Phosphate Sorption Capacities of Some Soils in Madhupur Tract of Bangladesh: Role of Aluminium, Iron and Clay Contents

Abstract

An experiment was carried out to study the phosphate sorption capacities of five soil series, namely Noadda, Salna, Gerua, Belabo and Chandra in the Madhupur Tract of Bangladesh. Nine composite soil samples representing the five soil series were found to exhibit varied sorption capacities. However, most of the soils showed the highest amount of phosphate sorption at 25 µg P/mL of application. The sorption data for most of the soils showed a good fit into Langmuir adsorption isotherm. The values of Langmuir-derived maximum sorption capacity (Q 0) varied widely and was in the following order: Belabo, site 9 (1000 µg/g) > Gerua, site 5 (995 µg/g) > Noadda, site 4 (500 µg/g) > Noadda, site 7 (250 µg/g) = Belabo, site 2 (250 µg/g) > Salna, site 6 (167 µg/g) = Gerua, site 8 (167 µg/g) > Salna, site 3 (143 µg/g) > Chandra, site 1 (56 µg/g). The highest sorption capacity of Belabo (site 9) soil could be attributed to its total Al, total Mn, and amorphous Fe contents. The calculated Langmuir coefficient b values were found to be higher than the threshold value of 0.07 mL/µg for the soils in question meaning the soils are not prone to loss via surface and subsurface flow. A strong relationship (r = 0.799**) was obtained between phosphate sorption at 10 µg P/mL of application and the clay content in the studied soils. The aluminium content was also found to have a significant relationship (r = 0.684*) with the phosphate sorption at 10 µg P/mL of application. However, the total free iron oxides and the amorphous Fe did not correlate well with the phosphate sorption at 10 µg P/mL of application. Therefore, the clay and total aluminium contents were primarily responsible for the phosphate sorption capacity of the acidic soils of the Madhupur tract under investigation.

Full text

National University Journal of Science Volume 3-5, 2018: 1-18
Phosphate Sorption Capacities of Some Soils in Madhupur Tract
of Bangladesh: Role of Aluminium, Iron and Clay Contents
Mohammad Enayet Hossain, Fatema Tuj Johra Sania and Md. Harunor Rashid Khan
Abstract
An experiment was carried out to study the phosphate sorption capacities of five
soil series, namely Noadda, Salna, Gerua, Belabo and Chandra in the Madhupur
Tract of Bangladesh. Nine composite soil samples representing the five soil series
were found to exhibit varied sorption capacities. However, most of the soils
showed the highest amount of phosphate sorption at 25 µg P/mL of application.
The sorption data for most of the soils showed a good fit into Langmuir adsorption
isotherm. The values of Langmuir-derived maximum sorption capacity (Q0) varied
widely and was in the following order: Belabo, site 9 (1000 µg/g) > Gerua, site 5
(995 µg/g) > Noadda, site 4 (500 µg/g) > Noadda, site 7 (250 µg/g) = Belabo, site
2 (250 µg/g) > Salna, site 6 (167 µg/g) = Gerua, site 8 (167 µg/g) > Salna, site 3
(143 µg/g) > Chandra, site 1 (56 µg/g). The highest sorption capacity of Belabo
(site 9) soil could be attributed to its total Al, total Mn, and amorphous Fe
contents. The calculated Langmuir coefficient b values were found to be higher
than the threshold value of 0.07 mL/µg for the soils in question meaning the soils
are not prone to loss via surface and subsurface flow. A strong relationship (r =
0.799**) was obtained between phosphate sorption at 10 µg P/mL of application
and the clay content in the studied soils. The aluminium content was also found to
have a significant relationship (r = 0.684*) with the phosphate sorption at 10 µg
P/mL of application. However, the total free iron oxides and the amorphous Fe did
not correlate well with the phosphate sorption at 10 µg P/mL of application.
Therefore, the clay and total aluminium contents were primarily responsible for
the phosphate sorption capacity of the acidic soils of the Madhupur tract under
investigation.
Keywords: Phosphate sorption capacity, maximum sorption capacity,
aluminium and clay contents, Madhupur tract.
Introduction
Phosphorus (P) is the tenth most abundant element in the earth’s crust. Total soil P
usually ranges from 100 to 2000 mg P/kg soil representing approximately 350 to 7000 kg
P/ha in the surface 25 cm of the soil, although only a small portion of this P is
immediately available for crop uptake [1]. The availability of P in soils varies greatly and
factors like the nature of parent material, the degree of weathering, and content of organic
matter, etc. control the availability of phosphorus.
Department of Soil, Water and Environment, University of Dhaka, Dhaka 1000, Bangladesh.
 Corresponding Author, E-mail:enayetswe@du.ac.bd

