Article pubs.acs.org/est
Sorption Selectivity in Natural Organic Matter Studied with Nitroxyl Paramagnetic Relaxation Probes Charisma Lattao,† Xiaoyan Cao,‡ Yuan Li,‡ Jingdong Mao,‡ Klaus Schmidt-Rohr,§ Mark A. Chappell,∥ Lesley F. Miller,∥ Albert Leo dela Cruz,⊥ and Joseph J. Pignatello*,† †
Department of Environmental Sciences, The Connecticut Agricultural Experiment Station, P.O. Box 1106, New Haven, Connecticut 06504, United States ‡ Department of Chemistry and Biochemistry, Old Dominion University, 4541 Hampton Boulevard, Norfolk, Virginia 23529, United States § Department of Chemistry, Iowa State University, Hach Hall, Ames, Iowa, 50010, United States ∥ Environmental Laboratory, U.S. Army Corps of Engineers, 3909 Halls Ferry Road, Vicksburg, Mississippi 39180, United States ⊥ LSU Superfund Research Center, Louisiana State University, Baton Rouge, Louisiana 70802, United States S Supporting Information *
ABSTRACT: Sorption site selectivity and mechanism in natural organic matter (NOM) were addressed spectroscopically by the sorption of paramagnetic nitroxyl compounds (spin probes) of different polarity, TEMPO (2,2,6,6-tetramethylpiperidine-1-oxyl) and HTEMPO (4-hydroxy-2,2,6,6-tetramethylpiperidine-1-oxyl). The sorbents were Pahokee peat, Beulah-Zap lignite, and a polystyrene−poly(vinyl methyl ether) (PS-PVME) polymer blend representing the mixed aliphatic− aromatic, polar−nonpolar character of NOM. Nuclear-electron spin interaction serves as an efficient relaxation pathway, resulting in attenuation of the 13C−CP/ TOSS NMR signal for 13C nuclei in proximity to the N−O· group (r−6 dependence). In the natural solids the spin probes sorbed more specifically (greater isotherm nonlinearity) and had lower rotational mobility (broader electron paramagnetic resonance signals) than in PS-PVME. Titration with spin probe indicated almost no selectivity for the different carbon functional groups of PSPVME, and little to no selectivity for the different carbon moieties of Pahokee and Beulah, including aromatic, alkyl, O-alkyl, di-O-alkyl, and O-methyl. In any case, sorption site selectivity of spin probes to NOM was always weaker than partition selectivity found in model solvent−water (toluene, hexadecane, anisole, octanol) and cellulose− water systems. The results indicate little or no preferential sorption in NOM based on functional group chemistry or putative microdomain character, but rather are consistent with the filling of pores whose walls have an average chemical environment reflecting the bulk chemical composition of the solid. This work demonstrates for the first time the use of paramagnetic probes to study sorption specificity.
■
INTRODUCTION The exact manner in which organic compounds sorb to soil or sedimentary organic matter (SOM) has been a topic of much study and debate. Among the central issues is the need for conceptual models that explain the high variability in the organic carbon-normalized partition coefficient (KOC) among whole soils and their isolated humic fractions,1−3 and the selectivity and nonideality often observed during sorption manifested as nonlinearity, competition, and hysteresis. A sound model must feature SOM as a heterogeneous phase that offers a hierarchy of sites with different affinities and capacities. Three major hypotheses have emerged to explain sorption selectivity as reviewed previously:3 bonding-based selectivity, domain-based selectivity, and selectivity based on the physical state of SOM and the conformations of its molecules. Bonding-based selectivity is possible for polar molecules capable of directional forces, such as hydrogen bonding, © 2012 American Chemical Society
coordination to bridging metal ions, etc., with complementary functional groups on SOM. While bonding-based selectivity seems intuitively valid, little conclusive direct evidence for it has appeared. Nonlinearity was relatively insensitive to solute polarity or H-bonding capability,4,5 yet some support exists for a transition to sites of lower polar character as concentration increases especially for compounds that are capable of strong specific interactions (e.g., strong H-bonding).5,6 Domain-based selectivity is possible if the organic phase is imagined to segregate into microscopic domains enriched according to functional group identity “carbohydrate-like domain”, “aromatic domain,” etc.that act independently as Received: Revised: Accepted: Published: 12814
May 30, October October October
2012 9, 2012 19, 2012 19, 2012
