SciELO - Scientific Electronic Library Online

Home Pagelista alfabética de revistas  

Journal of the Chilean Chemical Society

versión On-line ISSN 0717-9707

J. Chil. Chem. Soc. v.49 n.1 Concepción mar. 2004 



1Departamento de Química de los Materiales, Facultad de Química y Biología,
Universidad de Santiago de Chile, Santiago, Chile. E-mail:
2Department of Environmental Sciences, University of California, Riverside, CA 92521, USA.

(Received: April 25,2003 - Accepted: July 28, 2003)


The phosphorus distribution in soil (<2 mm) and in clay size separates (<2 m m) of two uncultivated volcanic soils from southern Chile (one Ultisol and one Andisol) was studied by phosphorus-31 nuclear magnetic resonance spectroscopy (31P-NMR). The total P content ranges of the Ultisol samples were 733-1108 mg kg-1 (soil) and 746-896 mg kg-1 (clay), lower than 2163-2216 mg kg-1 (soil) and 2683-3362 mg kg-1 (clay) obtained for Andisol samples. In both fractions the 31P-NMR spectra of a single alkaline extraction, using 0.25 M NaOH and Chelex 100, showed the presence of inorganic orthophosphate, monoester-P, diester-P and pyrophosphate. Phospholipids-teichoic acids P were observed, but only in the clay size separates of the Andisol. Important differences in spectra of soils and clay size separates were observed only in samples with high organic matter content. The clay size separates enriched with organic matter have more and better defined signals of organic P forms. This allows the identification of P forms that might not be detected by 31P-NMR in the whole soil extracts.

KEY WORDS: organic P forms, volcanic soil, particle size, 31P-NMR


In Chile, most agricultural activity occurs on Andisols and Ultisols, which are derived from volcanic ejecta. Andisols are characterized by a high organic matter content and mineralogy dominated by poorly organized aluminosilicates (Allophane). Ultisols have lower amounts of organic matter than Andisols and a mineralogy dominated by more crystalline compounds (1). These soils have a high P-retention capacity and a large proportion of P found in these soils is present as organic P, which is not readily available for plant uptake (2, 3).

Phosphorus-31 nuclear magnetic resonance spectroscopy (31P-NMR) is a simple and direct method which has been used to study the distribution of P chemical forms. Several classes of P-compounds can be identified on the basis of their chemical shifts. Orthophosphate monoesters and orthophosphate diesters, phosphonates, polyphosphates, and pyrophosphates have been reported to be present in soils (4-7).

Different chemical procedures have been proposed to extract P (4); the single alkaline extraction is the simplest way used to identify P-compounds. Also the effect of the extractant concentration in the hydrolysis of the sample during the extraction has been studied (8, 9).

P extractions are usually carried out on the whole soil (<2mm), even though the clay size separate (<2m m) is the most active fraction in ion exchange and adsorption. It has been reported that different P forms are present in particle size separates (9), and specifically in Chilean volcanic soils it has been reported an organic matter and organic P enrichment in the clay size fraction in relation with the soil (3). Thus, the objective of this work is to study the influence of particle size on the distribution of P chemical forms, studied by 31P-NMR, in two Chilean volcanic soils.


Soils collection

Samples of the 0-15cm and 15-30cm depth of two Chilean soils from uncultivated plots were used. Collipulli (geographical coordinates, 36º 58' S 72º 09' W, Fine, mesic Xeric Palehumult), and Diguillin (36º 53' S 72º 10' W, Medial, thermic Typic Dystrandept) were selected.

Sample preparation

Soil samples were screened in the field to pass a 2-mm sieve and stored at field moisture content. The <2m m particle size samples (clay size separates) were obtained by a sedimentation procedure based on Stoke`s law. The soil samples were suspended only in distilled water and sonicated with an ultrasonic vibrator. After 24 h, suspensions above 20 cm depth were siphoned. These samples were freeze dried for 48 h.

Samples characterization

Samples were characterized for: organic carbon content by the Walkley-Black method (10); pH using soil suspensions in double distilled water at 1:2.5 w/v ratio; and mineralogy by X-ray diffraction.

A full characterization of soil samples and methods description can be found in Escudey et al. (3).

