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


     

Published online 5 April 2007
Published in Soil Sci Soc Am J 71:711-719 (2007)
DOI: 10.2136/sssaj2006.0189
© 2007 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 HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Kaiser, K.
Right arrow Articles by Guggenberger, G.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Kaiser, K.
Right arrow Articles by Guggenberger, G.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Kaiser, K.
Right arrow Articles by Guggenberger, G.
Related Collections
Right arrow Soil Organic Matter
Right arrow Soil Surface Chemistry
Right arrow Biogeochemical Processes

SOIL CHEMISTRY

Increased Stability of Organic Matter Sorbed to Ferrihydrite and Goethite on Aging

K. Kaiser*, R. Mikutta and G. Guggenberger

Soil Sciences, Martin Luther Univ. Halle-Wittenberg, Weidenplan 14, 06108 Halle (Saale), Germany

* Corresponding author (klaus.kaiser{at}landw.uni-halle.de).


   ABSTRACT
TOP
 NOTES
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Sorption to micro- and mesoporous mineral phases can stabilize organic matter (OM) against microbial decay in soil. Formation of strong bonds that reduce desorbability is one plausible explanation for that effect. With time after sorption, sorbed OM may undergo changes in configuration or may migrate into intraparticle spaces. We tested the possible effects of residence time of OM sorbed to ferrihydrite and goethite. The minerals were loaded with different amounts of water-soluble OM from an Oa horizon, then stored moist (10% w/w water) for up to 1080 d at 4°C. We monitored the content of organic C, the desorbability and chemical stability (by extraction with 0.1 M NaOH–0.4 M NaF and treatment with 1 M NaOCl), and, after freeze-drying, the micro- and mesopore volume (by N2 and CO2 adsorption–desorption). Diffuse reflectance infrared Fourier transform (DRIFT) spectroscopy was used to characterize the OM on the mineral surfaces at the beginning and end of the experiment. There was no detectable decrease in sorbed organic C during the experiment; also, the micro- and mesoporosity of the samples remained unchanged. The proportion of desorbable organic C, however, decreased by up to 16%. This was paralleled by more pronounced bands indicative of complexed organic functional groups in the DRIFT spectra. We conclude that with increasing residence time, OM sorbed to porous minerals becomes decreasingly desorbable by the formation of additional chemical bonds to the surface via ligand exchange but not by diffusion into small pores. The decrease in desorbability was accompanied by a decrease in chemical destructibility with NaOCl. The stability of sorbed OM against biological degradation may similarly increase with residence time.

Abbreviations: DRIFT, diffuse reflectance infrared Fourier transform • OM, organic matter


   INTRODUCTION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Sorptive interactions with minerals stabilize organic matter (OM) efficiently against microbial degradation in soils (Nelson et al., 1994; Kalbitz et al., 2005). The stabilizing effect seems to be especially strong at micro- and mesoporous mineral surfaces (Mayer, 1994; Mayer and Xing, 2001). Evidence comes from the close relationship between concentrations of organic C and small pores (openings <10 nm) in natural samples (Kaiser et al., 2002; Mayer et al., 2004) as well as from laboratory tests with individual porous mineral phases. The latter showed OM to bind in small pores or at their mouths, thus reducing the micropore (<2 nm) and mesopore (2–50 nm) volume (and consequently the surface area) accessible for N2 by clogging (Mikutta et al., 2004; Filimonova et al., 2006). The decrease in small pores and surface area is most strong at smaller organic loadings, suggesting that small pores represent favorable binding sites for OM (Kaiser and Guggenberger, 2003; 2006). Stabilization of low-molecular-weight compounds such as citrate may result from sorption into pores small enough to prevent hydrolytic exoenzymes from entering (Zimmerman et al., 2004; Mikutta et al., 2006b). Although larger organic compounds cannot sorb into micropores, their openings allow strong multiple attachment, which hampers desorption of OM (Kaiser and Guggenberger, 2007)—a prerequisite for biological degradation (Jones and Edwards, 1998).

Much of the small pores and the surface area in soil is due to Al and Fe hydrous oxides (Borggaard, 1982; Eusterhues et al., 2005; Filimonova et al., 2006). Small pores, especially micropores, are typical features of multidomain minerals (e.g., goethite), located at the boundaries between single domains (Fischer et al., 1996). They may also result from aggregation of small crystals (e.g., ferrihydrite; Hofmann et al., 2004) or represent surface defects of crystals (Weidler et al., 1998).

Hydrous oxides of Al and Fe sorb OM strongly (e.g., Kaiser et al., 1997) and seem to account for much of the OM accumulation in acidic soils (e.g., Eusterhues et al., 2005). They stabilize sorbed organic compounds against chemical (Kaiser and Guggenberger, 2007) and biological degradation (Jones and Edwards, 1998), probably because of their porous nature (Kaiser and Guggenberger, 2007).

