Journal of Environmental Quality 31:388-392 (2002)
© 2002 American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America
Article
SYMPOSIUM PAPERS
Improvements in the Two-Dimensional Nuclear Magnetic Resonance Spectroscopy of Humic Substances
André J. Simpson*,a,
Myrna J. Sallouma,
William L. Kingeryb and
Patrick G. Hatchera
a Dep. of Chemistry, The Ohio State Univ., 100W. 18th Avenue, Columbus, OH 43210
b Dep. of Plant and Soil Sciences, Box 9555, Mississippi State University, MS 39762
* Corresponding author (asimpson{at}chemistry.ohio-state.edu)
Received for publication June 6, 2000.
 |
ABSTRACT
|
|---|
Understanding pollutant sorption, bioremediation of these pollutants, and their interactions with humic substances requires knowledge of molecular-level processes. New developments with nuclear magnetic resonance (NMR) experiments and labeled compounds have improved the overall understanding of these mechanisms. The advancements made with two-dimensional NMR show great promise, as structural information and hydrogen–carbon bond connectivity can be discerned. This communication presents the application of improved two-dimensional NMR methods, the double quantum filtered (DQF) correlation spectroscopy (COSY) and echo/anti-echo heteronuclear single quantum coherence (HSQC) experiments, for use in structural studies of humic substances. Both experiments were found to produce significant improvements over the conventional COSY and heteronuclear multiple quantum coherence (HMQC) experiments that have been previously employed in similar studies. The more sensitive echo/anti-echo HSQC experiment produced more cross-peaks with higher resolution when compared with the HMQC spectra. The DQF-COSY significantly suppressed the diagonal signals and allowed numerous signals previously hidden in the standard COSY experiment to be observed. These improvements will aid current characterization strategies of humic substances from soils, sediments, and water and their subsequent reactions with pollutants and microorganisms.
Abbreviations: COSY, correlation spectroscopy • DQF, double quantum filtered • HMQC, heteronuclear multiple quantum coherence • HSQC, heteronuclear single quantum coherence • NMR, nuclear magnetic resonance
|
INTRODUCTION
|
|---|
HUMIC SUBSTANCES ARE formed through the degradation of plant and animal matter and comprise a large portion of organic matter in soil and sedimentary environments that do not belong to any distinct class of compounds such as polysaccharides, nucleic acids, and polypeptides. Aiken et al. (1985) defined humic substances as "a general category of naturally occurring biogenic, heterogeneous organic substances that can generally be characterized as being yellow to black in color, of high molecular weight and refractory." Humic substances are found throughout the environment and are an essential component in many aspects of agriculture, pollutant transport, and environmental health (Stevenson, 1994; Hayes and Wilson, 1997). To understand the reactivity of humic substances in the environment, it is imperative to first understand their structures on a molecular level.
The application of two-dimensional NMR to study the structures of humic substances is still in its infancy and has only been reported in a small number of cases (Buddrus et al., 1989; Simpson et al., 1997a; Chien and Bleam, 1998; Schmitt-Kopplin et al., 1998; Kingery et al., 2000; Simpson et al., 2001). The diverse and sometimes low concentration of components often results in weak signals and/or overlapping chemical shifts. Hence, it is necessary to optimize the experimental sensitivity and suppress the breakthrough of background noise wherever possible. This communication presents the application of a double quantum filtered two-dimensional 1H–1H correlation spectroscopy (COSY) experiment and an echo/anti-echo two-dimensional 1H–13C heteronuclear single quantum coherence (HSQC) experiment to a fulvic acid isolated from a humus layer in an oak (Quercus spp.) forest. This paper does not aim to provide a guide to interpretation of the structures present in this fulvic acid sample but to highlight the advantages that these techniques, which are now used routinely to probe the structures of complex organic macromolecules such as proteins (Croasmun and Carlson, 1994), have over the more traditional techniques within the context of humic substances. For an in-depth discussion of the interpretation of humic structures from multidimensional NMR, readers are referred to Simpson et al. (2001). The main objective of this communication is to demonstrate the improvement in spectral quality that can be obtained by using the NMR experiments described here, consequently, only one fulvic acid sample is used as an example. However, these methods can be applied to characterize and make comparisons of humic substances from a variety of environments.
