An equipment and software for improved estimation of soil acidity

An equipment and software for improved
estimation of soil acidity
CZINKOTA Imre 1 , FILEP György 2 , RÉKÁSI Márk 2 , CZANIK Péter 1
1
Szent István University, Dept. of Soil Science and Agricultural Chemistry,
H-2103 Gödöllő, Páter K. u. 1, Hungary
2
Debrecen University, Dept. of Soil Science and Microbiology
H-4015 Debrecen, Böszörményi út 138, Hungary
Introduction
It is accepted in several countries to use hydrolytic acidity as suggested by Kappen
for estimating lime requirement (LR) of soils (Kappen, 1929). The essence of the
method is, that hydrogen and aluminum ions causing soil acidity are extracted with
calcium-acetate, and the acidity of the filtrate is measured with basic titration. It can
be observed, that if the remaining soil is treated again with calcium-acetate solution,
the filtrate shows again some acidity, which is lower than after the first extraction.
This phenomenon occurs in declining degree as the extraction is repeated multiple
times. (Filep and Filep, 1999).
Filep and Filep (1999) pointed out, that during repeated extraction with calciumacetate,
extractable acidity decreases exponentially with each repetition. Based on
this observation and knowing the parameters of the given function, acid quantity
can be calculated with good estimation for a large amount of extractant, and this
way also for more than once extraction.
The above mentioned phenomena can be traced back to two causes. One is that the
calcium-acetate, which is used for extraction, keeps the pH of the suspension only
approximately constant just like any other chemical buffer system. The pH of the
acetate – acetic acid system decreases with the growth of acetate/acetous acid ratio,
though it changes on a much smaller scale compared to non buffer systems. The
other cause is that during extraction, the rate of desorption from soil colloids –
mainly from inside pores – is limited. As a result of this, with limited time
extraction, only part of the reaction results show up in the extractum. For this
reason, there are always places deep in the pores, where the ratio of ions is different
from the macroscopic balance. In field the mentioned chemical processes might go
on for arbitrary long time, and pH of soil is determined by long term balance. For
this reason, currently used measurement methods can provide only rough
estimations for calculation of soil lime requirement.
1

In accordance with the previous paragraphs, the measurement method to determine
punctual potential acidity should have the following characteristics: it can be kept at
an exact pH value, it can follow a process for arbitrary long time, and it is exact
enough to allow reliable extrapolation of data to infinite time. It is quite difficult
and complicated to fulfill these requirements with a purely chemical system, so we
have built a combined, computer controlled measurement system to solve the
problem.
Materials and methods
Enhanced determination of hydrolytic acidity is based on the following theory:
During the desorption of ions causing acidity, pH is repeatedly measured in
negligible intervals compared to the full time of examination. When it is lower than
a given value, base is fed, so the system is artificially buffered. In a stand-alone
system, pH is continuously decreasing due to continuous movement of H + and Al 3+
ions into the solution, so an outer intervention to increase pH is sufficient. Since
solid and liquid stages are not separated during the use of this measurement method,
and there is no intervention to cause a sudden change in the system, the reaction can
be continued for arbitrary time. To evaluate results, it is enough to note time and
added quantities, and to extrapolate to infinite time using a function fitted on this
data. Lengthening the time of measurement can of course decrease the error of
extrapolation, so it is practical to apply longer measurement times for more exact
results.
Equipment to put the given theory in practice
The pH of soil suspension, which is continuously stirred with a proper speed, is
continuously measured with a pH-selective electrode. The soil suspension is
prepared with KCl solution, as the noise of pH-electrode is considerably higher
when distilled water is used to prepare the soil suspension. With a relatively
concentrated KCl solution, ionic strength can be kept at a relatively constant value.
Signals from the pH electrode are first amplified then digitized to provide input for
the computer, where they are transformed to pH values using preliminary
calibration. An outline of the equipment can be seen on Figure 1.
The software on the computer compares the incoming pH values in predefined time
intervals with a predefined pH limit value. If the measured pH value is lower than
the limit pH value, the computer sends a signal to an automatic burette, which adds
to the soil suspension the smallest amount of basic solution possible to add by the
burette. This method helps to keep the pH value of the system close to a predefined
2