Sorption processes of solutes at the surfaces of microporous minerals often exhibit a rapid and a slow reaction. The former is assumed to represent the instant binding to external surfaces, while the latter is attributed to diffusion-controlled sorption into intraparticle spaces, i.e., small pores, that may proceed for days or even longer (Strauss et al., 1997; Axe and Trivedi, 2002). Similarly, sorption of OM to oxide surfaces is initially rapid, often followed by a much slower binding at later stages of the reaction (e.g., Avena and Koopal, 1999). Most natural organic compounds are of a size that probably does not allow penetration of micropores (Buffle et al., 1998; Kaiser and Guggenberger, 2007). Diffusion into intraparticle space may therefore be limited to mesopores. Also, polymers or polyelectrolytes adsorbed to surfaces are known to undergo changes in their configuration with time toward more ligands being directly attached to the surface, which results in reduced desorbability (e.g., Stuart, 1991). Similar mechanisms can be expected for the sorptive interactions of natural OM with minerals. Another explanation for the slow reaction are competitive reactions, with the more strongly binding compounds replacing initially bound less sorptive ones with time (e.g., Ochs et al., 1994). In situations with no competition for binding sites, i.e., in contact with solutions containing no or little OM, and negligible concentrations of strongly sorbing inorganic anions (e.g., phosphate), diffusion and changes in configuration or spatial rearrangement will control the slow reaction.

Our objective was to evaluate the reactions of OM sorbed to porous Fe hydrous oxides with increasing residence time, as potential controls of the desorbability as well as of the chemical and probably biological degradability. Potential diffusion into small pores was tested by the determination of the micropore and mesopore volume (by N2 and CO2 adsorption–desorption), and the bonds involved were analyzed by DRIFT spectroscopy. The desorbability of OM was determined by extraction with NaOH–NaF, and its chemical degradation by treatment with NaOCl.


   MATERIALS AND METHODS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Materials
Minerals
Well-crystallized, multidomain goethite ({alpha}-FeOOH) was prepared by slowly neutralizing an FeCl3 solution with NaOH and aging the precipitate at 55°C for 3 d (Atkinson et al., 1967). Two-line ferrihydrite was produced as outlined by Schwertmann and Cornell (1991), by neutralizing a 0.1 M FeCl3 solution with NaOH. After thorough dialysis against deionized water (until the electrical conductivity was <50 µS cm–1), minerals were freeze-dried, and then passed through a 0.63-mm sieve. X-ray diffraction (D5000/5005, Siemens AG/Bruker AXS, Karlsruhe, Germany) showed that the goethite was pure and crystalline. Ferrihydrite exhibited solely reflexes indicative for the two-line type. The absolute densities of the minerals as determined by He pycnometry (Accupyc 1330, Micromeritics Instrument Corp., Norcross, GA) were 3.8 g cm–3 for the ferrihydrite and 4.2 g cm–3 for the goethite.

Organic Matter Solutions
Water-soluble OM was obtained from field-fresh samples of the Oa horizon of a Typic Haplorthod. The organic material was sieved field fresh to <2 mm to remove coarse fragments and stored frozen at –18°C before extraction. Extraction was performed by adding 2 L of deionized water to 500 g of organic material. After 15 min of stirring, the suspension was allowed to settle for 18 h at room temperature, then passed through a 0.7-µm pore-size glass fiber filter (GF/F, Whatman International Ltd., Maidstone, UK), and finally through a 0.1-µm pore-size polyethersulfone membrane filter (Supor-100, Pall Gelman Science, Ann Arbor, MI). The filtrate contained 175 mg organic C L–1 (determined by Pt-catalyzed high-temperature combustion; TOC-5050, Shimadzu Corp., Tokyo) and the pH was 4.0. We prepared 12 solutions of 3.8 to 175 mg organic C L–1 by diluting the extract with a solution of similar inorganic composition and the same pH (4.0) as the filtrate, and a solution containing no OM. The solution used to dilute the OM extract and the OM-free solution was made up from salts and acids, containing all cations and all inorganic anions except for Cl–, which was used to compensate for organic anions. The major cations in all solutions were NH4+ (0.50 mmol L–1) and K+ (0.27 mmol L–1); the major inorganic anions were SO42– (0.37 mmol L–1) and H2PO4– (0.09 mmol L–1). The carboxyl acidity of the OM was 169 nmol mol–1 C and the absolute density 1.4 g cm–3. For further details on the elemental composition and acidity of the OM see Kaiser and Guggenberger (2007).

Experiments
Sorption
Sorption experiments with ferrihydrite and goethite were performed at solid/solution ratios of 1:2000 and 1:4000 (g dry wt. mL–1) with the 12 initial OM solutions. The ratio of 1:4000 for the ferrihydrite was necessary to achieve maximum sorption. Additionally, the minerals were equilibrated with a solution containing no OM but of similar inorganic composition and pH as the OM solutions. After 8 h of shaking at 5°C, the suspensions were centrifuged at 2000 x g for 30 min. The supernatant was removed, filtered through 0.1-µm pore-size membranes (Supor-100), and analyzed for organic C (TOC-5050 analyzer), pH, Fe (by atomic absorption spectroscopy; AA-400, Varian Inc., Palo Alto, CA), K+, NH4+, Ca2+, Mg2+, and inorganic anions (by ion chromatography; DX-100, Dionex Corp., Sunnyvale, CA). The settled material was rinsed with 10 mL of a solution of similar inorganic composition as the initial solutions but without OM to remove unreacted OM, then either stored moist (water content 10% w/w) at 4°C or freeze-dried. The experiments were performed in triplicate. The variation in the organic C concentration of the equilibrium solutions between replicates was <3%. The amount of organic C sorbed by the minerals was calculated by difference between the concentration of organic C in solution before and after contact with minerals. The calculated amounts of C sorbed were verified by analyzing the C content of freeze-dried samples (elemental analyzer Vario EL, Elementar Analysensysteme GmbH, Hanau, Germany).

