SSSAJ Grow Your Career with SSSA
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 27 October 2006
Published in Soil Sci Soc Am J 70:2038-2048 (2006)
DOI: 10.2136/sssaj2006.0051
© 2006 Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Lombi, E.
Right arrow Articles by Paterson, D.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Lombi, E.
Right arrow Articles by Paterson, D.
Agricola
Right arrow Articles by Lombi, E.
Right arrow Articles by Paterson, D.
Related Collections
Right arrow Phosphorus
Right arrow Nutrients
Right arrow Nutrient Management
Right arrow Soil Chemistry

Soil Chemistry

Speciation and Distribution of Phosphorus in a Fertilized Soil

A Synchrotron-Based Investigation

E. Lombia,*, K. G. Scheckelb, R. D. Armstrongc, S. Forrestera, J. N. Cutlerd and D. Patersone

a CSIRO Land and Water, PMB 2 Glen Osmond, SA 5064, Australia
b USEPA, National Risk Management Research Lab., 5995 Center Hill Ave., Cincinnati, OH 45224
c Department of Primary Industries, Natimuk Rd., PB 260, Horsham, VIC 3400, Australia
d Canadian Light Source Inc., Saskatoon, SK S7N 0X4, Canada
e Advanced Photon Source, Argonne National Lab., Bldg. 431 B008, Argonne, IL 60439-4856

* Corresponding author (enzo.lombi{at}csiro.au)


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIAL AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Phosphorus availability is often a limiting factor for crop production around the world. The efficiency of P fertilizers in calcareous soils is limited by reactions that decrease P availability; however, fluid fertilizers have recently been shown, in highly calcareous soils of southern Australia, to be more efficient for crop (wheat [Triticum aestivum L.]) P nutrition than granular products. To elucidate the mechanisms responsible for this differential response, an isotopic dilution technique (E value) coupled with a synchrotron-based spectroscopic investigation were used to assess the reaction products of a granular (monoammonium phosphate, MAP) and a fluid P (technical-grade monoammonium phosphate, TG-MAP) fertilizer in a highly calcareous soil. The isotopic exchangeability of P from the fluid fertilizer, measured with the E-value technique, was higher than that of the granular product. The spatially resolved spectroscopic investigation, performed using nano x-ray fluorescence and nano x-ray absorption near-edge structure (n-XANES), showed that P is heterogeneously distributed in soil and that, at least in this highly calcareous soil, it is invariably associated with Ca rather than Fe at the nanoscale. "Bulk" XANES spectroscopy revealed that, in the soil surrounding fertilizer granules, P precipitation in the form of octacalcium phosphate and apatite-like compounds is the dominant mechanism responsible for decreases in P exchangeability. This process was less prominent when the fluid P fertilizer was applied to the soil.

Abbreviations: LCF, linear combination fitting • MAP, monoammonium phosphate • PC, principal component • PCA, principal component analysis • TG-MAP, technical-grade monoammonium phosphate • XANES, x-ray absorption near-edge structure • XRF, x-ray fluorescence


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIAL AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
PHOSPHORUS DEFICIENCY commonly limits crop production, with an estimated 5.7 billion ha of land having suboptimal P levels for crop growth worldwide (Hinsinger, 2001). This problem is particularly relevant in Australian soils due to the presence of highly P-fixing calcareous and weathered acidic soils, and naturally low P status. To overcome P deficiency, farmers in Australia, as in other parts of the world, have applied P fertilizers for >100 yr. The benefits of P fertilization, however, have been limited by its low agronomic efficiency (Olson et al., 1971; Holloway et al., 2001).

In highly calcareous soils, the low plant availability of applied P is principally due to the formation of sparingly soluble Ca minerals such as dicalcium phosphate dihydrate, octocalcium phosphate, and ultimately hydroxyapatite (Lindsay, 1979; Sample et al., 1980; Freeman and Rowell, 1981). Tunesi et al. (1999) reported that in addition to surface reactions between P and carbonates, precipitation reactions of P with Ca2+ is a key mechanism controlling P availability in soil. In a study of 22 alkaline and calcareous soils from southern Australia, Bertrand et al. (2003) found that P sorption behavior was a direct function of CaCO3 content; however, several other studies have shown that P behavior, even in calcareous soil, may be controlled by small amounts of Fe and Al oxides. For instance, Ryan et al. (1985) suggested that in soil systems, Fe oxides may also influence P availability either by direct reaction or by coating the carbonate phase. They found that P sorption was related to oxalate-extractable Fe in a range of Lebanese calcareous soils. Similar results were obtained for Spanish calcareous soils by Solis and Torrent (1989) and for Western Australia calcareous soils by Samadi and Gilkes (1998). The P reaction processes and the reaction products of P fertilizers in the heterogeneous soil environment are still controversial areas of research.

