Journal of Analytical and Applied Pyrolysis 126 (2017) 332–336

Contents lists available at ScienceDirect

Journal of Analytical and Applied Pyrolysis

journal homepage: www.elsevier.com/locate/jaap

Pyrolysis-gas chromatography–mass spectrometry Kováts retention index of MARK

pyrolysis products of lignocellulosic materials
⁎
Tamires O. Meloa,b, Francisco A. Marquesa, Fabricio A. Hanselb,
a
Departamento de Química, Universidade Federal do Paraná, Curitiba 81531 980, PR, Brazil
b
Embrapa Florestas, Estrada da Ribeira, km 111, C.P. 319, Guaraituba, Colombo 88411 000, PR, Brazil

A R T I C L E I N F O A B S T R A C T

Keywords: Pyrolysis – gas chromatography – mass spectrometry (Py-GC–MS) is a common hyphenated analytical technique
Wood for the characterisation of lignocellulosic materials, but such natural polymers generate highly complex total ion
Pulp current pyrograms. Recently, normalised retention time and automated deconvolution of chromatograms have
Lignin been shown to be a viable method for analyses containing a large number of peak components. However, the use
Cellulose
of Kováts retention index (RI) has not been examined in Py-GC–MS for lignocellulosic materials. In this study, we
Bio-oil
test the use of linear alkane polyethylene pyrolysis products to calibrate the Kováts RI for pyrolysis products of
Fast pyrolysis
AMDIS lignocellulosic materials in a Py-GC–MS system. Linear alkane polyethylene pyrolysis products yield a correct
Kováts RI, which means the method has the advantages of being able to apply the NIST/EPA/NIH mass spectra
library database for peak identification. In addition, using such an approach, Kováts RI of standard compounds
calibrated using a GC–MS system may be used for structural certification of unknown pyrolysis products due to
the similarities of RI calibration in both systems (GC–MS x Py-GC–MS).

1. Introduction Recently, Smits et al. [21] used a normalised RT and automated

Pyrolysis is a common analytical technique for the characterisation
of lignocellulosic materials, an important raw material for several
industrial branches (e.g. pulp and paper, biorefinary an metallurgy),
that use different products (e.g. cellulose, hemicellulose, lignin and
charcoal) [1–11]. Pyrolysis of natural or synthetic products generates a
set of complex mixtures of smaller pyrolysis product molecules. The
composition of such molecules is related to the composition of the
original sample, and due to their complexity, it is necessary to combine
pyrolysis with other techniques such as gas chromatography and mass
spectrometry (Py-GC–MS), to aid the separation and identification of
the species produced by pyrolysis [12,13].
Natural polymers generate pyrograms with several peaks, and mass
spectra and pyrogram libraries are only partially relevant to the
interpretation of a substantial amount of peak data [14]. Until now
the use of libraries has ignored the majority of the information available
in the pyrograms, since peak identification has been based on mass
spectra and retention time (RT) [15–19]. Various gas chromatography
(GC) parameters influence the RT during Py-GC–MS analysis, such as
the column length, film thickness, diameter, and size of the transfer line
[20]; this represents a challenge for the identification of large numbers
of compounds with similar mass spectra and slightly different RT.
deconvolution of pyrograms, providing a viable alternative method to
ease the analysis of pyrograms with very complex peak-components.
It is known that the use of the retention index (RI) is more useful
than using the RT for peak identification, since RI are quite independent
of GC parameters. The temperature programming Kováts RI has been
used for a long time [22,23], and a database using this index has been
updated, making its use interesting for the identification of compounds
resulting from a complex matrix [18,24,25,26]. The linear alkanes
series (e.g. C7–C40) are usually used as reference standards to calculate
the Kováts RI in GC–MS analysis. Therefore, the pyrolysis products of
polyethylene (i.e. dienes, alkenes and alkanes [27]) should provide the
linear alkanes needed to calculate the Kováts RI during pyrolysis
analysis. A previous study has demonstrated the possibility of using
polyethylene pyrolysis products for RI calibration and identification of
pyrolysis products from paper pulp [14]. Herein we describe, for the
first time, the use of polyethylene as a source of linear alkanes to
calibrate the Kováts RI for Py-GC–MS analysis of lignocellulosic
materials. The efficiency of the proposed methodology was validated
by comparing the Kováts RI generated from lignocellulosic pyrolysis
products with bio-oil compound Kováts RI calibrated using GC–MS
analysis.