2 Mohammad Enayet Hossain, Fatema Tuj Johra Sania and Md. Harunor Rashid Khan
The aspect which makes P sorption capacity of soils more important to soil and
environmental scientists is its nonrenewable nature. World phosphorus production rate is
projected to go down from around 2033, and if that happens the food security in the
world will be in jeopardy [2]. Accumulation of P in the soil results from applications of
animal manures, bio-solids or chemical fertilizer more than the requirement of crops;
phosphate accumulation in soils increases the risk of P movement to surface and ground
waters. Excessive phosphorus in the form of fertilizers or manures may find its way into
nearby water bodies giving rise to eutrophication phenomenon and toxic algal bloom
problem [3,4,5].
Many of the soils in Bangladesh are not fertile enough to maintain soil P concentrations
at an optimum. Studies conducted by the soil scientists of Bangladesh revealed that
Bangladesh soils are generally low in total phosphorus [6]. The capacity of the soil to
sorb phosphorus greatly influences plant yield responses to applied P fertilizers. To
correct phosphorus deficiencies, P is applied to agricultural fields in inorganic and/or
organic fertilizers. Soluble P-fertilizers applied to the soil are very rapidly changed to
less soluble compounds which, with time, become less and less available to plants [7].
Even though there is a fertilizer recommendation guide in Bangladesh for applying
phosphorus fertilizer, the farmers do not generally follow the fertilizer guide [8].
Consequently, fertilizers are applied arbitrarily and not based on any science, which not
only poses an environmental concern but also lowers the benefit/cost ratio. Therefore,
knowledge of the P sorption capacity of soils is required to provide an accurate estimate
on the P fertilizer requirements of soils.
Phosphate sorption data have been used to estimate P requirements for plant growth and
yield and for ensuring phosphorus is not in excess of requirements. The sorption
isotherms can be used to approximate the quantity of P that must be added in the soil to
raise the P concentration in the soil solution at equilibrium to a desired, or maximum
value [9]. Quantitative descriptions of P sorption by soils have often been made with the
Langmuir and Freundlich equations.
In Bangladesh, very few researches have been done on the phosphate sorption capacity of
soils. To date, only a few soils have been properly characterized in terms of phosphate
sorption requirements. Therefore, there is a need for bringing the un-investigated soils
under study which will help us to manage our rapidly-depleting phosphate resources. The
present experiment was conducted with a view to studying some Madhupur tract soils in
terms of phosphate sorption capacities. The specific objectives of the research were: (1)
to characterize some representative soils of Madhupur Tract of Bangladesh with respect
to phosphate sorption characteristics, and (2) to study the effects of different parameters
of soil on phosphate sorption characteristics.
Materials and Methods
2.1 Collection and Processing of Soil Samples
Nine sites were selected for soil sample collection. The sites covered five soil series of
the Madhupur Tract (Table 1). The soil series are Noadda, Salna, Gerua, Belabo, and

Phosphate Sorption Capacities of Some Soils in Madhupur Tract of Bangladesh 3
Chandra. The soils were designated as Chandra (site 1), Belabo (site 2), Salna (site 3),
Noadda (site 4), Gerua (site 5), Salna (site 6), Noadda (site 7), Gerua (site 8), and Belabo
(site 9). The soils selected from different locations had a wide range of characteristics
(Table 1). The soil samples were collected from the surface horizon at a depth of 0-15 cm
on the basis of the composite sampling method. Samples were air-dried and ground to
pass through 2 of mm and 0.2mm sieves and finally kept in clean plastic containers
carefully with avoidance of any kind of contamination. The 2 mm size samples were used
for mechanical and some chemical analyses while 0.2 mm size samples were used for
iron fractionation process.
Table 1: Correlation of the soil samples with the USDA soil classification system
Site
Soil Series
Location
Land Type
USDA Soil Family
1
Chandra
Bhaluka, Mymensingh
Medium High Land
Aeric Paleaquult
2
Belabo
Mirzapur, Tangail
High Land
Typic Paleudult
3
Salna
Shakhipur, Tangail
Medium High Land
Ultic Ustochrept
4
Noadda
Shakhipur, Tangail
High Land
Typic Paleudult
5
Gerua
Shakhipur, Tangail
High Land
Ultic Ustochrept
6
Salna
Kaliakoir, Gazipur
Medium High Land
Ultic Ustochrept
7
Noadda
Kaliakoir, Gazipur
Medium High Land
Typic Paleudult
8
Gerua
Kaliakoir, Gazipur
High Land
Ultic Ustochrept
9
Belabo
Kaliakoir, Gazipur
High Land
Typic Paleudult
Source: FAO/UNDP [10]
2.2 Methods for Soil Analysis
After collection, soils were processed, and the background analyses were done, and the
batch studies were conducted. The background analyses were done for the physical, and
chemical properties of the soil samples. Particle size analysis of the soils was performed
using the hydrometer method as described by Day [11]. Textural classes were determined
by Marshall’s Triangular Co-ordinates. Soil pH was measured electrochemically using a
combined glass electrode pH meter maintaining a soil-solution ratio of 1:2.5. EC of soil
samples was measured using an electrical conductivity meter maintaining a soil to water
ratio of 1:5. Soil organic carbon was determined by Walkley and Black’s wet oxidation
method as outlined by Jackson [12]. The organic matter percentage was determined by
multiplying the content of percent organic carbon with a factor of 1.724 (van Bemmelen
factor). Total aluminum and manganese were determined by digesting the soil samples
with HNO3-HClO4 acid and then measuring by flame atomic absorption spectroscopy
(AAS) with an Agilent Technologies 200 series AA. As the collected soil samples were
acidic (the pH ranged from 4.83 to 6.23) Bray and Kurtz no. 1 extractant was used for
available P extraction [13]. Phosphorus content in the extract was determined by the
ascorbic acid blue color method of Murphy and Riley [14] using a Spectrophotometer at

