Content uploaded by David Sebag
Author content
All content in this area was uploaded by David Sebag on Sep 14, 2014
Content may be subject to copyright.
Organic matter decomposition: bridging the gap between
Rock–Eval pyrolysis and chemical characterization
(CPMAS
13
C NMR)
R. Albrecht •D. Sebag •E. Verrecchia
Received: 27 March 2014 / Accepted: 3 September 2014
Springer International Publishing Switzerland 2014
Abstract Organic matter (OM) is a key component
of soils but information on its chemistry and behavior
in soils is still incomplete. Numerous methods are
commonly used to characterize and monitor OM
dynamics, but only a few include the qualities required
to become routine techniques i.e. simple, rapid,
accurate and at low cost. Rock–Eval pyrolysis (RE
pyrolysis) is a good candidate, as it provides an
overview of OM properties by monitoring four com-
ponents related to the main major classes of organic
constituents (from A1 for the labile biological constit-
uents to A4 for the mature refractory fraction).
However, a question is still pending: do these four
major classes used in the literature reflect a pertinent
compositional chemical counterpart?
13
C Nuclear
Magnetic Resonance Spectroscopy in the solid state
(
13
C CPMAS NMR) has been used to answer this
question by collecting information on structural and
conformational characteristics of OM. Moreover, in
order to avoid the blurring effect of pedogenesis on OM
dynamics, a ‘‘less complex OM’’ source, i.e. compost
samples, has been used. Results showed significant and
high determination coefficients between classes, indi-
ces (of transformation of plant biopolymers, humifi-
cation…) from RE pyrolysis, and the main classes of
OM characterized by
13
C NMR, e.g. A1 & A2 with
labile/easily degradable components (alkyl C et
O-alkyl C), A3 & A4 with humified OM (with aromatic
C and phenolic C). The R index (contribution of bio-
macromolecules) is correlated with phenolic and
aromatic C, whereas the I index (related to immature
OM) refers to labile––easily degradable components
(alkyl C et O-alkyl C). The results confirm the
pertinence of RE pyrolysis to monitor OM dynamics.
Responsible Editor: Matthew Wallenstein
R. Albrecht (&)E. Verrecchia
Institut des Sciences de la Terre, Universite
´de Lausanne,
Ge
´opolis, 1015 Lausanne, Switzerland
e-mail: remyalbrecht@gmail.com
R. Albrecht
Laboratory of Ecological Systems (ECOS), School of
Architecture, Civil and Environmental Engineering
(ENAC), Ecole Polytechnique Fe
´de
´rale de Lausanne
(EPFL), Station 2, 1015 Lausanne, Switzerland
R. Albrecht
WSL Swiss Federal Institute for Forest, Snow and
Landscape Research, Site Lausanne, Station 2,
1015 Lausanne, Switzerland
D. Sebag
Laboratoire Morphodynamique Continentale et Co
ˆtie
`re,
Universite
´de Rouen, CNRS, 76821 Mont-Saint-Aignan,
France
D. Sebag
IRD, HydroSciences Montpellier, Universite
´de
Ngaounde
´re
´, BP 1857 Yaounde
´, Cameroun
123
Biogeochemistry
DOI 10.1007/s10533-014-0033-8
Keywords Organic matter Compost Soils Rock–
Eval pyrolysis
13
C Nuclear magnetic resonance
spectroscopy
Introduction
Soil organic matter (SOM) benefits of a positive
perception since a long time. It is clearly associated
with the concept of soil fertility and, more recently, of
soil quality., Although a minor component of soil in
quantitative terms, organic matter influences many
soil properties, which in turn affects several soil
functions: for example, it provides nutrient reservoirs,
it is a substrate for microbial activity, it preserves the
environment, and it is critical for sustaining and
increasing agricultural productivity (Schnitzer 2005).
However, information on its chemistry and behavior in
soils is still incomplete. Numerous techniques have
been used to characterize and monitor SOM dynamics
(Kogel-Knabner 2000). Nonetheless, very few are
used routinely, because of the common need for
preliminary sample preparation (e.g. extraction or
purification), or due to their complexity and cost (e.g.
NMR techniques). Only a few techniques have the
necessary qualities to become routine techniques, i.e.
simple, rapid, accurate and at low cost, in order to be
used on a very large sampling design. Rock–Eval
pyrolysis (RE pyrolysis; Disnar et al. 2003; Sebag
et al. 2006) or Near-Infrared Spectroscopy (NIRS;
Albrecht et al. 2008a;2009) constitute two potential
candidates for routine techniques.
This study focuses on RE pyrolysis as a method to
evaluate both the nature and behavior of natural
organic matter (OM). Although initially designed for
petroleum exploration (Lafargue et al. 1998), RE
pyrolysis has been applied to a variety of other
materials such as soils or recent sediments because of
its simplicity.
RE method of pyrolysis consists of a programmed
heating to quantitatively and selectively determine (1)
free hydrocarbons contained in samples and (2)
hydrocarbon- and oxygen-containing compounds
(CO, CO
2
) that are volatilized during the cracking of
OE contained in the sample. Disnar et al. (2003)
estimated the various pools of organic compounds by
using their specific cracking temperatures. Several
authors referred to this approach for various objec-
tives, such as quantification of refractory OM,
assessment of humification processes, and also in
order to study the outcome of fossil OM in modern
environments, as well as diagenesis of sedimentary
OM and quantification of carbon content in soils
(Copard et al. 2006; Sebag et al. 2006; Marchand et al.
2008; Poot et al. 2009; Tambach et al. 2009; Graz et al.
2011; Saenger et al. 2013; Gillespie et al. 2014; Hare
et al. 2014). The main advantages of this method are
repeatability and rapidity in providing an overview of
OM properties and content.
