Investigation of marine and terrestrial humic substances by 1H and 13C, nuclear magnetic resonance and infrared spectroscopy

Or~4amc (,co(twmi.~trl, VoL 2. pp. 117 to 124 Pergamon Press Lid 1980. Printed in Great Britain

Investigation of marine and terrestrial humic substances by ~H and ~3C, nuclear magnetic resonance and infrared spectroscopy J E A N - M A R I E DEREPPE, C L A U D E T T E M O R E A U X

Physical Chemistry Department, University of Louvain, Place L. Pasteur 1, Louvain La Neuve, Belgium and Y. DEBYSER Geochemistry Department, Institut Francais du Petrole, l and 4 Avenue de Bois Preau, Rueil Malmaison, France

(Received 3 January 1979; accepted in revisedform 14 May 1979) Abstraet--Humic acids were isolated from 5 sediments in which the origin nature of the organic matter are both typical and different. The humic acids were characterized on the basis of elemental compositions, infrared spectra and ~H and 13C NMR. This last technique, especially 13C NMR, provides qualitative and semi-quantitative information regarding aromatic structure. Combined data from the three techniques permits differentiation of marine and terrestrial organic matter as well as identification of mixtures of humic acids from the two sources.

Sample 3

INTRODUCTION ANALYTICAL METHODS such as infrared absorption, electron paramagnetic resonance, and mass spectrometry have been used successfully to study the organic matter of soil and recent sediments. Proton and especially 13C nuclear magnetic resonance (NMR) have been used in a few cases only to characterize well-defined fractions of such organic matter obtained b y extraction with organic solvents or by degradative techniques. To the best of our knowledge, only two 13C NMR studies of untreated soil humic acids have appeared in the literature (Vila et al., 1976; Wilson and Goh, 1977). This paper reports an NMR study to compare the complex organic matter extracted from a soil and from a marine sediment. These results were complemented by infrared spectra and elementary analysis of the samples. SAMPLES

The following samples, because of their different origins and physical and chemical properties, were chosen for this study:

Sample 1 Sediment from the Norway Sea collected during the Orgon I cruise (Debyser and Gadel, 1977). Organic matter derived from plants of cold climates may contain little or no lignin (Pelet, 1977). Geochemical properties of this organic matter are similar to those of marine organic sediment (Debyser and Gadel, 1977).

Sample 2 Sediment from Bourgneuf Bay, France. This organic matter is for the most part of autochthonous planktonic origin (Gouleau, 1975). o(~. 2/3~I A

] 17

An A-1 podzol from 'La Roche aux F6es', France. This sample contains substances derived from lignin.

Sample 4 Peat moss from the USSR. This sample also contains substances derived from lignin.

Sample 5 Recent sediment from a core collected during the Orgon II cruise to the Cariaco Trench off the Venezuelan coast. The organic constituents were derived from marine substances but also contained terrestrial material. EXPERIMENTAL METHODS Each sample was treated with 2 N HCI at 4°C for 16 hr to remove carbonates. Humic acids were then dissolved in an aqeous mixture of 0.1 N NaOH and 1% sodium pyrophosphate (Kononova, 1966). Fulvic acids were separated from the humic acids by centrifugation of the extract after acidifying it to pH 2. The reprecipitated humic acids were then redissolved in 0.1 N NaOH, purified by dialysis and passed through a column of H + resin. The humic acids were finally freeze-dried and stored under vacuum in the dark. Elemental analyses were carried out by ATX (Suresnes, France) on samples previously dried under vacuum (2 Torr) at 80°C for 2 hr. Quantitative infrared spectra were obtained using a Beckman 42-40 spectrometer in the absorbance mode. The humic acids (1-2mg) were dispersed in 200 mg of KBr and the mixture was pelletized. The pellets were then dried by storing them in a dry atmosphere for 7 days (Robin, 1975). 13C NMR spectra were obtained with a Varian X L 100-15/Ft spectrometer using the following con-

