JEQ Grow Your Career With ASA
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (25)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Dria, K. J.
Right arrow Articles by Hatcher, P. G.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Dria, K. J.
Right arrow Articles by Hatcher, P. G.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Dria, K. J.
Right arrow Articles by Hatcher, P. G.
Related Collections
Right arrow Humic Substances
Right arrow Soil Analysis
Right arrow Soil Methods/Instrumentation
Right arrow Soil Chemistry
Journal of Environmental Quality 31:393-401 (2002)
© 2002 American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America

Article
SYMPOSIUM PAPERS

Solid-State Carbon-13 Nuclear Magnetic Resonance of Humic Acids at High Magnetic Field Strengths

Karl J. Driaa, Joseph R. Sachlebenb and Patrick G. Hatcher*,a

a Department of Chemistry, The Ohio State University, 100 W. 18th Ave., Columbus, OH 43210
b The Campus Chemical Instrumentation Center, The Ohio State University, 176 W. 19th Ave., Columbus, OH 43210

* Corresponding author (hatcher{at}chemistry.ohio-state.edu)

Received for publication April 2, 2001.

   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Use of solid-state 13C nuclear magnetic resonance (NMR) spectroscopy has become commonplace in studies of humic substances in soils and sediments, but when modern high-field spectrometers are employed, care must be taken to ensure that the data obtained accurately reflect the chemical composition of these complex materials in environmental systems. In an effort to evaluate the quality of solid-state 13C NMR spectra obtained with modern high-field spectrometers, we conducted a series of experiments to examine spectra of various humic acids taken under a variety of conditions. We evaluate conditions for obtaining semiquantitative cross polarization magic angle spinning (CPMAS) 13C NMR spectra of humic acids at high magnetic field and spinning frequency. We examine the cross polarization (CP) dynamics under both traditional and ramp CP conditions on Cedar Creek humic acid. Fitted equilibrium intensities from these CP dynamic studies compare to within 3.4% of the intensities determined from a Bloch decay spectrum of the same sample. With a 1-ms contact time, ramp CP and traditional CP spectra were acquired on this sample and were found to compare to within 5.4% of the Bloch decay spectrum; however, the signal-to-noise ratio per hour of data acquisition was found to double under ramp CP conditions. These results demonstrate the power of applying modern solid-state NMR techniques at high magnetic field strengths. With these techniques, high-quality, semiquantitative spectra can be quickly produced, allowing the application of solid-state NMR techniques to more environmentally relevant samples, especially those where the quantity is limited.

Abbreviations: BD, Bloch decay • CP, cross polarization • CPMAS, cross polarization magic angle spinning • CSA, chemical shift anisotropy • CT, contact time • CW, continuous wave • HA, humic acid • I0, thermal equilibrium magnetization • MAS, magic angle spinning • NMR, nuclear magnetic resonance • rf, radio frequency • rms, root mean square • SN, signal-to-noise ratio • SN/h, signal-to-noise ratio per hour • SOM, soil organic matter • T1, spin lattice relaxation time • TPPM, two pulse phase modulated


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
HUMIC ACID (HA), operationally defined as the base-soluble, acid-insoluble fraction of geoorganic materials, is a complex, polydisperse, polymeric mixture, whose properties derive from those of the parent material (Stevenson, 1994). A multitude of techniques have been used to examine the chemistry of plant precursors and various HAs (Schnitzer and Khan, 1972; Stevenson, 1994; Kögel-Knabner, 1997). All of these techniques have provided varying levels of structural detail; however, in recent years, solid-state 13C nuclear magnetic resonance (NMR) spectroscopy has made a tremendous impact on our knowledge of the structure of HAs (Wilson, 1987; Kögel-Knabner, 1997). This technique allows one to quickly and nondestructively obtain semiquantitative structural information on the carbon types present in soil organic matter (SOM) (Wilson, 1987; Kinchesh et al., 1995a; Preston, 1996; Smernik and Oades, 2000a,b).

In early studies of SOM, this semiquantitative structural information was extracted from 13C NMR spectra acquired with cross polarization magic angle spinning (CPMAS) and high power continuous wave (CW) 1H decoupling at a low magnetic field strength (100 MHz 1H, 25 MHz 13C) and low sample reorientation frequencies (3–4 kHz) (Wilson, 1987). Under these conditions, Hartmann–Hahn matching was easily obtained because the low spinning frequencies were too small to affect the heteronuclear dipolar coupling responsible for cross polarization (CP). Similarly, off-resonance decoupling effects were insignificant, and spectral interference due to spinning sidebands was avoided because the available spinning frequencies were greater than the frequency width of the chemical shift anisotropy (CSA) at this low magnetic field strength.