Research conducted in calcareous soils of southern Australia has shown that the agronomic efficiency of P fertilizers is strongly dependent on the form of fertilizer used. Holloway et al. (2001) showed that TG-MAP was 4 to 15 times more effective in increasing the grain yield of wheat than granular MAP. Similar results were obtained when a number of fluid and granular products were compared in both greenhouse and field studies (Holloway et al., 2004; McBeath et al., 2005). Subsequent investigations suggested that both chemical and physical factors caused fluid fertilizers to be more efficient than similar granular products. In particular, diffusion of P from granular products in soil was found to be more limited than when P was supplied with fluid fertilizers (Lombi et al., 2004a, 2004b). Using isotopic dilution techniques, we showed that P lability is lower when granular products are applied to calcareous soil than when a fluid P source is used. This finding may reflect the increased precipitation of P in or around the fertilizer granules; however, the specific mechanisms responsible for the differential response observed are still unclear.

In this study, we uses synchrotron-based techniques to directly assess the distribution and chemical speciation of P in a calcareous soil in the vicinity of the point of application of a fluid and a granular fertilizer source. X-ray absorption near-edge structure spectroscopy has been recently used to identify P species in soil (Hesterberg et al., 1999; Beauchemin et al., 2003). This technique has the advantage of being nondestructive and provides information on the local molecular environment of an element even when poorly ordered mineral phases are present (Beauchemin et al., 2003; Fendorf and Sparks, 1996; Schulze and Bertsch, 1995). We made use of both "bulk" P XANES and, to our knowledge for the first time in the soil environment, P nano- (n-) XANES that allows solid-phase speciation at high spatial resolution. Moreover, P distribution was investigated using µ- and n-x-ray fluorescence (XRF).


   MATERIAL AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIAL AND METHODS
RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Soil and Fertilizers
A gray calcareous sandy loam classified in accordance to the Soil Survey Staff (1992) as a Calcixerollic Xerochrept was collected from Warramboo, upper Eyre Peninsula, South Australia. The soil was air dried and sieved to <1 mm before analysis. The pH of a 1:5 soil/solution extract was 9.1 in water and 8.0 in 0.01 M CaCl2. The CaCO3 content was 770 g kg–1 and the total P content was 330 mg kg–1, as determined by inductively coupled plasma–atomic emission spectroscopy (Spectroflame Modula, Spectro Analytical Instruments, Fitchburg, MA) following digestion in aqua regia (1:3 HNO3/HCl; Zarcinas et al., 1996).

Two fertilizers were used in this study: commercial granular MAP (10:22:0 N–P–K) and commercial TG-MAP (12:26:0 N–P–K) dissolved in deionized water.

Experimental Design
The experimental design used was similar to that reported by Lombi et al. (2004a, 2004b). Plastic petri dishes (8.7 cm in diameter, 1.1 cm high) were filled with 78 g of dry soil per dish to obtain a soil density of 1.2 g cm3. The soil was then wetted to 60% of its water holding capacity (Jenkinson and Powlson, 1976) by dripping distilled water onto the soil. The petri dishes were closed, sealed with Parafilm (Alcan Packaging, Neenah, WI) and left to equilibrate overnight. The following day the petri dishes were opened and fertilizers containing the same amount of P were added. In the case of MAP (10:22:0), an intact granule (42 ± 0.5 mg) was placed in the center of each petri dish. In the case of TG-MAP (12:26:0), 36 mg was diluted with 200 µL of distilled water (equivalent to 1% of the total water added to each petri dish) and injected in the center of the petri dish using a needle. Control petri dishes that did not receive any fertilizer were also prepared. The petri dishes were then closed, sealed with Parafilm, and incubated in the dark for 5 wk in a controlled environment at 20°C. At the end of the equilibration period, the petri dishes were opened and concentric rings of soil, centered around the granule (for MAP) or the injection point, were removed using a series of plastic cylinders. These cylinders were driven into the soil one at a time, starting with the smaller one, and all soil inside the cylinder was removed. This procedure is similar to the method used by Izaurralde et al. (1986) and Kouboura et al. (1995) to sample soil at different distances from fertilizer bands. In this experiment, sections of soil were collected from 0- to 7.5- and 7.5- to 13.5-mm radius from the granule or injection point. The soil samples were air dried before further analysis.