dx.doi.org/10.1021/es302157j | Environ. Sci. Technol. 2012, 46, 12814−12822
Environmental Science & Technology
Article
sorbents.1,2 Evidence for the existence of such domains is mixed. Solid-state 13C NMR of humic substances reveals microdomains of amorphous and crystalline polymethylene, but carbohydratelike and lignin-like moieties appear to be intimately mixed.7−11 Many authors have suggested that black carbon, which is widely dispersed in soil and sediment as a result of fires and soot deposition, constitutes a polyaromatic domain in geosolids that dominates sorption, especially at low pollutant concentrations. Sorption to raw black carbon is typically strong and highly nonlinear, providing ample rationale for nonideality. Attributing sorption magnitude and selectivity to environmental black carbon in geosolids, however, is hampered by difficulties associated with separation, quantification, choice of reference standard (as black carbon is a continuum of materials that vary enormously in adsorbent properties), and anticipating the effects of weathering.3 13C NMR spectra of geosolids show signals for distinct organic domains of nonpolar aromatic carbon seemingly attributable to black carbon, but few studies have directly linked these domains to preferential sorption.12 Many attempts have been made to correlate sorption intensity with bulk functional group composition determined usually by 13C cross-polarization/magic angle spinning (CP/MAS) NMR spectroscopy. However, the results are conflicting or inconclusive on whether or not sorption correlates with aromaticity, aliphaticity, or polarity.1,2 It is noteworthy that, after combining data from the literature for natural geosolids, sorption nonlinearity, but not capacity, correlated with nonpolar aromatic carbon content. 2 The premise of the third hypothesis is that selectivity is a function of the physical state of the solid. Analogous to synthetic and certain natural polymers such as lignin, SOM is thought to exist in the stiff-chain, or glassy state, as opposed to the flexiblechain, or rubbery state. Unable to achieve the lowest energy conformation due to restricted relaxation, solids in their glassy state have internal micropores that preferentially attract solutes and could account for the selective behavior.1,3 If the micropore walls are of functional group composition similar to the bulk solid, functional group-based selectivity would not be observed. The objective of this study was to determine spectroscopically whether or not sorption is selective on the basis of domain or functional group identity by measuring the effect of paramagnetic spin probes on the 13C NMR spectra of two high-organic reference materials, Pahokee peat and Beulah-Zap lignite. These materials were chosen to represent soil organic matter and coaly particles in soil. Pahokee peat reportedly contains 8.1% “opaque” particles, attributable to woody charcoal.13 Beulah-Zap appears to be absent of fossilized charcoal (inertinite)14 consistent with leaving almost no carbon residue after combustion at 375 °C.15 However, it has a high aromatic-to-sp3 C ratio16 with an average polycyclic aromatic cluster size of nine carbons.17 We also include a reference blend of 1:1 polystyrene and poly(vinyl methyl ether) (PS-PVME), chosen because, like SOM, it contains contrasting aromatic−aliphatic and polar−nonpolar functionality that potentially offers specific sorption sites or the possibility for segregation as separate nano-domains. The probes are the thermally stable nitroxyl free radicals, TEMPO (2,2,6,6tetramethylpiperidine-N-oxyl) and HTEMPO (4-hydroxy2,2,6,6-tetramethylpiperidine-N-oxyl), whose structures are given in Figure 1. They are aliphatic compounds of similar size and structure but different polarity. Unpaired electrons provide fluctuating magnetic fields that cause nuclear relaxation, resulting in line-broadening and reduced relaxation times. Paramagnetic relaxation is highly efficient because the electron magnetic moment is 658 times the nuclear magnetic moment of 1H.18
Figure 1. Sorption isotherms of TEMPO and HTEMPO. Freundlich sorption parameters are as follows: TEMPO: Beulah (KF = 2.62 ± 0.02, N = 0.54 ± 0.01); Pahokee (KF = 1.91 ± 0.02, N = 0.68 ± 0.01); and PSPVME (KF = 1.26 ± 0.02, N = 0.99 ± 0.01). HTEMPO: Beulah (KF = 1.89 ± 0.04, N = 0.68 ± 0.02); Pahokee (KF = 1.09 ± 0.05, N = 0.82 ± 0.02); and PS-PVME (KF = 0.34 ± 0.05, N = 0.92 ± 0.01). Values of r2 are ≥0.99 in all cases.