JEAN-MARIE DEREPPE, CLAUDETTE MOREAUX a n d Y. DEBYSER

Table 1. Elemental compositions for humic acid isolates

Sample

C %

H %

0 %

N %

A %

Atomic H/C

Atomic O/C

Atomic N/C

Atomic S/C

1

53.3

5.8

33.8

4.9

2.1

1.3

0.48

0.08

0.015

2

52.9

5.9

33.4

5.3

2,5

1.4

0.47

0.10

0.020

3

51.1

4.5

41.8

1.9

0.8

1.0

0.61

0.03

0.006

4

53.5

4.7

38.2

2.6

0.0

0.9

0.71

0.05

0.017

5

53.8

5.6

32.0

5.5

3.0

1.3

0.45

0.09

0.021

ditions: spectral width, 5000Hz; flip angle, 30~; acquisition time, 0.Ssec; broadband decoupling; and 12-mm tubes. About 300,000 spectra were co-added for each run. The humic acid (200 mg) was dissolved in a mixture of D20 (2.5 ml) and NaOD (1 or 2 N). ~H NMR spectra were also recorded on the XL 100 spectrometer in the following conditions: sweep width, 1000 Hz; sweep rate, 4 Hz/sec.; time constant, 1 sec; sample tube, 5 mm dia. RESULTS AND DISCUSSION Elemental compositions for the humic acid preparations from each of the 5 samples are shown in Table 1. Differences in composition reflect the different origins of the organic matter. Thus, in Samples 1 and 2, the organic matter consists for the most part of proteins and glucosides. On the other hand, the

humic acids from the soil organic matter consist primarily of glucosides and lignin; the atomic H/C and N/H ratios for the soil humic acids are low corresponding to more aromatic structures. Infrared spectra for the humic acid show three rather well defined peaks at 2920, 1710 and 1540 cm- ~ (Fig. 1). These peaks, defined by the absorption coefficient K expressed in absorbance per mg of organic carbon (Robin, 1975), correspond to carbon hydrogen, carboxyl and amide groups, respectively (Table 2). Absorptions for the carbonhydrogen bonds are twice as large for the humic acids from the marine sediments (samples 1 and 2) as for those from the soils (samples 3 and 4). Absorption for the humic acid from the Cariaco Trench (sample 5) was intermediate between those for the other two groups. The amide absorption at 1540 cm-~ does not appear in the humic acid isolated from soil where,

Humic Acid Isolates

/k

/k/

i i

\_j

Norway Sea

V ,Aj'xJ L ~

'l

\

Bourgneuf Bay

A L/\ fx4' •

Podzol A1

Peat from USSR

,.A./ Cariaco Trench

2920

1710 1540 1050 1520 WAVE NUMBER (cm-1)

Fig. 1. Infrared spectra of samples 1-5.

Investigation of marine and terrestrial humic substances

119

Table 2. Infrared absorption coefficients, K, defined as absorbance per mg of organic carbon at indicated wave numbers K

Sample

K

2929 cm -I (Ch + CH 2 + CH 3)

K

1710 cm -I

1540 cm -I

(C=O)

Pepttde b O l l d 01~

amlde

1

58

77

18

2

56

61

24

3

30

80

0

4

28

73

0

5

42

69

15

instead, there is an absorption at 1520cm -1. This absorption can be correlated with aromatic double bonds of lignin moieties similar to vanillic or syringic acid (Farmer and Morrison, 1960). Typical 13C NMR spectra are shown in Fig. 2 where the signal-to-noise ratio is comparable to that previously obtained by Vila et al. (1976). In contrast to spectra of pure substances of low to moderate molecular weight that show well defined sharp peaks, spectra of complex mixtures such as humic acids are poorly resolved and show only 'absorption bands' rather than peaks. Various factors contribute to this broadening among which are the extreme molecular complexity of the samples and the presence of paramagnetic metal ions and free radicals, as well as broadening resulting from short relaxation time. Line broadening resulting from molecular complexity has been observed for other macromolecular systems such as proteins or mixtures of hydrocarbons. In certain

