SSSAJ Grow Your Career with SSSA
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 27 October 2006
Published in Soil Sci Soc Am J 70:2049-2056 (2006)
DOI: 10.2136/sssaj2005.0366N
© 2006 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 Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Mikutta, R.
Right arrow Articles by Mikutta, C.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Mikutta, R.
Right arrow Articles by Mikutta, C.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Mikutta, R.
Right arrow Articles by Mikutta, C.
Related Collections
Right arrow Forest Soils
Right arrow Soil Organic Matter

Soil Chemistry-Note

Stabilization of Organic Matter at Micropores (<2 nm) in Acid Forest Subsoils

Robert Mikuttaa,* and Christian Mikuttab

a Institut für Bodenkunde und Pflanzenernährung, Martin Luther Universität Halle-Wittenberg, Weidenplan 14, D-06108 Halle, Germany
b Institut für Ökologie, Technische Universität Berlin, Salzufer 12, D-10587 Berlin, Germany

* Corresponding author (robert.mikutta{at}landw.uni-halle.de)


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 REFERENCES
 
Binding of soil organic matter (OM) at micropore entrances within small mesopores (2–10 nm) has been suggested as a potential mechanism for the stabilization of OM against biodegradation. We hypothesized that the mineral-associated fraction of stable OM [OM resisting treatment with 6% sodium hypochlorite (NaOCl) and subsequently extracted by 10% hydrofluoric acid] is associated with pores <10 nm in 12 acid subsoil horizons. To study the coverage of micropores by stable OM sorbed in mesopores, we assumed that most micropores have entrances of mesopore size. We compared the accessibility of CO2 at 273 K to micropores after NaOCl treatment with that of N2 at 77 K. In contrast to N2, diffusion of CO2 into micropores is little affected by OM and the difference in both micropore volumes (MIV) is taken as a measure of micropore clogging and thus of the association of stable OM with mesopores. The MIV measured by CO2 adsorption was corrected for CO2 sorption in OM (MIV-CO2corr). In 7 out of 12 samples, the MIV-CO2corr equaled the MIV-N2, suggesting that micropore entrances are not blocked by stable OM. For four samples the results were ambiguous, whereas in an Eutric Hapludand Bw horizon, a threefold larger MIV-CO2corr compared with the MIV-N2 indicates that stable OM was associated with pores < 10 nm. Based on the findings that (i) mineral MIVs derived from CO2 and N2 adsorption were similar for most samples after exposure to NaOCl, (ii) the change of MIV-N2 on NaOCl treatment was small in all samples, and (iii) no relationship existed of the MIV-CO2corr and small mesopore volume with the content of mineral-associated stable organic C (OC), we conclude that the association of OM with micropore entrances in small mesopores does not primarily control the stabilization of OM in these acid subsoils.

Abbreviations: MIV, micropore volume • OC, organic carbon • OM, organic matter • SSA, specific surface area


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 REFERENCES
 
MINERAL MATTER OF SOILS AND SEDIMENTS generally comprises micro- and mesoporosity (Mayer, 1994; De Jonge et al., 2000; Mayer et al., 2004). Micropores are pores with a diameter of <2 nm and mesopores have pore sizes ranging from 2 to 50 nm (Rouquerol et al., 1994). Pores with diameters <2 nm may be associated with crystal defects, the stacking edges of expandable 2:1 clay minerals, dead-ends of pores at domain boundaries of Fe (hydr)oxides (intraparticle microporosity) (Fischer et al., 1996; Aringhieri, 2004), but may also arise from interparticle spaces of microcrystalline minerals like Fe and Al (hydr)oxides (Cabrera et al., 1981; Cornell and Schwertmann, 1996) (interparticle microporosity). In contrast, mesopores might largely reflect the interparticle porosity of clay domains and mineral aggregates (Mayer et al., 2004).