Since the establishment of these conditions for producing semiquantitative structural information, NMR technology has advanced considerably. Solid-state 13C CPMAS spectra are now routinely acquired at magnetic field strengths three to five times those used previously. While this brings considerable advantages in increased signal-to-noise ratio (SN) and resolution, it reintroduces interference from spinning sidebands (Fründ and Lüdemann, 1994) and increases off-resonance effects in 1H decoupling (Sachleben et al., 1996). New probe technology has increased available spinning frequencies to 30 kHz, which circumvent the sideband problems; however, CP efficiency decreases considerably as spinning frequencies exceed 10 kHz, seriously affecting the SN in fast spinning spectra. To circumvent this problem, many research groups have used higher magnetic field strengths but lower spinning frequencies and have attempted to account for the presence of sidebands in their spectra (Kinchesh et al., 1995b; Augris et al., 1997; Conte et al., 1997a,b; Dai and Johnson, 1999; Nierop et al., 1999, 2001). Unfortunately, many times, the sidebands can be buried under other peaks making quantification very difficult, especially since CSA parameters are not known a priori for most sites in SOM.

In the last decade, solutions to the problems introduced by collecting spectra at higher magnetic fields and spinning frequencies have been worked out in the NMR literature. Advances in probe design have lead to commercially available magic angle spinning (MAS) probeheads capable of spinning 4-mm rotors up to frequencies of 15 kHz. Variable amplitude cross-polarization techniques, ramp CP (Metz et al., 1994), and adiabatic sweep CP (Hediger et al., 1994) allow efficient CP at high spinning frequencies and have been applied to SOM samples (Cook et al., 1996; Cook and Langford, 1998; Chefetz et al., 2000; Knicker, 2000a,b). Two pulse phase modulated (TPPM) decoupling (Bennett et al., 1995) significantly improves decoupling efficiency at high magnetic field strengths and spinning frequencies. A more detailed description of solid-state NMR is given in the appendix to this paper.

In this paper, we examine MAS 13C NMR spectra acquired at high magnetic field strengths and spinning frequencies, revisiting the studies of Fründ and Lüdemann (1994), but using modern CP and decoupling methods. We measure the contact time dependence of the relative spectral intensities for both traditional CP and ramp CP on a HA sample. We compare the relative intensities acquired by Bloch decay (BD), traditional CP, and ramp CP and compare the efficiencies of these three techniques. Finally, CPMAS 13C NMR spectra of several different HA samples, taken under the established low field, low spinning frequency conditions, are compared with those taken with the new techniques. This study allows us to compare the quality of the spectra produced between the established low-field conditions and the ones we propose, and determine the conditions necessary and our ability to quantitate the spectra. These results directly influence routine 13C NMR acquisition of SOM.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Source and Isolation of Humic Acids
Humic acids were prepared by the standard method involving extraction with 0.1 M NaOH followed by precipitation after adjusting the solution pH to 2. They were redissolved and reprecipitated to purify. Details of the origin and preparation are given by Matsuda and Schnitzer (1971) for the Diluvial HA from Japan; by Hatcher et al. (1983) for both the Mangrove Lake HA, which was extracted from the sediment in Mangrove Lake, Bermuda, collected at a depth of 370 to 390 cm below the water sediment interface, and the Cedar Creek HA collected from a Minnesota peat; and by Hatcher (1987) for Mt. Rainier HA, which was extracted from a degraded wood sample collected on the slopes of Mt. Rainier, Washington. The elemental compositions are 61.42% C, 3.22% H, and 0.83% N for the Diluvial HA, 50.97% C, 6.55% H, and 4.21% N for the Mangrove Lake HA, and 54% C, 5.1% H, and 2.2% N for the Cedar Creek HA. Elemental data are not available for the Mt. Rainier HA.