Chemical P fractionation

Total phosphorus (Total P) was determined by the ascorbic acid-molybdenum blue method (11), after alkaline oxidation with sodium hypobromite (NaBrO). The distribution of organic and inorganic P was determined by the sequential extraction procedure of Steward and Oades (12), modified by Borie and Zunino (2). Briefly, 1.5 g of soil or 0.1 g of clay fraction was equilibrated with 15 mL of 1 M HCl, then the suspension was centrifuged and filtered to obtain the acid extractable P fraction. Inorganic P of the fraction was determined directly in this extract. Total P of the acid extract was determined after organic matter destruction with NaBrO (11). The organic P of the acid extract was estimated by difference between the total and inorganic P contents. The remaining solids, after the acid extraction, were ultrasonically dispersed and extracted three successive times with 0.5 M NaOH. In the first extraction 25 mL were used, followed by a second extraction with 15 mL, and finally with 15 mL of NaOH. The distribution of the inorganic and organic P of the combined alkaline extract was determined in the same manner as for the acid extract. Total inorganic P corresponds to the sum of inorganic P in acid plus alkaline extracts. Total organic P corresponds to the sum of fulvic and humic P. Total extracted P (designated as total P) corresponds to the sum of total inorganic and total organic extracted P (3).

31P-NMR spectroscopy

Five grams of sample (soil or clay) were mixed with 30 g of cation exchange resin (Chelex 100, Na form, Bio-rad 142-2832) and 100 mL 0.25 M NaOH. The samples were shaken on a reciprocal shaker overnight, centrifuged, and filtered through 0.45 m m pore size. The final extract was freeze dried and redissolved in 3.0 mL of D2O, shaken for 2 hs, centrifuged and transferred to NMR tubes. The 31P-NMR spectra for soil extracts were obtained at 202.5 MHz in a GE 500 MHz spectrometer, using a 45º pulse with a 3s delay and acquisition time of 0.506 s. The 31P-NMR spectra were proton decoupled using the standard waltz decoupling scheme. The decoupler was gated on during the acquisition time to avoid nuclear Overhauser enhancement. Approximately 7,000 scans were accumulated, and the chemical shifts were measured relative to external orthophosphoric acid (85%).


Mineralogy and characteristics of soils

The Collipulli soil is an Ultisol with a mineralogy dominated by kaolinite with traces of halloysite and vermiculite; Diguillin has the characteristic mineralogy of Andisols, dominated by allophane with traces of halloysite and vermiculite.

The OC and total P contents are given in Table 1. The 0-15 cm samples were similar or higher in OC than those of the 15-30 cm, with the exception of the Collipulli soil. In general, the OC content of clay size fraction is higher than and related to the OC of the corresponding soil; thus, when 4 Ultisols and 13 Andisols were considered, the direct relationship OCclays=2.53 OCsoils-2.64 (r=0.899, p<0.01) was obtained (OC expressed as wt. %). In this relationship the lower values correspond to the Ultisol samples.

Table 1.- Organic carbon content (OC), pH (1:2.5 soil:water ratio), inorganic P (Pi), organic P (Po), and total P content of soils and clay size separates.


0-15 cm
soil layer





(mg kg-1)

(mg kg-1)

Total P
(mg kg-1)



P i
(mg kg-1)

(mg kg-1)

Total P
(mg kg-1)


























15-30 cm
soil layer

























The P content of the 0-15 cm samples was higher than in the 15-30 cm samples. The inorganic P/organic P ratio decreases in depth in soils and clay size separates, probably due to the high inorganic P fixation in the upper horizon. The organic P of soils and clay size separates (Figure 1) was correlated with the OC content (r2=0.912, p<0.01), where Ultisol samples (clay and soil) present again the lower values. The forms and content of P are related more strongly to the organic matter content of samples than to particle size. Thus, because clay size separates have a higher OC content than soils, their P content and distribution seems to be modified relative to soils.

Fig. 1 .- Relationship between total P (mg kg-1) and organic matter content (OC, wt%) of soils and clay size separates.

31P-NMR spectroscopy

The NaOH extraction solubilizes organic matter, cations and P forms (inorganic and organic P) associated with the mineral fraction and the soil organic matter (13-15). The chelating resin helps to remove paramagnetic ions from solution, which interfere with the recording NMR spectra (8, 16, 17). The 31P-NMR spectra of the 0.25 M NaOH extract from soils and clay size separates can be observed in Figures 2 and 3. The spectra generally exhibited well signal-to-noise ratios. The assignment of peaks in the NMR spectra was made by measurements of chemical shift relative to orthophosphoric acid, and on the basis of previous reports (4, 6, 7, 14, 15, 18, 19).

The 31P-NMR spectra of soil and clay size separates of Collipulli (low OC content) at 0-15 cm and 15-30 cm depth present only slight differences in some signal intensities (Figure 2a and 2b). The signals in 31P-NMR spectra of clay size samples were well resolved and may be assigned to inorganic orthophosphate (at 7.0 ppm), monoester P (two signals in the range 6.5 - 5.0 ppm), and pyrophosphate (at -3.5 ppm). The pyrophosphate signal is better defined in the 15-30 cm depth. The 31P-NMR spectra of the 0-15 cm and 15-30 cm have a similar P-distribution pattern, however in the 31P-NMR spectra of clay fraction a better resolved signal associated to diester P (1.0 - 0.5 ppm) can be observed.