Soil Analyses
Total soil P was determined on subsamples of the soil sections as described above. Another portion of dry soil was used to measure the isotopically exchangeable P (E value) using an isotopic dilution technique (Salcedo et al., 1991). This technique allows quantification of the amount of P in the soil that is in rapid equilibrium with the soil solution P. Briefly, 2 g of soil was equilibrated for 24 h with 20mL of deionized water and the pH measured. Carrier-free 32P (30 kBq) was then added to each sample and the labeled suspension was equilibrated in an end-over-end shaker for a further 24 h. The soil suspensions were then filtered through 0.2-µm membrane filters (Sartorius AG, Goettingen, Germany) and analyzed for P in solution using the procedure of Murphy and Riley (1962). The 32P activity in the filtrates was measured by Cerenkov counting (RackBeta II, LKB Wallac). The exact total activity introduced in each sample was determined by analyzing spiked solutions, without soil, in parallel to the soil suspensions. The E values (mg P kg–1) were calculated according to Hamon et al. (2002). The E values and total soil P were measured in three replicates and analyzed using ANOVA and least significant difference with the Genstat 8 statistical package (VSN International, 2005).

"Bulk" P XANES were collected at the Synchrotron Radiation Center, Madison, WI. Air-dried samples were mounted as a thin layer on double-sided carbon tape and transferred into an ultra high vacuum solid-state x-ray absorption spectroscopy chamber. The K-edge data were collected using a double crystal monochromator (DCM) using InSb crystals with a photon resolution of ~0.9 eV and a nine-element solid-state Ge detector. The DCM beamline was calibrated using sodium pyrophosphate at 2152.5 eV. The P K-XANES were collected in the energy range 2133 to 2178 eV with 0.25-eV increments. All the spectra presented here represent the average of at least three scans.

Spatially resolved data were collected at beamline 2-ID-B at the Advanced Photon Source, Argonne, IL. Details of this beamline are reported by McNulty et al. (1995, 2003). The electron storage ring operated at 7 GeV and a 5.5-cm period undulator provided the coherent x-ray illumination. A Fresnel zone plate with a central stop focused the x-ray beam to a near diffraction-limited focal spot on the sample. A pinhole operating in the shadow of the central stop served as an order-sorting aperture to select the first-order focus of the zone plate (McNulty et al., 2003). The sample stage had precision translation axes for raster scanning. A silicon drift diode detector (KETEK Gmbh, Munich, Germany, 5-mm2 active area) was used for x-ray fluorescence measurements. The soil samples were ground and embedded in a hard mix of a low-viscosity resin. Thin sections (2.5 µm thick) were prepared and mounted on transmission electron micrograph (TEM) grids covered with a parllodion support film. The TEM grids were fixed to an aluminum holder and placed in the sample stage under a continuous flow of He. This experimental setup allowed collection of n-XANES spectra at the K-absorption edge of P and n-XRF elemental maps with a spatial resolution of <100 nm. Elemental (P, Al, Si, and Mg) distributions were collected using fluorescence data for 40- by 40-µm areas with a step size of 0.5 µm. A few highly resolved elemental distributions were collected for 10- by 10-µm areas with a step size of 0.06 µm (60 nm). Energy-resolved fluorescence spectra were collected at each measured location across the soil specimen. Global values for parameters—describing the detector response function (energy calibration, resolution, and peak shaping), spectrum background, and geometrical factors—and total elemental concentrations were determined by fitting the sum of all point spectra using a least squares fitting procedure. Elemental concentrations at each measurement location were then determined by fitting the individual point spectra with the global parameter values fixed. Four to six n-XANES spectra for each sample were collected in spots that showed high P fluorescence. These elemental distributions were refined by fitting the fluorescence spectrum obtained at each location using the MAPS package (Vogt, 2003). A spherical grating monochromator (McNulty et al., 1997) was used to decrease the energy of the incoming x-rays from 2190 to 2130 eV in 0.25-eV steps. All spectra presented are the average of three scans.