⁎
Corresponding author.
E-mail address: fabricio.hansel@embrapa.br (F.A. Hansel).

http://dx.doi.org/10.1016/j.jaap.2017.05.013
Received 7 February 2017; Received in revised form 9 May 2017; Accepted 27 May 2017
Available online 29 May 2017
0165-2370/ © 2017 Elsevier B.V. All rights reserved.
T.O. Melo et al. Journal of Analytical and Applied Pyrolysis 126 (2017) 332–336

Fig. 1. Partial total ion current chromatogram (A) and pyrogram (B) showing the characteristics of linear alkanes solution and pyrolysis products of polyethylene during GC–MS and Py-
GC–MS analysis, respectively. The dotted line indicates the positive difference in linear alkane Kováts RI during Py-GC–MS when calibrated using the linear alkanes solution in the GC–MS
analysis.

Fig. 2. Comparison of the correlation of the Kováts RI using an equivalent 5%-phenyl-95%-dimethylpolysiloxane stationary phase in the GC–MS and Py-GC–MS systems.

333
T.O. Melo et al. Journal of Analytical and Applied Pyrolysis 126 (2017) 332–336

Table 1 2.2. GC–MS and Py-GC–MS analysis

Average Kováts RI and main mass fragments for the compounds identified in CG-MS
analysis of the bio-oil and in Py-GC–MS analysis of lignocellulosic materials.
The extracts were injected (1 μL, Thermo Triplus AS) and the solids
ID RI Identity m/z were pyrolysed (c.a. 250 μg, CDS 5000 pyroprobe) into a Focus GC gas
chromatography tandem to a Polaris Q ion trap mass spectrometer
C01 838.0 Furfural 95 96 67 (Thermo), equipped with a DB5 ms capillary column
C02 916.9 Furanone 84 55 42 (30 m × 0.25 mm, 25 μm film thickness). Pyrolysis was carried out at
C03 963.2 5 - methyl - 2 - furaldehyde 109 81 53
P01 983.3 Phenol 94 66 74
550 °C for 10 s, and the coil heater was set at 10 °C ms−1. The other
C04 999.1 4 - hydroxy - 5,6 - dihydro - (2H) - pyran - 2 - 114 85 58 pyroprobe conditions were: temperature interface programme – 110 °C
one to 290 °C at 100 °C min−1 held for 120 s; oven temperature 240 °C; and
C05 1020.9 C1 - dihydro pyranone 112 84 55 transfer line temperature 240 °C. The GC oven temperature was
C06 1024.7 Hydroxy methyl cyclopentenone 112 84 55
programmed from 40 °C (held for 8 min) to 280 °C at 7 °C min−1, then
P02 1054.9 C1 - phenola 107 108 77
L01 1089.7 Guaiacol 124 109 81 held for 15 min. Helium, at a constant flow of 1.0 mL min−1, was the
P03 1147.0 C2 - phenola 122 107 77 carrier gas. The inlet in split mode 1:100 was set at 230 °C. The GC–MS
P04 1150.8 C2 - phenola 122 107 77 interface and ion source temperatures were 250 °C and 200 °C, respec-
L02 1179.5 C1 - guaiacol 138 123 95 tively. The ion trap mass spectrometer was operated in the positive
C07 1215.0 1,4:3,6 - dianhydro - α - D - glucopyranose 98 69 57
P05 1236.4 C3 - phenola 136 121 91
impact electronic mode at 70 eV scanning the range m/z 40–650 in a
L03 1256.7 Hydroxy guaiacol 140 125 97 total scan time of 0.59 s and emission current 250 mA.
L04 1275.6 4 - ethyl guaiacol 152 137 77
L05 1312.8 4 - vinyl guaiacol 150 135 107 2.3. Peak identification and data processing
L06 1348.7 Syringol 154 139 93
L07 1354.7 Eugenol 164 149 131
L08 1364.7 C3 - Guaiacola 166 137 122 Mass spectral deconvolution and automated calculation of RI was
L09 1398.6 Vanillin 151 123 109 performed by the automated mass spectral deconvolution and identi-
L10 1405.7 Isoeugenol (cis) 164 149 131 fication system (AMDIS, National Institute of Standards and
L11 1443.2 C1 - syringola 168 153 125 Technology, Gaithersburg, MD, USA). Standard solutions of linear
L12 1450.9 Isoeugenol (trans) 164 149 131
L13 1456.6 Homovanillin 166 137 122
alkanes (C7–C30, Sigma-Aldrich 49451-U) and polyethylene (CDS
C08 1505.1 Levoglucosan 60 73 97 Analytical, 0100-9014) were used for Kováts RI calibration in the
L14 1520.1 C2 - syringola 182 167 77 GC–MS and Py-GC–MS systems, respectively. Data deconvolution was
L15 1526.5 Guaiacyl acetone 180 137 122 performed with the following specifications: component width = 12;
L16 1561.8 4 - vinyl syringol 180 165 137
adjacent peak subtraction = 2; resolution = low; sensitivity = very
