Article pubs.acs.org/JPCA
Thermal Stability of Goethite-Bound Natural Organic Matter Is Impacted by Carbon Loading Wenting Feng,*,†,§ Jonatan Klaminder,†,∥ and Jean-François Boily‡,⊥ †
Department of Ecology and Environmental Science and ‡Department of Chemistry, Umeå University, SE-90187 Umeå, Sweden S Supporting Information *
ABSTRACT: Dissolved natural organic matter (NOM) sorption at mineral surfaces can significantly affect the persistence of organic carbon in soils and sediments. Consequently, determining the mechanisms that stabilize sorbed NOM is crucial for predicting the persistence of carbon in nature. This study determined the effects of loadings and pH on the thermal stability of NOM associated with synthetic goethite (α-FeOOH) particle surfaces, as a proxy for NOM−mineral interactions taking place in nature. NOM thermal stability was investigated using temperature-programmed desorption (TPD) in the 30−700 °C range to collect vibration spectra of thermally decomposing goethite−NOM assemblages, and to concomitantly analyze evolved gases using mass spectrometry. Results showed that NOM thermal stability, indicated by the range of temperatures in which CO2 evolved during thermal decomposition, was greatest in unbound NOM and lowest when NOM was bound to goethite. NOM thermal stability was also loading dependent. It decreased when loadings were in increased the 0.01 to 0.42 mg C m−2 range, where the upper value corresponds to a Langmuirian adsorption maximum. Concomitant Fourier transform infrared (FTIR) spectroscopy measurement showed that these lowered stabilities could be ascribed to direct NOM-goethite interactions that dominated the NOM binding environment. Mineral surface interactions at larger loadings involved, on the contrary, a smaller fraction of the sorbed NOM, thus increasing thermal stability toward that of its unbound counterpart. This study thus identifies a sorption threshold below which NOM sorption to goethite decreases NOM thermal stability, and above which no strong effects are manifested. This should likely influence the fate of organic carbon exposed to thermal gradients in natural environments.
■
INTRODUCTION Mineral particle surfaces are believed to play key roles in the long term resilience to abiotic and biotic stability of natural organic matter (NOM) in soils and sediments.1−3 NOM−mineral interactions are consequently considered for slowing CO2 release to the atmosphere from soils and sediments.4−6 As such, these efforts are calling for a detailed understanding of the mechanisms through which minerals alter NOM stability. The nature of NOM binding to minerals is predominantly controlled by the physicochemical and structural/steric properties of minerals and NOM,2,7,8 as well as the environmental conditions in which these reactions occur. Binding types and strength can, for instance, be considerably affected by pH, ionic strength, foreign dissolved ions, reaction temperature as well as NOM loading.9−13 The issue of NOM loading can be of particular importance especially in the light of the multilayer adsorption model,14 suggesting that increased sorption loadings shift mineral−NOM binding from direct (e.g., coordinative surface metal−carboxylate or hydrogen bonding to surface hydroxo groups) to indirect (e.g., outer-sphere, electrostatic, and van der Waals) interactions. Such shifts in binding modes may consequently impact the lability of NOM toward both abiotic (i.e., (thermo)chemical) and biotic (i.e., enzymatically catalyzed microbial decomposition) breakdown processes. © XXXX American Chemical Society
Though both processes proceed through mechanistically distinct pathways, a number of recent studies underscored potential links.15−18 These links stem from the facilitated pathway offered by the decomposition of the most labile fractions (e.g., low weight molecular organic compounds) prior hitting an energetic barrier, which is then either overcome by (i) enzymatically mediated biotic decomposition at low-temperatures or (ii) externally applied heat in thermochemical abiotic processes. Although comparisons of this type are not likely to go beyond this point, knowledge of the intrinsic abiotic (thermo)chemical stability of NOM remains of fundamental value in our quest to study the fate of organic carbon in the environment. For example, we recently showed how microbial decomposition altered the intrinsic thermochemical properties of organic matter of a recent podzols (up to 2150−2340 years before present) yet had little effect in the underlying paleosols (2750−7000 years before present).3 The potential links15−18 between biotic and abiotic degradation of NOM enable the application of thermal analysis to study biogeochemical stability of NOM. Recent studies show that the Received: October 7, 2015 Revised: November 15, 2015
A
DOI: 10.1021/acs.jpca.5b09821 J. Phys. Chem. A XXXX, XXX, XXX−XXX
Article
The Journal of Physical Chemistry A
