Tải bản đầy đủ - 0trang
III. Phosphorus in Soil Solution
liquid phases and also with the ratio between solid and solution. The
forms in which phosphorus exists in solution are governed by the reactions of protonation and complex formation. These forms are phosphoric acid, H,PO,, the corresponding ions, HLP04-,HP0,2-, and Po,”,
and the soluble complexes of these ions.
The distribution of phosphorus between phosphoric acid and the
phosphorus ions is determined primarily by p H in the following equilibrium:
From the law of mass action
where K is the dissociation constant of the acid and the brackets indicate
concentrations. This simple dissociation “constant” is influenced by the
temperature and concentration of the solution. The latter factor can be
accounted for by using activities, thus writing aH, etc., fbr active concentrations:
Using the notation pH = -log aH the equation can be written:
Phosphoric acid is a tribasic acid, and it has consequently three dissociation constants from which the proportions of the free acid and the
phosphorus ions which exist at various pH levels can be calculated. The
values presented in Fig. 1 are based on the values of the dissociation
FIG. 1. Distribution of ion species of phosphoric acid at different pH values at
infinite dilution (full line) and at ionic strength of 0.03 (broken line).
constants at 18°C. given by Bjerrum ( 1958):
pK’ = 2.120 - 0 499 4m -0.34~~
< f 1.04m
pK” = 7.227 - 1497 m
pK”’ = 12.465 - 2.495 4%
where m = ionic strength.
It is obvious from the graph that the ionic species H,PO,- and HP04’are the most abundant in the pH range encountered in soil and that the
ionic strength encountered in nonsaline soil does not alter this distribution significantly. It should be remembered that increased ionic strength
normally depresses soil pH.
2. Complex Formation
Phosphorus forms soluble complexes with many metallic ions, and
part at least of the phosphorus in the soil solution may be complexed.
The soluble complexes that are likely to occur and their equilibrium
constants, expressed as stability constants, are given in Table 11. The
greater this constant, the more stable is the complex.
A normal soil solution will contain several kinds of metallic ions
which will form complexes with ions other than phosphate (e.g., hydroxyl, carbonate, sulfate, and organic ions). In addition, phosphorus
ions may take part in other homogeneous reactions. It would be a very
formidable task to calculate the proportion of the phosphorus that is
likely to be present in a complexed form in a given soil solution. Even
in a simplified soil solution dominated by one electrolyte it is still difficult
to make more than a rough approximation.
However, there has been some recent work in which the presence of
complex ions have been investigated in soil extracts. Weir and Soper
(1963) found that phosphorus was held in a soluble complex of ferric
humate formed when humic acid was extracted from an acid soil by a
solution of ferric chloride. They discovered that the amount of phosphorus in the complex decreased slightly as the p H of the solution was
raised from 5 to 9. Taylor and Gurney (1962a) concluded that an acid
suspension of colloidal aluminum phosphate contained significant quantities of a soluble Al-P complex. Fordham (1963) found no evidence for
the presence of complex phosphate ions in a 0.01 M CaC1, extract of a
soil of pH 6.3. However, in a calcareous soil Larsen (1965, 1966a) found
that the increase in the solubility of phosphorus, relative to the calcium
chloride concentration of the extractant, could be explained by assuming
the formation of the soluble complex CaHPO,. He found no such effect in
an acid soil of pH 5.5 where the HPO, ion concentration is negligible.
No information concerning soluble phosphorus complexes in the true
Stability Constantsa of Metal-Phosphate Complexes
+ HP042-+ NaHP04K+ + HP042- + KHPO4Mg2+ + HPOn2- C MgHPO4
ca. 2 . 3
ca. 2 . 3
Compiled from Sillen and Martell (1964).
soil solution is available, but based on the meager experimental evidence
above and the known stability constants of phosphorus complexes (Table
11),it seems likely that a significant proportion of the phosphorus in the
soil solution may sometimes be present in this form. Further, it would be
expected that this proportion would be high both in acid soils (due to
the high stability of Fe-P and A1-P complexes) and in calcareous soils
(due to the high proportion of HPO, ions). Complex formation will be
lowest in slightly acid soils where H,PO, ions predominate and where
the concentrations of iron and aluminum are low.