High Field Carbon-13 Solid-State Nuclear Magnetic Resonance Experiments
High field solid-state 13C NMR experiments were performed on a Bruker (Billerica, MA) DMX 300 NMR spectrometer. This spectrometer operates at a 1H frequency of 300 MHz and a 13C frequency of 75.48 MHz. Depending on density of the sample and the amount of sample available, approximately 100 mg dry weight per sample was placed in a 4-mm (outside diameter) NMR rotor with a Kel-F cap (3M, Minneapolis, MN). Bloch decay experiments (Fig. 1A) were performed by applying a 90° 13C pulse of short duration (<5 µs) to excite the spins and were detected in the presence of TPPM decoupling. A recycle delay of 15 s was used to ensure relaxation of the carbons with the longest 13C spin lattice relaxation time (T1) as determined from a 13C inversion recovery experiment. The 13C inversion recovery experiment was performed by using a modification of the sequence proposed by Tortia (1978). Magnetization is transferred to the 13C from 1H by ramp CP and then the resulting 13C magnetization is stored antiparallel to the large external magnetic field. Longitudinal relaxation during an incremented delay brings this magnetization parallel to the magnetic field, which is probed by a 90° pulse on 13C (Torchia, 1978; Wilson, 1987). The incremented delay was varied logarithmically from 1 µs to 32 s. The 1H spin temperature alternation as well as the cyclops phase cycle (Ernst et al., 1987) on the 13C ramp pulse was used. This method can result in disappearance of carbons that have low CP efficiency and long T1 values (Barrie et al., 1993; Jurkiewicz and Maciel, 1995). Typically in geo-organic materials, such carbons would be graphitic, and the elemental analyses of our samples suggest that graphic carbon is not abundant in our samples. The BD experiment suffered from the interference of a 13C background, as discussed by Smernik and Oades (2000a)(b). To remove this interference, a background spectrum of an empty spinner was acquired and subtracted from the spectrum of the HA.



View larger version (10K):
[in this window]
[in a new window]
  		Fig. 1. Pulse sequences used in this study: (A) Bloch decay, (B) ramp cross polarization (CP), and (C) CP, where CT is contact time and TPPM is two pulse phase modulated.

 
High field ramp CPMAS 13C NMR experiments (Fig. 1B) were performed using high speed spinning with a ramp CPMAS pulse program and TPPM decoupling. The ramp CPMAS pulse program was employed as described by Metz et al. (1994) with the carbon spin-locked power ramped linearly from one half its final value. Background CP spectra were acquired and found to be insignificant. Traditional CPMAS experiments (Fig. 1C) were performed with constant amplitude spin lock fields that were reoptimized for maximum polarization transfer. Hartmann–Hahn matching and TPPM parameters were optimized on polycrystalline glycine under similar conditions as those employed on the HAs. The carboxyl shift of glycine (176.03 ppm) provided a convenient secondary reference for all the solid-state NMR spectra.

The contact time dependence of the CP was measured by performing a series of experiments in which the contact time (CT in Fig. 1B and 1C) was lengthened from 20 µs to 15 ms. As will be demonstrated for our samples, a contact time of 1 to 2 ms was found to give close to quantitative results, when a recycle delay time of 1 s and a sample spinning speed of 13 kHz were used. In all presented spectra, decoupling field strengths were greater than 70 kHz and spin locking field strengths were greater than 50 kHz. Free induction decays over a sweep width of 27778 Hz with 1024 complex data points were collected and zero filled to 8192 total complex data points. A 100-Hz exponential line broadening was applied before Fourier transformation and phasing.

Low Field Cross Polarization Magic Angle Spinning Experiments
Low field CPMAS 13C NMR spectra were obtained using the traditional CPMAS pulse sequence (Fig. 1C) and high power CW proton decoupling on a Chemagnetics (Fort Collins, CO) M-100 spectrometer operating at 25.2 MHz for carbon as described previously (Hatcher, 1987). Approximately 500 mg of sample was placed in a 9-mm (outside diameter) ceramic rotor and spun at 3.5 kHz at the magic angle in a Chemagnetics probehead. The spectra were acquired using a 1-ms contact time and 0.7-s recycle delay between scans. Decoupling field strengths were approximately 50 kHz. Over a sweep width of 10000 Hz, 512 complex data points were acquired and zero-filled to 4096 complex data points prior to filtering with 75-Hz line broadening and Fourier transformation.