Aging
Triplicate portions of selected samples were stored moist at 4°C under ambient atmosphere conditions for either 5, 42, 180, 360, 720, or 1080 d in loosely capped glass vials, to ensure aerated conditions throughout the experimental period. The ferrihydrite samples had 0, 0.5, and 1.1 mg of sorbed C m–2 initial surface area, which amounted to 0, 105, and 182 g C kg–1 (dry wt.). The goethite samples had 0, 0.9 and 1.9 mg sorbed C m–2, thus 0, 58, and 109 g C kg–1. The water content (10% w/w) was monitored biweekly (by weighing) and adjusted, if necessary, with deionized water to maintain a solution of similar inorganic composition as the OM solution used for the sorption. Before aging, portions of the three replicates of each of the samples were combined and well mixed to ensure similar initial conditions. At the end of the storage period, samples were immediately subjected to desorption and NaOCl treatment (see below), or freeze-dried for analysis of C (Vario EL) and porosity (see below).

Desorption
We extracted sorbed OM from all (moist) mineral samples by 0.1 M NaOH–0.4 M NaF at a solid/solution ratio of 1:200 (g dry wt. mL–1). After 24 h of gentle shaking at 5°C, the suspension was centrifuged at 2000 x g for 30 min, the supernatant filtered through 0.1-µm membranes (Supor-100), and analyzed for desorbed organic C (TOC-5050). Subsamples of the air-dried, settled material were additionally analyzed for C (Vario EL), showing that the two methods matched each other reasonably well (<10% difference). The reagent is a strong extractant of OM because of its capability of releasing even strong chemical bonds between OM and mineral constituents due to a high pH and displacement of complexed organic functionalities for OH– and F–.

NaOCl Treatment
To evaluate the stability of sorbed OM against chemical oxidation, the test materials were treated at 25°C with 1 M NaOCl (adjusted to pH 7.0 with HCl) at a solid/solution ratio of 1:50 (g dry wt. mL–1). After 6 h, the samples were centrifuged at 2000 x g for 30 min and the supernatants were removed. The treatment was repeated five times. After the last treatment, the centrifugation pellets were thoroughly washed with deionized water to remove salts (electrical conductivity <50 µS cm–1), freeze-dried, and analyzed for C (Vario EL). As a control, minerals without sorbed OM were also treated as described above. The NaOCl treatment for chemical degradation of OM was chosen because of its lack of effect on Fe oxides (e.g., Kaiser and Guggenberger, 2003; Mikutta et al., 2005).

Analytical Methods
Specific Surface Area and Porosity
The specific surface area (SSA) and porosity of freeze-dried minerals with and without sorbed OM were analyzed by 60-step adsorption–desorption of N2 at 77 K in the partial pressure range of 0.05 to 0.95 (Sing, 2001) with a Micromeritics ASAP 2010 surface area analyzer. Enough material was used for measurement to ensure a total surface area of >5 m2. We estimated the SSA by applying the Brunnauer–Emmett–Teller equation to the N2 sorption data obtained at partial pressures <0.3. Before analysis, water was removed from the samples by keeping them for 48 h at 20°C under a continuous stream of He (Kaiser and Guggenberger, 2003). The micropore (<2 nm) volumes were analyzed by the Dubinin–Astakhov method (Ravikovitch and Neimark, 2001). The estimation of the mesopore (2–50 nm) volume was done according to the Barrett–Joyner–Halenda method, using N2 desorption data obtained at partial pressures of 0.30 to 0.95 and assuming slit-type pore geometry (Reichenauer and Scherer, 2001).

We also determined the micropore volume of all samples by using CO2 as an adsorbate at 273 K in the partial pressure range of 0.001 to 0.035 and applying the Dubinin–Astakhov method. Carbon dioxide at 273 K, in contrast to N2 at 77 K, is assumed to pass through flexible OM and thus to reach mineral pores rendered inaccessible to N2 due to clogging by OM (de Jonge and Mittelmeijer-Hazeleger, 1996; Ravikovitch et al., 2005). Unlike N2, CO2 may also fill narrower pores (i.e., less than ~0.5 nm) due to its larger kinetic energy at 273 K (de Jonge and Mittelmeijer-Hazeleger, 1996; Echeverría et al., 1999). The CO2–based determination of volumes of mineral micropores clogged with OM may be biased, however, by the high affinity of CO2 for OM (Ravikovitch et al., 2005).