|
MATERIALS AND METHODS
|
|---|
The fulvic acid was isolated from the black humus layer under moss growing in an oak forest (Urah Wood, Lough Inchiquin, Kenmare, Co. Kerry, Ireland [Grid reference: V.83.62]). Detailed descriptions of the soil profile are given by Little (1994). The sequence of extracting solvents used was: water, 0.1 M sodium pyrophosphate (pH 7.0), 0.1 M sodium pyrophosphate (pH 10.6), and 0.1 M sodium pyrophosphate plus 0.1 M NaOH at pH 12.6. The humic and fulvic acids were isolated as described by Simpson et al. (1997b) and fractionated using the XAD-8 and XAD-4 resins (Rohm and Haas, Philadelphia, PA) in tandem (Hayes, 1996; Malcolm et al., 1992; Simpson, 1999). The fulvic acid fraction isolated at pH 12.6 is the sample used in this study. Approximately 60 mg of fulvic acid sample was dried at 55°C for 24 h over P2O5 and then dissolved in 1 mL of DMSO-d6 (Sigma–Aldrich, St. Louis, MO).
Nuclear magnetic resonance (NMR) experiments were carried out on a Bruker (Rheinstetten, Germany) 600 MHz DRX, a Bruker 600 MHz AMX, and a Bruker 800 MHz DRX spectrometer. Both the COSY (Aue et al., 1976) and DQF-COSY (Rance et al., 1983) were obtained with the 600 DMX instrument in 16 scans using gradient versions of the pulse sequences. We collected 1024 data points in F2 and 512 slices in F1 with a relaxation delay of two seconds. The gradient HSQC experiment was carried out on both the DMX 600 and DMX 800 spectrometers (32 scans) using echo/anti-echo–time proportional phase incrementation (TPPI) gradient selection with decoupling during acquisition (Palmer et al., 1991; Kay et al., 1992; Schleucher et al., 1994). The HMQC (128 scans) was carried out on the AMX 600 spectrometer using a standard TPPI HMQC sequence without the use of gradient refocusing (Bax et al., 1983). The heteronuclear experiments were collected with 1024 data points in F2 and 512 slices in F1, a 1J H-C of 145 Hz, and a relaxation delay of two seconds. In each case the spectra were processed with a sine squared window function (SSB 2) and projected just above the noise level.
|
RESULTS AND DISCUSSION
|
|---|
Double Quantum Filtered Correlation Spectroscopy
The COSY experiment identifies scalar or J coupling interactions between neighboring protons. Pseudo one-dimensional proton spectra of the compound under investigation run diagonally from the lower left hand corner to the upper right hand corner of the spectrum (Eich et al., 1982; Bax and Davis, 1985). The 1H–1H coupling is displayed in the spectrum as off-diagonal cross-peaks. For each cross-peak, a line can be drawn to the y axis and a second line can be drawn to the x axis. These two lines yield the chemical shifts of the interacting pair of protons. When the chemical shift of the two interacting protons is similar, they fall close to the diagonal and the cross-peaks may be masked. Double quantum filtering is a technique that can aid the suppression of signals from singlets (uncoupled protons) that fall on the diagonal and facilitates the identification of otherwise hidden couplings. In a conventional COSY spectrum, single quantum coherence (magnetization) is detected and contains information from coupled and uncoupled protons. However, during the second 90° pulse in a COSY sequence, zero and multiple quantum coherence are produced but are not normally recorded by the receiver coil. Double quantum coherence can exist in any coupled spin pair and can be observed by the receiver coil if it is converted to single quantum coherence by an additional 90° pulse. This reconverted double quantum coherence can then be observed alone if pulse-field gradients (weak pulses of the magnetic field) or phase cycling are used to defocus the original and unwanted single quantum coherence so it can no longer be detected. The resulting spectrum therefore contains information that has been "double quantum filtered" and only contains a signal from coupled protons. Although the DQF spectrum is "cleaner," the sensitivity of the experiment is less than that of conventional COSY spectrum and may require additional spectrometer time.
