SSSAJ Journal of Natural Resources and Life Sciences Education
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 8 June 2007
Published in Soil Sci Soc Am J 71:1111-1118 (2007)
DOI: 10.2136/sssaj2006.0300
© 2007 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 Google Scholar
Google Scholar
Right arrow Articles by Dougherty, W. J.
Right arrow Articles by Chittleborough, D. J.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Dougherty, W. J.
Right arrow Articles by Chittleborough, D. J.
Agricola
Right arrow Articles by Dougherty, W. J.
Right arrow Articles by Chittleborough, D. J.
Related Collections
Right arrow Soil Analysis
Right arrow Soil Chemistry

SOIL CHEMISTRY

On the Use of Hydrofluoric Acid Pretreatment of Soils for Phosphorus-31 Nuclear Magnetic Resonance Analyses

Warwick J. Dougherty*, Ronald J. Smernik, Else K. Bünemann and David J. Chittleborough

Soil and Land Systems, School of Earth and Environmental Sciences, Univ. of Adelaide, South Australia 5064, Australia

* Corresponding author (warwick.dougherty{at}dpi.nsw.gov.au).


   ABSTRACT
TOP
 NOTES
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Solid-state 31P nuclear magnetic resonance (NMR) has the potential to be used for investigating soil P forms without the need for extractions or pretreatment. Solid-state spectra typically suffer from poor resolution and low observability, however, in part due to interferences by paramagnetic species present in soil. Hydrofluoric acid treatment is routinely used in 13C and 15N NMR analysis of soil organic matter to remove these paramagnetic species and improve NMR spectra. We evaluated the use of HF pretreatment to improve 31P NMR spectra of four pasture soils. Hydrofluoric acid treatment of soils for 24 h resulted in almost complete removal of inorganic P (>90%), but also resulted in the loss of organic P (up to 49%). Mass-balance calculations revealed that much of the organic P removed was hydrolyzed. In contrast, most model organic P compounds were found to be resistant to acid-mediated hydrolysis. Little (<4%) of the P contained in phytic acid, ß-D-glucose phosphate, or DNA was hydrolyzed by HF in 24 h, although 29% of P in adenosine triphosphate was hydrolyzed. The observability of P (Pobs) by solid-state 31P NMR increased on HF treatment, but was still generally poor. Attempts to analyze the HF extracts by solution 31P NMR were unsuccessful due to the presence of high concentrations of paramagnetics. Comparison of solution 31P NMR spectra of NaOH–EDTA extracts of the soils before and after HF treatment indicated that specific organic P compounds, in particular inositol phosphates, were removed by HF. In this regard, HF treatment may have a role in the separation or fractionation of different organic P forms.

Abbreviations: CP, cross-polarization • DP, direct polarization • NMR, nuclear magnetic resonance • Pi, inorganic phosphorus • Po, organic phosphorus • Pobs, percentage of sample phosphorus observable by spin counting using 31P nuclear magnetic resonance • SSB, spinning side bands


   INTRODUCTION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
The forms and concentrations of P contained in soils vary widely. The various forms and their cycling in the soil system influence agricultural production and have off-site environmental impacts. Phosphorus contained in organic molecules (organic P) can constitute a significant proportion of soil P—often in excess of 50% of total soil P and in some cases up to 90% (Turner et al., 2003b)—and can play a major role in soil P dynamics (Oehl et al., 2001). Consequently, the characterization of soil organic P is an important step in understanding its role and behavior in the environment. There is a wide range of techniques available to study the forms of P in soil. These include relatively simple wet chemistry techniques, isotopic labeling, sequential fractionation, and a range of 31P nuclear magnetic resonance (NMR) techniques. In recent years, there has been a growing use of solution 31P NMR to study soil organic P such that it is now the most commonly used technique in the study of soil organic P.

Nuclear magnetic resonance spectroscopy is a powerful technique for characterization of soil organic components. It is used very differently depending on the nucleus of interest. Solid-state techniques are most commonly used for organic C and N (e.g., Skjemstad et al., 1994; Schmidt et al., 1997; Schmidt and Gleixner, 2005), whereas solution techniques are preferred for organic P (e.g., Cade-Menun, 2005; Turner et al., 2005). There are two reasons for this dichotomy. The NMR-active C and N nuclei have low magnetogyric ratios and low isotopic abundance, and thus can only be detected at relatively high concentrations. This is most easily achieved in the solid state. The NMR-active 31P nucleus has a higher magnetogyric ratio and is 100% abundant, so is readily detected at much lower concentrations. The lower magnetogyric ratios of 13C and 15N nuclei also benefit more than 31P from the sensitivity enhancement achieved with the solid-state cross-polarization (CP) technique. On the other hand, NMR signal dispersion is greater for organic C and N than for organic P. Resonances in 13C and 15N NMR spectra of organic matter span hundreds of parts per million, whereas most organic P species present in organic matter are monoesters of orthophosphate and resonate in a chemical shift range of no more than 5 ppm (Turner et al., 2003a). Thus 31P NMR spectra suffer much more from the broader line widths of the solid-state technique than do 13C and 15N NMR spectra. Therefore, the choice of NMR technique can be rationalized in terms of the relative strengths of solid-state and solution NMR: C and N require the greater sensitivity of the solid-state technique, whereas P requires the greater resolution of the solution technique.

