JEQ Grow Your Career With ASA
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 5 January 2006
Published in J Environ Qual 35:293-302 (2006)
DOI: 10.2134/jeq2005.0285
© 2006 American Society of Agronomy, Crop Science Society of America, and 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 Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (8)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by McDowell, R. W.
Right arrow Articles by Cade-Menun, B. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by McDowell, R. W.
Right arrow Articles by Cade-Menun, B. J.
Agricola
Right arrow Articles by McDowell, R. W.
Right arrow Articles by Cade-Menun, B. J.
Related Collections
Right arrow Nuclear Magnetic Resonance, NMR
Right arrow Phosphorus
Right arrow Soil Analysis
Right arrow Soil Chemistry

TECHNICAL REPORTS

Organic Compounds in the Environment

An Examination of Spin–Lattice Relaxation Times for Analysis of Soil and Manure Extracts by Liquid State Phosphorus-31 Nuclear Magnetic Resonance Spectroscopy

R. W. McDowella,*, I. Stewartb and B. J. Cade-Menunc

a AgResearch, Invermay Agricultural Centre, Private Bag 50034, Mosgiel, Otago, New Zealand
b Department of Chemistry, University of Otago, P.O. Box 56, Dunedin, New Zealand
c Department of Geological and Environmental Science, Stanford University, Building 320, Room 118, Stanford, CA 94305-2115

* Corresponding author (richard.mcdowell{at}agresearch.co.nz)

Received for publication July 26, 2005.

   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Phosphorous (P)-31 nuclear magnetic resonance (NMR) spectroscopy is used in the analysis of P forms in extracts of soils and manures for environmental and agronomic purposes. Quantitative spectra require knowledge about spin–lattice relaxation times (T1) to ensure adequate delays between pulses. This paper determined T1 values of P forms in reconstituted (0.2 g in 0.7 mL–1) samples of freeze-dried 0.25 M NaOH plus 50 mM EDTA extracts of eight diverse soils (Aquept, Dystrochrept x 2, Hapludand, Rendoll, Udand, Haplostoll, and Orthod), three different manures (dairy cattle, deer, and sheep), and one epiphyte moss. Total concentrations in the reconstituted samples ranged from 5 to 175 mg Fe mL–1, 2 to 62 mg Mn mL–1, and 72 to 837 mg P mL–1. Values of T1 for orthophosphate monoesters, orthophosphate diesters, and pyrophosphate varied from 0.42 to 1.69 s in soils and from 0.89 to 2.59 s in manures and the epiphyte. In contrast, T1 for orthophosphate varied from 0.78 to 1.94 s in soils and 1.45 to 5.82 s in manures and the epiphyte. For quantitative 31P NMR, delay times should be three to five times the T1 value, translating to delays of 3 to 5 s for soils and up to 25 s for manures. If the required delay is too long then strategies such as adding paramagnetics could shorten T1, provided this does not increase line-broadening too much. A regression relationship was obtained between orthophosphate T1 values and the ratio of P concentration to Fe and Mn concentration on a w/v basis (r2 = 0.97, P < 0.001), and between the T1 for all other compound classes and the ratio of P to Fe and Mn (r2 = 0.70, P < 0.01). By combining measurement of Fe, Mn, and P in the reconstituted extract and these relationships, T1 can be estimated and the appropriate delay time used. If T1 is not considered and the delay time is too short, some peaks will be under- or over-represented and the relative distribution of P forms not quantitative.

Abbreviations: NMR, nuclear magnetic resonance • P/(Fe + Mn), ratio of phosphorus concentration to the concentrations of iron and manganese • T1, spin–lattice relaxation rate


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
PHOSPHORUS-31 NMR spectroscopy has become popular for the analysis of P forms in extracts of soils and manures. The technique is used in both agronomic and environmental fields to examine the changes with time and management in concentration and distribution of P, especially organic P forms. Indeed, Cade-Menun (2005) cited more than 130 publications dealing with the characterization of P by NMR in agricultural and environmental samples. Of the many extractions that have been employed, such as water (Koopmans et al., 2003), 0.01 M CaCl2 (McDowell et al., 1998; McDowell, 2003), bicarbonate (Baldwin, 1996), and NaOH (Newman and Tate, 1980), the combination of NaOH and EDTA (Cade-Menun and Preston, 1996) has become the most readily used. This technique, based on the Bowman and Moir (1993) extraction procedure for soil organic P, utilizes NaOH to liberate humic material and the chelator EDTA to enhance the extraction of humic materials associated with metals. For manure samples, Turner (2004) showed that a NaOH concentration of 0.25 M and an EDTA concentration of 50 mM yielded a good compromise between efficient organic P extraction and minimizing hydrolysis to yield the greatest possible peak diversity in spectra. Unfortunately, this technique also extracts paramagnetic species such as Fe and Mn ions.

