Chemistry Reference
In-Depth Information
desorbing molecules in excess and to level the layer thickness. Also the availability
of realistic and reliable data from theory is not common. Finally, we should
consider that the three-layer model validity is sometimes questionable, as the
thickness of the layer and the strong absorption of some molecules (e.g., porphy-
rins) makes dubious the applicability of that approach. In conclusion, with respect
to the case of inorganic materials, for organics, often RAS is not used to get a deeper
knowledge of the layer electronic structure by a detailed interpretation of spectra,
but rather to connect a particular, significant spectroscopic feature, even a whole
spectrum, to some defined phase occurring at the layer in well-defined conditions,
thus becoming a signature of the existing state, that will be monitored to charac-
terize the evolution of the system with time, temperature, thickness, etc.
SDR and RAS spectroscopies are more similar than it can appear at first glance:
they have common origins and extremely similar apparatuses, and more meaning-
fully, they have often passed through similar applications and development. RAS
had a twofold derivation: on one side, it can be considered an evolution from
ellipsometry, where the working condition for light at normal incidence reduces
the otherwise formidably complex problem to be analyzed, to a heuristic and
simpler finding, that is, the existence of anisotropy [
28
]. On the other, it can be
interpreted as a particular case of SDR with polarized light, intentionally developed
for the investigation of III-V compound surfaces in UHV [
29
]. The former case
prototype was developed by D. Aspnes at Bell Labs (USA), the latter in
St. Petersburg's Ioffe Institute (USSR) by the group of V. (Slava) Safarov, in the
same period and independently, with some minor (although significant) technical
Table 1 Applicability of
SDR and RAS in different
experimental conditions
(“x” means the spectroscopy
can be profitably used)
Substrate
Layer
Isotropic
Anisotropic
Isotropic
Anisotropic
SDR
x
x
x
x
x
a
RAS
x
x
a
To be used carefully (see examples in main text)
differences in the experimental apparatuses [
30
].
In consequence of their resemblance, RAS and SDR data are similarly
interpreted within the same three-layer model and suffer common restrictions to
its applicability. However, sufficient differences exist to justify in some circum-
stances a different application of these two techniques, as reported in Table
1
. RAS
- being not dependent upon the comparison with a reference sample - is more
appropriate for investigating real-time processes. Then, its modulation at very high
frequency (usually more than 50 kHz that is the frequency of the light polarization
change between two orthogonal states) gives an unmatched time stability and
signal-to-noise ratio (down to 10
6
in recent apparatuses, commonly better than
10
4
). However, the necessary existence of a non-null resulting layer anisot-
ropy is sometimes a limit that cannot be overcome, for the intrinsic isotropy of the
1