4 Mohammad Enayet Hossain, Fatema Tuj Johra Sania and Md. Harunor Rashid Khan
880 nm. Total free iron oxide was extracted following the procedure of Holmgren [15].
Poorly crystalline or active iron oxide was determined by using NH4 oxalate-oxalic acid
(pH 3.0) [16]. The extracted total free iron oxide and poorly crystalline or active iron
oxide were then determined by flame atomic absorption spectroscopy (AAS).
2.3 Phosphate Sorption Experiment
Phosphate sorption experiment was carried out following the procedure described by
Nair et al. [17]. Briefly, one gram of 2-mm sieved air-dried soil was taken into a 50 mL
centrifuge tube containing 20 ml 0.01M calcium chloride (CaCl2). Nine initial P
concentrations, namely 0, 0.5, 1, 2.5, 5, 10, 20, 25 and 40 µg/mL were added separately
to each centrifuge tube using a soil to solution ratio of 1:20 (w/v). The resultant P
contents were 0, 10, 20, 50, 100, 200, 500, and 800 µg/g soil. The centrifuge tubes were
then shaken and equilibrated for 16 h. Three drops of toluene were added to the
centrifuge tubes to control the activity of microorganisms. After the equilibration period,
the mixtures were centrifuged, and the supernatants were analyzed for phosphate
following Murphy and Riley method [14]. The difference between the concentration of
PO4
3--P added in the initial solution and the concentration of PO4
3--P at equilibrium was
used to measure the sorbed P. Each treatment was replicated three times.
2.4 Sorption Models
The sorption isotherm data of P can be modeled into the Langmuir and Freundlich
equations. The Langmuir model assumes that there is no interaction between the
adsorbate molecules and the adsorption is localized in a monolayer [26-28].
𝑄𝑒=𝑄0𝑏𝑐𝑒
1+𝑏𝑐𝑒
(1)
Where, 𝐶𝑒 (µg/mL) is the equilibrium concentration of the remaining ions in the
solution; 𝑄𝑒 (µg/g) is the amount of ions adsorbed per mass unit of adsorbent at
equilibrium; 𝑄0 (µg/g) is the amount of ions at complete monolayer which is also called
the P sorption maximum, and 𝑏 (mL/µg) is the Langmuir constant related to the affinity
of binding sites which is a measure of the energy of adsorption. The Langmuir equation
can be described by the linearized form:
𝐶𝑒
𝑄𝑒
= 1
𝑄0
𝐶𝑒+1
𝑄0𝑏 (2)
The Langmuir constants 𝑄0 and 𝑏 are calculated from the slope (1/𝑄0) and intercept
(1/𝑄0𝑏) of the plots of 𝐶𝑒
𝑄𝑒
(Y-axis) and 𝐶𝑒 (X-axis). The maximum phosphorus buffering
capacity (MPBC) of the soil (the increase in sorbed P per unit increase in final solution P
concentration), is estimated from the product of Langmuir constants 𝑄0 and 𝑏.
The Freundlich isotherm model is an empirical relationship describing the adsorption of
solutes from a liquid to a solid surface, and the model assumes that in the adsorption
process different sites with different adsorption energies are involved [29]. The

Phosphate Sorption Capacities of Some Soils in Madhupur Tract of Bangladesh 5
Freundlich isotherm (Equation 3) [19] can be applied for non-ideal sorption on
heterogeneous surfaces and multilayer sorption. It also assumes that the adsorbent has a
different affinity for adsorption. The Freundlich equation is expressed as:
𝑄𝑒 = 𝐾𝐹𝐶𝑒1/𝑛 (3)
where 𝐾𝐹 and 𝑛 are Freundlich constants indicating adsorption capacity (µg/g) and
intensity or the favorability of the sorption process, respectively.
To determine the constants 𝐾𝐹 and 𝑛, the Freundlich equation can be described by the
linearized form:
𝑙𝑜𝑔 𝑄𝑒 = 𝑙𝑜𝑔 𝐾𝐹+1
𝑛𝑙𝑜𝑔 𝐶𝑒 (4)
Values of 𝐾𝐹 and 𝑛 are calculated from the intercept and slope of the linear plots of
𝑙𝑜𝑔 𝑄𝑒 vs 𝑙𝑜𝑔 𝐶𝑒.
2.5 Statistical Analysis
MS-Excel and Minitab17 computer programs were used to analyze the experimental
data.
Results and Discussion
Physical and Chemical Properties of Soils
Data obtained through analyses of soils are given in Tables 2 and 3.
Table 2: Some physical and chemical properties of the soil samples
Site
Soil
pH
Particle Size Distribution (%)
Textural Class
EC (µS/cm)
Sand
Silt
Clay
1
Chandra
5.53
24.25
59.5
16.25
Silt Loam
27.5
2
Belabo
4.95
21.75
42
36.25
Clay Loam
44
3
Salna
5.76
19.25
57
23.75
Silt Loam
33.5
4
Noadda
5.23
18.75
49.5
31.75
Silty Clay Loam
65.5
5
Gerua
5.38
21.75
49.5
28.75
Silty Loam
58.4
6
Salna
6.23
29.25
42
28.75
Clay Loam
30.5
7
Noadda
5.75
19.25
52
28.75
Silty Clay Loam
30.5
8
Gerua
5.47
26.25
42
31.75
Clay Loam
25.7
9
Belabo
4.91
29.25
34.5
36.25
Clay Loam
87