During an NMR experiment, nuclei are excited by a radio-frequency pulse. As the excited nuclei relax back to equilibrium, the emitted energy is detected as an emission signal and is recorded as a peak after Fourier transformation. Nuclei relax by exchanging energy with their surroundings (spin–lattice relaxation) or with each other (spin–spin relaxation): T1 and T2, respectively. In solutions, spin–spin relaxation is negligible; spin–lattice processes dominate. This excitation–relaxation is repeated until sufficient signal has been detected for all P species, relative to background noise, to produce a clear spectrum. If the time between pulses is not great enough then excited nuclei may not fully relax to equilibrium. The signals for these nuclei may become saturated, reducing their relative intensity compared to other, fully relaxed, peaks. The result is inaccurate quantification when determining peak area by integration and an over- or under-representation of certain P species (Cade-Menun et al., 2002). However, relaxation that is too rapid will produce broad peaks, because the line broadening of a peak is inversely proportional to its lifetime. Paramagnetic ions such as Fe and Mn can substantially shorten relaxation times by facilitating the relaxation of 31P nuclei, thus reducing the delay required between pulses, and decreasing the total length of the NMR experiment. However, they also increase line broadening, such that peaks that would otherwise be well resolved can overlap, possibly masking some P species. Clearly there is a need to understand the relationship of solution paramagnetic ion concentrations to the relaxation rates of P species before accurate conclusions can be drawn from 31P NMR experiments on soils and other samples such as manures.

Much work has already outlined the problems associated with paramagnetic ions for NMR experiments on a wide range of nuclei. Although solid state and for C, Arshad et al. (1988) reported that a C to Fe ratio of >1 in the soil was required to generate good 13C cross polarization magic angle spinning NMR spectra, whereas ratios below this caused too much line-broadening. Smernick and Oades (1999) also quantified T1 for 13C in soils in relation to Mn2+, Fe3+, and Cu2+. However, for solution 31P NMR of environmental samples, only two studies have examined T1. The first of these studies examined the 0.5 M NaOH extract of a native tussock grassland soil in New Zealand and reported T1 values of 2.0 s for orthophosphate diesters, 2.1 s for orthophosphate, and 3.0 s for orthophosphate monoesters (Newman and Tate, 1980). The second study examined 0.25 M NaOH, 0.25 M NaOH + 50 mM EDTA, and Chelex resin in 0.25 M NaOH extracts of soil and litter under western red cedar (Thuja plicata Don.) (Cade-Menun et al., 2002). Values of T1 for orthophosphate in the NaOH + EDTA extracts were 0.21 s for soil and 0.83 s for litter, while the corresponding values for orthophosphate monoesters and diesters ranged from 0.38 to 0.82 s. The study of Cade-Menun et al. (2002) also drew parallels between the concentrations of paramagnetic ions in the NMR solution and T1, concluding that Fe and Mn concentrations should also be determined with P in soil extracts to be analyzed by 31P NMR spectroscopy. For their samples the P to paramagnetic (Fe + Mn) ratio (w/v) was 0.4 for litter and 1.0 for soil and a delay time of 1.5 to 2 s was determined as adequate for quantitative spectra. However, for many environmental samples, especially manures, the concentration of paramagnetic species to P is more highly variable. For example, Ajiboye et al. (2004) showed that the P to Fe ratio of a dairy manure sample was 0.6, while that of a hog manure sample was 17.9. Clearly this will have an effect on T1 and thus the delay times necessary to quantitatively examine P species in these samples with 31P NMR. Consequently, our objective was to examine the concentration of paramagnetic ions and P species, and to determine T1 values for different classes of P compounds extracted by 0.25 M NaOH + 50 mM EDTA extracts of a range of soils and manures, and one plant sample.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Soil and Manure Sampling and Analysis
Topsoil (0–7.5 cm, Table 1) samples were taken from eight soils, representing the major New Zealand soil orders. The Hari Hari sandy loam (USDA Taxonomy, Aquent) is a young soil rich in primary minerals and was taken under native bush, while all others were taken under low intensity sheep-grazed pasture. They were: Eyre silt loam (Dystrochrept), Ngakuru loam (Udand), Oamaru clay loam (Rendoll), Rotoiti silt loam (Hapludand), Stewart silt loam (Dystrochrept), Taupo sandy silt (Udand), and Waharekohe silt loam (Orthod). Each soil received as part of an annual fertility program about 35 kg P ha–1 for at least the last 25 yr. All samples were air-dried, crushed, and sieved (<0.5 mm). For manure, bulked samples (1 kg) of sheep and deer manure were taken from the AgResearch Invermay farm located near Mosgiel, Otago, New Zealand, in March 2004. On the same day, 20 cattle dung pats were subsampled from a nearby (1 km away) dairy farm and bulked together (1 kg). All manure samples were taken within 1 h of excretion. Each manure was left to air-dry at 25°C in a temperature-controlled glasshouse for 20 d, ground, and sieved to <0.5 mm. One plant sample was also taken of an epiphyte moss (Macromitrium longipes) common in New Zealand native bush. This was included as it was suspected to be poor in paramagnetic ions like the manures and hence was treated the same.