The different ways NMR is used for the different elements in organic matter has resulted in emphasis on quite different aspects of organic matter chemistry. For C and N, NMR is mainly used to identify and quantify broad chemical classes (e.g., alkyl, O-alkyl, aromatic, and carbonyl) with an understanding that these functional groups are part of an organic "matrix" of polymers and large macromolecules. In contrast, for P, NMR has been used to identify the occurrence of P in specific, small, alkali-soluble molecules (e.g., phytate [Turner et al., 2003a; McDowell et al., 2005]). This difference in emphasis may be justified as small molecules constitute only a small proportion of organic C and N in most surface soils, whereas small molecules such as phytate may constitute a substantial proportion of organic P in soils. It is hard to conceive, however, that there is no organic P in the large and complex molecules that constitute the bulk of organic matter.

In a previous study, we observed that HF treatment, a technique widely used in the NMR analysis of organic C and N, greatly improved the solid-state NMR observability of organic P (Dougherty et al., 2005). We also found, however, that the recovery of organic P following HF treatment (41%) was substantially lower than is typical for organic C and N (e.g., >75% [Skjemstad et al., 1994; Schmidt et al., 1997; Schmidt and Gleixner, 2005]). There are two possible explanations. First, it may be that a large proportion of organic P is present as small molecules such as phytate that are held in soil through strong interactions with clays (Turner et al., 2002) such that once the clays are removed by HF, these molecules are lost in the HF solution. On the other hand, this loss of organic P may be caused by the hydrolysis of organic P molecules under the acidic conditions (Anderson, 1980; Makarov et al., 2002; Turner et al., 2003a) of HF treatment.

In this study, we examined the potential role of HF pretreatment of soil in the NMR analysis of soil organic P using four soils with a wide range of soil P composition. First, we investigated whether HF treatment can be used in the same way as it is for organic C and N NMR analysis, i.e., as a way of improving (i) sensitivity (by removing the mineral component), (ii) resolution (by decreasing paramagnetic broadening of resonances), and (iii) quantitation (by reducing selective signal loss of species associated with paramagnetic species) in solid-state NMR spectra. Second, we examined whether HF treatment can be used (in an entirely different way) as a means of differentiating organic P in small, water-soluble molecules and organic P in the macromolecular organic matter matrix. Third, we examined the stability of organic P under the acidic conditions associated with HF treatment. We considered this issue by examining the stability of model organic P compounds in HF solution, using relatively short HF treatment times and calculating mass balances of organic and inorganic P in the HF treatment process.


   MATERIALS AND METHODS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Soils
Four soils were collected from sites in southeast Australia under permanent pasture production (Table 1). They were selected to have a range of P contents and forms of P, viz., organic vs. inorganic. Total P content of the soils (determined by a nitric–perchloric acid digest [Olsen and Sommers, 1982]) ranged between 545 and 1299 mg kg–1 and organic P (determined by ignition [Saunders and Williams, 1955]) comprised between 28 and 61% of total P. Olsen P (Olsen et al., 1954) was determined as a measure of plant-available P. Taxonomic classifications of the soils are also indicated in Table 1.


View this table:
[in this window]
[in a new window]

  		Table 1. Physical and chemical properties of the soils used in this study.

