The presence of paramagnetic species is a major

Transcription

1 Paramagnetic Effects on Solid State Carbon-13 Nuclear Magnetic Resonance Spectra of Soil Organic Matter ABSTRACT The effects of paramagnetic species on solid state 13 C nuclear magnetic resonance (NMR) spectra were quantified in a series of doping experiments. The degree of signal loss caused by paramagnetic metals was shown to depend not only on the quantity, but also on the nature of the paramagnetic species, as well as the intimacy of contact with the organic substrate and the type of NMR experiment. Two mechanisms of signal loss were distinguished signal loss via loss of magnetic field homogeneity, which affects all 13 C nuclei in a sample, and signal loss via interaction between electronic and nuclear spins, the effects of which were localized to the close environment of the paramagnetic species. Loss of field homogeneity is important for manganese species, but not for copper species, and is equally important for both cross polarization and Bloch decay experiments. The interaction between electronic and nuclear spins is highly dependent on the spin-lattice relaxation rate constant of the free electron (T 1 e), as cations with very short T 1 e values (e.g., Pr 3 ) cause less signal loss than cations with longer T 1 e values (e.g., Cu 2,Mn 2 ). Cross polarization spectra are shown to be more susceptible than Bloch decay spectra to this mechanism of signal loss. Signal loss and increased relaxation rates brought about by paramagnetic species can be used to provide infor- mation on soil organic matter (SOM) heterogeneity in the submicron range. This is demonstrated for SOM doped with paramagnetic cations where selective signal loss and increased relaxation rates are used to determine the nature of cation exchange sites. The presence of paramagnetic species is a major impediment to the use of solid state 13 C NMR spectroscopy for the characterization of soil organic matter in whole soils and some soil fractions. For some soils, no NMR signal at all can be obtained. Even when a spectrum can be obtained, the question remains as to whether the relative intensities of the various reso- nances accurately reflect the distribution of functional groups present in the sample, given that some functional groups may be more affected than others by the presence of paramagnetic impurities. Most studies on the effects of paramagnetic species on solid state 13 C NMR spectroscopy of soil organic matter have concentrated on methods for improving the quality of spectra. Early attempts to reduce the influence of iron(iii) minerals via reduction with dithionite met with mixed success (Oades et al., 1987; Arshad et al., 1988). The use of HF, which removes the majority of the mineral fraction, has proven more reliable (Preston et al., 1989; Skjemstad et al., 1994; Schmidt et al., 1997). However, in soils such as the mineral horizons of Spodosols, in which much of the organic matter is present in the form of organo mineral complexes, HF treatment Department of Soil and Water, Waite Agricultural Research Institute, The University of Adelaide, Glen Osmond SA 5064, Australia. Received 2 June *Corresponding author (ronald.smernik@adelaide. edu.au). Published in J. Environ. Qual. 31: (2002). Ronald J. Smernik* and J. Malcolm Oades 414 can result in the loss of carbon as water-soluble material once the clay minerals have been removed (Dai and Johnson, 1999). In this paper, we present an overview of our recent work on the effects of paramagnetic species on solid state 13 C NMR spectroscopy of soil organic matter (Smernik and Oades, 1999, 2000a c; Smernik et al., 2000). We explore the various mechanisms through which paramagnetic materials affect solid state 13 C NMR spectra, in particular with respect to the effects on quan- titation. We also describe a number of experiments in which paramagnetic effects are actually turned to advan- tage, providing additional information on the structure of soil organic matter. MATERIALS AND METHODS Samples were prepared as described in Smernik and Oades (1999, 2000a,c) and Smernik et