Applications of Raman and Surface-Enhanced Raman Scattering to the Analysis of Eukaryotic Samples - Emerging Raman Applications and Techniques in Biomedical and Pharmaceutical Fields - page 83

Fig. 4.5. Ramanspectraoftreepollenthatisrichincarotenoid.Spectraofpollen
from horse-chestnut (a, b), sallow (c, d), large-leaved linden (e, f) before irradiation
with laser light of 633 nm wavelength for photodestruction of carotenoids (traces a,
c and e) and after 1 h photodestruction with 633 nm. All spectra were excited with
785 nm, 10 s accumulation time and laser power of 18 mW ∼ 1 . 8 × 10 6 W / cm 2
imposed with the contributions from other molecules. Originally, in HCA,
the spectra of these species formed one large cluster, obviously due to their
high resemblance with respect to the carotenoid spectral features (Fig. 4.4).
Therefore, the strong carotenoid features have to be taken into account ei-
ther prior to or post data acquisition. Interference from carotenoid pigments
needs to be considered before automated pollen identification procedures can
be established.
To enable the utilization of the full spectral range for classification also
in carotenoid-rich species, we propose a method for photodestruction of
carotenoid pigments in pollen using 633 nm light. In initial experiments that
were conducted at 633 nm excitation, we observed that the bands ascribed
to carotenoid molecules were not stable but diminished steadily, even at rel-
atively low excitation intensities of 10 mW. Figure 4.6 displays spectra of
horse-chestnut pollen excited with 633 nm as a function of exposure time to
the laser. The irreversible decrease of the intensities of the typical carotenoid
bands suggests the photodestruction of these molecules (see arrows in Fig. 4.6)
[72-74]. With excitation at 785 nm, the spectra of the pollen remained unal-