6 Mohammad Enayet Hossain, Fatema Tuj Johra Sania and Md. Harunor Rashid Khan
Table 3: Some chemical properties of the soil samples used in the present study
Site
Soil
OC
OM
Total Al
Av. P
Total Mn
Free Fe
Active Fe
%
µg/g
1
Chandra
1.16
1.99
1.13
11.73
54.11
905
2469.12
2
Belabo
1.85
3.19
3.16
2.01
139.13
1153
3495.12
3
Salna
1.36
2.35
1.64
2.98
86.03
1007
3321.12
4
Noadda
1.30
2.24
2.06
4.56
85.50
1244
2649.12
5
Gerua
1.34
2.32
2.12
3.38
90.30
1239
2833.92
6
Salna
1.11
1.92
2.80
2.56
524.68
875
5019.12
7
Noadda
0.20
0.34
2.25
1.94
325.69
987
5088.72
8
Gerua
0.94
1.61
2.43
2.32
301.86
1844
4188.72
9
Belabo
1.25
2.15
3.21
3.61
375.59
1090
4339.92
Phosphate Sorption Batch Studies
Phosphate sorption batch studies were done by equilibrating soils with 0.01 M calcium
chloride solution containing progressively higher concentrations (0 to 40 µg/mL) of
phosphorus. The sorbed phosphate content was then calculated from analysis of the
equilibrium solutions.
Phosphate Sorption in Soils
Results from phosphate sorption studies are presented in Table 4. Apart from 0 µg
P/mL application, phosphate was sorbed by the studied soils with all other rates.
Some desorption was noticed in all the soils when no P was applied. That phosphate
gets desorbed when equilibrated with phosphate -free water is a common phenomenon
and was observed by a number of researchers [8, 18 -20]. With few exceptions, the
amount of P sorption progressively increased with increasing P application in all the
soils. This finding is in agreement with the findings of other researchers [6, 18-21].
In Chandra series, phosphate sorption increased gradually up to the rate of 10 µg
PO4
3--P/mL. Phosphate sorption increased up to 20 µg PO 4
3--P/mL in Belabo (site 2),
Salna (site 6), Noadda (site 7) and up to 25 µg PO4
3--P/mL in Salna (site 3), and
Gerua (site 8) and beyond that concentrations, the amount of sorption decreased
substantially. The rest of the soil samples showed a trend of increasing P sorption
with increasing P application.

Phosphate Sorption Capacities of Some Soils in Madhupur Tract of Bangladesh 7
Table 4: Phosphate sorption at different rates of P application
Site
Soil series
P added (µg/mL)
0.5
1
2.5
5
10
20
25
40
P sorbed (µg/g)
1
Chandra
15.45
25.96
48.82
67.58
96.23
86.55
67.14
50.43
2
Belabo
17.95
38.17
98.18
181.46
280.56
374.88
350.61
226.07
3
Salna
18.35
37.76
82.08
185.7
242.60
247.37
267.63
119.61
4
Noadda
17.38
37.03
89.52
152.94
220.00
296.65
333.68
365.74
5
Gerua
18.07
35.85
90.22
112.32
222.50
452.75
530.21
614.89
6
Salna
18.13
36.62
71.91
98.58
196.44
213.24
157.28
136.16
7
Noadda
18.20
38.10
90.83
149.83
201.33
238.00
234.32
227.04
8
Gerua
18.20
37.97
87.10
123.91
195.77
232.99
234.52
135.65
9
Belabo
18.03
37.29
88.17
150.32
245.22
355.77
432.13
535.42
Phosphate Sorption Isotherms
Sorption isotherms are mathematical models that describe the distribution of the sorbate
species among liquid and sorbent making a set of assumptions that are mainly related to the
heterogeneity/homogeneity of sorbents, the type of coverage, and the possibility of
interaction between the sorbate species. The phosphate sorption isotherms of the studied
soils were constructed by plotting the equilibrium concentration of phosphate (Peq) against
the amount of phosphate sorbed (Psorb) (Figure 1). From the sorption isotherms, it is
obvious that the studied soil series showed different trends in sorbing phosphorus from the
soil solution. Almost all the soils behaved similarly exhibiting increased phosphate sorption
with an increase in P application up to 25 µg/mL. At the highest loading (40 µg/mL), P
sorption decreased in all the soils except in Noadda (site 4), Gerua (site 5), and Belabo (site
9). The different sorption mechanisms could be accounted for this phenomenon. There are
two mechanisms whereby phosphate anions get sorbed onto soil micelles. More
specifically, phosphate sorption takes place through specific and non-specific sorption
processes. Specific sorption is related to high-affinity sorption sites and micropore
diffusion sorption [22, 23, 24], whereas non-specific sorption is associated with low-
affinity sites. While specific sorption is predominant during the fast sorption stage, non-
specific sorption predominates during the slow sorption stage. The percentage of sorption
via non-specific sorption process increases when phosphate loading increases [25].
In the present study, at 40 µg/mL of PO4
3--P application, the amount of phosphate
sorption decreased in most of the soils than what was sorbed at 20-25 µg/mL of PO4
3--P
application. Non-specific adsorption process might have played the predominant role for
the additional P loading when P application was increased from 25 to 40 µg/mL.
Therefore, even though the amount of phosphate sorption was higher during the fast-
sorption stage at higher loading (40 µg/mL), in the slow sorption process, the phosphate
was initially sorbed and then desorbed. As a result, more phosphate was sorbed at 20-25
µg/mL than at 40 µg/mL of PO4
3--P application. The highest amount of phosphate sorbed
in the studied soils was compared and it is evident that each soil had a different capacity
to sorb P. Gerua, site 5 (614.89 mg/kg) had the highest maximum adsorption capacity,
followed by Belabo, site 9 (535.42 mg/kg) and Noadda, site 4 (365.74 mg/kg).

