On Composition, Morphology, and Size Distribution of Airborne Mineral Dust - Mineral Dust: A Key Player in the Earth System - page 28

to mineral dust is given by Kalashnikova and Sokolik ( 2004 ). Three-dimensional
information can be obtained by using laser scanning microscopy (Osada 2013 ),
shadowing techniques in electron microscopy (Okada et al. 2001 ), scanning electron
microscope stereogrammetry (Lindqvist et al. 2013 ), and atomic force microscopy
(Gwaze et al. 2007 ; Chou et al. 2008 ). The latter techniques are, however, quite work
intensive, and as a result only information on small particle numbers is available.
In contrast to shape, morphology is currently only described qualitatively, and no
measure for size and amount of small scale structures of roughness of mineral dust
particles is available, unlike for other aerosol types like, for example, soot (Yehliu
et al. 2011 ).
Aspect ratio measurements of several regions are summarized in Table 2.1 .
Note that the aspect ratio measurements for African dust are based on the longest
projection/area method mentioned above, while for the Asian dust the ratio of
the particle's longest projection to the orthogonally measured width is used; this
explains the generally lower aspect ratios found in Asian dust. Conversion between
the two definitions is not possible without further information on particle shape.
A considerable trend in aspect ratio as function of particle size is absent. Long-
range transported particles seem to have a higher aspect ratio than those found close
to the source, which would indicate a preferential settling of particles that are more
spherical. Where the particle length-to-height ratio has been determined, it exhibits
significantly larger values than the aspect ratio. For these cases we can conclude that
the majority of particles are elongated and platelike, resembling the shape of clay
minerals. In addition, from a methodical point of view, it means that the particles are
orientated “flat” on the substrate and that the commonly applied two-dimensional
projected area diameter is likely to overestimate the particle volume. Besides the
median values displayed here, dust aerosol consists usually of an ensemble of
particles having a broader distributions. This was illustrated, for example, by Okada
et al. ( 2001 ) and Kandler et al. ( 2007 ) who have shown that this aspect ratio
distribution can be readily parameterized by a modified log-normal distribution.
As mentioned above, shape factors are difficult to compare for different investiga-
tions and image resolutions. A general trend, however, seems to exist of increasing
shape factors with increasing particle size, indicating a more complex geometrical
structure of larger dust particles (i.e., aggregates).
Particle shapes are mostly a function of particle composition (Kandler et al.
2007 , 2009 , 2011a ; Coz et al. 2009 ; Scheuvens et al. 2011 ). In dry conditions,
soluble compounds not containing sulfate and iron and titanium (oxyhydr)oxides
as well as partly carbonates tend to have lower aspect ratios than silicates, whereas
crystallized sulfate-containing particles - except for ammonium (hydrogen)sulfate -
are usually elongated. Consequently, an internal mixture of dust with, for example,
sodium sulfate usually leads to a growth of the aspect ratio. Besides the atmospheric
alteration of the shape of single particles, the shape distribution of a dust aerosol
could also be changed by shape-sensitive atmospheric processes, for example,
heterogeneous processing due to higher surface area or material selectivity (Matsuki
et al. 2010b ) or preferential settling due to different settling velocities (Li and Osada
2007a ).