L17 1583.0 Propioguaiacone 180 165 151
L18 1596.6 4 - allyl syringol 194 81 179 low; shape requirements = medium. Compounds were identified from
L19 1603.7 C3 - syringola 196 167 151 the deconvoluted mass spectra by comparison with mass spectra
L20 1646.6 4 - propenyl syringol (cis) 194 81 179 published in the specialised literature [7,28,29,30,31].
L21 1657.8 Syringaldehyde 182 167 139
L22 1699.5 4 - propenyl syringol (trans) 194 81 179
L23 1728.3 Acetosyringone 196 181 153
3. Results and discussion
L24 1734.7 Coniferyl aldehyde (trans) 178 161 147
L25 1762.5 Syringyl acetone 210 167 123 Fig. 1 depicts the total ion current (TIC) chromatogram of linear
L26 1820.4 Propiosyringone 210 181 153 alkanes (C7–C30) and the TIC pyrogram of polyethylene pyrolysis
products. The pyrogram encompasses a pattern of triplet peaks includ-
a
Isomer abbreviations, C1 – methyl, C2 – dimethyl or ethyl, C3 – trimethyl, methyl –
ethyl or propyl.
ing the dienes, alkenes and alkanes (Fig. 1B); the linear alkanes range
from carbon numbers C6 to C25. The alkane distribution in the
2. Materials and methods polyethylene pyrolysis products afforded consistent Kováts RI in the
range 600–3000, which is suitable for the pyrolysis products of
2.1. Samples lignocellulosic materials [9,10,32,33]. Interestingly, when the Kováts
RI of the linear alkane pyrolysis products from polyethylene was
Bio-oil was produced from hardwood mulch in a pilot-scale calibrated using the linear alkanes solution (C7–C30) running in the
fluidised-bed fast pyrolysis reactor (20 kg/h, 500 °C, 5 s). Bio-oil was GC–MS system, the Kováts RI values for the linear alkane pyrolysis
separated into light (LF) and heavy (HF) fractions using molecular products were higher, indicating an increased RT (Fig. 1). Such
distillation (5000 Pa, feed flow 1 mL/min, evaporating temperature differences were interpreted as being caused by the length of transfer
70 °C, condensation temperature 23 °C, KDL-1, UIC, Germany). To the line of the pyrolysis equipment, and importantly, it suggests that linear
LF and HF fractions (100 μL) were added water (2 mL) and chloroform alkane pyrolysis products of polyethylene can be used for retention time
(2 mL). The mixture was stirred and centrifuged (5 min, 3000 rpm). The correction and RI calibration of the pyrolysis products of lignocellulosic
water phase was transferred to an Eppendorf tube and dried under material during the Py-GC–MS analysis.
vacuum for 18 h (Speed Vacuum–Eppendorf), and the chloroform phase In correlation analysis of Kováts RI, compounds identified in GC–MS
was transferred to a vial (2 mL) and dried under a gentle stream of were plotted against the same compounds identified in Py-GC–MS
nitrogen. The HF(aq) and LF (aq) were dissolved in acetone (100 μL) and (intercept equal to zero), and the slope was equal to one, with a high
HF(org) and LF(org) were dissolved in chloroform (100 μL) and analysed apparent linear correlation of RI values (Fig. 2). The Kováts RI of the
by gas chromatography–mass spectrometry (GC–MS). The solid materi- compounds in the GC–MS system were practically identical to the
als: (i) distilled bio-oil residue; (ii) charcoal; Eucalyptus (iii) pulp; (iv) Kováts RI of the pyrolysis products in Py-GC–MS analysis calibrated by
bleached pulp; (v) kraft lignin; and (vi) Pinus kraft lignin; were crushed using polyethylene, thus certifying the use of linear alkane pyrolysis
using a mortar and sieved through an 80 μm mesh screen before being products of polyethylene for Kováts RI calibration of pyrolysis products
analysed by pyrolysis (Py)-GC–MS (i and ii were produced by the pilot- of lignocellulosic material. Levoglucosan showed a slight deviation
scale fluidised-bed fast pyrolysis). from linearity (Fig. 2); such anhydrosugars are not well resolved by Py-
GC–MS due to the occurrence of polar OH groups in the molecule,
which cause broader peaks with loss of resolution in the non-polar GC
column [34]. However, the other polar compounds detected did not