The XANES data were background- and baseline-corrected using ATHENA (Version 0.8.049, Ravel and Newville, 2005). The energy scale was normalized to a reference energy of 2149 eV as reported by Beauchemin et al. (2003). The normalized XANES spectra were anayzed using principal component analysis (PCA) and linear combination fitting (LCF). In the case of PCA, spectral analysis was conducted using The Unscrambler (Camo Software, 2005). This analysis allows the discriminate grouping of samples, where principal components (PCs) are calculated in the same data space occupied by the samples by forming linear combinations of the original variables. The first PC contains the maximum variance from the data, with consequent PCs describing the next highest variance. The new orthogonal latent variables, called scores, can be used to determine which spectra are similar or dissimilar. For each PC, a loading is generated that describes the contribution of each energy for that PC. A target transformation was performed to identify the standard spectra that might contribute to the PCs. The SPOIL value (Malinowski, 1991; Manceau et al., 2002) was used as a criterion for determining whether the target candidate resembled the original standard spectrum. Linear combination fitting of the normalized sample XANES spectra was performed using all possible binary combinations of standards having SPOIL <3 (Malinowski, 1991). Only binary combinations were used since the PC analyses indicated that our system could be adequately described by two components. ATHENA (Ravel and Newville, 2005) was used to perform the LCF in a relative energy range of –18 to 16 eV using {chi}2 and residual values to assess the best fit.

Phosphorus XANES Standards
The "bulk" XANES spectra of eight P compounds were provided by Derek Peak (Dep. of Soil Science, Univ. of Saskatchewan) and Dean Hesterberg (Dep. of Soil Science, North Carolina State Univ.): monoammonium orthophosphate (NH4H2PO4), apatite [Ca5(PO4)3(OH,F,Cl)], octacalcium phosphate [Ca8H2(PO4)6·5H2O)], monetite (CaHPO4), phytic acid, phosphosiderite (FePO4·2H2O), vivianite [Fe3(PO4)2·8H2O], and wavellite [Al3[(OH,F)3(PO4)2·5H2O]).

The XANES spectra of seven P compounds were also collected at the Advanced Photon Source using the nano beamline: potassium orthophosphate (KH2PO4), monoammonium orthophosphate, fluoroapatite [Ca5(PO4)3F], hydroxyapatite, tetrasodium pyrophosphate, and freshly precipitated Fe and Al orthophosphate.


   RESULTS AND DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIAL AND METHODS
 RESULTS AND DISCUSSION
CONCLUSIONS
 REFERENCES
 
Total Phosphorus Distribution and Isotopic Exchangeability
The total P concentration was greater in the soil sampled immediately adjacent to the point of fertilizer application than in the soil sampled 7.5 to 13.5 mm from the fertilizers (Table 1). In the MAP treatment, the E value in the central section was approximately double that of the section farther away from the point of fertilizer application. In contrast, the isotopic exchangeability of P was similar in the two sections of the TG-MAP treatment. When the respective sections in the two fertilizer treatments were compared, the relative lability of P, as a percentage of total P, was significantly higher (P ≤ 0.001) in the TG-MAP than in the granular treatment. The percentage of P isotopically exchangeable in the central section of the granular treatment was not dissimilar to the lability of the control. These results are in agreement with previous findings obtained in calcareous soils by Lombi et al. (2004a, 2004b) and confirms that P derived from fluid fertilizers diffuses further in the soil and remains more labile (i.e., potentially available to plants) than P applied as a granular source. These results cannot be attributed to acidification near the fertilized zone since the pH in the granular treatment was slightly lower or similar to the pH in the fluid treatment (Lombi et al., 2004a).


View this table:
[in this window]
[in a new window]
  		Table 1. Soil pH, total P concentration, and P isotopic exchangeability (E value) at different distances from the point of fertilizer application. Data are the mean (±standard error) of three replicates.