al. (2000). Solid state 13 C NMR spectra were obtained at a 13 C frequency of 50.3 MHz on a Varian (Palo Alto, CA) Unity 200 spectrometer. Samples were packed in a 7-mm-diameter cylin- drical zirconia rotor with Kel-F end-caps and spun at Hz in a Doty Scientific (Columbia, SC) MAS probe. Free induction decays were acquired with a sweep width of 40 khz; 1216 data points were collected over an acquisition time of 15 ms. All spectra were zero-filled to 8192 data points and processed with a 50-Hz Lorentzian line broadening and a s Gaussian broadening. Chemical shifts were externally referenced to the methyl resonance of hexamethylbenzene at ppm. The 13 C cross polarization with magic angle spinning (CPMAS) and Bloch decay NMR spectra were acquired using a standard pulse sequence (Wilson, 1987). A 1-ms contact time was used for CPMAS spectra. Spin counting calculations were performed using the method of Smernik and Oades (2000a,b). Glycine (analytical reagent grade; Ajax Chemicals, Liverpool, NSW, Australia) was used as an external intensity standard. The T 1 H and T 1 H values were determined as described in Smernik and Oades (2000a). Proton spin relaxation editing (PSRE) was performed as described in Smernik et al. (2000). RESULTS AND DISCUSSION Quantifying Paramagnetic Signal Loss and Increased Relaxation Rates Paramagnetic materials have been shown to influence solid state 13 C NMR spectra by causing (i) signal loss and broadening (Arshad et al., 1988; Munson and Haw, 1990; Oades et al., 1987; Preston et al., 1989; Randall et al., 1995; Schmidt et al., 1997; Skjemstad et al., 1994; Snape et al., 1989; Vassallo et al., 1987), (ii) increased Abbreviations: CP, cross polarization; CPMAS, cross polarization with magic angle spinning; NMR, nuclear magnetic resonance; PSRE, proton spin relaxation editing; SOM, soil organic matter.

2 SMERNIK & OADES: PARAMAGNETIC EFFECTS ON SOLID STATE 13 C NMR SPECTRA 415 relaxation rates (Brown, 1982; Ganapathy et al., 1981; Table 1. Relative observability of cellulose 13 C nuclear magnetic Preston et al., 1984; Sullivan and Maciel, 1982), and/or resonance (NMR) signal in physical mixtures of cellulose and paramagnetic salts (adapted from Smernik and Oades, 2000a). (iii) changes in the chemical shift of resonances (Brough et al., 1993; Chacko et al., 1983; Ganapathy et al., 1986). Relative Concentration observability Relative Some of these paramagnetic effects are interrelated. Organic Paramagnetic of paramagnetic (cross observability For example, increased relaxation rates can bring about substrate species metal polarization) (Bloch decay) signal loss and signal broadening, although not all signal % loss and broadening is due to increased relaxation rates. Cellulose By doping organic samples with paramagnetic species, Cellulose MnCl 2 4H 2 O Cellulose MnCl 2 4H 2 O we have been able to quantify signal loss and increases Cellulose CuCl 2 2H 2 O nd in relaxation rates (Smernik and Oades, 1999; Smernik Observability (signal/mass) relative to neat cellulose sample. and Oades, 2000a,c). Through these studies we have Not determined. shown that the type and size of the paramagnetic effect depends on a number of factors including (i) the amount From these results it is clear that: of paramagnetic material, (ii) the type of paramagnetic material, (iii) the distribution of the paramagnetic matecentration of paramagnetic species, since there (i) The degree of signal loss is dependent on the con- rial and the intimacy of contact between paramagnetic species and the organic material, and (iv) the type of was a greater degree of signal loss with the higher NMR experiment. Signal broadening was also observed concentration of manganese(ii) chloride. in some of the experiments, but the degree of broadentype of paramagnetic species, since the presence (ii) The degree of signal loss is dependent on the ing was not quantified. The final paramagnetic effect listed, changes in chemical shift, has been observed of 9.8% copper resulted in negligible signal loss widely in solution NMR and is the basis for so-called while the presence of 8.1% manganese resulted lanthanide shift reagents. Lanthanide shifts have been in the loss of more than half of the NMR signal observed in solid state 13 C NMR spectra (Brough et al., of the cellulose. 