Figure 2. 31P-NMR spectra of Collipulli soil (a) and caly saparate (b) samples.

The 31P-NMR spectra of Diguillin (high OC content) have similar pattern between 0-15 cm and 15-30 cm. The signals found in the soil fraction (Figure 3a) correspond to inorganic orthophosphate (at 7.0 ppm) and monoester P (6.5 - 5.0 ppm), a signal related with pyrophosphate (-3.5 ppm) was found only in the 0-15 cm sample. Soils and clay size separates spectra present significant differences. The Figure 3b shows the 31P-NMR spectra of Diguillin clay fractions; the signals found may be assigned to inorganic orthophosphate (7.0 ppm), monoester P (6.5 - 5.0 ppm) and diester P (in the range 2.5 - (-1.0) ppm), these were subdivided into phospholipids-teichoic acids P (2.5 - 1.0 ppm) (7), and into other diester P species ( -1.0 - 1.0 ppm)(6, 7, 20-23). The phospholipids-teichoic acids P and other diester P were more abundant in clay separates than in soils. The pyrophosphate signal was observed only in the 0-15 cm samples.

Figure 3. 31P-NMR spectra of Diguillin soil (a) and caly saparate (b) samples.

Inorganic orthophosphate was present in all samples. It is probably produced by the extraction of orthophosphate associated with iron and aluminum, rather than from the hydrolysis of organic P compounds (15, 24).

A small pyrophosphate signal was observed in all extracts (except in Diguillin soil 15-30 cm) in both size fractions; probably it is present in the soil as inorganic forms or, alternatively, as an ester that is hydrolyzed during the extraction procedure. Signals can not be considered as a result of the action of alkali on orthophosphate in aqueous solution (25).

Organic P of NaOH extracts of clay fraction in both soils was mainly monoester P; similar results were previously reported by Leinweber et al. (9). Multiple peaks were observed in the monoester P region. These could include inositol phosphate, sugar phosphates and mononucleotides (26, 27). However, because the chemical shift of these compounds overlap, they cannot be distinguished from one another using 31P-NMR spectroscopy.

The occurrence of phospholipids-teichoic acids P (2.5 - 1.0 ppm), a diester P form (7), and probably inosine monophosphate (0 - (-1.5) ppm) (18), was observed only in the clay fraction of Diguillin sample. Teichoic acids are acidic polysaccharides attached to the cell wall of Gram-positive bacteria; thus, the high bacterial activity near and at clay surfaces may lead to the accumulation of teichoic acids in the clay-size separates (20, 21). The presence of diesters P, a P-form readily biodegradable in the soil, was associated with high organic matter content, especially in the clay fraction of Andisols. This relationship of particle size, organic matter enrichment and diester P forms was confirmed by the31P-NMR spectra of Diguillin samples (Figure 3a and b). The weak signal of diester P in Collipulli soil, with a mineralogy dominated by more crystalline compounds and with low organic matter content, is also indicative that most of the accumulated P is not readily available for plant absorption. It also confirms that diester P is a more easily mineralized form of organic P in the soil environment than monoester P (15, 18).

In general, no significant differences between horizons were observed in 31P-NMR spectra. The Collipulli and Diguillin soils have different mineralogy, organic matter and P contents, and their 31P-NMR spectra have shown also some differences with respect to particle size. Andisols have a higher organic matter and P content than Ultisols (3); consequently, the clay fraction of the Andisol is enriched in both organic matter and P with respect to the soil, and the clay from the Ultisol has similar OC and P content than the soil. As result of this, while 31P-NMR spectra of Ultisol clay and soil showed similar patterns, important differences were observed in the Andisol. In its clay fraction more and better defined signals were observed; P forms as phospholipids-teichoic acids and some diester P are only observed in this clay fraction.


Soils and clay size separates present a similar 31P-NMR pattern, both show a linear correlation between organic P forms and OC content. More and better defined signals were observed in the clay fraction of the soil with the higher organic matter content (Diguillin); P forms as phospholipids-teichoic acids and some diester P are only observed in this clay fraction.

The clay size fraction may not represent the soil behavior, but allows the identification of P-forms that normally cannot be detected in soil extracts by 31P-NMR.


This study was supported by DICYT-USACH, FONDECYT 2003 Nº1030778, and the University of California, Riverside. M. Briceño acknowledges the financial supports provided by CONICYT to her graduate program.