 
"Bulk" Phosphorus XANES
The P K-XANES of the standards showed interesting features in the spectra, and these are characteristic of different compounds (Fig. 1 ). The position of the main peak, a 1s -> 3p electronic transition, revealed slight variations in energy in the different compounds. The MAP did not show any shift in the main peak, whereas a slight shift to higher energy was observed for wavelite, vivianite, and phosphosiderite (+0.4 eV). The main peak of apatite and octacalcium phosphate were shifted toward a lower energy (–0.2 eV). The most characteristic XANES features, however, were a pre-edge feature between –5.5 and –1 eV (relative energy) for the Fe compounds (phosphosiderite and vivianite) and a postedge shoulder between 1.5 and 5.5 eV in the case of the Ca phosphates. Both of these features have been observed previously (Hesterberg et al., 1999; Peak et al., 2002; Beauchemin et al., 2003; Khare et al., 2005). The pre-edge feature has been attributed to a P 1s electron transition into an Fe(3d)–O(2p)–P(3p) antibonding molecular orbital (Franke and Hormes, 1995). Hesterberg et al. (1999) reported that the postedge shoulder was more pronounced for compounds containing many Ca atoms (apatite and octacalcium phosphate) than for monocalcium phosphates (monetite and brushite).


Figure 1
View larger version (26K):
[in this window]
[in a new window]
  		Fig. 1. Phosphorus XANES "bulk" spectra for selected P standards. Data are background- and baseline-corrected and energy normalized to a reference energy of 2149 eV.

 
The relative changes in intensity of the phosphorus K-edge spectra decreased from the MAP sample collected in the vicinity of the fertilizer to the control sample (data not shown) and were well correlated with the total P concentrations reported in Table 1. The noise of the P XANES spectrum for the control soil is due to the low total P concentration in the unfertilized soil. The spectra for the soil samples reported in Fig. 2 are clearly different from the P XANES of MAP. The presence of a distinct shoulder on the high-energy side of the white line and a shift of the main peak to slightly lower photon energy indicate that P is associated with Ca. The pre-edge feature observed in the Fe-containing standards is absent in the soil spectra, confirming that P is coordinated with Ca and not Fe in the second shell. This is in agreement with the results reported by Beauchemin et al. (2003) for slightly calcareous soils, where approximately 83% of the P was in the form of Ca compounds.


Figure 2
View larger version (20K):
[in this window]
[in a new window]
  		Fig. 2. Phosphorus XANES "bulk" spectra of soil samples collected at two distances from the point of fertilizer application (0–7.5 and 7.5–13.5 mm). Two fertilizer treatments were investigated: fluid technical-grade monoammonium phosphate (TG-MAP) and granular monoammonium phosphate (MAP). The control soil was not fertilized. Data are background- and baseline-corrected and energy normalized to a reference energy of 2149 eV.

 
A first PCA was performed on the preprocessed data for the entire energy range and results are shown in Fig. 3A , where the score plot for PCs 1 and 2 is presented. The PC1 and PC2 explained 71 and 12% of the sample spectral variation, respectively. All the fertilized soil samples were grouped on the negative side of the PC1 axis. In this plot, the more similar samples or spectra are closer to each other. Consequently it can be concluded that the samples on the negative side of PC1 were different from those grouped together on the negative side with the Al and Fe phosphate standards and the other P standards that did not contain Ca. Analysis of the loading spectra, which indicate the contribution of each energy point in the spectra for each PC, was then performed (data not shown). For PC2, the contributing factors were due to the noise of the spectrum from the control (unfertilized sample) and the main-peak energy, which accounted for almost all of the positive effect. For PC1, the samples on the positive side showed loadings due to spectral contributions for energies below the energy of the main peak. Further investigation indicated that the large postedge shoulder in the region 1 to 5.5 eV (relative energy) had the largest effect on the variability of samples to the left-hand side of PC1. To assess subtle differences between the spectra of the soil samples, this spectral region (1–5.5 eV) was used for a further PCA of the samples. In this second PCA, PC1 and PC2 explained almost all the spectral variation in the samples (87 and 7%, respectively). The results reported in Fig. 3B indicated that the P XANES spectra of the Ca phosphates follows the order monetite, octacalcium phosphate, and apatite along the PC1 axis, which follows the visual observation of an increase in intensity of the postedge shoulder of the Ca-phosphate standards with increasing complexation with Ca. The sample collected in the vicinity of the MAP fertilizer granule was very similar to the spectra of the apatite and octacalcium phosphate standard. In contrast, the TG-MAP samples and the sample collected between 7.5 and 13.5 mm from the granule were grouped together.