Mercury porosimetry was applied to freeze-dried ferrihydrite and goethite without and with sorbed OM. Analyses were performed with a Micromeritics AutoPore 9410 at pressures up to 412 MPa, using sample quantities of 200 mg. Porosity was calculated according the method of Washburn (1921). In samples with no OM clogging or filling pores, mesopore volumes derived from N2 desorption and from Hg porosity should fairly match each other, while in organic-rich materials, where the accessibility of pores to N2 is limited, Hg is assumed to fill these pores by deforming the pore-blocking or filling OM (Echeverría et al., 1999). Samples from the aging experiment were not analyzed by Hg porosimetry because of high cost and environmental concerns.

Diffuse Reflectance Infrared Fourier
Transform Spectroscopy
The DRIFT spectra of freeze-dried samples of minerals with and without sorbed OM before and at the end of the aging period were recorded with an IFS 66v spectrophotometer (Bruker GmbH, Karlsruhe, Germany), at a spectral range of 1000 to 4000 cm–1. Spectra obtained from 256 scans were corrected for water, atmospheric CO2, and the baseline drift. Difference spectra of minerals with and without sorbed OM were calculated to provide spectra of the sorbed OM. Identification of the DRIFT bands of the OM was done according to Niemeyer et al. (1992). Changes in the degree of complexation of carboxyl groups were determined by relating the intensity of bands of protonated carboxyl groups (at ~1715 cm–1) and carboxylate groups (i.e., dissociated carboxyl groups; at ~1605 cm–1) to that of complexed carboxylate groups (i.e., groups involved in complexes with metals; at ~1400 cm–1). Therefore, the spectra were scaled to the size of the carboxylate band. The ratios presented thus indicate the increase or decrease in intensity of the other bands relative to that of carboxylate.

Scanning Electron Microscopy
Scanning electron imaging of minerals with and without sorbed OM before and at the end of the aging experiment was done using a high-resolution microscope equipped with a field-emission cathode (LEO 1530, LEO Electron Microscopy Ltd., Cambridge, UK). Freeze-dried subsamples were spread out on an adhesive tape fixed to the sample holder, then sputtered with Pt. During imaging under high-vacuum conditions (<10–8 hPa), the tension voltage was limited to 10 kV to avoid charging-induced movement of the particles.

Statistical Evaluation
All experiments were performed in triplicate. Data are presented as means of replicates, with standard errors as direct indicators of variability, thus as a measure for the significance of differences. Temporal trends where not analyzed because of the interdependences among samples along time series.


   RESULTS AND DISCUSSION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
CONCLUSIONS
 REFERENCES
 
Properties of Test Minerals
The minerals used exhibited surface areas (Table 1) in the range typical for synthetic ferrihydrite and goethite (e.g., Schwertmann and Cornell, 1991). The large surface area of ferrihydrite suggests the predominance of small particles <10 nm, which is consistent with the literature (Schwertmann and Cornell, 1991) and the results from scanning electron imaging (Fig. 1 ). Much of the surface area is due to micropores; however, most of it derives from mesopores, which possibly are a result of aggregation (Hofmann et al., 2004). The goethite consisted of multidomain crystals of different sizes, with the domains being small and closely packed (Fig. 1). Consequently, the surface area is mainly due to the large micropore volume, which equals that of ferrihydrite (Table 1). The mesopore volume, in contrast, was smaller than that of the micropores. We explain this with the little aggregation and narrow pores at the boundaries of domains. The micropore volumes obtained from N2 and CO2 sorption (Table 1) for goethite as well as for ferrihydrite match each other well, suggesting little contribution of pores less than ~0.5 nm (Ravikovitch et al., 2005). For both minerals, the mesopore volumes obtained either by N2 desorption or by Hg porosimetry were largely similar, which is in agreement with the findings for OM-poor soil materials (Echeverría et al., 1999).


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

  		Table 1. Properties of the test minerals (without sorbed organic matter).

 

Figure 1
View larger version (75K):
[in this window]
[in a new window]

  		Fig. 1. High-resolution scanning electron microscope images of the ferrihydrite and goethite samples. Images result from detection of secondary electrons (Everhart–Thornley detector/in-lens detector).

 
Sorption of Organic Matter and Resulting Changes in Porosity
Both minerals sorbed organic C strongly, with rather steep slopes at smaller additions and leveling off at the largest ones, indicating distinct maximums for the sorption of OM (Fig. 2 ). This is consistent with previous findings on the sorption of OM by ferrihydrite and goethite (e.g., Tipping, 1981; Chorover and Amistadi, 2001). On a mass basis, the sorption of OM to ferrihydrite (maximum 318 g C kg–1) clearly exceed that of goethite (maximum 138 g C kg–1), whereas sorption to goethite was stronger when normalized to the initial surface area. This in agreement with previous findings on the sorption of OM to hydrous Fe oxides (e.g., Tipping, 1981). The achieved maximum surface loadings (ferrihydrite: 1.1 mg C m–2; goethite: 1.9 mg C m–2; Fig. 1) are at the upper end of the range of reported values (Tipping, 1981; Chorover and Amistadi, 2001). The large surface loadings can be explained by the high reactivity of the used OM (carboxyl acidity: 169 mmol mol–1 C) and the microporous character of the two minerals tested. The obtained maximum surface loadings match well the estimated maximum surface loadings of Al and Fe hydrous phases in acidic subsoils (1.1 mg C m–2; Eusterhues et al., 2005). Assuming no change in density of both minerals and OM during the sorption, the volume ratios of OM to mineral phase at maximum sorption were 2.0 for the ferrihydrite and 0.9 for the goethite system. The organic component in the produced organic–mineral associations, at least for those with the larger OM load, were fairly great if not dominant.