8 Mohammad Enayet Hossain, Fatema Tuj Johra Sania and Md. Harunor Rashid Khan
The sorption isotherm data of P was modeled into the Langmuir and Freundlich
equations. When the phosphate sorption data were plotted in the Langmuir adsorption
isotherm by taking Ce/Qe against Ce (Figure 2, Table 5), all the soils showed a good fit
except Gerua soil series (site 5) and Belabo soil series (site 9). R2 value was the highest
in Noadda soil series (site 7), followed by Noadda (site 4)>Salna (site 6)>Belabo (site
2)>Belabo (site 9)> Chandra (site 1)>Gerua (site 8)>Salna (site 3)>Gerua (site 5).
Sorption parameters were calculated from the Langmuir equation (Table 6).
Table 5: Fitted Langmuir and Freundlich equations using sorption data from the studied soils
Site
Soil Series
Langmuir equation
Freundlich equation
Estimated Equation
R2
Estimated Equation
R2
1
Chandra
y = 0.018x - 0.025
0.951
y = 0.245x + 1.564
0.678
2
Belabo
y = 0.004x - 0.003
0.961
y = 0.330x + 2.128
0.703
3
Salna
y = 0.007x - 0.012
0.907
y = 0.294x + 1.996
0.633
4
Noadda
y = 0.002x + 0.004
0.991
y = 0.432x + 2.017
0.915
5
Gerua
y = 0.001x + 0.007
0.833
y = 0.538x + 2.067
0.947
6
Salna
y = 0.006x - 0.002
0.973
y = 0.319x + 1.849
0.854
7
Noadda
y = 0.004x + 0.001
0.998
y = 0.335x + 1.987
0.865
8
Gerua
y = 0.006x - 0.007
0.937
y = 0.311x + 1.932
0.813
9
Belabo
y = 0.001x + 0.005
0.959
y = 0.489x + 2.077
0.969
Table 6: Sorption parameters derived from Langmuir and Freundlich isotherm models
Site
Soil Series
Langmuir equation
Freundlich equation
Q0 (µg/g)
b (mL/µg)
MPBC (mL/g)
Kf
1/n
1
Chandra
56
0.72
40.00
36.64
0.25
2
Belabo
250
1.33
332.50
134.28
0.33
3
Salna
143
0.58
82.36
99.08
0.29
4
Noadda
500
0.50
250.00
103.99
0.43
5
Gerua
995
0.14
140.00
116.68
0.54
6
Salna
167
-3.00
-50.00
70.63
0.39
7
Noadda
250
4.00
1000.00
97.05
0.34
8
Gerua
167
-0.86
-143.33
85.51
0.34
9
Belabo
1000
0.20
200.00
119.4
0.49
The maximum amount of sorbate sorbed by the soil (Q0) was calculated for all the soils
and Belabo series (site 9) was found to sorb the highest amount of P from a solution
according to Langmuir equation. For the soils in question, the Langmuir-equation-
derived sorption maximum (Q0) ranged from 56 to 1000 µg/g. The maximum phosphate
sorption capacity was in the following order: Belabo (site 9) > Gerua (site 5) > Noadda
(site 4) > Noadda (site 7) = Belabo (site 2) > Gerua (site 8) = Salna (site 6) > Salna (site
3) > Chandra (site 1). The highest sorption capacity of Belabo soil series (site 9) could be
attributed to its Al, Mn and amorphous iron content. Belabo series (site 9) had the highest
amount of Al (3.21%) and much higher Mn (375.99 µg/g) and amorphous iron content
(4339.92 µg/g) among the investigated soils. On the other hand, Chandra series (site 1)