Figure 3
View larger version (16K):
[in this window]
[in a new window]
  		Fig. 3. Principal component analyses (PCA) of "bulk" P K-XANES of standard and soil samples: (A) PCA of all standards and samples across the full spectral range; (B) PCA of non-Fe and -Al standards used with the spectral range limited to the postedge shoulder (1–5.5-eV relative energy). MAP is granular monoammonium phosphate and TG-MAP is fluid technical-grade monoammonium phosphate.

 
Based on the results of the PCA discussed above, LCF was performed using every possible combination of the three standards that had a SPOIL <3: monetite, octacalcium phosphate, and apatite. Since the PCA analyses showed that most of the variability in the soil samples could be explained by two components, the maximum number of components was set to two. The goodness of fit indicated by {chi}2 was typically <0.4 except for the control soil, which was too noisy to be fitted with a low degree of uncertainty. The LCF results (Table 2) showed that the spectra of the soil sample closer to the fertilizer could be adequately fitted by a combination of apatite and octacalcium phosphate. In contrast, all the other samples could be fitted by combinations of apatite and monetite. These results corroborate the interpretation obtained using PCA. These findings also support and provide a mechanistic explanation for the results from work on P lability obtained in this study and in previous work (Lombi et al., 2004a, 2004b). Phosphate compounds containing large amounts of Ca, such as octacalcium phosphate and apatite, have very limited solubility in alkaline soils and therefore their contribution to the isotopically exchangeable P pool is limited. This explains the small relative lability (E value as a percentage of total P) of the sample collected in the vicinity of the fertilizer granule. The relationship between the contribution of the most soluble Ca phosphate (monetite, monocalcium phosphate) to the soil spectra reported in Table 2 and the relative P exchangeability (Table 1) is presented in Fig. 4 . The relationship suggests that, even though monocalcium phosphate may be largely isotopically exchangeable and therefore contributes substantially to the lability of P, other isotopically exchangeable P species are probably present in the soil. It should be remembered that the LCF procedure is determined by the standard used. Other forms of P such as dicalcium phosphate and sorbed P, for which XANES spectra were not available, could contribute to the isotopical exchangeability of P.


View this table:
[in this window]
[in a new window]
  		Table 2. Linear combination fitting of "bulk" P K-XANES spectra showing the relative proportion of P compounds in soil samples.

 

Figure 4
View larger version (14K):
[in this window]
[in a new window]
  		Fig. 4. Relationship between isotopically exchangeable P in soil (as a percentage of total P, Table 2) and the proportion of monetite (monocalcium P) as obtained by linear combination fitting of the "bulk" P XANES spectra.

 
Spatially Resolved Phosphorus Distribution and Speciation
"Bulk" P XANES gave an indication of the "average" molecular environment surrounding P in soil but the use of this technique alone is not sufficient to directly assess whether a small proportion of P forms complexes with Fe. Therefore, we used spatially resolved XRF and XANES to probe P speciation at the micro or nano scale.

The XANES spectra of the standards collected at the nano beamline are reported in Fig. 5 . The main features of these spectra are similar to those discussed above even though some differences exist between the "bulk" standard spectra and those obtained with the nano beamline. These differences are possibly due to self-absorption or spatial heterogeneity in the samples but were not investigated. Both hydroxy- and fluoroapatite spectra were characterized by a broad postedge shoulder, while freshly precipitated Fe phosphate showed a pre-edge feature.


Figure 5
View larger version (16K):
[in this window]
[in a new window]
  		Fig. 5. Phosphorus n-XANES spectra for selected P standards. Data are background- and baseline-corrected and energy normalized to a reference energy of 2149 eV.