Relative Intensity Determination and Comparisons
All NMR spectra were integrated over the regions of 0 to 45, 45 to 60, 60 to 90, 90 to 120, 120 to 140, 140 to 160, 160 to 190, and 190 to 220 ppm and these regions are assigned to different chemical moieties in Table 1 (after Knicker and Lüdemann, 1995). However, minimal signals were observed in the 190 to 220 region for these samples, so these areas are not reported. The spectra were also integrated for aliphatic and aromatic regions for other comparisons. The aliphatic region includes the integral between 0 and 90 ppm and a portion of the integrated region between 90 and 120 ppm. Because the signal for the anomeric carbon of the carbohydrates (105 ppm) overlaps that of proton-substituted aromatic carbons, the correction suggested by Bates et al. (1991) was applied to the peak areas in this region to account for anomeric carbon atoms. If we assume that carbohydrates are mostly of the six-carbon variety and contain one anomeric carbon with a resonance at 105 ppm, and the other five carbons resonate in the 60 to 90 ppm region, then the ratio of anomeric to other carbohydrate carbons is 1:5. Thus, if we measure the area in the 60 to 90 ppm region and divide by 5, then this value corresponds to the calculated area for anomeric carbons and should represent the contribution anomeric carbons make to the area between 90 and 120 ppm. The remainder of the area in the 90 to 120 ppm region and the area between 120 and 160 ppm represent the aromatic carbons in the spectra. This correction was only applied to the reported aromatic–aliphatic areas.


View this table:
[in this window]
[in a new window]
  		Table 1. Integration region assignments (after Knicker and Lüdemann, 1995).

 
The relative intensities of two spectra were compared by calculating the root mean square (rms) deviation:

[1]
between the intensities determined from the two spectra. Ii1 and Ii2 are the two integrated intensities of the same regions, labeled by i, in the two spectra and n is the number of regions integrated. This number acts as a figure of merit for determining the relative abundances of the different types of carbon sites with the methods studied. This rms deviation suggests the size of the systematic errors in the intensity determination and can be used as an estimate of the error in the measurement.


   RESULTS AND DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
CONCLUSIONS
 REFERENCES
 
To produce semiquantitative 13C NMR spectra with high SN in the least amount of time, the effect of polarization transfer dynamics and different experimental procedures need to be evaluated. We examined the polarization dynamics under normal CP and ramped CP conditions in order to find a contact time where systematic intensity errors are minimized. We also examined the efficiency of spectral acquisition by examining the SN per hour (SN/h), defined as the signal intensity of the tallest peak in the spectrum divided by the rms noise and the acquisition time in hours, for a series of experimental conditions.

Polarization Transfer Dynamics
Polarization transfer dynamics for CP and ramped CP were examined and are shown in Fig. 2 . Two clear differences are seen in these magnetization buildup curves. The maximum 13C intensity occurs at shorter contact times in normal CP than in ramped CP and normal CP shows the presence of transient oscillations due to the quantum mechanical C–H dipolar coupling (Müller et al., 1974). These transient oscillations are known to disappear under adiabatic sweep conditions (Heidiger et al., 1994), as seen in Fig. 2, and they hinder the analysis of the normal CP buildup curves. For this reason, we only fit the result of the ramped CP experiment. Polarization transfer dynamics are commonly described as a balance between the exponential increase in the 13C magnetization with a time constant TCH and the exponential decrease in the source 1H magnetization with a time constant TH1{rho} given by the expression:



View larger version (22K):
[in this window]
[in a new window]
  		Fig. 2. Graph of region areas versus contact time for Cedar Creek humic acid (HA), obtained using (A) cross polarization magic angle spinning (CPMAS) and (B) ramp CPMAS. See Table 1 for region assignments.

 
[2] where I0 is the thermal equilibrium magnetization of 13C, CT is the contact time, and the factor 4 is the maximum theoretical enhancement in a single scan (Wind et al., 1993). Fits to this expression showed that some regions of the spectra, specifically the regions between 140 and 0 ppm, were not adequately described by a simple monoexponential polarization transfer. This is probably due to different overlapping peaks that experience different local dynamics. For this reason, we extended Eq. [2] to include biexponential polarization transfer:

where P and (1 - P) are the percent of the peak that is experiencing fast and slow polarization transfer, respectively. Our fits to the ramped CP data taken on Cedar Creek HA are shown in Table 2.


View this table:
[in this window]
[in a new window]
  		Table 2. The TCH and TH1{rho} (Eq. [2] and [3]) results from 13C ramp cross polarization magic angle spinning (CPMAS) contact time experiment of Cedar Creek humic acid (HA). The root mean square deviation between I0 and IBD = 3.4.

 
Examination of the fits presented in Table 2 show the following trends. The TH1{rho} values range from 6 to 14 ms (Table 2), with the aliphatic type carbons (0–90 ppm) having a shorter TH1{rho} than the aromatic (90–160 ppm), carboxyl (160–190 ppm), and carbonyl (190–220 ppm) carbons. The TCH values follow a similar trend. These results are consistent with the fact that both parameters decrease with increasing presence of hydrogen atoms.

Comparison of Bloch Decay and Cross Polarization Experiments
Bloch decay experiments are single pulse experiments traditionally believed to produce the most quantitative results given that appropriate parameters are chosen (Snape et al., 1989). Cedar Creek HA was again used; however, before the BD experiment could be performed, the 13C T1 had to be determined in order to prevent saturation.

The 13C T1 of Cedar Creek HA was determined using an inversion recovery pulse sequence with ramp CP and TPPM decoupling. The T1 values for all substantial peaks were calculated and are found to range from 0.69 to 2.80 s (Table 3). Peaks at 24 and 32 ppm represent the methyl and methylene carbons and had the shortest T1 values, equaling 0.69 s. The second set of peaks consists of protonated carbon peaks at 56, 104, 115, and 127 ppm, with T1 values averaging 1.4 ± 0.3 s. The third and last set of peaks comprise both protonated and unprotonated carbons with T1 values averaging 2.6 ± 0.2 s, including peaks at 72, 154, and 174 ppm. A recycle delay of 15 s was chosen on the basis of these data for the BD experiment.


View this table:
[in this window]
[in a new window]
  		Table 3. Carbon spin lattice relaxation time (T1) data of Cedar Creek humic acid.

 
The thermal equilibrium value or limiting intensity (I0) of Eq. [1] and Eq. [2] gives an estimate of the relative intensities of the carbon environments in the solid as determined by a contact time dependence experiment. Fitted values of I0 from the contact time dependence study for the ramped CP experiment on Cedar Creek HA (Fig. 2B) for each region are given in Table 2. These intensities are found to approximately equal the respective regions measured in the BD experiment (last column of Table 2). However, a few exceptions can be pointed out. Larger carboxyl region intensity (190–160 ppm), but smaller carbohydrate (90–60 ppm) and methoxy region (60–45 ppm) intensities were observed with the BD experiment. These differences may be due to area calculation errors in the BD experiment resulting from a noisy baseline or insufficient time for full relaxation of the carbohydrate and methoxy carbons. Errors in the I0 values resulting from the biexponential calculation could also cause this discrepancy. The rms deviation between the BD spectrum and fit values of I0 was 3.4% (percent of total 13C intensity in the spectra), which provides a quantitative estimate of the systematic error in this method of quantifying the data.

The BD spectrum is compared with the CP spectra at three contact times in Fig. 3 . In the rest of this paper, we assume that the BD spectrum shown in Fig. 3 correctly represents the relative 13C abundance, given in Table 2, of the carbon species observed. We compare the CP data of Cedar Creek HA with this BD spectrum to determine the systematic intensity errors in the CP spectra at different contact times.



View larger version (21K):
[in this window]
[in a new window]
  		Fig. 3. Comparison of ramp cross polarization magic angle spinning (CPMAS) 13C nuclear magnetic resonance (NMR) with different contact times (CT) to a Bloch decay NMR spectrum of Cedar Creek humic acid (HA). SN/h is the signal-to-noise ratio per hour of acquisition time.

 
From the contact time dependence data for the ramped CP data of Cedar Creek HA, we have attempted to define an optimum contact time. Our data show that the maximum total integrated area is observed using a contact time of 1 ms; however, this time is insufficient for the buildup of full aromatic, carboxyl, and carbonyl intensities, although by 2 ms aliphatic intensities have dropped considerably. When compared with the BD spectrum, a spectrum acquired with a 1 ms contact time is found to be slightly less quantitative, with an rms deviation from the BD spectrum of 5.9%, than the experiment run with a contact time of 2 ms, with an rms deviation of 5.4%. With a 3-ms contact time, the signal-to-noise ratio significantly decreases (Fig. 3). We note that more quantitative results were produced by fitting the buildup curves, where the rms deviation was 3.4%, than by using any single contact time.