190

I 200

I

160

112

, 150

'

macromolecular systems, however, it is still possible to observe well resolved resonance peaks where there is a limited number of repeating sub-units. Line broadening because of short relaxation time cannot explain observed broadening on the basis of values reported for TI in systems of equivalent molecular weight. Broadening of the spectra, therefore, must be mainly attributable to the extreme complexity and non-repeating nature of the sample. Line broadening caused by paramagnetic centers cannot be ignored. Several studies (Atherton et al., 1967) have shown the presence of quinone-type free radicals in humic acid and EPR measurements have confi'rned the presence of paramagnetic centers in our samples. Interpretation of the 13C NMR spectra has been mainly based on chemical shift data and several general correlation tables reported in the literature. Correlations more specifically applicable to the study of complex organic substances thought to be building

90

, I00

J

I

t

50

0

Fig. 2. Carbon-13 NMR spectra of samples 1, 3 and 4.

~ PPM

JEAN-MARIE DEREPPE. CLAIJIIETTE MOREAUX and Y. DEBYSEK

120

‘P,R-CXYi-

mdN.0

R-COOR m

-cowl-c?E

-

m

1B-2 :c=c;

1

t

‘$0

?O

200 Fig.

3. Carbon-13

aliphatic

group.

+ eC+%R

up

f

I

Chemical

shift regions.

I 0

I

loo

150

50

BPPM

The symbol v, stands for a phenyl group and R for an

The symbol s( refers to x-position relative to an aromatic or carboxyl group. The carbon

atom in resonance is noted t. blocks of humic acids are shown in Fig. 3. This table is based on recent publications that are listed separately in the Bibliography. Considering the probable nature of humic acids (Felbeck, 1971; Haworth, 1971), all spectra can be divided into five sections as shown in Table 3. The absence of any resonance between 190 and 210ppm, where aldehydic and ketonic carbonyl groups are observed, shows that most carbonyl groups occur in Table 3. Assignment

acid, ester, amide, or quinone structures. The carbony1 band for samples 3 and 4 extends from 175 to 190ppm suggesting carboxylic acids and peptide or amide structures. This interpretation is in accord with the nature of the original organic matter which consists primarily of lignin and glucosides for samples 3 and 4, and of proteins and glucosides for samples 1 and 2. The interpretations are also in agreement with those based on infrared spectra where the characterof “C

NMR bands

Band

Range

of band,

ppm

Majar

.xssLgnment -__

1

190-160

Carbon

doubly

one oxygen

160-112

112-

90

Aromatic

90-

50

atom

(and

carbon

(CA)

Carbon

singly

two oxygen Aliphattc bonded

bonded

(>C = 0)

unsaturated)

bonded

atoms carbon

0

Saturated

singly

to one oxygen

carbon

to

(-0-C-O)

(-0-C-j 50-

to

(CS)