Thus although our knowledge about soluble phosphorus complexes
in the soil solution is very limited, it seems an oversimplification to assume that even the bulk of the soluble phosphorus in soil is always present as the two ion species H2P0,- and HPO,*-.
B. H E ~ O G E N E OEQUILIBRIA
The upper limit for the phosphorus concentration in solution is set by
the heterogeneous equilibria in which it takes part, The reactions involved are the dissolution and precipitation of sparingly soluble phosphorus salts, controlled by the solubility product principle and by the
adsorption of phosphorus on the surface of soil particles. Because the
phosphorus concentration in solution is governed by reactions with the
solid phase, the equilibrium level can be used to characterize the energy
state of phosphorus in the whole system. This has led to a considerable
amount of work on the chemical potential of phosphorus in soil.
1. Solubility Product Principle
The solubility product of a salt AaBb taking part in the following
A,& G aA
is defined as
When the salt is present in the solid phase, AaBb is constant and the
solubility product simplifies to
K q = [Ala[Blb
the value of which depends on the ionic strength of the solution. A true
constant can be obtained by using activities instead of concentrations,
and using the convention that p = --log,,
the solubility constant can
then be expressed:
PK., = a pA
+ b pB
The relevant sparingly soluble phosphorus salts are those of magnesium, calcium, aluminum, and iron. Solubility products for these have
been tabulated by Sillen and Martell ( 1964).
Wild (1954) examined a large number of early analytical data but
found no agreement between the concentration of phosphorus in the soil
solution and the concentration predicted from solubility products. More
recently, several other workers in this field (Chakravarti and Talibudeen,
1962; Hagin and Hadas, 1962; Bache, 1963) have also found that the
phosphorus concentration of the soil solution did not conform to solubility product principles.
In neutral and calcareous soils the absence of agreement between the
observed phosphorus concentrations and the solubility of calcium phos-
phates may be explained by the incomplete understanding of the solubility product of these compounds. Thus Bjerrum (1949) found two
solubility products for octocalcium phosphate and two for hydroxylapatite, one when the equilibrium was approached by precipitation and
the other when it was approached by dissolution.
A further complication is the strong influence that impurities seem
to have on the solubility of basic calcium phosphate. Thus Greenwald
( 1942) and Ericsson (1949) found that pure hydroxylapatite obeyed the
solubility principle but in the presence of small amounts of calcium
carbonate it did not. The equilibrium solutions were apparently supersaturated with respect to hydroxylapatite.
The deviation from pure hydroxylapatite behavior can be given by
the “saturation index” n suggested by Bjerrum ( Schmidt-Nielsen, 1946) :
log n = (pL - pZ)/9
where I is the apparent ionic product and L is the true solubility product
of hydroxylapatite. The factor 1/9 appears because there are 9 ions
involved in the hydroxylapatite formula Ca5( PO, ),OH.
Calculating this index for the solubility of hydroxylapatite in the
presence of calcium carbonate, Ericsson found that it varied with pH
according to the equation:
0.44pH - 2.33
Studying the solubility of soil phosphorus in 0.01 M CaCI,, Larsen and
Court (1961) found the relationship between log n. and p H shown in
Fig. 2, from which it may be seen that over the pH range 5.0 to 7.5 the
solubility was consistent with that of impure hydroxylapatite. Above
pH 6.0 the solutions were supersaturated, and below p H 6.0 they were
undersaturated with respect to pure hydroxylapatite. Thus it is not
surprising that Clark and Peech (1955) observed a lower solubility of
phosphorus in acid soil solution than corresponds to the solubility of
hydroxylapatite. Their statement that “At intermediate and low p H values, it is obviously necessary to postulate the existence, in soils, of solid
phosphate phases that are less soluble than the calcium phosphates” can
thus be disputed.
