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


     

Published online 24 October 2007
Published in J Environ Qual 36:1856-1863 (2007)
DOI: 10.2134/jeq2006.0563
© 2007 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 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 Google Scholar
Google Scholar
Right arrow Articles by Güngör, K.
Right arrow Articles by Karthikeyan, K. G.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Güngör, K.
Right arrow Articles by Karthikeyan, K. G.
Agricola
Right arrow Articles by Güngör, K.
Right arrow Articles by Karthikeyan, K. G.
Related Collections
Right arrow Phosphorus
Right arrow Nutrients
Right arrow Animal Waste

TECHNICAL REPORTS

Waste Management

Determination of Phosphorus Speciation in Dairy Manure using XRD and XANES Spectroscopy

Kerem Güngöra, Astrid Jürgensenb and K. G. Karthikeyana,*

a Biological Systems Engineering Dep., Univ. of Wisconsin-Madison, 460 Henry Mall, Madison, WI 53706
b Canadian Synchrotron Radiation Facility, Synchrotron Radiation Center, 3731 Schneider Dr., Stoughton, WI 53589-3097

* Corresponding author (kkarthikeyan{at}wisc.edu).

Received for publication December 29, 2006.

   ABSTRACT
TOP
 NOTES
 ABSTRACT
INTRODUCTION
 Materials and Methods
 Results and Discussion
 Conclusions
 REFERENCES
 
Intensive manure application is an important source of diffuse phosphorus (P) pollution. Phosphorus availability from animal manure is influenced by its chemical speciation. The major objective of this study was to investigate the P speciation in raw and anaerobically digested dairy manure with an emphasis on the calcium (Ca) and magnesium (Mg) phosphate phases. Influent and effluent from an on-farm digester in Wisconsin were sampled and sieved, and the 25 to 53 µm size fraction was dried for X-ray powder diffraction (XRD) and P K-edge X-ray absorption near edge structure (XANES) analyses. Struvite (MgNH4PO4·6H2O) was identified in both the raw (influent) and anaerobically digested (effluent) manure using XRD. Qualitative analysis of P K-edge XANES spectra indicated that the Ca orthophosphate phases, except dicalcium phosphate anhydrous (DCPA) or monetite (CaHPO4), were not abundant in dairy manure. Linear combination fitting (LCF) of the P standard compounds showed that 57.0 and 43.0% of P was associated with DCPA and struvite, respectively, in the raw manure. In the anaerobically digested sample, 78.2% of P was present as struvite and 21.8% of P was associated with hydroxylapatite (HAp). The P speciation shifted toward Mg orthophosphates and least soluble Ca orthophosphates following anaerobic digestion. Similarity between the aqueous orthophosphate (aq-PO4), newberyite (MgHPO4·3H2O), and struvite spectra can cause inaccurate P speciation determination when dairy manure is analyzed solely using P K-edge XANES spectroscopy; however, XANES can be used in conjunction with XRD to quantify the distribution of inorganic P species in animal manure.

Abbreviations: ACP, amorphous calcium phosphate • ß-TCP, beta tricalcium phosphate • DCPA, dicalcium phosphate anhydrous or monetite • DCPD, dicalcium phosphate dihydrate or brushite • HAp, hydroxylapatite • OCP, octacalcium phosphate • Pi, inorganic phosphorus • Po, organic phosphorus • PyroP, pyrophosphate • PWEP, water-extractable phosphorus to total phosphorus percentage • TP, total phosphorus • WEP, water-extractable phosphorus • XANES, X-ray absorption near edge structure • XRD, X-ray powder diffraction


   INTRODUCTION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
Materials and Methods
 Results and Discussion
 Conclusions
 REFERENCES
 
AGRICULTURE is the primary contributor to the pollution of more than 40% of the impaired lakes in the United States (USEPA, 2000). Nitrogen and phosphorus (P) are the major cause of impairment for half of these lakes. As a consequence of intensive fertilizer and manure application practices, P is exported to water bodies from agricultural areas with high runoff or erosion potential (Sharpley et al., 2001). Agricultural P transport contributes to eutrophication and water quality deterioration in lakes. Therefore, significant amount of research has been performed to understand P dynamics in animal manure and the interaction of manure P components with soils.

Sharpley and Moyer (2000) proposed water-extractable P (WEP) as an indicator of P availability from animal manure to runoff and leachate. Subsequent studies have shown that the ratio of WEP to total P (TP) can vary significantly depending on manure type and storage conditions (Brandt et al., 2004; Kleinman et al., 2005). For example, dairy and poultry manure with identical amounts of TP can have different P availability when applied to the same soil. Under this context, P speciation in the manure matrix has garnered greater attention, since it is one of the factors influencing P availability and fate in the environment. Recent studies have shown that inorganic P (Pi) constitutes more than 60% of TP in dairy manure (Dou et al., 2000; Sharpley and Moyer, 2000; Ajiboye et al., 2004). Inorganic and organic P (Po) fractions of manure have been further investigated to elucidate their major constituents.