atom

Investigation of marine and terrestrial humic substances

121

Table 4. Integrated intensities for '3C NMR bands Band Ratios

Band, ppm

Sample I

2

3

4

5

2/5

4/5

3+4/5

i

14.0

14. i

2.2

19. i

50.6

0.27

0.37

0.42

2

7.9

12.7

7.8

23.1

48.5

0.27

0.47

0.63

3

i0. i

21.5

4.2

36,5

27.7

0.75

1.32

1.47

4

12.9

33.3

6.7

26.3

20.8

1.58

1.26

1.59

5

17.1

34.1

4.2

17.9

26.7

1.28

0.67

0.82

istic amide band was observed only for samples 1, 2 and 5. However, this interpretation based on the NMR spectra is risky because the signal-to-noise ratio is poor for spectra 1, 2 and 5. Also, if peptides were the main contributors to the carbonyl group, the nitrogen content of the humic acid would be expected to be much greater. Third, a broad peak of nearly equal intensity would be expected to be present between 40 and 60 ppm for carbon atoms alpha to the nitrogen atom of the peptide bond. With these points in mind, a detailed assignment of the carbonyl resonances must be deferred until better spectra become available. The band between 112 and 160 ppm, representing aromatic structure, is clearly observable in all the spectra. The shoulder at 155 ppm that appears in the spectra of samples 3 and 4 may represent aromatic carbon linked to an oxygen atom as in phenols or albrylaromatic ethers. This assignment is in agreement with the lignin component of these samples. The band between 50 and 112 ppm contains three characteristic peaks corresponding to glucosidic structures; the resonances at 65, 72 and 108 ppm correspond, respectively to --CH2OH, ~H2--O--, and - - O - - ~ H 2 - - O ~ groups of the glucosides. The simultaneous occurrence of these resonance bands, and especially of the - - O ~ H 2 - - - O - - - band, which does not interfere with other bands from lignin structures, permits detection and identification of glucosides. The spetral region between 0 and 50 ppm provides but little information as to the nature of the aliphatic groups; the band at 28 ppm represents aliphatic side chains. Based on normal experimental conditions used for 13C NMR, it is hazardous to conclude that the intensities of the signals reflect absolute concentrations of various classes of carbon atoms. The relationship between the intensities obtained in the FT mode and the number of nuclei at resonance is actually influenced by several factors such as (1) spin lattice relaxation times with respect to pulse repetition rates, (2) effects of the limited pulse strength (HI) on the spectral band width to be covered, (3) nonlinearity in the recording system and (4) Nuclear Overhauser Enhancement (NOE). In the case of humic acids, the high molecular weight, the presence

of paramagnetic centers and the high complexity of the molecules considerably shorten the relaxation time and the NOE. In this case, the measured integrated intensities can be used to characterize and compare humic acids from various origins. Even where the intensity measurements are not strictly quantitative, the method will, nevertheless, provide a picture of the nature of a sample and will permit comparison with other samples characterized by the same method. The integrated intensities corresponding to various resonance bands, expressed as percent of total intensity, are shown in Table 4. The proportion of aromatic carbon atoms is low for samples 1 and 2, and much higher for samples 3 and 4. The opposite is true for the aliphatic carbon atoms. These observations agree well with the infrared absorption peaks at 2920cm-1 and with what is to be expected for the origin of the organic matter in these four samples. Sample 5 has a very high proportion of aromatic carbon atoms and, conversely, a fairly low percentage of aliphatic carbon atoms. These observations, which agree with the conclusions based on the infrared spectra, confirm the dual origin of the organic matter of sample 5. The number of carbon atoms associated with ~--O-and - - O - - C - - O - - structures is greater for samples 3 and 4 in agreement with the high atomic O/C ratio of these two humic acids and in accord with the lignin relationship of this organic matter thus offering an explanation for the high proportion of carbon atoms :t to oxygen atoms. Typical I H NMR spectra are shown in Fig. 4. Considering the probable nature of the humic acids and the chemical shift table of Fig. 5, which is taken for the most part from the data published by Chamberlain (1974), it is possible to account for four main resonance bands (Table 4). Maxima at 0.8 and 1.2 ppm for sample 4 appear as shoulders for the other samples. These can be attributed to terminal methyl groups and unbranched aliphatic ~ H 2 - groups. The maximum at 6.5 ppm has been attributed to quinonic structures. The well defined peak at 8.5 ppm, which has been observed in samples 3 and 5 as well as in spectra of lignin (Ludeman, 1973), has not been identified. This peak actually appears in a spectral region characteristic of protons linked to oxygen and

122

JEAN-MARIEDEREPPE,CLAUDETTEMOREAUXand Y. DEBYSER ,'

