Published in Soil Sci. Soc. Am. J. 69:120-135 (2005).
© 2005 Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
Review and Analysis
Review
Organic Matter Removal from Soils using Hydrogen Peroxide, Sodium Hypochlorite, and Disodium Peroxodisulfate
R. Mikutta*,
M. Kleber,
K. Kaiser and
R. Jahn
Institut für Bodenkunde und Pflanzenernährung, Martin-Luther-Universität Halle-Wittenberg, Weidenplan 14, D-06108 Halle, Germany
* Corresponding author (mikutta{at}landw.uni-halle.de).
 |
ABSTRACT
|
|---|
We compare the performance of three most accepted reagents for organic matter removal: hydrogen peroxide (H2O2), sodium hypochlorite (NaOCl) and disodium peroxodisulfate (Na2S2O8). Removal of organic matter from soil is mostly incomplete with the efficiency of removal depending on reaction conditions and sample properties. Generally, NaOCl and Na2S2O8 are more effective in organic C removal than H2O2. Alkaline conditions and additives favoring dispersion and/or desorption of organic matter, such as sodium pyrophosphate, seem to be crucial for C removal. Pyrophosphate and additives for pH control (bicarbonate) may irreversibly adsorb to mineral surfaces. In soils with a large proportion of organic matter bound to the mineral matrix, for example subsoils, or rich in clay-sized minerals (Fe oxides, poorly crystalline Fe and Al phases, expandable phyllosilicates), C removal can be little irrespective of the reagents used. Residual organic C seems to seems to represent largely refractory organic matter, and comprises mainly pyrogenic materials and aliphatic compounds. If protected by close association with minerals, other organic constituents such as low-molecular weight carboxylic acids, lignin-derived and N-containing compounds may escape chemical destruction. For determination of mineral phase properties, treatment with H2O2 should be avoided since it may promote organic-assisted dissolution of poorly crystalline minerals at low pH, disintegration of expandable clay minerals, and transformation of vermiculite into mica-like products due to NH4+ fixation. Sodium hypochlorite and Na2S2O8 are less harmful for minerals than H2O2. While the NaOCl procedure (pH 9.5) may dissolve Al hydroxides, alkaline conditions favor the precipitation of metals released upon destruction of organic matter. Prolonged heating to >40°C during any treatment may transform poorly crystalline minerals into more crystalline ones. Sodium hypochlorite can be used at 25°C, thus preventing heat-induced mineral alteration.
Abbreviations: Ac, acetate • CEC, cation exchange capacity • IR, infrared • PAH, polynuclear aromatic hydrocarbons • SSA, specific surface area • XRD, x-ray diffraction
|
INTRODUCTION
|
|---|
REMOVAL OF organic matter by chemical reagents is a common pretreatment for analyses of the soil mineral phase such as particle-size distribution (Gee and Bauder, 1986), mineral composition (Tributh and Lagaly, 1991), cation exchange capacity (CEC), and specific surface area (SSA; Kahle et al., 2003). Chemical destruction of organic matter is also used to uncover mineral surfaces for subsequent sorption experiments (Jardine et al., 1989; Kaiser and Zech, 2000; Pagel-Wieder et al., 2004) and to assess organically bound metals (e.g., Shuman, 1983). Other fields for the use of chemical degradation of soil organic matter include the isolation of refractory organic fractions (Balesdent, 1996; Righi et al., 1995; Eusterhues et al., 2003) and the exploration of organic matter properties (Cuypers et al., 2002). Determination of soil mineral properties requires complete removal of organic matter without modification of the mineral phase. An incomplete removal of organic matter hampers the evaluation of phyllosilicate clay mineralogy (Tributh and Lagaly, 1991), SSA determination (de Jonge et al., 2000) and may also affect infrared (IR) and differential thermal analysis (Mitchell and Farmer, 1962).
Hydrogen peroxide was introduced by Robinson (1922) for soil texture analysis and became the most widely used chemical reagent for organic matter destruction. Some alternative reactants have been proposed to increase C removal efficiency and to reduce possible effects of H2O2 on minerals. Sodium hypochlorite (Anderson, 1963) and Na2S2O8 (Meier and Menegatti, 1997) have emerged as reagents with the greatest potential to replace H2O2. These reagents have been recently applied to soils (Kaiser et al., 2002; Kiem and Kögel-Knabner, 2002; Eusterhues et al., 2003) and sediments (Mayer, 1999).
To date, there is a vast number of oxidation protocols and much uncertainty regarding the effects of oxidative reagents on soil constituents. A thorough understanding of treatments for soil organic matter removal with oxidants is the key to a correct interpretation of experimental results. Thus, our objective is to review and compare the suitability of H2O2, NaOCl, and Na2S2O8 for organic matter removal from soils, sediments, and minerals. Special emphasis was put on (i) the evaluation of individual treatment procedures including reaction conditions required, (ii) mechanisms of interaction between the reagents and organic matter, (iii) the organic C removal efficiency and reasons for organic matter resistance, (iv) the effects of reactants on soil minerals and (v) metal precipitation, and (vi) implications for the use of chemical treatments for soil analysis.
|
PROCEDURES, REACTION CONDITIONS, AND APPLICATION
|
|---|
Removal of organic materials from soil by aqueous reactants is controlled by multiple factors, including the reaction conditions (pH, temperature, contact time, chemical additives) and soil properties (mineralogy, organic matter content, and quality). Attempts have been made to optimize reaction conditions to increase the efficiency of organic C removal from soils. This resulted in a variety of treatment protocols compiled in Table 1, which will be briefly discussed here.
Hydrogen Peroxide
Most protocols for sample preparation before textural and mineralogical analysis propose the use of 30% (wt/wt) H2O2 but 10 and 50% (wt/wt) H2O2 have been suggested as well (Table 1). McLean (1931b) found that doubling the H2O2 concentration from 6 to 12% resulted in little increase in C removal from soils (Fig. 1)
. This suggests that the concentration of H2O2 is not decisive for C removal efficiency, probably because of the decomposition of H2O2 (REACTION OF OXIDANTS WITH INORGANIC AND ORGANIC MATTER section), and the presence of chemically stable and mineral-protected organic compounds (EFFICIENCY OF ORGANIC CARBON REMOVAL DEPENDS ON SAMPLE PROPERTIES section).
