heating plate and a continuous wave
532 nm laser source (Fig. 11.12). On heating, the Stokes Raman peak shifts to lower
vibrational energies, broaden and decrease its total intensity. The spectral “hot” ROI
for detection in the simulation is marked. In our simulations, we modeled these
spectral changes using an analytical model for the temperature dependence of the
Raman peak [46], while scanning different sample geometries (point, line, and surface) formed images according to the intensity inside the spectral ROI. We assumed
277
respectively. This represents reduction of ~16% in the PSF pp width, which is in close
agreement with the theoretical prediction.
11.6 Using Nonlinear Response in Other Label-Free
Microscopy Modalities
11.6.1 Temperature-Dependent Raman Scattering for SR
Raman microscopy, in which specific spectral features discriminate between distinct materials, provides a label-free counterpart to fluorescence microscopy. Tip
enhancement methods [42], combining scanning probe microscopy, such as STM
[43] or AFM [44] and Raman spectroscopy, provide rich information with nanometric resolution on the studied system. However, the use of tip-enhanced Raman
microscopy is limited to the surface. Katsumasa Fujita introduces in this book the
combination of structured illumination and micro-Raman microscopy to achieve farfield Raman-based SR. We present the use of nonlinear response to achieve SR by
monitoring photo-modulated Raman scattering [45].
As in NPMR, an ultra-short laser pulse (pump) is focused on the sample, and
induces (within ~1 ps) a diffraction-limited temperature profile. The integral intensity
of the Raman peaks, as well as the peak frequency (energy) and shape, changes with
the local temperature. An overlapping probe laser records specific spectral regions
of interest (ROI) in the Raman spectrum. Our SR approach relies on measuring the
changes in the Raman spectra, induced by the pump beam.
Simulations showed improvement of two to three times in resolution over the
diffraction limit. Experimental validation of the concept by monitoring the changes
in the total intensity of the Raman peaks shows mild resolution enhancement (
√
2).
The experimental setup is illustrated in Fig. 11.11. As in NPMR, the pump
(392 nm) and probe (785 nm) beams, with a controlled delay, are merged using
a dichroic mirror and enter the Raman microscope. Scattered light is removed by a
notch filter and the Raman spectra are detected by a spectrograph coupled to a CCD
camera. The sample is scanned by an x–y stage with 10 nm resolution. The difference
in Raman signal is taken at 0.1 Hz square wave modulation of the delay in the probe
beam.
Calibration measurements of the Raman spectra of silicon in the temperature
range of 300–900 K were performed using a heating plate and a continuous wave
532 nm laser source (Fig. 11.12). On heating, the Stokes Raman peak shifts to lower
vibrational energies, broaden and decrease its total intensity. The spectral “hot” ROI
for detection in the simulation is marked. In our simulations, we modeled these
spectral changes using an analytical model for the temperature dependence of the
Raman peak [46], while scanning different sample geometries (point, line, and surface) formed images according to the intensity inside the spectral ROI. We assumed