Nevertheless, there is still a question pending: do
the major classes of OM defined by this approach
accurately reflect and with pertinence OM chemical
composition?. To answer this question,
13
C Nuclear
Magnetic Resonance Spectroscopy in the solid state
(
13
C CPMAS NMR) has been used to collect direct
information on structural and conformational charac-
teristics of OM. NMR resonances were assigned to
chemical structures, according to five main forms:
alkyl C, O-alkyl C, aromatic C and phenolic C, and
carbonyl–carboxyl C. The pertinence of RE pyrolysis
is evaluated using a comparison between pyrolysis
results and those of the
13
C CPMAS NMR reference
tool.
In this study, natural OM from compost samples is
used for the following reasons: (1) contrary to soils,
compost variability only comes from transformation
of OM and type of inherited parent materials; (2)
composting processes are well described and refer-
enced in the literature (De Bertoldi et al. 1983; Epstein
1997); (3) the samples used have already been studied
in previous works (Albrecht et al. 2008b;2009) and
are well constrained; (4) finally, this type of natural
OM allows complex ‘‘interferences’’ due to pedoge-
netic processes occurring in soils to be avoided.
Materials and methods
Experimental materials
Composts are obtained from local dewatered digested
municipal sewage sludge, green wastes, and pine barks
at a 1:1:1 v/v ratio (Company Biotechna, Ensue
`s,
Bouches du Rho
ˆne, France). Pine barks are incorpo-
rated into other biowastes to improve aeration during
the process. The mixture was composted for 20 days
in impervious boxes (100 m
3
) with forced aeration,
and then stored in windrows (10 m long, 4 m high, and
Biogeochemistry
123
5 m deep) on a composting platform for 6 months.
The piles are mixed several times to promote humi-
fication of OM. Approximately 1 kg of homogenized
compost was collected from each windrow at nine
different stages of composting (4, 18, 40, 57, 67, 84,
101, 114 and 128 days) with four replicates for 36
samples in total. All samples were sieved (\20 mm
mesh). Samples were freeze-dried and ground with a
Cyclotec 1,093 mill (FOSS) to 1 mm.
Chemical and physical analyses of OM
RE pyrolysis has been performed with a ‘Turbo model
Rock–Eval 6 pyrolyser’ (Vinci Technologies, Rueil-
Malmaison, France). A homogenized sample of about
100 mg undergoes a pyrolysis step followed by a
complete oxidation of the residual material. A FID
detector measures hydrocarbons released during pyro-
lysis, while CO
2
and CO are detected by infrared
absorbance during both steps. Pyrolysis starts isother-
mally at 200 C for 3 min, after which, the sample is
heated to 650 C. The oxidation step starts isother-
mally at 400 C (3 min) and then heats up to 850 C.
Deconvolution of signals results in four main peaks:
the S1 peak (hydrocarbons released during the
isothermal phase), the S2 peak (hydrocarbons pro-
duced between 200 C and 650 C), the S3 peak (CO
2
from pyrolysis of OM up to 400 C), and the S4 peak
(CO
2
released from residual OM below around 550 C
during the oxidation step). Mineral carbon decompo-
sition is recorded by the S30peak (pyrolysis–CO
2
released above 400 C), and the S5 peak (oxidation–
CO
2
released above about 550 C).
The present work focuses on hydrocarbon com-
pounds released between 200 and 650 C. The thermal
maturity (TpS2––temperature peak of the S2 pyro-
gram in C) was determined as the temperature
corresponding to the maximum of the S2 curve. This
study refers to the method developed by Disnar et al.
(2003) to estimate the contribution of each pool of
organic compounds defined by their specific cracking
temperature. S2 curves are deconvoluted into a
combination of elementary components. The residual
method, which consists of subtracting the major
Gaussian elementary distribution centered on the main
mode, has been used first on the initial S2 signal, and
then on successive residual distributions (Sebag et al.
2006; Hete
´nyi et al. 2010). This empirical approach
arbitrarily reduces the number of elementary
components. Five elementary Gaussian distributions
as ‘‘components’’ were chosen according to the
following ranges (from A1 to A5): 205–340,
340–400, 400–460, 460–550 or 550–600 C.
Humic (HA) and fulvic acid (FA) fractions were
analyzed using a precipitation method (Swift 1996).
Compost samples were diluted with NaOH (0.1 M) in
a conical flask and shaken for 4 h. Then, samples were
centrifuged (Sorval Super T21), to recover the super-
natant fraction containing the humic material. The
supernatant was acidified to pH 1.0 using HCl (6 M),
and then centrifuged to separate HA (precipitate) and
FA (supernatant) fractions. Both fractions were dried
(105 C, 24 h) and weighed. HA and FA substances
were expressed in mg kg
-1
DM.
All solid-state
13
C CPMAS NMR spectra were
obtained on a Bruker Avance-400 MHz spectrometer
(Bruker, Bremen, Germany) operating at a
13
C
resonance frequency of 100.7 MHz. Samples were
placed in a 7 mm zirconium rotor and spun at the
magic-angle at 6 kHz. All measurements were made at
room temperature. The
13
C chemical shifts were
referenced to tetramethylsilane and calibrated with
the glycine carbonyl signal, set at 176.5 ppm. Decon-
volution of the NMR spectra was performed using the
DmFit software (Massiot et al. 2002). This software
adjusts the spectra to obtain the line-width and the
peak positions (in ppm), allowing each peak to be
integrated to get the percentage of each contribution.
NMR resonances were assigned to chemical structures
according to previous NMR studies on compost and
SOM (Inbar et al. 1991; Vinceslas-Akpa and Loquet
1997). The chemical shift range of
13
C NMR spectra
for compost corresponds to the following dominant
forms: alkyl C, e.g. from amino acids, lipids and waxes
(0–45 ppm), O-alkyl C, e.g. from cellulose and
hemicelluloses (45–110 ppm), aromatic C (110–145
ppm) and phenolic C, e.g. from lignin (145–165 ppm),
and finally carbonyl–carboxyl C (165–210 ppm).
