Techniques for the dierentiation of carbon types
present in lignite-rich mine soils
Cornelia Rumpel
a,*, Jan O. Skjemstad
b, Heike Knicker
c,
Ingrid KoÈgel-Knabner
c, Reinhard F. HuÈttl
aaDepartment of Soil Protection and Recultivation, Brandenburg Technical University, PO Box 10 13 44, D-03013 Cottbus, Germany bCSIRO Land & Water, Waite Road, Urrbrae, PMB 2, Glen Osmond Adelaide, SA 5064, Australia
cLehrstuhl fuÈr Bodenkunde, Technische UniversitaÈt MuÈnchen, D-85350 Freising-Weihenstephan, Germany
Received 7 June 1999; accepted 2 March 2000 (returned to author for revision 23 November 1999)
Abstract
The objective of this study was to assess high energy UV photo-oxidation,13C CP/MAS NMR spectroscopy and14C
activity measurements for their suitability in dierentiating carbon derived from plant material and that derived from lignite. The conceptual approach included analysis of bulk soil samples and physical fractions of lignite-rich mine soils. Additionally, the <53mm fraction of selected samples was subjected to 2 hours of high-energy UV photo-oxidation. Quanti®cation of the lignite carbon content by14C activity measurements showed that a mixture of lignite and recently
formed organic matter occurs in the bulk surface soil horizon as well as a number of particle size and density fractions. Using13C CP/MAS NMR spectroscopy, carbon species, characteristic for lignite- and plant-derived organic matter can
be identi®ed. Both methods yield corresponding results. It was also shown that lignite-derived carbon is more resistant to the high-energy UV photo-oxidation treatment than carbon derived from recent plant material. This method pro-vides qualitative indications of lignite contribution to soils.#2000 Elsevier Science Ltd. All rights reserved.
Keywords:Lignite; Humus; Mine soil;14C activity measurements;13C CP/MAS NMR spectroscopy; High energy UV photo-oxidation
1. Introduction
Lignite mining in the Lusatian mining district, in the eastern part of the Federal Republic of Germany is carried out in open-cast mines. After excavation and relocation, the sandy overburden is deposited at spoil banks. The overburden material contains lignite carbon of up to 5% w/w. After aorestation with deciduous or coniferous trees, plant litter accumulates and is incor-porated into the mineral soil. Because of the dark colour of the parent substrate, humi®ed materials formed dur-ing the decomposition of plant litter cannot be dis-tinguished by macromorphological observation from lignite. Dierentiation of lignite and recent carbon is
necessary to quantify the accumulation of recent carbon during soil genesis and to assess the degree of humi®ca-tion in these soils (Rumpel et al., 1999).
While fresh plant material is mainly composed of carbohydrates, protein, lignin and lipids, lignite con-tains a considerable portion of aliphatic and aromatic structures, which were formed during coali®cation.13C
CP/MAS NMR spectroscopy was used to study changes occurring during humi®cation (Preston, 1996) as well as coali®cation (Wilson et al., 1987; Bates and Hatcher, 1989). It was also shown that structural analysis by13C
CP/MAS NMR can provide an indication of lignite carbon contribution to soils (Schmidt et al., 1996; KoÈgel-Knabner, 1997; Rumpel et al., 1998a,b, 1999). However, overlapping signals in the spectra do not allow for quantitative estimates of the lignite contribu-tion to the spectra. Therefore, addicontribu-tional techniques are necessary in order to distinguish the two carbon types.
0146-6380/00/$ - see front matter#2000 Elsevier Science Ltd. All rights reserved. P I I : S 0 1 4 6 - 6 3 8 0 ( 0 0 ) 0 0 0 2 6 - 7
www.elsevier.nl/locate/orggeochem
* Corresponding author. Tel.: 355-78-1162; fax: +49-355-78-1170.
Lignite is composed of carbon which was deposited several million years ago and consists only of stable carbon isotopes. Therefore, radiocarbon measurements can be used to quantify the lignite carbon contribution to the organic matter mixture of lignite-rich mine soils (Rumpel et al., 1998a). Another promising approach to dierentiate carbon types which developed under dier-ent conditions is high-energy UV photo-oxidation. This technique was applied by Skjemstad et al. (1996) and Golchin et al. (1997) to detect charred organic matter in soils.
In the present study, solid-state 13C NMR
spectro-scopy was applied to bulk soil samples and physical fractions of the surface soil horizon containing humi®ed plant material (Ai horizon) of lignite-rich mine soils to elucidate structural dierences of the carbon types. Additionally the samples were analyzed for radiocarbon activity to quantify the lignite content. The results were correlated with structural characteristics as revealed by
13C CP/MAS NMR spectroscopy. Selected samples were
subjected to high-energy UV photo-oxidation.