:

s

®

I

10

9

8.5 I

I

I

1

5.8 I

I

4.7 4.1 I ,

I

2.8 ,

I

8

7

6

5

4

3

2

1.7 I

I

I

1

0

8PPM

Fig. 4. Proton NMR spectra of samples 1-5. nitrogen hetero-atoms which cannot be seen under the experimental conditions used in this work. Most resonances in band 2 are hidden by the broad HDO peak. Moreover, using a DzO/NaOD solvent, protons linked to hetero-atoms and several other types of protons activated by NaOD do not appear in the spectra because of hydrogen-deuterium exchange (Becker, 1969). In this case, data obtained by proton spectroscopy are not quantitative. Nevertheless, it is possible to compare ratios of band intensities, as for bands 3, 4 and 5, that represent aliphatic protons, inasmuch as they are not affected by hydrogendeuterium exchange. It is also possible to compare, but with caution, intensities of band 1. Actually, only a few types of aromatic hydrogen can undergo the hydrogen/deuterium exchange (Massicot, 1967), and the intensity of band 1 is essentially quantitative. Integrated intensities of bands 1, 3, 4 and 5, is expressed in percent of total intensity for all 4 bands,

as listed in Table 6. Samples 1 and 2, which are of marine origin, are characterized by low aromaticity that is clearly reflected in the 1:5 ratio of aromatic to aliphatic hydrogen. The 3:5 ratio for (--O----CH2--1 aliphatic) hydrogens and the 3:4 ratio for (---OCH2--1 hydrogen to aromatic or/C~------O) structures are higher for soil samples 3 and 4. This is interpreted to indicate that besides the glucosides that are common to all humic acids, the soil humic acid also contains aikylaromatic esters from its lignin-derived component. The ratios for sample 5 are of intermediate values as would be expected for a dual origin. CONCLUSIONS Proton and 13C NMR spectra of humic acids from different origins differ considerably. Qualitative interpretations may be used to suggest or confirm the origin of the organic matter. Quantitative interpretation

Investigation of marine and terrestrial humic substances 1

123

H3C-N L~'

R-CH2-C(~)O-1R-CH~-C(~)CO1 I

~ - C H 2-

~P-CH2-O CO-R,~

I

R-O-CH2-O-R 1

~-CH2-O-R,~

1

R2-CH-O-R '

R-CH 2 - O - R , ~ 1

R-CH2-O-CO-R ,

CH3-O-R,~I 1

H,R-CH2 -COOH,R

CH3-O-CO-R,~

Pyrones 1

CH ,CH2 ('~) 1CH3(~)

I

I

~ R COND. I A R

1-CH-OH,R

85 l 10

[ 9

I

58 ~

I

I 7

6

47 41 I I 5

28

II

I I 3

4

ICH

3

I 1

0

17 I 2

I

&PPM

Fig. 5. Proton chemical shift regions. The symbol ~# stands for a phenyl group and R for an aliphatic group. The symbol ~ refers to ~-position relative to an aromatic or carboxyl group.

Table 5. Assignment of proton NMR bands Band

Range of band, ppm

Major assignment

I

8.5 - 5.8 ppm

Aromatic H

2

5.8 - 4.1 ppm

H in saturated groups ~ to two oxygen atoms

3

4.1 - 2.8 ppm

H in satucated groups ~ to one oxygen atola

4

2.8 - 1.7 ppm

H in saturated groups ~ to aromatic ring or (>C=O) fuact[onal group

5

1.7 - 0 ppm

H in saturated group B or further from aromatic ring aqd (>C=O) functional group

Table 6. Integrated intensities of proton NMR bands Sample

Band, ppm 1

3

4

Band Ratios 5

1/5

3/5

3/4

I

ii.i

25.5

21.9

41.6

0.26

0.61

1.16

2

13.1

20.1

16.6

48.6

0.27

0.41

1.20

3

14.4

44.0

17.2

24.4

0.59

1.80

2.56

4

23.0

35.7

14.8

26.5

0.87

1.35

2.42

5

18.3

24.7

21.5

35.7

0.51

0.70

1.15

124

JEAN-MARIE DEREPPE, CLAUDETTE MOREAUX and Y. DEBYSER

of the spectra leads to the differentiation of hydrogen and carbon atoms into different structural parameters, such as the aromaticity factor, which are not readily obtained by other methods. This first effort needs confirmation a n d refinement based on study of more samples of various origins or belonging to a diagenetic series. The a m o u n t and quality of information thus far obtained from 13C N M R spectra have been restricted by a poor signal-to-noise ratio. Inasmuch as higher concentrations of samples cannot readily be used, higher field strengths or larger volumes will be required to obtain better spectra.