Statistical analysis
The relationships between major classes of organic
constituents given by RE pyrolysis and
13
C CPMAS
NMR were calculated using Pearson coefficients of
determination (r
2
). Significant coefficients were
retained for pvalue \0.05. Principal component
analysis (PCA) was performed to get dimension
reduction in RE pyrolysis and
13
C NMR dataset. The
Biogeochemistry
123
key idea of PCA is to represent the variation of a data
matrix in a reduced number of dimensions to obtain
the general structure of the variability. A circle of
correlation expresses the influence of chemical com-
pounds and RE pyrolysis main parameters on the PCA
sample ordination. All statistical analyses were per-
formed using XLSTAT 2012.
Results
Elementary components obtained with RE
pyrolysis
RE pyrolysis results of compost samples are given in
Table 1. S2 curves were deconvoluted into a combi-
nation of the elementary components A1–A5. Trends
displayed by the four elementary components A1–A4
are particularly interesting. Two major and opposite
trends are easily identified (they are partly co-linear as
A1–A5 components constitute a closed dataset––their
sum is equal to 1). The first group (biopolymers),
which comprises A1 and A2 components, steadily
decreases from 32 to 33 % (after 4 days) to a final
value of 23 and 26 % after 128 days of composting,
for A1 and A2 respectively. The other group (medium
to high mature OM) comprises A3 and A4 compo-
nents, and increases from 20 to 26 % and 14 to 23 %,
for A3 and A4, respectively. TpS2 increases from
312.5 to 327.3 C between samples of four and
128 days, respectively.
13
C NMR and chemical results
from previous studies on the same samples (Albrecht
et al. 2008b;2009) are given in Table 2.
Statistical analysis of RE pyrolysis versus
13
C CPMAS NMR indices
Pearson’s correlation coefficients calculated using RE
pyrolysis and
13
C CPMAS NMR indices are given in
Table 3. Significant figures are presented in bold.
Main significant and interesting high r
2
are the
following:
– Alkyl C is significantly and positively correlated
with A1 component (r =0.478) but significantly
and negatively correlated with A4 component
(r =-0.406);
– O-alky C is not correlated with RE pyrolysis
components, but when O-alky C and alkyl C
fractions are added together, this new ‘‘labile OM
and plant biopolymers’’ component is significantly
and positively correlated with A1 and A2 compo-
nents (r =0.621 and 0.709, respectively), but also
significantly and negatively correlated with A3
and A4 components (r =-0.431 and -0.756,
respectively);
– Aromatic C is significantly and negatively corre-
lated with A2 component (r =-0.508) and sig-
nificantly and positively correlated with A4
component (r =0.450);
– Phenolic C is significantly and negatively corre-
lated with A1 and A2 components (r =-0.606
and -0.458, respectively) and significantly and
Table 1 Rock-Eval pyrolysis results on compost samples
Age (d) Tmax (C) A1 (%) A2 (%) A3 (%) A4 (%) A5 (%) R-Index I-Index
4 312.5 (2.1) 33.4 (1.3) 32.1 (2.3) 19.9 (1.3) 13.9 (1.6) 0.5 (0.6) 0.34 (0.01) 0.52 (0.02)
18 319.0 (4.2) 30.1 (2.6) 33.4 (0.8) 20.5 (1.1) 15.3 (2.1) 0.7 (0.5) 0.37 (0.04) 0.49 (0.05)
40 317.8 (2.2) 31.1 (1.6) 30.7 (2.9) 20.5 (0.4) 16.1 (1.9) 1.5 (1.1) 0.38 (0.03) 0.48 (0.03)
57 322.6 (4.0) 27.6 (0.2) 27.8 (0.5) 21.6 (0.2) 20.2 (0.6) 2.7 (0.1) 0.45 (0.01) 0.41 (0.01)
67 324.2 (1.6) 26.3 (0.1) 28.8 (0.1) 21.4 (0.0) 21.1 (0.0) 2.2 (0.0) 0.45 (0.02) 0.41 (0.02)
84 328.3 (1.1) 25.8 (0.3) 28.3 (1.9) 22.3 (0.5) 21.1 (1.2) 2.4 (1.1) 0.46 (0.02) 0.38 (0.01)
101 329.1 (1.4) 25.3 (0.3) 27.6 (0.3) 22.7 (0.9) 21.1 (0.6) 3.2 (0.2) 0.47 (0.01) 0.37 (0.02)
114 326.0 (1.7) 24.6 (0.6) 27.3 (0.9) 24.3 (0.3) 22.6 (0.5) 1.0 (0.3) 0.48 (0.01) 0.33 (0.01)
128 327.3 (3.3) 23.4 (0.8) 26.2 (0.7) 26.1 (0.7) 22.9 (1.3) 1.4 (1.3) 0.50 (0.02) 0.28 (0.01)
Values in parentheses are standard error (n=4)
A1 (205–340 C), A2 (340–400 C), A3 (400–460 C), A4 (460–550 C) or F5 (550–600 C), R-index =A3 ?A4 ?A5,
I-index =log[(A1 ?A2)/A3]
Biogeochemistry
123
positively correlated with A3 and A4 components
(r =0.623 and 0.422, respectively);
– Tmax (TpS2) is significantly and negatively
correlated with A1 and A2 components (r =
-0.884 and -0.719, respectively) and obviously
significantly and positively correlated with A3 and
A4 components (r =0.643 and 0.869,
respectively);
– HA (humic acids), as a HA/FA ratio (i.e. ratio
between humic and fulvic acids) displays a highly
significant and negative correlation with A1 and
A2 components and is significantly and positively
correlated with A3 and A4 components;
– OM (organic matter content) is significantly and
positively correlated with A1 and A2 components
(r =0.499 and 0.471, respectively) and finally
significantly and negatively correlated with A4
component (r =-0.519).