View this table:
[in this window]
[in a new window]
  		Table 1. Classification and general chemical characteristics of each soil and manure used.

 
Analyses included pH (in water, 1:2.5 w/v), organic C by a LECO (St. Joseph, MI) C/N analyzer, total P by aqua regia (4:1 concentrated HCl to HNO3) digestion (Crosland et al., 1995), and organic P by ignition (Saunders and Williams, 1955). A list of these data is given in Table 1.

Soil and Manure Extraction for Phosphorus-31 Nuclear Magnetic Resonance
Analysis of soils, manures, and the epiphyte by 31P NMR was made on resuspended NaOH + EDTA extracts (Cade-Menun and Preston, 1996). Briefly, 5 g of ground samples were shaken with 100 mL of 0.25 M NaOH + 0.05 M Na2EDTA for 16 h and centrifuged (4500 x g for 20 min), and the supernatant was filtered (<0.45-µm cellulose acetate syringe filter). Total P and organic P were determined in a subsample of the extract by colorimetry (Watanabe and Olsen, 1965) for orthophosphate before and after digestion by K2S2O8 (Eisenreich et al., 1975). Each extract was frozen and then freeze-dried.

Phosphorus-31 Nuclear Magnetic Resonance Spin–Lattice Relaxation (T1) Experiments
Solution 31P NMR spectra were obtained using a Varian (Palo Alto, CA) 500-MHz Inova NMR spectrometer with a 51-mm standard Oxford (Witney, UK) superconducting magnet, FTS (Stone Ridge, NY) temperature controller, and a 5-mm Varian z-axis PFG direct detection probe. Each sample was prepared to a pH of >13 by taking 0.2 g of the dried extract, and adding 600 µL of D2O and 100 µL of 10 M NaOH. Samples were ultrasonicated (Crest [Trenton, NJ] Model 175T) for 3 min, equilibrated for 20 min, then centrifuged at 7000 rpm (Qualitron 6 place mini-centrifuge) for 5 min. The supernatant was transferred to a 5-mm NMR tube and 31P NMR spectra obtained at 202.298 MHz at 20°C. Chemical shifts were recorded relative to an external phosphoric acid standard ({delta} = 0 ppm) in a capillary tube. Spectra were processed with 5-Hz line-broadening, using Mestre-C software (Gómez and López, 2004). After 31P NMR spectroscopy, the reconstituted sample was analyzed for total P, Fe, and Mn by inductively coupled plasma–mass adsorption spectroscopy after digestion by aqua regia.