The ramp CPMAS method produces results consistent with the BD spectrum, but with much higher sensitivity. Figure 3 shows the relative efficiencies of acquiring the ramped CP and BD data by reporting the SN/h. The SN/h is found to increase from 2.6 ± 0.2 for the BD data to 189 ± 5 for the 2-ms contact time ramped CP experiment.

Comparison of Ramp Cross Polarization Magic Angle Spinning and Cross Polarization Magic Angle Spinning at a High Field Strength
The results from the contact time experiment using Cedar Creek HA show that the ramped pulse sequence is a powerful technique that greatly enhances the signal-to-noise ratio. The CPMAS has a lower efficiency than ramp CPMAS (Fig. 4) under otherwise identical instrumental conditions, as can be seen by the increase of SN/h from 98 ± 5 to 186 ± 5, respectively. The areas under each of the integrated peaks in these spectra are not significantly different and both compare favorably to the BD spectrum (Table 4); however, in normal CPMAS care must be taken to avoid transient oscillations in the areas. For this reason, ramp CPMAS produces more consistent results over a wider range of contact times than normal CP and provides increased sensitivity. Cook et al. (1996) reported differences in the relative intensities with application of these two techniques, which is possibly due to transient oscillations. The agreement between our spectra is possibly fortuitous.



View larger version (18K):
[in this window]
[in a new window]
  		Fig. 4. Comparison of ramp cross polarization magic angle spinning (CPMAS) and CPMAS 13C nuclear magnetic resonance (NMR) of Cedar Creek humic acid (HA) using a DMX300 spectrometer. Contact time (CT) = 1 ms.

 

View this table:
[in this window]
[in a new window]
  		Table 4. Integrated areas from humic acid nuclear magnetic resonance (NMR) spectra. HA, humic acid; CPMAS, cross polarization magic angle spinning.

 
At high spinning frequencies, the Hartmann–Hahn matching condition breaks up into a series of sidebands with the most efficient CP occurring when {omega}H1 = {omega}C1 ± {omega}r, where {omega}H1 and {omega}C1 are the precession frequencies about the 1H and 13C spin lock fields and {omega}r is the rotation rate (Stejskal et al., 1977; Wind et al., 1988). Unfortunately, these sidebands in the Hartmann–Hahn matching condition are very narrow, making it difficult to optimize the experiment under these conditions. In addition, mismatches in the Hartmann–Hahn condition will occur when the output power of the amplifiers and the spinning frequency fluctuate, degrading signal intensity. Ramp CP is more immune to these problems and additionally corrects for the effect of radio frequency (rf) magnetic field (B1) inhomogeneity across the sample, allowing more of the sample to contribute to the observed signal. These arguments explain the observed increase in SN/h shown in Fig. 4.

Comparison of High Field Ramp Cross Polarization Magic Angle Spinning versus Low Field Cross Polarization Magic Angle Spinning
Ramp CPMAS at a high field strength (300 MHz, 1H) using high sample spinning frequencies (13 KHz) and a 4-mm-o.d. rotor are compared with CPMAS spectra obtained at low field strengths (100 MHz, 1H) with a 9-mm-o.d. rotor at optimized conditions. The 13C NMR spectra were compared at both field strengths for many different types of samples including a dominantly aromatic HA (Diluvial HA from Japan, Fig. 5A ; described by Matsuda and Schnitzer, 1971), a dominantly aliphatic HA (Mangrove Lake HA from Bermuda, Fig. 5B; described by Hatcher et al., 1983), and a Mt. Rainier HA extracted from degraded wood (Fig. 5C; described by Hatcher, 1987).



View larger version (22K):
[in this window]
[in a new window]
  		Fig. 5. Comparison of 13C NMR spectra obtained on M-100 and DM-X300 spectrometers of (A) Diluvial humic acid (HA) from Japan (contact time [CT] = 1 ms), described by Matsuda and Schnitzer (1971); (B) Mangrove Lake HA from Bermuda (CT = 1 ms), described by Hatcher et al. (1983); and (C) HA from Mt. Rainier degraded wood, Washington (CT = 2 ms), described by Hatcher (1987).