2. Materials and methods
2.1. Sampling
Prior to soil sampling an investigation of rehabilitated mine sites was carried out. A grid sampling and inven-tory of soil chemical parameters showed a great varia-bility (Rumpel et al., 1998a,b). For soil sampling, soil pro®les were obtained from lignite containing mine soils where the chemical parameters were found in the med-ium range of all samples taken in the grid. The soil pro®les are located under a chronosequence of pine stands (Black pine, 11 years old, Scots pine, 17 and 32 years old) and a red oak stand (36 years old). Samples were taken from the forest ¯oor horizons (L, Oh) and the Ai horizons of lignite-containing mine soils. The subsoil (Cv horizon) at 1 m depth was also sampled.
2.2. Sample pre-treatment and chemical analysis
Roots and visible plant remains were mechanically removed from the samples where possible. The samples were air dried and the fraction >2 mm was removed by dry sieving. For chemical analysis an aliquot was ground. Carbon and nitrogen contents were determined with a Leco CHN 1000 analyzer.
2.3. Particle-size fractionation
Particle-size fractionation of the Ai horizon under 36-year-old red oak was carried out after ultrasonic dispersion (Christensen, 1992). The amount of ultra-sonic energy has been calibrated following the
proce-dure suggested by Schmidt et al. (1999). 150 J/ml were used to obtain a complete dispersion of aggregates. Three sand fractions (2000±630, 360±200 and 200±63
mm) were obtained by sieving. The clay fraction (<2
mm) and the three silt fractions (63±20, 20±6.3 and 6.3±2
mm) were recovered after sedimentation in a sedimenta-tion cylinder. The recovered soil suspension was ®ltered using polysulfon membrane ®lters (<0.45 mm). The solid material remaining on the ®lter paper was recov-ered, freeze-dried and stored in a brown-glass bottle.
2.4. Density fractionation
Density fractionation of the Ai horizon under 36-year-old red oak, to obtain three density fractions (<1.6, 1.6±2.0 and >2 g/cm3), was carried out using
sodium polytungstate Na6(H2W12O40) according to
Golchin et al. (1994). 5 g of soil were placed in a 250 ml centrifuge tube and 50 ml of sodium polytungstate solution was added. Afterwards, the aggregates were dispersed ultrasonically. Finally the solution was cen-trifuged (3000 RPM) for 20 min and suction ®ltered (Whatman GFA). The solid material on the ®lter was recovered, washed and dried at 40C.
2.5. Radiocarbon dating
Radiocarbon (14C) ages were obtained using the
con-ventional macro-technique of liquid scintillation as described by Becker-Heidmann et al. (1988). A soil sample containing 6 g carbon was burned and the car-bon was subjected to benzene synthesis. After 6 weeks, the 14C activity of the benzene was recorded with a
scintillation spectrometer (Packard Tri Carb Model 3320). The 14C activity must be corrected for isotopic
fractionation, according to Stuiver and Pollach (1977). Therefore part of the sample was combusted, the CO2
puri®ed and itsd13C determined in an isotope ratio mass
spectrometer (MTA 250, Fa Finnigan). 14C dates are
corrected with thed13C values for isotopic eects. The
measuring time was chosen to keep the error below1 pmC (percent modern carbon). The14C activity data are
referenced to the measured 14C activity of oxalic acid
dating from 1950 (Mann, 1983). Assuming that the sample contains a mixture of recently formed SOM, derived from plant litter, and dead carbon (carbon without14C activity), the amount of lignite C present in
the sample can be obtained as a percentage of the total organic carbon. To do so the measured 14C activity
must be corrected for the elevated14C activity of recent
plant material, which increased with regards to the standard, dating from 1950 (14C activity=100 pmC) due
to nuclear testing during the 1960s. This can be done by dividing the measured 14C activity by the activity of
planted at the most 36 years ago (Geyh, personal com-munication). The lignite contribution can be calculated by subtracting the corrected14C activity from the total
carbon which is 100%:
M1ÿÿ14C activity=115
100 1
The error introduced by choosing 115 pmC is greatest for samples containing small amounts of lignite. How-ever, in no sample is it greater than5%.
2.6. High energy UV photo-oxidation
For high energy UV photo-oxidation, 10 g of sample were Na+saturated using 5 M NaCl, dialyzed until free
of Clÿ and dispersed by mild ultrasonic treatment.
