Page 157 - Vibrational Spectroscopic Imaging for Biomedical Applications
P. 157
sFTIR, Raman, and SERS Imaging of Fungal Cells 133
5.3.2 Spatial Resolution
Different spectroscopies provide information on different scales:
magnetic resonance imaging (MRI) can provide ex vivo images of
brain morphology; however, the spatial resolution is on the order
3
of mm , much larger than the size of a hippocampal neuron (~20 μm)
or a fungal hypha (3 to 10 μm). Scanning electron microscopy (SEM)
provides images of molecular organization on the order of nanome-
ters, while x-ray crystallography can reveal accurate molecular struc-
tures on the scale of bond lengths (Å). The best spatial resolution
possible with Raman and IR microspectroscopy falls between 0.5 and
1 μm, respectively. The physical differences between these two meth-
ods result in different issues, which we deal with separately.
Spatial resolution in any IR map ultimately depends on how the
individual pixels are apertured or on how an area is imaged onto an
array detector; the lower limit is ultimately determined by diffrac-
−1
tion. The mid-IR region nominally ranges from 4000 to 400 cm ; this
translates into wavelengths of 2.5 to 25 μm. The diffraction limit
becomes an issue because the wavelengths of infrared light impose a
physical limit on the spatial resolution that may be achieved. The
mid-IR region encompasses most of the fundamental vibrational
energies of chemical bonds. If the tissue components of interest are
lipids, they may be identified through the CH stretch vibrational
−1
modes around 2900 cm . In this case, the diffraction limit to spatial
resolution would be on the order of 3 μm. However, if the primary
interest is sugar composition, the vibrational region of interest is
likely the sugar ring vibrational modes around 1000 cm , and the
−1
associated spatial resolution is about 10 μm. Even if the spatial resolution
chosen is on the order of 3 μm, it must be remembered that the thickness
of the sample represents the third dimension to the sampling volume,
and the IR light may be internally scattered slightly on passage
through the sample. Finally, experimental considerations and sample
condition can affect spatial resolution.
In Raman microspectroscopy, the wavelength of the laser source
is much shorter, typically in the near IR to visible spectrum (though
deep UV is also being used); spatial resolution of 1 μm is easily
feasible. There is a trade-off between IR and Raman, in that the latter
spectra are less intense and, particularly in biosamples, occur in
conjunction with broadband fluorescence that gives an intense back-
ground signal that must be removed though postprocessing of the
data. However, Raman spectroscopy is ideal for analysis of micro-
crystalline inclusions (see below).
5.4 sFTIR Spectra of Fungi
The information obtained from the spatially resolved sFTIR spectra
of fungi relates primarily to differences as cells mature and sporulate.