NOM Batch Adsorption. Batch adsorption experiments were conducted in goethite suspensions at pH 3.0, 4.0, and 6.0. A 0.5 mL aliquot of a goethite (11.23 g L−1, specific surface area: 69.59 m2 g−1) aqueous suspension was mixed with 20 mL diluted NOM solutions (1.77 to 692 mg C L−1) in 50 mL polyethylene centrifuge tubes to achieve C/Fe (w/w) ratios in the 0.01−3.91 range. These ratios are, for example, comparable to those of most soils worldwide (0.01−6.45)26 but lower than those of Swedish rivers (5.26−74.84).27 The pH values of the resulting suspensions were adjusted to pH 3.0, 4.0, and 6.0 with dilute HCl or NaOH and then mixed on an end-to-end rotator at 25 ± 1 °C for 48 h. This equilibration time was previously shown to yield stable soluble NOM concentrations, suggesting that (near) equilibrium was reached.28 Next, the suspension was centrifuged (4000 rpm, 15 min) and decanted into glass vials for final soluble C concentrations using a TOC analyzer (HACH IL500). Finally, the centrifuged solids were freeze-dried for TPD experiments. Separate goethite and NOM (692 mg C L−1) solutions were also equilibrated at pH 3.0, 4.0, and 6.0 with dilute HCl and NaOH, went through the same adsorption procedure as goethite-NOM assemblages, and freeze-dried for these experiments. NOM uptake data were modeled using a two-term model accounting for Langmuirian-type adsorption at low loading and condensation/precipitation/coagulation events at higher loadings. The analytical expression describing these processes is derived from the Do-Do adsorption model:29
thermal stability of NOM could vary in the absence and presence of minerals among soil types, 18−21 land use, 20,21 and ecosystems.18−23 To our best knowledge, few studies24 have explored how the biological and thermal stability of NOM sorbed on soil minerals changed with carbon loading. Still, this research area suffers from a knowledge gap concerning the mechanisms affecting the thermal stability of NOM bound to mineral surfaces. In an effort to further our knowledge on the nature of mineral− NOM interactions, this study is focused on testing the hypothesis that adsorption and variations in loading alter the intrinsic thermochemical stability of NOM. This study is consequently not an attempt to emulate conditions of biotically mediated NOM decomposition reactions but, more specifically, is focused on resolving the thermochemical stability of mineral−NOM assemblages. The temperature-programmed desorption (TPD) technique used in this study allows us to monitor the temperature-resolved release of NOM moieties, from the most weakly bound at low temperature up to the most resilient fractions at high temperature. These processes are monitored via thermally evolved H2O and CO2 gases by mass spectrometry (MS), and via concomitant changes in molecular vibrations of the mineral−NOM assemblages by Fourier transform infrared (FTIR) spectroscopy. These efforts notably build upon a previous study from our group25 where we resolved the distinct thermal decomposition pathways of oxalic acid bound to goethite (α-FeOOH). In this study, we extend those previous efforts to thermal decomposition pathways to NOM and demonstrate a strong loading dependence on the thermal stability.
n=α+1
Kμ ∑1 xn Kfx Cμ = So + Cμs n=α+1 n n = α + 1 n − α1 1 + Kfx Kμ ∑1 x + Kμ ∑1 x
■
EXPERIMENTAL METHODS Materials. Synthetic goethite was prepared by dropwise addition of 2.5 M NaOH to a vigorously stirred 0.15 M Fe(NO3)3·9H2O solution until the suspension reached pH 12. The resulting product was placed in an oven and converted to goethite at 50 °C for 24 h. The solids were then washed with deionized water and dialyzed. Dialysis water was changed three times per day until its conductivity was on the same order of magnitude as that of deionized water. Powder X-ray diffraction analysis (Bruker D8 Advance) confirmed that the final product was solely composed of crystalline goethite. Attenuated total reflectance (Golden Gate, diamond cell, single bound, 45° angle of incidence) FTIR spectroscopic measurements (Bruker Vertex 70/V) showed that no other iron oxyhydroxides were present. X-ray photoelectron spectroscopy (Kratos Axis Ultra electron spectrometer) confirmed that particle surfaces contained no atoms and compounds other than Fe(III), O, OH, and traces of aliphatic C (285.0 eV), the latter being commonly found in this in vacuo analysis technique. Finally, a Brunauer−Emmett−Turner (BET) analysis of a 55-point adsorption/desorption N2(g) isotherm (TriStar, Micromeritics), on samples previously dried in situ under a stream of N2(g) at 100 °C for 16 h, returns a specific surface area of 70 m2 g−1. The isoelectric point of the resulting material, measured by electrophoretic mobility (Zetasizer) in CO2-free aqueous suspensions of 0.01−0.1 M NaCl, is 9.4. Natural organic matter was collected from a bog on the top of Bräntberget hill in Umeå, Sweden (N 63° 50′ 19″, E 20° 18′ 4″). The sample was filtered through a 200 μm sieve and then concentrated by partial freezing. After volumes of initial NOM were reduced from ∼20 to ∼1.5 L, the final NOM was passed through 0.45 μm membrane filter (Whatman). The final NOM was enriched to 692 mg C L−1 and stored at 4 °C for further analyses.