Figure 1
compares the standard gradient COSY (Fig. 1A) with the DQF-COSY (Fig. 1B) for the fulvic acid sample. The advantages of the double quantum filter are immediately apparent. A clean spectrum is produced that exhibits a reduction in diagonal influences and signals that were previously hidden are now apparent. This is most clearly seen in the region between 3.0 and 4.0 ppm (Fig. 1C). This region was totally masked by diagonal signal in the standard COSY experiment and interpretation of this region was not previously possible. However, the application of the double quantum filter produced a number of peaks that are consistent with sugar residues (Simpson, 1999; Simpson et al., 2001). This new information can then be used, in combination with other experiments such as total correlation spectroscopy (TOCSY), heteronuclear multiple bond coherence (HMBC), and nuclear Overhauser effect spectroscopy (NOESY), to formulate accurate structural assignments of humic substances (Simpson et al., 2001).
View larger version (29K):
[in this window]
[in a new window]
|
Fig. 1. Correlation spectroscopy (COSY) spectra of the fulvic acid acquired at 600 MHz. Standard COSY experiment (A), double quantum filtered correlation spectroscopy (DQF-COSY) (B), and expanded central region of the DQF-COSY highlighting couplings from sugars (C). Boxes 1 and 2 highlight regions where couplings from sugars and aromatic compounds in humic substances fall. The couplings in these structures tend to be close to the central diagonal. Note how the coupling information from these moieties is difficult to decipher in spectrum (A) but are clearly visible in spectrum (B).
|
|
The Echo/Anti-Echo Heteronuclear Single Quantum Coherence Experiment
The heteronuclear multiple quantum coherence (HMQC) experiment is one of the oldest heteronuclear experiments (Bax et al., 1983) and has been the prevalent experiment to study single H–C bond correlations in humic substances (Simpson et al., 1997a; Schmitt-Kopplin et al., 1998; Simpson et al., 2001; Kingery et al., 2000). Heteronuclear experiments such as the HMQC experiment are inverse-detection experiments that transfer the chemical shift and coupling information of the relatively insensitive nuclei, in this case 13C, to the more sensitive nuclei, such as 1H (Bax et al., 1987). As a result, sensitivity is substantially increased due to the greater natural abundance of 1H relative to 13C. Heteronuclear two-dimensional spectra do not exhibit a spectrum diagonal. Cross-peaks arise from coupling of a 13C with a 1H rather than coupling between adjacent protons. A 1H bonded directly to a 13C atom will produce a cross-peak at the point of intersection of the 1H and 13C chemical shifts.
The heteronuclear single quantum coherence (HSQC) experiment has one significant advantage over its HMQC analogue in that the 1H–1H multiplets broaden the cross-peaks in either dimension. This leads to a sensitivity and resolution enhancement in the HSQC experiment and has been elaborately demonstrated by Reynolds et al. (1997) on the natural product clionasterol. However, the HMQC experiment employs a relatively short pulse sequence. Alternatively, the HSQC requires multiple refocusing periods and the experiment necessitates very precise pulse calibration to minimize artifacts and optimize sensitivity. Moreover, these proton-detected correlation experiments can be further enhanced via J couplings. In the conventional experiments, two types of coherence (magnetization) can be produced depending on the relative phases of the pulses that occur at the end of T1 (the waiting period between pulses that allow the nuclear interactions to evolve). Yet only one of these coherences is refocused, and consequently half of the observable magnetization is lost. With a sensitivity enhancement, an additional refocusing period is incorporated such that the second coherence can be observed, doubling the signal to noise. This is often referred to as echo/anti-echo selection (Palmer et al., 1991; Kay et al., 1992; Schleucher et al., 1994).