Afterwards, the sand fractions (>53mm) were recovered by sieving. The suspension of the < 53mm fractions was made up to a total volume of 500 or 1000 ml and the carbon content was measured by combustion. Aliquots of the < 53mm suspensions containing ca. 2±2.5 mg C were photo-oxidized at 2.5 kW for a period of 2.0 h. After photo-oxidation, the suspensions were washed, centrifuged, the pellets weighed and 2 ml of water added. Aliquots of the ®nal suspensions were analysed for their carbon content (Skjemstad et al., 1996). The photo-oxidized fraction of selected samples was accu-mulated by UV oxidation of several subsamples.
2.7. 13C CP/MAS NMR analysis
Samples of lignite-rich mine soils contain large amounts of paramagnetic material. Therefore, they need to be pretreated in order to obtain interpretable spectra. HCl and HF are commonly used to remove para-magnetic material from soil samples (Preston et al., 1989). The spectra of the soil under 36-year-old red oak, before and after the photo-oxidation treatment, were recorded after 2% HF treatment as proposed by Skjemstad et al. (1994). For recording spectra of the bulk soil of the other sites, which were recorded later, a 10% HF treatment was used. This concentration was shown to remove paramagnetic material in a shorter time. The method published by Schmidt et al. (1997) can thus be regarded as an improvement of the ®rst method by Skjemstad et al. (1994). However, both studies showed that HF pretreatment eciently removes para-magnetic compounds with no alteration of the organic matter.
13C CP/MAS NMR analysis of the soil samples from
the soil pro®les under the chronosequence of pine and the red oak stand were carried out with a Bruker MSL 100 NMR spectrometer at a frequency of 25.6 MHz. Cross polarization with magic angle spinning at 4±5 kHz (Schaefer and Stejskal, 1976) was applied. The contact time was 1 ms and a pulse delay between 200
and 600 ms was chosen. The chemical shift was refer-enced to the TMS scale. All samples were treated with 10% HF (Schmidt et al., 1997) prior to analysis to remove paramagnetic compounds.
13C CP/MAS NMR analysis of soil fractions, before
and after high energy UV photo-oxidation (red oak site), was carried out on a Varian Unity 200 NMR spectrometer. The instrumental conditions were iden-tical to those reported by Skjemstad et al. (1996). Magic angle spinning was applied at 5 kHz. A contact time of 1 ms was used and a pulse delay of 350 ms was chosen. The chemical shift was referenced to the TMS scale. Each set of the photo-oxidized fractions was combined and treated with HF (2%) before being subject to a13C
CP/MAS NMR experiment (Skjemstad et al., 1994). The 13C CP/MAS NMR spectra were integrated to
obtain the relative contribution of the signal intensities to dierent chemical shift regions. The chemical shift regions 0±45, 45±110, 110±165 and 165±220 ppm were assigned to alkyl C, O-alkyl C, aromatic C and car-boxylic C, respectively (Wilson, 1987).
2.8. Statistical analysis
Data sets were tested for normal distribution using the Kolmogorov±Smirnov Test (Lorenz, 1988). Regres-sion analysis was performed with Sigma Plot computer software with signi®cance tested using the studentt-test.
3. Results and discussion
3.1. Radiocarbon measurements of the soil organic matter
Measurements of the radiocarbon activity of the sampled soils show that in the bulk mineral soil 47 to 100% of all carbon is derived from lignite (Tables 1 and 2). Lignite carbon contributions of less than 100% as is observed in the Ai horizons indicate that the carbon derived from lignite and from plant litter are mixed in this horizon (Rumpel et al., 1998a). In the Ai horizons of the pine chronosequence, the contribution of recent carbon derived from plant litter increases with stand age from 16 % under the 11-year-old pine to 36% under its 32-year-old counterpart (Table 2). This observation is consistent with the reported increase of the recent car-bon in the Ai horizon during mine soil development (Rumpel et al., 1999). In the forest ¯oor, the contribu-tion from dead carbon can be as high as 41% (Table 2). This shows the organic carbon input into the mine soils from airborne contamination in the vicinity of either coal-®red power plants or briquette factories (Schmidt et al., 1996; Rumpel et al., 1998b, 1999).
contribution of lignite carbon from 11 to 54% (Table 1), indicating that all the fractions contain a mixture of lignite and recent carbon.
3.2. 13C CP/MAS NMR analysis of the bulk soil and physical fractions
13C CP/MAS NMR spectra of the two carbon types
present in lignite-rich mine soils are presented in Fig. 1. The 13C CP/MAS NMR spectrum of the parent
substrate (Cv horizon) is characterised by signals indicating a high contribution of aromatic and ali-phatic carbon species (0±45 and 110±160 ppm). Only a small signal can be observed in the O-alkyl region. The spectrum is characteristic of lignite (Meiler and Meu-singer, 1991). The 13C CP/MAS NMR spectrum of
plant litter (L-horizon) shows a dominant peak at 72 ppm. Together with the shoulder at 105 ppm this signal is assigned to carbohydrate structures. Signals at 119, 130, 150 and 56 ppm may originate from lignin com-pounds and represent protonated, C-substituted, O-substituted aromatic C and methoxyl C. The peak at 172 ppm can be assigned to carbonyl groups (KoÈgel-Knabner, 1997).