334
T.O. Melo et al.
Fig. 3. RI and RT coefficients of variation (%) for the compounds used in the comparison analysis (A), RI compound percent relative errors of GC–MS compared with Py-GC–MS (B), and
RT compound percent relative errors of GC–MS compared with Py-GC–MS (C). For the compound abbreviations see Table 1.
show such strong effects on their resolution (Fig. 2 and Table 1).
Table 1 shows the 39 compounds detected in both systems, and the
matched compounds identified are members of the compound classes
commonly detected in the pyrolysis products of lignocellulosic materi-
als (i.e. furans, pyrans, anhydrosugars, phenols and lignin-derived
phenols). These compounds were identified by their main mass frag-
ments (m/z) using published mass spectrum libraries [7,28,29,30,31]
and RIs [35]. The compound RIs were very similar for both systems
(GC–MS and Py-GC–MS) using the 5%-phenyl-95%-dimethylpolysilox-
ane GC column. The general RI average for each compound (i.e.
considering all replicates in both systems) is also shown in Table 1.
The coefficients of variation (CV) were measured considering the
whole RI and RT averages for each single compound (Table 1, Fig. 3A).
The CVs ranged from 0.1 to 4.2% and 0.0 to 1.3% for RT and RI,
respectively. In fact, the RI CV was smaller than the RT CV for all
compounds, except for C05, which indicates greater precision for the
identification of compounds. Although it is known that RT accumulates
chromatographic errors during chemical identification [17–19,36], it
seems here that it simply follow the same pattern as seen in Fig. 1 in
which the linear alkane pyrolysis products of polyethylene were
compared with a linear alkane solution using the same GC programmed
oven. The effect was clearly shown by the percentage relative error of
RT and RI of GC–MS analysis compared with Py-GC–MS (Fig. 3B, C).
The RT error gives high negative values for most compounds, indicating
335
Journal of Analytical and Applied Pyrolysis 126 (2017) 332–336
an early RT for the bio-oil compounds during GC–MS analysis compared
with the same pyrolysis products in Py-GC–MS analysis of lignocellu-
losic materials (Fig. 3C). Conversely, the RI errors were smaller and
showed overall slight negative values, certifying similar RIs for the
same compounds in the GC–MS and Py-GC–MS systems (Fig. 3B). Two
factors have been suggested to contribute to the percent relative error:
(i) the error is a function of the RI and the boiling point of the analytes
(i.e. higher error at earlier RT), and (ii) single analytes may exhibit
abnormally high deviations (e.g. levoglucosan [37]). Interestingly,
given the similarities of the RIs in both systems, RIs of standard
compounds calibrated in the GC–MS system can be used for structural
certification of unknown pyrolysis products calibrated via polyethy-
lene, including isomers with similar mass spectra and slightly different
RTs.
4. Conclusions
Linear alkane polyethylene pyrolysis products yielded correct
calibration of Kováts RI in the Py-GC–MS system. Recently developed
methods for deconvolution and retention time normalisation enable
automatic data treatment for peak identification in Py-GC–MS, provid-
ing highly detailed information on the complex pyrograms [14,21]. Our
results show that the use of linear alkane pyrolysis products of
polyethylene can afford correct Kováts RI on the Py-GC–MS system,
T.O. Melo et al.
thus allowing the use of the NIST/EPA/NIH mass spectrum library
database for peak identification (i.e. target mass fragments and Kováts