 
Hydrofluoric Acid Extraction of Soil
One gram of each soil was shaken with 50 mL of HF (2%) for up to 24 h to examine the P release. At each of 1, 2, 4, 8, and 24 h, subsamples (1 mL) of the HF–soil solution supernatant (after 10-min centrifugation at 1200 x g) were vacuum evaporated at –20°C. Vacuum evaporation was necessary to improve the safety of handling for subsequent analysis and to reduce the acidity of the samples for subsequent colorimetric determination of P in the solutions. When the samples were ready to be analyzed, they were reconstituted to 10 mL with deionized water (10x dilution), mixed thoroughly on a vortexer, and analyzed for total P (TP) and inorganic P (Pi) immediately. Preliminary experiments involving analysis of spiked and unspiked solutions after vacuum evaporation and subsequent dilution showed P recoveries of >95% (data not presented). An estimate of Pi was made by analyzing samples for molybdate-reactive P using molybdenum blue colorimetry (Murphy and Riley, 1962). Although we acknowledge the limitations of this estimate of Pi (i.e., it may include some hydrolyzed organic P and will not include inorganic P in the form of poly- and pyrophosphates), it is the most commonly used and practical estimate of Pi in solution (e.g., Turner et al., 2005; McDowell et al., 2006). Total P in the samples was determined using molybdenum blue colorimetry after digestion of the samples using an acidic persulfate procedure (American Public Health Association, 1995). The difference between TP and Pi is defined as unreactive P and provides an estimate of organic P (Po). As with Pi, this estimate is subject to a number of errors because it includes some inorganic P in the form of poly- and pyrophosphates (Turner et al., 2005) but again it is the most widely used estimate of Po in solution. Soil sample extractions and subsequent P determinations were all performed in duplicate.

Stability of Model Organic Phosphorus Compounds in Hydrofluoric Acid
To investigate the possible hydrolysis of organic P compounds by HF during the extraction procedure, a number of model organic P compounds representing important organic P classes were added to HF and subsequently analyzed in the same manner as the soils, as described above. The model P compounds examined were phytic acid (Na salt from corn, Sigma P 8810, Sigma-Aldrich Corp., St. Louis, MO), ß-D-glucose phosphate (Na salt, Sigma G 7879), DNA (Na salt from salmon testes, Sigma D 1626), and adenosine triphosphate (ATP; disodium salt from microbial source, Sigma A 3377). These were made up so that on addition of HF to make 2% HF solution, P concentrations were 10 mg L–1. Immediately following addition of HF, the solutions were shaken at 20°C and 1-mL aliquots taken for analysis at 1, 2, 4, 8, and 24 h. These samples were evaporated, reconstituted, and analyzed as described above for the soil extractions. Again, all incubations and analyses were performed in duplicate.

Sodium Hydroxide–Ethylenediaminetetraacetic Acid Extraction of Organic Phosphorus
Standard procedures were used for extraction of organic P from both the whole soils and HF-treated residues (Cade-Menun and Preston, 1996). This involved shaking 2.5 g of soil with 50 mL of 0.25 M NaOH and 0.05 M Na2EDTA for 16 h, followed by centrifugation (1300 x g) and filtration of the supernatant though a Whatman no. 42 filter paper. Total P in the supernatant was measured by inductively couple plasma–atomic emission spectrometry after digestion with HNO3. The filtered supernatant was analyzed for Pi using colorimetry following a 100x dilution (Turner et al., 2003b). The difference between TP and Pi was again defined as Po (see the discussion above of the limitations of these definitions). The remaining supernatant was frozen (–20°C) and freeze-dried in preparation for solution 31P NMR analysis as described below.

Solution Phosphorus-31 Nuclear Magnetic Resonance Analysis
The freeze-dried NaOH–EDTA extracts were ground in a mortar and pestle to optimize their subsequent dissolution. Approximately 500 mg of the ground extract was dissolved in 5 mL of deionized water. The reconstituted extracts had pH >13, ensuring chemical shifts in the NMR spectra consistent with those in the literature and maximizing differentiation of species (Turner et al., 2003a). The solution was centrifuged (1300 x g) for 20 min to remove particles >0.1 µm in diameter. The supernatant solution (3.5 mL) and D2O (0.3 mL) were placed in a 10-mm NMR tube. Solution 31P NMR spectra were acquired at 24°C on a Varian INOVA 400 NMR spectrometer (Varian, Palo Alta, CA) at a 31P frequency of 161.9 MHz. Recovery delays for the whole-soil extracts were in the range of 15 to 20 s, and were set to at least five times the T1 value (Canet, 1996) of the orthophosphate resonance determined in preliminary inversion-recovery experiments (data not shown). The T1 value for orthophosphate is generally greater than those of organic P resonances (McDowell et al., 2006). A recovery delay of 30 s was used for the HF residue extracts. This value was greater than five times the T1 value for the orthophosphate resonance of the Camden HF-residue extract. Given the much lower orthophosphate concentration of the other HF residues, inversion-recovery experiments were not performed on the other HF-residue extracts. We used a 90° pulse of 60 to 80 µs, an acquisition time of 1.0 s, and broadband 1H decoupling. Between 3000 and 3600 scans were acquired for the whole-soil extracts, representing an acquisition time of 15 to 18 h, and between 2700 and 5000 scans were acquired for the HF-treated soil extracts, representing an acquisition time of 23 to 42 h.

