Advances in Chemical Engi neering and Science , 2011, 1, 252-255
doi:10.4236/aces.2011.14036 Published Online October 2011 (http://www.SciRP.org/journal/aces)
The Influence of Surface Energy on the Washing
Quality of Filter Cakes
Markus Wilkens, Urs A. Peuker
Technische Universität Bergakademie Freiberg Institute of Mechanical Process Engineering and
Mineral Processing, Freiberg, Germany
Received May 1, 2011; revised May 30, 2011; accepted June 9, 2011
Abstract
The washing of filter cakes, especially the displacement washing, represents an important aspect in science
researches and industrial applications. A lot of ongoing researches are focussed on impurities, which are
dissolved in the mother liquid (e.g. sodium chloride) and washed out with the identical pure liquid without
impurities. The project flushing focuses on systems with two chemically different liquids. The main aim is to
exchange an organic solvent by water. This article focuses on the adsorption effects during a washing proc-
ess with solid systems of different wetting behaviours.
Keywords: Filter Cake Washing, Surface Energy Effects, Residual Content
1. Introduction
Washing processes and in this case also flushing proc-
esses always involve the elution of impurities. The aim is
the preferably complete replacement of the mother liquid
by the lowest possible amount of washing liquid.
A possibility to characterise a washing process is the
washing curve (Figure 1), here a concentration ratio c/c0
is plotted as a function of a washing ratio W. The con-
centration ratio is related to the concentration c0 of the
impurity at the beginning of the washing process (for a
flushing process the concentration of the organic solvent
in the mother liquid (normally 100%)). The concentra-
tion c represents the actual concentration of the impurity.
This concentration c can be related to the filtrate or the
filter cake itself. The washing ratio W represents the ratio
of the added washing liquid volume related to the vol-
ume of the pores in a fully saturated filter cake. In case
of a flushing process a mass washing ratio is used, be-
cause of the difficult calculation of the washing liquid
volume for two-liquid-systems. Due to this, the mass of
the added washing liquid MW is related to the mass of
mother liquid in the filter cake pores MV at the beginning
of the flushing process.
W
MV
M
W
M
Because of the fact, that the concentration of an impu-
rity within the filter cake is not detectable, the concentra-
tion ratio in the filter becomes an inaccessible factor. For
the industrial application a possibility has to be found, to
calculate the concentration ratio in the filter cake out of
an accessible factor (e.g. the concentration ratio in the
filtrate). Due to this Heuser [1] defined the residual con-
tent R as an integral factor for the concentration ratio in
the filter cake.
0
1d


W
W
Rc*W
The concentration factor c* represents the concentra-
Figure 1. Ideal and exemplary real washing curve (concen-
tration ratio in the filter cake depending on the washing
ratio).

M. WILKENS ET AL.
253
tion ratio c/c0 in the filtrate and a factor for the adsorp-
tion of the mother liquid in the filter cake.
In Figure 1 an ideal washing curve and an exemplary
real washing curve are shown. For an ideal washing
process or flushing process a complete exchange of the
mother liquid by the washing liquid happens at a wash-
ing ratio of 1. This means it is possible to clean the filter
cake with the same amount of washing liquid as the
amount of mother liquid in the filter cake.
The real washing curve is divided in three characteris-
tic areas [2,3]; the displacement regime (I), the intersti-
tial regime (II) and the diffusion regime (III). During the
displacement regime the washing liquid displaces the
mother liquid in a kind of a plug flow. Due to this only
mother liquid is detected in the washing filtrate. The end
of displacement regime is indicated by the detection of a
concentration ratio lower than one in the washing filtrate.
At this point a mixture of mother and washing liquid
leaves the filter cake and the interstitial regime starts.
During the last regime of the washing process (III) the
removal of the mother liquid is dominated by diffusion
mechanisms (diffusion regime). The remaining impuri-
ties in the filter cake diffuse out of stagnant areas into the
primary stream. This removal often occurs slowly be-
cause of sterical barriers. Quite long time scales are
needed to achieve a significant impurity removal [4].
2. Materials
For the research of the adsorption properties on a wash-
ing process an aluminosilicate (Dorkafill, x50 = 1.65 µm)
in three different surface modifications is used. The sur-
faces vary from hydrophilic (DF600) over hydrophilic-
hydrophobic (DF601) to hydrophobic (DF602). Because
of the identical shape and particle size distribution the
filtration and flushing are only influenced by structure
effects, which result from the different particle-fluid in-
teractions [5].
The two liquids water and methanol are used as
mother and washing liquid. Characteristic of this system
is that the progression of the dynamic viscosity depends
on the mixing ratio of water and methanol, with a maxi-
mum at a methanol percentage of 40 vol.-percent.
3. Methods
3.1. Filtration
The filter cake formation is performed with a lab scale
suction filter (nutsch filter) according to VDI 2762. The
filtration area is 19.64 cm² (diameter 50 mm). The proc-
ess parameters are a solid volume concentration of 0.2
and a filtration pressure of 3 bar. For the displacement
washing a fully saturated filter cake is needed, because
of this the filtration is stopped when the suspension
reaches the filter cake. With the known porosity of each
system it is possible to calculate the amount of liquid
staying in the filter cake, which is necessary for the cal-
culation of the washing ratio during the flushing step.
3.2. Flushing
For the flushing process the filter cake (inside a cake
formation unit) is removed from the suction filter into a
combined compression-washing unit.
Before and during the washing the filter cake is com-
pressed by a pneumatic piston with a constant pressure of
3.5 bar to prevent fissuring and crack formation. The
washing liquid flow is provided by a pressure difference
of 2 bar.
For the composition analyses of the washing liquid
(concentration ratio in the filtrate) samples in an interval
of ΔW = 0.05 are taken. A Karl-Fischer-Titrator and a
gas chromatograph (GC) are used for the analysis.
4. Results
For the technical control of a flushing process the con-
centration of the impurity (organic solvent) in the filtrate
is in most cases the only possibility to get online infor-
mations of achieved washing qualities after a certain
time or washing ratio.
Figure 2 shows the concentration ratio c/c0 in the fil-
trate depending on the mass washing ratio for the hy-
drophilic DF600.
Due to the hydrophilic surface of DF600 the surface
energy between water and DF600 is higher than between
methanol and DF600. The definition of the surface en-
ergy as the ratio of the needed work to create a new sur-
face and the new surface itself shows, that there is a
Figure 2. Flushing process of DF600 (hydrophilic) with
water and methanol.