A combination of enzymatic hydrolysis and solution-state 31P nuclear magnetic resonance (NMR) methods have been used successfully to determine Po species in animal manure and attempts are underway to standardize these methods (Turner, 2004; Turner and Leytem, 2004; Cade-Menun, 2005; He et al., 2007; McGrath et al., 2005; Toor et al., 2005a). Orthophosphate monoesters (mainly phytic acid) and diesters have been identified as the major organic P (Po) species in dairy and beef cattle manure (He and Honeycutt, 2001; He et al., 2004; Turner, 2004). Enzymatic hydrolysis and solution-state 31P NMR do not reveal any information on Pi speciation. Inorganic P is mostly associated with particulate solids larger than 0.45 µm in dairy manure (Güngör and Karthikeyan, 2007). Elemental composition, particularly calcium (Ca) and magnesium (Mg) levels, and slightly alkaline pH conditions support the hypothesis that these particulates are Ca and/or Mg orthophosphate solid phases. Indeed, our recent chemical modeling study, supported by water extraction data, indicated that dicalcium phosphate dihydrate (DCPD, CaHPO4·2H2O), dicalcium phosphate anhydrous (DCPA, CaHPO4), octacalcium phosphate (OCP, Ca4H(PO4)3·3H2O), beta tricalcium phosphate (ß-TCP, ß-Ca3(PO4)2), and struvite (MgNH4PO4·6H2O) were the probable phases in dairy manure (Güngör and Karthikeyan, 2005). Thermodynamic probability of a certain phase to exist does not necessarily guarantee its presence or predominance in manure. Therefore, microscopic and spectroscopic techniques have been employed to directly identify the Pi solid phases. Struvite has been identified in pig, cattle, sheep manure, and poultry litter using qualitative scanning electron microscopy/energy dispersive X-ray spectroscopy (SEM/EDX), X-ray powder diffraction (XRD), and solid-state 31P NMR techniques (Fordham and Schwertmann, 1977; Bril and Salomons, 1990; Shand et al., 2005; Toor et al., 2006). Shand et al. (2005) showed that DCPD and struvite accounted for 63% of TP in the sheep feces using quantitative XRD.

X-ray absorption near-edge structure (XANES) spectroscopy is a non-invasive element specific technique that is being increasingly applied to investigate P speciation in animal manures. In this technique, synchrotron radiation with varying energy is applied on a sample and its electrons are excited to create an absorbance spectrum. Energy region of XANES is determined according to the "absorption edge" or "whiteline" of the selected element. Absorption edge refers to the energy at which core electrons of the element are ejected. For example, P K-edge XANES uses an absorption edge around 2150.0 eV, which corresponds to the energy necessary to eject electrons from the K shell (1s orbital) of P atoms (Toor et al., 2005b). X-ray absorption near-edge structure spectra include information on oxidation state, coordination number, and local chemical environment of the element. Recently, XANES has been used to investigate the solid phase speciation in environmental samples. However, limited accessibility of synchrotron beamlines prevents XANES from becoming a routinely used method. Therefore, there are limited XANES studies addressing P speciation in animal manure (Peak et al., 2002; Sato et al., 2005; Toor et al., 2005b; Shober et al., 2006; Maguire et al., 2006). These studies have used a "fingerprinting" approach for data analysis: for example, K-edge XANES spectra of the known standards have been compared with the unknown sample spectrum to investigate P speciation. A detailed review of application of NMR (both liquid- and solid-state) and XANES to elucidate P speciation in animal manures can be found elsewhere (Toor et al., 2006).

Three main groups of standards have been used successfully in the P K-edge XANES analysis of animal manure: Ca phosphates, phytic acid, and weakly bound/sorbed orthophosphates. Calcium phosphates and phytic acid were selected to represent Pi solid phases and Po, respectively. Peak et al. (2002) proposed aqueous orthophosphate (aq-PO4) as a standard to account for free and weakly bound orthophosphates in "as is" (i.e., not subjected to any drying) poultry litter. Aqueous-PO4 was also used to represent adsorbed Pi in dried manure (Sato et al., 2005; Toor et al., 2005b). Recently, orthophosphate sorbed on Fe- and Al-oxides has been used as a standard to represent Pi sorbed on these minerals in the XANES analysis of dairy and layer manure (Maguire et al., 2006; Shober et al., 2006). After comparing the given set of standards with the sample spectra, the previous XANES studies concluded Ca orthophosphate solids and weakly bound/sorbed orthophosphates constituted most of the TP in chicken, dairy, and turkey manure.