At present, the importance of micro- and small mesopores for OM stabilization is unclear. Mayer (1994) introduced the concept by which OM contained in pores <10 nm is protected against microbial decay, because microorganisms and their extracellular enzymes are excluded due to their size. Micropore entrances are regarded as preferential sorption sites for OM (Kaiser and Guggenberger, 2003) while filling of micropores in soils is unlikely given the size constraints of OM (Buffle et al., 1998; Kaiser and Guggenberger, 2006). Common sizes of dissolved OM components, fulvic and humic acids were reported to fall into the range of 2 to 8000 nm depending on solution pH, ionic strength and ionic composition (Jones and Bryan, 1997; Myneni et al., 1999; Plaschke et al., 2000; Namjesnik-Dejanovic and Maurice, 2001; Tan, 2003; Alvarez-Puebla and Garrido, 2005; Kawahigashi et al., 2005). Sorption experiments with ill-defined dissolved OM and citrate added to soils and minerals show that there is no complete reduction in MIV and thus no complete filling of micropores (Kaiser and Guggenberger, 2003; Mikutta et al., 2006a). Consequently, a portion of the mineral MIV can be considered bare of OM. For this reason, Kaiser and Guggenberger (2003, 2006) suggested that organic polyelectrolytes sorb to the mouth of micropores via multiple site attachments, that is, in small mesopores (<10 nm), rather than fill micropores. Indeed, porosity measurements of Fe (hydr)oxides equilibrated with humic or non-humic substances (Mikutta et al., 2006a, 2006b) show that the porosity of >10-nm pores is far less reduced than that of pores <10 nm. Compared with nonporous mineral surfaces, mesoporous sorbents like SiO2 and Al2O3 exhibit larger surface-normalized sorption capacities of OC, and reduced desorption and mineralization rates for organic sorbates in batch experiments (Goyne et al., 2004, 2005; Zimmerman et al., 2004a, 2004b). In contrast, Mayer et al. (2004) showed that in marine sediments and soil Ap horizons only a small fraction of OM (<10–20%) was actually contained in small mesopores, suggesting that 2- to 10-nm pores are not significantly involved in the stabilization of OM.

In a previous experiment we isolated stable OM in 12 subsoil horizons (below-A horizons) by treatment with 6% NaOCl. This pool comprised 2.4 to 20.6 mg OC per gram soil and had a significant lower 14C content compared with untreated OM (Mikutta et al., 2006c). Up to 96% of this stable OC (73% on average) was protected against oxidation predominantly by association with ferrihydrite, Al (hydr)oxides, and crystalline Fe (hydr)oxides. Here we hypothesize that stable OM is protected against destruction by NaOCl because of its sorption to micropore entrances in small mesopores. The resistance of OM associated with small mineral pores against chemical attack can be assumed to derive from multiple bonds of organic polyelectrolytes at narrow pore walls that likely impede the oxidative breakdown and the desorption of OM (Kaiser and Guggenberger, 2006). To test our hypothesis, we conducted gas adsorption experiments with N2 at 77 K and CO2 at 273 K by using samples from the 12 acid subsoil horizons, which considerably differ in their mineral assemblage (Mikutta et al., 2006c). We used the accessibility of both sorbates for micropores pitting small mesopores as a measure of the spatial arrangement of OM at micropore entrances. We took advantage of the fact that at equilibration times commonly applied in gas adsorption studies (several minutes), diffusion of N2 at 77 K into micropores via passage of OM-clogged mesopores is kinetically restricted whereas the accessibility of CO2 to micropores is hardly affected by OM at 273 K. Despite similar molecular diameters (CO2: 2.8 Å; N2: 3.0 Å), CO2 is capable of penetrating mineral and organic pores < ~0.5 nm in diameter whereas N2 is not (Garrido et al., 1987; Pignatello, 1998; De Jonge et al., 2000; Ravikovitch et al., 2005). Consequently, if stable OM clogs small mesopores, a larger CO2 micropore volume (corrected for CO2 sorption in OM) compared with the N2 micropore volume will be observed. Our reasoning is based on several assumptions: (i) mineral micropores are not entirely filled with OM due to size and electrostatic restrictions, (ii) both gases access mineral micropores to the same extent, that is, the contribution of < ~0.5-nm pores to the total micropore volume of mineral matter is negligible, and (iii) in our sample pool most micropores have entrances of mesopore size. The validity of assumption (i) has been justified above. Assumption (ii) was validated by comparing the N2 and CO2 micropore volumes (MIV-N2, MIV-CO2) of an ensemble of microporous minerals commonly found in soils (ferrihydrite, bentonite, and vermiculite). While assumptions (i) and (ii) rely on experimental evidence, the validity of assumption (iii) is more difficult to ascertain because of lacking data on pore geometries in soils. For Fe and Al (hydr)oxides and nonparallel stacked sheets of layer silicates assumption (iii) seems appropriate. However, in case of parallel stacked sheets of layer silicates assumption (iii) is not valid. As most of the stable OC of our soil samples is associated with (hydr)oxides, a rather wedge-like pore geometry may be regarded as predominant in terms of OM stabilization. Ink-bottle pore geometry, which leads to a hysteresis loop of the H2 type according to IUPAC classification in N2 sorption experiments (Sing et al., 1985), was neither observed for calcareous soils (Echeverría et al., 1999) nor in our samples (Fig. 1 ).