Chemical shifts were referenced to external 85% H3PO4 (Turner et al., 2003a). The spectra presented have a line broadening of 2 Hz. The percentage of the total 31P NMR detectable P that each of four broad classes of P accounted for was quantified by integration of the entire spectra and subsequent apportioning based on chemical shift. Peak assignments were as follows: orthophosphate, 6.2 to 5.6 ppm; P monoester, 5.6 to 3.6 ppm; P diester, 0.5 to –1.0 ppm; pyrophosphate, –4.5 to –5.5 ppm.

Solid-State Phosphorus-31 Nuclear Magnetic Resonance Analysis and Spin Counting
Solid-state 31P NMR spectra were acquired with magic angle spinning (MAS) and high-power 1H decoupling on a Varian Unity INOVA 400 spectrometer (Varian, Palo Alta, CA) with a Doty Scientific supersonic MAS probe (Doty Scientific, Columbia, SC) at a 31P frequency of 161.9 MHz. Samples were packed into 7-mm-diam. cylindrical zirconia rotors with Kel-F end caps (Doty Scientific) and spun at 5 kHz at the magic angle. Free induction decays were acquired with a sweep width of 50 kHz. A total of 1216 data points was collected for all spectra, representing an acquisition time of 12 ms. All spectra were zero-filled to 131072 data points and processed with a 100-Hz Lorentzian line broadening and a 0.010-s Gaussian broadening. Chemical shifts were externally referenced to NH4H2PO4 at 0.72 ppm (Frossard et al., 2002).

Phosphorus-31 CP NMR spectra of the whole soils and HF residues represent the accumulation of 20000 scans and were acquired using a 1-ms contact time and a 1-s recycle delay (Dougherty et al., 2005). The total acquisition time for 31P CP NMR spectra was around 6 h (3 h for the HF-treated residue of the Flaxley soil).

Phosphorus-31 direct polarization (DP) NMR spectra of the whole soils and HF residues represent the accumulation of 3400 to 10000 scans and were acquired using a 20-s recycle delay (Dougherty et al., 2005). The total acquisition time for 31P DP NMR spectra was around 19 to 56 h. The 31P DP NMR spectra were corrected for a broad background signal by subtracting the 31P DP NMR spectrum acquired for an empty rotor (there was no such background signal for 31P CP NMR spectra).

Phosphorus-31 spin-counting experiments were performed using the method of Dougherty et al. (2005). Ammonium dihydrogen phosphate was used as an external intensity standard (i.e., the NH4H2PO4 spectrum was acquired separately from those of the samples). For CP spin-counting experiments, differences in spin dynamics between the sample and the NH4H2PO4 standard were accounted for using the method of Dougherty et al. (2005). The T1{rho}H relaxation rate for NH4H2PO4 was found to be 120 ms. The T1{rho}H relaxation rate for the Camden whole soil was found to be 2.46 ms; this value was used in the T1{rho}H correction for all of the other whole soils. The T1{rho}H relaxation rate for the Camden HF residue was found to be 2.95 ms; this value was used in the T1{rho}H correction for all of the other HF residues.


   RESULTS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Hydrofluoric Acid Extraction of Soils
The majority of the P that was extracted in 24 h was extracted within the first hour (Fig. 1), with 54 to 72% of total soil P being solubilized after 1 h, increasing to 61 to 85% after 24 h (i.e., on average, 86% of the P extracted in 24 h was extracted in the first hour). Whereas the concentration of total and inorganic P in the HF extract increased slightly with extraction time (1–24 h), the concentration of organic P decreased by 15 to 23% between 1 and 24 h.


Figure 1
View larger version (22K):
[in this window]
[in a new window]

  		Fig. 1. Cumulative amount of organic P (black) and inorganic P (gray) extracted during the 24-h extractions of soils with HF.

 
Table 2 summarizes the distribution of P between extract and HF residue after 24 h of HF extraction. The HF was very efficient at solubilizing inorganic P, with only 2 to 3% of the inorganic P found in the residue after 24 h for three of the soils, and only 9% remaining for the Waite soil. Conversely, a relatively large proportion (51–62%) of the organic P was found in the residue. Total P recovery, summed across the extract and residue fractions, was 93 to 112%. For three of the four soils, however, only 72 to 80% of the organic P was recovered in the combined fractions, whereas there was 25 to 39% more inorganic P in the combined fractions than in the whole soil. For the Waite soil, recoveries of total, organic, and inorganic P were all in the range of 96 to 99%.


View this table:
[in this window]
[in a new window]

  		Table 2. Phosphorus fractions{dagger} following 24-h HF extraction.