Figure 2
View larger version (15K):
[in this window]
[in a new window]

  		Fig. 2. Sorption of organic matter to ferrihydrite and goethite. The sorption is given as the relation between organic C (OC) sorbed and the added mass of OC, both normalized to the initial surface area of the minerals. The error bars (standard errors for each three replicate treatments) were smaller than the symbols. Arrows indicate the samples selected for the aging experiment.

 
As observed in recent studies (Mikutta et al., 2004; Kaiser and Guggenberger, 2007), the sorption of OM to the minerals resulted in a strong decrease in the N2–accessible micro- and mesopore volume of the minerals (Fig. 3 ). The decrease of the micropore volume was strongest at smaller additions and leveled off at larger additions of OM. In the case of ferrihydrite, the sorption of OM resulted in an almost complete loss of N2–accessible micropores. The stronger response of micropores than of mesopores to increasing loads of sorbed OM indicates preferential sorption of OM in, but more likely at the openings of, micropores, thus clogging and rendering them inaccessible to N2 (Kaiser and Guggenberger, 2003; Mikutta et al., 2004).


Figure 3
View larger version (22K):
[in this window]
[in a new window]

  		Fig. 3. Response of micropore (<2 nm) and mesopore (2–50 nm) volumes accessible to N2, CO2 (only micropores), or Hg (only mesopores) to increasing amounts of organic matter (OM) sorbed to ferrihydrite and goethite. The sorbed OM is given as organic C. For the mineral–OM complexes, the pore volumes are normalized to minerals therein. Bars are standard errors; n = 3.

 
The CO2–accessible micropore volume of both minerals was little affected by the sorption of OM (Fig. 3). This supports the assumption that CO2 at 273 K is capable of filling micropores clogged by OM and thus inaccessible to N2 at 77 K (de Jonge and Mittelmeijer-Hazeleger, 1996; Ravikovitch et al., 2005). The unchanged volumes of CO2–accessible micropores suggest that sorption of OM did not result in a major dispersion of ferrihydrite aggregates and goethite crystals and little CO2 was sorbed to mineral-associated OM across a wide range of OC contents (Fig. 3). The CO2–micropore volume, however, slightly increased at the largest OM loadings, reflecting the uptake of CO2 by sorbed OM (Ravikovitch et al., 2005). The existence of CO2–accessible molecular-size pores within OM has been suggested by de Jonge and Mittelmeijer-Hazeleger (1996). Since the CO2–micropore volume did not respond to small to moderate sorption of OM, we conclude that the extra micropore volume at the larger OM loadings was derived from narrow pores within organic aggregates at the mineral's surfaces (Kaiser and Guggenberger, 2007).

Similar to the CO2–micropore volume, the mesopore volume determined with Hg porosimetry was not affected by increasing loadings of OM (Fig. 3). That means the original porosity of the test minerals was not affected by the sorption of OM but just rendered inaccessible to N2. The Hg-porosimetry-derived mesopore volumes independent of OM loadings supports the assumption that Hg under high pressure deforms OM, thus filling pores inaccessible to N2 due to clogging with organic compounds (Echeverría et al., 1999).

Aging of Organic Matter–Mineral Complexes: Organic Carbon and Sample Structure
Filtration of the OM <0.1 µm before contact with the minerals should have removed most of the microorganisms. Glassware was not sterilized however, and thus the experiments were not under strictly sterile conditions. In the dissolved state, water-soluble OM from similar Oa horizons as the one used here showed some mineralization (~30% of organic C) during longer periods (up to 365 d, Kalbitz et al., 2003, 2005). Despite that, the concentrations of organic C of the four samples remained constant (Fig. 4 ), indicating insignificant mineralization during the experiment. This matches well the view of stabilization of OM against biodegradation on sorptive interactions with minerals (Jones and Edwards, 1998; Kalbitz et al., 2005). The mineralization of, e.g., citrate dropped almost to zero when sorbed to Fe hydroxide (Jones and Edwards, 1998). Scanning electron imaging revealed no detectable changes in the morphology of samples after the 1080-d aging compared with samples analyzed before the experiments (not shown).


Figure 4
View larger version (15K):
[in this window]
[in a new window]

  		Fig. 4. Concentrations of organic C (OC) in ferrihydrite and goethite samples, each with two different initial loadings of sorbed organic matter (OM), during the 1080-d aging. The initial OM loadings are given as OC concentrations in the mineral–OM complexes. Bars are standard errors; n = 3.