Figure 1
View larger version (24K):
[in this window]
[in a new window]
  		Fig. 1. Nitrogen adsorption–desorption isotherms at 77 K of selected subsoil samples used in our study showing a range of observed hystereses. The adsorption–desorption isotherms of Samples 7 and 11 are representative whereas Sample 6 exhibits the largest hysteresis among the samples due to the prevalence of smectite. Note that the isotherms are stacked by a value of 20 cm3 g–1 for better visualization.

 

   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 REFERENCES
 
Twelve subsoil samples from forest sites varying in type and assemblage of minerals were used in our study (Table 1). Soil pH (KCl) ranged from 3.3 to 5.8. Detailed information about sites and sample properties are given in Kleber et al. (2005) and Mikutta et al. (2006c). Stable OM in the studied samples was determined by treatment with 6% NaOCl (pH 8) at 25 ± 1°C. A 10-g sample was treated three times with 100 mL of 6% NaOCl for a total of 18 h. Samples were centrifuged (2574 x g for 5 min) and the supernatant was decanted. The residues were washed twice using 100 mL 1 M NaCl and shaken overnight with deionized water before dialysis (Medicell International; 12–14 000 Daltons). When the electrical conductivity was <40 µS cm–1, the samples were freeze-dried. Mineral-associated stable OM was subsequently released by dissolution of minerals with 10% HF and quantified by mass balance calculations utilizing the sample weights before and after HF treatment, and the OC content after NaOCl treatment and those of the residuum left after HF treatment (Mikutta et al., 2006c).


View this table:
[in this window]
[in a new window]
  		Table 1. Selected properties of subsoil samples.

 
Crystalline Fe (hydr)oxides and poorly crystalline Fe and Al phases were extracted with Na-dithionite-citrate and acid oxalate, respectively (Blakemore et al., 1987). Element concentrations in extraction solutions were determined by inductively coupled plasma atomic-emission spectroscopy (ICP–AES; Jobin Yvon JY 70 plus). The cation exchange capacity (CEC) of silicates was estimated after treatment with NaOCl and Na-dithionite-citrate, and was corrected for the CEC of residual OM (Mikutta et al., 2006c).