1993; Chacko et al., 1983; Ganapathy et al., 1986), but (iii) The degree of signal loss is independent of the not for soil organic matter samples, and will not be distion (CP) and Bloch decay (BD) observabilities type of NMR experiment, since cross polariza- cussed further here. were virtually identical in each case. Signal Loss The potential for signal loss through interaction between the nuclear spins ( 1 H and 13 C) and the electronic There are two mechanisms via which paramagnetic spins of paramagnetic impurities is illustrated in Table materials cause signal loss in solid state 13 C NMR experi- 2. Solutions of metal chloride salts (Zn 2,Cu 2, and ments: (i) field inhomogeneity brought about by the Pr 3 ) were reacted with organic substrates (chitin and bulk magnetic properties of the material (Aime et al., pectin) to produce samples in which the paramagnetic 1996; Alemany et al., 1984; Brough et al., 1993; Dereppe cations were in close contact with the organic material and Moreaux, 1987; Preston et al., 1989) and (ii) interac- (Smernik and Oades, 2000c). As expected, no signal tion between the nuclear and electronic spins (Bovey, loss was observed for the samples cation-exchanged with 1988; Wilson, 1987). Nuclear magnetic resonance spec- nonparamagnetic Zn 2. The concentrations of Cu 2 troscopy requires a strong, stable, and homogeneous (1.7% and 11.9%) in these samples would not be exapplied magnetic field. Naturally, this field will be af- pected to cause significant signal loss through loss of fected by the magnetic properties of paramagnetic and field homogeneity, based on the results in Table 1 above. ferromagnetic minerals that occur in soil samples. The The same is also true for the Pr 3 amended samples result is loss of spectral information through signal loss (Smernik and Oades, 2000a). Thus, all signal loss in (decreased sensitivity) as well as signal broadening (de- these samples can be attributed to interaction between creased resolution). Nuclear spins are influenced by the the nuclear and electronic spins. spin of unpaired electrons in paramagnetic species, just The results in Table 2 show that: as they are influenced by the spins of neighboring nuclei. The difference is that the magnetic moment of the elecresonance (NMR) signal in samples doped with various cations Table 2. Observability of chitin and pectin 13 C nuclear magnetic tron is much greater than that of any nuclei, on account (adapted from Smernik and Oades, 2000c). of the much smaller size of the electron, and hence the effects are greater. Electron nuclear coupling generally Observability Organic Paramagnetic Concentration (cross Observability results in rapid relaxation of the nuclear spins, which substrate species of cation polarization) (Bloch decay) can cause signal loss and signal broadening. % The potential for signal loss through loss of magnetic Chitin Zn nd field homogeneity brought about by paramagnetic im- Chitin Cu purities is illustrated in Table 1. In these samples, the Chitin Pr nd Pectin Zn nd organic material (cellulose) and the paramagnetic salts Pectin Cu were mixed dry. There was minimal change in T 1 H and Pectin Pr T 1 H relaxation rates of the cellulose, precluding any signal loss through decreased relaxation rates. Observability (signal/mass) relative to glycine. Not determined.

