Fractionations and Speciation Techniques
Bremner (1996), Mulvaney (1996) and Stevenson (1996) described the basic methodologies, including discussions of errors and uncertainties, that remain
the basis for soil Norg studies. Total Norg concentrations are frequently deter- mined using the Kjeldahl method. This involves heat- and catalyst-supported digestion in concentrated H2SO4 (to convert the Norg into NH4+), followed by distillation of the digest with NaOH, and analysis of the distillate NH4+–N by titration. Principal modifications to the method involve digestion with HF to include mineral-fixed-NH4+, and the salicylic acid-thiosulfate or alkaline reduction methods to include NO3−–N and NO2−–N. Although still widely used, the Kjeldahl Norg method has been gradually supplanted by the Dumas (dry combustion) method. In general, this involves high- temperature combustion of the soil in a pure-O2 atmosphere (converting Norg into NOx), followed by reduction of the NOx to N2, and detection/
quantification of the N2 using a thermal conductivity detector. Several types of Dumas systems that employ elemental analyzers (EAs) are commercially available and offer the simplicity of modern automated instruments capable of the simultaneous determination of C, N, S (and/or H and/or O) together with high sample throughput. These systems also offer the benefit of elimi- nating the need for sample pretreatment and avoiding the use of hazard- ous reagents (Craft et al., 1991; Pérez et al., 2001; Rutherford et al., 2008;
Yeomans and Bremner, 1991). However, dry combustion methods deter- mine total soil N (Nt), with Norg determined by subtraction, i.e. Norg = Nt − (NH4+ + NO3− + NO2−). This requires a separate determination of the exchangeable and soluble NH4+, NO2−, and NO3−, which generally account for <2% of Nt in surface soils. If required (e.g. after mineral fertilizer application), N-ions in solution and those easily desorbed can be separated using mild extractions (e.g. with K2SO4 or CaCl2 solution) and quantified using ion chromatography. The determination of nonexchangeable, or fixed, NH4+ is more time-consuming—requiring extraction of the sample with KOBr to remove exchangeable NH4+ and labile organic N compounds, followed by digestion with HF–HCl solution (Mulvaney, 1996). Though time-consuming, these steps are necessary to avoid the significant errors that can arise when Nt is used as a proxy for Norg, for example, when using the calculated C:Nt ratio (as opposed to the C:Norg ratio) as an indicator of SOM quality. Indeed, data compiled by Stevenson (1986) show that the presence of fixed-NH4+ causes C:Nt ratios in surface and subsurface soils to deviate systematically (6–37%, respectively) from the corresponding C:Norg ratios. Not surprisingly, the effects of fixed-NH4+ can be pronounced in investigations of Norg in different particle-size fractions of SOM (Olk, 2008;
Schulten and Leinweber, 2000). As seen in Fig. 2.1, Nt values determined by dry combustion strongly reflect the contribution of fixed-NH4+, whereas
the corresponding total-Kjeldahl N (Norg) values were essentially unaf- fected by the presence of fixed-NH4+. Likewise, in a comparison of the Nt and Norg in organic-mineral clay fractions from a long-term (34-yr) humus formation study (Leinweber and Reuter, 1992), Nt concentrations deter- mined by dry combustion (4.5–23.4 g kg−1) exceeded the corresponding Kjeldahl Norg concentrations (0.40–20.4 g kg−1) by 4–17% (Leinweber, unpublished). Clearly, there is a need to exercise great care in determining Norg concentrations, and when using C:N ratios to evaluate SOM qual- ity. Surprisingly, however, there is little evidence in the recent literature to suggest that potential contributions from fixed-NH4+ are often taken into account when making C and N determinations for SOM characterization.
Stevenson (1994, 1996) compiled data on the various organic forms of N obtained by liberating Norg constituents from organic and inorganic colloids using hot, 6 M HCl followed by direct distillation and separate determination of NH3–N, or Kjeldahl N digestion, distillation and titra- tion of the 6 M HCl extract. In this relatively simple way, three different Norg fractions were obtained (Table 2.1). Comparisons of the Norg fractions determined in this way show a great similarity among samples—irrespective of soil type, management or particle-size fractions.
Figure 2.1 Impact of interlayer NH4+ on N concentrations as determined by (■) dry combustion (CNS-analyzer) and (▴) the Kjeldahl method. NH4+ smectites and ver- miculites with determined concentrations of sorbed NH4+–N were kindly provided by Dr. S. Dultz, Institute of Soil Science, Leibniz University, Hannover, Germany.