As test minerals for microporosity measurements, we chose ferrihydrite, bentonite (S&B Industrial Minerals), and vermiculite (UBM, União Brasileira de Mineração S.A.). Ferrihydrite was produced by neutralization of a 0.1 M Fe(NO3)3 solution with 1 M KOH at pH 7 to 8 as described in Cornell and Schwertmann (1996). Bentonite and vermiculite were washed with doubly deionized water (pH 5) and 2% NaOCl (pH 8); then the clay fraction (<2 µm) was obtained by sedimentation in Atterberg cylinders. Thereafter, the clays were Na-saturated by washing with 1 M NaCl, dialyzed until the electrical conductivity was <40 µS cm–1, and freeze-dried. For transmission electron microscopy coupled with energy-dispersive X-ray spectroscopy (JOEL JEM 1210), we used the sonicated fine clay fraction (<0.2 µm) dried on SiO2–coated Cu-nets (200 mesh).

Gas adsorption and desorption was measured with a Nova 4200 analyzer (Quantachrome Corp., Boynton Beach, USA) using N2 and CO2 as adsorbates. Predried samples (24 h drying at 0.06 kPa) were weighed into sample cells and outgassed at 298 K and 10–4 kPa for 24 h at the instrument degas stations to avoid structural transformation of OM and poorly crystalline minerals, which may start at temperatures as low as 313 K (Kaiser and Guggenberger, 2003). We are aware that this is a compromise between efficient water removal and the necessity to prevent OM and mineral transformations. However, we are confident that the degassing procedure applied had no significant impact on measured MIVs due to residual water, because the MIV of bentonite derived from CO2 was much larger than that obtained by N2 adsorption (Fig. 2 ), indicating that water was successfully removed from small micropores < ~0.5 nm. Nitrogen adsorption and desorption isotherms were recorded at 77 K in the partial pressure region 0.01 to 0.995 P/P0 using 25 adsorption and 20 desorption points. The specific surface area (SSA) was determined from the relative pressure range of 0.05–0.3 by the BET equation (Brunauer et al., 1938). The mineral SSA (m2 g–1) was corrected for the weight of residual OM: SSA = SSAmeasured/[1– 2 x OC/1000)], where SSAmeasured refers to the measured SSA (m2 g–1) and 2 is a conversion factor to transform OC into OM contents (g kg–1). The small mesopore volume (MEV, mm3 g–1) was determined by the BJH model (Barrett et al., 1951) using the desorption leg of the isotherm and corrected for the weight of residual OM: MEV = MEVmeasured/[1– 2 x OC/1000)]. Carbon dioxide adsorption was performed at 273 K in an ice bath filled with 2/3 finely crushed ice and 1/3 water from 0.001 to 0.03 P/P0 using 25 adsorption points. For both adsorbates, MIVs were determined according to the Dubinin-Radushkevich equation (Gregg and Sing, 1982):

Formula 1[1]
where V is the volume of gas adsorbed per mass of sample (mm3 g–1) that is calculated from V = n{rho}–1, where n is the mass of gas adsorbed (g g–1) and {rho} is the liquid density of the gas used (g mm–3); V0 is the total MIV (mm3 g–1), and D is a constant related to the structure of the adsorbent and the adsorbate-adsorbent affinity. The MIV was obtained as the intercept in a plot of log(V) versus log2(P0/P) after extrapolation from the linear region in the Dubinin-Radushkevich plot. All analyses were performed at least in duplicate; therefore, the error of measurements is given as mean range. Regression analyses were performed with Statistica 5.0 (StatSoft Inc., Tulsa, OK).


Figure 2
View larger version (18K):
[in this window]
[in a new window]
  		Fig. 2. Micropore volumes of ferrihydrite, Na-bentonite, and Na-vermiculite determined by N2 adsorption at 77 K and CO2 adsorption at 273 K.