The ratio of signal intensities A=(alkyl C+aromatic C)/(O-alkyl C+carboxylic C) is also shown in Fig. 1. The parameter A was proposed as an indicator for the lignite contribution to soils (Schmidt et al., 1996). In this study it will be tested for its potential to elucidate the contribution of lignite to the mineral soil as well as the
forest ¯oor. However, biases can arise from using the cross polarisation technique. In a solid state13C NMR
experiment with coal and related materials the eciency of magnetisation transfer is dependent on the 1H±13C
dipole±dipole interactions (Packer et al., 1983). Carbon nuclei, which are not in close proximity to protons may not undergo cross-polarisation and therefore could not be `seen' in a CP experiment by the NMR technique (Snape et al., 1989). For samples that contain signi®cant amounts of carbon isolated from protons with a highly aromatic structure, CP/MAS 13C NMR may not give
quantitative data (Kinchesh et at., 1995; Skjemstad et al., 1997). For those samples the single pulse or Bloch decay technique may yield quantitative results (Zhang and Maciel, 1990; Skjemstad et al., 1996). This techni-que is too expensive to be applied routinely (Preston, 1996). Lignite is an early stage of coali®cation and does not contain polycyclic aromatics (Hatcher, 1988). Therefore, quantitative information may be obtained with the CP/MAS technique with a reasonable degree of accuracy (Rumpel et al.,1998a).
In the spectrum of plant litter the contribution of O-alkyl C and carboxylic C is high and the value of A is low in comparison to the Cv horizon (Fig. 1). The rela-tive intensity distributions of the four carbon species in the13C CP/MAS NMR spectra of the bulk soil samples
of the soil pro®les under the three pine stands are pre-sented in Table 2. In the L-horizons, the contribution of O-alkyl carbon ranges from 66 to 69% in the Cv hor-izons its contribution is at most 32%, whereas much of
Table 1
Total carbon, lignite content obtained by14C activity measurements and ratio of signal intensity A of the forest ¯oor and the mineral
soil under 36-year-old red oak
Horizona OC
(g/kg)
Lignite (% total C)
A=(alkyl C+aromatic C)/ (O-alkyl C+carboxylic C)
L 462 n.d. 0.490.05
Oh 224 21.40.5b 0.960.10
Ai 110 47.00.4 0.960.10
Cv 37 96.20.3 1.380.15
Particle size fractions of the Ai horizon
2000±630mm 6 10.80.6 0.960.10
630±200mm 8 41.30.6 1.130.12
200±63mm 21 39.60.6 1.270.13
63±20mm 40 54.60.7 1.500.16
20±6.3mm 185 30.30.5 1.270.13
6.3±2mm 130 36.61.9 1.220.13
< 2mm 74 50.00.5 1.380.15
Density fractions of the Ai horizon
<1.6 g/cm3 358 32.10.3 0.960.10
1.6-2-.0 g/cm3 218 45.40.5 0.940.10
>2.0 g/cm3 5 54.30.4 1.020.11
a L=unaltered plant litter; Oh=humic layer on the mineral soil; Ai=upper mineral soils, where accumulation of recent organic
matter took place; Cv=parent substrate for soil development.
the signal intensity is found in the aromatic spectral region (110±160 ppm, 33±41%). For the parent sub-strate (Cv horizon) of the three sites, the values of A range between 1.66 and 1.93 (Table 2). For plant
mate-rial as sampled in the L-horizon, A was found to be between 0.38 and 0.44. In the Of horizons, A increases up to 0.84 under the 17-year-old pine (Table 2). This may be explained by the structural changes occurring
Fig. 1. 13C CP/MAS NMR spectra of pine litter (L-horizon, recent carbon) and the parent substrate (Cv-horizon, lignite carbon).
Field strength: 25 MHz.
Table 2
Integrals of the13C CP/MAS NMR spectra and lignite content as determined by14C activity measurements of the forest ¯oor and
mineral soil sampled in three soil pro®les under pine
Chemical shift (ppm) Alkyl C O-alkyl Aromatic Carboxylic C A=(alkyl C+aromatic C)/ (O-alkyl C+carboxylic C)
Lignite
0±45 45±110 110±160 160±220
(% Ca) (% total C)
Black pine, 11 years
L 16 69 12 3 0.380.04 n.d.