(1)
The first term of this equation accounts for adsorption and includes a Langmuirian adsorption maximum density (So) and its association constant (Kf). The second term for condensation/ precipitation/coagulation and includes a saturation concentration (Cμs) in the second plateau of a Type IV isotherm,30 the association constant (Kμ), and size of the NOM cluster (α) involved in the reaction. The NOM loading dependence of this equation is given by the term x, which is given by the fraction of NOM relative to the maximal soluble value (Cmax), such that x = C/Cmax. Co-optimization of parameters (So, Kf, Cμs, Kμ, and Cmax) was carried out using all experimental data with a trust region reflective algorithm in the computational environment of Matlab (R2014b, The Mathworks). TPD-FTIR. The relative thermal stabilities of the freeze-dried solids were evaluated by TPD experiments during which time the vibrational modes of the solid sample and the evolution of gaseous H2O and CO2 were monitored. About 2 mg of the samples was pressed onto a tungsten mesh (Unique wire weaving, 0.002 in. diameter) and then emplaced into a copper heating shaft in direct contact with a K-type thermocouple. The sample was then introduced into a reaction chamber (from AABSPEC, model #2000-A) equipped with CaF2 optical windows enabling the collection of FTIR spectra during TPD. The chamber was pumped to pressures below 2.5 mTorr, the detection limit of the capacitance manometer (from MKS, model 1179A), using a vacuum pump (HiCube, Pfeiffer). The TPD experiment was thereafter initiated by raising the temperature from 30 to 700 °C at a rate of 10 °C min−1. FTIR spectra were continuously collected with a Bruker Vertex FTIR 70/v spectrometer from 600 to 4000 cm−1 at a working resolution of 4.0 cm−1 and a forward/reverse scanning rate of 10 Hz and were obtained by coadding 50 scans. The Blackman−Harris B
DOI: 10.1021/acs.jpca.5b09821 J. Phys. Chem. A XXXX, XXX, XXX−XXX
Article
The Journal of Physical Chemistry A
loadings and precipitation/condensation/coagulation at higher loadings. The model adopted for this work (eq 1) predicts an adsorption maximum of 0.44 mg C m−2. Considering that surface waters have molecular weights of the order of 1250−2190 Da,37 this Langmuirian maximum could correspond to loadings of 0.3 μmol m−2 (0.2 sites m−2). This loading consequently is well within the crystallographic density of sites at goethite surfaces (e.g. 3~ sites/nm2 of the highly exchangeable singly-coordinated OH groups), is in the range of previous fundamental studies of NOM binding,38 and likely denotes the multidentate nature of NOM−goethite surface interactions. Thermal Decomposition of NOM. FTIR spectra of NOM reveal a rich signature 1000−2000 cm−1 region arising from vibrational modes of NOM-bearing functionalities and exhibit four predominant bands that are used to track NOM protonation state and thermal stability (Figure 2). The 1221, 1377, 1585, and 1711 cm−1 bands dominate this region and undergo changes in intensity with pH. Comparison with the vibration modes of low
three-term apodization function was used to correct for phase resolution. Background spectra were obtained on the same tungsten mesh used for the experiment. A mass spectrometer (Pfeiffer Vacuum, PrismaPlus QMG 220 F) also monitored the thermally evolved CO2 and H2O effluent gases during the experiments. Evolved gas levels are reported in terms of the ion current (A) detected under Faraday analysis mode. Chemometric analyses of the FTIR spectra in the 1000−2000 cm−1 region were used to identify correlations and lack of correlations of band sets corresponding to NOM functional groups and moieties. The multivariate curve resolution (MCR)31 method applies the Beer−Lambert law (A = εC) to extract linearly independent spectral components (ε) and their relative concentrations (C) from a matrix of experimental data (A) generated from TPD experiments. These matrices were first offset to zero absorbance at 2000 cm−1, where absorbances are negligible. Their eigenvectors were then inspected by single value decomposition (SVD)32 and rotated into a real chemical space to obtain ε using the MCR-ALS program.31 The CO2 TPD data of the mixed NOM−goethite samples were treated to remove contributions from any decarbonation reactions involving a minor (≪2 mol %) carbonate species of the goethite bulk. Details on the nature of this goethite−carbonate solid solution can be found in a previous study from our group.33 The goethitederived CO2 contributions were removed by scaling the CO2 trace generated by the thermal decomposition of a NOM-free goethite sample to the dehydroxylation peak of goethite. All calculations were performed with MATLAB 8.4 (The Mathworks, Inc.).