Transformation of OM (R vs. I index map)
Sebag et al. (2006) proposed a novel Rock–Eval
based-diagram, using two indices defined as
I=log(A1 ?A2)/A3 related to immature OM and
R=(A3 ?A4 ?A5)/100, which represents the
mature OM contribution to the S2 signal. Figure 1
shows a R versus I indices plot with two opposite
trends throughout the time of composting. The
R-index increases with values of 0.66 after 4 days
and 0.75 after 128 days of composting. In contrast,
Table 2
13
C NMR and chemical results from previous studies (Albrecht et al. 2008b;2009)
Time/
day
carbonyl–carboxyl
C 165–210 ppm
phenolic C
145–165 ppm
aromatic C
110–145 ppm
O-alkyl C
45–110 ppm
alkyl C
0–45 ppm
OM %
DM
HA
mg kg
-1
DM
FA
mg kg
-1
DM
4 8.3 (1.4) 4.3 (0.4) 10.2 (1.1) 54.8 (1.7) 22.4 (1.2) 58.5 (3.0) 28.3 (7.5) 54.9 (6.6)
18 8.3 (1.1) 4.3 (0.3) 9.8 (0.7) 58.7 (2.0) 19.0 (1.0) 57.5 (3.6) 28.3 (1.9) 51.9 (9.1)
40 8.1 (0.9) 4.1 (0.2) 10.2 (0.7) 57.9 (1.0) 19.8 (1.0) 60.3 (3.0) 31.5 (5.4) 52.0 (6.3)
57 8.7 (0.5) 4.6 (0.3) 10.8 (0.4) 54.9 (0.7) 21.0 (1.1) 48.3 (2.1) 39.3 (1.8) 38.2 (4.5)
67 8.4 (1.8) 4.3 (0.7) 10.0 (1.1) 57.6 (1.8) 19.7 (1.8) 53.5 (5.9) 26.6 (8.0) 47.7 (6.7)
101 8.9 (0.7) 4.9 (0.3) 10.9 (0.5) 56.9 (2.6) 18.5 (1.5) 52.6 (2.7) 53.5 (4.7) 46.8 (5.1)
114 9.3 (1.9) 4.9 (0.6) 11.5 (1.5) 56.5 (3.6) 17.9 (0.5) 50.8 (4.1) 59.5 (7.7) 47.9 (4.4)
128 8.5 (1.3) 4.4 (0.3) 10.4 (0.6) 57.1 (2.5) 19.6 (1.5) 50.5 (4.3) 51.9 (4.6) 47.2 (4.2)
Values in parentheses are standard error (n=4)
OM Organic matter content, HA humic, FA fulvic acids
Table 3 Pearson’s correlation coefficients between Rock–Eval pyrolysis and
13
C CPMAS NMR indices
Al % A2 % A3 % A4 % A5 % R-Index I-Index
Alkyl C 0.478 0.287 -0.243 20.406 20.160 0.231 0.323
O-Alkyl C 20.038 0.198 0.057 -0.101 -0.146 -0.240 0.000
Alkyl ?O-Alkyl C 0.621 0.709 20.431 20.756 20.483 0.154 0.601
Aromatic C 20.221 20.508 0.107 0.450 0.315 0.178 -0.248
Phenolic C 20.606 20.458 0.340 0.623 0.422 -0.372 20.488
Carboxyl C -0.176 -0.326 0.106 0.338 0.092 0.089 -0.185
Tmax 20.884 20.719 0.643 0.869 0.572 0.886 20.803
C/N 0.295 0.318 -0.282 -0.306 -0.194 0.092 0.320
HA 20.796 20.611 0.740 0.763 0.193 20.523 20.799
HA/FA 20.728 20.592 0.571 0.730 0.352 20.444 20.683
OM 0.499 0.471 -0.388 20.519 -0.346 0.235 0.480
Value in bold for p-value \0.05
HA humic acid, FA fulvic acid, OM organic matter, R-index =A3 ?A4 ?A5, I-index =log[(A1 ?A2)/A3]
Biogeochemistry
123
I-index decreases from 0.27 to 0.24 (Table 1) during
the same time span. Moreover, another presentation of
the I- and R-indices versus time of composting
using box plots brings up gaps in OM evolution
with a threshold at 57/67 days for I and R indices
(Fig. 1).
PCA was performed on RE pyrolysis and
13
CNMR
measured parameters to confirm correlation between
thermal resistance and OM transformation. PCA made
it possible to ordinate RE pyrolysis data according to
the stage of composting on both first and second
principal components PC1 and PC2 (Fig. 2). PC1 and
PC2 explained 37.2 and 27.4 % of total variance in RE
pyrolysis data, respectively. Three clusters are iden-
tified in the factorial projection. Young composts
(4–57 days) are clearly separated from all others on
PC1 and form a first cluster (left hand side of the
diagram). Fifty-seven and 101 day-old composts form
a second cluster (middle/right hand side of the
diagram). The third cluster is composed by oldest
composts of 114 and 128 days at the right hand side of
the plot. From this specific ordination of samples, a
chronological distribution of composts on PC1 is
clearly emphasized. On PC1, A1 and A2 factors have
an opposite trend compared to A3 and A4: A1 and A2
point to youngest composts, whereas A3 and A4 point
to the oldest ones.
Discussion
Composting processes
Numerous chemical and biological parameters (mois-
ture, pH, C
org
,N
org
, C/N, OM, HA and FA, respiration,
cellulase, protease, and phenoloxidase activities) were
R-Index = (A3+A4+A5)/100
I-index = log[(A1+A2)/A3]
Rindex versus time Iindex versus time
0.35 0.40 0.45 0.50
0.45
0.35
0.25
4d 40d 67d 101d 128d 4d 40d 67d 101d 128d
Fig. 1 I-index versus R-index plot (top) and I-index and R-index box plots (bottom;dfor day). Box plots display gaps in OM evolution
with a threshold at 57/67 days for both the I- and the R-indices
Biogeochemistry
123
monitored on samples during a complete composting
process of lasting six months (Albrecht et al. 2009).