Paramagnetic ions of relatively long electron spin−lattice relaxation time T1e (>10−8 s), as opposed to those of relatively short T1e ( KOC(HTEMPO), presumably due to the difference in hydrophobicity. For both spin probes, KOC follows the order Beulah > Pahokee > PS-PVME (except
free radicals. Extraction removed virtually all the spins (Figure S5). The EPR spectra of Pahokee and Beulah pure solids both display a signal with g-factor of 2.0036−2.0038 characteristic of carbon-centered radicals vicinal to oxygen, at 1.08 × 1017 and 1.20 × 1019 spins g−1, respectively. While nitroxyl compounds such as TEMPO and HTEMPO that have no α-hydrogens are not very reactive with respect to coupling with carbon-centered radicals (the rates depend greatly on resonance delocalization and steric effects),33 it is, nevertheless, not possible to tell whether the matrix carbon radical signal is reduced upon addition of spin probe, as it is obscured by the spin probe signal. We identified the oxidation product, 4-oxo-2,2,6,6-tetramethylpiperidine-N-oxyl in supernatants and solvent extracts after sorption of HTEMPO to Pahokee, as described in more detail in the SI (Figures S6−S7). This product results from oxidation at the hydroxyl, not the nitroxyl position. Sorption isotherms of the spin probes for Pahokee, Beulah, and PS-PVME, all corrected for recovery, along with their Freundlich fits are shown in Figure 1. Linearity (N) follows the same order for TEMPO (Beulah (0.536) < Pahokee (0.678) < PS-PVME (0.991)) as for HTEMPO (Beulah (0.676) < Pahokee (0.822) < PS-PVME (0.921)). Linearity trends inversely with specific surface area, and inversely with micropore volume for 12817
dx.doi.org/10.1021/es302157j | Environ. Sci. Technol. 2012, 46, 12814−12822
Environmental Science & Technology
Article
Figure 3. Peak suppression plots for integrated regions in the 13C−CP/TOSS/MAS spectrum for HTEMPO in Pahokee (a), Beulah (b), and PS-PVME (c). Error bars: 95% confidence level.
Pahokee ∼ PS-PVME for TEMPO at Ce = 300 mg/L). For both TEMPO and HTEMPO the KOC does not trend with aromatic content, but trends inversely with alkyl content. The KOC trends with CO2-micropore volume at each Ce evaluated:, suggesting the importance of pore-filling in the natural sorbents. Electron Paramagnetic Resonance (EPR) Analyses. The EPR spectra of adsorbed- and solution-phase TEMPO are shown in Figure S10. Spectra of nitroxyl radical solutions typically consist of three symmetrical hyperfine lines corresponding to interaction between the nuclear spin of nitrogen (+1, 0, −1) and the free electron spin.35 Adsorbed TEMPO gives a broad high field line (h−1) and the central hyperfine line (h0) is of higher intensity than the low and high field lines. The line width of the central peak follows the order: Beulah > Pahokee > PS-PVME. Furthermore, PS-PVME exhibits three hyperfine lines similar to solution spectra. This indicates that rotation is hindered in the order PS-PVME < Pahokee < Beulah. The more restricted motion of TEMPO in Pahokee and Beulah compared to PSPVME suggests a more confined location or a more rigid phase, or both. Suppression of 13C NMR Signals by Spin Probe Sorption. The CP/TOSS spectra of control sorbents (zero spin probe) and sorbents with TEMPO/HTEMPO are shown in
Figure S11. Titration with TEMPO and HTEMPO resulted in decreases in signal intensity with increasing spin probe concentration throughout most of the spectrum. TEMPO was more effective than HTEMPO due to its stronger sorption. The signal loss is attributed to the shortening of T1ρH, T1H, and T1C, especially T1ρH. Signal broadening was not observed. Percent peak suppression plots appear in Figure 2 for TEMPO and Figure 3 for HTEMPO. The axis is spin probe concentration corrected for extractive recovery. Percent peak suppression generally increases with spin probe loading, not leveling off until suppression exceeds ∼80%. Peak enhancement (negative suppression) is observed for Pahokee and Beulah at low spin probe loading. Since it seems greater than the random error, it may be due to spin probe quenching of organic free radicals or paramagnetic metals inherent to these natural solids. If so, it helps explain the poor solvent recovery of the spin probe. Note that while spin probe degradation products could be paramagnetic and also cause relaxation, plotting the axis as uncorrected spin probe concentration merely shifts the curves to the right, having no effect on the interpretation. Overall, Figures 2 and 3 show remarkably little selectivity for the different spectral regions. To help interpret the results, we compared NMR peak suppression with solvent−water partition 12818
dx.doi.org/10.1021/es302157j | Environ. Sci. Technol. 2012, 46, 12814−12822
Environmental Science & Technology
Article
Figure 4. Comparison of spin probe partitioning in solvents and putative sorbent microdomains for TEMPO. Mole fraction partition coefficient between a given organic solvent (or cellulose) and hexadecane (a). Mole fraction partition coefficient between a given domain and the alkyl domain of Pahokee (b), Beulah (c), and PS-PVME (d). Error bars are absolute errors at 95% confidence level. Only points where peak suppression was observed are included.