 
Aging of Minerals and Organic Matter–Mineral Complexes: Porosity
Aging of ferrihydrite and goethite without sorbed OM did not alter their N2–accessible mesopore volume (Fig. 5 ), indicating that their state of aggregation or dispersion was not affected. The unchanged N2– and CO2–micropore volumes of goethite implies no major changes in the crystallinity of the mineral, whereas the N2– as well as CO2–micropore volume of ferrihydrite decreased slightly (by ~10%) with time. This can be attributed to a gradual increase in crystal order, which typically results in larger crystals (Schwertmann and Cornell, 1991). Due to the limited amounts of sample, we were not able to justify that assumption by x-ray diffraction.


Figure 5
View larger version (16K):
[in this window]
[in a new window]

  		Fig. 5. Micropore (<2 nm) and mesopore (2–50 nm) volumes accessible to either N2 or CO2 (only micropores) of ferrihydrite and goethite without sorbed organic matter during the 1080-d aging. Bars are standard errors; n = 3. Horizontal lines indicate the initial micropore and mesopore volumes as determined by N2 sorption.

 
Aging of organic complexes of ferrihydrite and goethite for 1080 d did not alter their N2–accessible porosity (Fig. 6 ), irrespective of the amount of sorbed OM. This is not surprising given the strong reduction in micro- and mesopore volume on sorption of OM and the negligible mineralization of organic C. For the smaller OM loadings, where a portion of micro- and mesopores remained accessible to N2, and thus not clogged by OM, this result suggests no diffusion or migration of OM into smaller pores. That probably is due to preferential sorption, i.e., the concentration of OM at binding sites located in or at the openings of micropores (Mikutta et al., 2004; Kaiser and Guggenberger, 2007). Once bound there, OM does not seem to migrate to other sites.


Figure 6
View larger version (30K):
[in this window]
[in a new window]

  		Fig. 6. Micropore (<2 nm) and mesopore (2–50 nm) volumes accessible to either N2 or CO2 (only micropores) of ferrihydrite and goethite, each with two different initial loadings of sorbed organic matter (OM), during the 1080-d aging. The initial OM loadings are given as organic C (OC) concentrations in the mineral–OM complexes; the pore volumes are normalized to minerals within the mineral–OM complexes. Bars are standard errors; n = 3. Horizontal lines indicate the initial micropore and mesopore volumes as determined by N2 sorption.

 
The unchanged N2– and CO2–micropore volumes of the OM–ferrihydrite complexes are in contrast to the slight decrease in micropore volumes observed for the pure ferrihydrite (Fig. 5). Probably, the sorbed OM prevented or retarded the crystallization of the ferrihydrite. This is in agreement with the findings that sorbed OM hinders or delays the transformation of ferrihydrite into more crystalline forms (Schwertmann and Cornell, 1991).

Aging of Organic Matter–Mineral Complexes: Extractability of Organic Carbon
Although diffusion or migration of sorbed OM into narrower pore spaces with time is not supported by the unchanged pore volumes (Fig. 6), the portion of sorbed organic C desorbable by NaOH–NaF decreased with time (Fig. 7 ). At the beginning of the aging period, the portion of desorbable C appeared larger for the larger surface loadings, suggesting a less tight binding of OM at the more crowded surfaces. The difference in the desorbability diminished at later stages of the experiment. We interpret both the general decrease in desorbability and the vanishing of the difference in desorption at differing OM loadings to be the result of a slow rearrangement of sorbed organic molecules toward configurations that allow a maximum of ligands to attach. That would mean the sorption of natural OM is analogous to that of synthetic polymers or polyelectrolytes, which form additional bonds to the sorbing surface with increasing residence time by spatial rearrangement and changes in configuration (e.g., Stuart, 1991).


Figure 7
View larger version (16K):
[in this window]
[in a new window]

  		Fig. 7. Changes in the portion of sorbed organic C (OC) extractable by 0.1 M NaOH–0.4 M NaF in samples of ferrihydrite and goethite, with two different initial loadings of sorbed organic matter (OM), during the 1080-d aging. The initial OM loadings are given as concentrations of OC in the mineral–OM complexes. Bars are standard errors; n = 3.

 
At the end of the experimental period, the desorbability of organic C from ferrihydrite and goethite was down to ~50% of the sorbed C (Fig. 7). Both minerals seem equally well suited to preserve sorbed OM. The poor extractability of mineral-associated OM in many acidic subsoils by alkaline reagents may thus result from close interaction with both poorly ordered and crystalline Fe (and Al) hydrous oxides. Long-term residence of OM at the surfaces of such minerals seemingly diminishes the efficiency of alkaline extraction.