The primary objective of this study was to directly evaluate P speciation in dairy manure using XANES and XRD methods, with an equal emphasis on both Ca and Mg phosphates. To our knowledge, there is no other XANES study focusing on the presence of Mg phosphates (i.e., newberyite and struvite) in animal manure.


   Materials and Methods
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and Methods
Results and Discussion
 Conclusions
 REFERENCES
 
Sample Collection and Characterization
Influent and effluent samples from an on-farm anaerobic digester were collected as part of a larger-scale study initiated in July 2005 (Güngör and Karthikeyan, 2007). Dairy manure was collected by a scrape system and fed into the digester, which was designed for 1200 heads and 20 d hydraulic retention time. The digester was plug-flow and operated under mesophilic temperature (Kramer, 2004). One-liter plastic bottles were used to collect influent and effluent samples from the inlet and outlet pits of the digester, respectively. Total solids (TS) content was determined by pre-weighing the subsamples (directly transferred from 1-L containers) and removing moisture in an oven at 105°C for 24 to 48 h (Peters et al., 2003). Total P concentrations of the wet samples were measured using acid persulfate digestion followed by the molybdate colorimetric method (QuikChem 10-115-01-4-E) with a flow injection analyzer (FIA) (Lachat QuikChem 8000; Hach, Loveland, CO).

Influent and effluent TP were 0.77 and 0.84%, respectively, on a dry weight basis. The samples were separated into different size fraction using a sieve set of 1000, 500, 250, 106, 53, and 25 µm on an analytical sieve shaker (AS200; Retsch, Newtown, PA). Subsamples retained on the 25 µm sieve were dried below 45°C in an incubator to avoid possible phase changes in the Pi solid phases (e.g., struvite) (Sarkar, 1991). The total P content of the samples was measured using the FIA following nitric-sulfuric acid digestion (Eaton et al., 2005). Phosphorus concentrations were 1.17 and 1.24% in the 25 to 53 µm size fraction of the influent and effluent samples, respectively. Phosphorus concentrations in the 25 to 53 µm size fraction were approximately 50% greater than those of the bulk samples. The subsamples were preferred over bulk samples to improve the signal-to-noise ratio generated during spectroscopic analysis. Signal-to-noise ratio is a power ratio between signal (meaningful information) and background noise and stronger signals can be expected from samples with higher P concentration. After grinding with a mortar and pestle, the dried subsamples, addressed as samples hereafter, were saved for detailed elemental characterization and analysis by XANES spectroscopy and XRD. Total content of K, Ca, and Mg was determined in the concentrated nitric acid digestates of the samples using inductively coupled plasma–mass spectrometry at the University of Wisconsin (UW) Soils and Plant Analysis Laboratory (Table 1 ). To determine WEP, pre-weighed samples were placed in 50-mL centrifuge tubes and extracted with 30 mL deionized water at 200:1 water/manure ratio for 16 h using a horizontal shaker (Kleinman et al., 2002). The extracts were centrifuged at 18,000 rpm for 10 min. The centrifugates were passed through 0.45-µm filters and saved for further analysis. The filtrates were analyzed for Pi using FIA and the resultant concentrations were used to calculate the fraction of TP available in a water-extractable form (PWEP) (Table 1).


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

  		Table 1. Characterization of influent and effluent (anaerobically-digested) manure samples used in this study.

 
X-ray Powder Diffraction Analysis
X-ray diffractograms (Cu K{alpha} radiation) of the influent and effluent were collected using a Scintag PAD-V X-ray diffractometer at the Department of Geology, UW-Madison. The samples were mounted on a silica wafer using commercial finger nail polish. Before sample analysis, the background (polish+wafer) diffractogram was checked for possible interferences and verified for absence of crystalline phases. Sample diffractograms were collected between 5 and 70° 2{theta} with a step size and dwell time of 0.015° and 15 s, respectively, to improve the signal-to-noise ratio (Shand et al., 2005). Each sample scan lasted approximately 18 h. Phase identification was performed using XRD data analysis software and its powder diffraction file (PDF) database (Jade (v6.5); Materials Data Inc., Livermore, CA).