Phosphate Sorption Capacities of Some Soils in Madhupur Tract of Bangladesh 9
possessed the lowest amount of Al (1.13%), Mn (54.11 µg/g) and amorphous iron
(2469.12 µg/g) content and accordingly the sorption capacity was the lowest in that soil.
The Langmuir b coefficient, expressing the bonding energy of P sorption, ranged
from -3.00 to 4.00 mL/µg. Varied values of b indicate that bonding energy associated
with P sorption are different for the studied soils. The binding energy (b) was the
highest for soil Noadda (site 7) (4.00 mL/µg) and the lowest for Salna series (site 6)
(-3.00 mL/µg). Binding energy has enormous implications because it determi nes the
susceptibility of a given soil vis-à-vis phosphate loss via surface and subsurface flow.
The binding energy values for Salna (site 6) and Gerua (site 8) were below the
threshold value of 0.07 mL/µg. However, negative values of coefficient b are
physical nonsense and impossible. It might have arisen from the deviations in the
experiment at lower concentrations. Apart from those two soils (Salna, site 6 and
Gerua, site 8), the rest of the soils are not prone to phosphate loss through the surface
and sub-surface runoff.
The adsorption isotherms were also examined by the linear form of the Freundlich
adsorption equation by plotting log Qe against log Ce (Figure 3, Table 5). Belabo (site
9), Gerua (site 5), and Noadda (site 4) soil series’ showed a good fit to the equation.
The R2 value was the highest for Belabo series (site 9) (0.969), followed by Gerua
(site 5) (0.947), Noadda (site 4) (0.915), Noadda (site 7) (0.865), Salna (site 6)
(0.854), Gerua (site 8) (0.813), Belabo (site 2) (0.703), Chandra (site 1) (0.678) and
Salna (site 3) (0.633).
Sorption parameters were also calculated from the Freundlich equation (Table 6). The
Freundlich exponent, 1/n, represents the heterogeneity of surface sites having
different affinities for phosphate retention by soil. The highest 1/n was found in
Gerua series (site 5) (0.538), followed by Belabo (site 9) (0.489), Noadda (site 7)
(0.432), Noadda (site 7) (0.335), Gerua (site 8) (0.335), B elabo (site 2) (0.33), Salna
(site 6) (0.319), Salna (site 3) (0.294) and Chandra (site 1) (0.245) (Table 6).
Freundlich coefficient Kf, which is a measure of the amount of P sorption sites,
ranged from 36.64 to 134.28 µg/g in the studied soils. According to Freundlich
parameter Kf, Belabo (site 2) had the highest sorption sites followed by Belabo (site
9), Gerua (site 5), Noadda (site 4), Salna (site 3), Noadda (site 7), Gerua (site 8),
Salna (site 6) and Chandra (site 1). The fact that Belabo soils have the highest
amount of sorption sites can be explained by the amount of total aluminium content
present in the soil. This finding can be substantiated by the findings of Singh and
Gilkes [30]. In a phosphate sorption experiment, the researchers found all fo rms of
extractable aluminium were significantly and positively correlated with the
Freundlich parameter, K f.
Maximum P buffering Capacity (MPBC)
Maximum P buffering capacity (MPBC) is a product of Langmuir model Q0 and b [31,
32]. The MPBC data shown in Table 6 indicate that Noadda (site 7) will need the highest

10 Mohammad Enayet Hossain, Fatema Tuj Johra Sania and Md. Harunor Rashid Khan
amount of phosphate fertilizer compared to other soils. Gradually decreasing amount of P
fertilization should be maintained in Belabo (site 2), Noadda (site 4), Belabo (site 9),
Gerua (site 5), Salna (site 3) and Chandra (site 1) for optimum plant growth. Following
this trend, someone should be able to maintain an optimum concentration of phosphorus
for crop growth with less expense.
a)
b)
c)
d)
Figure 1(a-d): Phosphate sorption isotherm of studied soil samples.
Chandra (Site 1)
Belabo (Site 1)
Salna (Site 3)
Noadda (Site 4)

Phosphate Sorption Capacities of Some Soils in Madhupur Tract of Bangladesh 11
e)
f)
g)
h)
i)
Figure 1(e-i): Phosphate sorption isotherm of studied soil samples.
Gerua (Site 5)
Salna (Site 6)
Noadda (Site 7)
Belabo (Site 8)
Belabo (Site 9)

12 Mohammad Enayet Hossain, Fatema Tuj Johra Sania and Md. Harunor Rashid Khan
Figure 2: Sorption isotherm fitted to Langmuir equation

Phosphate Sorption Capacities of Some Soils in Madhupur Tract of Bangladesh 13
Figure 3: Sorption isotherms fitted to Freundlich equation

14 Mohammad Enayet Hossain, Fatema Tuj Johra Sania and Md. Harunor Rashid Khan
Phosphate Sorption in Relation to Soil Properties
Different soil properties and phosphate sorption parameters were also tested among
themselves for relationships. Pearson correlation analysis was performed between the
amount of phosphate sorbed at 10 µg P/mL of application and different sorption
parameters and different soil properties (Table 7). The 10 µg P/mL application rate was
chosen for correlation analysis because at this concentration, the sorption behavior was
more or less similar across the soils selected for the study. Correlation analysis was also
done among the sorption parameters and the different soil properties. Significant
correlations with r > 0.5 (therefore explaining 25% of total variation) were only
considered for further discussion. Significance was determined based on whether p-
values are <0.05 or not.
There is a strong relationship (r = 0.799**) between P sorption at 10 µg P/mL of
application and the clay content. Statistically significant relationship was also found
between P sorption at 10 µg P/mL of application and the aluminum content (r = 0.684*).
These relationships are confirming that P sorption was primarily governed by the amount
of clay and aluminum content in the studied soils. The phosphate sorption coefficient Kf
correlated strongly with phosphate sorption implying that phosphate sorption increases as
sorption site increases (Kf is a measure of the number of P sorption sites in a soil). The
amount of clay content was found to correlate strongly with P sorption coefficient Kf
meaning that the clay content controls the number of sorption sites in the studied soils.
Two phosphate sorption coefficients Q0 (maximum sorption capacity) and 1/n correlated
strongly (r = 0.927***) which means that as the sorption capacity of soil increases, the
slope 1/n increases, and more sorption occurs; this relationship also ascertains the fact
that these two coefficients are expressions of the same soil property. No significant
relationships were observed between phosphate sorption and other soil properties
considered in this study.
The aluminum content and the clay content of the studied soils correlated strongly (r =
0.906***) meaning the percentage of aluminum was higher than usual. Organic matter,
active Fe (total free iron oxides) and active Fe (amorphous Fe) did not correlate well with
the phosphate sorption at 10 µg P/mL of application.