■
RESULTS AND DISCUSSION Sorption of NOM on Goethite. NOM sorption data (Figure 1; Table S1) covers loadings in the 0.01−4.44 mg C m−2
Figure 1. Adsorption isotherms of NOM on goethite at pH 3 (square), 4 (triangle), and 6 (circle). Lines denote model predictions from the DoDo model (eq 1) explaining NOM uptake data in terms of an adsorption (direct NOM−goethite interactions) and a condensation regime. Model parameters are as follows: So = 0.44 sites nm−2; Kf = 25.9; Cμs = 406 mg C L−1; Kμ = 0.99; α = 4.7 sites; Cmax = 520 mg C L−1.
range. They consequently cover considerably broader loadings than in previous NOM sorption studies: e.g., 0.04 mg C m−2 in ref 1, 0.12−0.62 mg C m−2 in ref 34, 0.69 mg C m−2 in ref 35, and 1.9 mg C m−2 in ref 36. All NOM adsorption isotherms are of Type IV,30 starting with a steeply increasing loading to 0.28 mg C m−2 goethite, a plateau at the loading of 0.28−0.42 mg C m−2, followed by a precipitation/condensation/coagulation-like domain with loading above 0.42 mg C m−2. The uptake data can be described using a two-term adsorption model (eq 1) predicting adsorption by goethite surface sites at low NOM
Figure 2. FTIR spectra of dry NOM previously equilibrated at pH 3, 4, and 6. Spectra collected under TPD (30−700 °C at a rate of 10 °C min−1). C
DOI: 10.1021/acs.jpca.5b09821 J. Phys. Chem. A XXXX, XXX, XXX−XXX
Article
The Journal of Physical Chemistry A molecular weight carboxylic acids suggests that the 1711 cm−1 (υCO) and 1221 cm−1 (δCOH) bands arise from protonated carboxyl groups whereas the 1585 cm−1 (υCO asymmetric) and 1377 cm−1 (υCO symmetric) bands are from deprotonated groups. This band assignment falls in line with the predominance of the 1585/1377 cm−1 band pair at pH 6 and their replacement by the 1711/1221 cm−1 band pair at low pH. These, we stress, are predominant spectral features superimposed with a variety of unresolvable additional modes (e.g., ring modes, amines, and phosphates) that contribute to the overall signature of the region. We also note that the OH and CH stretching regions were monitored in all experiments but did not provide any further information than what can be extracted from the 1000−2000 cm−1 region, and therefore, they are not reported in this study for the sake of conciseness. MCR analysis of these data required three components (Figure 3a) that were associated with distinct ranges of temperature (Figure 3b). MCR component I dominates below 300 °C and corresponds to the low temperature spectra. MCR
component II dominates the 300−650 °C range and exhibits none of the low temperature band pairs, although broad intensities in the 1100−1600 cm−1 region are to be noted. Comparison with the mass spectrometric-derived CO2 and H2O data (Figures 3c and S1) shows that this component correlates with a first set of NOM decarbonation and dehydration events. MCR component III gains importance only above 300 °C and becomes the dominant component above 650 °C. Again, comparison with the mass spectrometric data (Figure 3c) shows that MCR component III correlates not only with a secondary dehydration peak above 500 °C but also, most importantly, with the release of CO2 from the samples. This latter correlation can be appreciated in Figure 3b where it is shown in terms of cumulative released CO2. Finally, we note that differences in pH, and therefore protonation state, did not produce any distinct effects on the intrinsic thermal stability of the NOM considered for this study. Thermal Decomposition of Sorbed NOM. FTIR spectra of NOM sorbed to goethite particle surfaces at 30 °C (Figure 4a)
Figure 4. FTIR spectra of NOM−goethite samples at pH 6 measured at 20 °C.