 

   RESULTS AND DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
REFERENCES
 
Figure 2 reveals that the MIVs of ferrihydrite and Na-vermiculite determined by CO2 and N2 were similar but the amount of sorbed N2 at 77 K slightly exceeded that of sorbed CO2 at 273 K. The same has been observed for N2– and CO2–derived MIVs of calcareous soils (Echeverría et al., 1999) but the reason for the difference in gas uptake is unclear. Probably it arises from reduced diffusion of CO2 into micropore voids. Compared with the less polarizable N2 molecules, CO2 has a large quadrupole moment (1.03–1.13 x 10–35 C [3.1 to 3.4 x 10–26 esu]) that may cause interactions with polar inorganic surfaces (Lemcoff and Sing, 1977). Under the experimental conditions chosen, there was no significant abundance of < ~0.5 nm-pores in Na-vermiculite and ferrihydrite. In contrast, Na-bentonite exhibited a significant fraction of micropores accessible to CO2 but not to N2, leading to a nearly twofold difference in MIVs (Fig. 2). Therefore, in our sample pool, evaluation of micropore clogging by association of OM with small mesopores by means of gas adsorption is only possible for samples that contain no or only minor amounts of smectite.

Micro- and Mesoporosity in Relation to Mineralogical Properties
To determine the mineral microporosity (MIV-CO2corr; mm3 g–1), we corrected the measured CO2 volumes for the amount of CO2 likely to sorb in stable OM according to

Formula 2[2]
where MIV-CO2 is the measured MIV (mm3 g–1), 2 is a conversion factor, OC is the content of stable OC (mg g–1), and 0.043 ± 0.022 corresponds to the average ± mean range of published CO2 volumes sorbed to various OM types including peat, humic acid, and materials collected from O layers of different soils (mm3 mg–1, De Jonge et al. (2000) and references given therein). Note, that the amount of CO2 that sorbs to OM was determined on freeze-dried samples and is thus probably not representative for mineral-bound OM in soils. Sorbed OM may be present in a dense configuration (Wang and Xing, 2005; Kaiser and Guggenberger, 2006), that is, the microporosity of sorbed OM can be expected to be at the lower end of the range given in De Jonge et al. (2000). In accordance, the CO2–MIVs of Fe (hydr)oxides are only insignificantly affected by OM up to C contents of ~120 g kg–1 (Klaus Kaiser, personal communication, 2006). For the 12 subsoil samples, the volume of CO2 sorbed to stable OM was estimated using Eq. [2] and was found to be between 0.1 to 2.6 mm3 per gram soil, representing 1 to 35% (11% on average) of the total MIV measured by CO2 adsorption (not shown). This suggests that in the subsoil samples, the microporosity was due predominantly to minerals rather than OM. The calculated MIV-CO2corr was positively correlated with the SSA (r2 = 0.78; P < 0.001), the content of dithionite-extractable minus oxalate-extractable Fe (Fed–o; r2 = 0.39; P < 0.05), and clay content (r2 = 0.36; P < 0.05) (n = 12). These correlations indicate that crystalline, clay-sized Fe (hydr)oxides merely explain 40% of the variability of the measured MIV of mineral matter. The microporosity associated with Fe (hydr)oxides likely arises from boundaries between individual crystal domains or from crystal defects (Fischer et al., 1996; Weidler et al., 1998). Noteworthy, the mineral microporosity did not relate to the CEC of the silicate matrix (P = 0.26) and to oxalate-extractable Fe plus Al (Feo+Alo; P = 0.90). Poor or lacking relations between mineral microporosity and the CEC of silicates, content of crystalline Fe (hydr)oxides, and poorly crystalline Fe and Al phases indicate that some microporosity must result from agglomerations of soil particles or aggregates (interparticle microporosity). Slightly better correlations were observed between the volume of small mesopores and SSA, dithionite-extractable minus oxalate-extractable Fe, and clay content (r2 = 0.95; P < 0.001; r2 = 0.81; P < 0.001; r2 = 0.78; P < 0.001, respectively), suggesting that substantial surface area of small mesopores is associated with crystalline Fe (hydr)oxides. According to the BJH model, the SSA contained in small mesopores <10 nm in the subsoil samples accounted for 42 ± 7% of the total SSA.