coefficients for solvents selected to represent certain functional group domains of NOM: toluene for ArC-C,H, anisole to represent ArC-O,N, hexadecane for alkyl, and cellulose for Oalkyl domains of NOM. We also included n-octanol because of the extensive database on free energy relationships with NOM sorption that exists for this solvent. Note that the solvent phase at equilibrium includes water; the mole fraction water contents are octanol (0.1940), anisole (0.0076), toluene (0.0024), and hexadecane (0.0006).36 Cellulose is highly hydrated in aqueous media.37 The partition coefficient of TEMPO or HTEMPO between a solvent or cellulose and water, expressed as mole fraction ratio, is given by the following:
′ − hexadecane = Korg
′ −w Korg ′ Khexadecane −w
(4)
The K′org‑w are listed in Table S2 and the K′org‑hexadecane are shown in bar-graph in Figure 4a for TEMPO and Figure 5a for HTEMPO. ′ (TEMPO) > Korg‑w ′ (HTEMPO), consistent In general, Korg‑w with the greater hydrophobic character of TEMPO than HTEMPO as reflected in the octanol−water and micelle− water partition coefficients.38,39 For both spin probes, Koctanol‑w > Khexadecane‑w, despite that most of the volume of both solvents is occupied by linear alkyl chain. This may be due to the H-bonding ability of the octanol phase, which contains abundant H2O plus the hydroxyl group.40 HTEMPO favors polar solvents and cellulose over nonpolar solvents more so than does TEMPO most likely because of the additional H-bonding afforded by the −OH group of HTEMPO.38 Lastly, K′toluene‑w > K′hexadecane‑w; this result possibly reflects either a preference for an aromatic over an aliphatic environment (even though both probes are aliphatic), or the 10-fold greater abundance of water in toluene. For sorption we assume that the spin probe concentration (moles cm−3) associated with a given functional group resonance, z, and a given sorbed concentration, Se, is equal to the peak suppression (eq 1) times a sensitivity factor (az):
′ −w = Korg moles of solute in organic phase/total moles of organic phase moles of solute in water/total moles of aqueous phase (3)
Cellulose was regarded as a collection of individual monomers and assumed to have a density of 1 g/cm3. Comparison between sorption and solvent partitioning is facilitated by choosing the “inert” alkane state as the reference state. Thus, we normalize Korg‑w ′ to that for hexadecane,
Sz = az ·σz 12819
(5) dx.doi.org/10.1021/es302157j | Environ. Sci. Technol. 2012, 46, 12814−12822
Environmental Science & Technology
Article
Figure 5. Comparison of spin probe partitioning in solvents and putative sorbent microdomains for HTEMPO. Mole fraction partition coefficient between a given organic solvent (or cellulose) and hexadecane (a). Mole fraction partition coefficient between a given domain and the alkyl domain of Pahokee (b), Beulah (c), and PS-PVME (d). Error bars are absolute errors at 95% confidence level. Only points where peak suppression was observed are included.
partition domains contribute little to spin probe sorption to PSPVME. Next we consider the natural samples (Figures 4 and 5). Neither probe shows strong preference for relaxation of ArC− C,H compared to alkyl resonances. TEMPO slightly prefers alkyl in Pahokee, but ArC−C,H in Beulah. HTEMPO gives mixed results (depending on concentration) for Pahokee, but prefers ArC−C,H in Beulah. In any case, the selectivity is small and less than the selectivity for toluene over hexadecane. TEMPO prefers alkyl over ArC−O,N in Pahokee, and the opposite in Beulah; nevertheless, there is a large uncertainty in the values, and at some concentrations there is no preference. HTEMPO prefers ArC−O,N over alkyl in Beulah, but gives mixed results in Pahokee. In all cases, the preference for ArC−O,N over alkyl is small, as is the preference for anisole over hexadecane. TEMPO shows weak preference for alkyl over O-alkyl in Pahokee, but the opposite in Beulah. HTEMPO prefers alkyl over O-alkyl in Beulah, but gives mixed results in Pahokee. Selectivity for O-alkyl over alkyl regions of SOM is weaker than, or in the opposite sense as, selectivity for partitioning between cellulose or octanol and hexadecane.