3 416 J. ENVIRON. QUAL., VOL. 31, MARCH APRIL 2002 Table 3. Relative observability of 13 C cross polarization nuclear magnetic resonance (CP NMR) signal of de-ashed soil organic matter (SOM) samples in sample doped with various cations (adapted from Smernik and Oades, 1999). Cation added Na Ca 2 Zn 2 Pr 3 Eu 3 Co 2 Ni 2 Cu 2 Fe 3 Mn 2 Concentration, % by mass Concentration, mol/g Relative CP observability, % Observability (signal/mass) relative to unamended sample. (i) The degree of signal loss is dependent on the con- to electronic spin-lattice relaxation rates (T 1 e) (Bovey, centration of paramagnetic species, since there 1988). Species with very short T 1 e values such as Pr 3, was a greater degree of CP signal loss with the Eu 3, and Co 2 are not efficient relaxation agents and higher concentration of Cu 2 and Pr 3 cations in hence caused less signal loss than species with relatively the pectin samples compared with the corre- long T 1 e values such as Cu 2,Fe 3, and Mn 2, which sponding chitin samples. have a large effect on T 1 H. Note that the presence of (ii) The degree of signal loss is dependent on the as little as 1.66% copper in the exchange sites of the type of paramagnetic species, since CP signal loss de-ashed soil organic matter resulted in the loss of 50% was much greater for the Cu 2 amended samples of the NMR signal. This contrasts with the lack of any than for the Pr 3 amended samples, despite the signal loss when 9.8% copper was present as a salt in a similar concentrations of cations. physical mixture with cellulose (Table 1). Also of note (iii) The degree of signal loss is dependent on the is the loss of 77% of the NMR signal when just 1.02% type of NMR experiment, with much greater sig- manganese is present in cation exchange sites. nal loss observed for the cross polarization (CP) spectra than for the Bloch decay (BD) spectra. Increased Relaxation Rates This difference is due to the involvement of 1 H As discussed above, paramagnetic centers must be nuclei in the CP experiment. In a CP experiment, 1 in intimate contact with organic materials in order to H nuclei are irradiated with a radiofrequency pulse. Polarization is then transferred from the influence relaxation rates. Physical mixing of paramag- 1 Htothe 13 C population during the contact time. netic salts with cellulose did not bring about such close The 1 H relaxation during this time (T contact, and did not affect cellulose relaxation rates. 1 H relaxation) reduces the intensity in the resulting 13 C Table 4 illustrates the influence of cations on T 1 H CP spectrum. Since 1 H is an isotopically abunamended with Zn 2,Cu 2, and Pr 3. and T 1 H relaxation rates for chitin and pectin samples dant nucleus, rapid relaxation of a 1 H nucleus close to a paramagnetic center can induce rapid These results show that: relaxation of neighboring 1 H nuclei through the (i) Cation amendment does not always result in inprocess of spin diffusion, hence spreading the creased relaxation rates (decreased T 1 H and signal loss effect. The Bloch decay experiment T 1 H values). Amendment with nonparamagnetic does not involve polarization transfer from the Zn 2 1 resulted in increased T 1 H and T 1 H values H population, and spin diffusion is not an impor- for both chitin and pectin, which can be attribtant process for the isotopically rare ( 1%) 13 C uted to decreases in molecular mobility brought nucleus. about by the chelating effects of the divalent We also determined the degree of signal loss brought cation. Divalent Cu 2 should also cause similar about by cation exchange of a de-ashed soil with a numeven chelation effects and trivalent Pr 3 may cause ber of nonparamagnetic and paramagnetic cations (Smerof stronger chelation effects. Thus, the strength nik and Oades, 1999). Table 3 confirms that cation exand any paramagnetic effect (to decrease T 1 H change with nonparamagnetic cations (Na,Ca 2, and T 1 H values) should be gauged against the Zn 2 ) did not affect observability. For the other paramagnetic Zn 2 amended samples, rather than against the cations, the degree of signal loss was related unamended samples. Table 4. The T 1 H and T 1 H relaxation rates of chitin and pectin 13 C nuclear magnetic resonance (NMR) signal in samples doped with various cations (adapted from Smernik and Oades, 2000c). Organic Paramagnetic Concentration of substrate species paramagnetic metal T 1 H T 1 H % ms Chitin 7.59 ( ) 203 ( ) Chitin Zn ( ) 310 ( ) Chitin Cu ( ) 42.9 ( ) Chitin Pr ( ) 370 ( ) Pectin 2.63 ( ) 264 ( ) Pectin Zn ( ) 650 ( ) Pectin Cu ( ) 0.90 ( ) Pectin Pr ( ) 92 (80 108) Values in parentheses represent lower and upper bounds, respectively, of the 95% confidence interval.