X-ray Absorption Near-Edge Structure Analysis
Phosphorus Standards
Certified DCPD, DCPA, OCP, and hydroxylapatite (HAp) were obtained from Clarkson Chromatography Products Inc. (South Williamsport, PA). Certified amorphous Ca phosphate (ACP) and ß-TCP (unsintered) were purchased from HiMed Bioactive Materials Resource Center (Old Bethpage, NY). A solution comprising reagent-grade NH4Cl, Na2HPO4·7H2O, and MgCl2·6H2O was mixed and its pH was increased to values above 8.0 to synthesize struvite. After XRD phase confirmation (PDF#15-0762), the precipitate was used as the struvite standard. Corn derived dodecasodium salt (C6H6O24P6Na12), Ca pyrophosphate (Ca2P2O7), and Mg phosphate dibasic (MgHPO4·3H2O) were used as phytic acid, pyrophosphate (PyroP), and newberyite standards, respectively (Sigma-Aldrich Co., St. Louis, MO). An aqueous phosphate standard solution (1.0 mol L–1) was prepared using reagent grade Na2HPO4·7H2O in deionized water for investigating adsorbed orthophosphate in the manure samples (Sato et al., 2005; Toor et al., 2005b). The pH was maintained at 7.2, wherein the major aqueous orthophosphate species are H2PO4– and HPO42–, during XANES analysis.

Data Collection and Pre-treatment
Phosphorus K-edge XANES spectra of the dried manure samples were collected at the Double Crystal Monochromator beamline of the Canadian Synchrotron Research Facility using the UW Synchrotron Research Center storage ring in January 2006. Toor et al. (2005b) also used this beamline for the XANES analysis of turkey manure and broiler litter. A thin layer of sample was mounted on graphite tape and placed in the vacuum chamber. Since the penetration depth is only a few microns in P K-edge XANES, data collection in transmission mode is not possible (Toor et al., 2006). Therefore, data acquisition was performed using three regions in the fluorescence mode: (1) 2125–2144 eV, step size = 1.00 eV, (2) 2145–2185 eV, step size = 0.25 eV, (3) 2186–2200 eV, step size = 1.00 eV with a constant dwell time of 3.00 s/point. The XANES spectra were obtained by measurement of the total electron yield (TEY) as a function of incident photon energy. Total electron yield data is typically free of self-absorption effect that may affect fluorescence yield (FY) data by distorting XANES spectrum around the whiteline (Toor et al., 2006). Each sample was scanned five consecutive times. The same procedure was followed for data acquisition for the powder P standards. However, the gain factor was set ten times lower for the P standards as compared to that of the manure samples. A lower gain factor for P standards was selected to enable the detector to capture the whole whiteline feature. Spectra of aq-PO4 were collected in the fluorescence mode using a liquid specimen holder (Jürgensen, unpublished data, 2006) with a 6 µm Lexan window. Since the TEY signal originates from the outside window surface instead of from the solution, FY was used for constructing the XANES spectra of aq-PO4. The rest of the aq-PO4 data acquisition procedure was identical to that used for the samples and powder standards.

The following data pre-treatment routine was performed on both standards and samples using the Athena software (v0.8.049) (Ravel and Newville, 2005). Three to five scans of each sample and standard were averaged using the merge function. The TEY scans that increased the average relative standard deviation (RSD) of the sample and standard spectra were excluded from the averaging process. Here, RSD means percent ratio of standard deviation to arithmetic mean of the absorbances observed at a given photon energy. Average RSD is the arithmetic mean of all RSD values determined for photon energies between 2136.0 and 2198.0 eV. Average RSDs were less than 0.6 and 2.0% for the samples and the standards, respectively, excluding newberyite and aq-PO4. Newberyite and aq-PO4 had average RSDs of 5.9 and 3.5%, respectively. The pre-edge region of the spectra was set between –14.0 and –4.0 eV relative to the reference energy, Eo (2149.5–2150.0 eV) for background removal. Then, the spectra were normalized using the post-edge energy range of +31 and +48 eV relative to Eo. It is important to use a consistent normalization procedure for both the standards and the samples (Shober and Hesterberg, personal communication, 2006), since an inconsistent approach can change the results of quantitative analysis.

Contribution of the P standards to the manure spectra was investigated using the linear combination fitting (LCF) function of Athena. The fit range was between –9.0 and +41.0 eV relative to Eo and included 170 data points. An energy shift was allowed for the standard spectra, since energy calibration was not performed during data pre-treatment. The weight fraction of each reference was limited between 0.0 and 1.0, and the sum of the weight fractions was forced to be 1.0. Binary and ternary combinations of eleven normalized P standard spectra were fitted to the normalized spectrum of each sample. Goodness-of-fit was assessed both visually and quantitatively. Residual (Eq. [1]), {chi}2 (Eq. [2]), and R (Eq. [3]) values were determined for quantitative assessment:

Formula 1[1]

Formula 2[2]

Formula 3[3]
where ymeasured(i) = ith normalized absorbance data point measured for the sample, yfitted(i) = ith normalized absorbance data point fitted for the sample, {sigma} = uncertainty (or experimental error) estimate, and N = number of data points (Newville, 1998; Ressler, 2004). The reported {chi}2 and R values were obtained from the Athena software outputs, and the residual (Eq. [1]) was calculated using the fitted and measured absorbance values.