To compare with hexadecane−normalized solvent (or cellulose)−water partitioning, we select the reference domain to be the alkyl domain of each sorbent. It may be assumed that sensitivity to the electron spin depends primarily on separation distance and only weakly on the nature of the functional group. Therefore, az ≅ aalkyl and the mole fraction partition coefficient of spin probe between a given functional group domain and the chosen reference (alkyl) domain is given by σ K z′− alkyl = z σalkyl (6) In Figures 4 and 5 we compare log K′org‑hexadecane with log K′z‑alkyl at various spin probe concentrations. First we consider PSPVME. For this sorbent, TEMPO shows little selectivity for any group relative to the alkyl group. The observed selectivity is, in fact, much less than the selectivity of these probes for toluene over hexadecane and for the polar solvents over hexadecane. For HTEMPO, there is likewise almost no selectivity for nonpolar aromatic (ArC1−6) relative to alkyl carbons. There is a small selectivity for the spectral region represented by (O-alkyl and O−CH3 + −CH−Ar) relative to alkyl. But such selectivity was much less than the selectivity for cellulose over hexadecane or for octanol over hexadecane. Taken together, these results indicate that specific functional group interactions or domains segregated on a functional group identity basis (aromatic, aliphatic, nonpolar, polar) on a scale large enough for them to serve as
■
DISCUSSION Sorption to PS-PVME approaches linearity and there is almost no selectivity of spin probe-induced relaxation on the basis of functional group identity. For a sorbent that is a penetrable 12820
dx.doi.org/10.1021/es302157j | Environ. Sci. Technol. 2012, 46, 12814−12822
Environmental Science & Technology
Article
phase, a Freundlich N approaching 1.0 indicates site homogeneity and implies solid-phase dissolution, also called partitioning. At room temperature the 1:1 PS-PVME blend cast from toluene is rubbery41 and has properties consistent with largely homogeneous molecular mixing.41,42 At 320 K, the 2-dimensional wide-line separation NMR spectrum shows the structure to be nanoheterogeneous, with heterogeneities of at least 3.5 ± 1.5 nm.43 Our results, however, are consistent with the polymer blend acting as a homogeneous, rubbery sorbent, in which sorption “sites” are fleeting and molecules are “free” to explore the entire solid phase. There appears to be little microporosity associated with a pure PS phase that could confer sorption nonlinearity. Since PS-PVME is about 50% aromatic carbon, the results further teach that aromatic content alone is an insufficient condition for nonlinearity in the sorption isotherm. Sorption of the spin probes strengthens and becomes more nonlinear in the order PS-PVME < Pahokee < Beulah. Nonlinearity alone does not distinguish among the three proposed mechanisms for sorption selectivity. However, the weak selectivity for spin probe relaxation of 13C signals indicates that most spin probe molecules see a roughly average chemical environment and therefore argues against bond-based and domain-based selective interactions. Both sorption intensity (KOC) and nonlinearity (N−1) trend with specific surface area and microporosity of the sorbent, consistent with pore filling as the cause of selectivity. It is possible that the black carbon particles detected in Pahokee peat13 provide a source of micropores. However, spin probe relaxation shows little if any selectivity for nonpolar, nor polar, aromatic carbons. The use of crosspolarization instead of direct polarization excludes some (perhaps 20%19) of the nonpolar aromatic C signal from our view. Nevertheless, these results argue against a dominant role for the small black carbon component of the geosolids. The adsorption potential of any black carbon that might be in these samples may have been suppressed by “fouling” with humic substances, which in laboratory weathering experiments with soil or dissolved organic matter has been shown to decrease the apparent distribution coefficient of the weathered black carbon by as much as 2 orders of magnitude compared to the corresponding raw black carbon.44−46 In this regard, the results serve as a cautionary statement to researchers who attempt to evaluate sorption following harsh chemical, thermochemical, or extractive treatments because such treatments can remove the humic substances that coat particles. Ruling out black carbon, the results are consistent with the source of nonlinearity being the glassy state of the OM in the geosolid samples used in this study. Larsen47 proposed that Beulah-Zap contains internal closed micropores akin to those observed in glassy polymers, and DuBose and Wertz17 show that these micropores are flexible but do not collapse reversibly when the solute leaves. The evidence for the glassy, or glassy-like, character of Pahokee and Beulah with respect to various behaviors has been presented in numerous publications, and recent reviews are available.1,3 The correlation of nonlinearity with soil aromaticity found in the literature survey,2 rather than reflecting the black carbon content, could be due instead to an increase in matrix stiffness (manifested as glass-to-rubber transition temperature) resulting from incorporation of aromatic rings into the backbone, as is known for polymers. The overall conclusion of this study is that the source of sorption specificity of these compounds is the configurations and conformations of the organic matter molecules in the sample that leave micropores in the solid, which create preferential sorption sites. While our
conclusions apply to the two geosolids used in this work, there is no reason to believe these two samples are unique. Clearly, studies of other types of geosolids are warranted in order to affirm the relative importance of one sorption mechanism over another or the coexistence of multiple mechanisms. For instance, one might anticipate that soils with high inputs of black carbon will show preferential sorption to aromatic structures.