4 SMERNIK & OADES: PARAMAGNETIC EFFECTS ON SOLID STATE 13 C NMR SPECTRA 417 (ii) Paramagnetic decreases in T 1 H and T 1 H values are dependent on the concentration of paramagnetic cation, since much larger decreases were observed for the Cu 2 amended pectin (8.3% Cu) than for the Cu 2 amended chitin (1.7% Cu). (iii) Pr 3 is much less efficient than Cu 2 at decreasing T 1 H and T 1 H values. This is because Cu 2 and Pr 3 have very different electronic spin-lattice relaxation rates (T 1 e). Species with very short T 1 e values such as Pr 3 are relatively inefficient at decreasing nuclear relaxation rates compared with species with relatively long T 1 e values such as Cu 2. (iv) T 1 H values are more sensitive to paramagnetic species than are T 1 H values. For example, amendment of pectin with Pr 3 resulted in a decrease in T 1 H from 264 to 92 ms, whereas T 1 H increased from 2.63 to 3.48 ms, indicating that the paramagnetic effect was not large enough to overcome the chelation effect. Turning Paramagnetic Effects to Advantage In this section we describe how we can turn signal loss and increased relaxation rates caused by paramagnetic species to advantage in the study of soil organic matter (SOM). This is based on the premise that some paramagnetic effects are localized and that SOM is heteroge- neous. As we saw above, signal loss is brought about by two mechanisms. One of these mechanisms, loss of field homogeneity, affected 13 C nuclei remote from the paramagnetic species and hence will not produce selective effects in a heterogeneous material. However, the other mechanism of signal loss was shown to operate only when there is intimate contact between the paramagnetic species and the organic substrate. Paramagnetic effects on relaxation rates were also shown to require such intimate contact. Thus in a heterogeneous sample, components with a higher concentration of paramagnetic centers will tend to have lower observabilities and more rapid relaxation rates. Selective increases in relaxation rates brought about by a nonuniform distribution of paramagnetic cations have been reported for a model sewage sludge (Pfeffer et al., 1984) and a soil humin (Preston and Newman, 1992). Selective signal loss was observed when samples of SOM were doped with paramagnetic cations via cation exchange (Smernik and Oades, 1999). Figure 1 shows that cation amendment with nonparamagnetic Zn 2 did not significantly alter the distribution of 13 C NMR resonances, whereas amendment with paramagnetic cations (Pr 3,Cu 2, and Mn 2 ) did produce observable changes. By subtracting spectra for the cation-amended samples from the spectrum for the unamended sample (Fig. 1 difference spectra), we can observe the portion of signal that is lost through paramagnetic effects. Signal loss in the Pr 3 amended sample is restricted to the immediate environment of the cation exchange sites, since Pr 3 has a very short T 1 e value. The difference spectrum for the Pr 3 amended sample, which represents the 14% Fig. 1. Carbon-13 cross polarization with magic angle spinning (CPMAS) nuclear magnetic resonance (NMR) spectra of unamended and cation-amended de-ashed soil organic matter (SOM). Difference spectra were generated by subtracting the cationamended spectra from the unamended spectrum (adapted from Smernik and Oades, 1999). of signal lost on cation amendment (Table 3), contains strong carbonyl and O-alkyl peaks, suggesting that a significant number of cation exchange sites in this mate- rial are uronic acid structures. The CP spectra for the Cu 2 and Mn 2 amended samples are more affected by signal loss. These cations have relatively long T 1 e values, and hence cause signal loss to more remote 13 C nuclei by affecting T 1 H relaxation rates. The distance over which T 1 H relaxation rates are influenced by a paramagnetic center is determined by the extent of 1 H spin diffusion and is probably 3 nm in this case (New- man and Condron, 1995; Zumbulyadis, 1983). The re- maining CP signal for the Cu 2 and Mn 2 amended samples (Fig. 1) represents structures least affected, and hence farthest from, cation exchange sites. These spectra contain proportionately more alkyl signal than the corresponding