Faculty of Mechanical Engineering
Institute of Process Engineering and Environmental Technology
Research Group Mechanical Process Engineering
Dipl.-Ing. Lars Hillemann, PD Dr.-Ing. Michael Stintz
TU Dresden, Inst. of Process Engineering and Environmental Technology
Research Group Mechanical Process Engineering, D-01062 Dresden
Phone: +49 351 463 32914 Fax: +49 351 463 37058
Prof. Dr.-Ing. Christoph Helsper
University of Applied Science Aachen,
Ginsterweg 1, D-52428 Jülich
Phone: +49 241 6009-53114 Fax: +49 241 6009-53199
Determination of the charge distribution
of unipolar charged aerosols
Lars Hillemann, Michael Stintz, Christoph Helsper
Diffusion charging in mixing flows
Bipolar chargers are widely used in aerosol
science because their charge distribution is well
understood [1]. They produce a high ratio of
singly charged particles. Therefore they are
suitable for aerosol conditioning in mobility
spectrometers like the SMPS. Unfortunately,
these chargers employ a radioactive source to
generate an ion atmosphere. These sources are
problematic in terms of transportation and
charging efficiency. Thus, unipolar charging can
be an alternative solution. For the employment
of this technique as a part of a mobility
spectrometer the charge distribution has to be
Measurement of charge distributions
One main advantage of diffusion charging of particles
is the independency of this process from the particle
material. This requires the separation of the ion
generation process from the charging zone, where
the ions are attached to the particles. This prevents
them from undergoing field charging, which is much
stronger but material-dependent. Thus the available
types of chargers differ in the strategy of ion
transport into the charging zone.
The Figure shows the schematic setup of the
charger under consideration. Its main part is a mixing
chamber, where particles get in contact with an
unipolar ion atmosphere during convective
circulation. The ions are produced by a corona wire
and swept into the mixing chamber with the clean
bypass airflow.
The discrete charge distribution
was determined by operating the
charger with monodisperse aerosol
and subsequently classifying these
particles according to their mobility.
This means separating the particles
with respect to the number of
charges they are carrying.
The mobility analysis was done
with a DMPS-setup. In parallel, the
particle output of the charger was
quantified by a condensation
particle counter and an
electrometer. The data of these
devices deliver the generated mean
charge and this gives a good
possibility to verify the results.
A slightly modified setup was used to quantify the fraction of uncharged
particles. The mobility of these particles is zero, thus they are unaffected by
the electrical field in an electrostatic classifier. Connecting a CPC to the
exhaust port of the classifier running at maximum voltage enables for detecting
the uncharged particles (see figure above).
For the calculation of the charge
distribution from the particle
mobility spectrum the model of
the DMA by Wang & Flagan was
used [2]. This model determines
the response of a DMA for a
known aerosol input and its
charge distribution. Comparing the
model-data with the measurement
a minimization algorithm was used
to evaluate the charge distribution.
A. Wiedensohler, J. Aerosol Science, 19 (1988), 387-389.
S. C. Wang, R. C. Flagan, J. Aerosol Science, 20 (1989), 1485-1488.
[3] A. Medved, F. Dorman, S. L. Kaufman, J. Aerosol Science, 31 SUPP/1
(2000), 616-617.
Internal set-up of the corona-jet-charger
Measured concentration spectrum due to
different numbers of charges
Experimental set-up to detect
uncharged particles
The bar graph below shows an example of the calculated results recorded for
200-nm-particles. The charge distribution has a maximum at charge number 9.
The mean charge calculated from this distribution is about 7.1 charges per
particle. This gives a good compliance with the mean charge of the device
measured by Medvet et. al. [3].
The influence of the concentration was
tested by varying its value while
measuring the mean charge. For the
selected particle sizes in the range
mentioned above the mean charge was
independent from the particle
concentration in the considered span
(< 2000 p/cm3).
As standard particle material
ammonium sulfate was chosen.
Additionally several particle
materials were employed for
measuring the mean charge.
Among them DEHS-oil and soot
was used to generate
monodisperse aerosols in different
size ranges. Despite the different
surface properties and the surface
area only minimal differences in
the mean-charge-value were
Comparison to Medvet et. al. [3]
Calculated charge distribution for 200-nm-