 
The experimental setup allowed collection of x-ray fluorescence (K edges) maps for various elements at the nano scale. The distribution of P, Si, and Al across an area of 10 by 10 µm with a resolution of 60 nm is shown in Fig. 6 , where the spatial distribution of elements is depicted within the sample using the software MAPS (Vogt, 2003). This image clearly shows the highly heterogeneous distribution of P in soil with what appear to be small crystals highly enriched in P. The colocation of P, Si, and Al can be seen in the compound picture, in which the fluorescence of the three elements is presented in one image. In this section, as in all the other sections investigated, Si and Al were generally colocated, probably in the form of aluminosilicate (clays). Some hot spots of Si alone were identified, probably reflecting the presence of silicates such as quartz. Phosphorus was not associated with either Al or Si. This is confirmed by the correlation diagrams between fluorescence x-ray photons of P, Al, and Si (Fig. 7 ), where the counts for each element for each pixel of Fig. 6 were used (10 200 data points). A good correlation between Al and Si was observed, with only a few data points showing high Si but relatively low Al counts. In contrast, P fluorescence counts were uncorrelated with Al counts. Unfortunately the energy used to generate this fluorescence map does not produce fluorescence of Ca, thereby preventing the use of n-XRF to directly assess any association between Ca and P; however, we performed P XANES analysis of a few hotspots (the position of the XANES analysis is indicated in Fig. 6 with an arrow). The results obtained (Fig. 8 ) strongly suggest that a Ca phosphate was responsible for the high-P hotspots. In fact, the P XANES obtained is very similar to the spectrum of fluoroapatite.


Figure 6
View larger version (86K):
[in this window]
[in a new window]
  		Fig. 6. Nano x-ray fluorescence maps of Si, P, and Al in a thin section of technical-grade monoammonium phosphate (TG-MAP) treated soil (0–7.5-mm sample). In the small images, the individual signals of each element are shown with concentration proportional to brightness. The large image is the composite of the x-ray fluorescence signals of the three elements shown in the small images. The arrow indicates the location for which n-XANES analysis was conducted (Fig. 8 ).

 

Figure 8
View larger version (16K):
[in this window]
[in a new window]
  		Fig. 8. Representative P n-XANES spectra of P hotspots in samples collected at two distances from the point of fertilizer application (0–7.5 and 7.5–13.5 mm). The n-XANES of an area selected in Fig. 6 (P nanoparticle) are also shown. Data are background- and baseline-corrected and energy normalized to a reference energy of 2149 eV. MAP is granular monoammonium phosphate and TG-MAP is fluid technical-grade monoammonium phosphate.

 

Figure 7
View larger version (33K):
[in this window]
[in a new window]
  		Fig. 7. Correlations between the fluorescence signals of P vs. Al and Al vs. Si. Each point represents a pixel in Fig. 6.

 
All P hot spots analyzed (n = 25) using n-XANES in soil samples indicated the presence of Ca phosphates. An example of the spectra obtained in each soil sample is reported in Fig. 8. In all the spectra collected, the main feature was the broad shoulder in the postedge region between 15 and 5.5 eV. These spectra were similar to those reported by Beauchemin et al. (2003) for slightly alkaline and P-enriched soils and by Peak et al. (2002) for poultry litter (not alum amended). It can be concluded, therefore, that in the soil we tested, the formation of Ca phosphate seems to dominate P precipitation and that Al and Fe complexes, if present, were of secondary importance.

Due to the high intrasample heterogeneity, and the tendency to analyze hotspots when XRF maps are collected, it was difficult to observe differences between fertilizer treatments at the micro or nano scale; however, this technique provided clear evidence of interactions between P and other elements and indicated that Ca controlled P chemistry at the micro or nano scale. A combination of "bulk" and spatially resolved spectroscopic investigations provided a powerful approach to assess the fundamental mechanisms controlling P behavior in soil and their spatial occurrence in the highly heterogeneous soil environment.


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIAL AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
REFERENCES
 
This study confirms that P applied as fluid fertilizer diffuses more and remains more potentially available (more isotopically exchangeable) to plants in calcareous soil than P applied as a granular product. Direct spectroscopic evidence was also obtained that P is highly heterogeneously distributed in soil and that, at least in this highly calcareous soil, seems to be invariably associated with Ca rather than Fe at the nano scale. The "bulk" XANES analyses also revealed that in the vicinity of fertilizer granules, P precipitation in the form of octacalcium phosphate or apatite-like compounds is the dominant mechanism responsible for decreases in P lability. In contrast, when a fluid P fertilizer is applied to soil, more P remains in a form similar to that of monocalcium phosphate. As suggested by other researchers (Matar et al., 1992; Tunesi et al., 1999), low P concentrations would favor adsorption over precipitation, whereas in the vicinity of P fertilizer granules, higher P concentrations could be conducive to precipitation reactions. The limited diffusion of P from fertilizer granules (which leads to higher P concentration around the granules) may be due to a mass flow of water toward the highly hygroscopic P granule (Lawton and Vomocil, 1954). This mass flow of water occurs in a direction opposite to that of dissolved P diffusion, as has recently been demonstrated using an x-ray computed microtomography technique (Hettiarachchi et al., 2006).