spectrum for the unamended sample. A heterogeneous distribution of paramagnetic cen- ters can also result in differentiation of relaxation rates between components. We have exploited this effect to distinguish the char component (which naturally contains a high concentration of paramagnetic organic free radicals) within ultraviolet (UV) photo-oxidized soil samples (Smernik et al., 2000). Spectra from inversion recovery experiments for a typical sample (Fig. 2) clearly show that aromatic structures in these samples undergo more rapid T 1 H relaxation than do O-alkyl and alkyl structures. By fitting plots of total integrated intensity versus recovery delay to a two-component model, we determined that, in this sample, the rapidly relaxing

5 418 J. ENVIRON. QUAL., VOL. 31, MARCH APRIL 2002 Fig. 2. Inversion-recovery spectra of a typical ultraviolet (UV) photooxidized soil. component comprises 66% of the 13 C NMR signal and is characterized by a T 1 H value of 21.0 ms. The slowly relaxing component comprises the remaining 34% of Fig. 4. Proton spin relaxation editing (PSRE) subspectra for Cu 2 the signal and is characterized by a T 1 H value of 150 ms. amended de-ashed soil organic matter (SOM) (Smernik and Oades, Subspectra for the two components were then generated 1999). using proton spin relaxation editing (PSRE). This method involves taking linear combinations of the fully relaxed netic center on T 1 C is restricted to those 13 C nuclei close and a partially relaxed inversion recovery spectrum enough to be directly affected. The effect of cation (Fig. 3). The subspectrum characterized by the shorter amendment on T 1 C relaxation rates of a de-ashed soil T 1 H value is highly aromatic and can be primarily attriba was investigated by acquiring Bloch decay spectra with uted to the char component of the sample. very short recycle delay (300 ms) (Smernik and Oades, We have also used PSRE in conjunction with paraacquired 1999). Bloch decay spectra for SOM samples are usually magnetic cation amendment in order to probe further using a recycle delay of up to 90 s (Smernik the distribution of cation exchange sites in a de-ashed and Oades, 2000b), to avoid signal loss through satura- SOM sample (Smernik and Oades, 1999). Figure 4 tion. In these experiments, we were investigating the shows that the slowly relaxing PSRE subspectrum of the Cu 2 amended SOM sample is dominated by the alkyl resonance. This subspectrum represents a highly alkyl and presumably hydrophobic component of the soil organic matter that contains relatively few cation exchange sites, and which may play an important role in the binding of organic pollutants. The scale of heterogeneity that can be detected by this method is of the order of 10 to 30 nm (Newman and Condron, 1995; Zumbulyadis, 1983). Another relaxation parameter that is affected by the presence of paramagnetic centers is T 1 C, the 13 C spinlattice relaxation rate constant. T 1 C differs from T 1 Hin that it is not affected by spin diffusion, since 13 Cisan isotopically rare nucleus. Thus the effect of a paramag- effect of cation amendment on the degree of satura- tion observed. Fig. 5. Short recycle delay 13 C Bloch decay nuclear magnetic resonance (NMR) spectra of unamended and cation-amended de-ashed Fig. 3. Generation of proton spin relaxation editing (PSRE) subspec- soil organic matter (SOM). Difference spectra were generated by tra from the linear combination of the fully relaxed and a partially subtracting the unamended spectrum from the cation-amended relaxed inversion-recovery spectrum. spectra (adapted from Smernik and Oades, 1999).