Of 24 51 20 5 0.790.08 41.20.5
Ai 26 40 28 5 1.200.13 84.00.3
Cv 30 32 33 6 1.660.18 100.0
Scots pine, 17 years
L 16 67 14 3 0.420.04 n.d.
Of 24 48 22 7 0.840.09 29.80.5
Ai 21 35 33 10 0.840.09 68.90.3
Ai2 22 34 36 8 1.380.15 82.60.2
Cv 19 28 41 12 1.930.21 n.d.
Scots pine, 32 years
Kiefer
L 15 66 16 3 0.440.04 n.d.
Of 20 52 21 7 0.690.07 n.d.
Ai 26 48 21 4 0.900.09 64.00.4
Cv 22 24 41 13 1.700.18 99.50.3
a The variation of integration data of signals due to the treatment of a well resolved FID (Fourier transformation, phasing and
during humi®cation of plant material. It was demon-strated during the humi®cation process that the peak at 72 ppm decreases slightly and that the peaks in the car-boxylic region increase due to oxidation processes (KoÈgel-Knabner, 1993). An additional increase was noted in the alkyl region (KoÈgel-Knabner, 1993) and attributed either to selective preservation of cutan/sub-eran (Augris et al., 1998) or the contribution of micro-bial material (Golchin et al., 1996). Taking into account that this horizon has been subject to airborne con-tamination, lignite carbon contribution could also be responsible for the increase of A.
For the Ai horizons of the three sites, values of A range between 0.90 and 1.38 (Tables 1 and 2). The ratio A is 1.20 for the Ai horizon of the 11-year-old pine site, indicating that the SOM is dominated by lignite. An A value of 0.90 was recorded for the Ai horizon of the oldest mine soil (32 years; Table 2). A decrease in the values of A in the Ai horizons with age indicates a higher contribution of O-alkyl C and carboxylic C car-bon species, characteristic for recent plant material. These values are in agreement with data obtained by14C
activity measurements.
To investigate if the structural dierences as observed by 13C CP/MAS NMR spectroscopy are related to a
lignite contribution, the values for A were correlated with the lignite contents determined by the14C activity
measurements. A linear relationship (r2=0.79***)
between the lignite content and ratio A can be obtained (Fig. 2) for the bulk soil samples and physical fractions. The correlation may have been biased by the under-estimation of the aromatic fraction of the lignite in CP spectra. If Bloch decay were used this correlation may be even better.
3.3. Characterisation of the chemical composition of the organic matter before and after high-energy UV photo-oxidation
The carbon recoveries after dierent times of photo-oxidation treatment of the three horizons of the 36-year-old mine soil are shown in Fig. 3. For all three horizons, a plateau is reached after 2 h of treatment and thereafter only minimal carbon loss occurs even when the sample was UV-treated for up to 8 h. The pattern of carbon oxidation is similar for all soil samples studied and is in agreement with results obtained by Skjemstad et al. (1996) for some natural soils. Because of this plateau, the authors de®ned the carbon fraction remaining after 2 h of treatment as protected. Carbon can resist high energy UV photo-oxidation if it is associated with the mineral matrix (physical protection) or if it consists of polycyclic aromatics (chemical protection). Physical protection was observed when inorganic cementing agents occur (Golchin et al.,1997). In the studied soils, lignite-derived carbon may be more stable to
photo-oxidation than organic matter formed during the decomposition of plant material.
To study the behaviour of lignite, present in the par-ent substrate for soil developmpar-ent, to high energy UV photo-oxidation, samples from the Cv horizons of the four sites containing mainly lignite-derived carbon (Tables 1 and 2) were UV-treated. Pure lignite cannot be taken as a reference because the composition of Lusa-tian lignite, which is exploited by the mining industry, diers from the lignite-rich overburden (Rumpel et al., 1998a). After 2 h of treatment, 35±53% of the original carbon is still present in this horizon (Table 3). The data indicate that lignite carbon in the Cv horizon is partly susceptible to photo-oxidation. For the samples of the
Fig. 2. Relationship between data from13C CP/MAS NMR
spectroscopy and radiocarbon dating.