contain the aforementioned NOM bands, alongside goethite O− H stretching (υOH = 3120 cm−1; not shown), bending (γOH = 794 cm−1; δOH = 896 cm−1; not shown), and bending overtone/ combination (δOH = 1663 cm−1, δOH + γOH = 1798 cm−1) modes (Figure S2). At high NOM loadings, spectra appear as a simple linear combination of NOM and goethite bands. High loadings consequently do not alter the vibrational modes of NOM because only a minor, unresolvable, fraction of the moieties are in contact with goethite surface. Low loadings, on the contrary, give rise to considerably more distinct bands, and particularly the lowest (C/Fe = 0.01) where a collection of finer bands appears. These more distinct attributes arise from direct interactions with the goethite surface perturbing the original configurations (e.g.,
Figure 3. (a) MCR spectral components (I, II, and III) of dry NOM under TPD, (b) MCR concentration profiles and cumulative normalized CO2 collected by MS, and (c) MS-derived CO2 and H2O profiles. MCR spectral components and concentration profiles were derived from the data of Figure S1. The similarity between the concentration profile of MCR component III with the cumulative CO2 (b) suggests a link between the consumption of components I and II with the release of CO2 from the samples. No effects of pH are noted. D
DOI: 10.1021/acs.jpca.5b09821 J. Phys. Chem. A XXXX, XXX, XXX−XXX
Article
The Journal of Physical Chemistry A intramolecular hydrogen bonding network) of NOM. This vibrational feature therefore provides a first line of evidence for surface loading effects on the nature of goethite−NOM interactions. This evidence is revealed further in the TPD experiments (Figure 5 for pH 6 data at C/Fe = 0.14, cf. Figure S2 for full data
Figure 6. (a, b) Mass spectrometrically derived gaseous H2O (a) and CO2 (b) evolved from dry NOM−goethite samples at pH 6 during TPD (full data sets in Figures S3−S5). Units are arbitrary and used to compare peak positions (temperature) between samples, and thus not peak height/area. (c, d) MCR components from FTIR spectra (Figure S5) showing the correspondence of component ② (c) with the dehydroxylation event at 275−300 °C (a) and of components ③ and ④ with the decarbonation events above 300 °C. (c,d) Profiles in gray correspond to those of unbound NOM from Figure 3b.
Component ① of goethite−NOM had a comparable profile to that of Component I of NOM and is associated with the loss of the 1585/1377 cm−1 and 1711/1221 cm−1 band pairs. Component ② of goethite−NOM corresponded to a bulk goethite dehydroxylation event, which was also recovered in the mass spectrometric H2O data (Figure 6a) through a peak centered at 275−300 °C. This peak corresponds to a single dehydroxylation event of goethite and is independent of pH, again as discussed in a previous study.33 Components ③ and ④ of goethite−NOM are comparable with Components II and III of NOM, respectively, in their appearance with the release of CO2 (Figure 6b) from the samples. The cumulative released CO2 data (Figure 7) offers an insightful perspective on the impacts of loading on NOM thermal stability. Increasing C loadings up to the adsorption maximum of 0.42 mg C m−2 (C/Fe = 0.14) at all three pH values considered in this study effectively decreased the thermal stability of NOM. For instance, unbound NOM equilibrated at pH 3 loses 50% of its CO2 at 577 °C during a TPD experiment in the 30− 700 °C range. This temperature, however, decreases to 474 °C in goethite−NOM assemblages at C/Fe 0.01 and down to 402 °C at C/Fe 0.14, namely, our Langmuirian sorption maximum of 0.42 mg C m−2. The trend of decreased thermal stability is, however, reversed when loadings exceed this C/Fe ratio with the temperature of 50% CO2 loss nearly achieving those of unbound NOM at the largest C/Fe ratio under study (Figure 7). Our observed trends for the thermal stability hold for all three pH values considered in this study. The controlling factor for the loading-dependent thermal stabilities lies in the nature of NOM−goethite interactions, as first noted for the 30 °C spectra of Figure 4. At low loadings, a greater proportion of the moieties of NOM interacts with goethite surfaces, a condition that alters the intrinsic NOM
Figure 5. TPD-FTIR spectra of NOM−goethite samples at pH 6 (full data set in Figure S2).
set). Just as in the NOM-only system, intensities of the 1585/ 1377 and 1711/1221 cm−1 band pairs were systematically attenuated with temperature. The bending combination (δ(OH)′) and overtones (δ(OH) + γ(OH)) of goethite also underwent concomitant losses in intensity from bulk dehydroxylation reactions to hematite but these reactions were independent of NOM TPD reactions. Details on the impact of thermal decomposition of the goethite bulk on the FTIR spectra are shown in Figure S2 and discussed in greater detail in a previous study from our group.33 MCR analyses of the FTIR spectra for the mixed NOM− goethite samples under TPD generated the following four components (see Figure 6 for pH 6 data, and Figures S3−S5 for full data set). As in pure NOM, all components were largely unaffected by the preparation pH. The low temperature MCR E
DOI: 10.1021/acs.jpca.5b09821 J. Phys. Chem. A XXXX, XXX, XXX−XXX
Article
The Journal of Physical Chemistry A
soils2,5,24,39 and sediments.40−42 Although our observations may seemingly challenge this point of view, a few words of caution are in order. This study was solely focused on the intrinsic thermochemical stability of NOM under dry conditions, and therefore not biotic decomposition per se. We therefore cannot directly ascertain whether biotic decomposition mechanisms of NOM would be disfavored in mineral−NOM assemblages from our observations alone. It could, for example, be possible that although the close association of NOM to mineral surfaces decreases its thermal stability, it may hinder enzymatically mediated microbial decomposition, as expected from the literature. This and related possibilities, for example, enzyme docking mechanisms at or within NOM, would have to be more specifically tested in the context of microbial respiration assay studies.