6 SMERNIK & OADES: PARAMAGNETIC EFFECTS ON SOLID STATE 13 C NMR SPECTRA 419 Figure 5 shows that the short recycle delay Bloch ten detrimental to the study of SOM using solid state decay spectrum for the unamended sample is very different 13 C NMR, they can actually be turned to advantage. from the CP spectrum of the same sample (Fig. 1). Selective paramagnetically induced signal loss and in- This difference is due to differences in T 1 C values for the creased relaxation rates can be used to probe aspects various resonances. In this sample, the alkyl resonance is of the structure of SOM, especially those related to less affected by saturation than are other resonances, heterogeneity on the submicron scale. indicating that (at least some) of the alkyl carbon is characterized by a shorter T 1 C value. Amendment with ACKNOWLEDGMENTS nonparamagnetic Zn 2 did not cause any significant This work was funded by an Australian Research Council change to the spectrum. Amendment with Pr 3 caused (ARC) grant. some signal loss in the carbonyl region. This lost signal can be ascribed to the carboxylate cation binding sites. It should be noted that the degree of signal loss is less REFERENCES than for the CP spectrum of the sample (Fig. 1), probaresolution solid state NMR in paramagnetic molecules. Coord. Aime, S., I. Bertini, and C. Luchinat Considerations on high bly because for the CP spectrum, spin diffusion through Chem. Rev. 150: the 1 H population increases the sphere of influence. Alemany, L.B., D.M. Grant, R.J. Pugmire, and L.M. Stock Amendment with Cu 2 had quite a different effect (Fig. Solid state magnetic resonance of Illinois No. 6 coal and some 5). Signal intensity increased in most spectral regions reductive alkylation products. Fuel 63: other than the alkyl region. This increase in signal intenimprove Arshad, M.A., J.A. Ripmeester, and M. Schnitzer Attempts to solid state 13 C NMR spectra of whole mineral soils. Can. 1 C that result in sity can be ascribed to decreases in T J. Soil Sci. 68: less signal saturation. The most intense resonances in Bovey, F.A Nuclear magnetic resonance spectroscopy. 2nd ed. the difference spectrum are those due carbonyl and Academic Press, San Diego, CA. O-alkyl carbons, confirming the importance of uronic Brough, A.R., C.P. Grey, and C.M. Dobson Paramagnetic ions acids as binding sites. Amendment with Mn 2 produced as structural probes in solid-state NMR: Distance measurements a broad spectrum with increased intensity in the car- in crystalline lanthanide acetates. J. Am. Chem. Soc. 115: Brown, C.E Effects of chemical modification and CoCl 2 addibonyl, aromatic, and O-alkyl regions, but a decrease tion on the 13 C NMR spectra of carnosine as a solid powder. J. Am. in the intensity of the alkyl region. Increases in signal Chem. Soc. 104: intensity can be ascribed to decreases in T 1 C, as ob- Chacko, V.P., S. Ganapathy, and R.G. Bryant C CP-MAS served for the Cu 2 amended sample. Signal loss is most NMR spectra of paramagnetic solids. J. Am. Chem. Soc. 105:5491 likely due to loss of field homogeneity, as seen for the manganese salt doped cellulose samples (see above). Dai, K.H., and C.E. Johnson Applicability of solid-state 13 C CP/ MAS NMR analysis in Spodosols: Chemical removal of magnetic materials. Geoderma 93: CONCLUSIONS Dereppe, J.-M., and C. Moreaux A limitation of 13 C CP-MAS NMR spectroscopy for the study of treated coals. Fuel 66:1008 The presence of paramagnetic materials can have a range of effects on solid state 13 C NMR spectra including Ganapathy, S., V.P. Chacko, R.G. Bryant, and M.C. Etter Carbon CP-MASS NMR and X-ray crystal structure of paramagsignal loss and increased relaxation rates. These effects netic lanthanide acetates. J. Am. Chem. Soc. 108: were quantified by doping organic materials with para- Ganapathy, S., A. Naito, and C.A. McDowell Paramagnetic magnetic impurities. The degree of signal loss was shown doping as an aid in obtaining high-resolution 13 C NMR spectra of to be dependent on the amount and type of paramag- biomolecules in the solid state. J. Am. Chem. Soc. 103: netic material, as well as the intimacy of contact with Munson, E.J., and J.F. Haw Effects of paramagnetic lanthanides on the study of carbonaceous deposits on zeolite catalysts by carthe organic substrate, and the type of NMR experiment. bon-13 solid-state nuclear magnetic resonance spectroscopy. Anal. Two separate mechanisms of signal loss were identified; Chem. 62: loss of magnetic field homogeneity and interaction be- Newman, R.H., and L.M. Condron Separating subspectra from tween nuclear and electronic spins. cross-polarisation magic-angle spinning NMR spectra by proton This study confirms that iron, being the most abun- spin relaxation editing. Solid State NMR 140:1 8. Oades, J.M., A.M. Vassallo, A.G. Waters, and