Results revealed the existence of two phases within the
composting processes. The initial phase (4 to
50–60 days) has been characterized by an intensive
degradation and a rapid increase in temperature. Both
C/N ratio and OM content sharply decreased. The
second phase (up to 146 days) was characterized by
the stabilization of C/N ratio and OM content, a
decrease in all biological activities (respiration, cellu-
lase, phenoloxidase, and protease activities), and an
increase in the humification process occurring within
the OM with notably a high increase of HA/FA ratio,
which tripled from 0.54 to 1.61 between 4 and
146 days. This decline in biological activity was
explained by a quantitative and qualitative reduction
of nutrient sources, which became a limiting factor.
Indeed, as shown by
13
C CPMAS NMR on the same
samples (Albrecht et al. 2008b), peaks at 40 and
35 ppm, assigned to the CH
2
groups of proteins and
lipids, which are easily degradable compounds,
decrease rapidly during composting. Similarly, an
increase in aromaticity (aromatic C ?phenolic C by
13
C NMR) emphasizes a clear preference of easily
biodegradable C compounds for microorganisms.
Lignin transformation and its structural changes,
observed through the S/G (syringyl to guaiacyl) ratio
by
13
C NMR, revealed the same inclination (Albrecht
et al. 2008b).
Statistical analyses of
13
C NMR and NIRS data
using principal components analysis (PCA) show, in
both cases, a chronological distribution of composts
according to the degree of compost maturation and
clearly indicate two distinct phases. To summarize,
previous works on the same samples revealed a clear
link between the degree of transformation of OM
during composting processes and a succession of
specific bacterial communities within two phases
(Albrecht et al. 2010). The first phase is characterized
by a fast decomposition of non-humic biodegradable
OM and the second phase by the modification of
recalcitrant biopolymers (cellulose, lignin and hemi-
celluloses), accompanied by the formation of humic-
like substances (Albrecht et al. 2008b;2009). Conse-
quently, compost represents a perfect system to study
OM evolution (mineralization and humification pro-
cesses) and is considered in this study as a model with
clear and well identified transformations of OM
without the complex ‘‘interference’’ due to pedoge-
netic processes occurring in soils.
4
4
18
18
40
40
40
40 57
57
57
67 84
84
84 101
101
101
101
114
114
114
128
128
128
128
R-Index
I Index
carboxyl C
phenolic C
aromatic C
O-Alkyl C
methoxyl C
alkyl C
-8
-6
-4
-2
0
2
4
6
8
-8 -6 -4 -2 0 2 4 6 8
PC2 (27.40 %)
PC1 (37.20 %)
Fig. 2 Principal
component analysis (PCA)
using Rock–Eval pyrolysis
indices (I and R) and
13
C
NMR data (functional
chemical groups). The
percentage for each axis of
the explained variance is
given in axis labels. The
black dotted arrow
expresses the time variable
(number of days of
composting) on the PCA
plane and numbers refer to
these numbers of days
Biogeochemistry
123
Contribution of RE pyrolysis
Behaviour of elementary components
RE pyrolysis analysis of compost samples showed two
opposite trends for the four main elementary compo-
nents (A1–A4) generated by deconvolution of the S2
curves (Table 1). While A1 and A2 decreased during
composting, A3 and A4 increased. Sebag et al. (2006)
proposed that these four components are related to
major classes of organic constituents differing in
origin and their resistance to pyrolysis: labile biolog-
ical constituents (A1), resistant biological constituents
(A2), immature non-biotic constituents (A3), and a
mature refractory fraction (A4).
Regarding RE pyrolysis parameters, it appears that
labile and resistant biological constituents (A1 and
A2) decrease as observed by Albrecht et al. (2008b,
2009). Indeed, these authors found a decrease in the
C/N ratio linked to the decrease of biopolymers during
the composting process. Similarly, they showed an
intensive biological activity (respiration, cellulase,
phenoloxidase, and protease activities) during the first
phase of composting. This was explained by the
abundance of labile OM and plant biopolymers, such
as cellulose and hemicelluloses in the sewage sludge
and green wastes used. This labile OM component has
also been identified and characterized by
13
C NMR as
alkyl-C, which are commonly assigned to amino acids,
lipids, and waxes (Inbar et al. 1991; Vinceslas-Akpa &
Loquet, 1997). Moreover, specific peaks at 40 and
35 ppm were assigned to the CH
2
groups of proteins
and lipids. Alkyl C, known to be easily degradable
compounds (Albrecht et al. 2008b), may be linked to
the first component identified by RE pyrolysis: labile
biological constituents (A1). The Pearson correlation
coefficient corroborates this hypothesis with a signif-
icant and positive correlation between A1 and alkyl C
fraction with r =0.478 (Table 3).
According to Disnar et al. (2003) and Sebag et al.
(2006), thermoresistant biopolymers during RE pyro-
lysis may be found in the A2 component. To verify this
hypothesis, correlation between O-alkyl C and A2
fractions was tested according to the postulate that the
O-alkyl C fraction with a
13
C NMR chemical shift
ranging between 45 and 110 ppm is assigned to
cellulose and hemicelluloses (Inbar et al. 1991;
Vinceslas-Akpa & Loquet, 1997). Although O-alkyl
C––A2 correlation is positive, it is not significant
(Table 3). It is possible that the above-noted decreas-
ing trend for both A1 and A2 throughout composting
(Table 1) could be at the source of this lack of
significant correlation. The principle of RE pyrolysis
is based OM segregation depending on its thermal
resistance and not directly on its biomolecular com-
position. Thus, plant biopolymers (cellulose, hemicel-
luloses, etc.) may occur in both A1 and A2
components at the same time, depending of their
thermal resistance. Based on this hypothesis, alkyl C
and O-alkyl C fractions can be combined to build a
new ‘‘labile OM and plant biopolymers’’ component.
