Extinction spectrum

Figure 3.2 Extinction spectra of colloidal water solutions of gold nanospheres and nanorods. Dotted curve nanospheres (diameter 15-25 nm). Solid curve nanorods, low aspect ratio. Dashed curve nanorods, high aspect ratio. Extinction is normalized at about 520 nm. (Reproduced with permission from Royal Society of Chemistry [10]).

Figure 6. Difference extinction spectra of the silver nanoparticles on the rutile XiO2(100) single crystal after irradiation with monochromatic visible light (wavelength = 480 nm, light intensity 5.0mW cm , irradiation time = 30 min 550 nm, 5.0mW cm , 30min 600nm, 3.0mWcm , 60min (FWHM = lOmn)).

Steele, H. M., and R. P. Turco, Retrieval of Aerosol Size Distributions from Satellite Extinction Spectra Using Constrained Linear Inversion, J. Geophys. Res., 102, 16737-16747 (1997a). 

Because the particles must be very small, not more than a few hundred angstroms for most substances, there have been few laboratory observations of structure in ultraviolet extinction spectra of small, nonmetallic particles. Particles of the required size are difficult to make. Grinding them from the bulk, with the attendant problem of separation, is almost hopeless. Various smokes can be made, MgO particularly easily, but not of all substances that might be of interest. 

Measured extinction spectra for aqueous suspensions of polystyrene spheres—the light scatterer s old friend—are shown in Fig. 11.19. Water is transparent only between about 0.2 and 1.3 jam, which limits measurements to this interval. These curves were obtained with a Cary 14R spectrophotometer, a commonly available double-beam instrument which automatically adjusts for changing light intensity during a wavelength scan and plots a continuous, high-resolution curve of optical density. To reproduce the fine structure faithfully, the curves were traced exactly as they were plotted by the instru-... 

Extinction spectra could be used to size particles by matching measured features with those calculated from Mie theory provided that the size distribution is narrow and the particles are nearly spherical. 

For any real material, the frequency at which (12.27) is satisfied is complex—the surface modes are virtual. However, its real part is approximately the frequency where the cross sections have maxima, provided that the imaginary part is small compared with the real part. We shall denote this frequency by us. For a sphere, o>s is the Frohlich frequency wF. If used intelligently, always keeping in mind its limitations, (12.27) is a guide to the whereabouts of peaks in extinction spectra of small ellipsoidal particles but it will not necessarily lead to the exact frequency. 

Figure 12.20 Extinction spectra calculated for small aluminum spheres and a continuous distribution of ellipsoids (CDE) in air (---) and in a medium with c = 2.3 (—). The circles show data...

Dayawansa, I. J., and C. F. Bohren, 1978. The effect of substrate and aggregation on infrared extinction spectra of MgO particles, Phys. Status Solidi (b), 86, K27-K30. 

As noted above, the LUT algorithm assumes a unimodal lognormal functional form to describe stratospheric aerosols. This approximation is well suited for most non-volcanic stratospheric aerosols as shown by Pueschel et al. [7] and Yue et al. [8]. Volcanic size distributions, however, are typically bi- or trimodal. This raises the question of whether the assumption of unimodality in the LUT can introduce bias into retrieved values of Rt//, S and V. Russell et al. [1] have shown that retrieved unimodal distributions accurately describe the second, larger mode of several measured bimodal size distributions, but fail to account for the smaller particles in the first mode. The smaller particles, which contribute little to the measured extinction spectra, are not accounted for in the LUT retrievals. Unless this bias is accounted for, the values of Rtff retrieved under the assumption of a unimodal distribution will be overestimated. 

A procedure to estimate and remove any bias due to assuming a unimodal functional form has been incorporated into the LUT. The bias is determined by testing the LUT performance with extinction spectra calculated from measured bi- and tri-modal size distributions obtained from wire-impactor and dustsonde measurements which coincide in space and time with the SAGE D/CLAES composites. The bi- and tri-modal distributions were reported by Pueschel et al. [9], Goodman et al. [10] and Deshler et al. [11,12], The calculated bias is then subtracted from the retrieved Reff values to obtain bias-corrected values of RtJf. Similar procedures are used to estimate and remove bias in retrieved S and V. Figure 5 compares a measured post-Pinatubo bimodal... 

Figure 5. Left Frame Comparison between a measured post-Pinatubo bimodal size distribution (solid line) and that retrieved by the LUT from its best-fit extinction spectrum (i.e., at = 1.6) (dashed line). The fitting parameters of the measured bimodal and the LUT retrieved uni-modal are shown in the table. Right Frame Calculated extinction spectra for size distributions in the left frame The error bars on the spectrum calculated from the measured bimodal (open circles) are derived from the relative errors on coincident SAGE II and CLAES measure-...

Figure 6. Results of applying the LUT algorithm to synthetic extinction spectra calculated from measured pre- and post-Pinatubo size distributions obtained from Pueschel el al. [9], Goodman el al [10] and Deshler el al. [11,12]. R,/bimodal) is the effective radius of the measured bimodal size distribution, and R /uni modal) is the corresponding effective radius returned by the LUT. The dots are results obtained when a range of distribution widths are considered in the LUT calculations and the crosses are results obtained when at is restricted to the value that yields the best fit between calculated and measured extinction spectra. The solid and dashed curves are second order polynomial fits to the dots and crosses, respectively.

Figure 2.5 Reproduced with permission from 7. Phys. Chem. C 2007, III, 16076-16079. Copyright 2007 American Chemical Society, (a-b) Extinction spectra (0ext( s)) 8° nanoparticle arrays Al (100nm...

An example of the results for such an approach is shown in Fig. 2.5 (from Ref [1]), which we now discuss. Methylene Blue (MB) is used here as a probe molecule and is transferred onto gold NP arrays by dipping them into a 10 // M MB solution for 5 minutes. The arrays are plasma-cleaned before any dipping, to ensure that no contaminants are previously adsorbed on the NPs. The MB molecules are therefore adsorbed directly onto the gold surfaces, with possibly some molecules further away from the surface if several monolayers are present. Three NP arrays are used with distinct LSP resonances at 612 nm (array Al), 667 nm (A2) and 713 nm (A3), as evidenced in the extinction spectra of Fig. 2.5(b). The resonance of A2 is close to the peak absorption and fluorescence of MB and would be considered as the standard situation for most MEF experiments (except for the direct adsorption onto the metal). Al and A3 have resonances much further away on either side of the MB fluorescence spectrum. The MEF spectra (corrected for the ITO background), shown in Fig. 2.5(c), exhibit a broad spectrum underneath the SERS (Raman) peaks. The SERS peaks clearly confirm the presence of MB on the NPs surface. The accompanying broad signal is attributed to the modified MB fluorescence (MEF), initially for two reasons ... 

In all the results shown here, the extinction spectra were measured on bare substrates (not covered with molecules). However, it has been shown that molecular adsorption may affect the underlying LSP resonances and therefore change slightly the extinction profiles [39]. This aspect can be improved upon by measuring the extinction spectrum of the arrays for each molecule. 

Figure 4.2 (A) Series of normalized extinction spectra of chemically synthesized colloidal Ag nanoparticle solution showing the tuning of the plasmon resonance across the visible region. (B) Images of colloidal nanoparticle solutions. The differences in color are due to variations in the size and shape of the nanoparticles within each solution.

---