M.A. Wilson dant paramagnetic metal in most soils, is the one most Characterisation of organic matter in particle size and density fraclikely to be a problem for NMR spectroscopy. However, tions from a red-brown earth by solid-state 13 C NMR. Aust. J. Soil this study also shows that manganese can have large Res. 25: detrimental effects, even at low concentrations, and that Pfeffer, P.E., W.V. Gerasimowicz, and E.G. Piotrowski Effect of copper, due to its affinity for organic matter, can also paramagnetic iron on quantitation in carbon-13 cross polarisation cause serious problems in CP spectra at low concenmental matrices. Anal. Chem. 56: magic angle spinning NMR spectroscopy of heterogeneous environ- trations. The form of the paramagnetic species is also Preston, C.M., R.L. Dudley, C.A. Fyfe, and S.P. Mathur Effects important; discrete particles of iron minerals will un- of variations in contact times and copper concentrations in a 13 C doubtedly cause signal loss, but this signal loss will be CP/MAS NMR study of samples of four organic soils. Geoderma nonselective and hence will not affect the relative sizes 33: of NMR resonances. If, however, the iron is present in Preston, C.M., and R.H. Newman Demonstration of spatial heterogeneity in the organic matter of de-ashed humin samples by organo mineral complexes, or occupies cation exchange solid-state 13 C CPMAS NMR. Can. J. Soil Sci. 72: sites, the NMR signal of carbons in close contact will Preston, C.M., M. Schnitzer, and J.A. Ripmeester A spectrobe selectively affected. scopic and chemical investigation on the de-ashing of a humin. Soil Although the effects of paramagnetic species are of- Sci. Soc. Am. J. 53:

7 420 J. ENVIRON. QUAL., VOL. 31, MARCH APRIL 2002 Randall, E.W., N. Mahieu, D.S. Powlson, and B.T. Christensen natural organic materials. Commun. Soil Sci. Plant Anal. 31:3011 Fertilization effects on organic matter in physically fractionated soils as studied by 13 C NMR: Results from two long-term field Smernik, R.J., J.O. Skjemstad, and J.M. Oades Virtual fractionexperiments. Eur. J. Soil Sci. 46: ation of charcoal from soil organic matter using solid state 13 C Schmidt, M.W.I., H. Knicker, P.G. Hatcher, and I. Kogel-Knabner. NMR spectral editing. Aust. J. Soil Res. 38: Improvement of 13 C and 15 N CPMAS NMR spectra of bulk Snape, C.E., D.E. Axelson, R.E. Botto, J.J. Delpuech, P. Tekely, soils, particle size fractions and organic material by treatment with B.C. Gerstein, M. Pruski, G.E. Maciel, and M.A. Wilson % hydrofluoric acid. Eur. J. Soil Sci. 48: Quantitative reliability of aromaticity and related measurements Skjemstad, J.O., P. Clarke, J.A. Taylor, J.M. Oades, and R.H. Newon coals by 13 C NMR: A debate. Fuel. 68: man The removal of magnetic materials from surface soils. A Sullivan, M.J., and G.E. Maciel Spin dynamics in the carbonsolid state 13 C CP/MAS NMR study. Aust. J. Soil Res. 32: nuclear magnetic resonance spectrometric analysis of coal by Smernik, R.J., and J.M. Oades Effects of added paramagnetic cross polarization and magic-angle spinning. Anal. Chem. 54:1615 ions on the 13 C NMR CP/MAS NMR spectrum of a de-ashed soil Geoderma 89: Smernik, R.J., and J.M. Oades. 2000a. The use of spin counting for Vassallo, A.M., M.A. Wilson, P.J. Collin, J.M. Oades, A.G. Waters, determining quantitation in solid state 13 C NMR spectra of natural and R.L. Malcolm Structural analysis of geochemical samples organic matter. 1. Model systems and the effects of paramagnetic by solid-state nuclear magnetic resonance spectrometry. Role of impurities. Geoderma 96: paramagnetic material. Anal. Chem. 59: Smernik, R.J., and J.M. Oades. 2000b. The use of spin counting for Wilson, M.A NMR techniques and applications in geochemistry determining quantitation in solid state 13 C NMR spectra of natural and soil chemistry. Pergamon Press, Oxford. organic matter. 2. HF treated soil fractions. Geoderma 96: Zumbulyadis, N Selective carbon excitation and the detection Smernik, R.J., and J.M. Oades. 2000c. Effects of paramagnetic cations of spatial heterogeneity in cross-polarisation magic-angle-spinning on solid state 13 C nuclear magnetic resonance (NMR) spectra of NMR. J. Magn. Res. 53: