EXPERIMENTAL INVESTIGATION OF TRANSPORT OF STRONGLY

RETAINED SPECIES BY SOIL COLUMNS

MARCO PETRANGELI PAPINI



and MAURO MAJONE

University of Rome “La Sapienza” – Department of Chemistry P. le Aldo Moro, 5–00185 Rome, Italy

(Received 14 August, 1995; accepted 26 March, 1996)

Abstract. Column experiments have been extensively used in transport studies of major cations
but few investigations are available on migration through soils of strongly retained species that
are environmentally relevant (like heavy metals). By presenting some selected experiments (lead
and proton step-breakthrough tests in different conditions), this study shows that the soil-column
technique is also applicable in the case of species which exhibit very large retention factors. The use
of very small soil columns (about 0.4 mL of pore volumes) combined with relatively high flow rates
(0.1–0.3 mL min

,

1

) allows to observe the entire breakthrough curve (adsorption and desorption

front up to 5000 pore volumes) in reasonable experimental time, in reproducible conditions and
without experimental drawbacks. In the adopted experimental conditions no kinetic effects, related
to diffuse transport and sorption reaction were recognized; moreover, Peclet number was higher than
60. Consequently, it was possible to calculate the equilibrium isotherms from the diffuse fronts of the
breakthrough. Knowledge that can be derived, concerning the reversibility of the adsorption process,
the influence of complexation on the adsorption, the kinetics of complex formation, and the effect of
dissolution on proton transport, is also discussed.

Key words: transport, soil columns, breakthrough curves, lead, proton

1. Introduction

The ability to predict the transport of hazardous chemicals in soils and subsoils is
an important task either for the correct management of waste disposal and for the
contamination assessment of polluted sites (e.g. uncontrolled dumps). The move-
ment of a solute through soils is a complicated phenomenon which involves several
different mechanisms. While the rate of percolation is influenced by hydrogeolog-
ical parameters, physical and chemical characteristics of both soil and liquid phase
are responsible for the partitioning of the solute between the two phases in contact
and thus for its retardation during the transport.

Generally, the evaluation of solute movement through soils should be performed

by three strictly related steps, essentially comprising batch reactor experiments,
laboratory column experiments and subsequent scaling up of the results to the
natural soil of a given site if the preceding steps are satisfactory (Schweich and
Sardin, 1981, Amoozegar-Fard et al., 1984).

In this framework, batch experiments are used to obtain the equilibrium dis-

tribution function of a solute between the solid and liquid phases while column
experiments are performed to reproduce, as well as possible, the field conditions
and to validate the transport modelling.



Author for all correspondence.

Water, Air, and Soil Pollution 95: 337–351, 1997.

c

1997 Kluwer Academic Publishers. Printed in the Netherlands.

338

M. PETRANGELI AND M. MAJONE

Some inconsistencies among batch results obtained in different experiments are

often encountered, especially dealing with heterogeneous solids. These differences
can be essentially attributed to different experimental procedures and effects due
to shaking, phase separation and solid/liquid ratio are widely reported (Sparks and
Rechcigl, 1982; Gschwend and Wu, 1985; Voice and Weber, 1985; Miller et al.,
1989). With respect to the latter, during the adsorption different species are released
from the solid phase depending both on the adsorbate and solid concentration in
the suspension. The resulting change in the composition of the liquid phase during
the adsorption tests can affect the adsorption equilibria in a batch (closed) sys-
tem. However, if the experimental conditions are carefully controlled and sources
of differences are taken into account, batch experiments allow to obtain a great
number of information about the equilibrium distribution functions (i.e. adsorption
isotherms) to be included in the transport models (Roy, 1991 and 1993).

