Analysis and Characterization of Manufactured Nanoparticles in Aquatic Environments - Environmental and Human Health Impacts of Nanotechnology - page 241

X-ray analysis in the SEM: The basic assumption for a quantitative analysis of the
X-rays within the SEM is that the electron beam loses all its energy within the
specimen. This propagation of the electron probe within the specimen results in a
so-called interaction volume, which describes the distribution of the primary elec-
trons within the specimen. Further assumptions are that the sample is homoge-
neous, fl at and infi nitely thick with regard to the interaction volume. The interaction
volume depends on the energy of the primary electron beam (acceleration voltage)
and the atomic number of the specimen. The interaction volume ranges from about
one to a few m 3 for typical acceleration voltages of 15 kV and increases with

increasing acceleration voltage. If the above stated criteria are met, then a full
quantifi cation based on the X-ray signal is possible using correction algorithms.
Mostly, ZAF correction methods are applied, where effects of atomic number (Z),
absorption of X-rays within the sample (A) and the generation X-rays of a lower
energy by X-rays of higher energy (secondary fl uorescence) (F) are taken into
account. EDX analysis of NPs in the SEM will be dominated by the contributions
for the background (underlying substrate) and remains qualitative.

µ

X-ray analysis in the (S)TEM: The interaction volume in a TEM sample is essen-
tially determined by the beam diameter and the sample thickness. In the TEM it is
assumed that the electron beam penetrates the samples and the electron beam is
only scattered once (single scattering regime). Specimens in the order of 100 nm
and less fulfi l these criteria, and for quantifi cation the Cliff- Lorimer thin foil
approximation can be used. This essentially means that absorption and fl uorescence
effects can be neglected and the ZAF correction procedure reduces to a Z correc-
tion procedure. This is done by relating elemental concentration ratios (of known
standards) to measured intensity ratios of the standards. The obtained factors
(k-factors) can then be used to convert measured intensity ratios of unknown
samples into elemental concentrations. The procedure is clearly outlined in Williams
and Carter (1996). The k-factors depend on parameters such as acceleration voltage
and the beam size, and also on the optical setup of the microscope. Therefore, every
TEM should be calibrated using respective standard materials in order to get reli-
able data.

Applications and limits. The combination of conventional TEM with a qualita-
tive elemental analysis (EDX) is straightforward, and thus most frequently used.
These studies are generally performed on aquatic colloids with a lower size that
extends well into the nanoscale size range ( 100 nm). Examples include rather

general descriptions of aquatic colloids from various environments (Leppard, 1992;
Webb et al. , 2000 ; Mondi et al. , 2002 ; Lienemann et al. , 1997; Chanudet and Filella,
2008) as well as detailed investigations on colloids such as iron colloids (Perret et
al. , 2000 ; Leppard et al. , 1989 ; Lienemann et al. , 1999 ; He et al. , 1996 ; Tipping et al. ,
1981, 1982; Fortin et al. , 1993) or marine snow (Leppard et al. , 1996 ).
Although the quantifi cation of elemental abundances is rather straightforward,
it requires adequate standards and puts more severe constraints on the maximal
thickness of the particles, which explains why quantitative TEM-EDX data are only
rarely found in the literature of environmental colloids. Buffl e et al. (1989) report

<