Figure 2A
displays the HMQC spectrum and Fig. 3A
illustrates the echo/anti-echo HSQC spectrum for the fulvic acid sample. In fewer scans, the HSQC (32 scans) displays a comparable number of cross-peaks as that in the HMQC (128 scans). This may partly result from a sensitivity enhancement of the HSQC experiment itself, but the application of gradient refocusing used in the HSQC described here will result in a "cleaner" spectrum. Gradients in the inverse experiments can be used to help defocus signals from protons attached to 12C. However, if they are not employed (as in the HMQC experiment, described here), the unwanted proton signals are suppressed by a less efficient method of phase cycling, and "T1 noise streaks" from very intense signals are observed. This explains the breakthrough of noise from the solvent peak extending from the top to the bottom on the HMQC spectrum that begins at 2.5 ppm. The resolution gain in the HSQC spectrum can be best observed in the central region where the crowding of sugars, amino acid side chains, methoxy, and cross-peaks from units adjacent to ester and ether occur and complicate the interpretation (Fig. 3B). The peaks on the HMQC (Fig. 2A) tend to form regions rather than the more discrete singularities observed in the HSQC (Fig. 3).
View larger version (20K):
[in this window]
[in a new window]
|
Fig. 2. Full spectrum (A) and expanded central region (B) of the heteronuclear multiple quantum coherence (HMQC) spectrum acquired at 600 MHz using 128 scans. Region 1 is consistent with C–H bonds from various aliphatic structures including fatty acids and amino acids side chains. Region 2 is consistent with units in amino acids, sugars, methoxy, and methylene groups next to ester and ether groups. Region 3 is consistent with aromatic fragments (Simpson et al., 2001).
|
|
View larger version (17K):
[in this window]
[in a new window]
|
Fig. 3. Echo/anti-echo heteronuclear single quantum coherence (HSQC) of the fulvic acid acquired at 600 MHz using 32 scans. Full spectrum (A) and expanded central region (B).
|
|
The echo/anti-echo HSQC experiment was also carried out at 800 MHz. In this experiment, the hard pulse used in the second refocusing period was replaced by four adiabatic (frequency-swept) "soft" pulses (Titman et al., 1990) to provide a more efficient refocus of the second coherence. At 800 MHz, the carbon atoms are spread over a wide range of frequencies and the adiabatic pulses cover a wide frequency range that allows efficient refocusing of all the different carbon nuclei. At 600 MHz the width of a hard pulse is sufficient to refocus all the carbons and therefore the adiabatic pulses are not employed at this field strength. As apparent from Fig. 4A
, the sensitivity of the higher field produces a far superior spectrum in which a multitude of cross-peaks not apparent at 600 MHz are observed. Closer inspection of the central region (Fig. 4B) demonstrates that although these extra cross-peaks are congested and the contours become partially agglomerated, the vast majority of cross-peaks are visible as resolved singularities that can be distinguished within this region. This is a major advantage over the HMQC experiment, which tends to produce clusters of cross-peaks that are agglomerated and difficult to discriminate (see Fig. 2B).