 
Stability of Model Organic Phosphorus Compounds in Hydrofluoric Acid
Little hydrolysis of organic P was detected for phytic acid and ß-D-glucose phosphate (i.e., 4 and 0%, respectively). The DNA sample was insoluble in HF solution (although it is water soluble) and did not release any orthophosphate to solution after 24 h. Only ATP was found to partially hydrolyze, with 29% of total P detected as orthophosphate after 24 h in HF solution. In the first hour, however, only 5% of the ATP-P was hydrolyzed.

Solid-State Phosphorus-31 Nuclear Magnetic Resonance Analysis of Hydrofluoric Acid Residues
Solid-state 31P NMR CP and DP spectra of the whole soils and the residues following 24 h of HF extraction are shown in Fig. 2 and 3. The CP spectra of the whole soils are similar to those we have published previously (Dougherty et al., 2005), and consist of a broad resonance centered at 0 to –2 ppm and a series of strong spinning side bands (SSBs) at 31-ppm (5-kHz) spacings. The SSBs are a consequence of the rate of magic angle spinning (5 kHz) being smaller than the 31P chemical shift anisotropy. The CP spectra of the HF residues are similar in appearance to those of the whole soils, except for the HF residue of the Waite soil, which has less prominent SSBs. The DP spectra of the whole soils are also similar to those we have published previously (Dougherty et al., 2005). In addition to the broad resonances with prominent SSBs observed in the CP spectra, the DP spectra each contain a sharp resonance at around 2.5 ppm, which is associated with small first-order SSBs. These additional resonances in the DP spectra have been assigned to inorganic P (Dougherty et al., 2005), which is not detected in the CP spectra due to the absence of neighboring 1H nuclei. The DP spectra of the HF residues do not contain the sharp resonances observed in the DP spectra of the whole soils, reflecting their low inorganic P content (Table 2).


Figure 2
View larger version (46K):
[in this window]
[in a new window]

  		Fig. 2. Solid-state 31P cross-polarization nuclear magnetic resonance spectra of whole soils and residues following 24-h extractions with HF.

 

Figure 3
View larger version (30K):
[in this window]
[in a new window]

  		Fig. 3. Solid-state 31P direct polarization (DP) nuclear magnetic resonance spectra of whole soils and residues following 24-h extractions with HF. The Hamilton DP spectrum contained no discernible features and thus is not presented.

 
Table 3 shows the results of spin counting on the solid-state 31P NMR spectra of the whole soils and HF residues. Spin-counting measurements express the amount of NMR signal detected per 31P nucleus as a percentage of that detected for a standard (NH4H2PO4 in this case). Interpretation of solid-state NMR spectra is complicated by the fact that 31P nuclei in some environments (e.g., those close to paramagnetic centers) are not detected by NMR or give signals of reduced intensity. Spin counting provides a gauge of the extent of signal nondetection or underdetection (Dougherty et al., 2005).


View this table:
[in this window]
[in a new window]

  		Table 3. Observability (Pobs) of 31P in solid-state nuclear magnetic resonance spectra of whole soils and residues following 24-h HF extraction.

 
The CP and DP 31P NMR observabilities (Pobs) for the Camden and Flaxley soils (Table 3) were similar to those we had reported earlier for a series of soils collected from the Flaxley site (Dougherty et al., 2005). The higher observability for the DP spectra mainly reflect the detection of inorganic P, which is not detected by the CP technique. Both CP and DP observabilities were lower for the Waite and Hamilton whole soils, with the Hamilton soil producing no discernible signal in the DP spectrum. The generally low Pobs values for the whole soils are probably due to the presence of paramagnetic species that cause a range of problems with spectra (Cade-Menun, 2005). Differences in the amount, nature, and distribution of paramagnetic species probably account for the differences in Pobs values between the soils, with the Pobs for the Waite and Hamilton soils being the lowest (24 and 12%, respectively).

The 31P CP and DP observabilities were in all cases higher for the HF residues than for the corresponding whole soils. This is to be expected, given that HF removes paramagnetic species (Skjemstad et al., 1994). The NMR observabilities for the HF residues were not as high, however, as the values of 73 to 77% that we reported in a previous study in which HF treatment was extended to several days (Dougherty et al., 2005). It may be that the 24-h HF extraction was not sufficient to remove paramagnetic species to an extent where they no longer affected NMR observability. This is especially true for the Hamilton soil, which even after HF treatment gave a poor-quality CP spectrum with a low Pobs value of 12%, and no discernible signal in the DP spectrum.

Solution Phosphorus-31 Nuclear Magnetic Resonance Analysis of Hydrofluoric Acid Extracts
The 31P NMR spectrum of a neat HF extract of the Camden soil is shown in Fig. 4. It consists of only one broad resonance despite containing a mixture of inorganic and organic P (Table 2). The chemical shift of this resonance (0.8 ppm) is much lower than those of orthophosphate and monoester P observed in the NaOH–EDTA extract of the same soil.