   Results and Discussion
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and Methods
 Results and Discussion
Conclusions
 REFERENCES
 
X-ray Powder Diffractograms of Manure Samples
The highest intensity peak (2{theta} = 28.43°) in the diffractogram of influent sample belonged to a KCl mineral, namely sylvite (PDF#73-0380) (Fig. 1 ). Sylvite was also identified in the effluent (Fig. 1). This phase identification was in agreement with the elemental composition of the influent and effluent samples (Table 1). Elevated K+ concentration in manure slurries combined with the high solubility of both K+ and Cl– strongly support the possibility that sylvite must be an artifact of the sample drying process.


Figure 1
View larger version (12K):
[in this window]
[in a new window]

  		Fig. 1. X-ray powder diffractograms of the influent and effluent manure samples (S: struvite (PDF#15-0762), Syl: sylvite (PDF#73-0380)). The influent diffractogram is offset by 10,000 units.

 
Importantly, struvite (PDF#15-0762) was identified in both the influent and effluent samples. Seven struvite peaks with the highest relative intensities (100 to 40%) were assigned to the peaks in the diffractograms of the influent and effluent samples (Fig. 1). Unlike K and Cl, P is mainly associated with particulate solids (i.e., solids larger than 0.45 µm) in the dairy manure slurry (Güngör and Karthikeyan, 2007). Therefore, presence of struvite in the influent and effluent cannot be attributed as an artifact of the sample drying process. As mentioned earlier, struvite was previously identified in cattle, pig, sheep manure, and poultry litter samples (Fordham and Schwertmann, 1977; Bril and Salomons, 1990; Shand et al., 2005; Toor et al., 2006). Shand et al. (2005) also showed dissolution of struvite had a significant influence on the WEP levels in sheep feces.

A few noticeable differences among the samples were the two sharp peaks observed at 25.72 and 36.59° 2{theta} in the effluent sample, and a peak observed at 68.98° 2{theta} in the influent (Fig. 1). These peaks could not be assigned to any P solid phases. No P solid phases other than struvite could be determined in the samples due to the complexity of the manure matrix.

Qualitative XANES Analysis of Phosphorus Standards and Manure Samples
The Ca orthophosphate standards had two features in the XANES region (Fig. 2 ) in addition to the whiteline (Fig. 2a): a post-edge shoulder around 2152.0 eV (Fig. 2b) and a peak around 2160.0 eV (Fig. 2c). The post-edge shoulder was not as noticeable in the DCPA spectrum as compared to the other Ca phosphate spectra. As the stability of a Ca phosphate phase increases and its solubility decreases, the spectral features may appear more distinctly (Sato et al., 2005). Spectra of the Mg phosphate phases, namely newberyite and struvite, lacked the features observed in Ca phosphates and they had a broad feature around 2160.0 eV (Fig. 3b ), more pronounced in struvite as compared to newberyite. The pyroP standard also has a broad feature around 2160.0 eV (Fig. 3b). The shape of this feature was different from that present in struvite and newberyite, and the pyroP standard also had a shoulder-like feature around 2152.0 eV. The aq-PO4 and phytic acid standards did not have significant spectral features; however, the whiteline of phytic acid salt was slightly broader as compared to that of aq-PO4 (Peak et al., 2002).


Figure 2
View larger version (17K):
[in this window]
[in a new window]

  		Fig. 2. Normalized X-ray absorption spectra of the Ca orthophosphate standards (ACP: amorphous calcium phosphate; ß-TCP: beta tricalcium phosphate; DCPA: dicalcium phosphate anhydrous; DCPD: dicalcium phosphate dihydrate; HAp: hydroxylapatite; OCP: octacalcium phosphate) (a: whiteline, b: shoulder, c: distinct peak, d: oxygen oscillation).

 

Figure 3
View larger version (16K):
[in this window]
[in a new window]

  		Fig. 3. Normalized X-ray absorption spectra of the phosphorus standards (Aq-PO4: aqueous orthophosphate; phytic: phytic acid; pyroP: pyrophosphate) (a: whiteline, b: spectral feature, c: oxygen oscillation).