surface soil horizons (Ai), 20±55% of all carbon is recovered after 2 h of photo-oxidation (Table 3). In comparison to the Cv horizons, the slightly lower resis-tance to UV photo-oxidation can be explained by the higher contribution of carbon derived from decompos-ing plant material. The most rapid and, at the same time, the greatest carbon decline is observed for the sample of the forest ¯oor (Oh) (Fig. 3). Here, only 17% of the original carbon remains after 2 h of photo-oxidation (Table 3). This result is not surprising as this sample contains large amounts of natural humic mate-rial, which is removed quickly by high-energy UV photo-oxidation (Skjemstad et al., 1993, 1996). A linear relationship can be obtained between the carbon recov-ered after two hours of high energy UV photo-oxidation and the lignite content as determined by 14C activity
measurements (r2=0.67**) (Fig. 4). Our data suggest
that the presence of lignite in soil can be detected by high energy UV photo-oxidation. However, this exam-ination cannot be quantitative because in a sample con-taining mainly lignite, half of it is lost by the UV treatment. To support these results changes of the chemical structure after 2 h of photo-oxidation have been observed using13C CP/MAS NMR spectroscopy.
The13C CP/MAS NMR results of the photo-oxidized
fractions, after dierent times of treatment of the Cv horizon containing 96% lignite as determined by 14C
activity measurements, are presented in Fig. 5. Our data indicate that in this horizon containing lignite as a single carbon source there is no signi®cant dierence in struc-tural composition after the photo-oxidation treatment. No loss of any speci®c fraction was detectable after 2 h treatment. However, these data may be biased by
selec-tive enrichment of aromatic rings, which were found by other authors during the oxidation of coal (Fredericks et al., 1983). These structures may not be observed com-pletely by CP/MAS NMR. The reported dierences may also be due to dierences in the organic structure of the coals, which is generally related to its rank (Rausa et al., 1994). We conclude that it was not possible with the CP/MAS technique to observe changes occurring in the chemical structure of lignite after photo-oxidation.
Table 3
Organic carbon and NMR data of the <53mm fraction before and after high energy UV photo-oxidation
Before UV treatment After 2 h UV treatment
Horizon Ratio of signal intensity (A)a C recovery Ratio of signal intensity (A)a
(% original OC)
Red oak, 36 years
Oh 1.1 172 1.50.16
Ai 1.2 200 1.60.17
Cv 2.2 532 2.60.28
Black pine, 11 years
Ai n.d. 532 n.d.
Cv n.d. 356 n.d.
Scots pine, 17 years
Ai n.d. 551 n.d.
Ai n.d. 391 n.d.
Cv n.d. 368 n.d.
Scots pine, 32 years
Ai n.d. 491 n.d.
Cv n.d. 4820 n.d.
a A = (alkyl C + aromatic C)/(O-alkyl C + carboxylic C).
Fig. 4. Relationship between the lignite contribution of the bulk soil determined by 14C activity measurements and the
In the Oh and Ai horizon, where recently formed organic matter is a major contributor (79 and 53%, Table 1) to the organic carbon pool, the composition of the organic matter does change after the high energy UV photo-oxidation treatment (Fig. 5) in that the con-tribution of aromatic carbon species (110±116 ppm) increases relatively with time in both samples and the contribution of O-alkyl carbon species (50±110 ppm) decreases. These data show that changes in the chemical structure after photo-oxidation can be observed with the CP/MAS technique when samples contain high amounts of recent organic matter derived from plant litter.
The ratio of signal intensities A=(alkyl C+aromatic C)/(O-alkyl C+carboxylic C), increases after 2 h of high energy UV photo-oxidation of samples from the Oh horizon from 1.1 to 1.5 (Table 3). In the Ai horizon an increase from 1.2 to 1.6 can be noted. These data indi-cate that carbon species characteristic for lignite are preserved relative to carbon species characteristic for plant litter. The data suggest that lignite is generally more resistant to high energy UV photo-oxidation than organic matter formed during the decomposition of plant material. Thus, high energy UV photo-oxidation may provide a qualitative indication of the presence of lignite carbon in soils.
4. Conclusion
It was shown that14C activity measurements and13C
NMR spectroscopy can be used to dierentiate lignite carbon and recent carbon in soils containing mixtures of both carbon types. Both methods yield corresponding results for a whole range of soil samples. 14C activity
measurements give quantitative information about the lignite-derived carbon in soil. Using13C CP/MAS NMR
spectroscopy chemical structures derived from plant material can clearly be distinguished from carbon spe-cies derived from lignite. A linear relationship was established, indicating that a higher contribution of lig-nite carbon to soil increases alkyl and aromatic carbon species in the NMR spectra.
The carbon recovery after 2 h UV photo-oxidation of soil samples was compared to the lignite content as determined by radiocarbon dating. The UV recovery data showed that a portion of the lignite is susceptible to photo-oxidation and this approach can therefore not be quantitative and is at best indicative. 13C CP/MAS
NMR analysis indicated that the oxidation behaviour of lignite is not conferred to the chemical stability of one speci®c carbon species. In contrast to the observations made by other authors, our results show that high energy UV photo-oxidation is not chemically selective and oxidises the lignite structure of the overburden material as a whole.