Bulk α-amino-N concentrations in the HCl extracts can be determined colorimetrically using the ninhydrin method. However, Stevenson (1996) emphasized that this method can be applied exclusively for amino acids with –NH2 adjacent to –COOH or with NH–CH2 group—thus under- estimating the amino-acid-N content of the sample. Amino acids in soil hydrolyzates can be determined by ion exchange-, liquid- (High pressure liquid chromatography (HPLC)) and gas chromatography—the latter two
Table 2.1 Distribution of Norg Fractions (%) in Different Reference Soil Groups and Organic-mineral Particle-size Fractions from Unfertilized and Fertilized Experimental Soils (uh−N = unidentified hydrolyzable N)
References Soil units Amino-N
+ uh−N NH3–N Nonhydrolyzed N
Sulỗe et al. (1996) Arenosol 43 19 37
Cambisol 37 29 34
Fluvisols 50–54 27–29 20–22
Vertisols 43–48 27–33 19–29
Zhang et al. (2006) Anthrosols 30–39 14–17 24–39 Leinweber et al. (1999) Vertisols 43–53 32–54 15–20 Leinweber and Schulten
(1998a, 1998b) Cambisols 51–54 26–29 13–17
Luvisols 41–53 21–26 8–13
Phaeozems 42–51 28–31 20–27
Chernozems 48–50 28–30 21
Clay,
unfertilized 88–95 46–58 13–18 Fine silt,
unfertilized
80–97 40–50 15–27
Medium silt,
unfertilized 76–94 40–45 14–27 Coarse silt,
unfertilized
42–97 46–68 11–28
Sand,
unfertilized 43–94 20–59 13–31 Clay, fertilized 81–97 43–53 14–18 Fine silt,
fertilized 72–88 31–44 18–25 Medium silt,
fertilized 73–82 33–39 19–27 Coarse silt,
fertilized 69–97 41–50 15–25 Sand, fertilized 58–95 14–47 10–25
often in combination with mass spectrometric detection. A review of 46 research articles published between 2000 and 20121 indicates the prevalence of HPLC methods. Olk (2008) summarized advances in the extraction of amino acids using 4 M methanesulfonic acid and quantification by anion chromatography. Using this approach, Martens and Loeffelmann (2003) and Martens et al. (2006) determined that amino acid N (consisting of 17 amino acids) accounted for 51–52% of soil N, with NH4–N and nonamide Norg accounting for 34% and 14–15%, respectively. Warren (2008) devel- oped a capillary electrophoresis with laser-induced fluorescence (CE-LIF) detection method as a more rapid alternative to GC and LC methods for measuring individual amino acids in soil KCl extracts. The CE-LIF method separated 17 common amino acids in crude 1 M KCl extracts, with detec- tion limits between 7 and 250 nM and a run time of only 12 min. The relative standard deviation of migration times for replicate analyses was
<0.2% and relative standard deviations for peak areas was <5%. Hou et al.
(2009) proposed an HPLC-MS method based on a 6-aminoquinolyl-N- hydroxysuccinimidyl carbamate (AQC) derivatization; baseline separation of 17 amino acid–AQC-derivatives on an XTerraR MS C18 column, using ammonium formate as a mobile-phase modifier; and optimized MS detec- tion. Detection limits were 0.20–0.60 pmol àL−1 on column (i.e. 0.07–
0.24 mg kg−1 soil) under the optimized conditions.
Amelung (2003) used amino acid enantiomers—especially the d:l ratios of alanine, aspartic acid, lysine and proline—as biomarkers to assess the mechanisms of Norg transformations and aging in soil. He also discussed the relevant analytical methods needed to reliably determine amino sugars and amino acid enantiomers in soil. Furthermore, Zhang et al. (2007) published a GC/MS method for the assessment of 15N and 13C incorporation into soil amino acid enantiomers using labeled 15NH4+ or 13C-glucose substrates.
In the mass spectra, 15N-enrichment of the amino acids was estimated by calculating the atom percentage excess from intensity increases of fragment peaks [F + 1 for neutral (alanine, proline) and acidic (aspartic acid) and F + 2 for basic amino acids (lysine)].
Although they do not occur in significant amounts in plant residues, amino sugars—such as glucosamine, galactosamine, mannosamine, and muramic acid, as well as more than 20 other minor amino sugars—account for about 5–10%
of the Norg in soil (Amelung, 2003) and can be used as biomarkers for soil Norg transformations affected by natural and anthropogenic influences. The
1 Obtained by searching Science Direct (http://www.sciencedirect.com/) for “amino acid soil”.
method for amino sugar determinations in 6 M HCl hydrolyzates was based on derivatization with acetonitrile–acetate, purification, GC separation and flame ionization detection (Zhang and Amelung, 1996).
Various mild extractants are used to isolate available N from soil. In addition to inorganic N (i.e. NO3− and NH4+), however, these extracts con- tain dissolved Norg (Kuzyakov and Siniakina, 2001; Landgraf et al., 2006).
Concentrations of Norg in these extracts are usually obtained by subtracting the available N (NO3− + NH4+; determined independently) from the Nt
determined using either the Kjeldahl method or thermal–catalytical oxida- tion (C, N EA). Various forms of Norg in these extracts can be qualitatively examined using the above-described Norg fractionations and/or amino acid and amino sugar determinations, though more detailed information can be gained using the different spectrometric and spectroscopic methods described in Sections 2.2, 2.3, and 2.4.