■
CONCLUSIONS Our results point to a threshold organic C loading value below which sorption decreases the thermal stability of NOM. Direct NOM−mineral surface interactions would be responsible for the decreased resilience of NOM to thermal gradients, perhaps in destabilizing internal networks of hydrogen bonds and van der Waals-type interactions that protect the NOM bulk from decomposition. Above this threshold, thermal stability increases with NOM loadings but only to reach stabilities comparable to that of unbound NOM. Our findings do not agree with the concept that NOM sorption to mineral surfaces increases its stability, and uncovered a potentially complex interplay between stability and loadings. Our work highlights the importance of studying changes both thermal and biological stability of organomineral complexes simultaneously as well their composition and structure in a variety of environments. This should help improve our understanding of the roles that minerals play on the fate of NOM in nature.
■
Figure 7. Cumulative CO2 released from the dry NOM−goethite samples during TPD. Red arrows are shifts in temperature profile for increasing NOM loadings, namely, toward low temperature from C/Fe = 0.01 to 0.14 and toward high temperature from C/Fe = 0.14 to 3.91.
ASSOCIATED CONTENT
S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jpca.5b09821. Raw and cumulative TPD traces (Figure S1), FTIR-TPD data (Figure S2), and TPD trace, MCR components, and cumulative TPD trace (Figure S3), tables of composition data (Table S1) and temperatures from pyrolysis (Table S2) for natural dissolved organic matter sorption on goethite (PDF)
properties. The collection of sharper bands (e.g., at C/Fe = 0.01) likely denote a complex range of direct iron−carboxylate group bonding, and hydrogen bonding of carboxylic moieties directly with surface (hydr)oxo groups, as for instance noted for the case of goethite-bound oxalic acid in our previous TPD study.19 These low loadings are, in our view, conditions prone to destabilizes internal networks of hydrogen bonds and van der Waals-type interactions that are otherwise present in unbound NOM. In contrast, NOM loadings exceeding our Langmuirian maximum of 0.42 mg C m−2 are associated with greater fractions of NOM that are unperturbed by goethite. Again, this result can be understood by the expectation that only a minor portion of the NOM is involved in surface binding, and therefore that the NOM thermal stability is predominantly characterized by its unbound portion. From these results, we conclude that NOM binding to goethite surfaces compromise its intrinsic thermal stability in dry environments, and that the temperature dependence of NOM thermal decomposition is strongly loading dependent. Still, it should be reemphasized that because this study has solely focused on the thermal stability of NOM it does call for comparative thermal and biological decomposition studies to improve our understanding of the fate of NOM in nature. In fact, environmental science research generally assumes that minerals increase the chemical and biological stability of NOM in
■
AUTHOR INFORMATION
Corresponding Author
*W. Feng. E-mail:
[email protected]. Phone: +1 405 325 6519. Present Address §
101 David L. Boren BLVD, Norman, OK 73019-5300, USA.
Notes
The authors declare no competing financial interest. ∥ E-mail:
[email protected]. ⊥ E-mail:
[email protected].