View larger version (23K):
[in this window]
[in a new window]
|
Fig. 4. Echo/anti-echo heteronuclear single quantum coherence (HSQC) of the fulvic acid acquired at 800 MHz using 32 scans. Full spectrum (A) and expanded central region (B).
|
|
|
CONCLUSIONS
|
|---|
The incorporation of double quantum filtering to the COSY experiment and use of the sensitivity enhanced HSQC in place of the HMQC experiment provide significant improvements in the two-dimensional NMR of humic substances. The removal of signals from uncoupled protons in the COSY experiment reduces the diagonal signal and reveals numerous peaks that were previously masked. The enhanced HSQC illustrates better sensitivity and resolution than the HMQC experiment and allows the identification of peaks that were either cluttered in the HMQC spectra or not apparent. These techniques show considerable improvements for the study of humic substances over the more commonly used experiments and the authors recommend they be used as replacements for the HMQC and COSY experiments in the study of humic substances from soil, sedimentary, and aquatic environments.
|
ACKNOWLEDGMENTS
|
|---|
We thank Dr. Charles Cottrell of the Campus Chemical Instrumentation Center, The Ohio State University, for NMR data acquisition and advice, and Dr. Michael Hayes, Limerick University, for guidance in sample fractionation and isolation. The Natural Science and Engineering Research Council (NSERC) of Canada provided a Postdoctoral Fellowship to M.J. Salloum. Financial support for this research was provided by the U.S. Department of Energy, Office of Biological and Environmental Research, Joint Bioremediation program, Grant no. DE-FG02-97ER62356 and the Mississippi Agricultural and Forestry Experiment Station and Remote Sensing Technology Center, Mississippi State.
|
REFERENCES
|
|---|
- Aiken, G.R., P. MacCarthy, R.L. Wershaw, and D.M. McKnight. 1985. An introduction to humic substances in soil, sediment and water. p. 1–9. In G.R. Aiken et al. (ed.) Humic substances in soil, sediment, and water. Geochemistry, isolation and characterization. John Wiley & Sons, New York.
- Aue, W.P., E. Bartholdi, and R.R. Ernst. 1976. Two-dimensional spectroscopy. Application to nuclear magnetic resonance. J. Chem. Phys. 64:2229–2246.
- Bax, A., and D.G. Davis. 1985. MLEV-17 based two-dimensional homonuclear magnetization transfer spectroscopy. J. Magn. Reson. 65:355–360.[ISI]
- Bax, A., R.H. Griffey, and B.L. Hawkins. 1983. Correlation of proton and nitrogen-15 chemical shifts by multiple quantum NMR. J. Magn. Reson. 55:301–315.[ISI]
- Bax, A., L.G. Marzilli, and M.F. Summers. 1987. New insights into the solution behavior of coblamin studies of the base-off form of coenzyme B12 using modern two-dimensional NMR methods. J. Am. Chem. Soc. 109:566–574.
- Buddrus, J., P. Burba, H. Herzog, and J. Lambert. 1989. Quantitation of partial structures of humic substances by one and two dimensional solution 13C nuclear magnetic resonance. Anal. Chem. 61: 628–631.
- Chien, Y., and W. Bleam. 1998. Two dimensional NOESY nuclear magnetic resonance study of pH dependent changes in humic acid conformation in aqueous solution. Environ. Sci. Technol. 32:3653– 3657.
- Croasmun, W.R., and R.M.K. Carlson. 1994. Two dimensional NMR spectroscopy: Applications for chemists and biochemistry. VHC Publ., New York.
- Eich, G., G. Bodenhausen, and R.R. Ernst. 1982. Coherence transfer by isotropic mixing: Application to proton correlation spectroscopy. J. Am. Chem. Soc. 104:3731–3732.
- Hayes, M.H.B., and W.S. Wilson. 1997. Humic substances, peats and sludges, health and environmental aspects. The Royal Soc. of Chem., Cambridge.
- Hayes, T.M. 1996. Isolation and characterization of humic substances from soil and the soil solution, and their interactions with anthropogenic organic chemicals. Ph.D. thesis. The Univ. of Birmingham, England.
- Kay, L.E., P. Keifer, and T. Saarinen. 1992. Pure absorption gradient enhanced heteronuclear single quantum correlation spectroscopy with improved sensitivity. J. Am. Chem. Soc. 114:10663–10665.