value, with the best precision possible. After each lapse of the time interval the
software saves the time, pH and added quantity data on the disk. Saved data are
evaluated automatically using nonlinear regression analysis, when a predefined
measurement time is over, or when the measurement process is stopped by the
experimenter. Figure 2. shows the theoretical outline of the software. The
measurement system was built using Radelkis OP-0808P pH electrode, Schott
Titronic 96 automatic burette, ALTAIR BT AAD2816S amplifier and analog digital
converter and an I486 personal computer.
Properties of the examined soils are shown in Table 1.
4g of soil sample was suspended in 160 cm 3 of 0.1 M KCl solution. 0.1 M NaOH
solution was used for titration. The limit value was pH 8.2 and the time interval
between measurements was 10 sec. Maximum time of measurements was 12 hours.
Results and discussion
On Figure 3. data from sample 1. can be seen, pH values and volume of fed basic
solution in function of time.
The figure shows that at the given resolution, pH value can be considered as
constant after an initial rising phase. The almost linear rise of pH at the beginning
of measurements is to be interpreted in the following way: H + and Al 3+ ions already
in the solution or desorbing instantly, react to the fed base solution immediately, so
in this linear phase base solution is fed in each cycle of measurement evaluation. In
effect, steepness of the rising pH curve is characterized by feeding speed,
determined by cycles of the equipment. For this reason, this part cannot be used to
kinetically analyze the curve, as it increases the error of measurement. Therefore,
this part of the data is omitted during evaluation.
However, looking at the pH curve at a better resolution (Figure 4.) the essence of
the examination can be seen. This is that when base solution is added, pH value
rises sharply (even if this rise is quite small), then it decreases under the
predetermined pH limit more and more slowly with lapse of time and increase of
added base solution.
It can be seen on Figure 3. that the volume of added base solution grows almost like
an exponential associate function. Because of the sudden change of rise in the first
section, we adopt the mathematical description of Filep and Csubák (1997) to
define this growth. This description is a first order kinetic equation, which was
fitted with two different speed constants: one faster, belonging supposedly to ions
in the solution and quickly desorbable ions from the outer surface, and one slower,
belonging supposedly to ions desorbed in the inner pores. The applied formula has
the following form:
3

−c1t
−c2t
y = A ⋅( 1−
e ) + A ⋅(1
− e )
1
y amount of fed base, cm 3
t time, sec
A 1 base consumption of faster process, cm 3
A 2 base consumption of slower process, cm 3
k 1 rate constant of faster process, s -1
k 2 rate constant of slower process, s -1
2
The curve fitted on the data is shown on Figure 5. It can be seen on this figure that
measured values are approximately equal to the fitted function’s graph.
It can be seen on Table 2. that the error of fitting is around 0.3 %. Therefore these
processes can be estimated very accurately with the use of measurement data.
Measurement results after two hours (7200 sec) of reaction time account near half
of the full acidity. This corresponds mainly to the fast process (1.6 cm 3 /sample 1/).
The rest (3.05 cm 3 /sample 1/) is available only with a multiple day measurement or
curve fitting (Figure 6.).
Full acidity, and on its basis lime requirement (CaCO 3 g.kg soil -1 ) is to be calculated
with the two base consumption parameters using the following formula:
LR =
( A + A )
1
2
⋅c
2⋅
m
soil
base
⋅c
LR
-1
Lime Requirement, g.kg soil
m soil Amount of examined soil, g
c base Concentration of NaOH solution, mol.dm -3
c l CaCO 3 content, % of lime.
l
In Table 3. the lime requirement calculated by the new method is compared with
that calculated by the method of Kappen. It can be seen that the lime amounts
calculated with the new formula are considerabely different from those calculated
with the method of Kappen.
Conclusions
We built a measurement equipment to eliminate problems of such balance systems
that are created with chemical extractants effective for a limited time. The
measurement is fully automated, requires no manual intervention with the exception
4

of sample preparation, and might last for arbitrary time. To increase the efficiency
of measurement method, it is possible to collect measurement data from multiple
sources and control multiple burettes with one data collecting / controlling
computer. Precision of calculated results is very good, the error is about 0.3 %. This
is due to the large amount of data obtained with frequent measurement cycles, and
evaluation of data with nonlinear fitting. When pH is kept at a constant value,
movement of adsorbed H + and Al 3+ ions into the solution can be satisfyingly
modeled as the addition of two first order kinetic equation, where a quick process
taking place for one-two hours and a slow process taking place probably for days
can be distinguished.
The measurement theory and equipment elaborated are suitable to kinetically
examine any process of sorption or ion exchange taking place on surface of soil
colloids, if one condition is fulfilled. This condition is, that the examined matter or
parameter to measure can be somehow transformed into electric signals without
changing the system. For example, long-term exchange of polluting or nutritive
materials can be measured with ion selective electrodes, or redox properties of soils
can be measured with platinum electrodes.
References
Filep, Gy. and Csubák, M. 1997.: Kinetics of surface reactions involving proton transfer in
soil/aqueous solution systems. (in Hung.) Agrokémia és Talajtan No. 1-4. pp.159-170.
Filep, Gy. and Filep, T. 1999.: Characterization of forms potential soil acidity. (in Hung.)
Agrokémia és Talajtan No. 1-2. pp. 33-48.
Kappen, H., 1929. Die Bodenazidität. Springer Verlag. Berlin.
5

Summary
In some countries it is accepted for estimating the lime requirement (LR) to use
hydrolytic acidity suggested by Kappen, which is based on a single time extraction
of hydrogen- and aluminum-ion with calcium-acetate. We could achieve more
accurate results, if we measure the total surface acidity (TSA) of soils. For this
reason it is an improvement to use a direct measurement method and equipment,
which is suitable to estimate the results of long-term processes and TSA, via
investigating the kinetic properties of desorption.
The method of measurement: A pH electrode is dipped into continuously stirred soil
suspension, containing background salt (e.g. KCl), and it is connected to a computer
using an amplifier and A/D converter. A computer program has been developed that
controls an automatic burette, which adds the base solution into the system if pH is
less than the pre-adjusted value (e.g. pH 6.5 or pH 6.8) and stops adding if pH
reaches this value.
For evaluation, the amount of added base vs. time data series can be used. With
increasing time the amount of added base keeps to a constant (asymptotic) value.
The program fits an exponential associate function on measured data, and outputs
the asymptotic value connecting to infinite time, which can be used to calculate LR.
Let us suppose that there are two processes with different rate, in this case the
function can be created as the addition of two first order kinetic sub-processes.
−c1t
−c2t
y = b1
⋅ ( 1−
e ) + b2
⋅ (1 − e )
The faster process that takes place on the outer surfaces features easily removable
acidity, and the slower process probably describes processes inside the deeper
pores. The fitting error of parameters is about 0.3 %, which means, that the TSA
value, based on these measured data and method can be estimated with high
accuracy.
The measurement is fully automated. The evaluation is based on extrapolation so
the precision of results increases with the number of measurement points and the
length of measurement time. Depending on the application, a quick measurement
with approximate results or a longer measurement with more precise results can be
chosen.
6

Stirrer
Elektrode voltage (analog signal) pH electrode Soil suspension
Amplifyer &
AD converter
Elektrode voltage (digital signal)
Reagent solution
Computer
Automatic
burette
Burette control (digital signal)
Feeded volume (digital signal)
Figure 1. Theoretical outline of the equipment
Start
Calibrate
No
Yes
Measurement
Measurement
No
Standard1
Yes
No
Timer
Yes
Ask standard2
Ask standard1
No
End of meas.
Yes
Meas. voltage
Meas. voltage Meas. voltage
No
The other standard OK
Yes
Calculating pH-voltage function
Save caibration data
Store all data
Close files
Fitting function
Output parameters
End
Feed one step
Calculating pH
No
Yes pH

Figure 3. Graphical display of data saved during measurement (Sample 1)
Figure 4. Magnified view of a part of the measured curves (Sample 1)
8

5,0
4,5
4,0
3
Added NaOH, cm
3,5
3,0
2,5
2,0
1,5
1,0
0,5
0,0
Measured value
Calculated value
0 5000 10000 15000 20000 25000 30000
Time, sec
Figure 5. Graph of the fitted function (Sample 1)
9

5,0 A
4,5
Asymptotic value
4,0
3,5
2 hours
Added NaOH, cm 3
3,0
2,5
2,0
1,5
1,0
0,5
Total kinetic process
Faster kinetic process
Slower kinetic process
Measured value
0,0
0 5000 10000 15000 20000 25000 30000
Time, sec
B
2,0
Asymptotic value
Added NaOH, cm 3
1,5
1,0
0,5
2 hours
Total kinetic process
Faster kinetic process
Slower kinetic process
Measured value
0,0
0 5000 10000 15000 20000 25000 30000 35000 40000
Time, sec
Figure 6. The two parts of the fitted function. A: Sample 1; B: Sample 3
10

N o of sample pH (KCl) OM% Clay+Silt*%
1 4.7 1.2 39.7
2 4,0 1,9 69,0
3 3,7 0,9 11,0
4 4,1 1,1 20,1
5 4,2 3,1 68,0
*Clay+Silt% < 0.02 mm
Table 1. Some properties of the examined soils
N o of sample Parameter Value Error Error %
A 1 cm 3 1.6054 0.007 0.43
1 k 1 s -1 0.00336 0.00004 1.19
A 2 cm 3 3.0572 0.0105 0.34
k 2 s -1 0.00008 8.41 10 -7 1.05
R 2 0.9976
A 1 cm 3 2.25674 0.00435 0.19
2 k 1 s -1 0.00167 6.6182 10 -6 0.39
A 2 cm 3 1.63498 0.00385 0.23
k 2 s -1 0.0001 8.3434 10 -7 0.83
R 2 0.9988
A 1 cm 3 0.63475 0.00195 0.30
3 k 1 s -1 0.00536 0.00007 1.30
A 2 cm 3 1.61516 0.00668 0.41
k 2 s -1 0.00005 4.4444 10 -7 0.88
R 2 0.9986
A 1 cm 3 1.13408 0.00362 0.31
4 k 1 s -1 0.00313 0.00003 0.95
A 2 cm 3 1.27671 0.00601 0.47
k 2 s -1 0.00007 1.0355 10 -6 1.47
R 2 0.9971
A 1 cm 3 1.88475 0.00652 0.34
5 k 1 s -1 0.00196 0.00002 1.02
A 2 cm 3 1.3581 0.00662 0.48
k 2 s -1 0.00008 1.5049 10 -6 1.88
R 2 0.9951
Table 2. Parameter values of curve fitting
11

N o of sample LR 1 LR 2
1 5.82 3.65
2 4.86 5.27
3 2.81 3.70
4 3.01 4.96
5 4.05 6.51
Table 3. The lime requirement calculated with the two different methods
LR1: The lime requirement calculated with the new formula
LR2: The lime requirement calculated with the method of Kappen
12