Phosphate Sorption Capacities of Some Soils in Madhupur Tract of Bangladesh 15
Table 7: Correlation coefficient (r) of phosphate sorption at 10 µg/mL with different properties of soil and phosphate
sorption parameters (n = 9)
Kf
0.543NS
Significant at * P<0.05, ** P<0.01 and *** P<0.001, respectively; NS = not significant
P app @ 10 = Phosphorus application at 10 µg P/mL application
Q0
0.580NS
0.927***
OM
0.180NS
0.359NS
0.125NS
Clay
content
0.209NS
0.473NS
0.814**
0.546NS
Active Fe
0.381NS
-0.563NS
-0.108NS
0.067NS
0.029NS
Free Fe
-0.047NS
0.392NS
0.011NS
0.086NS
0.219NS
0.149NS
Mn
-0.047NS
0.918***
0.379NS
-0.418NS
0.001NS
-0.058NS
0.182NS
Al
0.624NS
0.156NS
0.581NS
0.906***
0.197NS
0.363NS
0.640NS
0.447NS
P app@ 10
0.684*
0.075NS
0.156NS
0.191NS
0.799**
0.415NS
0.387NS
0.940***
0.402NS
Al
Mn
Free Fe
Active Fe
Clay
content
OM
Q0
Kf
1/n

16 Mohammad Enayet Hossain, Fatema Tuj Johra Sania and Md. Harunor Rashid Khan
Conclusion
A sorption experiment was done with some representative acidic soils of Madhupur Tract
of Bangladesh to observe the role of clay content, iron and aluminum content on the
phosphate sorption capacities of soils. The soils having higher aluminum and clay
content sorbed more phosphate compared to other soils. A strong relationship was
obtained between P sorption at 10 µg P/mL of application and the clay content and the
aluminum content. However, the amount of total and free iron oxides did not correlate
well with the phosphate sorption capacity of soils. These relationships ascertain the
much-studied fact that phosphate sorption is primarily governed by the amount of clay
and aluminum content in acidic soils.
Acknowledgements
We are grateful to the Ministry of Science and Technology, the Government of the
People’s Republic of Bangladesh for the grant under the 2017-2018 special allocation
and the University Grants Commission of Bangladesh for their financial support in the
2017-2018 financial year. We are also thankful to the people in the Advanced Lab of the
Department of Soil, Water and Environment, the University of Dhaka for their immense
help in the analysis of the samples.
References
[1] Morel, C. 2002. Caractérisation de la phytodisponiblité du P du sol par la modélisation du
transfert des ions phosphate entre le sol et la solution. Habilitation à Diriger des
Recherches. INPL-ENSAIA Nancy. 80 pp.
[2] Cordell, D., Rosemarin, A., Schroder, J.J., and Smit, A. L. 2011. Towards Global
Phosphorus Security: A Systems Framework for Phosphorus Recovery and Reuse Options.
Chemosphere, 84(6): 747-758.
[3] Kawasaki, N., Ogata, F., and Tominaga, H. 2010. Selective adsorption behavior of
phosphate onto aluminum hydroxide gel. Journal of Hazardous Materials, 181:574–579.
[4] Pratt, C. and Shilton, A. 2009. Suitability of adsorption isotherms for predicting the
retention capacity of active slag filters removing phosphorus from wastewater. Water
Science and Technology, 59:1673–1678.
[5] Hossain, M.E, Ritt, C. R., Almeelbi, T.B., and Bezbaruah, A. N. 2018. Biopolymer Beads
for Aqueous Phosphate Removal: Possible Applications in Eutrophic Lakes. Journal of
Environmental Engineering, 144(5): 04018030
[6] Islam, A., and Mandal, R. 1977. Vertical distribution of phosphate in some humid tropical
soils. Dhaka University Student (Part E), 26: 39-51.
[7] Ozanne, P.G., and Shaw, T. C. 1967. Phosphate sorption by soils as a measure of the
phosphate requirement for pasture growth. Australian Journal of Agricultural Research,
18: 601-612.