■
ACKNOWLEDGMENTS This work was supported by the Kempe Foundation as well as by the Swedish Research Council (grant no 2009-3282 to J.K. and no. 2012-2976 to J.-F.B.). F
DOI: 10.1021/acs.jpca.5b09821 J. Phys. Chem. A XXXX, XXX, XXX−XXX
Article
The Journal of Physical Chemistry A
■
(21) Saenger, A.; Cécillon, L.; Sebag, D.; Brun, J.-J. Soil organic carbon quantity, chemistry and thermal stability in a mountainous landscape: A Rock−Eval pyrolysis survey. Org. Geochem. 2013, 54, 101−114. (22) Saenger, A.; Cécillon, L.; Poulenard, J.; Bureau, F.; De Daniéli, S.; Gonzalez, J.-M.; Brun, J.-J. Surveying the carbon pools of mountain soils: A comparison of physical fractionation and Rock-Eval pyrolysis. Geoderma 2015, 241−242, 279−288. (23) Rovira, P.; Kurz-Besson, C.; Couteaux, M.-M.; Vallejo, V. R. Changes in litter properties during decomposition: A study by differential thermogravimetry and scanning calorimetry. Soil Biol. Biochem. 2008, 40 (1), 172−185. (24) Feng, W.; Plante, A. F.; Aufdenkampe, A. K.; Six, J. Soil organic matter stability in organo-mineral complexes as a function of increasing C loading. Soil Biol. Biochem. 2014, 69, 398−405. (25) Boily, J. F.; Szanyi, J.; Felmy, A. R. Effects of surface coordination on the temperature-programmed Desorption of oxalate from goethite. J. Phys. Chem. C 2007, 111 (45), 17072−17081. (26) Wagai, R.; Mayer, L. M. Sorptive stabilization of organic matter in soils by hydrous iron oxides. Geochim. Cosmochim. Acta 2007, 71 (1), 25−35. (27) Kritzberg, E. S.; Ekström, S. M. Increasing iron concentrations in surface waters - a factor behind brownification? Biogeosciences 2012, 9 (4), 1465−1478. (28) Jönsson, J. Phase transformations and surface chemistry of secondary iron minerals formed from acid mine drainage; Umeå University: Umeå, Sweden, 2003. (29) Do, D. D.; Do, H. D. A model for water adsorption in activated carbon. Carbon 2000, 38 (5), 767−773. (30) Gregg, S. J.; Sing, K. S. W. Adsorption, surface area and porosity; Academic Press: London, 1992. (31) Muik, B.; Lendl, B.; Molina-Diaz, A.; Valcarcel, M.; AyoraCanada, M. J. Two-dimensional correlation spectroscopy and multivariate curve resolution for the study of lipid oxidation in edible oils monitored by FTIR and FT-Raman spectroscopy. Anal. Chim. Acta 2007, 593 (1), 54−67. (32) Golub, G. H.; Reinsch, C. Singular value decomposition and least squares solutions. Numer. Math. 1970, 14 (5), 403−420. (33) Boily, J.-F.; Szanyi, J.; Felmy, A. R. A combined FTIR and TPD study on the bulk and surface dehydroxylation and decarbonation of synthetic goethite. Geochim. Cosmochim. Acta 2006, 70 (14), 3613− 3624. (34) Saidy, A. R.; Smernik, R. J.; Baldock, J. A.; Kaiser, K.; Sanderman, J. The sorption of organic carbon onto differing clay minerals in the presence and absence of hydrous iron oxide. Geoderma 2013, 209, 15− 21. (35) Gu, B. H.; Schmitt, J.; Chen, Z.; Liang, L. Y.; McCarthy, J. F. Adsorption and desorption of different organic-matter fractions on ironoxide. Geochim. Cosmochim. Acta 1995, 59 (2), 219−229. (36) Kaiser, K.; Guggenberger, G. Sorptive stabilization of organic matter by microporous goethite: sorption into small pores vs. surface complexation. Eur. J. Soil Sci. 2007, 58 (1), 45−59. (37) Metsämuuronen, S.; Sillanpaa, M.; Bhatnagar, A.; Manttari, M. Natural organic matter removal from drinking water by membrane technology. Sep. Purif. Rev. 2014, 43 (1), 1−61. (38) Filius, J. D.; Meeussen, J. C. L.; Lumsdon, D. G.; Hiemstra, T.; Van Riemsdijk, W. H. Modeling the binding of fulvic acid by goethite: The speciation of adsorbed FA molecules. Geochim. Cosmochim. Acta 2003, 67 (8), 1463−1474. (39) Kalbitz, K.; Schwesig, D.; Rethemeyer, J.; Matzner, E. Stabilization of dissolved organic matter by sorption to the mineral soil. Soil Biol. Biochem. 2005, 37 (7), 1319−1331. (40) Keil, R. G.; Montlucon, D. B.; Prahl, F. G.; Hedges, J. I. Sorptive preservation of labile organic matter in marine sediments. Nature 1994, 370 (6490), 549−552. (41) Ransom, B.; Kim, D.; Kastner, M.; Wainwright, S. Organic matter preservation on continental slopes: Importance of mineralogy and surface area. Geochim. Cosmochim. Acta 1998, 62 (8), 1329−1345. (42) Mayer, L. M.; Xing, B. S. Organic matter-surface area relationships in acid soils. Soil Sci. Soc. Am. J. 2001, 65 (1), 250−258.