Lower rates of lignite oxidation in comparison to recently formed organic matter indicate that this method has the potential of dierentiating the two types of organic matter and could also provide a qualitative guide to contamination of soil from other chemically resistant organic material (e.g. charcoal, soot, etc.), especially when combined with methods for the identi-®cation of the remaining carbon species.
Acknowledgements
The authors would like to thank the Deutsche For-schungsgemeinschaft for ®nancial support. J. Taylor and S. Grocke, CSIRO, Adelaide, Australia are acknowl-edged for their help with the photo-oxidation work. We thank Dr. P. Clarke who recorded the13C NMR spectra
of photo-oxidised fractions. Dr. P. Becker-Heidmann, University of Hamburg, is acknowledged for the radio-carbon dating.
Associate EditorÐC.E. Snape
References
Augris, N., Balesdent, J., Mariotti, A., Derenne, S., Largeau, C., 1998. Structure and origin of insoluble and non-hydro-lyzable, aliphatic organic matter in a forest soil. Organic Geochemistry 28, 119±125.
Fig. 5. 13C CP/MAS NMR spectra of the <53mm fractions
Bates, A.L., Hatcher, P.G., 1989. Solid-state13C NMR studies of
a large fossil gymnosperm form the Yallourn Open Cut, Lat-robe Valley, Australia. Organic Geochemistry 14, 609±617. Becker-Heidmann, P., Liang-wu, L., Scharpenseel, H.W., 1988.
Radiocarbon dating of organic matter fractions of a Chinese mollisol. Zeitschrift fuÈr P¯anzenernaÈhung und Bodenkunde 151, 37±39.
Christensen, B.T., 1992. Physical fractionation of soil and organic matter in primary particle size and density separates. In: Stewart, B.A. (Ed.), Advances in Soil Science 20. Springer, New York, pp. 2±76.
Fredericks, P.M., Warbrooke, P., Wilson, M.A., 1983. Chemi-cal changes during natural oxidation of a high volatile bitu-minous coal. Organic Geochemistry 5, 89±97.
Golchin, A., Oades, J.M., Skjemstad, J.O., Clarke, P., 1994. Study of free and occluded particulate organic matter in soils by solid state 13C CP/MAS NMR spectroscopy and scanning electron microscopy. Australian Journal of Soil Research 32, 385±409. Golchin, A., Clarke, P., Oades, J.M., 1996. The heterogeneous
nature of microbial products as shown by solid-state 13C CPMAS NMR spectroscopy. Biogeochemistry 34, 71±97. Golchin, A., Clarke, P., Baldock, J.A., Higashi, T., Skjemstad,
J.O., Oades, J.M., 1997. The eects of vegetation and burn-ing on the chemical composition of soil organic matter in a volcanic ash soil as shown by 13C NMR spectroscopy. I.
Whole soil and humic acid fraction. Geoderma 76, 155±174. Hatcher, P.G., 1988. Dipolar-dephasing13C NMR studies of
decomposed wood and coali®ed xylem tissue: evidence for chemical structural changes associated with defunctionaliza-tion of lignin structural units during coali®cadefunctionaliza-tion. Energy Fuels 1/2, 40±58.
Kinchesh, P., Powlson, D.S., Randall, E.W., 1995. 13C NMR studies of organic matter in whole soils: I. Quantitation pos-sibilities. European Journal of Soil Science 46, 125±138. Klouda, B., Lewis, C.W., Rasmussen, R.A., Rhoderick, G.C.,
Sams, R.L., Stevens, R.K. et al., 1996. Radiocarbon mea-surements of atmospheric volatile organic compounds: quantifying the biogenic contribution. Environmental Sci-ence and Technology 30, 1098±1105.
Knicker, H. 1993. Quantitative 15N-und 13C-CPMAS-FestkoÈr-per und 15N-FluÈssigkeits-NMR-Spectroskopie an P¯anzen-komposten und natuÈrlichen BoÈden. PhD thesis, Regensburg. KoÈgel-Knabner, I., 1993. Biodegradation and humi®cation
processes in forest soils. In: Bollag, J.-M., Stotzky, G. (Eds.), Soil Biochemistry 8. Marcel Dekker, New York, pp. 101±137. KoÈgel-Knabner, I., 1997.13C and15N NMR spectroscopy as a
tool in soil organic matter studies. Geoderma 80, 243±270. Lorenz, R.J., 1988. Grundbegrie der Biometrie. Gustav
Fischer, Stuttgart.