The lack of agreement between the phosphorus concentration of a
soil solution and the solubility of pure hydroxylapatite does not necessarily imply that hydroxylapatite is not determining the phosphorus
A degree of acidity at which all calcium phosphates are so. soluble
that they cannot possibly control the phosphorus concentration will, of
course, eventually be encountered. This may well be from pH 5.0 downward, but it will certainly be true for pH levels below 4.0. At this p H
value the clay fraction will yield significant amounts of aluminum ions
which will then be present in the cation exchange complex and in the
log n =0.701pH
r = 0.981
FIG.2. The logarithm of hydroxylapatite saturation index ( n ) as a function of
soil pH. (From Larsen and Court, 1961.)
solution. Solubility of aluminum phosphates are likely, in this situation,
to determine the upper limit of the phosphorus concentration of the soil
solution. But more recent evidence has challenged the earlier view that
this limit is governed by the simple solubility product of variscite. Complications of incongruent dissolution and complex formation, analogous
to the chemistry of hydroxylapatite have been pointed out (Taylor and
Gurney, 1962a; Bache, 1963; Raupach, 1963). These are discussed later
in Section IV, B, 2.
Even in the absence of phosphorus precipitating ions, phosphorus will
still be removed from solution by adsorption onto the surface of soil
particles such as clay or calcium carbonate. Adsorption is therefore
another possible mechanism which can determine the phosphorus concentration in solution.
In practice as the adsorption system becomes more saturated by the
addition of phosphorus, the concentration in solution rises and a point
will ultimately be reached when precipitation of a sparingly soluble
phosphorus compound will occur. The solubility of this compound will
then determine the upper limit of the phosphorus concentration; conversely, if the phosphorus concentration is lowered, sparingly soluble
phosphate will dissolve until the adsorption complex has been saturated
to a degree which corresponds to the solubility of the least stable
phosphorus compound present.
3. Chemical Potentials
It is a well known principle of physical chemistry that in any multiphase system at equilibrium, the chemical potentials or partial molar free
energies of all diffusible chemical components are equal. Thus for a
system consisting of a solution phase in equilibrium with a solid phase,
the chemical potential of all components is the same, so that the potential of the solid phase is then easily calculated from activity measurements made in solution.
Schofield (1955) suggested that this approach could be used to
obtain an index of soil phosphate availability. He proposed the “phosphate potential,” the negative chemical potential of monocalcium phosphate (?h pCa pH,PO,) determined in a 0.01 M CaCl, soil extract.
The determination and use of the phosphate potential of soil solutions is,
however, beset with several practical and fundamental difficulties. The
most obvious practical difficulty is the low phosphorus concentration of
the solution which can only be partly overcome by improved analytical
methods. In addition, the pH measurement may give unduly variable
results due to the low buffer capacity of soil extracts.
Less obvious, but very real difficulties are ( a ) lack of equilibrium,
( b ) microbial activity, ( c ) influence of the soil:solution ratio, ( d )
formation of soluble complexes.
a. Lack of equilibrium. Following Schofield and Taylor ( 1955),
Aslyng (1954, 1964) used the chemical potential of calcium hydroxide,
the “lime potential” defined as (pH - %pCa),and the phosphate potential in an attempt to assess the presence and nature of calcium phosphate
compounds in soils. He apparently adopted a short but unspecified period
for equilibration. Similar procedures have been adopted by several other
workers who have recorded their equilibration time. These periods and
other details of the procedures reported are compiled in Table 111.
Methods of Measuring Phosphate Potential
Method of shaking
Andersen and Mogensen (1962)
Barrow et al. (1965)
Chakravarti and Talibudeen
Clark and Peech (1955)
Clark and Peech (1960)
0-90 hr. (17 hr.)
30 sec.-17 hr.
B y hand
Soil: solution ratio
(9. soi1/100 ml.)
Solution conc. (CaC12)
2 X 10- M KCl
2 . 5 x 10-3 M ,
5 x 10-3 M ,
x lo-* M
Larsen and Court (1961)
Larsen and Widdowson (1964)d
10, 20, 40, 80
2, 4, 8, 16, 32
Moreno et al. (1960)e
Moser et al. (1959)d
Olsen et al. (1960)d
RamaMoorthy and Subramanian
Taylor and Gurney (1965)
White and Beckett (1964)
10, 20, 40, 80
1, 2, 5, 10, 20
b Acid soils only; calculated (+ pM
pH~P04)where M = Ma+ or Fe3+.