Association of Stable Organic Matter with Mineral Micro- and Small Mesopores
The MIV-CO2corr correlated positively with the MIV-N2 determined after NaOCl treatment (r2 = 0.97; P < 0.001; n = 10; samples 6 and 11 omitted as outliers). A plot of MIV-N2 versus MIV-CO2corr shows that almost all samples are grouped around the 1:1 line and that for five samples the observed MIV-CO2corr were smaller than the MIV-N2 (Fig. 3a ). For Samples 6 and 11, the MIV-CO2corr was significantly larger than the MIV-N2, suggesting the blockage of micropores by OM. However, when considering a possible underestimation of MIV-CO2corr by up to 17% due to the presence of vermiculite-type minerals as shown for the vermiculite standard (Fig. 2), Samples 7, 8, and 10 would also exhibit larger MIV-CO2corr than the respective MIV-N2. The results thus indicate that for the remaining seven samples, the accessibility of both sorbates to micropores was not restricted by OM, suggesting that the mineral-associated stable OM was not associated with micropore entrances. This means that small mesopores were not blocked by stable OM. Equal MIV measured by N2 and CO2 could also accrue from partial intrusion of chains of OM into micropores so that OM entirely impairs the micropore diffusion of CO2. However, Kwon and Pignatello (2005) demonstrated for a black carbon loaded with up to 35 wt.% soybean lipids that CO2 can pass the organic barrier in micropores at 273 K due to the flexible movement of lipid chains. The observation that micropores in the majority of samples were not blocked by stable OM is astonishing given the fact that micropore entrances represent preferential sorption sites for OM in laboratory experiments (Kaiser and Guggenberger, 2003, 2006).


Figure 3
View larger version (17K):
[in this window]
[in a new window]
  		Fig. 3. Nitrogen micropore volume (MIV-N2) after NaOCl treatment versus the CO2 micropore volume after NaOCl treatment corrected for CO2 sorption to stable OM (MIV-CO2corr) (a); plot of the N2–MIVs determined before and after NaOCl treatment (b).

 
Mayer et al. (2004) showed that in 10 of 21 marine sediments and soil Ap horizons, the volume of <8-nm pores was potentially large enough to hold half of the total OM. Their results imply that sediments and topsoils offer enough pore volume to host and eventually protect OM from degradation by exclusion of microbes and exoenzymes. Figure 4 and Table 2 show that the volume of mineral-associated stable OM is in the same order of magnitude as the MIV-CO2corr and small mesopore volume of mineral matter when assuming OM densities of 1.4 and 1.9 g cm–3. Using these densities for OM, on average 86 and 93% of mineral-associated stable OM in all samples could be contained in small mesopores, respectively. Thus, the calculations are in line with Mayer et al. (2004), implying that mineral pores < 10 nm in subsoil horizons theoretically have the potential to hold the majority of mineral-associated stable OM. Provided that stable OM was primarily located in small mesopores due to preferential sorption to micropore entrances, one would expect a correlation of MIV-CO2corr or small mesopore volume with the content of mineral-associated stable OC. However, due to smaller OM inputs, subsoil horizons have a minor portion of their SSA covered with OM when compared with topsoils (Kaiser and Guggenberger, 2003). Therefore, mineral surfaces and thus micro- and mesopores of subsoils are probably unsaturated with regard to OM accumulation, which could bias the correlation analyses. To account for such a disequilibrium, we excluded all samples where the amount of mineral-associated stable OC normalized to SSA was <0.3 mg OC m–2. We assumed that at OC loadings ≥0.3 mg C m–2 most mineral surfaces are near their saturation capacity as has been shown in sorption experiments with dissolved OM (Chorover and Amistadi, 2001; Kaiser, 2003). No statistical relations existed between MIV-CO2corr and small mesopore volume with the content of mineral-associated stable OC (P = 0.20 and P = 0.27, respectively; n = 7). This finding agrees with our gas adsorption data, suggesting that the majority of stable OM was not bound to micropore entrances. Additionally, this reasoning is supported by the observation that even in the non-oxidized samples, micropores were already accessible to N2 even though the potential volume of mineral-associated OM exceeded the MIV-CO2corr by an average factor of three (not shown). Here, we assumed an OC fraction of 10% that is not associated with minerals (Eusterhues et al., 2005), an OM density of 1.4 g cm–3, and an OC content of OM of 50 wt.% (not shown). The larger dimensions of OM than those of micropores suggests that OM could have entirely clogged mineral micropores by sorption in mesopores if micropores entrances were preferential binding sites and accessible for OM. But as the MIV-N2 largely equaled those of MIV-CO2corr and the change of N2–MIVs during NaOCl treatment was small (Fig. 3b), we conclude that already in the non-oxidized samples, OM was not predominantly associated with micropore entrances.