Mann, W.B., 1983. An international reference material for radiocarbon dating. Radiocarbon 25, 519±527.
Meiler, W., Meusinger, R., 1991. NMR of coals and coal pro-ducts. Annual Reports on NMR spectroscopy 23, 376±410. Packer, K.J., Harris, R.K., Kenwright, A.M., Snape, C.E.,
1983. Quantitative aspects of solid state 13C n.m.r. of coals and related materials. Fuel 62, 999±1002.
Preston, C.M., Schnitzer, M., Ripmeester, J.A., 1989. A spec-troscopic and chemical investigation on the de-aching of a humin. Soil Science Society American Journal 53, 1142±1147. Preston, C.M., 1996. Application of NMR to soil organic matter
analysis: history and prospects. Soil Science 161, 144±166.
Rausa, R., Girardi, E., Calemma, V., 1994. Humic acids from coal. Production, characterization and utilization. In: Senesi, N., Miano, T.M. (Eds), Humic Substances in the Global Environment and Implications on Human Health, pp. 1225± 1244.
Rumpel, C., KoÈgel-Knabner, I., Knicker, H., Skjemstad, J.O., HuÈttl, R.F., 1998a. Types and chemical composition of organic carbon in reforested lignite-rich mine soils. Geo-derma 86, 123±142.
Rumpel, C., Knicker, H., KoÈgel-Knabner, I., HuÈttl, R.F., 1998b. Airborne contamination of forest soils by lignite-derived materials: its detection and its impact on the composition of soil organic matter. Water, Air and Soil Pollution 150, 481±492. Rumpel, C., KoÈgel-Knabner, I., HuÈttl, R.F., 1999. Organic matter composition and degree of humi®cation in lignite-rich mine soils under a chronosequence of pine. Plant and Soil 213, 161±168.
Schaefer, J., Stejskal, E.O., 1976. Carbon-13 nuclear magnetic resonance of polymers spinning at magic angle. Journal of the American Chemical Society 98, 1031±1032.
Schmidt, M.W.I., Knicker, H., Hatcher, P.G., KoÈgel-Knabner, I., 1996. Impact of brown coal dust on organic matter in particle size fractions of a Mollisol. Organic Geochemistry 25, 29±39. Schmidt, M.W.I., Knicker, H., Hatcher, P.G., KoÈgel-Knabner,
I., 1997. Improvement of 13C and 15CPMAS NMR spectra of bulk soils, particle size fractions and organic material by treatment with hydro¯uoric acid (10%). European Journal of Soil Science 48, 319±328.
Schmidt, M.W.I., Rumpel, C., KoÈgel-Knabner, I., 1999. Parti-cle size fractionation of soils comprising coal and combusted particles. European Journal of Soil Science 50, 515±522. Skjemstad, J.O., Janik, L.J., Head, M.J., McClure, S.G., 1993.
High energy ultraviolet photo-oxidation: a novel technique for studying physically protected organic matter in clay- and silt-sized aggregates. Journal of Soil Science 44, 485±499. Skjemstad, J.O., Clarke, P., Taylor, J.A., Oades, J.M.,
New-man, R.H., 1994. The removal of magnetic materials from surface soils. A solid-state 13C CP/MAS n.m.r. study. Aus-tralian Journal of Soil Research 32, 1215±1229.
Skjemstad, J.O., Clarke, P., Taylor, J.A., Oades, J.M., McClure, S.G., 1996. The chemistry and nature of protected carbon in soil. Australian Journal of Soil Research 34, 251±271. Skjemstad, J.O., Clarke, P., Golchin, A., Oades, J.M., 1997.
Characterisation of soil organic matter by solid-state 13C
NMR spectroscopy. In: Cadisch, G., Giller, K.E. (Eds.), Driven by Nature: Plant Litter Quality and Decomposition. CAB International, Oxon, pp. 253±271.
Snape, C.E., Axelson, D.E., Botto, R.E., Delpuech, J.J., Tekely, P., Gerstein, B.C. et al., 1989. Quantitative reliability of aromaticity and related measurements on coals by 13C NMR. A debate. Fuel 68, 547±560.
Stuiver, M., Polach, H.A., 363. Discussion: Reporting of 14C data. Radiocarbon 19, 355±363.
Wilson, M.A., Verheyen, T.V., Vassallo, A.M., Hill, R.S., Perry, G.J., 1987. Selective loss of carbohydrates from plant remains during coali®cation. Organic Geochemistry 11, 265±271. Wilson, M.A., 1987. NMR-Techniques and Application in
Geochemistry and Soil Chemistry. Pergamon Press, Oxford. Zhang, M., Maciel, G.E., 1990. Large sample 13C MAS NMR