Aging of Organic Matter–Mineral Complexes: Surface Complexation
Carboxyl groups are the most important functionalities involved in the sorption of OM to hydrous oxides (e.g., Gu et al., 1994). Although binding of dissociated, thus negatively charged, carboxylate groups to positively charged sites at the mineral surface (anion exchange) is possible, the predominate mode of interaction is the formation of surface complexes where the carboxyl group replaces a surface hydroxyl group (ligand exchange; e.g., Gu et al., 1994). Sorbed OM, when compared with the OM before contact with the minerals, exhibits a narrowing in the infrared (IR) band intensity ratios of complexed carboxylate to carboxylate groups (Table 2). That indicates an increase in the proportion of complexed carboxylate groups on sorption. In turn, the ratios of carboxyl to carboxylate groups drastically decreased, which means that protonated carboxyl groups were not abundant in sorbed OM. For both test minerals, the IR band intensity ratios of complexed carboxylate to carboxylate groups was smaller for the larger than for the smaller OM loading while the IR band intensity ratios of carboxyl to carboxylate groups was larger. This suggests that fewer carboxyl groups are involved in surface complexes at more populated surfaces and more of them are left unreacted. These results are in full agreement with previous results of IR studies on the sorptive interaction of OM with hydrous oxides (Gu et al., 1994; Kaiser et al., 1997; Chorover and Amistadi, 2001).


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

  		Table 2. Ratios of intensity of carboxyl (1715 cm–1) and complexed carboxylate (1400 cm–1) bands to that of the carboxylate band (1605 cm–1) in diffuse reflectance infrared Fourier transform (DRIFT) spectra of organic matter (OM) sorbed to ferrihydrite and goethite, before and at the end of the 1080-d aging. The initial OM loadings of the tested samples are given as organic C concentrations of the mineral–OM complexes. The spectra of sorbed OM were calculated by the difference between the DRIFT spectra of minerals with and without sorbed OM. As a reference, the ratios for OM before contact with minerals are also shown. The data are the means of three replicates ± the standard error.

 
At the end of the aging period, for both minerals and both OM loadings, IR band intensity ratios of complexed carboxylate to carboxylate groups were larger and those of carboxyl to carboxylate groups were smaller than at the beginning (Table 2). It seems that with time, more (unreacted, protonated) carboxyl groups become involved in surface complexes, which is supported by the decrease in desorbability (Fig. 7). The increase in reacted carboxyl groups supports the idea of changes in the configuration of the sorbed OM with time. The reversibility of the sorption of OM to hydrous oxides thus is not only controlled by rapidly formed surface complexes but also by slow reactions.

Aging of Organic Matter–Mineral Complexes: Removal of Organic Matter by Sodium Hydrochloride
The portion of sorbed organic C removable by NaOCl decreased, similar to the extractability by NaOH–NaF, on aging of the OM–minerals complexes (Fig. 8 ), suggesting that sorptive stabilization can increase with the residence time of sorbed OM. The increased stabilization may be induced by the increase in strong bonds, thus the decrease in desorbability of OM; however, the decrease in desorbability was less pronounced than that in destructibility. Additional effects such as changes in the configuration, leading to a larger portion of functional groups being complexed and denser surface packing, may contribute to the differential response of desorbability and chemical stability to residence time.


Figure 8
View larger version (16K):
[in this window]
[in a new window]

  		Fig. 8. Changes in the portion of organic C (OC) removal by 0.1 M NaOCl (pH 7) in samples of ferrihydrite and goethite, each with two initial different loadings of sorbed organic matter (OM), during the 1080-d aging. The initial OM loadings are given as concentrations of OC in the mineral–OM complexes. Bars are standard errors; n = 3.

 
The portion of organic C removable by NaOCl was always smaller than that extractable by NaOH–NaF. This could mean that only weakly attached OM was prone to oxidative destruction, while tightly attached OM survived the chemical attack.

The NaOCl-induced removal of organic C sorbed to ferrihydrite at the smaller loading (0.5 mg C m–2) was larger than for the goethite at the smaller loading (0.9 mg C m–2) throughout the entire experimental period. This suggests a better stabilization of sorbed OM by goethite than by ferrihydrite. For the larger OM loading, however, the removal of organic C sorbed to goethite was much larger than for the smaller OM loading. In a previous study (Kaiser and Guggenberger, 2007), this was attributed to the presence of bulky organic agglomerates that form at large surface concentrations and disappear on NaOCl treatment. At smaller surface loadings where OM binds tightly to the mineral surface, NaOCl seemed less capable of removing organic C. The destructibility of OM sorbed to ferrihydrite at the larger loading (1.1 mg C m–2), although exceeding that for the smaller loadings, was less than for the goethite at the larger loading (1.9 mg C m–2). Another possible explanation is that, at larger OM loadings, no bulky organic agglomerations form but the sorbed OM is incorporated in closely packed ferrihydrite aggregates. Such aggregates have been identified in fine clay fractions of acidic subsoils by transmission electron microscopy (Mikutta et al., 2006a). Scanning electron microscopy, however, did not reveal any differences in the morphology of differently loaded ferrihydrite samples and no organic agglomerations were found (results not shown). The third possible explanation for the smaller chemical degradability of OM at smaller surface loadings is selective sorption of inherently more stable constituents. Sorption of OM to oxides is known to be strongly selective for the more acidic, aromatic constituents (e.g., Chorover and Amistadi, 2001). The selection for such strong binding moieties, however, should be strongest close to the sorption maximum, i.e., at the largest surface loadings. Therefore, we assume other factors than sorptive fractionation, such as the strength of bonds, to be of greater importance in the sorptive stabilization of OM.

The decrease in extractability and destructibility of OM sorbed to ferrihydrite and goethite related to each other but the relations were different for the two minerals and the loadings (Fig. 9 ). Destructibility, compared with desorbability, was greater for the larger OM loadings, especially in the case of goethite. We conclude that a less tight attachment of OM to the minerals leaves more OM vulnerable to the NaOCl attack. For goethite, the mentioned bulky organic agglomerations forming at larger loadings seem to be easily destructible (Kaiser and Guggenberger, 2007).


Figure 9
View larger version (19K):
[in this window]
[in a new window]

  		Fig. 9. Relation between the portions of organic C (OC) extractable by 0.1 M NaOH–0.4 M NaF and removable by 0.1 M NaOCl (pH 7) in samples of ferrihydrite and goethite, each with two initial different loadings of sorbed organic matter (OM), during the 1080-d aging. The initial OM loadings are given as concentrations of OC in the mineral–OM complexes. Bars are standard errors; n = 3.

 
These findings suggest that goethite, although stabilizing sorbed OM strongly at low to moderate organic loadings, is less effective for the accumulation of OM in soil than ferrihydrite, which can bind much more C per mass unit and stabilizes OM more efficiently at loadings close to the sorption maximum. Recent studies of acidic soils also pointed out the greater importance of poorly ordered than of crystalline mineral phases on both the concentrations and the residence time of organic C (Eusterhues et al., 2003, 2005; Mikutta et al., 2006a). Goethite (and probably other well-crystalline Fe oxides), however, probably controls OM accumulation and stabilization in soils poor in poorly ordered hydrous phases (e.g., Mikutta et al., 2006a).


   CONCLUSIONS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
REFERENCES
 
The results presented give rise to the following conclusions about the consequences of residence time of OM sorbed to the surfaces of ferrihydrite and goethite (and probably also other hydrous phases) in the soil environment.

Longer residence time increases the number of strong bonds between sorbed OM and mineral surfaces. The main driving factor for the increasing formation of strong bonds may be a rearrangement of organic molecules that allows additional organic functional groups to form complexes with surface metals and thus tighter attachment. As a direct consequence, the desorbability decreases and the stability against chemical, and probably also against biological, degradation increases.

Enhanced stability of sorbed OM by aging probably requires situations where the formed OM–mineral complexes are not exposed to further interactions with inorganic and organic solutes. Otherwise, exchange for more effectively binding organic substances (e.g., Ochs et al., 1994) will prevent the slow changes in the configuration necessary to allow more bonds per molecule. Also, highly competitive inorganic ions such as phosphate (Gu et al., 1994) may interfere in the aging process. Changes in flow pathways of water in soils, e.g., by formation of aggregates, can create the conditions necessary for long-term aging.

The stability of sorbed OM against chemical and probably biological degradation is related to, but not exclusively explainable by, the desorbability. Other factors such as the configuration of sorbed OM, which relates to the organic loading of sorbents, may additionally modify the extent of the sorptive stabilization.

Sorptive association with both ferrihydrite and goethite effectively stabilizes OM. The less crystalline ferrihydrite, however, has a much greater potential to contribute to the accumulation and stabilization of OM in soil than goethite (and probably other more crystalline hydrous oxides) because of its larger mass-normalized sorption capacity. Also, it may stabilize sorbed OM more efficiently at larger loadings, probably because of incorporation into ferrihydrite aggregates.


   ACKNOWLEDGMENTS
 
The study was funded by the Deutsche Forschungsgemeinschaft priority program 1090 "Soils as source and sink for CO2—mechanisms and regulation of organic matter stabilization in soils." Specific surface areas and porosity were determined in the Institute of Material Sciences (Univ. of Bayreuth). We are grateful to I. Otto, T. Schubert, R. Stallforth, and K. Urmann for their help with the measurements. The DRIFT spectra were recorded in the Institute of Inorganic Chemistry (Univ. of Bayreuth). The scanning electron images were recorded at the Institute for Experimental Physics II, Univ. of Bayreuth, under the guidance of C. Abetz. For comments and discussions we thank K. Eusterhues and K. Kalbitz. The aging experiment was set up following suggestions by G.W. Brümmer.


   NOTES
TOP
 NOTES
ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Permission for printing and for reprinting the material contained herein has been obtained by the publisher.

Received for publication May 16, 2006.


   REFERENCES
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 




This article has been cited by other articles:


Home page
 Vadose Zone JHome page
T. J. Reilly, N. S. Fishman, and A. L. Baehr
Effect of Grain-Coating Mineralogy on Nitrate and Sulfate Storage in the Unsaturated Zone
Vadose Zone J., February 10, 2009; 8(1): 75 - 85.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow 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 HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Kaiser, K.
Right arrow Articles by Guggenberger, G.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Kaiser, K.
Right arrow Articles by Guggenberger, G.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Kaiser, K.
Right arrow Articles by Guggenberger, G.
Related Collections
Right arrow Soil Organic Matter
Right arrow Soil Surface Chemistry
Right arrow Biogeochemical Processes

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