The spin–lattice relaxation time for resolvable peaks was measured using an inversion recovery pulse sequence. The method used a two-pulse sequence where spin populations were inverted with a 180° pulse (p1). The recovery was monitored by waiting a variable time {tau} (0.0625, 0.25, 0.5, 2, and 6 s) followed by a 90° observation pulse (pw90 = 14.3 µs). The relaxation delay was 8 s. The same number of scans was taken at each {tau} value until a sufficient signal to noise (S/N) ratio (>150 calculated on the orthophosphate peak) was obtained.

At {tau} = 0 the spectrum was 180° out of phase. As {tau} increased the relaxation process resulted in a change in the intensity and phase of the signal. The signal passed through a null ({tau} = T1) and the full intensity of the signal was restored at {tau} > T1 x 5. The T1 analysis involved plotting {tau} vs. intensity for each peak followed by regression analysis using standard Varian software. The null point was calculated, which equaled T1.

Peak assignments for spectra (Fig. 1 ) were derived from the literature and ascribed to between 21 and 18 ppm for phosphonates (Newman and Tate, 1980), between 6.6 and 6.5 ppm for orthophosphate (Newman and Tate, 1980), between 3 and 6 ppm for orthophosphate monoesters (Newman and Tate, 1980), between –0.5 and 2.0 for orthophosphate diesters (including DNA assigned to between –0.3 and –0.25 and phospholipids assigned to between 1 and 2: McDowell and Stewart, 2005; Turner et al., 2003a, 2003b), and about –3.5 ppm for pyrophosphate (Newman and Tate, 1980). No polyphosphates, usually found at about –19 ppm, were evident in any spectra. The peaks representing orthophosphate, orthophosphate diesters, and pyrophosphate were used to calculate T1 values for their respective compound class in each sample. For orthophosphate monoesters, a single T1 value was determined by taking the mean of four peaks common among all spectra in the region between a chemical shift of 6.0 and 4.1 ppm. These peaks are largely those of inositol-hexakisphosphate (McDowell and Stewart, 2005; Turner and Richardson, 2004).



View larger version (16K):
[in this window]
[in a new window]
  		Fig. 1. Phosphorus-31 nuclear magnetic resonance (NMR) spectra of each reconstituted soil and manure sample at the longest {tau} employed. The term S/N is the signal to noise ratio. Regions within a spectrum for P compound classes are given for the Oamaru soil.

 

   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 CONCLUSIONS
 REFERENCES
 
General chemical characteristics of each sample are given in Table 1 and show the wide range of pH, organic C, organic P, and total P in the samples used for this study.

On average, organic P represented 50% of soil total P (Table 1). However, NaOH-EDTA extracted 70 to 96% of soil organic P and 51 to 95% of total soil P (Table 2). Organic P represented 12 to 15% of total P for the manures and 66% for the epiphyte, and NaOH-EDTA extracted 90 to 97% of organic P and 56 to 84% of total P from the manure and epiphyte samples. In general, relative to soils, concentrations of P and Mn in the reconstituted NMR samples were much greater in the manures (Table 2). However, Fe concentrations were much greater in the soils than manures. The concentrations of other paramagnetic ions such as Co2+, Ni2+, and Cu2+ were not measured, as the concentrations of these in soils, manures, and pastures are usually an order of magnitude less than Fe and Mn (Morton et al., 1999). The high concentration of P and low Fe and Mn concentration in deer and sheep manures led to the greatest P/(Fe + Mn) ratios. Ratios in most soils were in the 1.0 to 2.5 range, except for the Oamaru and Hari Hari soils at 4.5 and 5.3, respectively (Table 2).


View this table:
[in this window]
[in a new window]
  		Table 2. Organic and total P in the NaOH-EDTA extracts of each soil or manure, and the total Fe, Mn, P, and ratio of P to the sum of Fe and Mn concentration in the reconstituted sample.