Column experiments are performed at solid/solution ratios closer to the field

conditions and allowing the released species to be removed by the continuous flow
of the mobile phase (Griffioen et al., 1992). The mobile phase can be adjusted
to the composition of interest and from the pattern of the breakthrough curve it
is possible to obtain informations about the shape (linear, convex or concave) of
the adsorption isotherm followed by the adsorbate. Moreover, column experiments
allow to observe the desorption processes, just changing the feeding to a solu-
tion without the adsorbate. From the diffuse front (adsorption or desorption front
depending on the isotherm shape) the adsorption isotherm can be calculated by a
numerical integration, if multicomponent effects can be neglected (Borkovec et al.,
1991; Burgisser et al., 1993).

Column experiments have been extensively used in transport studies of major

cations (Na, K, Ca and Mg) but few investigations are available about transport
through soils of strongly retained chemicals like radionuclides, PCB, PCH, pro-
tons and heavy metals, that are environmentally relevant (Doner, 1978; Selim et
al., 1989; Dudley et al., 1991; Dunnivant et al., 1992; Selim et al., 1992; Wagner
et al., 1994; Hinz and Selim, 1994). In general, these studies have been performed
in such conditions that 10–100 pore volumes were sufficient to observe the entire
breakthrough curve. However, from published batch investigations, it can be cal-
culated that, in cases where the affinity is very strong (Brownawell et al., 1990;
Copenhaver et al., 1993), retention factors would be greater than thousands.

The main problem in carrying out column experiments with such high retention

factors is probably the long time required to observe the breakthrough curve.
Performing column experiments in a reasonable time scale can be made either by
increasing the flow rate of the leaching solution or reducing the size of the column
(i.e. decreasing pore volumes). In both cases, experimental problems could arise
concerning possible kinetic effects, high pressure in the column, unhomogeneous
packing of the sorbing material, etc.

The aim of this paper is to show how, by using very small soil columns (about

0.4 mL of pore volumes), it is possible to carry out experiments even with species

TRANSPORT OF STRONGLY RETAINED SPECIES BY COLUMN EXPERIMENTS

339

strongly retained by the sorbing material. Some selected experiments (lead and
proton step-breakthrough tests in different conditions) are presented to demonstrate
how typical breakthrough curves can be observed up to 5000 pore volumes in
controlled conditions (negligible kinetic and dispersion effects). Knowledge that
can be derived from such experiments, such as the adsorption isotherm from a
single column experiment, the reversibility of the adsorption process, the influence
of complexation on the adsorption, the kinetics of complex formation, and the
effect of dissolution on proton transport, is also discussed.

2. Materials and Methods

2.1. S

ORBENT MATERIAL

The solid phase used in the column experiments was a volcanic soil (red pozzolan)
collected in a quarry situated in the neighbourhood of Rome (Italy). The typical
composition of these materials (essentially incoherent tuff produced directly by
volcanic eruption) is: SiO

2

45%, Al

2

O

3

16%, Fe

2

O

3

9%, FeO 1%, MgO 4%, CaO

9%, K

2

O 4%, Na

2

O 2%. The specific surface area was 56.8 m

2

/g as resulted by

BET measurement. The cation exchange capacity (C.E.C.) was determined by the
BaCl

2

method and resulted 11.4 meq/100 g. The total organic carbon content of

the soil was measured by the CMAT 550 PC Strolheim apparatus and resulted 0.04
%. The column experiments were performed on the soil previously air dried and
sieved down to the size fraction

<

200



m (which represents 20% of the total). The

fraction so obtained was about 50%

<

100



m and 30%

<

75



m.

2.2. C

OLUMN PREPARATION

The columns used in this study were glass chromatographic columns (OMNI) of
3 mm internal diameter and 10 cm length. The endpieces were provided with
0.25



m filters. The choice of such small columns was necessary to perform

experiments up to 5000 pore volumes in a reasonable time. The columns were
uniformly packed with air-dried soil, purged with CO

2

and then saturated by

upward flow with 0.05 M CaCl

2

solution. Several pore volumes of CaCl

2

solution

were applied to each column prior to perform the tracer experiments in order to
stabilize the soil aggregates and to obtain the complete removal of the presorbed
ions (Grolimund et al., 1995). The average quantity of soil used to fill the columns
was 0.9



0.03 g. The preparation of soil columns was quite straightforward and

non time-consuming (about one day was necessary to have the column completely
characterised and ready for a breakthrough experiment). For each experiment a
fresh column was used.

340

M. PETRANGELI AND M. MAJONE

Figure 1. Schematic drawing of the apparatus used for column experiments.

2.3. E

XPERIMENTAL SYSTEM

Figure 1 shows a schematic drawing of the apparatus used for the column experi-
ments. Basically, the feeding solution (S1 or S2) were pumped through the column
by HPLC pumps (P1 or P2) after be flowed through a degasser (D). For the tracer
experiments an injection loop of 25



L (L) was inserted before the column inlet

while the outlet was coupled to an UV/VIS flow-through detector (UV) connected
to a PC for data accumulation. For lead breakthrough experiments samples were
collected by a fraction collector (FC) and analysed either by ICP-AES or flame-
AAS. In the pH breakthrough experiments a pH-flow through system was inserted
before the fraction collector (PHFT) and the measurements were collected by a
recorder (R). In all the experiments the flow rate was measured by weighting. The
experiments were carried out in a thermostated room at 25.0



0.5



C.

2.4. T

RACER EXPERIMENTS

The characteristic parameters of the columns (pore volumes, porosity and Peclet
number) were determined by pulse experiment using conservative (i.e. non retained)
tracers. Sodium nitrate and potassium bromide (



10

,

3

M) were injected by the

sample loop into the flow of medium (CaCl

2

). The outlet concentration of nitrate

and bromide was measured on line by the UV/VIS flow through detector. Both
nitrate and bromide have shown the same breakthrough response confirming to

TRANSPORT OF STRONGLY RETAINED SPECIES BY COLUMN EXPERIMENTS

341

be conservative tracers. The outlet response of the tracer can be described by the
average and the standard deviation of the travel time,

t

o

and



, respectively. These

quantities can be related to the column Peclet number by



2

=t

2

o

=

2

=P

e

. The pore

volume

V

o

is calculated by

V

o

=

q

t

o

, where q is the volumetric flow rate. The

apparent porosity of the medium,



, is calculated from

V

o

=V

tot

where

V

tot

is the

total volume of the empty column. Since the tubing volume was comparable with
the column one, tracer experiments were carried out both with and without the soil
column. Thus, the response due to the external tubing was subtracted to the total
signal and the column parameters were calculated according to (Villermaux, 1981)

t

e

=

t

o

+

t

t



2

e

=



2

o

+



2

t

where

t

e

and



e

are the average and the standard deviation of the travel time,

respectively, measured for the whole system and

t

t

and



t

those measured in the

experiments performed without the soil column (

t

o

and



o

being the “net” column

parameter consequently calculated).

The typical column parameters were: pore volume = 0.42



0.02 mL,



=

0.60



0.02 and Pe always greater than 60. In Figure 2, typical tracer respons-

es (for the whole system and for the tubes as well) are reported along with the
physical characteristics and the flow data.

2.5. L

EAD EXPERIMENTS

Lead step-breakthrough experiments were carried out for different background
electrolytes and at different pH. Either sodium perchlorate, nitrate and chloride at
concentration of 0.1 M were used at two different pH (3.0 and 5.0). After the tracer
experiment, at least 500 pore volumes of the chosen solution were passed through
the column to equilibrate the soil with Na

+

and H

+

. Then, the column feeding was

switched to a solution of the same composition but in the presence of lead (spiked
either at 10 or 100 mg L

,

1

level). The experiments were carried out until the

effluent concentration was equal to the feeding one. In some experiments, after the
adsorption front was completely developed, the desorption of lead was measured
changing the feeding solution back to the preequilibration composition (without
lead). In this case the experiments were stopped when the lead concentration in
the effluent was below the detection limit of the analytical technique used (50



g

L

,

1

). The area of the normalised breakthrough curve, which represents the value

of the overall retention factor R, was calculated by integrating the spline smoothed
curve for each experiment. From the R value and the initial metal concentration,
the quantity of metal adsorbed on or desorbed per pore volume of the soil column
(respectively for the adsorption or desorption breakthrough front) can be calculated.

Theoretical speciation in the liquid phase was performed to distinguish between

free and complexed Pb in the different background electrolytes and pH, by the

342

M. PETRANGELI AND M. MAJONE

Figure 2. Typical tracer responses for the whole system (with column) and for tubing (without
column). Volumetric flow rate = 0.089 mL min

,

1

.

use of MINTEQA2/PRODEFA2 software (1991). Table I reports the chemical
reactions considered for the speciation calculation along with the logarithm of the
formation constants (25



C, ionic strength = 0 mol L

,

1

).

2.6. pH

EXPERIMENTS

A pH breakthrough experiment was performed following the same procedure
already described for the lead experiments. In this case 0.1 M sodium nitrate
solutions at pH = 11.0 and 2.0 were used. During the experiment the effluent solu-
tion was collected by the fraction collector and measured for the Al, Si and Fe
contents.

3. Results and Discussion

3.1. L

EAD EXPERIMENTS

Figures 3a and 3b show the lead-breakthrough curves for step experiments per-
formed with different background electrolytes (NaClO

4

, NaNO

3

and NaCl 0.1 mol

TRANSPORT OF STRONGLY RETAINED SPECIES BY COLUMN EXPERIMENTS

343

Table I

Equilibrium

considered

for

the

calculation

of

lead speciation in the different solution (MINTE-
QA2/PRODEFA2, 1991)

Reaction

Log K

Pb

2

+

+ OH

,

,

PbOH

+

–7.710

Pb

2

+

+ 2OH

,

,

Pb(OH)

2

(

aq

)

–17.120

Pb

2

+

+ 3OH

,

,

Pb(OH)

,

3

–28.060

2Pb

2

+

+ OH

,

,

Pb

2

OH

3

+

–6.360

3Pb

2

+

+ 4OH

,

,

Pb

3

(OH)

2

+

4

–23.880

Pb

2

+

+ 4OH

,

,

Pb(OH)

2

,

4

–39.699

Pb

2

+

+ NO

,

3

,

PbNOH

+

3

1.17

Pb

2

+

+ Cl

,

,

PbCl

+

1.600

Pb

2

+

+ 2Cl

,

,

PbCl

2

(

aq

)

1.800

Pb

2

+

+ 3Cl

,

,

PbCl

,

3

1.699

Pb

2

+

+ 4Cl

,

,

PbCl

2

,

4

1.380

Figure 3a. Breakthrough curves for lead-step experiments with different background electrolytes
(C

feed

= 10 mg L

,

1

Pb, electrolyte concentration 0.1 mol L

,

1

) at pH = 3.0 (3a) and pH = 5.0 (3b).

344

M. PETRANGELI AND M. MAJONE

Figure 3b.

Table II

Summary of the lead breakthrough results for the different con-
ditions (R = retention factor)

pH

[Pb

2

+

]

free

R

R/R

NaClO

4

NaCl

3.0

38.7%

339

0.70

NaNO

3

3.0

65.2%

453

0.94

NaClO

4

3.0

99.9%

484

1.00

NaCl

5.0

38.7%

1257

0.70

NaNO

3

5.0

65.2%

1632

0.90

NaClO

4

5.0

99.9%

1806

1.00

L

,

1

, respectively) at pH 3 and 5, respectively. The concentration of lead in the

feeding solution was always 10 mg L

,

1

(4.83 10

,

5

mol L

,

1

). To observe the

whole adsorption front, it was necessary to leach the soil column with a great
number of pore volumes of the feeding solution (from 600 in the case of sodium
chloride at pH 3 to about 3000 for sodium perchlorate at pH 5). In Table II the
results are summarized for the different conditions in terms of retention factors (R)
along with the percentage of free lead, as calculated by the theoretical speciation
(MINTEQA2/PRODEFA2, 1991). As it can be seen, there is a good correspon-

TRANSPORT OF STRONGLY RETAINED SPECIES BY COLUMN EXPERIMENTS

345

dence, at constant pH, between the observed retention factor and the concentration
of free lead in solution (generally assumed to be the species more available to be
sorbed) that is, the bigger the free lead concentration the higher the retention factor.
Considering the ratio (R/R

NaClO

4

) between each retention factor and that observed

in the NaClO

4

solutions where no lead complexation occurred, it can be noted how

this is practically invariant for the two experimental pH. This despite the retention
factor values at pH = 5.0 were about 4 times bigger than at pH = 3.0.

Moreover, it is noteworthy that the faster breakthrough front observed in the

NaCl solution is also clearly spreader than in the other matrixes. Being the flow
rate the same in all the experiments, the spread-out of the breakthrough curve can
not be explained in terms of kinetic effects due to lower transport and sorption
rate with respect to advection and dispersion (Schwarzenbach and Westall, 1981).
Considering that in all the matrixes the sorbing species is the same (free lead),
the observed effect can be explained in terms of different kinetics of complexation
equilibria in the liquid phase: according to this hypothesis, the spread-out of the
breakthrough front should indicate a slower dissociation kinetics of lead chloride
complexes (essentially PbCl

+

and PbCl

2

(

aq

)

) than of lead nitrate one. As a matter

of fact, slow exchange kinetics of Pb chloride complexes has been observed also
in batch tests with Chelex–100 chelating resin (Majone et al., 1996).

In Figure 4 the breakthrough curves of lead step input for 0



V/V

o



800

(change time to the preequilibration solution without lead) for two different flow
rates are reported (background electrolyte NaNO

3

). In this case after the column

has reached equilibrium with the feeding solution, the desorption of lead was
followed until its concentration in the effluent was below the detection limit of
the measurement equipment. Even though the lead concentration in the feeding
solution was quite high (100 mg L

,

1

) and the pH low (pH = 3.0) it was necessary

to leach the column with about 5000 pore volumes to observe the whole adsorption
and desorption front of the breakthrough curve.

From the figure it can be observed the typical sharp adsorption front and diffuse

desorption front of an adsorbate obeying to a convex isotherm (Schweich and
Sardin, 1981). In the range of flow rates used the two breakthrough curves superpose
so indicating that no relevant kinetic effects are related to the diffuse transport of
lead from the liquid to the solid phase and/or to the sorption reaction. This means
that local adsorption equilibrium is reached even if the flow rates utilised are quite
high (Burgisser et al., 1993). As already pointed out, this is not verified when lead
is in the chloride form.

The overall retention factor was practically equal for the adsorption and des-

orption front (R

adsorption

= R

desorption

= 126



10) and this indicates a reversible

adsorption process. Independent measurements of the lead content in the soil col-
umn at the end of the experiment confirmed that all the lead initially adsorbed was
subsequently leached out.

From the diffuse desorption front, the entire adsorption isotherm can be cal-

culated by an integration procedure (Borkovec et al., 1991) if some simplifying

346

M. PETRANGELI AND M. MAJONE

Figure 4. Breakthrough curve of lead-step experiments (C

feed

= 100 mg L

,

1

Pb, pH = 3.0, background

electrolyte 0.1 mol L

,

1

NaNO

3

)

for two different flow rates. The arrow indicates the change to the

equilibration solution. The small insert reports the calculated isotherm from the desorption front
along with the equilibrium adsorption data obtained from the experiments shown in Figure 3a.

hypotheses are assumed, such as local equilibrium conditions, negligible multi-
component effects and negligible dispersion effects. This is reasonable in the above
experiments because different flow rates did not affect breakthrough curves, the
proton concentration was enough higher than Pb concentration to maintain the pH
constant, and Pe number was higher than 50 (Burgisser et al., 1993), respectively.

