178 I.R. Sambles et al 1·2 S 08 Gold 323334353637383940 Angle/deg Figure 5. Form of the reflectivity, frss fe Coupling gap urve for p-polarised and Reflectivity curves obtained from a palladium film on s-polarised (.=6328mm) thick gold and silver prism(n= 1- 699) using 3 391 um radiation, the films with a prism(n=1·766) asmon occurs at about 45. Coupling gap for (a)is is 0-5 um for gold and 1.0 um for silver. 445 m and for(b)is90μn which normally oxidize[4, 5]. Of course there is no simple manner in which it may be changed to use as a sensor or to optimize coupling at other wavelengths. It is for this reason and also because of the small air gap required for coupling in the visible that the Otto geometry has received rather limited attention over the years Fortunately this has not severely impeded progress in he area of the optical excitation of surface plasmons This is because there is an alternative and much simpler y. Rather than dielectric Gretsch mann and Raether[6] realized that the metal itself could be used as the evanescent tunnel barrier provided it was thin enough to allow radiation to penetrate to the other figure 6. Variation of the surface plasmon resonance (at side. All that is now needed is a prism with a thin coating 0-75 um and(e)1-0 um. The s upling gaps of (a)o-s ymw hes It some suitable metal. This is illustrated in figure 4(6) a= 632. 8 nm) in gold for ce rapidly with increasing gap. It is an easy matter to deposit a thin film(<50 nm)of a metal such as silver or gold on to a prism and create a suitably smooth film which may support a very strong surface plasmon resonance. A typical result for silver in of the spectrum, yet here surprisingly little this geometry is shown in figure 8. The continuous line in mental work has been conducted. We illustrate in this figure is, as in figure 7, the fit obtained using simple 7 results for palladium in this region of the Fresnel reflectivity theory for a 2-interface system. For the Otto geometry the real and imaginary parts of the The spacing problem created by the Otto geometry metal permittivity and the air gap thickness are unknown may be addressed in quite a different manner by using, variables in the fitting procedure, while for the layer(or perhaps a spun polymer). This of course gives are unknown plus the thickness of the metal film. By a non-adjustable gap but at least the fabrication is simple carefully recording data and comparing with theoretic for now it is only necessary to evaporate, on top of the ally generated curves it is possible to obtain useful appropriate thickness dielectric, a thick metal layer to information as regards the dielectric response of the give the resonance. This particular procedure may indeed metal which supports the surface plasmon. (In table I we be very beneficial in the study of protected interfaces as list reported a values for gold and silver using the for example in the case of magnesium or aluminium Kretschmann-Raether technique