   ACKNOWLEDGMENTS
 
We thank M.D. de Jonge for further analysis of spatially resolved fluorescence maps, R.E. Hamon for suggestions and for working with us at the APS, G. Hettiarachchi and Mike McLaughlin for their useful discussions, and Peter Self for sample preparation. This work was supported by the Australian Synchrotron Research Program, which is funded by the Commonwealth of Australia under the Major National Research Facilities Program, by the Victorian Department of Infrastructure and Regional Development under the Industry Access to Overseas Synchrotron Facilities program, and by a linkage grant from the Australian Research Council (LP0454086), the South Australian Grains Industry Trust and CSBP Ltd. E. Lombi gratefully acknowledges the support of Grains Research and Development Corporation. Use of the Advanced Photon Source was supported by the U.S. Department of Energy, Office of Science, Basic Energy Sciences, under Contract no. W-31-109-Eng-38. We are grateful to Drs. K. Tan and A. Jurgensen from CSRF and the staff of the Synchrotron Radiation Center, University of Wisconsin, Madison, for their technical support and the National Science Foundation for supporting the SRC under Award no. DMR-0084402.

Received for publication February 5, 2006.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIAL AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 




This article has been cited by other articles:


Home page
 Soil Sci.Home page
B. Ajiboye, O. O. Akinremi, Y. Hu, and A. Jurgensen
XANES Speciation of Phosphorus in Organically Amended and Fertilized Vertisol and Mollisol
Soil Sci. Soc. Am. J., September 1, 2008; 72(5): 1256 - 1262.
[Abstract] [Full Text] [PDF]

Home page
 Soil Sci.Home page
G. M. Hettiarachchi, M. J. McLaughlin, K. G. Scheckel, D. J. Chittleborough, M. Newville, S. Sutton, and E. Lombi
Evidence for Different Reaction Pathways for Liquid and Granular Micronutrients in a Calcareous Soil
Soil Sci. Soc. Am. J., January 11, 2008; 72(1): 98 - 110.
[Abstract] [Full Text] [PDF]

Home page
 J. Environ. Qual.Home page
B. Ajiboye, O. O. Akinremi, Y. Hu, and D. N. Flaten
Phosphorus Speciation of Sequential Extracts of Organic Amendments Using Nuclear Magnetic Resonance and X-ray Absorption Near-Edge Structure Spectroscopies
J. Environ. Qual., October 16, 2007; 36(6): 1563 - 1576.
[Abstract] [Full Text] [PDF]

Home page
 Soil Sci.Home page
B. Ajiboye, O. O. Akinremi, and A. Jurgensen
Experimental Validation of Quantitative XANES Analysis for Phosphorus Speciation
Soil Sci. Soc. Am. J., June 29, 2007; 71(4): 1288 - 1291.
[Abstract] [Full Text] [PDF]

Home page
 Soil Sci.Home page
Z. He, C. W. Honeycutt, T. Zhang, P. J. Pellechia, and W. A. Caliebe
Distinction of Metal Species of Phytate by Solid-State Spectroscopic Techniques
Soil Sci. Soc. Am. J., May 16, 2007; 71(3): 940 - 943.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Lombi, E.
Right arrow Articles by Paterson, D.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Lombi, E.
Right arrow Articles by Paterson, D.
Agricola
Right arrow Articles by Lombi, E.
Right arrow Articles by Paterson, D.
Related Collections
Right arrow Phosphorus
Right arrow Nutrients
Right arrow Nutrient Management
Right arrow Soil Chemistry

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Crop Science
Journal of Natural Resources
and Life Sciences Education		 Vadose Zone Journal
Journal of Plant Registrations Journal of
Environmental Quality		 The Plant Genome