 
In all cases, the SN/h and resolution were much improved in the spectra acquired on the 300-MHz NMR spectrometer with high speed spinning and ramped CP. The improvement in SN/h is surprising considering that the total amount of sample being analyzed has decreased by approximately a factor of five. This can be attributed to several causes. First, improvements in spectrometer design, especially with regard to low-noise amplifiers, and the implementation of time domain oversampling and digital filtering are a major portion in this difference. The factor of three increase in field strength gives rise to a factor of 37/4 {cong} 6.84 (Hoult and Richards, 1976) increase in the SN/h. The spectra acquired on the 100-MHz NMR spectrometer were acquired in a 9-mm rotor. This implies a very large excitation–detection coil, which is very inefficient and limits spin lock and decoupling powers. Decreased spin lock radio-frequency (rf) fields reduce TH1{rho} (Wind et al., 1993) and result in decreased polarization transfer. Lower decoupling power also decreases the intensity of the decoupled spectrum, which is improved both by the higher decoupling strengths available on the 300-MHz spectrometer and the implementation of TPPM decoupling (Bennett et al., 1995; Sachleben et al., 1996). Finally, the large coil used on the 100-MHz NMR spectrometer has much larger rf inhomogeneity, resulting in only a fraction of the sample being Hartmann–Hahn matched. Ramp CP allows more of the sample to be observed. All these effects conspire to significantly reduce the SN/h observed under the old low-field conditions.

Table 4 shows a comparison of the relative intensities of these samples taken under the established low-field conditions and the new higher field conditions that we propose. Notice that the relative areas change little between data sets on equivalent samples with rms deviations less than 3.0%. This occurs even though the chemical makeup changes dramatically from sample to sample. The largest rms difference is observed with the Diluvial HA, which can be attributed to linewidth narrowing at the high field conditions, allowing more accurate area integrations. This indicates that the newly proposed conditions are providing spectra as quantitative as that from the older established conditions, but with greater SN/h of acquisition.


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
REFERENCES
 
All these results show that by using high magnetic field strengths, ramp CP, TPPM decoupling, and fast spinning, semiquantitative results with higher SN and enhanced resolution than those acquired under the older conditions can be produced. We estimate that the errors in our intensity measurements under the conditions we present here are approximately 5% of the total intensity in the spectrum. The spectra shown in this paper demonstrate that high quality experiments can be performed at high magnetic field without the complications of overlapping sidebands and poor CP and decoupling efficiency. These improved spectra allow more detailed investigation of the functional groups present in a SOM with a considerable savings in acquisition time, and they allow the investigation of smaller quantities or higher inorganic content SOM values than was previously possible. For instance, application of these techniques has allowed the analysis of samples with less than 1% organic carbon contents that would not have otherwise been possible (Dria, 2000).

Overview of Solid-State Nuclear Magnetic Resonance
Nuclear magnetic resonance of solid samples is complicated by the presence of interactions that depend upon the orientation of the molecule with respect to the magnetic field. Two of these interactions, the chemical shift anisotropy (CSA) and the C–H heteronuclear dipolar coupling, are especially important in 13C NMR of SOM. Both of these interactions broaden the NMR lines in the carbon spectrum; however, the C–H dipolar coupling can be used to improve the sensitivity of the 13C NMR experiment.

Resonant irradiation of 1H during 13C spectral acquisition is called 1H decoupling and results in the removal of the effects of the dipolar coupling from the 13C spectrum. The traditional method of performing this in the solid state is high power CW decoupling, where the resonant 1H rf field is simply turned on during 13C acquisition (Slichter, 1989). Two pulse phase modulated decoupling (Bennett et al., 1995) has recently been developed to improve the decoupling effect. At higher magnetic field strengths, a larger range of 1H chemical shifts occurs and leads to broader lines in CW decoupled spectra (Sachleben et al., 1996). Two pulse phase modulated decoupling helps remove these off-resonant effects.

Magic angle spinning removes the broadening due to the CSA and results, when combined with 1H decoupling, in "liquid-like" NMR spectra. Magic angle spinning is performed by rapidly reorienting the sample about an axis oriented at 54.7° with respect to the magnetic field. If the frequency at which the sample is spun is much greater than the frequency width of the CSA broadened line, a single narrow NMR resonance at the isotropic chemical shift results for each type of carbon in the SOM. A good rule of thumb is that the spinning frequency (in Hz) needs to be greater than {nu}C(MHz) x 100 ppm. If a sample is spun at a much lower frequency than this, additional peaks, called spinning sidebands, occur at integer multiples of the spinning frequency from the isotropic chemical shift (Maricq and Waugh, 1979; Herzfeld and Berger, 1980). These additional peaks take intensity away from the isotropic peak, can overlap with other NMR resonances, and make quantification difficult. Thus, at higher magnetic field strengths, larger spinning frequencies are necessary to remove this effect. Unfortunately, these higher spinning frequencies interfere with 1H decoupling and CP.