View larger version (27K):
[in this window]
[in a new window]
|
Fig. 1. (a) Effect of added H2O2 volume on the efficiency of organic C removal from Welsh carbonate-free Madryn soil (•, 5.1% C) and Bodrwyn soil ( , 4.3% C) using 6%(wt/wt) H2O2. (b) Organic C removal efficiency from bulk soils with increasing H2O2 concentration. Soils contained between 1.3 and 6.5% organic C. Sixty milliliters H2O2 were added equivalent to 0.08 g C. The samples were heated gently then boiled (data adopted from McLean, 1931a, 1931b).
|
|
Typically, the treatments start at room temperature because of a strong initial reaction with easily decomposable organic matter. If frothing subsides, the sample is commonly heated to 60 to 90°C (Table 1). Increasing temperature accelerates the decomposition of H2O2 but shortens the reaction time necessary to oxidize organic matter (Schultz et al., 1999). At temperatures >70°C, H2O2 is rapidly consumed and additional H2O2 is needed to perpetuate organic matter destruction. At lower temperatures, the contact time needs to be extended (Schultz et al., 1999). In practice, removal of organic C from soils using H2O2 often requires several days and no reliable indicator exists showing completion of the reaction. Lack of visible frothing and bleached soil color indicate a complete reaction, but frothing may continue due to the decomposition of excess H2O2 at mineral surfaces (REACTION OF OXIDANTS WITH INORGANIC AND ORGANIC MATTER section).
Sometimes H2O2 is used in combination with acetate buffer (pH 5) to prevent acidic conditions resulting from formation of acid oxidation products (Douglas and Fiessinger, 1971; Pennell et al., 1995). The pH of unbuffered soil–H2O2 suspensions may drop by up to three units and final pH values between 2 and 4 have been reported (Douglas and Fiessinger, 1971; Lavkulich and Wiens, 1971; Griffith and Schnitzer, 1977). Using acetate buffer, H2O2 is additionally consumed due to the oxidation of acetate. Acetate (Ac) may also adsorb to minerals (e.g., van Hees et al., 2003) and thus contribute to residual C (17–41% C; Pennell et al., 1995).
There is little consistency in H2O2 use before SSA determination of the mineral phase. Hydrogen peroxide concentrations vary from 6 to 30% (Sequi and Aringhieri, 1977; Theng et al., 1999) and reaction temperatures from 50 to 80°C (Feller et al., 1992; Theng et al., 1999) (Table 1). Similarly, H2O2 concentration and contact time proposed for isolation of stable organic matter vary strongly (Righi et al., 1995; Theng et al., 1992). These examples demonstrate little agreement on protocols to achieve given research objectives.
Sodium Hypochlorite
The use of NaOCl for organic matter removal was first proposed by Anderson (1963) for mineralogical analysis of clays (Table 1). The method utilizes 6% (wt/wt) NaOCl at pH 9.5 and three consecutive cycles including boiling for each 15 min. The restriction of reaction time to 15 min is due to the fast decomposition of NaOCl at high temperatures. Heat-induced mineral changes (TREATMENTS INDUCE MODIFICATIONS OF MINERAL CONSTITUENTS section) can be minimized at room temperature, which requires extended contact times (Table 1; Kaiser et al., 2002). If applied at pH 9.5, NaOCl may partly dissolve Al secondary phases (TREATMENTS INDUCE MODIFICATIONS OF MINERAL CONSTITUENTS section). To avoid this, NaOCl can be used at lower pH (Table 1). Treating soils with NaOCl avoids vigorous frothing and boiling over as often experienced when using H2O2. Modified NaOCl protocols have been used for multiple purposes such as for metal extraction from soil organic matter (Shuman, 1983), before metal sorption studies (McDowell and Condron, 2001), and to investigate stabilization of organic C by mineral phases (Kaiser et al., 2002) (Table 1).
Disodium Peroxodisulfate
Meier and Menegatti (1997) proposed the use of Na2S2O8 to remove organic matter from mineral phases in one single step (Table 1). Thermal decomposition of Na2S2O8 did not decrease C removal at temperatures >80°C. By using a NaHCO3 buffer, the pH of reaction suspensions with clay minerals and soils can be kept between 7 and 8.5 (Menegatti et al., 1999; Kiem and Kögel-Knabner, 2002), thus preventing acid-mediated mineral dissolution. After treatment, Meier and Menegatti (1997) applied a hot wash (98°C, 1 min) with formic acid to remove any traces of salts. However, SO42– and HCO3– have a high affinity for hydroxyl-bearing mineral surfaces (Ali and Dzombak, 1996; Su and Suarez, 1997) and are likely to become attached to minerals such as Fe and Al oxides or short range order minerals when applied in high amounts (Table 1). Peroxodisulfate oxidation was employed for predicting the bioavailability of polynuclear aromatic hydrocarbons (PAH) in soils and sediments since both, oxidation and biodegradation, removed similar portions of PAH (Cuypers et al., 2000). Recently, the Na2S2O8 procedure has been adopted to soils to isolate stable organic matter and to study the influence of mineral surfaces on organic matter storage (Kiem and Kögel-Knabner, 2002; Eusterhues et al., 2003).
|
REACTION OF OXIDANTS WITH INORGANIC AND ORGANIC MATTER
|
|---|
In theory, CO2 and H2O are the end products of oxidative organic matter destruction but a variety of by-products form during the chemical treatments. Intermediate degradation products adsorb to mineral surfaces, form precipitates (Martin, 1954), complex trace metals, and thus hinder their redistribution to solid phases (Hoffman and Fletcher, 1981) and have the potential to promote mineral dissolution at low pH (Cornell and Schwertmann, 1996; Zhang and Bloom, 1999). The interaction of oxidants with organic matter involves complex reactions, depending on either one of the following factors: (i) reaction conditions (pH, temperature), (ii) the presence of inorganic catalysts that are capable to transform reagents into more reactive forms, and, in case of H2O2, (iii) the presence of enzymes that reduce the oxidant's reactivity.
Hydrogen Peroxide
Hydrogen peroxide is thermodynamically unstable and decomposes into O2 and H2O according to Eq. [1] (Pardieck et al., 1992). Decomposition of H2O2 increases with pH, being maximal close to the reagent's pKa at 11.6 (Xiang and Lee, 2000).
 | [1] |
When present in soils, reduced metal species can catalyze the decomposition of H2O2 and thereby produce OH radicals that are much more powerful oxidants than H2O2 (Strukul, 1992). The Fe2+–catalyzed decomposition of H2O2 into OH species according to Eq. [2] resembles the classical Fenton reaction. Surface sites of oxides and montmorillonite (Fe2+ or Fe3+) may similarly produce OH radicals by Fenton-like reactions (Huang et al., 2001; Gournis et al., 2002; Kwan and Voelker, 2002; Petigara et al., 2002). Equations [2] through [6] represent the metal-catalyzed Haber-Weiss mechanism, where the metal (M) is cycled between a lower [M(II)] and higher [M(III)] oxidation state and where H2O2 is gradually decomposed into O2 and H2O (Eq. [1]) via production of odd-electron (radical) intermediates (Evans and Upton, 1985). Fenton-like chemistry has been applied in remediation of soils (Watts et al., 2002; Kanel et al., 2003) and in wastewater treatment (Waite, 2002).
 | [2] |
 | [3] |
 | [4] |
 | [5] |
 | [6] |
Hydrogen peroxide can also be consumed by catalase-positive bacteria, which are common members of the microbial community in soil environments (Pardieck et al., 1992). Catalase decomposes H2O2 into O2 and H2O without formation of OH radicals (k 
4 x 107 M s–1; Pardieck et al., 1992). Manganese oxides can also induce the catalase-type reaction resulting in reduced efficiency of organic C removal in Mn-containing soils. When the soil reaction is acid, protonation of H2O2 may yield electrophilic OH cations (Eq. [7]) while at alkaline conditions nucleophilic perhydroxyl ions (HOO–) form (Eq. [8]).
 | [7] |
 | [8] |
The possible decomposition pathways of H2O2 in soils imply that several reactive H2O2 species can interact with organic matter. In the simplest case, H2O2 directly oxidizes organic compounds in a peroxidic-type reaction by a two-electron process without O2 formation (Schumb et al., 1955). Reaction of organic matter with OH radicals are far more complex. Under laboratory conditions, OH radicals react with alkenes and aromatics at diffusion-controlled rates (Watts et al., 2002). Hydroxyl radicals can also initiate radical reactions among organic radicals produced by hydrogen abstraction from C–H bonds. There is evidence that OH radicals preferentially attack aromatic compounds: Westerhoff et al. (1999) noted that aliphatic structures of natural organic matter react slower with OH radicals than aromatic moieties. Xie and Barcelona (2003) reported a higher chemical resistance of low-molecular weight aliphatic hydrocarbons (C5–C8) compared with aromatic compounds in jet-fuel contaminated sediments treated with H2O2. These results are consistent with the general order of reactivity: aromatic > –CH2– > –CO– > –COOH (Peyton, 1993) and may account for the enrichment of aliphatic compounds in H2O2–treated soils (Schulten et al., 1996; Leifeld and Kögel-Knabner, 2001; COMPOSITION OF ORGANIC MATTER section).
In addition, certain aromatic compounds may selectively be removed by H2O2. Andreozzi et al. (2002) observed that phenols adsorbed to goethite containing two adjacent OH groups (or one OH and one NH2) were oxidized by H2O2 while those containing two COOH groups or one COOH plus one OH group were not reactive toward H2O2.
Degradation of lignin strongly depends on the reaction milieu. At alkaline pH, degradation of lignin is more complete than at acid pH, possibly because of more reactive H2O2 species present and larger lignin solubility (Sun et al., 2000; Xiang and Lee, 2000). This suggests that under acid conditions lignin components may contribute to residual organic C in H2O2–treated soil. Chemical degradation of humic substances, lignin, and simple carbohydrates by H2O2 yields a variety of water-soluble compounds such as mono- and dicarboxylic low-molecular-weight organic acids (e.g., formic, acetic, oxalic, and malonic acid), phenols and benzenecarboxylic acids (Craik, 1924; Küchlein, 1932; Griffith and Schnitzer, 1977; Xiang and Lee, 2000; Goldstone et al., 2002). Martin (1954) estimated that in two soil clays reacted with H2O2, 30 to 40% of the initial organic C was transformed into oxalate. Similarly, Harada and Inoko (1977) using redox titration have calculated that 100 g of soil yielded between 0.01 and 0.2 moles of oxalate during H2O2 treatment.
Less is known about the interaction of H2O2 with N-containing compounds. Herriott (1947) reported that at room temperature 5% (wt/wt) H2O2 caused a loss of 60% of amino-N from virus protein. Harada and Inoko (1977) showed that after treating soils with H2O2, residual N was mainly water-soluble (60–100%) and dominated by inorganic N species (>70%). Ammonia is a product of the oxidative deamination of amino acids (Schnitzer and Hindl, 1980) and has been found after oxidation of humic acid (14 mg NH3 out of 46 mg total N) (Miles et al., 1985).
Sodium Hypochlorite
In aqueous solution, NaOCl principally exists in the following equilibria (Eq. [9] and [10]).
 | [9] |
and
 | [10] |
When natural organic matter is treated with hypochlorite species (HOCl, OCl–), high-molecular-weight chlorinated compounds form in a first step that is followed by the cleavage of benzene rings into trihalomethanes and haloacetic acids (Jimenez et al., 1993; Li et al., 2000; Pomes et al., 2000). Sodium hypochlorite-treated humic acid yielded more chloroform than treated fulvic acid (Peters et al., 1980). Trihalomethane and haloacetic acids likely derive from 3,5-dihydroxybenzene structures. In accordance, Norwood et al. (1987) noted that chlorination of aquatic fulvic acid produced chlorinated aliphatics and resulted in preferential removal of lignin phenols. Chlorinated phenols were also found after reaction of hypochlorite with humic and fulvic acid (Quimby et al., 1980), substituted benzoic acids (Larson and Rockwell, 1979), and after chlorination of lignin (Rajan et al., 1996). Westerhoff et al. (2004) observed a higher reactivity of aqueous chlorine for aromatic compounds than for aliphatic compounds in natural organic matter. Chakrabartty et al. (1974) used 12% (wt/wt) NaOCl (pH 12) to investigate the structural composition of humic acids extracted from soil and coal. They reported that reaction results in aliphatic carboxylic acids (malonic, succinic, glutaric, and adipic acid) and aromatic polycarboxylic acids (>50% CO2, up to 32% non-volatile acids). Hypochlorite is presumed to cleave all methine, methylene, and methyl goups activated by electron-withdrawing heteroatoms without simultaneously disrupting aromatic systems.
Disodium Peroxodisulfate
In alkaline to slightly acid solutions, S2O82– reacts with H2O according to Eq. [11] (Kolthoff and Miller, 1955) to finally give SO42– by dissociation of HSO4–.
 | [11] |
When heated, S2O2–8 undergoes thermal homolysis or reacts with reduced metals, for example, Fe2+ ions, to yield SO4– radicals that can react with organic matter (Edwards and Curci, 1992). Hydroxyl radicals can also form when SO4– radicals react with H2O and with OH– ions at alkaline conditions (Eq. [12] and [13]). Reaction [12] becomes negligible at dissolved organic C concentrations >1 mg L–1, where organic compounds are the major sink for SO4– radicals (Peyton, 1993).
 | [12] |
 | [13] |
Like OH radicals, SO4– radicals have a similar reactivity toward organic structures (Peyton, 1993). Larger, more resistant phenolic compounds may be synthesized by coupling of phenols during radical reactions induced by SO–4 radicals.
EFFICIENCY OF ORGANIC MATTER REMOVAL
Organic matter cannot completely be removed from soils by wet oxidative treatments (Fig. 1). Early work took advantage of this fact by using H2O2 for determination of the degree of humification in soils (Robinson and Jones, 1925) and to gain information about the composition of organic matter (McLean, 1931a, 1931b). Progressing research revealed a number of factors responsible for incomplete C removal such as soil reaction, presence of carbonates, chemically resistant organic compounds, and protection of organic matter by mineral surfaces.
Hydrogen Peroxide
The extent of organic C removal by H2O2 varies with soils and particle-size separates, ranging from <20% (Bartlett et al., 1937) to >93% (Kahle et al., 2003). Typical efficiencies of C removal are listed in Table 1. Hosking (1932) noted for a number of calcareous Australian Black soils that 20% (wt/wt) H2O2 removed only 5 to 20% of organic matter when the soil reaction was alkaline (pH 9–10) while 50 to 90% were removed when the soil pH was between 6 and 7.5. In soils containing Mn oxides, only <30% of organic matter was removed because of rapid H2O2 decomposition. Hosking (1932) and Anderson (1963) increased the C removal by removing carbonates first, using HCl (pH 2–3) or a Na acetate buffer (pH 5). In calcareous soils, carbonate coatings favor occlusion and thus physical protection of organic matter against destruction. Oxalate, which is a common product of organic matter destruction (HYDROGEN PEROXIDE under section REACTION OF OXIDANTS WITH INORGANIC AND ORGANIC MATTER), can form insoluble complexes with Ca2+, thus leaving a residual organic C fraction behind after H2O2 treatment. Moreover, HCO–3 and CO2–3 ions are known to inhibit organic matter degradation during water purification by scavenging OH radicals (Wang et al., 2001). The same may hold true for H2O2 treatment of calcareous soils where carbonates likely dissolve due to the production of organic acids.
Using Na4P2O7 as a dispersing agent in the H2O2 treatment, organic C removal is more complete than by applying H2O2 alone (Simon et al., 1992). Sequi and Aringhieri (1977) increased the average organic matter removal by using H2O2–Na4P2O7 from 79 to 96%, likely by disruption of aggregates, thereby releasing occluded organic matter, and by displacing sorbed organic compounds from mineral surfaces. Pyrophosphate and also PO3–4 sorb strongly and hysteretically to soil minerals (Varadachari et al., 1995; Celi et al., 2000). Hence, when carrying out a sorption experiment with soil treated with a mixture of H2O2 and Na4P2O7, occupation of reactive sorption sites by pyrophosphate needs to be considered. Furthermore, organo–mineral complexes and chemically stable compounds contribute to the resistance of organic matter (EFFICIENCY OF ORGANIC CARBON REMOVAL DEPENDS ON SAMPLE PROPERTIES section).
Sodium Hypochlorite
According to Anderson (1963), the C removal efficiency of 6% (wt/wt) NaOCl at pH 9.5 was less affected by the presence of carbonates (up to 9% CaCO3) than when using 30% (wt/wt) H2O2. Several studies revealed that 6% (wt/wt) NaOCl at pH 9.5 removed more C from bulk soils and clays than 30% (wt/wt) H2O2 (Anderson, 1963; Omueti, 1980; Cheshire et al., 2000). The NaOCl treatment does not require additional dispersing reagents because of dispersion by NaOCl itself. In spite of a higher electric potential of NaOCl with decreasing pH, C removal efficiency is maximal at pH 9.5 (Lavkulich and Wiens, 1970). This indicates that desorption from mineral surfaces under alkaline conditions is decisive for the removal of organic compounds. Desorption may be of similar importance for organic matter removal as the oxidative breakdown of organic compounds, suggesting the term "wet oxidation" to be misleading. Kaiser and Guggenberger (2003) tested a modified NaOCl treatment (pH 8, room temperature, five repetitions) on 196 heavy soil fractions (density >1.6 g cm–3) and found that between 77 to 95% of the initial organic C could be removed by this procedure. Reduction of the number of repetitive treatments may decrease the removal of organic matter in soils rich in poorly crystalline Fe and Al phases (Siregar et al., 2004).
Disodium Peroxodisulfate
The efficiency of Na2S2O8 to remove organic matter has been proposed to be superior to H2O2 and NaOCl (Meier and Menegatti, 1997; Table 1). When the authors tested NaH2PO4 as buffer, C removal was less than when using NaHCO3 buffer. This can be attributed to the lower pH (drop from approximately 8 to 5) in case of the Na2S2O8–NaH2PO4 mixture, which disfavors desorption of mineral-bound organic matter. Kiem and Kögel-Knabner (2002) applied Na2S2O8 to particle-size fractions of loamy and sandy surface soils and found a contact time of 16 h sufficient to reduce organic C concentrations by about 93% (Table 1). Oxidative destruction of humic acid (Martin et al., 1981) and organic matter from aquifer sediments (Powell et al., 1989), however, not exceeded 50%. Thus, desorption of organic matter by HCO–3 or SO2–4 is likely the key step in removal of organic C from soils. In contrast, Eusterhues et al. (2003) found that in some cases even 2 d of reaction time were insufficient to destroy organic C in acid soils rich in secondary minerals (Table 1). This indicates that either chemically stable compounds represented a major organic fraction (not tested) or strong interactions with mineral surfaces limited desorption of organic matter (PROTECTION OF ORGANIC MATTER BY SOIL MINERALS section under EFFICIENCY OF ORGANIC CARBON REMOVAL DEPENDS ON SAMPLE PROPERTIES).
|
EFFICIENCY OF ORGANIC CARBON REMOVAL DEPENDS ON SAMPLE PROPERTIES
|
|---|
Sample-specific properties like the presence of chemically stable organic compounds and protective mineral phases are also crucial for organic C removal efficiency.
Composition of Organic Matter
Pyrogenic materials (black carbon) represent a group of chemically resistant compounds in soils. They comprise randomly oriented condensed aromatic rings located in graphite-like structures (Schmidt and Noack, 2000). Such materials can make up a significant portion of organic matter in soils, especially in soils adjacent to coal-processing industries (Schmidt et al., 1999) or in soils from technogenic materials (Zikeli et al., 2004). Hydrogen peroxide is ineffective to degrade graphite, anthracite, lignite, charcoal, and ash (Robinson, 1927; Schmidt et al., 1999).
Low-molecular-weight organic acids produced during oxidative treatments are relatively resistant against further degradation (Luft and Stöffler, 1998) and thus become enriched during oxidative treatments. Analyzing H2O2–treated clay and bulk soil fractions by differential thermal and infrared analysis, Farmer and Mitchell (1963) and Harada and Inoko (1977) found no evidence for resistant organic compounds other than oxalate. Formation of insoluble calcium oxalates is favored under alkaline conditions, while in acidic soils adsorption of oxalate to variable charge minerals is more likely. Farmer and Mitchell (1963) noted that in some cases complexed oxalate couldn't be extracted by H2O but by 5% ethylenediaminetetraacetate. In contrast, Escudey et al. (1999) suggested extensive washing with H2O after the H2O2 treatment to remove oxalate from soils completely but did not quantify their results.
Several studies confirmed that organic compounds other than black carbon and oxalate could survive oxidative treatments (Righi et al., 1995; Schulten et al., 1996; Cuypers et al., 2002). Aliphatic compounds are a significant portion of chemically resistant organic matter. Griffith and Schnitzer (1977) estimated that n-alkanes and n-fatty acids accounted for up to 40% of the H2O2–resistant organic matter, while a substantial fraction of residual C could not be extracted by organic solvents. Studying the H2O2–resistant organic matter in mica–beidellite interstratified clay by Py-FIMS, Schulten et al. (1996) found it to be enriched in N-containing compounds, n-C22–C26 carboxylic acids, n-alkanes, n-diols and alkyl-substituted aromatic esters. Leifeld and Kögel-Knabner (2001) showed by CPMAS 13C-NMR analysis that H2O2 preferentially removed sugars (O-alkyl C) and lignin compounds (mainly aromatic C) from <20-µm particle-size separates of agricultural soils leaving an aliphatic residue behind. In accordance, persulfate treatment evidenced the presence of aliphatic biopolymers in sedimentary and terrestrial humic acids (Saiz-Jimenez, 1992). Martin et al. (1981) found that n-C16–C18 fatty acids were the most abundant compounds after reaction of humic acid with acid K2S2O8. However, a variety of benzenecarboxylic and phenolic acids were also detected. Cuypers et al. (2002) reported that persulfate preferentially removed labile and more amorphous organic matter while the residuum was enriched with long-chain aliphatics. The resistant organic matter was proposed to have a more condensed structure, a higher affinity for hydrophobic compounds and being more thermostable than amorphous organic matter.
Lower C/N ratios of H2O2–resistant organic matter indicate a higher chemical stability of some N compounds. Cheshire et al. (2000) found that some amino acids were protected against degradation possibly within microaggregates or by interaction with mineral surfaces. Interpretation of C/N ratios after organic matter removal is, however, questionable since mineral-bound NH+4 or NH+4 produced during organic matter degradation and subsequently fixed to minerals can also affect C/N ratios (Miles et al., 1985; Leifeld and Kögel-Knabner, 2001).
Protection of Organic Matter by Soil Minerals
Hosking (1932) recognized that more organic matter resisted the H2O2 treatment in soils with higher clay content. This suggests that organic matter can be protected against oxidative treatments by interaction with soil minerals. Contemporary work distinguishes between the mechanisms of intercalation and sorptive protection.
Intercalation includes replacement of hydrated inorganic interlayer cations of expandable 2:1 phyllosilicate clay minerals by organic molecules. This requires either cationic or neutral organics and a strongly acidic soil reaction. Theng et al. (1992) and Righi et al. (1995) assumed that organic matter intercalated in expandable clay minerals is hardly decomposable by H2O2 due to a limited accessibility toward the reagent. In contrast, Kodama and Schnitzer (1971) found no evidence that organic matter sorbed into the interlayer regions of mica-vermiculite-montmorillonite-interstratified clay resisted 15% (wt/wt) H2O2.
Eusterhues et al. (2003) observed that the concentration of Na2S2O8–resistant organic matter related positively to the clay content (r = 0.93) in twelve horizons from two acid soil profiles (Typic Haplorthod and Typic Dystrochrept). In the Spodosol, a significant relationship of residual organic C to dithionite–citrate–bicarbonate-extractable Fe (r = 0.90) suggests that the resistant organic matter was likely associated with Fe (hydr)oxides (sorptive protection). Similar results were reported by Mikutta et al. (2004) for NaOCl-resistant organic matter in clay subfractions from subsurface horizons of smectitic, vermiculitic, illitic, kaolinitic, and chloritic soils (Table 1). Despite the heterogeneous sample set, the concentration of residual organic C correlated well with oxalate-extractable Fe (Feo) and Al (Alo) (Fig. 2a)
. Acid oxalate primarily extracts Fe and Al from poorly crystalline aluminosilicates, ferrihydrite and Fe and Al humus complexes (Wada, 1989).
View larger version (28K):
[in this window]
[in a new window]
|
Fig. 2. (a) Relation between the concentrations of residual organic C and oxalate-extractable Fe plus Al in fine (•, <0.2 µm) and coarse clay (, 0.2–2 µm) fractions of 12 acid subsoil samples with different mineralogy after treatment with 6% (wt/wt) NaOCl. (b) Relationship between the organic C loading of mineral surfaces and the amount of removable organic C (data adopted from Mikutta et al., 2004).
|
|
The efficiency of poorly crystalline minerals to protect organic matter against oxidative destruction was confirmed in batch experiments. Singer and Huang (1993) and Huang (1995) showed that humic, tannic, and citric acid associated with non-crystalline Al hydroxide were only partly destructible by H2O2 and NaOCl, while tannic and citric acid were completely oxidized in the absence of Al hydroxide. The higher resistance of mineral-associated organic compounds was explained by incorporation of organic compounds by coprecipitation or by formation of strong surface complexes.
Mikutta et al. (2004) showed that the capability of minerals in clay subfractions to protect organic matter decreases as more organic C is associated with mineral surfaces. The C removal efficiency related closely to the initial organic C concentration normalized to the mineral SSA (C loading; Fig. 2b). This result is consistent with the finding that at small C loadings, organic matter occupies a larger surface area with more functional groups involved in multiple-site attachments with mineral surfaces (Kaiser and Guggenberger, 2003). Organic matter in direct contact to the mineral surface thus appears less susceptible to chemical destruction. Eusterhues et al. (2003) provided further evidence that the C removal efficiency depends on the protective capability of mineral surfaces. They observed that C removal by Na2S2O8 decreased with soil depth. Nearly all organic C was removed from A and EA horizons (>98%) while in the deeper B and C horizons 9 to 84% of the initial organic C resisted Na2S2O8. This effect can be explained by larger amounts of mineral-bound organic matter and by stronger association of organic matter with mineral surfaces with increasing soil depth. In surface horizons, there is much unprotected, particulate organic matter and binding sites at mineral surfaces may already be occupied by organic matter resulting in weaker bindings. In deeper horizons where less of the mineral surface area is covered by organic compounds (small C loading), organic matter may sorb in a more spread-out, uncoiled conformation with more ligands involved in direct contact with the mineral surface (Kaiser and Guggenberger, 2003). As a result, desorption of organic matter from mineral surfaces during oxidative treatments is more difficult in subsoils than in topsoils.
|
TREATMENTS INDUCE MODIFICATIONS OF MINERAL CONSTITUENTS
|
|---|
The use of chemical destruction of organic matter is based on the supposition that minerals are unaffected by the treatments. Here we summarize reports on mineral changes likely induced by the treatments.
Thermal Effect on Minerals
Most treatments are conducted under elevated temperatures (60–100°C) (Table 1). Kaiser and Guggenberger (2003) showed that temperatures >40°C lasting several hours can transform moist amorphous Al hydroxide into gibbsite while temperatures >80°C converted ferrihydrite into hematite. The total SSA of both mineral phases decreased on heating by about 90%, partly due to the entire loss of microporosity (<2 nm). So even moderate heating can alter oxide surface properties and thus gives rise to artifacts, at least in soils rich in poorly crystalline components. The only procedure reported to be efficient under ambient temperature is the modified NaOCl treatment proposed by Kaiser et al. (2002) (Table 1). However, the method is time-consuming and laborious.
Hydrogen Peroxide
Phyllosilicates
Many studies on the effects of H2O2 on individual minerals were conducted outside the field of soil research (Hayashi and Oinuma, 1964; Muromtsev et al., 1990; Üçgül and Girgin, 2002). The use of unweathered minerals and the variety of procedures applied, complicates the transferability to soil systems. Despite this, some general remarks can be made.
Drosdoff and Miles (1938) first noted destruction of mica and some vermiculite samples when treated with 6 and 30% (wt/wt) H2O2. Mineral exfoliation resulted from catalytic decomposition of H2O2 by Mn oxides located in mineral interlayers. Amonette et al. (1985) showed that H2O2 penetrates biotite interlayer spaces expanded by tetraphenylborate and oxidizes 85% of the structural Fe2+ ions within 24 h. Also, H2O2 diffuses into the interlayer spaces of phlogopite (mica with low Fe content) and vermiculite via exchange with H2O and cations and decomposes into O2 and H2O (Üçgül and Girgin, 2002; Obut and Girgin, 2002). The gas evolved can disrupt individual silicate layers. Increasing thickness of the minerals (80–120 fold) was noted with increasing H2O2 concentration (1–50%), temperature (40–60°C), and contact time (1–30 h). Above 60°C, phlogopite started to exfoliate as described by Drosdoff and Miles (1938), possibly because of accelerated decomposition of H2O2 with increasing temperature. However, treatment with 30% (wt/wt) H2O2 (60°C, 70 min) caused no phase change of phlogopite according to x-ray diffraction (XRD) (Üçgül and Girgin, 2002).
Douglas and Fiessinger (1971) showed by XRD that the (001) signals of smectite and vermiculite decreased after reaction with H2O2 in presence of large amounts of sucrose (168 g C kg–1 clay) while little effect was observed without sucrose or in presence of sodium acetate buffer (pH 5). They inferred that both minerals, especially vermiculite, were partly destroyed due to the low pH (pH 1.8–3) induced by incomplete oxidation of sucrose. However, decreasing (001) basal peak reflections following H2O2 treatment may also result from residual organic matter producing less than perfect orientation of clay specimen (Dohrmann, 2003).
Miles et al. (1985) observed NH4+ produced during organic matter degradation to sorb to exchange sites of vermiculite, inducing collapse of the 1.48-nm spacings to the 1.03-nm position characteristic for mica-like phases. Infrared spectroscopy confirmed NH4+ fixation by the treated vermiculite. Therefore, the H2O2 treatment is not suitable for XRD-based identification of vermiculite in soils.
Van Langeveld et al. (1978) inferred from reduced XRD reflections that montmorillonite (6% organic C) was structurally altered during H2O2 treatment. Figure 3a and 3b
display the transformation of bentonite on treatment with 30% (wt/wt) H2O2 at 80°C for 6 h in presence of dissolved organic matter. Compared with untreated particles, the treated ones show distinctively frayed structures suggesting corrosion likely caused by decomposing H2O2 and acidic organic matter degradation products.
View larger version (288K):
[in this window]
[in a new window]
|
Fig. 3. (a) Scanning electron microscope (SEM) image of oxide-free (dithionite–citrate-treated) Prassa-Kimolos bentonite (Greece), (b) oxide-free bentonite treated with 30% (wt/wt) H2O2 (6 h, 80°C) in the presence of 230 mg L–1 dissolved organic C. Bars represent 100 µm.
|
|
In contrast, Hayashi and Oinuma (1964) found no indications for major alterations of repository clays (vermiculite, montmorillonite, illite, chlorite) treated with 30% (wt/wt) H2O2 at 80 to 90°C over several days. Muromtsev et al. (1990) recognized no structural alteration of pure vermiculite treated with 30% (wt/wt) H2O2 by applying XRD analysis, nevertheless the CEC doubled because of layer expansion. Rosenberg and Hooper (1997) showed that deposit illites were not attacked by H2O2. We studied illite treated with 10% (wt/wt) H2O2 (80°C, 120 h) in the presence and absence of organic matter. Before treatment, clay specimens were subjected to drying–wetting cycles in the presence of manure and straw (0.1 mm) to facilitate the formation of clay–organic associations. Using XRD, we observed no structural change of illite induced by H2O2 treatment (Fig. 4)
.
View larger version (21K):
[in this window]
[in a new window]
|
Fig. 4. X-ray diffractograms of Mg-saturated, glycerated, and oriented Fithian illite (I) (<2 µm) before and after treatment with 10% (wt/wt) H2O2 (Fithian illite: 85% illite, 15% mixed layers, kaolinite, quartz). The treatment was conducted in the presence of straw and manure (S, M, diam. 0.1 mm). Before treatment, organo–mineral associations (5% organic matter) were prepared by three drying and wetting cycles. Treatments: (a) no, (b) I + H2O2, (c) I + H2O2 + S, and (d) I + H2O2 + M.
|
|
The results compiled above indicate that the alteration of phyllosilicates by H2O2 depends on (i) the presence of catalyzing materials, (ii) the existence of accessible interlayer mineral surfaces, and (iii) the abundance of organic matter. The results imply that pure depository minerals not weathered in the soil environment and thus lacking structural modifications such as cracks, microfissures, and expansion zones as well as sorbed organic matter are little affected by treatment with H2O2.
Carbonates, Oxides, and Sulfides
In calcareous soils, H2O2 treatment may cause dissolution and corrosion of carbonate minerals because of the low pH of H2O2 (Pingitore et al., 1993). Typically, Mn oxides are destroyed during the H2O2 treatment (Shuman, 1983; Papp et al., 1991; Table 2). Manganese (III, IV) oxides are affected by reduction to Mn2+, which is subsequently oxidized to yield MnOOH (Pardieck et al., 1992), Mn3O4 (Jackson, 1958), or a mixture of different Mn oxides (Moon et al., 1999). Sulfide minerals are readily dissolved by reaction with H2O2 (Mukherjee et al., 2001). Papp et al. (1991) showed that H2O2 dissolved 21 to 99% of the total metals present in sulfides (Table 2). Consequently, H2O2 should not be used to estimate organically bound metals when sulfide minerals are present, for example in acid sulfate soils from coastal lowlands (Shamshuddin et al., 2004).
Iron and Al (hydr)oxides may be affected during the H2O2 treatment by heat-induced transformation into more crystalline forms and by organic-assisted dissolution at low pH. At alkaline pH, for example, in calcareous soils or when using H2O2 in combination with Na4P2O7, oxides remain unaltered (Marzadori et al., 1991). At low pH (pH 2–4), large concentration of oxalate, and high temperature, the H2O2 treatment is comparable or even more aggressive than the acid-oxalate method (pH 3, room temperature, darkness, 4 h) designed for extraction of poorly crystalline minerals and Fe and Al bound to organic matter (Blakemore et al., 1987). For acid soils treated with 30% (wt/wt) H2O2, Lavkulich and Wiens (1970) reported that the amount of oxalate-extractable Fe and Al decreased by 34 to 80% and 1 to 90% (II–H2O2 treatment, Table 3). They assumed that inorganic Fe and Al phases dissolved during the H2O2 treatment. However, mass balance calculations utilizing selective extractions prior and after H2O2 treatment suggest that dissolution of poorly crystalline Fe and Al components alone does not entirely explain the smaller Feo and Alo concentrations after H2O2 treatment (Table 3). More likely, they result from heat-induced recrystallization of poorly crystalline oxides. However, after mild H2O2 treatment (I-H2O2 treatment; Table 3), the mass balance was positive indicating higher extraction efficiency of acid oxalate after the H2O2 treatment due to disaggregation and removal of organic matter coatings from oxide surfaces (Marzadori et al., 1991). When testing the influence of 10% (wt/wt) H2O2 on six acid subsoil samples, we found an average decrease of Feo and Alo concentrations by approximately 90% (Table 3). Aluminum and Fe in reacted H2O2 solutions were little (Table 3), suggesting the decrease in oxalate-extractable components to be related to the transformation of minerals at high temperature (80°C).
View this table:
[in this window]
[in a new window]
|
Table 3. Effect of H2O2 and NaOCl treatment on oxalate-extractable Fe and Al fractions in acid bulk soils. Negative values in the ‘mass balance’ column indicate that poorly crystalline components were transformed during heating into more crystalline forms. Data for the H2O2 (30%) and the NaOCl treatment adapted from Lavkulich and Wiens (1970).
|
|
Pyrophosphate, when used in conjunction with H2O2 as a dispersing reagent to enhance C removal, can cause loss of peptized fine mineral particles, especially in presence of adsorbed organic matter, or dissolution of certain mineral constituents like amorphous Al(OH)3 (Parfitt and Childs, 1988; Kaiser and Zech, 1996). In addition, pyrophosphate may induce precipitation of Ca phosphates (Hawke et al., 1989; Celi et al., 2001) and Al phosphates (e.g., poorly crystalline Al phosphate, taranakite, wavellite) (Kim and Kirkpatrick, 2004).
Sodium Hypochlorite
In general, Mn oxides and sulfide minerals are dissolved to a lesser extent by NaOCl compared with H2O2 (Table 2). Lavkulich and Wiens (1970) found that, at pH 9.5, NaOCl dissolved significantly less oxalate-extractable Fe and Al than H2O2 (Table 3). This can be explained by the alkaline pH of NaOCl, which prevents acid-induced mineral dissolution and may probably support hydrolysis and precipitation of hydroxides (METAL PRECIPITATION section). However, some Al can be dissolved at pH 9.5, for example in the Alouette Ap horizon (Table 3). Using the method of Lavkulich and Wiens (1970), Osei and Singh (1999) reported that no Fe was released from tropical surface soils by NaOCl while extracted Al and Si accounted for up to 0.3 g kg–1. Negligible amounts of Fe and Al were dissolved during treatment of a Mollisol surface soil with 5.3% (wt/wt) NaOCl (Qiang et al., 1994; Table 1). In soils rich in poorly crystalline minerals, 6% (wt/wt) NaOCl at pH 8 (room temperature, three repetitions) dissolved no Fe and <3% of dithionite-citrate extractable Al and Si (Siregar et al., 2004). These findings support the view that crystalline oxides and silicates are not affected by NaOCl.
However, NaOCl treatments at higher temperatures may induce changes in extractable pedogenic Al and Fe (e.g., Langley Ap; Table 3) due to transformation into more crystalline forms. In contrast, Marzadori et al. (1991) studying calcareous soils found an increase in Feo and Alo after destruction of organic matter by 7% (wt/wt) NaOCl despite heating to 80°C (Table 1). In that case, the thermal transformation of a poorly crystalline oxide fraction was probably overcompensated by an increased extractability of poorly crystalline phases due to the removal of organic coatings. Mayer (1999) found an increased enthalpy for the adsorption of N2 on marine sediments treated with 13% (wt/wt) NaOCl at pH 9 to 9.5. The change in adsorption enthalpy was attributed to microtopographic or chemical changes of mineral surfaces due to the highly alkaline reaction conditions. This effect warrants further research.
Disodium Peroxodisulfate
Using XRD, IR spectroscopy and N2 adsorption, Menegatti et al. (1999) found no structural alteration of Na2S2O8–treated illite, kaolinite, and montmorillonite reference minerals (<0.2 and <2 µm). Minor losses of Ca, K, and Mn were explained by partial removal of fine-grained accessory minerals such as feldspars and carbonates. The SSA of most treated samples remained unaltered while the SSA of montmorillonite increased by >20 m2 g–1. This was attributed to the dispersion of large aggregates. The CEC of the treated minerals remained unchanged. However, HCO–3 as used for pH control may partly be converted into CO2–3 on heating, which at 80°C, can extract Al from allophanes and hydroxides (Wada, 1989). Thus, if used for organic matter removal from soils where these phases are abundant, for example volcanic ash soils and Spodosols, this potential defect requires further attention.
|
METAL PRECIPITATION
|
|---|
The question whether metal ions complexed with organic matter are released and precipitate during oxidation is relevant for the reliable assessment of mineral phase properties and organically bound metal fractions. Sequi and Aringhieri (1977) noted that Fe and Al associated with organic matter are released during H2O2 oxidation and subsequently precipitate as hydroxides on inorganic and organic components. These precipitates are assumed to alter SSA and charge of the mineral phase and to protect organic matter against further degradation. The formation of metal precipitates during the H2O2 treatment was inferred from the increase in SSA (N2–BET) and positive surface charge, and the concomitant decrease of negative charge. However, it is more likely that the alteration of surface charge derived from the removal of organic matter carrying much negative charges located in carboxylic groups and from exposure of positively charges on oxide surfaces. Despite that, precipitation of metals formerly complexed with organic matter cannot be ruled out. Schultz et al. (1999) preferred H2O2 instead of alkaline NaOCl to assess organically bound metals in carbonate-free sediments to avoid metal precipitation (Table 1). Addition of NH4 acetate during the H2O2 treatment was intended to prevent readsorption of metals to the solids. Hoffman and Fletcher (1981) showed that NaOCl (6.5%, pH 9.5) induced little precipitation of trace metals (Cu, Zn, Fe, Mn, Mo) when inorganic precipitates from soils and sediments were redissolved by acidified H2O. In spite of the alkaline conditions, soluble organic compounds can effectively scavenge metals released and thus hinder their precipitation. In accordance, Shuman (1983) showed for six bulk soils that 5% (wt/wt) NaOCl (pH 8.5) dissolved similar amounts of metals (Cu, Fe, Zn) from the organic fraction as 30% (wt/wt) H2O2 while H2O2 extracted significantly more Mn. In summary, there is no consistent perception on the degree of precipitation of organically bound metals released during organic matter destruction and on the potential effects on mineral properties.
|
SYNTHESIS
|
|---|
This review shows that a large number of protocols for removal of organic matter by H2O2, NaOCl, and Na2S2O8 is available. Removal of soil organic matter is never complete and largely relies on the reaction conditions chosen and the sample properties. Sodium hypochlorite and Na2S2O8 are more effective in organic C removal than H2O2, especially in calcareous soils where oxidation-resistant Ca oxalates may form. Carbon removal efficiency can be increased by desorbing reagents such as NaHCO3 and Na4P2O7, but those compounds strongly sorb to minerals, which needs to be considered in sorption experiments with treated samples. In soils with large portions of mineral-bound organic matter, Fe and Al (hydr)oxides and other poorly crystalline constituents, even NaOCl and Na2S2O8 fail to reduce the organic C concentrations effectively. Clay-sized Fe and Al phases provide large SSA and reactive hydroxyls favoring ligand exchange reactions with organic matter (Gu et al., 1995). The resulting mineral–organic attachments are difficult to break, especially when a single molecule is attached to multiple mineral surface sites, thus hampering desorption of organic matter, which is an important prerequisite for organic matter removal. However, sorptive protection of organic matter by minerals is hardly discernable from protection of organic matter due to occlusion in microaggregates or coprecipitation with hydrous oxides. Since intercalation is confined to expandable clay minerals and strongly acid conditions, the relevance of protected organic compounds in interlayer spaces is presumably small. Organic matter resistant to oxidative treatment likely represents a refractory C pool since it seems to be older than bulk organic matter before treatment (Theng et al., 1992). Similar composition of residual organic matter after chemical and biological degradation suggests the treatment with oxidative reagents to mimic biodegradation (Rihani et al., 1995; Cuypers et al., 2002). However, selective degradation of certain structures of organic matter by the three reagents (aromatics versus aliphatics) and, in case of H2O2, the likely adsorption of organic intermediates deriving from labile organic matter (e.g., oxalate) to minerals needs to be considered.
Based on the evidence shown, changes of mineral phase properties when treated with oxidative reagents seem inevitable. This attributes some ambiguity to studies concerned with soil mineralogical composition or surface-controlled processes (e.g., adsorption, diffusion). Hydrogen peroxide has been shown to disintegrate micas, vermiculites, and smectites. Phyllosilicate disintegration may become more significant with increasing H2O2 concentration, temperature and in the presence of decomposable organic matter. In soils, destruction of phyllosilicates might change the surface properties but little experimental evidence has been given. For example, Theng et al. (1999) hypothesized the SSA increase of a smectitic soil following H2O2 treatment to result from destruction of smectite. Treatment with NaOCl produced only a minor increase in SSA although both reagents removed similar amounts of organic matter. Compared with H2O2, the effects of NaOCl or Na2S2O8 on phyllosilicates seem negligible, but more studies on soils are needed to confirm that.
Poorly crystalline constituents are most susceptible to alteration during treatments for organic matter removal. In organic matter-rich soils treated with H2O2, organic oxidation products like low-molecular-weight organic acids may assist mineral dissolution at low pH. During the alkaline NaOCl procedure (pH 9.5), Al from hydrous oxides can dissolve. This effect can be avoided by using NaOCl at lower pH. Moreover, temperatures >40°C applied during organic matter removal involves the risk of recrystallization of poorly crystalline Al phases, while temperatures >80°C may convert poorly crystalline into more crystalline Fe oxides. Since poorly crystalline phases significantly contribute to the physical and chemical properties of soils (Percival et al., 2000; Kiem and Kögel-Knabner, 2002), more effort should be put on that question. At present, only the NaOCl treatment has been shown to remain efficient at room temperature and thus avoids heat-induced transformation of sensitive mineral phases.
Metals released during degradation of organic matter may precipitate, and thereby possibly reduce the C removal efficiency and the amount of organically bound trace metals in sequential extraction studies, and may also affect surface properties of minerals. However, little awareness exists of that problem. Hydrogen peroxide and NaOCl seem to induce similar metal precipitation, with the degree of precipitation depending on the pH, metal content and the amount of organic matter. Inorganic precipitates may probably become relevant for soils and particle-size separates rich in organically complexed metals such as Andisols and Spodosols.
|
IMPLICATIONS
|
|---|
At present, the knowledge on oxidative removal of soil organic matter is incomplete with respect to the effects of oxidants on soil minerals. Therefore, systematic studies on changes of mineral properties on oxidative removal of soil organic matter are required. These studies should involve pure minerals (poorly crystalline oxides, expandable clay minerals), defined organo–mineral complexes and soils with a wide range of properties (mineral composition, pH, organic matter content). Conclusions on the feasibility of a certain oxidant/protocol to remove organic C are possible only when several oxidants are comparatively applied to the same sample collective. When modifying oxidation protocols, mineral alterations should be elucidated by using combined methods like XRD, IR spectroscopy, gas adsorption (e.g., N2, CO2), and selective dissolution techniques.
Table 4 may serve as a guide for the use of reagents for organic matter removal. For textural analysis, H2O2 is adequate when the soil mineral phase is not dominated by mica, vermiculite or smectites. This recommendation must be used with caution since the impact of mineral disintegration on particle-size distribution has not yet been tested. Sodium hypochlorite is impractical for this purpose because of the formation of haloorganics that have to be disposed separately. Similarly, the large amounts of oxidant needed render the Na2S2O8 procedure unsuitable for organic matter removal before textural analysis. For organic C removal before sorption experiments and when surface properties of soils and minerals (SSA, CEC) are assessed, we recommend the use of NaOCl (pH 8, 25°C) instead of 30% (wt/wt) H2O2 since phyllosilicate disintegration and heat-induced transformations of minerals are kept to a minimum. For quantification of clay minerals, H2O2 should be avoided because of the transformation of vermiculites into mica-like minerals by NH+4 fixation. The use of Na2S2O8 is appropriate for clay preparation before XRD analysis since it seems not to alter the properties of reference clays. The application of NaOCl (pH 9.5) should be avoided for soils containing interlayered Al and Al hydroxides. Before further analysis, a test for the residues of organic C and the reagents used is recommended.
TOP
ABSTRACT
INTRODUCTION