Phosphate Sorption Capacities of Some Soils in Madhupur Tract of Bangladesh 17
[8] Hossain M. E., Hoque, S. and Osman K.T. 2012. Phosphate Sorption in Some
Representative Soils of Bangladesh. Archives of Agronomy and Soil Science, 58(9):959-
966.
[9] Pierzynski G. M., Sims, J.T. and Vance, G. F. 1994. Soils and environmental quality. Lewis
Publishers, Boca Raton.
[10] FAO/UNDP 1988. Land Resources Appraisal of Bangladesh for Agricultural Development
2 FAO, Rome.
[11] Day, P. R. 1965. Particle Fractionation and Particle Size Analysis. In C. A. Black (Ed.),
Methods of Soil Analysis. Part 1. Agronomy Monograph 9, ASA and SSSA.
[12] Jackson, M. L. 1973. Soil Chemical Analysis. Prentice Hall of India Pvt. Ltd., New Delhi.
[13] Bray, R. and Kurtz, L. 1945. Determination of Total, Organic and Available Forms of
Phosphorus in Soils. Soil Science, 50:39-45.
[14] Murphy, J. and Riley, J. 1962. A Modified Single Solution Method for Determination of
Phosphate in Natural Water. Analytica Chimica Acta, 27:495-501.
[15] Holmgren, G. 1967. A Rapid Citrate-Dithionite Extractable Iron Procedure. Soil Science
Society of America Proceedings, 31:210-211.
[16] Schwertmann, U. 1964. Differenzierung Der Eisenoxide Des Bodens Durch Extracktion
Mit Ammoniumoxalat. L ö sung. Z. Planz. D ü ng. Bodenk. 105:194-202.
[17] Nair, P. S., Logan, T. J., Sharpley, A. N., Sommers, L. E., Tabatabai, M. A. and Yuan, T.
L., 1984. Interlaboratory Comparison of a Standardized Phosphorus Adsorption Procedure.
Journal of Environmental Quality, 13(4):591-595.
[18] Vaananen, R., Hristov, J., Tanskanen, N., Hartikainen, H., Nieminen M. and Ilvesniemi, H.
2008. Phosphorus Sorption Properties in Podzolic Forest Soils and Soil Solution
Phosphorus Concentration in Undisturbed and Disturbed Soil Profiles. Boreal Environment
Research, 13(6):553-567.
[19] Yousuf, T. S., Hossain, M. E., Afsar, M. Z., Osman, K. T. 2019. Phosphate Sorption
Indices as Affected by the Calcareousness of Soils. Dhaka University Journal of Biological
Sciences, 28(1):93-110.
[20] Afsar, M. Z. and Hossain, M. E. 2012. Characterization of Some Representative Calcareous
Soils of Bangladesh with Respect to Soil Phosphorus Requirements. International Journal
of Agricultural Research, 7:388-397.
[21] Naseri, A. A., Hoseini, Y., Moazed, H., Abbasi, F. and Samani, H. M. V. 2010.
Determining of Soil Phosphorus Requirement with Application of Freundlich Adsorption
Isotherm. Asian Journal of Agricultural Research, 4(4):226-231.
[22] Cabrera, F., Arambarri, Pd., Madrid, L. and Toga, C. G. 1981. Desorption of Phosphate
from Iron Oxides in Relation to Equilibrium Ph and Porosity. Geoderma, 26(3):203-216.
[23] Strauss, R., Brummer, G. W. and Barrow, N. J. 1997. Effects of Crystallinity of Geothite: Ii.
Rates of Sorption and Desorption of Phosphate. European Journal of Soil Science, 48:101-
114.

18 Mohammad Enayet Hossain, Fatema Tuj Johra Sania and Md. Harunor Rashid Khan
[24] Wilson, G., Rhoton, F. E. and Selim, H. M. 2004. Modeling the Impact of Ferrihydrite on
Adsorption-Desorption of Soil Phosphorus. Soil Science, 169(4):271-281.
[25] Wang, X., Liu, F., Tan, W., Li, W., Feng, X. and Sparks, D. L. 2013. Characteristics of
Phosphate Adsorption-Desorption onto Ferrihydrite: Comparison with Well-Crystalline
Fe(Hydr)Oxides. Soil Science, 178(1):1-11.
[26] Langmuir, I. 1918. The Adsorption of Gases on Plane Surfaces of Glass, Mica and
Platinum. Journal of the American Chemical Society, 40:1361-1403.
[27] Sodré, F. F., Lenzi, E. and Costa, A. C. 2001. Applicability of adsorption models to the
study of copper behaviour in clayey soils. Quimick Nova, 24:324-330.
[28] Limousin, G., Gaudet, J. P., Charlet, L., Szenknect, S., Barthes, V. and Krimissa, M. 2007.
Sorption isotherms: A review on physical bases, modeling and measurement. Applied
Geochemistry, 22:249–275.
[29] Freundlich, H. 1906. Concerning Adsorption in Solutions. Zeitschrift Fur Physikalische
Chemie--Stochiometrie Und Verwandtschaftslehre 57(4):385-470.
[30] Singh, B. and Gilkes, R. 1991. Phosphorus Sorption in Relation to Soil Properties for the
Major Soil Types of South-Western Australia. Australian Journal of Soil Research,
29(5):603-618.
[31] Holford, I. C. R. 1979. Evaluation of Soil Buffering Indices. Australian Journal of Soil
Research, 17(3):495-504.
[32] Agbenin, J. O. and Tiessen, H. 1994. The Effects of Soil Properties on the Differential
Phosphate Sorption by Semiarid Soils from Northeast Brazil. Soil Science, 157(1):36-45.