REFERENCES
(1) Kaiser, K.; Guggenberger, G. The role of DOM sorption to mineral surfaces in the preservation of organic matter in soils. Org. Geochem. 2000, 31 (7−8), 711−725. (2) Kalbitz, K.; Kaiser, K. Contribution of dissolved organic matter to carbon storage in forest mineral soils. J. Plant Nutr. Soil Sci. 2008, 171 (1), 52−60. (3) Feng, W.; Boily, J.-F.; Erhagen, B.; Nilsson, M.; Klaminder, J. Millennia-old organic carbon in a boreal paleosol: chemical properties and their link to mineralizable carbon fraction. J. Soils Sediments 2015, 1−10. (4) Torn, M. S.; Trumbore, S. E.; Chadwick, O. A.; Vitousek, P. M.; Hendricks, D. M. Mineral control of soil organic carbon storage and turnover. Nature 1997, 389 (6647), 170−173. (5) Kramer, M. G.; Sanderman, J.; Chadwick, O. A.; Chorover, J.; Vitousek, P. M. Long-term carbon storage through retention of dissolved aromatic acids by reactive particles in soil. Glob. Change Biol. 2012, 18 (8), 2594−2605. (6) Lalonde, K.; Mucci, A.; Ouellet, A.; Gelinas, Y. Preservation of organic matter in sediments promoted by iron. Nature 2012, 483 (7388), 198−200. (7) Kaiser, K.; Guggenberger, G. Mineral surfaces and soil organic matter. Eur. J. Soil Sci. 2003, 54 (2), 219−236. (8) Scheel, T.; Haumaier, L.; Ellerbrock, R. H.; Ruehlmann, J.; Kalbitz, K. Properties of organic matter precipitated from acidic forest soil solutions. Org. Geochem. 2008, 39 (10), 1439−1453. (9) Jardine, P. M.; Weber, N. L.; McCarthy, J. F. Mechanisms of dissolved organic carbon adsorption on soil. Soil Sci. Soc. Am. J. 1989, 53 (5), 1378−1385. (10) Gu, B. H.; Schmitt, J.; Chen, Z. H.; Liang, L. Y.; McCarthy, J. F. Adsorption and desorption of natural organic matter on iron-oxide mechanisms and models. Environ. Sci. Technol. 1994, 28 (1), 38−46. (11) Feng, X. J.; Simpson, A. J.; Simpson, M. J. Chemical and mineralogical controls on humic acid sorption to clay mineral surfaces. Org. Geochem. 2005, 36 (11), 1553−1566. (12) Feng, X. J.; Simpson, A. J.; Simpson, M. J. Investigating the role of mineral-bound humic acid in phenanthrene sorption. Environ. Sci. Technol. 2006, 40 (10), 3260−3266. (13) Mikutta, C.; Lang, F.; Kaupenjohann, M. Citrate impairs the micropore diffusion of phosphate into pure and C-coated goethite. Geochim. Cosmochim. Acta 2006, 70 (3), 595−607. (14) Kleber, M.; Sollins, P.; Sutton, R. A conceptual model of organomineral interactions in soils: self-assembly of organic molecular fragments into zonal structures on mineral surfaces. Biogeochemistry 2007, 85 (1), 9−24. (15) Lopez-Capel, E.; Bol, R.; Manning, D. A. C. Application of simultaneous thermal analysis mass spectrometry and stable carbon isotope analysis in a carbon sequestration study. Rapid Commun. Mass Spectrom. 2005, 19 (22), 3192−3198. (16) Plante, A. F.; Fernández, J. M.; Haddix, M. L.; Steinweg, J. M.; Conant, R. T. Biological, chemical and thermal indices of soil organic matter stability in four grassland soils. Soil Biol. Biochem. 2011, 43 (5), 1051−1058. (17) Broström, M.; Nordin, A.; Pommer, L.; Branca, C.; Di Blasi, C. Influence of torrefaction on the devolatilization and oxidation kinetics of wood. J. Anal. Appl. Pyrolysis 2012, 96 (0), 100−109. (18) Peltre, C.; Fernandez, J. M.; Craine, J. M.; Plante, A. F. Relationships between Biological and Thermal Indices of Soil Organic Matter Stability Differ with Soil Organic Carbon Level. Soil Sci. Soc. Am. J. 2013, 77 (6), 2020−2028. (19) Boily, J. F.; Szanyi, J.; Felmy, A. R. Effects of surface coordination on the temperature-programmed Desorption of oxalate from goethite. J. Phys. Chem. C 2007, 111 (45), 17072−17081. (20) Lopez-Capel, E.; Sohi, S. P.; Gaunt, J. L.; Manning, D. A. C. Use of thermogravimetry-differential scanning calorimetry to characterize modelable soil organic matter fractions. Soil Sci. Soc. Am. J. 2005, 69 (1), 136−140. G
DOI: 10.1021/acs.jpca.5b09821 J. Phys. Chem. A XXXX, XXX, XXX−XXX