In Figure 4, along with the breakthrough curve, the calculated isotherm for the

experiment is reported (small insert) in terms of moles of sorbed lead per liter
of pore volumes as function of free lead equilibrium concentration (as calculated
by theoretical speciation). In the figure, the single points refer to equilibrium data
obtained by the lead step breakthrough experiments at 10 mg L

,

1

(Figure 3a).

Adsorption at the equilibrium was obtained by the retention factors multiplied by
the initial concentration while the equilibrium concentration in the liquid phase
are expressed in terms of free lead by taking into account the different background
electrolytes. A good agreement between the calculated isotherm and the single data
can be observed.

TRANSPORT OF STRONGLY RETAINED SPECIES BY COLUMN EXPERIMENTS

347

Figure 5. Proton breakthrough curve (background electrolyte 0.1 mol L

,

1

NaNO

3

; equilibration

solution pH = 11.0; feeding solution pH = 2.0). The arrow indicates the change to the equilibration
solution. The small insert is the expansion scale of the first part of the breakthrough curve.

3.2. pH

EXPERIMENTS

As in the case of heavy metals, also H

+

generally exhibits a large retention factor.

A proton breakthrough experiment from pH = 11.0 to pH = 2.0 has been carried
out with 0.1 mol L

,

1

NaNO

3

solutions. The soil column was initially equilibrated

with the solution at pH 11 and then the feeding was switched to the pH 2 solution.
When the outlet pH did not vary more than 0.01 unit per hour, the feeding was
switched again to the pH = 11 solution. The typical result of such experiments is
shown in Figure 5 in terms of pH units (the small insert is the scale expansion of
the first part of the breakthrough front). From the figure it can be noted a sequence
of both diffuse-sharp adsorption and diffuse-sharp desorption fronts. Similar pH-
breakthrough curves have been presented for other sorbent materials where silica
was the main component (Burgisser et al., 1994; Scheidegger et al., 1994). This
behaviour corresponds to an adsorption isotherm that is S-shaped in terms of proton
concentration (Schweich and Sardin, 1981), like it is generally in the case of soils.
Moreover, a spread tail is observed at the end of both adsorption and desorption
front. This behaviour can be attributed to proton consumption in the soil dissolution
reactions. To confirm this effect, during the experiment, samples were collected by

348

M. PETRANGELI AND M. MAJONE

Figure 6a. Adsorption front for the pH breakthrough experiment reported in Figure 5 (6a) compared
with measured concentration of Al, Si and Fe in the collected samples (6b).

fraction collector at the outlet of the column. In Figure 6, the proton adsorption
front (6a) along with the measured concentration of Al, Si and Fe in the collected
samples (6b) are reported. The high Al and Si content in the collected samples,
in correspondence of the front tail, demonstrates the distinct role of dissolution
processes on the shape of the proton breakthrough curve. The order of appearance
detected in the column effluent well agrees with soil dissolution patterns reported
in the literature (Bruggenwert et al., 1991). Experiments carried out at different
flow rates (not reported) have shown that the shape of the tail was affected from the
flow rate, thus indicating that the dissolution reactions are not at the equilibrium.

4. Conclusions

Column experiments are an essential step to be carried out in the assessment of
a transport model. Generally, such studies have been extensively performed on
weakly retained species (such as major cations) or not strongly sorbing materials.
The lack of investigations, by column procedures, about the movement of strongly
retained species can be attributed to the difficulty to perform experiments which
require to be carried out for a long time under controlled conditions.