■
ASSOCIATED CONTENT
* Supporting Information S
pH dependence of TEMPO and HTEMPO UV/visible spectra; functional group composition of sorbents; solvent−water partitioning experimental technique; procedure and data for spin probe recovery; identification of oxidation product; pore volume histogram of sorbents; organic-carbon normalized distribution coefficient of TEMPO and HTEMPO; EPR spectra of dissolved and sorbed TEMPO; 13C CP/TOSS NMR spectra of control sorbents and sorbents sorbed with TEMPO and HTEMPO; and cellulose−water and solvent−water partition coefficients. This material is available free of charge via the Internet at http://pubs.acs.org.
■
AUTHOR INFORMATION
Corresponding Author
*Phone: 203-974-8518; fax: 203-974-8502; e-mail: Joseph.
[email protected]. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS This work was supported by the National Science Foundation (CBET 0853682 and 0853950). We thank Dr. Barry Dellinger and the LSU Superfund Research Center for access to the EPR Spectroscopy Facility.
■
REFERENCES
(1) Pignatello, J. J. Dynamic interactions of natural organic matter and organic compounds. J. Soils Sediments 2012, No. 10.1007/s11368-0120490-4. (2) Chefetz, B.; Xing, B. Relative role of aliphatic and aromatic moieties as sorption domains for organic compounds: A review. Environ. Sci. Technol. 2009, 43, 1680−1688. (3) Pignatello, J. J. Interactions of anthropogenic organic chemicals with natural organic matter and black carbon in environmental particles. In Biophysico-Chemical Processes of Anthropogenic Organic Compounds in Environmental Systems; Xing, B., Senesi, N., Huang, P. M., Eds.; J. Wiley & Sons: NJ, 2011; pp 3−50. (4) Zhu, D.; Pignatello, J. J. A concentration-dependent multi-term linear free energy relationship for sorption of organic compounds to soils based on the hexadecane dilute-solution reference state. Environ. Sci. Technol. 2005, 39, 8817−8828. (5) Endo, S.; Grathwohl, P.; Haderlein, S. B.; Schmidt, T. C. Compound-specific factors influencing sorption nonlinearity in natural organic matter. Environ. Sci. Technol. 2008, 42, 5897−5903. (6) Borisover, M.; Graber, E. R. Classifying NOM-organic sorbate interactions using compound transfer from an inert solvent to the hydrated sorbent. Environ. Sci. Technol. 2003, 37, 5657−5664. (7) Mao, J.-D.; Hundal, L. S.; Thompson, M. L.; Schmidt-Rohr, K. Correlation of poly(methylene)-rich amorphous aliphatic domains in humic substances with sorption of a nonpolar organic contaminant, phenanthrene. Environ. Sci. Technol. 2002, 36, 929−936. (8) Gunasekara, A. S.; Simpson, M. J.; Xing, B. Identification and characterization of sorption domains in soil organic matter using structurally modified humic acids. Environ. Sci. Technol. 2003, 37, 852− 858.