RI [18]). In addition, by using such an approach, the Kováts RI of
standard compounds calibrated in a GC–MS system may be used for
structural certification of unknown pyrolysis products due to the
similarities of the RI in both systems (GC–MS x Py-GC–MS).
Acknowledgments
Funding was provided by a CNPq project (406897/2013-1) led by
Dr. Washington Magalhães. The authors also thank the anonymous
reviewers for comments.
References
[1] P.C. Uden, Nomenclature and terminology for analytical pyrolysis (IUPAC recom-
mendations 1993), Pure Appl. Chem. 65 (1993) 2405–2409.
[2] A.V. Bridgwater, G.V.C. Peacocke, Fast pyrolysis processes for biomass, Renew.
Sustain. Energy Rev. 4 (2000) 1–73, http://dx.doi.org/10.1016/S1364-0321(99)
00007-6.
[3] A. Demirbaş, Mechanisms of liquefaction and pyrolysis reactions of biomass, Energy
Convers. Manag. 41 (2000) 633–646, http://dx.doi.org/10.1016/S0196-8904(99)
00130-2.
[4] T. Ohra-Aho, M. Tenkanen, T. Tamminen, Direct analysis of lignin and lignin-like
components from softwood kraft pulp by Py-GC/MS techniques, J. Anal. Appl.
Pyrolysis 74 (2005) 123–128, http://dx.doi.org/10.1016/j.jaap.2004.11.010.
[5] M.F. Nonier, N. Vivas, N. Vivas De Gaulejac, C. Absalon, P. Souli, E. Fouquet,
Pyrolysis-gas chromatography/mass spectrometry of Quercus sp. wood: application
to structural elucidation of macromolecules and aromatic profiles of different
species, J. Anal. Appl. Pyrolysis 75 (2006) 181–193, http://dx.doi.org/10.1016/j.
jaap.2005.05.006.
[6] F.O. Silvério, L.C.A. Barbosa, D. Piló-Veloso, A pirólise como técnica analítica,
Química Nov. 31 (2008) 1543–1552.
[7] J.C. del Río, J. Rencoret, G. Marques, J. Li, G. Gellerstedt, J. Jiménez-Barbero,
Á.T. Martínez, A. Gutiérrez, Structural characterization of the lignin from jute
(Corchorus capsularis) fibers, J. Agric. Food Chem. 57 (2009) 10271–10281, http://
dx.doi.org/10.1021/jf900815x.
[8] A. Lourenço, J. Gominho, A.V. Marques, H. Pereira, Reactivity of syringyl and
guaiacyl lignin units and delignification kinetics in the kraft pulping of Eucalyptus
globulus wood using Py-GC–MS/FID, Bioresour. Technol. 123 (2012) 296–302,
http://dx.doi.org/10.1016/j.biortech.2012.07.092.
[9] Z. Chen, M. Hu, X. Zhu, D. Guo, S. Liu, Z. Hu, B. Xiao, J. Wang, M. Laghari,
Characteristics and kinetic study on pyrolysis of five lignocellulosic biomass via
thermogravimetric analysis, Bioresour. Technol. 192 (2015) 441–450, http://dx.
doi.org/10.1016/j.biortech.2015.05.062.
[10] C.F. Lima, L.C.A. Barbosa, M.N.N. Silva, J.L. Colodette, F.O. Silvério, In situ
determination of the syringyl/guaiacyl ratio of residual lignin in pre-bleached
eucalypt kraft pulps by analytical pyrolysis, J. Anal. Appl. Pyrolysis 112 (2015)
164–172, http://dx.doi.org/10.1016/j.jaap.2015.02.002.
[11] W. Jin, D. Shen, Q. Liu, R. Xiao, Evaluation of the co-pyrolysis of lignin with plastic
polymers by TG-FTIR and Py-GC/MS, Polym. Degrad. Stab. 133 (2016) 65–74,
http://dx.doi.org/10.1016/j.polymdegradstab.2016.08.001.
[12] T.P. Wampler, Practical applications of analytical pyrolysis, J. Anal. Appl. Pyrolysis
71 (2004) 1–12.
[13] M. Blazsó, Recent trends in analytical and applied pyrolysis of polymers, J. Anal.
Appl. Pyrolysis 39 (1997) 1–25 http://www.sciencedirect.com/science/article/pii/
S0165237096009564.
[14] P. Wenig, J. Odermatt, Efficient analysis of Py-GC/MS data by a large scale
automatic database approach: an illustration of white pitch identification in pulp
and paper industry, J. Anal. Appl. Pyrolysis 87 (2010) 85–92, http://dx.doi.org/10.
1016/j.jaap.2009.10.007.
[15] T.A. Lee, L.M. Headley, J.K. Hardy, Noise reduction of gas chromatography/mass