 
An example of the spectra produced in an inversion recovery experiment to estimate T1 for the Taupo soil reconstituted extract is given in Fig. 2 . A plot of magnetization versus time for each peak was used to calculate T1 values, which for most peaks fell between the 0.25- and 2.0-s pulse delays (i.e., zero magnetization is where no peaks are present). The calculated T1 values for all peaks in soils were between 0.42 and 1.94 s: for manures and the epiphyte the range was wider, from 0.89 to 5.82 s (Table 3). Figure 3 gives examples of spectra of the Stewart and Taupo soils obtained using a similar number of scans but with either a 0.1- or 5-s delay, while Fig. 4 gives spectra of the sheep dung extract with delays of 0.1 and 15 s. Where present, obvious changes were evident for the orthophosphate diester (DNA), phosphonate, and pyrophosphate peaks: the decreased signal intensity in the 0.1-s spectrum relative to the 5- or 15-s delay indicates that nuclei were not fully relaxed between pulses with a delay time of 0.1 s. When peak areas were calculated and expressed relative to total NaOH-EDTA-extractable P, differences were noted among all compound classes between these spectra (Table 4). However, for feed and manure samples, analytical error has been calculated at 5% for large signals and 10% for small signals (Kemme et al., 1999; Leinweber et al., 1997). Assuming these errors translate to soils, then only differences in some monoester regions of both soils, the diester region of the sheep manure and Stewart soil, and pyrophosphate in the Taupo soil were significant. These changes could not be accounted for by alkaline hydrolysis during analysis. A comparison of spectra taken after 2 h and after 16 h (at the end of the T1 experiment) for the Taupo soil extract using a delay five times > T1 showed no difference.



View larger version (11K):
[in this window]
[in a new window]
  		Fig. 2. Inversion recovery stacked spectra of the Taupo NaOH-EDTA reconstituted soil extract. The numbers at the base of each spectrum indicate the delay time used.

 

View this table:
[in this window]
[in a new window]
  		Table 3. Measured spin–lattice relaxation rate (T1) values(s) and, in parentheses, percentage of total spectral area for each peak or compound class. Note, the T1 values for phosphonates (18–21 ppm) in the Stewart and Hari Hari soil extracts were 0.71 and 0.94 s, respectively; the T1 for a peak between 1 and 2 ppm in the Stewart soil extract was 0.79 s.

 


View larger version (17K):
[in this window]
[in a new window]
  		Fig. 3. Phosphorus-31 nuclear magnetic resonance (NMR) spectra of the Stewart and Taupo soil extracts using a delay of 0.1 and 5 s. The term S/N is the signal to noise ratio. Assignments are given for each class of P compound above their respective peak. Numbers refer to concentrations given in Table 4.

 

View this table:
[in this window]
[in a new window]
  		Table 4. Concentrations of P compound classes determined in the Stewart and Taupo soil extracts (percentage of spectral area given in parentheses) by 31P nuclear magnetic resonance (NMR) using a 0.1- and 5-s delay, and a 0.1- and 15-s delay for sheep dung.

 


View larger version (20K):
[in this window]
[in a new window]
  		Fig. 4. Phosphorus-31 nuclear magnetic resonance (NMR) spectra of the sheep dung extract using a delay of 0.1 and 15 s. The term S/N is the signal to noise ratio. Assignments are given for each class of P compound above their respective peak. Numbers refer to concentrations given in Table 4. The miniature spectra show the full intensity of orthophosphate relative to other peaks.

 
Low concentrations of Fe and Mn decrease line-broadening, but also lengthen T1. Relationships were generated between T1 and the concentrations of Fe or Mn in the reconstituted samples (e.g., Fig. 5 ). Not all relationships are presented, as most, such as those between Mn and T1 values, were not significant (P > 0.05). Importantly, for all P compound classes the best relationship was gained by comparing T1 to P/(Fe + Mn) (Fig. 6 ). Relationships were only generated between T1 and the concentrations of Fe and Mn in the reconstituted samples, as only the constituents soluble in the reconstituted sample are relevant to the NMR. However, for P a linear regression existed between the concentration in the reconstituted sample (y, mg mL–1) and that in the soil extract (x, mg kg–1): y = 0.121x + 59, r2 = 0.92 (significant at the 0.001 probability level). For Fe the relationship was y = 0.314x – 70, r2 = 0.81 (significant at the 0.01 probability level). No data were available for Mn in the soil extract. The relationship is not 1:1 as a precipitate is always left behind while trying to maximize the concentration of P in the NMR solution.



View larger version (14K):
[in this window]
[in a new window]
  		Fig. 5. Relationship between the concentration of Fe in the reconstituted sample and spin–lattice relaxation rate (T1) for orthophosphate. Manure samples are plotted with white fill.