Effect of CaCh concentration also studied.
d Partial pressure of COZ controlled.
MCP or DCPD added to suspensions.
The position seems to be that after a comparatively short shaking
period of a few minutes an apparent state of equilibrium is obtained,
followed by a slow increase of the phosphorus concentration in solution.
This slow increase goes on for months, perhaps years, and equilibrium
In this situation a choice must be made. By choosing a short period
of equilibration, a phosphate potential relevant to the most reactive
phosphorus in the solid phase may be achieved. The assumption that
this potential is a measure of the partial molar energy (free energy) of
all the solid-phase phosphorus is not valid, as this is based on the condition that full equilibrium is achieved.
b. Microbial activity. When soil suspensions are shaken for only a
few minutes, microbial activity can probably be ignored, but this is not
the case when the soil is shaken for hours or days. There will then be a
cumulative effect of the microbial activity which, after a short initial
period of adaptation, shows a peak that is particularly marked when
air-dry soil is wetted. This latter effect can be reduced by storing the soil
moist for a prolonged pre-period (White and Beckett, 1964).
There are two main ways in which the microbial activity can influence the amount of phosphorus in the solution and the phosphate potential: ( 1) biological immobilization of phosphorus; ( 2 ) solubilization of
phosphorus by the acidic compounds produced by the microorganisms.
The former effect is generally insignificant, unless the microbial
activity is boosted by addition of organic material. This is because under
normal circumstances the small amount of phosphorus removed from the
soil solution is replaced from the solid phase, so that the effect is to delay,
rather than disturb, the equilibrium.
Solubilization of phosphorus by organic acids from microbial activity
is also a rare occurrence, since such acids are normally quickly decomposed. Accumulation of strong inorganic acids could affect the equilibrium, for example, when oxidizable materials such as sulfur are present.
The most likely cause of solubilization, however, is the accumulation of
carbon dioxide produced by microbial respiration. This difficulty may be
overcome by using a germicide (Barrow et al., 1965), but this has the
disadvantage of possible side effects. Aeration with moist air is the safest
method of preventing accumulation of CO, (Larsen and Widdowson,
c. Soi1:solution ratio. There have been several attempts to overcome the marked influence that soil: solution ratio has on phosphate
potential. Aslyng ( 1954), for example, extrapolated his experimental results to “zero dilution,” and others have adopted this procedure. This is
not justifiable, however, since both his own data and those of Larsen and
Court (1960) showed that there is no approach to a limiting value at
zero dilution. White (1966) overcame the effect by allowing the soil
phosphorus to achieve equilibrium by prolonged storage under constant
environmental conditions. Larsen and Widdowson ( 1964) suggested that
an important factor for the soil: solution ratio effect was the accumulation
of CO, during the shaking of the soil suspension in stoppered bottles.
Their results are presented in Fig. 3, from which the marked effect of
g. soill50ml. 0.01M CaCI,
FIG.3. Effect of soi1:solution ratio on phosphate potential where COXaccumulation is ( a ) prevented, ( b ) allowed. (From Larsen and Widdowson, 1964.)
CO, accumulation can be seen, an effect which is proportional to the
amount of soil in the suspension. The phosphate potential of this slightly
calcareous soil became completely independent of soil :solution ratio
when CO, accumulation was prevented by aeration. In the light of this
information, the previous work on the effect of soil :solution ratio where
no precaution against CO, accumulation has been taken, should be
reexamined. It is to be expected that different results will be obtained
when the suspension is aerated, except perhaps in the most acid soils
where the biological activity is restricted by the acidity and also where
the carbonic acid formed is largely undissociated.
For example, Salmon (1965) found no effect attributable to CO, in
the two acid soils that he studied. H e ascribed the decrease in potential
which he found with increased soi1:solution ratio to the fact that the soil
was changed more from its original state in attaining equilibrium with
smaller ratios. This would only seem a plausible explanation where the
soil was poorly buffered with respect to phosphorus.
d. Formation of soluble complexes. The calculation of phosphate
potentials is based on the activities of free ions, and no account is taken