This component is significantly, highly and positively
correlated with A1 and A2 components (r
2
of 0.621
and 0.709), respectively. Consequently, these results
confirmed the hypothesis about the nature of A1 and
A2 components being easily degradable fractions of
OM.
Moreover, both correlation coefficients between
‘‘labile OM and plant biopolymers’’ and A3 and A4
components have negative signs (r =-0.431 and -
0.756, respectively), compared to the positive corre-
lation coefficients with A1 and A2. As shown by
results in Table 1, these opposite trends in correlation
coefficients confirm the hypothesis by Albrecht et al.
(2010) on the adaptation of metabolic capacities by
compost microbial communities between the two
well-known phases of composting process. These
authors showed that composting process operated into
two phases with the same samples. The first phase was
characterized by a fast decomposition of non-humic
biodegradable OM (\67 days), as it is observed with
decreasing A1 and A2 and emphasized by their
relationships with alkyl and O-alkyl C, which are
easily degradable compounds. The second phase
corresponds to the formation of humic-like substances
by the adaptation of microbial communities. More
precisely, the first phase is focused on easily degrad-
able substrate utilization, whereas the maturation
phase consists of multiple metabolisms, which induce
the release of metabolites and humification processes
(Albrecht et al. 2008b).
These humification processes have been identified
through the behavior of the more thermal resistant
biomolecules comprised in both A3 and A4 compo-
nents. According to Disnar et al. (2003)andSebagetal.
(2006), A3 and A4 are designed for immature non-biotic
constituents (A3) and mature refractory fractions (A4).
A3 and A4 are both positively correlated with aromatic
Biogeochemistry
123
and phenolic C (Table 3) during the time of composting.
Moreover, significant correlation coefficients only
existed for A4 (r =0.623 and 0.422, respectively).
Similarly to A1–A2 and aromatic–phenolic C, A3–A4
were negatively correlated with alkyl–O-alkyl C
(Table 3). Consequently, it is assumed that aromatic–
phenolic C compounds increase during the composting
process. In addition, increasing A3 and A4 during
composting (Table 1) can be related to a trend showing
the increase of several indices noted in previous studies
on the same samples and used to assess the humification
processes, i.e. increase of humic substances or degra-
dation of plant biopolymers (Albrecht et al. 2008b). The
first index is the monitoring of HA, and more precisely,
the monitoring of the HA/FA ratio during composting.
HA/FA is a tool usually used to assess humification
processes. Indeed, increase of HA and decrease of FA
raise the HA/FA ratio from values of 0.5 after 4 days to
1.1 after 128 days (Albrecht et al. 2008b). This change
can be attributed to the selective preservation of
aromatic compounds and high molecular size com-
pounds and/or to the formation of humic-like substances
with molecular structures and properties similar to that
of HA. Thus, in this case, both increasing A3 and A4
may be related to humification processes during com-
posting. Similarly, the aromaticity index provides an
overall view of the evolution of aromatic compounds
and enables humification in compost to be characterized
with respect to the accumulation of aromatic com-
pounds (Vinceslas-Akpa and Loquet, 1997). This index
increased (Albrecht et al. 2008b) in direct proportion to
A3 and A4 components.
One point to clarify is, even if the prevalence of
humic substances as polymers in soils has been
assumed for decades, a large body of evidence shows
instead an alternative understanding of their confor-
mational nature. Humic substances should be regarded
as supramolecular associations of self-assembling
heterogeneous and relatively small molecules deriving
from the degradation and decomposition of dead
biological material (Piccolo 2001). Recent studies
underlined the importance of ecosystem properties in
SOM stabilization processes, such as physical discon-
nection between SOM and microbes or organo-
mineral associations (Schmidt et al. 2011). This is
indeed supported by progress in analytical and visu-
alization techniques. Methods used to classify soil OM
into active, slow, and passive pools are no longer
chemically, but rather physically performed, based on
the various degrees of physical protection of SOM by
either aggregates, association force between SOM and
minerals, or different particle sizes (von Lu
¨tzow et al.
2007). However, in compost science, the ecosystem
properties of OM remain totally different and deter-
mining humic-like substance contents is a frequently
used tool to assess the degree of maturity during the
composting process (Sanchez-Monedero et al. 1999;
Tomati et al. 2000). This is why significant and
positive correlations between HA/FA ratio and A3 and
A4 give additional evidence that RE pyrolysis is
capable of providing information about changes of
OM during the composting process.
Even if it is often difficult to assess changes in
content of resistant biopolymers, several indices
reveal their transformations. Increase of cellulose
crystallinity expressed by the ordered cellulose
(89 ppm)/disordered cellulose (84 ppm) ratio and the
S/G (syringyl to guaiacyl) ratio revealed transforma-
tions of cellulose and lignin, respectively (Albrecht
et al. 2008b). Thus, increasing A3 and A4, related to
immature non-biotic constituents (A3) and mature
refractory fraction (A4) by Disnar et al. (2003) and
Sebag et al. (2006), during the composting process,
can be related to these transformations. Therefore,
these results confirm the nature of A3 and A4
components as being complex OM, still under humi-
fication or having started to undergo humification
processes.
Rock–Eval pyrolysis indices
TpS2 is a well-known maturity indicator for OM
(Espitalie et al. 1985a,b; Disnar 1994; Carrie et al.
2012) and its increase (312.5–327.3 C between 4 and
128 days––Table 1) is a good parameter of OM
maturity produced by aforementioned humification
processes. According to Saenger et al. (2013), TpS2
can be regarded as a proxy of the thermal energy
required by microorganisms to decompose OM.
Increase of TpS2 during the composting process
confirms the presence of easily degradable compo-
nents and/or plant biopolymers in both A1 and A2
components and the presence of more complex and/or
humified and/or recalcitrant compounds in A3 and A4.