TRANSPORT OF STRONGLY RETAINED SPECIES BY COLUMN EXPERIMENTS

349

Figure 6b.

By presenting some selected experiments, this study has shown that the soil-

column technique is applicable also in the case of species which exhibit very large
retention factors (up to 5000). For these species, the use of very small soil columns
allows to observe the entire breakthrough curve (adsorption and desorption front)
in controlled conditions and in an acceptable experimental time scale.

The small size of the column and small quantity of soil did not introduce any

relevant experimental drawbacks. Typical breakthrough curves were obtained and
the results were highly reproducible. No kinetic effects (in terms of transport of
lead from the mobile to the immobile phase and sorption reaction)were observed
despite volumetric flows corresponding to high pore water velocities (0.023 +
0.072 cm sec

,

1

) were required. Even Pe number was always high enough to

allow the interpretation of diffuse fronts in terms of equilibrium isotherms. As
a consequence, the equilibrium partitioning curve could be predicted by a single
column experiment.

Lead step breakthrough experiments were carried out in different background

electrolytes, pH and lead concentrations (10 and 100 mg L

,

1

). The quantity of

metal retained by the soil column as a function of the background electrolyte was in
qualitative correspondence with the free-lead (Pb

+

2

) concentration as calculated by

theoretical speciation. By considering the free lead as the sorbed species, adsorption
data were calculated by the retention factors of the adsorption front for the different
electrolytes. These data were in good agreement with the equilibrium isotherm

350

M. PETRANGELI AND M. MAJONE

calculated from the integration of the diffuse desorption front, so confirming the
reliability of the procedure

The free lead was found as the only adsorbed species, also in batch experiments

on another solid phase, like kaolinite (Majone et al., 1993). Under this assumption,
the comparison of the shape of the breakthrough adsorption front for different
background electrolytes, allowed to obtain informations about the dissociation
kinetics of the different complexes in solution.The sensible spreadness of the
adsorption front observed in NaCl solution (compared to the NaNO

3

and NaClO

4

experiment fronts), has been ascribed to a slow dissociation of lead complexes
with chloride. However, a deeper insight about this aspect has to be achieved by
using appropriate transport codes in which both kinetics and thermodynamics of
speciation in the liquid phase are accounted for (work in progress).

Another interesting feature of such small soil columns is the possibility to

investigate the reaction of soil with protons under flow conditions. This is very
important as pH controls many soil chemical characteristics. The interpretation of
batch titration curves of soils, even in the case of a well characterized material, is
generally very difficult as the proton interaction with soil involves both adsorption
and dissolution reactions (Bruggenwert et al., 1991). By column studies it is pos-
sible, simply varying the flow rate of the feeding solution, to obtain information
about the effective acid neutralizing capacity of the soil (which is strictly related
to the rate of the flowing solution). Moreover, the possibility to measure the con-
centration of some selected ions in the column effluent offers further indications
about reactions occurring during the throughflow of acid solutions.

Acknowledgements

The experimental work has been carried out in the Institute of Terrestrial Ecology,
Soil Chemistry, Federal Institute of Technology, ETH, Schlieren, Switzerland as a
part of the Ph.D. thesis of one of the authors. We thank Prof. H. Sticher and M.
Borkovec for essential comments and enjoyable discussions on the manuscript. We
also acknowledge K. Bartmettler, C. Burgisser, M. Cernik and A. Scheidegger for
skilful lab assistance.

References

Amoozegar-Fard, A., Fuller, W. H. and Warrik, A. W.: 1984, J. Environ. Qual. 13(2), 290.
Borkovec, M., Buchter, B., Sticher, H., Behra, P. and Sardin, M.: 1991, Chimia 45, 221.
Brownawell, B. J., Hua Chen, Collier, J. M. and Westall, J. C.: 1990, Environ. Sci. Technol. 24, 1234.
Bruggenwert, M. G. M., Hiemstra, T. and Bolt, G. H.: 1991, Soil Acidity, Ulrich, B. and Sumner, M.