Cross polarization is the process of transferring magnetization from abundant, highly magnetic 1H to rare, less-magnetic 13C nuclei. This is performed by applying a 90° pulse to 1H, which rotates the 1H magnetization into the plane perpendicular to the large external magnetic field, followed by a long spin-lock pulse applied in the direction of the 1H magnetization. This spin-lock pulse prevents the 1H magnetization from precessing and locks it along the direction of the rf field. Simultaneous with the 1H spin-lock field, another long rf pulse is applied to the 13C such that the precession frequency of 13C about its rf field equals the precession frequency of 1H about its rf field: {gamma}HBH1 = {gamma}CBC1 or {omega}H1 = {omega}C1. Under this Hartman–Hahn condition, flip-flop transitions occur that change a 1H from the +1/2 to -1/2 state while 13C changes from -1/2 to +1/2 state. This allows magnetization to flow from 1H to 13C during this contact time. In a single acquisition, this process can lead to a maximum of a factor of four increase in the SN of the 13C spectrum. Typically, a factor of two increase in SN is seen per scan. However, since multiple scans are typically needed to acquire a 13C NMR spectrum of SOM and 1H T1 values are significantly shorter than those for 13C, the increased repetition rate of the CP experiment results in much larger SN/h.

Unfortunately, fast MAS interferes with the CP process. If one plots the intensity of a carbon peak in a static solid as a function of the mismatch from the Hartmann–Hahn condition, a broad Hartmann–Hahn matching curve results, which is centered at the Hartmann–Hahn condition and whose width is approximately the H–H dipolar coupling (Marks and Vega, 1996). Finding the Hartmann–Hahn condition is relatively easy in such circumstances because of the broadness of the matching curve. However, when a sample experiences fast MAS, the matching curve breaks up into a series of sidebands with peaks at {omega}H1 = {omega}C1 ± k{omega}r, where {omega}r is the rotation frequency in angular units and k is an integer. The intensity of the band occurring at the Hartmann–Hahn condition (k = 0) becomes small and most efficient CP occurs at plus or minus the spinning frequency (k = ±1) from this condition (Stejskal et al., 1977; Wind et al., 1988). These sidebands tend to be very narrow, making optimum Hartmann–Hahn matching difficult and susceptible to fluctuations in rf power and spinning frequency. Metz et al. (1994) and Hediger et al. (1994) both suggested variable amplitude CP as a solution to this problem. Under the ramp CP conditions (Metz et al., 1994), the amplitude of the carbon pulse during the contact time is ramped linearly from 1/2 its final value. This effectively sweeps across the k = ±1 sideband of the Hartmann–Hahn matching curve allowing efficient magnetization transfer to occur. This permits effective CP even if the Hartmann–Hahn condition is slightly mismatched, making this method less sensitive to power fluctuations in the amplifiers and allowing a larger volume of the MAS rotor to be observed.


   ACKNOWLEDGMENTS
 
We would like to thank Jacqueline M. Bortiatynski and Heike Knicker for analyzing the samples at The Pennsylvania State University using the 100-MHz spectrometer. This research was supported by the National Science Foundation by Grants DEB-9727057 and DEB-9727056


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




This article has been cited by other articles:


Home page
 Soil Sci.Home page
J. M. De la Rosa, H. Knicker, E. Lopez-Capel, D. A. C. Manning, J. A. Gonzalez-Perez, and F. J. Gonzalez-Vila
Direct Detection of Black Carbon in Soils by Py-GC/MS, Carbon-13 NMR Spectroscopy and Thermogravimetric Techniques
Soil Sci. Soc. Am. J., January 11, 2008; 72(1): 258 - 267.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (25)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Dria, K. J.
Right arrow Articles by Hatcher, P. G.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Dria, K. J.
Right arrow Articles by Hatcher, P. G.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Dria, K. J.
Right arrow Articles by Hatcher, P. G.
Related Collections
Right arrow Humic Substances
Right arrow Soil Analysis
Right arrow Soil Methods/Instrumentation
Right arrow Soil Chemistry

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Crop Science
Vadose Zone Journal Journal of Plant Registrations
Journal of Natural Resources
and Life Sciences Education		 Soil Science Society of America Journal