To have a more explicit view of these results, a PCA
on RE pyrolysis (I and R- indices) and
13
C NMR data
was performed (Fig. 2). Time of composting (i.e.
humification of OM) is shown by a black dashed arrow
Biogeochemistry
123
in Fig. 2. This chronological distribution of composts
is negatively linked with the I-index (left hand side of
the diagram) but positively linked with R-index
related to the oldest compost (right hand side of the
diagram). Among all PCA parameters, I- and R-indi-
ces accounted for 10.6 and 12.4 %, respectively, of the
variance explained on PC1 and 23.4 and 19.9 % on
PC2. These large percentages of variance explained on
both axes reinforce the above-mentioned trend. This
representation of data highlights links between the
I-index and easily degradable fractions of OM, which
both decrease during composting (negatively corre-
lated), and R-index and complex OM, which both
increase (positively correlated).
In order to follow OM evolution using RE pyrolysis
and during the composting process, study results have
been projected on a I versus R plot built using a soil
database (in grey, Fig. 1). R and I have two opposite
behaviors along the timeline of composting. The
R-index increases from a value of 0.66 after 4 days to
0.75 after 128 days of composting. In contrast, the
I-index decreases during the same time span from 0.27
to 0.24 (Table 1). By using both indices, a comparison
of SOM evolution between different horizons in a
given profile, as well as between profiles from
different contexts, can be conducted (Copard et al.
2006; Sebag et al. 2006; Marchand et al. 2008; Poot
et al. 2009; Tambach et al. 2009; Graz et al. 2011;
Saenger et al. 2013). A similar comparison can be
drawn with compost samples: they are distributed
along the curve referring to organic soils and horizons
(in grey), and all compost samples are ranked accord-
ing to their degree of maturity. Young composts
(4–67 days) are plotted on the left hand side of the
diagram, near organic tissues and forest litter samples.
They are characterized by fresh and/or slightly
evolved OM. More mature composts (67–128 days)
are plotted on the right hand side of the diagram,
among organic horizons characterized by more com-
plex and/or evolved OM. This chronological distribu-
tion of compost samples along the curve (in grey)
characterizing the degree of transformation of OM
(from organic tissues/forest litter samples to organic
horizons on the right hand side of the diagram)
confirms the pertinence of RE pyrolysis to assess OM
transformation, i.e. humification and mineralization
processes. Indeed, an IR plot may be a valuable tool
for monitoring bio-wastes composting maturation. In
addition, box plots (Fig. 1) display a gap, delimited by
a threshold after 57/67 days, in the I-index as well as
in the R-index. This threshold is related to the increase
of maturity by changes in the compost chemical
composition: a decrease of aliphatic compounds and a
concomitant increase of humic substances and aro-
maticity (Albrecht et al. 2008b), as well as a change in
the microbial community structures and functional
diversity (Albrecht et al. 2010). To conclude, these
gaps, observed as well by NIRS (Albrecht et al. 2008a)
and
13
C NMR (Albrecht et al. 2009) confirm that RE
pyrolysis and the associated I and R-indices constitute
a valuable approach to an accurate and rapid estima-
tion of the degree of OM transformation.
Conclusions
Rock–Eval pyrolysis and the use of I (for immature
OM) and R (for recalcitrant OM) indices constitute a
valuable approach to a fast and accurate assessment of
transformation degrees of OM. This has been demon-
strated by using a model of composting as it is a
revealing model to study OM evolution (mineraliza-
tion and humification processes) and its by-products
are well documented. In addition, in a first step to
explore the pertinence of RE pyrolysis for SOM,
compost excludes the complex ‘‘interferences’’ due to
pedogenetic processes occurring in soils. The preci-
sion of RE pyrolysis is good enough to identify
thresholds (57/67 days). These thresholds have been
demonstrated to be related to an increase in maturity
due to changes in chemical composition, as well as in
microbial community structures and functional diver-
sity changes. The presented results show that relation-
ships between indices from RE pyrolysis and the main
classes of OM characterized by
13
C NMR can be
significantly correlated. The R index (contribution of
bio-macromolecules) is mainly correlated with phe-
nolic and aromatic C, whereas the I index refers to
labile––easily degradable components (alkyl C et
O-alkyl C). In other words, these results confirm the
pertinence of RE pyrolysis to monitor OM dynamics.