1. Besoaín, E. Mineralogía de los suelos volcánicos del centro-sur de Chile. In Suelos Volcánicos de Chile; Tosso, J., Ed.; Instituto de Investigaciones Agropecuarias: Santiago, Chile, 1985; 107-302.        [ Links ]

2. F. Borie, H. Zunino. Soil Biol. Biochem., 15 (5), 599 (1983).        [ Links ]

3. M. Escudey, G. Galindo, J.E. Forster, M. Briceño, P. Díaz, A. Chang. Commun. Soil Sci. Plant Anal., 32 (5 & 6), 601 (2001).        [ Links ]

4. R.H. Newman, K.R. Tate. Commun. Soil Sci. Plant Anal., 11 (9), 835 (1980).        [ Links ]

5. Z.R. Hinedi, A.C. Chang, R.W.K Lee. Soil Sci. Soc. Am. J., 52, 1593 (1988).        [ Links ]

6. M.B. Turrión, B. Glaser, D. Solomon, A. Ni, W. Zech. Biol. Fertil. Soils, 31,134 (2000).        [ Links ]

7. M.I. Makarov, L. Haumaier, W. Zech. Biol. Fertil. Soils, 35, 136 (2002).         [ Links ]

8. B.J. Cade-Menun, C.M. Preston. Soil Sci., 161, 770 (1996).        [ Links ]

9. P. Leinweber, L. Haumaier, W. Zech. Biol. Fertil. Soils, 25, 89 (1997).        [ Links ]

10. L.E. Allison, Organic carbon. In Methods of Soil Analysis. Part 2. Agronomy 9; Black, C.A.; Evans, D.D.; White, J.L.; Ensminger, L.E.; Clark, F.E., Eds.; American Society of Agronomy: Madison, WI, 1965; 1367-1396.        [ Links ]

11. W.A. Dick, M.A. Tabatabai. J. Environ. Qual., 6, 82 (1977).        [ Links ]

12.- J.H. Steward, J.M. Oades. J. Soil Sci., 23 (1), 38 (1972).        [ Links ]

13. K.R.Tate, R.H. Newman. Soil Biol. Biochem., 14, 191 (1982).        [ Links ]

14. G.E. Hawkes, D.S. Powlson, E.W. Randall, K.R. Tate. J. Soil Sci., 35, 35 (1984).        [ Links ]

15. M. Escudey, G. Galindo, J.E. Förster, I. Salazar, A.L. Page, A.C. Chang. Commun. Soil Sci. Plant Anal., 28 (9 & 10), 727 (1997).        [ Links ]

16. M.A. Adams, L.T. Byrne. Soil Biol. Biochem., 21 (4), 523 (1989).        [ Links ]

17. L.M. Condron, M.R. Davis, R.H. Newman, I.S. Cornforth. Biol. Fertil. Soils, 21, 37 (1996).        [ Links ]

18. H.K. Pant, P.R. Warman, J. Nowak. Commun. Soil Sci. Plant Anal., 30 (5-6), 757 (1999).        [ Links ]

19. A. Moller, K. Kaiser, W. Amelung, C. Niamskul, S. Udomsri, M. Puthawong, L. Haumaier, W. Zech. Aust. J. Soil Res., 38, 1017 (2000).         [ Links ]

20. M. Sumann, W. Amelung, L. Haumaier, W. Zech. Soil Sci. Soc. Am. J., 62, 1580 (1998).        [ Links ]

21. G.H. Rubaek, G. Guggenberger, W. Zech, B.T. Christensen. Soil Sci. Soc. Am. J., 63, 1123 (1999).        [ Links ]

22. G. Guggenberger, B.T. Christensen, G.H. Rubaek. J. Plant. Nutr. Soil Sci., 163, 151 (2000).        [ Links ]

23. W. Amelung, A. Rodionov, I.S. Urusevskaja, L. Haumaier, W. Zech. Geoderma, 103, 335 (2001).        [ Links ]

24. M.L. Bishop, A.C. Chang, R.W.K. Lee. Soil Sci., 157 (4), 238 (1994).        [ Links ]

25. K.H. Dai, M.B. David, G.F. Vance, A.J. Krzyszowska. Soil Sci. Soc. Am. J. 60, 1943 (1996).        [ Links ]

26. L.M. Condron, E. Frossard, H. Tiessen, R.H. Newman, J.W.B. Steward. J. Soil Sci., 41, 41 (1990).        [ Links ]

27. F. Gil-Sotres, W. Zech, H.G. Alt. Soil Biol. Biochem. 22 (1), 75 (1990).        [ Links ]

Creative Commons License Todo el contenido de esta revista, excepto dónde está identificado, está bajo una Licencia Creative Commons