 


View larger version (27K):
[in this window]
[in a new window]
  		Fig. 6. Linear regressions between spin–lattice relaxation rate (T1) values for orthophosphate (solid line) or orthophosphate monoesters (from the four peaks representing largely inositol hexakisphosphate), orthophosphate diesters, and pyrophosphate (dashed line), and the ratio of P to Fe and Mn in each reconstituted nuclear magnetic resonance (NMR) sample (on a mg L–1 basis). Note, the 95% confidence intervals for each relationship are given as the dotted lines and the number of asterisks denote significance at the 0.01 (**) and 0.001 (***) probability levels. Data from Cade-Menun et al. (2002) are included as solid black symbols. Manure samples are plotted with white fill.

 
An F test indicated that the slope of the linear regression between P/(Fe + Mn) and T1 values was significantly greater for orthophosphate than for all other peaks, which, with no detectable difference among classes (orthophosphate monoesters, orthophosphate diesters, or pyrophosphate; Fig. 6), were pooled together in one regression function. Exclusion of the two longest T1 values for orthophosphate in deer and sheep manures indicated these values did not significantly affect the slope of the relationship between T1 and P/(Fe + Mn).


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
CONCLUSIONS
 REFERENCES
 
Compared to other studies of T1 in alkaline soil extracts (Newman and Tate, 1980; Cade-Menun et al., 2002), concentrations of total P in most of our samples were greater, but still typical of the wide range of soils and agricultural systems studied by 31P NMR. The efficiency of extraction in NaOH-EDTA was good, ranging from 51% of total P in the Hari Hari sandy loam to 95% in the Wharekohe soil. In general, this compares well to other studies that have used NaOH-EDTA as an extractant. For example, Chen et al. (2004) were able to extract between 46 and 86% of total P from seven New Zealand soils in pasture. Similarly, Turner et al. (2003b) extracted between 45 and 88% of total P from 29 soils under pasture in the UK. Dai et al. (1996), Cade-Menun et al. (2002), and Koopmans et al. (2003) also reported good P recoveries in NaOH-EDTA extracts. In contrast, Turner et al. (2003a) could only extract between 14 and 45% of total P from 18 semiarid high-pH soils from the western United States. The low efficiency of extraction from these high-pH soils may be attributed to the inefficient extraction of Ca-P by NaOH-EDTA. Among our soils, Ca was not measured, but only the Oamaru soil (Rendoll) is expected to contain much Ca, and indeed exhibited poorer extraction efficiency (69% of total P) than most of the other soils of this study. Similarly, poor P extraction from the Hari Hari soil may be due to an increased concentration of P as the primary Ca-P mineral apatite, which is frequently found in young soils (Syers and Walker, 1969), but may be removed from older soils by weathering. Alternatively, due to a very high organic matter concentration the Hari Hari soil could retain much of its P in recalcitrant organic complexes.

Examination of our reconstituted NaOH-EDTA extracts indicated that P could be classified as phosphonates, orthophosphate, orthophosphate monoesters, orthophosphate diesters, and pyrophosphate. No polyphosphates were found in any spectrum. Changes due to hydrolysis may have occurred during extraction or during the long analysis time required to generate T1 values (Makarov et al., 2002). However, comparison of spectra at the end of the experiment to previous analysis of these soils showed a similar number and distribution of peaks. Furthermore, it is likely that most degradation would have occurred during soil extraction. Degradation during analysis by NMR was minimized and peak resolution made consistent by performing analyses at 20°C (Crouse et al., 2000).

In terms of T1 values, those determined here are of similar magnitude to those reported by Newman and Tate (1980) of between 2.0 s for orthophosphate diesters and 3.0 s for orthophosphate monoesters. However, the T1 values determined here are generally greater than those of Cade-Menun et al. (2002), which ranged from between 0.137 s for teichoic acid (defined here as phospholipids, a subset of orthophosphate diesters) and 0.797 s for phosphonates, with the majority < 0.5 s. Factors that affect T1 values include paramagnetic ion concentration, pH, and temperature. Common among the data of Newman and Tate (1980), Cade-Menun et al. (2002), and ours is that pH is held > 13, yielding consistent chemical shifts among samples. No paramagnetic ion concentration or temperature data are given in the study of Newman and Tate (1980). Cade-Menun et al. (2002) measured paramagnetic ion concentrations, but determined their T1 values at 32°C. While we are aware that this warmer temperature will increase relaxation times (Ramarajan et al., 1981), we calculated P to paramagnetic ion ratios and plotted the T1 data of Cade-Menun et al. (2002) with ours. The result is given in Fig. 6 and shows that the T1 values for orthophosphate fit within the 95% confidence intervals of our regression relationship if it is extended to the y axis. Data for orthophosphate monoesters and diesters were also aligned with their respective regression relationship, but one value for pyrophosphate was not. Data for phosphonates were not included because we had too few data for comparison.

Using the established regression relationship, delay times can be determined so that 31P NMR experiments can be conducted that will generate quantitative spectra. The general rule is that the total delay between pulses should be three to five times T1; due to the exponential nature of relaxation, 99% of nuclei will be fully relaxed with a delay of five times T1 (Canet, 1996). Depending on the concentration of P and paramagnetic ions in the sample, our data show that appropriate delays should be around 3 to 5 s for the soils but upward of 18 to 25 s for the deer and sheep manure. However, data in Fig. 3 and 4 shows that for a spectrum with a short delay « T1, all but the diester, some monoester, and pyrophosphate peaks were within analytical error of 5% for large peaks and 10% for small peaks. If spectrophotometer time is limited and the researcher wishes to only know the presence or absence of peaks or is less interested in diesters or pyrophosphate than other peaks, then using delays < T1 may be acceptable. However, if the researcher is interested in these peaks then spectra should be run with delays three to five times > T1. Clearly delays of 18 to 25 s may prohibit long runs so it may be possible to add paramagnetics to the sample to facilitate relaxation of nuclei provided this does not increase line-broadening too much that peaks cannot be separated from one another.

While some studies have delays of 2 to 3 s for soils, very few studies of manures have delays of more than this. For example, Leinweber et al. (1997) used a delay of 0.2 s to study NaOH extracts of swine and poultry manure. Crouse et al. (2000) give details of acquisition time (1.36 s) only: Fe or Mn concentrations are not reported. If insufficient delays between pulses were used in their study to account for T1 then their recommendations for optimal pH and temperature may have been based on semi-quantitative spectra.

Spin–lattice relaxation times have been used to inform the user of the relationship between compounds and paramagnetic metal complexes. For example, Elgavish and Granot (1979) altered the Fe concentration in solution with different model P compounds and used the measured T1 values to infer their degree of association. As mentioned above, the researcher could add paramagnetics to the solution to shorten T1 and the required delay. The quantity of paramagnetics could be determined from the relationship in Fig. 6. However, care must be taken to not increase line-broadening and thus the ability to separate peaks. Bortiatynski et al. (1997) used T1 values to determine the degree and type of association between 13C-phenol and humic acid. In our extracts, P was solubilized by OH– and further aided by the chelation of metals with EDTA. Metals, including paramagnetic ions, stay complexed with EDTA in solution (Owens et al., 2000), which increases the T1 of previously associated compounds since the ability of metals to dissipate energy from these compounds is decreased.

Many P species are thought to be bound with paramagnetic ions such as Fe and Mn in the soil, especially in acid soils. These include orthophosphate, orthophosphate monoesters, and pyrophosphate (Miltner et al., 1998; Celi et al., 1999; Leytem et al., 2002; Sutter et al., 2002). Cade-Menun et al. (2002) suggested that the relationship of metals to P species could be inferred from T1. Our data show two points: first, the slope of the relationship between orthophosphate and the ratio of P to paramagnetic ions is much greater than for any other compound class; and second, there is no statistical difference in slope among any other compound class and hence only one relationship is given. However, this does not mean that with addition samples, statistical differences could be found between these classes. Provided that chelation of paramagnetic metals was efficient one possibility is that orthophosphate that was directly associated with Fe and Mn via sorption with hydrous-oxides is now dependant on soluble paramagnetic ions to determine T1. In contrast, all other compound classes had a shorter T1 with increasing P to Fe and Mn ratio, which suggested they were less influenced by metal chelation and were in closer association with paramagnetic ions. This could, as Hawkes et al. (1984) suggest, be the result of the interaction of these compounds via organic moieties, but could also be the result of physical protection of metals from chelation by humic macromolecules. Trivalent ions such as Fe3+ are strongly held onto the organic exchange sites of humic material and may also form cationic bridges enabling humic material to adopt a condensed formation that excludes solvents (Swift, 1996). These interactions commonly include lignin and partly decomposed plant materials such as inositol phosphates, a major group of orthophosphate monoesters (Swift, 1996). However, further work should be directed at clarifying this interaction.