References
Albrecht R, Joffre R, Gros R, Le Petit J, Terrom G, Perissol C
(2008a) Efficiency of near-infrared reflectance spectros-
copy to assess and predict the stage of transformation of
Biogeochemistry
123
organic matter in the composting process. Bioresour
Technol 99(2):448–455
Albrecht R, Ziarelli F, Alarcon-Gutierrez E, Le Petit J, Terrom
G, Perissol C (2008b) (13)C solid-state NMR assessment of
decomposition pattern during co-composting of sewage
sludge and green wastes. Eur J Soil Sci 59(3):445–452
Albrecht R, Joffre R, Le Petit J, Terrom G, Perissol C (2009)
Calibration of Chemical and Biological Changes in Co-
composting of Biowastes Using Near-Infrared Spectros-
copy. Environ Sci Technol 43(3):804–811
Albrecht R, Perissol C, Ruaudel F, Le Petit J, Terrom G (2010)
Functional changes in culturable microbial communities
during a co-composting process: carbon source utilization
and co-metabolism. Waste Manag 30(5):764–770
Carrie J, Sanei H, Stern G (2012) Standardisation of Rock-Eval
pyrolysis for the analysis of recent sediments and soils. Org
Geochem 46:38–53
Copard Y, Di-Giovanni C, Martaud T, Albe
´ric P, Olivier J-E
(2006) Using Rock-Eval 6 pyrolysis for tracking fossil
organic carbon in modern environments: implications for
the roles of erosion and weathering. Earth Surf Proc Land
31(2):135–153
De Bertoldi M, Vallini G, Pera A (1983) The biology of com-
posting: a review. Waste Manage Res 1(2):157–176
Disnar JR (1994) Determination of maximum paleotempera-
tures of burial (MPTB) of sedimentary rocks from pyro-
lysis data on the associated organic matter: basic principles
and practical application. Chem Geol 118(1–4):289–299
Disnar JR, Guillet B, Keravis D, Di-Giovanni C, Sebag D (2003)
Soil organic matter (SOM) characterization by Rock-Eval
pyrolysis: scope and limitations. Org Geochem 34(3):327–343
Epstein E (1997) The science of composting. Technomic Pub-
lishing Company, Pennsylvania
Espitalie J, Deroo G, Marquis F (1985a) La pyrolyse Rock-Eval
et ses applications. Premie
`re partie. Oil Gas Sci Technol
40(5):563–579
Espitalie J, Deroo G, Marquis F (1985b) La pyrolyse Rock-Eval
et ses applications. Deuxie
`me partie. Oil Gas Sci Technol
40(6):755–784
Gillespie AW, Sanei H, Diochon A et al (2014) Perennially and
annually frozen soil carbon differ in their susceptibility to
decomposition: analysis of Subarctic earth hummocks by
bioassay, XANES and pyrolysis. Soil Biol Biochem
68:106–116
Graz Y, Di-Giovanni C, Copard Y, Elie M, Faure P, Defarge FL,
Leveque J, Michels R, Olivier JE (2011) Occurrence of
fossil organic matter in modern environments: optical,
geochemical and isotopic evidence. Appl Geochem
26(8):1302–1314
Hare AA, Kuzyk ZZA, Macdonald RW et al (2014) Charac-
terization of sedimentary organic matter in recent marine
sediments from Hudson Bay, Canada, by Rock-Eval
pyrolysis. Org Geochem 68:52–60
Hete
´nyi M, Nyilas T, Sajgo
´C (2010) Organic geochemical
evidence of late Pleistocene-Holocene environmental
changes in the Lake Balaton region (Hungary). Org Geo-
chem 41(9):915–923
Inbar Y, Chen Y, Hadar Y (1991) C-13 Cpmas Nmr and Ftir
Spectroscopic Analysis of Organic-Matter Transformations
During Composting of Solid-Wastes from Wineries. Soil Sci
152(4):272–282
Kogel-Knabner I (2000) Analytical approaches for character-
izing soil organic matter. Org Geochem 31(7–8):609–625
Lafargue E, Marquis F, Pillot D (1998) Rock-Eval 6 applica-
tions in hydrocarbon exploration, production, and soil
contamination studies. Revue de l’Institut Francais du
Petrole 53(4):421–437
Marchand C, Lallier-Verges E, Disnar JR, Keravis D (2008)
Organic carbon sources and transformations in mangrove
sediments: a Rock-Eval pyrolysis approach. Org Geochem
39(4):408–421
Massiot D, Fayon F, Capron M, King I, Le Calve S, Alonso B,
Durand JO, Bujoli B, Gan ZH, Hoatson G (2002) Model-
ling one- and two-dimensional solid-state NMR spectra.
Magn Reson Chem 40(1):70–76
Piccolo A (2001) The supramolecular structure of humic sub-
stances. Soil Sci 166:810–832
Poot A, Quik JTK, Veld H, Koelmans AA (2009) Quantification
methods of Black Carbon: comparison of Rock-Eval ana-
lysis with traditional methods. J Chromatogr A
1216(3):613–622
Saenger A, Ce
´cillon L, Sebag D, Brun J-J (2013) Soil organic
carbon quantity, chemistry and thermal stability in a
mountainous landscape: a Rock-Eval pyrolysis survey. Org
Geochem 54:101–114
Sanchez-Monedero MA, Roig A, Cegarra J, Bernal MP (1999)
Relationships between water-soluble carbohydrate and
phenol fractions and the humification indices of different
organic wastes during composting. Bioresour Technol
70:193–201
Schmidt MWI, Torn MS, Abiven S, Dittmar T, Guggenberger
G, Janssens IA, Kleber M, Kogel-Knabner I, Lehmann J,
Manning DAC, Nannipieri P, Rasse DP, Weiner S,
Trumbore SE (2011) Persistence of soil organic matter as
an ecosystem property. Nature 478:49–56
Schnitzer M (2005) Organic Matter - Principles and Processes.
In: Daniel H (ed) Encyclopedia of Soils in the Environ-
ment. Elsevier, Oxford, pp 85–93
Sebag D, Disnar JR, Guillet B, Di Giovanni C, Verrecchia EP,
Durand A (2006) Monitoring organic matter dynamics in
soil profiles by ‘Rock-Eval pyrolysis’: bulk characteriza-
tion and quantification of degradation. Eur J Soil Sci
57(3):344–355
Swift RS (1996) Organic matter characterization. In Methods of
Soil Analysis. Part 3. Chemical Methods; Soil Science
Society of America. Madison, WI 1996:1011–1069
Tambach TJ, Veld H, Griffioen J (2009) Influence of HCl/HF
treatment on organic matter in aquifer sediments: a Rock-
Eval pyrolysis study. Appl Geochem 24(11):2144–2151
Tomati U, Madejon E, Galli E (2000) Evolution of humic acid
molecular weight as an index of compost stability. Com-
post Sci Util 8:108–115
Vinceslas-Akpa M, Loquet M (1997) Organic matter transfor-
mations in lignocellulosic waste products composted or
vermicomposted (eisenia fetida andrei): chemical analysis
and
13
C CPMAS NMR spectroscopy. Soil Biol Biochem
29(3–4):751–758
von Lu
¨tzow M, Ko
¨gel-Knabner I, Ekschmitt K, Flessa H,
Guggenberger G, Matzner E, Marschner B (2007) SOM
fractionation methods: relevance to functional pools and to
stabilization mechanisms. Soil Biol Biochem
39:2183–2207
Biogeochemistry
123