REFERENCES

Atherton, N. H., Cranwell, P. A., Floyd, A. J., and Haworth, R. D., 1967, ESR spectra of humic acids: Tetrahedton, v. 23, p. 1653-1667. Becket, E. D., 1969, High Resolution NMR, Academic Press. Berger, S., and Rieker, A., 1972, Zur Kenntnis des chinoiden Zustandes: Tetrahedron, v. 28, p. 3132. Breitmaier, E., and Voolter, W., 1974, ~3C NMR Spectroscopy. Verlag Chemie. Chamberlain, N, F., 1974, The Practice of NMR Spectroscopy, Plenum Press. Debyser, Y., and Gadel, F., 1977, Mission Orgon I, Geochemical Study of Humic Substances and Kerogens, Editions du CNRS. Farmer, V. C., and Morrison, R. I., 1960, Chemical and infrared studies of phragmites peat and its humic acid: Sci. Proc. R. Dublin Soc., p. 85-104. Felbeck, C. T., 1971, Structural hypothesis of soil humic acids: Soil Sci., v. 111, p. 42. Gagnaire, D., and Robert, D., 1977, A polymer model of lignin (D. H P.) ~3C selectively labeled at the benzylic

positions. Synthesis and NMR study: Makromol. Chem., v. 178, p. 1477. Gouleau, L., 1975, First stages of sedimentation of litoral muds, Thesis, University of Nantes, France. Haworth, R. D., 1971, The chemical nature of humic acid: Soil Sci., v. 111, p. 71. Kononova, M. M., 1966, Soil Organic Matter, Pergamon. Lounasmaa, M., 1977, Derives phloroglucinoliques d'Hagenia Abyssinica: Acta Chem. Scand., v. B31, p. 77. Ludeman, H. D., 1973, Protonenresonenzspektroskopie von ligninen und humic sauren bei t00 Megahertz: Erdoel Kohle, v. 26, p. 50(~508. Ludeman, H. D., and Nimz, H., 1974, ~3C-Kernresonanzspekten von Ligninen. 1 and 2: Makromol. Chem., v. 175, p. 2393 and 2409. Massicot, J., 1967, Etude par R.M.N. de la reaction d'echange des protons aromatiques des phenols en milieu alcalin: Bull. Soc. Chim. France, p. 2204-2208. Nimz, H., 1974, Beech lignin--proposal of a constitutional scheme: Anyew. Chem., Int. Ed. Enyl., v. 13, p. 313. Pelet, R., 1977, Mission Orgon I, Organic Geochemistry of Pelagic Marine Sediments from the Norwegian Sea: Overall aspect, Editions du CNRS. Robin, P., 1975, Characterization of kerogens and their evolution by infrared spectroscopy, Thesis, University of Louvain, Belgium. Simpson, T. J., 1977, ~3C NMR spectra and biosynthetic studies of xanthomegnin and related pigments from Aspergillus sulphureus and mellus: J. Chem. Soc. Perkin Trans., p. 592. Vila, F. J. G., Lentz, H., and Ludeman, H. D., 1976, FT-13C nuclear magnetic resonance spectra of natural humic substances: Biochem. Biophvs. Res. Commun., v. 72, p. 1063. Wehrli, F. W., and Wirthin, T., 1976, Interpretation of ~3C NMR Spectra. Heyden. Westerman, P. W., et al., 1977, 13C NMR study of naturally occurring xanthones: Org. Magn. Res., v. 9, p. 631. Wilson, M. A., and Goh. K. M., 1977, Proton decoupled pulse Fourier transform ~3C nuclear magnetic resonance of soil organic matter: J. Soil. Sci.. v. 28. p. 645-652.