E. (eds.), p. 8.

Burgisser, C. S., Cernik, M., Borkovec, M. and Sticher H.: 1993, Environ. Sci. Technol. 27, 5, 943.
Burgisser, C. S., Scheidegger, A. M., Borkovec, M. and Sticher, H.: 1994, Langmuir 10, 855.
Copenhaver, S. A., Krishnaswami, S., Turekian, K. K., Epler, N. and Cochran, J. K.: 1993, Geochimica

et Cosmochimica Acta 57, 597.

TRANSPORT OF STRONGLY RETAINED SPECIES BY COLUMN EXPERIMENTS

351

Doner, H. E.: 1978, Soil Sci. Soc. Am. J. 42, 882.
Dudley, L. M., McLean, J. E., Furst, T. H. and Jurinak, J. J.: 1991, Soil Science 151(2), 121.
Dunnivant, F. M., Jardine, P. M., Taylor, D. L. and McCarthy, J. F.: 1992, Environ. Sci. Technol. 26,

360.

Griffioen, J., Appelo, C. A. J. and van Veldhuizen, M.: 1992, Soil Sci. Soc. Am. J. 56, 1429.
Grolimund, D., Borkovec, M., Federer, P. and Sticher, H.: 1995, Environ. Sci. Technol. 29, 2317.
Gschwend, P. M. and Shian-Chee Wu.: 1985, Environ. Sci. Technol. 19, 90.
Hinz, C. and Selim, H. M.: 1994, Soil Sci. Soc. Am. J. 58, 1316.
Majone, M., Petrangeli Papini, M. and Rolle, E.: mar-apr. 1993, Ingegneria Sanitaria, 59.
Majone, M., Petrangeli Papini, M. and Rolle, E.: 1996, Environmental Technology 17, 587.
Miller, D. M., Summer, M. E. and Miller, W. P.: 1989, Soil Sci. Soc. Am. J. 53, 373.
MINTEQA2/PRODEFA2, A Geochemical Assessment Model for Environmental Systems: version 3.0

EPA/600/3–91/021, 1991.

Roy, W. R.: 1993, Migration and Fate of Pollutants in Soils and Subsoils, Petruzzelli, D. and Helfferich,

F. G. (eds.), NATO ASI Series. Series G: Ecological Sciences. Vol. 32, p. 169.

Roy, W. R., Krapac, I. G., Chou, S.F. J. and Griffin, R. A.: 1991, Batch-Type Procedures for Estimating

Soil Adsorption of Chemicals, Technical Resource Document EPA/530–SW–87–006–F.

Scheidegger, A. M, Burgisser, A. S., Borkovec, M., Sticher, H., Meeussen, J. C. L. and Van Riemsdijk,

W. H.: 1994, Water Resources Research 30(11), 2937.

Schwarzenbach, R. P. and Westall, J.: 1981, Environ. Sci. Technol. 15(11), 1360.
Schweich, D. and Sardin, M.: 1981, J. Hydrol. 50, 1.
Selim, H. M., Amacher, M. C. and Iskandar, I. K.: 1989, Soil Sci. Soc. Am. J. 53, 996.
Selim, H. M., Buchter, B., Hinz, C. and Ma, L.: 1992, Soil Sci. Soc. Am. J. 56, 1004.
Sparks, D. L. and Rechcigl, J. E.: 1982, Soil Sci. Soc. Am. J. 46, 875.
Villermaux, J.: 1981, Percolation Processes, Rodrigues, A. E. and Tondeur, D. (eds.), Sijthoff:

Amsterdam, p. 83.

Voice, T. C. and Weber W. J., Jr.: 1985, Environ. Sci. Technol. 19, 789.
Wagner, J., Hua Chen, Brownawell, B. J. and Westall, J. C.: 1994, Environ. Sci. Technol. 28, 231.