12821
dx.doi.org/10.1021/es302157j | Environ. Sci. Technol. 2012, 46, 12814−12822
Environmental Science & Technology
Article
(31) Guidelines for The Testing of Chemicals, OECD 107, Partition Coefficient (n-octanol/water) (Flask-shaking Method); Organization for Economic Cooperation and Development: Paris, France. (32) Dixon, W. T. Total suppression of sidebands in CPMAS C-13 NMR. J. Magn. Reson. 1982, 49, 341−345. (33) Bowry, V. W.; Ingold, K. U. Kinetics of nitroxide radical trapping. 2. Structural effects. J. Am. Chem. Soc. 1992, 114, 4992−4996. (34) Ottaviani, M. F.; Garcia-Garibay, M.; Turro, N. J. TEMPO radicals as EPR probes to monitor the adsorption of different species into X xeolite. Colloids Surf., A 1992, 72, 321−332. (35) Dumestre, A.; McBride, M.; Baveye, P. Use of EPR to monitor the distribution and availability of organic xenobiotics in model soil systems. Environ. Sci. Technol. 2000, 34, 1259−1264. (36) Demond, A. H.; Lindner, A. S. Estimation of interfacial tension between organic liquids and water. Environ. Sci. Technol. 1993, 27, 2318−2331. (37) Akim, E. L. Cellulose-bellwether or old hat. Chem. Tech. 1978, 8, 676−682. (38) Almeida, L. E.; Borissevitch, I. E.; Yushmanov, V. E.; Tabak, M. Different micellar packing and hydrophobicity of the membrane probes TEMPO and TEMPOL influence their partition between aqueous and micellar phases rather than location in the micelle interior. J. Colloid Interface Sci. 1998, 203, 456−463. (39) Pegi, A.; Julijana, K.; Slavko, P.; Janez, S.; Marjeta, S. The Effect of lipophilicity of spin-labeled compounds on their distribution in solid lipid nanoparticle dispersions studied by electron paramagnetic resonance. J. Pharm. Sci. 2003, 92, 58−66. (40) Abraham, M. H.; Chadha, H. S.; Whiting, G. S.; Mitchell, R. C. Hydrogen bonding. 32. An analysis of water-octanol and water-alkane partitioning and the Δlog P parameter of Seiler. J. Pharm. Sci. 1994, 83, 1085−1100. (41) Bank, M.; Leffingwell, J.; Thies, C. The influence of solvent upon the compatibility of polystyrene and poly (vinyl methyl ether). Macromolecules 1971, 4, 43−46. (42) Caravatti, P.; Neuenschwander, P.; Emst, R. R. Characterization of polymer blends by selective proton spin-diffusion nuclear magnetic resonance measurements. Macromolecules 1986, 19, 1889−1895. (43) Schmidt-Rohr, K.; Clauss, J.; Spiess, H. W. Correlation of structure, mobility, and morphological information in heterogeneous polymer materials by 2-Dimensional Wideline-Separation NMR spectroscopy. Macromolecules 1992, 25, 3273−3277. (44) Pignatello, J. J.; Kwon, S.; Lu, Y. Effect of natural organic substances on the surface and adsorptive properties of environmental black carbon (char): Attenuation of surface activity by humic and fulvic acids. Environ. Sci. Technol. 2006, 40, 7757−7763. (45) Kwon, S.; Pignatello, J. J. Effect of natural organic substances on the surface and adsorptive properties of environmental black carbon (char): Pseudo pore blockage by model lipid components and its implications for N2-probed surface properties of natural sorbents. Environ. Sci. Technol. 2005, 39, 7932−7939. (46) Oen, A. M. P.; Beckingham, B.; Ghosh, U.; Kruså, M. E.; Luthy, R. G.; Hartnik, T.; Henriksen, T.; Cornelissen, G. Sorption of organic compounds to fresh and field-aged activated carbons in soils and sediments. Environ. Sci. Technol. 2012, 46, 810−817. (47) Larsen, J. W.; Hall, P.; Wernett, P. C. Pore structure of the Argonne premium coal. Energy Fuels 1995, 9, 324−330.