 
Oxidation state of P in animal manure is +5 (Peak et al., 2002; Sato et al., 2005; Toor et al., 2005b; Maguire et al., 2006; Shober et al., 2006). Consequently, the XANES spectra of the manure samples do not include distinct features imposed by a mixture of P species with different oxidation states (Fig. 4 and 5 ). The whiteline or absorption edge and the oxygen oscillation were observed around 2150.0 and 2166.0 eV, respectively, and the region between the whiteline and the oxygen oscillation contains information on the local environment of P and its speciation. Since the manure P K-edge XANES spectra contained contributions from various P species, it was difficult to determine the presence of a particular P phase only by visual inspection. On the other hand, absence of the features distinct for Ca phosphates (Fig. 2) in the manure spectra (Fig. 4 and 5) implied that these phases, excluding DCPA, were not likely to be the predominant P species in manure slurries.


Figure 4
View larger version (12K):
[in this window]
[in a new window]

  		Fig. 4. Measured (solid line) and fitted (dashed lines) X-ray absorption spectra of the influent sample. The DCPA+struvite combination yielded the best fit (DCPA: Dicalcium phosphate anhydrous; aq-PO4: aqueous orthophosphate).

 

Figure 5
View larger version (12K):
[in this window]
[in a new window]

  		Fig. 5. Measured (solid line) and fitted (dashed lines) X-ray absorption spectra of the effluent sample (after anaerobic digestion). The struvite+HAp combination yielded the best fit (aq-PO4: aqueous orthophosphate; HAp: Hydroxylapatite).

 
Least Squares Linear Combination Fitting of the XANES Data
Binary and ternary combinations of eleven standard spectra were fitted to the influent and effluent manure spectra using the Athena software. Since struvite was determined in the samples using XRD (Fig. 1), this phase was selected as a required standard for LCF. Therefore, the total number of (binary and ternary) standard combinations fitted to the sample spectrum was 55.

Combinations of struvite with aq-PO4 or newberyite yielded low {chi}2 values, ranging between 1.27 and 2.34. However, all of these combinations were plagued with two serious issues: (i) the energy shift of struvite was approximately 30.0 eV, and (ii) the percentage of TP in the form of struvite in the manure sample was only between 2 and 8%. The energy shift of struvite was unacceptably high and the fraction of TP as struvite was also in disagreement with the XRD result (Fig. 1), which indicated that the struvite concentration should be significant in the influent. Also, struvite (PDF#15-0762) was identified in both the influent/effluent samples (Fig. 1) whereas the corresponding diffractograms clearly lacked the second highest intensity peak (18.82° 2{theta}) for newberyite (PDF#72-0023). Therefore, these combinations were regarded as unlikely for our manure samples. Struvite+DCPA combination had the next lowest {chi}2, 2.40, and a higher fraction of TP as struvite, 43%, as compared to the previous combinations (Table 2 ). This struvite fractional composition appeared to be more consistent and in better agreement with the XRD result. Hence, this combination was selected to be the most plausible for the influent sample. In this fitting exercise, the combinations with high aq-PO4 or newberyite weight fractions yielded better {chi}2 values than the one with high ratio of TP as struvite (i.e., 43%). As shown in Fig. 3, there were only subtle differences between the spectral features of struvite, newberyite, and aq-PO4. If struvite had not been included in the LCF analysis, aq-PO4 or newberyite could have been reported as a major P species in the influent. Binary combinations of DCPA with aq-PO4 and newberyite were performed to investigate this probability. Both these combinations produced lower {chi}2 as compared to the DCPA+struvite combination (Table 2). In other words, both aq-PO4 and newberyite could replace struvite without substantially affecting the {chi}2 values. Fitted combinations (three binary combinations) of DCPA with (i) struvite, (ii) newberyite, and (iii) aq-PO4 and measured (influent) spectra are shown in Fig. 4. All combinations captured the post-edge feature observed around 2160.0 eV in the influent spectrum.


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

  		Table 2. Results from linear combination fitting (LCF) analysis of influent and effluent manure samples.

 
Another Ca and Mg orthophosphate binary combination provided the best fit to the effluent spectrum. This combination consisted of struvite+HAp with a {chi}2 of 2.24. The fraction of TP existing as struvite was 78.2%; thus, it was the dominant P species in the effluent (Table 2). As performed for the influent sample, struvite was replaced by aq-PO4 and newberyite in binary combinations with HAp. The results were quite similar to those observed for the influent spectrum. Binary combinations of HAp with aq-PO4 and newberyite had lower {chi}2 values as compared to the HAp+struvite combination (Table 2). Combinations of HAp with struvite and newberyite captured the 2166.0 eV spectral feature of the effluent spectrum quite well, whereas the HAp+aq-PO4 combination missed this feature due to the extremely high fraction of TP existing as aq-PO4 (i.e., 91.1%) (Fig. 5).

The LCF results show that manure P speciation shifted toward Mg (i.e., struvite) and the least soluble Ca (i.e., HAp) orthophosphate phases following anaerobic digestion. The fraction of TP present in the form of struvite increased from 43.0 to 78.2% and HAp, which did not appear as a significant component in the influent, constituted 21.8% of the TP in the effluent. The decrease in PWEP (Table 1) after anaerobic digestion also supported these LCF results. Goodness-of-fit indicator of the plausible binary combinations (influent: DCPA+struvite; effluent: struvite+HAp) was not as low as the {chi}2 and residual values published previously (Table 2). For instance, our {chi}2 values are significantly higher than the 0.10 to 0.30 range reported in a previous study on dairy manure (Shober et al., 2006). Similarly, residuals obtained in this study are higher as compared to the 0.8 to 1.0% range for the LCF analysis performed on broiler litter and turkey manure spectra (Toor et al., 2005b). It must be noted that ternary combinations were used in both of these studies. In our study, ternary combinations did not improve the goodness-of-fit (data not shown). A third minor component, probably a Po phase, could also be present in the manure samples. Absorption spectrum of this probable Po phase might be significantly different from those of the eleven P standards used in the LCF analysis. Exclusion of this minor component could be one of the factors that increased residuals or {chi}2 values. On the other hand, our {chi}2 and R-values (Table 2) were comparable with those of 1.82 and 0.008, respectively, reported for the ternary fits of two chicken manure spectra (Sato et al., 2005; Maguire et al., 2006).

As mentioned above, Mg orthophosphate phases were not considered in the previous XANES studies investigating animal manure. We found that struvite could be an important Pi solid phase in dairy manure. Our previous chemical modeling work (Güngör and Karthikeyan, 2005) and Mg/P molar ratio of the manure samples (Table 1) are also in agreement with the spectroscopic data included above. Identification of struvite by XRD has been very helpful in the selection of realistic linear combinations and quantification of P species during XANES analysis. Our results highlight the importance of using: (i) a complete standard set for XANES analysis and (ii) other complementary spectroscopic methods in conjunction with XANES.

Another noticeable difference between our study and the previous XANES work on dairy manure is in the Po concentration. Shober et al. (2006) determined Po concentrations in their samples to be between 11.2 and 34.1% of TP using a Na phytate standard. Although we used the same standard to represent Po, the LCF analysis did not yield phytate as a significant species. This difference may be due to: (i) intrinsic variability in the nature of Po species among the manure samples, (ii) insensitivity of XANES spectroscopy in discriminating Po from other Pi species, and (iii) Po mineralization during the sample preparation and/or storage. Drying has been shown to increase orthophosphate levels by up to 8.4% of TP in dairy manure through Po mineralization (He et al., 2007).

Environmental Implications of Manure Phosphorus Speciation
Our results highlight the fact that dissolution of DCPA and struvite could be an important mechanism involved in orthophosphate release to runoff from soils receiving dairy manure. A similar P release mechanism involving DCPD and struvite dissolution has been proposed for pastures impacted by sheep feces (Shand et al., 2005). These P solid phases have been reported to dissolve within weeks following land application with up to 50% solubilization during the first week of land application (Shand et al., 2005). Therefore, DCPA and struvite introduced via dairy manure addition are not likely to persist in agricultural soils.

Orthophosphates released from the dissolution of these Ca and Mg phosphates may sorb onto Al- and Fe-oxides and/or re-precipitate as more stable Ca-P solid phases over the long term (Beauchemin et al., 2003). Relatively insoluble Ca-P solid phases such as OCP, TCP, and HAp have been directly identified in both acidic and slightly alkaline soils receiving long-term manure inputs (Beauchemin et al., 2003; Sato et al., 2005). As an exception, some sandy agricultural soils may not favor the predominance of stable Ca-P solid phases (Harris et al., 1994). Our results also indicate that struvite and HAp are the major P species in anaerobically digested dairy manure. Soils amended with anaerobically digested manure may initially release less orthophosphate to runoff due to the presence of HAp. Anaerobic digestion appears to decrease PWEP values via transformation of more soluble P solid phases into less soluble ones. A complete understanding of the underlying mechanisms involved in this chemical phase transformation would facilitate manipulation of the design and operational conditions of anaerobic digesters to create favorable conditions for the formation of inorganic P phases with low solubility.

Probable Improvements in the XANES and XRD Methods
Greater attention has been devoted to investigate the P K-edge XANES characteristic features of Ca orthophosphate solid phase compared to Mg orthophosphates (Franke and Hormes, 1995; Hesterberg et al., 1999). Identification of the diagnostic spectral features for Mg orthophosphates would result in better characterization of Pi speciation in animal manure and other environmental samples. Furthermore, quantitative P speciation in animal manure may be performed more rigorously by simultaneously using Ca, Mg, and P K-edge XANES analyses. Peak et al. (2002) followed a similar approach and performed qualitative Ca, P, and S K-edge XANES on poultry litter. The P concentration in animal manure (<1% on a dry basis) is generally below the detection limit of XANES for in situ analysis. Liquid manure samples are, therefore, dried before XANES analysis and their spectra are collected using the fluorescence mode. Drying is likely to affect manure P speciation and, hence, PWEP values (Chapuis-Lardy et al., 2003, 2004; Ajiboye et al., 2004; He et al., 2007). Future advancements in X-ray absorption spectroscopy, especially sensitivity improvements in the transmission mode, may enable direct investigation of P speciation in liquid manure or similar environmental samples.

Although XRD instrumentation is more accessible compared to XANES, longer scan times (i.e., more than half a day) are required due to low P content of animal manure. Also, relatively low peak intensities in the diffractograms hinder identification of the crystalline P solid phases. To overcome these limitations, methods to improve the signal-to-noise ratio in the XRD diffractograms of animal manures are required.


   Conclusions
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and Methods
 Results and Discussion
 Conclusions
REFERENCES
 
Struvite was unambiguously identified in both the raw and anaerobically digested dairy manure samples using XRD. Linear combination fitting of the XANES spectra showed that 57.0 and 43.0% of TP was associated with DCPA and struvite, respectively, in the raw manure. In the anaerobically digested sample, 78.2% of TP was present as struvite and 21.8% of P was associated with HAp. Importantly, P speciation shifted toward Mg and the least soluble Ca orthophosphates following anaerobic digestion. The LCF analysis also showed that the similarity of the aq-PO4, newberyite, and struvite spectra can cause inaccurate determinations of P speciation when manure samples with a significant amount of Mg orthophosphates are analyzed solely by XANES spectroscopy. X-ray absorption near edge structure analysis can be used to quantify P speciation in dairy manure in conjunction with XRD. Improvement of the XANES and XRD methods can be helpful to investigate the effect of different treatment alternatives on P speciation in animal manure and mechanisms influencing P release from soils receiving manure.


   ACKNOWLEDGMENTS
 
The authors express their sincere gratitude to Dr. Huifang Xu (Univ. of Wisconsin-Madison), Dr. Amy L. Shober (Univ. of Florida), Dr. Derek Peak (Univ. of Saskatchewan), Dr. Dean L. Hesterberg (North Carolina State Univ.), Dr. Shinjiro Sato (Univ. of Florida), Dr. Gurpal Toor (Univ. of Arkansas), and Dr. William F. Bleam (Univ. of Wisconsin-Madison) for their contributions. The authors also thank the Canadian Synchrotron Research Facility for granting access to their beamline. This research was partially supported by the USDA-CSREES (Wisconsin) Hatch Program, Project No. WIS04655. This work is based on research conducted at the Synchrotron Radiation Center, Univ. of Wisconsin-Madison, which is supported by the NSF under Award No. DMR-0537588.


   NOTES
TOP
 NOTES
ABSTRACT
 INTRODUCTION
 Materials and Methods
 Results and Discussion
 Conclusions
 REFERENCES
 
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.


   REFERENCES
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and Methods
 Results and Discussion
 Conclusions
 REFERENCES
 




This article has been cited by other articles:


Home page
 J. Environ. Qual.Home page
Z. He, C. W. Honeycutt, T. S. Griffin, B. J. Cade-Menun, P. J. Pellechia, and Z. Dou
Phosphorus Forms in Conventional and Organic Dairy Manure Identified by Solution and Solid State P-31 NMR Spectroscopy
J. Environ. Qual., July 23, 2009; 38(5): 1909 - 1918.
[Abstract] [Full Text] [PDF]

Home page
 J. Environ. Qual.Home page
S. Hunger, J. T. Sims, and D. L. Sparks
Evidence for Struvite in Poultry Litter: Effect of Storage and Drying
J. Environ. Qual., June 23, 2008; 37(4): 1617 - 1625.
[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 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 Google Scholar
Google Scholar
Right arrow Articles by Güngör, K.
Right arrow Articles by Karthikeyan, K. G.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Güngör, K.
Right arrow Articles by Karthikeyan, K. G.
Agricola
Right arrow Articles by Güngör, K.
Right arrow Articles by Karthikeyan, K. G.
Related Collections
Right arrow Phosphorus
Right arrow Nutrients
Right arrow Animal Waste

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
Soil Science Society of America Journal Journal of Plant Registrations The Plant Genome