254 M. WILKENS ET AL.
preferential exchange of a liquid with a poor wetting
behaviour by a liquid with a good wetting behaviour.
This fact correlates with the lower reachable concen-
tration ratios of the washing process with methanol as
mother liquid and water as washing liquid (grey plot).
Between the two washing directions (1 mother liquid
water, 2 mother liquid methanol) a shift of the washing
regimes can be observed. For methanol as mother liquid
the displacement regime ends at a washing ratio of 0.81,
for water at a washing ratio of 0.96. Because of the dif-
ferent structures of the filter cakes (porosities, pore size
distribution) this shift can not be directly ascribed to the
different surface energies. If no preferred adsorption of
the two liquids exists, there should be no shift between
the washing regimes and the washing quality, repre-
sented by the concentration ratios.
The system DF601 with a hydrophilic-hydrophobic
surface shows such behaviour.
Both flushing steps show an identical washing curve,
with the end of the displacement regime in a range of W
= 0.87.
This fact leads to the assumption that the better the
wettability of a system the better is the washing result,
independent on the structure behaviours.
Until now only the concentration ratios in the washing
filtrate are presented. The more interesting concentration
in the filter cake can be calculated by the residual content.
For the experimental work a possibility is found to ex-
tract the liquid of a flushed filter cake by a freezing unit,
which is based on the principle of a vacuum cooling trap.
Figure 4 shows the residual content of the hydrophilic
system DF600 with methanol as mother liquid and water
as washing liquid.
A comparison between the residual content according
to Heuser (black) and the residual content experimentally
determined shows a difference of nearly one dimension
for high washing ratios. This would lead to a wrong
technical dimensioning of flushing processes. The resid-
ual content calculated in consideration of the influence of
the volume flow fluctuation [6,7] during a flushing proc-
ess leads to a relative good fit between calculation and
experiment. This new kind of residual content is based
on the residual content expressed by Heuser.
 
 


max
0
,max
,max
,max 0
d
1
d










Wt Filtrate mm
m
new WFiltrate m
m
m
VWt
Wt
Wt
RR VW W
W





Thought has been given to an amplification of the re-
sidual content R by a normalised volume flow. In future
works more attention will be paid to the volume flow
fluctuation and the effects which cause the fluctuation.
Figure 3. Flushing process of DF601 (hydrophilic- hydro-
phobic) with water and methanol.
Figure 4. Residual content/concentration ratio in the filter
cake for DF600 (hydrophilic); mother liquid methanol, wa-
shing liquid water.
5. Conclusions
The flushing of filter cakes is an important technique for
industrial applications. This works shows the importance
of wetting behaviours between the solid system and the
mother liquid and washing liquid involved in a flushing
process. It has been shown that the washing quality and
the washing regimes depend on the surface energy of the
system.
Another interesting and not to be undervalued fact is
the difference between the detectable concentrations
within a filter cake and the calculated concentration (re-
sidual content) out of the concentration ratio in the fil-
trate. An amplification of the residual content by a nor-
malised volume flow eliminates the shift between ex-
periment and calculation to a minimum.
6. References
[1] J. Heuser, “Filterkuchenwaschprozesse unter Besonder-
er Berücksichtigung Physikalisch-Chemischer Einflüsse,”
Shaker-zgl. Diss. U. Karlsruhe, Aachen, 2003.
[2] W. Bender, “Filtrieren, Auswaschen und Entfeuchten
feindisperser Feststoffe,” Chemie Ingenieur Technik, Vol.

M. WILKENS ET AL.
255
48, No. 4, 1976, pp. 270-281. doi:10.1002/cite.330480404
[3] F. Hardekopf, “Experimentelle und Theoretische Unter-
suchungen zum Waschen von Filterkuchen auf Filter-
pressen,” Chemie Ingenieur Technik, Vol. 64, No. 11,
1992, pp. 1041-1044. doi:10.1002/cite.330641123
[4] F. Russlim, “Flow and Mass Transfer Phenomena in Fil-
ter Cake Washing Driven by Mass Force,” Dissertation
University, Karlsruhe, 2009.
[5] S. Neubauer and U. Peuker, “Flüssigkeitsmischungen als
Suspensionsflüssigkeit-Wirkung auf die Filtrationspara-
nmeter,”Chemie Ingenieur Technik, Vol. 79, No. 2, 2007,
pp. 1753-1758. doi:10.1002/cite.200700128
[6] M. Wilkens and U. Peuker, “Flushing-Washing Behav-
iour of Filter Cakes,” 13th Nordic Filtration Symposium,
Iappeenranta, 10-11 June 2010.
[7] M. Wilkens and U. Peuker, “Filter Cake Washing with
Non-Aqueous Liquids—The Influence of the Surface
Properties,” Filtech Conference 2011, Wiesbaden, 22-24
March 2011.