(9) Mao, J.; Schmidt-Rohr, K. Absence of mobile carbohydrate domains in dry humic substances proven by NMR, and implications for organic-contaminant sorption models. Environ. Sci. Technol. 2006, 40, 1751−1756. (10) Lattao, C.; Birdwell, J.; Wang, J. J.; Cook, R. L. Studying organic matter molecular assemblage within a whole organic soil by nuclear magnetic resonance. J. Environ. Qual. 2008, 37, 1501−1509. (11) Newman, R. H.; Tate, K. R. 13C NMR characterization of humic acids from soils of a development sequence. J. Soil Sci. 1991, 42, 39−46. (12) Golding, C. J.; Smernik, R. J.; Birch, G. F. Investigation of the role of structural domains identified in sedimentary organic matter in the sorption of hydrophobic organic compounds. Environ. Sci. Technol. 2005, 39, 3925−3932. (13) Karapanagioti, H. K.; Childs, J.; Sabatini, D. A. Impacts of heterogeneous organic matter on phenanthrene sorption: Different soil and sediment samples. Environ. Sci. Technol. 2001, 35, 4684−4690. (14) http://web.anl.gov/PCS/pcshome.html. (15) Gustafsson, O.; Bucheli, T. D.; Kukulska, Z.; Andersson, M.; Largeau, C.; Rouzaud, J.-N.; Reddy, C. M.; Eglinton, T. I. Evaluation of a protocol for the quantification of black carbon in sediments. Global Biogeochem. Cycles 2001, 15, 881−890. (16) Solum, M. S.; Pugmire, R. J.; Grant, D. M. 13C Solid-state NMR of Argonne premium coals. Energy Fuels 1989, 3, 187−193. (17) DuBose, S. B.; Wertz, D. L. X-ray analysis of coals treated with organic liquids. Detailed study of the adduct formed between pyridine molecules and Beulah-Zap lignite. Energy Fuels 2002, 16, 669−675. (18) Bertini, I.; Luchinat, C.; Aime, S. Chapter 3. Relaxation. Coord. Chem. Rev. 1996, 150, 77−110. (19) Smernik, R. J.; Oades, J. M. Effects of added paramagnetic ions on the 13C CP/MAS NMR spectrum of a de-ashed soil. Geoderma 1999, 89, 219−248. (20) Koptyug, I. V.; Bossmann, S. H.; Turro, N. J. Inversion-recovery of nitroxide spin labels in solution and microheterogeneous environments. J. Am. Chem. Soc. 1996, 118, 1435−1445. (21) Franchi, P.; Lucarini, M.; Pedrielli, P.; Pedulli, G. F. Nitroxide radicals as hydrogen bonding acceptors. An infrared and EPR study. Chem. Phys. Chem. 2002, 3, 78−793. (22) Endo, K.; Morishima, I.; Yonezawa, T. Use of a stable free radical as a NMR spin probe for studying intermolecular interactions. XIV. A proton relaxation study of the hydrogen bond involving a stable free radical. J. Chem. Phys. 1977, 67, 4760−4767. (23) Pace, M. D.; Snow, A. W. Nitroxide spin probe/label study of hydrogen bonding and probe size effects in a linear epoxy polymer. Macromolecules 1995, 28, 5300−5305. (24) Satterlee, J. D. Fundamental concepts of NMR in paramagnetic systems part II: Relaxation effects. Concept Magnetic Res. 1990, 2, 119− 129. (25) Mao, J. -D.; Xing, B.; Schmidt-Rohr, K. New structural information on a humic acid from two-dimensional 1H-13C correlation solid-state nuclear magnetic resonance. Environ. Sci. Technol. 2001, 35, 1928−1934. (26) Morishima, I.; Endo, K.; Yonezawa, T. Interaction between closed-shell and open-shell molecules. 6. H-1 and C-13 contact shifts and molecular-orbital studies on hydrogen-bond of nitroxide radical. J. Chem. Phys. 1973, 58, 3146−3154. (27) Preston, C. M.; Dudley, R. L.; Fyfe, C. A.; Mathur, S. P. Effects of variations in contact times and copper contents in a 13C CPMAS NMR study of samples of four organic soils. Geoderma 1984, 33, 245−453. (28) Lee, S.; Sung, C. S. P. Surface chemical composition analysis in polystyrene/poly(vinyl methyl ether) blend films by UV reflection spectroscopy. Macromolecules 2001, 34, 599−604. (29) Braida, W. J.; Pignatello, J. J.; Lu, Y.; Ravikovitch, P. I.; Neimark, A. V.; Xing, B. Sorption hysteresis of benzene in charcoal particles. Environ. Sci. Technol. 2003, 37, 409−417. (30) Forrester, A. R.; Hay, J. M.; Thomson, R. H. Organic Chemistry of Stable Free Radicals; Academic Press: London and New York, 1968; p 225. 12822
dx.doi.org/10.1021/es302157j | Environ. Sci. Technol. 2012, 46, 12814−12822