Figure 4
View larger version (26K):
[in this window]
[in a new window]
  		Fig. 4. Small mesopore volume (2–10 nm) corrected for the weight of residual OM after NaOCl treatment versus the average volume of the mineral-associated stable OM assuming a minimal and maximal OM density of 1.4 and 1.9 g cm–3 (vertical error bars). The amount of mineral-associated stable OM was determined as the fraction that was released by 10% HF following the NaOCl treatment.

 

View this table:
[in this window]
[in a new window]
  		Table 2. Content of initial, stable, and mineral-associated stable organic carbon (OC); estimated volumes of mineral-associated stable OM assuming an OC content of OM of 50 wt.% and a density of 1.4 and 1.9 g cm–3, respectively; micropore volumes (MIV) present after NaOCl treatment as determined with N2 and CO2 adsorption at 77 and 273 K, respectively, and small mesopore volumes (2–10 nm) after NaOCl treatment. All errors in parentheses denote mean range of duplicate measurements.

 
Based on the dominance of smectite in the clay fraction of the AB horizon of a Typic Hapludoll (Sample 6), the twofold larger CO2 uptake compared with N2 after NaOCl treatment can be ascribed to the abundance of CO2–accessible micropores < ~0.5 nm. Based on the potential experimental error, only for the Bw horizon of the Eutric Hapludand (Sample 11) we can be sure that stable OM covered the micropore entrances as indicated by a threefold larger MIV-CO2corr compared with MIV-N2. The sample was dominated by hydroxy-interlayered vermiculite and kaolinite, and contained the largest amount of oxalate-extractable Al among all samples (Table 1). The stable OC of this sample was found to be located in nanometer-sized aggregates and related to the concentration of Al, Si, and Fe (Fig. 5 ); its MIV-N2 changed little on NaOCl treatment, implying that OM clogging the micropores was not removed by NaOCl (Fig. 3b). The finding thus supports the view that OM associated with micropore entrances is stabilized against desorption and oxidation (Kaiser and Guggenberger, 2003). Assuming that OM was not present in micropore voids, the difference in gas uptake in this Bw horizon must be ascribed to the incorporation of stable OM within small mesopores.


Figure 5
View larger version (87K):
[in this window]
[in a new window]
  		Fig. 5. Transmission electron images of a small aggregate in the fine clay fraction (<0.2 µm) of the Eutric Hapludand (Sample 11) after treatment with NaOCl. The A-B line represents the spatial distribution of Si, Al, Fe, and stable OC as analyzed by energy-dispersive X-ray spectroscopy.