Figure 4
View larger version (11K):
[in this window]
[in a new window]

  		Fig. 4. Solution 31P nuclear magnetic resonance spectrum of neat HF extract of Camden soil.

 
Attempts to transfer the sample into NaOH–EDTA were unsuccessful. Hydrofluoric acid was removed under vacuum and the residue extracted with water. The addition of NaOH–EDTA resulted in the formation of large amounts of gelatinous precipitate, presumably a mixture of Al, Si, and Fe hydroxides. The supernatant was separated from the precipitate by centrifugation, but it contained <10% of the P present in the initial HF extract. An alternate method was tried in which the residue following removal of HF was dissolved in water to get soluble P into solution before the addition of NaOH–EDTA and subsequent precipitation of the hydroxides. The rationale for this was that the EDTA would complex the soluble paramagnetics. Still, however, <10% of the P was recovered in the supernatant.

Solution Phosphorus-31 Nuclear Magnetic Resonance Analysis of Sodium Hydroxide–Ethylenediaminetetraacetic Acid Extracts of Whole Soils and Hydrofluoric Acid Residues
Of the total P in the whole Camden, Flaxley, Waite, and Hamilton soils, 73, 70, 54, and 66%, respectively, was extracted by NaOH–EDTA, whereas 69, 78, 70, and 102% of the total P in the HF residues of the Camden, Flaxley, Waite, and Hamilton soils, respectively, was extracted by NaOH–EDTA.

Figure 5 shows the 31P NMR spectra of the NaOH–EDTA extracts of the whole soils (top traces of each pair) and the corresponding NaOH–EDTA extracts following 24 h of HF extraction (bottom traces of each pair). There were no discernable resonances present outside the range of chemical shift we have presented in these figures. The spectrum obtained from the HF residue of the Waite soil was of very poor quality with no identifiable features and is not shown. This was probably because there was relatively little Po in the Waite soil (193 mg kg–1) compared with our other soils (≥332 mg kg–1) and others presented in the literature that often contain >400 mg kg–1 (e.g., Cade-Menun and Preston, 1996; Rheinheimer et al., 2002). Table 4 shows the distribution of signal among chemical shift regions derived from the spectra presented in Fig. 5. As expected, orthophosphate, which comprises the majority (51.0–72.0%) of signal in the spectra of the NaOH–EDTA extracts of the whole soils, represents a much smaller proportion (12.2–16.8%) of total signal in the spectra of the NaOH–EDTA extracts of the HF residues (Table 4). In contrast, the proportions of signal in the organic P regions (monoester and diester) are much higher in the 31P NMR spectra of the HF-treated soils (Table 4). Pyrophosphate represents similar proportions of total signals for extracts of both whole soils and HF-treated residues.


Figure 5
View larger version (24K):
[in this window]
[in a new window]

  		Fig. 5. Solution 31P nuclear magnetic resonance spectra of NaOH–EDTA extracts of whole soils (top traces of each pair) and HF residues (bottom traces). Resonances assigned to phytate are marked with an asterisk (*).The spectrum obtained from the HF residue of the Waite soil was of very poor quality and is not shown.

 

View this table:
[in this window]
[in a new window]

  		Table 4. Percentage of total 31P nuclear magnetic resonance signal in diagnostic chemical shift regions of spectra presented in Fig. 4.{dagger}

 
There are consistent differences between the spectra of the whole-soil and HF-residue extracts in the monoester region. The two strongest monoester resonances in the spectra of the whole-soil extracts (at approximately 5.0 and 4.7 ppm) are preserved in the spectra of the HF-residue extracts (Fig. 5). On the other hand, minor resonances in the spectra of the whole-soil extracts at 4.8 to 4.9, 4.5, and 4.3 to 4.4 ppm, which we have previously identified as phytate (Smernik and Dougherty, 2007), are not present in the spectra of the HF residue extracts. A prominent resonance at approximately 4.0 ppm in the spectra of the whole soil extracts is also not present in the spectra of the HF-residue extracts. This resonance has been assigned previously to an isomer of phytate (Turner and Richardson, 2004). The other main difference between the spectra of the whole-soil and HF-residue extracts is that the broad hump that extends from 4 to 6 ppm is stronger in the former than the latter.


   DISCUSSION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
CONCLUSIONS
 REFERENCES
 
The primary objective of this study was to determine whether HF pretreatment of soil could be useful in aiding the NMR characterization of soil organic P. There are three possible fates for soil organic P on HF treatment—it can be extracted into solution, it can remain in the residue, or it can be hydrolyzed to produce orthophosphate. Table 2 shows that after 24-h extraction in 2% HF at room temperature (20°C), 18 to 21% of the soil organic P was detected in the HF extract for three of the four soils, and 43% for the Waite soil. Table 2 also shows that 51 to 62% of the organic P was found in the residue. This means that 72 to 81% of the organic P was recovered in three of the four soils (96% for the Waite soil), with the remainder presumably being hydrolyzed. A mass balance of orthophosphate supports this assumption, with 25 to 39% more orthophosphate detected in the combined fractions than in the whole soils for all soils except Waite (99% recovery of orthophosphate). Figure 1 provides further evidence for hydrolysis of organic P, with the organic P content decreasing with increasing exposure to HF. It has previously been reported that a range of organic P forms is susceptible to some hydrolysis under acidic conditions (e.g., Anderson, 1980; Makarov et al., 2002; Turner et al., 2003a).

The apparent ease of hydrolysis of a substantial proportion of soil organic P during HF treatment contrasts with the stability in HF of the model organic P compounds we examined. There was no hydrolysis detected on 24-h exposure to HF for phytic acid and ß-D-glucose phosphate. Only the triphosphate ester ATP was partially hydrolyzed, with 29% of P released as orthophosphate, although only ~5% was hydrolyzed in the first hour; however, ATP typically constitutes only a very small proportion of soil organic P. Thus it is likely that only a small fraction of the P hydrolyzed in the soils can be explained by the presence of ATP or similar triphosphate esters.

Ideally, it would be very valuable to be able to obtain NMR characterization of organic P in both the HF extract and residue fractions. Unfortunately, we were not able to obtain any meaningful NMR characterization of the HF extract. The 31P NMR spectrum of the neat HF extract of the Camden soil shown in Fig. 4 contained only one very broad resonance and provided no useful discrimination of P forms. There are two possible reasons why the NMR spectrum of the HF extract was so broad. First, resonances in solution 31P NMR spectra tend to be broader at low pH than in alkaline NaOH–EDTA solution (Crouse et al., 2000). Second, the HF extract contains high concentrations of paramagnetic ions (HF is used specifically to remove these species from the soil and is very effective at this [Mathers et al., 2002]), which will also contribute to the broadening of resonances.

Attempts were made to isolate the organic P in the HF extract and to transfer it into an alkaline solution for NMR analysis. These were unsuccessful, however, due to the formation of a gelatinous precipitate under alkaline conditions—most likely a mixture of Al, Fe, and Si hydroxides. Approximately 50% of the soil mass for the Camden soil (and similar proportions for the other soils), was lost on HF treatment, presumably as a result of the solubilization of aluminosilicate clay minerals. The addition of hydroxide solution resulted in the immediate conversion of these compounds into insoluble hydroxides. The HF-treatment supernatant was separated from the precipitate by centrifugation, but it contained very little of the P in the whole supernatant (<10%). Alternative attempts to isolate organic P were also tried and found to be unsuccessful.

The HF residue fraction was analyzed using both solid-state and solution 31P NMR spectroscopy, although the solid-state NMR characterization proved to be of little value due to poor resolution and quantitation. As was the case in our previous study (Dougherty et al., 2005), the only differentiation of P types provided by solid-state 31P NMR was between orthophosphate, which gave rise to a sharp resonance at 2.5 ppm with small SSBs in the DP spectra and no signal in the CP spectra, and organic monophosphate esters, which gave rise to a broad resonance centered at 0 to –2 ppm with prominent SSBs in both the DP and CP spectra (Fig. 2 and 3). The solid-state spectra were broadly consistent with the chemical analyses, in that HF treatment resulted in the loss of the orthophosphate resonance. The solid-state 31P NMR spectra could not be used to quantify organic and inorganic P, however, because the NMR observabilities were generally low, especially for the whole soils. It is interesting to note that although HF treatment improved both CP and DP NMR observability, the Pobs values did not reach the values of 73 and 77% for CP and DP 31P NMR, respectively, that we reported in a previous study in which a more exhaustive HF treatment was used (Dougherty et al., 2005). The 24-h HF extraction resulted in a loss of ~50% of the mass of the original soil, whereas the normal extended HF extraction for 7 d typically results in losses of 70 to 90% (Mathers et al., 2002). Therefore, we hypothesize that although 24 h of HF extraction is sufficient to remove almost all of the inorganic P, it is not long enough to remove a sufficient amount of the constituents (presumably paramagnetic species) that interfere with NMR observability in the solid state. This is particularly true for the Hamilton soil, where even after HF treatment, P was detected with only 12% efficiency with the CP technique and no signal was detected with DP. It is likely that longer exposure to HF would result in lower recoveries of organic P. Consequently, it would appear that solid-state 31P NMR may only be useful for soil P characterization in limited circumstances, such as in polyphosphate fertilizer studies (e.g., McBeath et al., 2006).

Comparison of 31P NMR spectra of NaOH–EDTA extracts before and after HF treatment (Fig. 5) shows that some species are removed by HF treatment while others are preserved in the residue, (i.e., HF treatment results in a chemical fractionation of soil organic P). As we have previously proposed (Smernik and Dougherty, 2007), it would be tempting to assign the most prominent peaks in the monoester region at approximately 5.0 and 4.7 ppm to phytate, as it is commonly cited as being a major form of monoester organic P; however, these resonances are still prominent in the solution 31P NMR spectra of HF-treated soils. This supports our previous contention that these resonances are not due to phytate (Smernik and Dougherty, 2007). Resonances assigned to diester P of DNA are retained in the spectra of the extracts of the HF residue. This is consistent with the result for the model organic P compounds in which DNA was found to be insoluble in HF and resistant to acid-mediated hydrolysis. The other main difference between the 31P NMR spectra of the NaOH extracts of the whole soils and HF extracts is that the broad hump between 4 and 6 ppm is more prominent in the former than the latter. We hypothesize that there are two possible reasons for this. First is that HF treatment resulted in a loss of paramagnetics from the soil such that the subsequent NaOH–EDTA extract also contained less paramagnetics. The presence of paramagnetics can result in broadening of peaks (e.g., Cade-Menun et al., 2002). Second, the broad feature may be due to complexed organic P in a wide range of chemical environments in the organic matter matrix (as opposed to the sharp resonances, which represent specific, relatively small P-containing molecules). A number of early researchers in this field noted that difficulty in accurately quantifying phytate in alkaline soil extracts was related to its existence in complexed forms (Anderson and Hance, 1963; Hong and Yamane, 1981; Borie et al., 1989). As noted above, the idea that organic P occurs exclusively in small molecules that will be solubilized under alkaline conditions appears to be an assumption and not a proven fact. There is an apparent disconnect between the accepted model of organic C and N—which are mainly tied up in large, "non-biopolymers"—and the accepted model of organic P, which exists exclusively as small and recognizable biomolecules. That the broad feature is smaller following HF treatment suggests that organic P in these environments is susceptible to acid hydrolysis. The nature of this organic P warrants further investigation.


   CONCLUSIONS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
REFERENCES
 
The HF-treatment technique we used resulted in improved Pobs in solid-state 31P NMR and was very efficient at removing inorganic P (>90%). It also removed up to 50% of organic P, however, despite experiments with model organic P compounds showing that they were largely resistant to acid-mediated hydrolysis in 2% HF. In the solid-state 31P NMR spectrum of HF-treated soils, only a single central broad resonance was observed that has previously been identified as organic P. No differentiation of organic P forms within this broad resonance was possible, however.

Attempts to examine P species in the HF extract using solution 31P NMR were unsuccessful, presumably because of the presence of high concentrations of paramagnetics. Various attempts to remove these paramagnetics were unsuccessful. Hydrofluoric acid treatment did provide a form of chemical fractionation and resulted in the removal of inositol phosphate and at least one of its isomers and provided evidence in support of our contention that large prominent peaks in the solution 31P NMR spectrum of NaOH–EDTA extracts of our soils were not phytate (Smernik and Dougherty, 2007). Our experiments indicate that the use of HF treatment for 31P NMR analysis as is routinely used for 13C and 15N NMR experiments is of limited value. Alternative techniques to improve the resolution and Pobs of solid-state 31P NMR spectra require investigation.


   ACKNOWLEDGMENTS
 
W.J. Dougherty was supported by New South Wales Primary Industries and a Dairy Australia study scholarship (UA11189).


   NOTES
TOP
 NOTES
ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Current Address: New South Wales Dep. of Primary Industries Locked Bag 4 Richmond, New South Wales 2753 Australia

Current Address Institute of Plant Sciences ETH Zurich, Eschikon 33 CH-8315 Lindau Switzerland

All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Permission for printing and for reprinting the material contained herein has been obtained by the publisher.

Received for publication August 28, 2006.


   REFERENCES
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 





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 Google Scholar
Google Scholar
Right arrow Articles by Dougherty, W. J.
Right arrow Articles by Chittleborough, D. J.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Dougherty, W. J.
Right arrow Articles by Chittleborough, D. J.
Agricola
Right arrow Articles by Dougherty, W. J.
Right arrow Articles by Chittleborough, D. J.
Related Collections
Right arrow Soil Analysis
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