- Kingery, W.L., A.J. Simpson, M.H.B. Hayes, M.A. Locke, and R.P. Hicks. 2000. Structural analysis of humic substances by homonuclear and heteronuclear two-dimensional NMR. Soil Sci. 165:483– 494.
- Little, D.J. 1994. Occurrence and characteristics of podzols under oak woodland in Ireland. Ph.D. thesis. The Natl. Univ. of Ireland, Dublin.
- Malcolm, R.L., E.M. Thurman, and G.R. Aiken. 1992. Quantitative evaluation of XAD-8 and XAD-4 resins in tandem for removing organic solutes from water. Environ. Int. 18:597–607.
- Palmer, A.G., J. Cavanagh, P.E. Wright, and M. Rance. 1991. Sensitivity improvement in proton-detected two-dimensional heteronuclear correlation spectroscopy. J. Magn. Reson. 93:151–170.
- Rance, M., O.W. Soerensen, G. Bodenhausen, G. Wagner, R.R. Ernst, and K. Wuethrich. 1983. Improved spectral resolution in COSY proton NMR spectra of proteins via double quantum filtering. Biochem. Biophys. Res. Commun. 117:479–485.[ISI][Medline]
- Reynolds, W.F., M. Stewart, L. Tay, M. Yu, R.G. Enriquez, D.M. Estwick, and K.O. Pascoe. 1997. Comparison of 13C resolution and sensitivity of HSQC and HMQC sequences and application of HSQC-based sequences to the total 1H and 13C spectral assignment of clionasterol. Magn. Reson. Chem. 35:455–462.
- Schleucher, J., M. Schwendinger, M. Sattler, P. Schmidt, O. Schedletzky, S.J. Glaser, O.W. Sorensen, and C. Griesinger. 1994. A general enhancement scheme in heteronuclear multidimensional NMR employing pulsed field gradients. J. Biomol. NMR 4:301–306.[ISI][Medline]
- Schmitt-Kopplin, P., N. Hertkorn, H.-R. Schulten, and A. Kettrup. 1998. Structural changes in a dissolved soil humic acid during photochemical degradation processes under O2 and N2 atmosphere. Environ. Sci. Technol. 32:2531–2541.
- Simpson A.J. 1999. The structural interpretation of humic substances isolated from podzols under varying vegetation. Ph.D. thesis. The Univ. of Birmingham, England.
- Simpson, A.J., R.E. Boersma, W.L. Kingery, R.P. Hicks, and M.H.B. Hayes. 1997a. Applications of NMR spectroscopy for studies of the molecular compositions of humic substances. p. 46–63. In M.H.B. Hayes and W.S. Wilson (ed.) Humic substances, peats and sludges, health and environmental aspects. The Royal Soc. of Chem., Cambridge.
- Simpson, A.J., J. Burdon, C.L. Graham, N. Spencer, M.H.B. Hayes, and W.L. Kingery. 2001. Interpretation of heteronuclear and multidimensional NMR spectroscopy of humic substances. Eur. J. Soil Sci. 52:495–509.
- Simpson, A.J., B.E. Watt, C.L. Graham, and M.H.B. Hayes. 1997b. Humic substances from podzols under oak forest and a cleared forest site. I. Isolation and characterization. p. 73–83. In M.H.B. Hayes and W.S. Wilson (ed.) Humic substances, peats and sludges, health and environmental aspects. The Royal Soc. of Chem., Cambridge.
- Stevenson, F.J. 1994. Humus chemistry: Genesis, composition, reactions. 2nd ed. John Wiley & Sons, New York.
- Titman, J.J., A.L. Davis, E.D. Laue, and J. Keeler. 1990. Selection of coherence transfer pathways using inhomogeneous adiabatic pulses: Removal of zero-quantum coherence. J. Magn. Reson. 89: 176–183.
TOP
ABSTRACT
INTRODUCTION