 
Based on our premises, equal CO2corr– and N2–MIVs may result from two scenarios: stable OM is only partly or not associated with micropore entrances. Both scenarios cannot be distinguished by means of gas adsorption. This would only be possible by comparison of the mesopore volume of OM-coated samples with that of samples where the OM has been removed. Unfortunately, oxidative treatments (wet oxidation, muffling) are supposed to affect the pore structure of subsoil samples due to the removal of organic coatings and the dehydroxylation of poorly crystalline minerals. Apart from this methodological limitation, it is not reasonable to assume that all mesopores, especially those of <10-nm size, are free of stable OM given the association of stable OC within nanometer-sized aggregates (Fig. 5; Mikutta et al., 2006c). It seems more appropriate to assume that stable OM partly filled mesopores to varying extents depending on the sample properties such as the type of minerals present, and the content and chemical composition of OM.

In summary, in one sample we found evidence for micropore blockage by OM whereas three other samples are likely candidates when the potential underestimation of sorbed CO2 is accounted for. For another sample, micropore blockage could not be examined due to the large content of smectite clay. Based on the findings that for seven samples (i) the MIVs determined by N2 and CO2 after NaOCl treatment were similar, (ii) micropores in untreated samples, even in those that contained large amounts of OM, were accessible to N2, and (iii) no correlation between both, MIV-CO2corr and small mesopore volume with the content of mineral-associated stable OC existed, we conclude that stable OM in these subsoil horizons appears not to reside at micropore entrances in small mesopores. Thus, we have to reject our hypothesis that the association of stable OM with micropore entrances was the primary cause for the resistance of OM during NaOCl treatment. Our data, however, do not generally rule out a stabilizing effect of mineral micropores for associated OM. Yet, the results are in line with Mayer et al. (2004) stating that other factors than association with small mineral pores control OM stabilization. Several causes might account for the discrepancy between results obtained from batch experiments, showing that micropores are preferential sorption sites of OM, and analyses of the in situ association of stable OM with mineral matter in acid subsoil samples. Due to aggregation of mineral matter, a large fraction of micro- and mesopores in subsoil horizons is probably inaccessible to dissolved OM, microorganisms and their metabolites but accessible to CO2 and N2 during porosity measurements. Also, a substantial fraction of mineral surfaces in subsoils might not have had contact with OM, either due to small OM inputs or due to the movement of dissolved OM along preferential flow paths. Sorption processes in upper horizons may also cause chemical fractionation of OM and thus leave a fraction behind that is probably less capable to interact with porous mineral surfaces in subsoil horizons. We are aware that the conclusions drawn from N2 and CO2 sorption data rely on our premises (different accessibilities of adsorbates for organic micropores, negligible initial filling of micropores by OM, prevalence of wedge-like pore geometry, HF-soluble portion of NaOCl-resistant OM as a measure of mineral-associated stable OM). Here, more research is needed to further quantify the effect of mineral micro- and mesopores on the stabilization of OM in subsoil horizons. Specifically, we would like to encourage further research on (i) the quantification of stable OC either free or attached to mineral surfaces, (ii) the determination of densities of mineral-associated OM for precise volume calculations, and (iii) the evaluation of predominant pore geometries in soils, for example, by gas adsorption studies and advanced microscopy techniques.


   ACKNOWLEDGMENTS
 
This study was funded by the German Research Fund (DFG) as part of the priority program SPP. 1090 "Soils as Source and Sink for CO2–Mechanisms and Regulation of Organic Matter Stabilization in Soils". We thank Maren Kahle (Swiss Federal Institute of Aquatic Science and Technology) for providing the ferrihydrite sample, and Markus Kleber and Klaus Kaiser for valuable comments on the early draft of the manuscript. We also gratefully acknowledge the helpful comments of three anonymous reviewers.

Received for publication November 7, 2005.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 REFERENCES
 





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 Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Mikutta, R.
Right arrow Articles by Mikutta, C.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Mikutta, R.
Right arrow Articles by Mikutta, C.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Mikutta, R.
Right arrow Articles by Mikutta, C.
Related Collections
Right arrow Forest Soils
Right arrow Soil Organic Matter

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
Journal of Plant Registrations Journal of
Environmental Quality		 The Plant Genome