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
REFERENCES
 
The concentrations of the paramagnetic ions Fe and Mn influence the relaxation rate of P species and hence their T1 values in extracts of soil, manure, and one epiphyte. In reconstituted NMR samples, there was a relationship between orthophosphate T1 values and P/(Fe + Mn). A similar relationship existed between this ratio and pooled T1 values for orthophosphate monoesters, orthophosphate diesters, and pyrophosphate, but the slope was significantly less. For spectra to be quantitative, full relaxation must be achieved. Using delay times of three to five times the longest T1 for P species in a sample is considered best practice. If insufficient delays are used, then peaks may be over- or under-represented, producing information on the relative distribution of P compounds that is not quantitative. For most of the soils used in this study, this translated to a 3- to 5-s delay. However, for the manures, delays of 18 to 25 s were necessary for full relaxation of all P species. Using the generated relationship, the delay times needed to produce quantitative spectra can be estimated by measuring the relative concentrations P, Fe, and Mn in the reconstituted NMR sample before NMR analysis. If the required delay is too long to generate a reasonable number of scans within the required time frame then paramagnetics could be added to decrease T1, but this requires further work. This may be especially important for extracts of manures and other materials such as wood or plants where, unlike soils, paramagnetic ion concentrations may be poor relative to P.


   ACKNOWLEDGMENTS
 
Funding for this work was provided by the New Zealand Foundation for Research Science and Technology under Contract AGRX002. We thank Dr. Mervyn Thomas for helpful comments.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 




This article has been cited by other articles:


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. L. Turner, S. Newman, A. W. Cheesman, and K. R. Reddy
Sample Pretreatment and Phosphorus Speciation in Wetland Soils
Soil Sci. Soc. Am. J., August 9, 2007; 71(5): 1538 - 1546.
[Abstract] [Full Text] [PDF]

Home page
 Soil Sci.Home page
W. J. Dougherty, R. J. Smernik, E. K. Bunemann, and D. J. Chittleborough
On the Use of Hydrofluoric Acid Pretreatment of Soils for Phosphorus-31 Nuclear Magnetic Resonance Analyses
Soil Sci. Soc. Am. J., June 8, 2007; 71(4): 1111 - 1118.
[Abstract] [Full Text] [PDF]

Home page
 J. Environ. Qual.Home page
Z. He, B. J. Cade-Menun, G. S. Toor, A.-M. Fortuna, C. W. Honeycutt, and J. T. Sims
Comparison of Phosphorus Forms in Wet and Dried Animal Manures by Solution Phosphorus-31 Nuclear Magnetic Resonance Spectroscopy and Enzymatic Hydrolysis
J. Environ. Qual., May 25, 2007; 36(4): 1086 - 1095.
[Abstract] [Full Text] [PDF]

Home page
 J. Environ. Qual.Home page
G. F. Koopmans, W. J. Chardon, and R. W. McDowell
Phosphorus Movement and Speciation in a Sandy Soil Profile after Long-Term Animal Manure Applications
J. Environ. Qual., January 9, 2007; 36(1): 305 - 315.
[Abstract] [Full Text] [PDF]

Home page
 Soil Sci.Home page
Z. He, T. Ohno, B. J. Cade-Menun, M. S. Erich, and C. W. Honeycutt
Spectral and Chemical Characterization of Phosphates Associated with Humic Substances
Soil Sci. Soc. Am. J., August 22, 2006; 70(5): 1741 - 1751.
[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 Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (8)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by McDowell, R. W.
Right arrow Articles by Cade-Menun, B. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by McDowell, R. W.
Right arrow Articles by Cade-Menun, B. J.
Agricola
Right arrow Articles by McDowell, R. W.
Right arrow Articles by Cade-Menun, B. J.
Related Collections
Right arrow Nuclear Magnetic Resonance, NMR
Right arrow Phosphorus
Right arrow Soil Analysis
Right arrow Soil Chemistry

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