spectrometry data using principal component analysis, Anal. Chem. 63 (1991)
357–360, http://dx.doi.org/10.1021/ac00004a011.
[16] B. Spangenberg, A new proposal for a parameter-free integration software,
Fresenius, J. Anal. Chem. 360 (1998) 148–151, http://dx.doi.org/10.1007/
s002160050665.
336
Journal of Analytical and Applied Pyrolysis 126 (2017) 332–336
[17] F.C.-Y. Wang, P.B. Smith, Composition and microstructure analysis of chlorinated
polyethylene by pyrolysis gas chromatography and pyrolysis gas chromatography/
mass spectrometry, Anal. Chem. 69 (1997) 618–622, http://dx.doi.org/10.1021/
ac960947c.
[18] S.E. Stein, P. Ausloos, C.L. Clifton, J.K. Klassen, S.G. Lias, A.I. Mikaya,
O.D. Sparkman, D.V. Tchekhovskoi, V. Zaikin, D. Zhu, Evaluation of the NIST/EPA/
NIH Mass Spectral Library, Abstr. Pap. Am. Chem. Soc. 218 (1999) 368–368.
[19] P. Kusch, G. Knupp, W. Fink, D. Schroeder-Obst, V. Obst, J. Steinhaus, Application
of pyrolysis–gas chromatography–mass spectrometry for the identification of
polymeric materials, LCGC North Am. 32 (2014) 210212–210217.
[20] P.J. Linstrom, W.G. Mallard, The NIST chemistry WebBook: a chemical data
resource on the internet, J. Chem. Eng. Data 46 (2001) 1059–1063, http://dx.doi.
org/10.1021/je000236i.
[21] M.M. Smits, R. Carleer, J.V. Colpaert, PYQUAN. A rapid workflow around the
AMDIS deconvolution software for high throughput analysis of pyrolysis GC/MS
data, J. Anal. Appl. Pyrolysis 118 (2016) 335–342, http://dx.doi.org/10.1016/j.
jaap.2016.01.006.
[22] E. Kováts, Gas-chromatographische Charakterisierung organischer Verbindungen.
Teil 1: Retentionsindices aliphatischer Halogenide, Alkohole, Aldehyde und Ketone,
Helv. Chim. Acta 41 (1958) 1915–1932, http://dx.doi.org/10.1002/hlca.
19580410703.
[23] H. van Den Dool, P. Dec. Kratz, A generalization of the retention index system
including linear temperature programmed gas—liquid partition chromatography, J.
Chromatogr. A 11 (1963) 463–471, http://dx.doi.org/10.1016/S0021-9673(01)
80947-X.
[24] Q. Lu, X. Yang, X. Zhu, Analysis on chemical and physical properties of bio-oil
pyrolyzed from rice husk, J. Anal. Appl. Pyrolysis 82 (2008) 191–198, http://dx.
doi.org/10.1016/j.jaap.2008.03.003.
[25] V. Behrends, G.D. Tredwell, J.G. Bundy, A software complement to AMDIS for
processing GC–MS metabolomic data, Anal. Biochem. 415 (2011) 206–208, http://
dx.doi.org/10.1016/j.ab.2011.04.009.
[26] A. Mastrangelo, A. Ferrarini, F. Rey-Stolle, A. Garcia, C. Barbas, From sample
treatment to biomarker discovery: a tutorial for untargeted metabolomics based on
GC-(EI)-Q-MS, Anal. Chim. Acta 900 (2015) 21–35, http://dx.doi.org/10.1016/j.
aca.2015.10.001.
[27] T.P. Wampler, Introduction to pyrolysis-capillary gas chromatography, J.
Chromatogr. A 842 (1999) 207–220, http://dx.doi.org/10.1016/S0021-9673(98)
00921-2.
[28] O. Faix, D. Meier, I. Fortmann, Thermal degradation products of wood, Holz. Als.
Roh.-Und. Werkst. 48 (1990) 281–285, http://dx.doi.org/10.1007/BF02626519.
[29] O. Faix, I. Fortmann, J. Bremer, D. Meier, Thermal degradation products of wood
gas chromatographic separation and mass spectrometric characterization, Holz. Als.
Roh.-Und. Werkst. 49 (1991) 213–219.
[30] D. Meier, O. Faix, Pyrolysis-gas chromatography-mass spectrometry, in: S.Y. Lin,
C.W. Dence (Eds.), Methods Lignin Chem. Springer, Berlin, Heidelberg, 1992, pp.
177–199, , http://dx.doi.org/10.1007/978-3-642-74065-7_13.
[31] I. Pastorova, R.E. Botto, P.W. Arisz, J.J. Boon, Cellulose char structure: a combined
analytical Py-GC–MS, FTIR, and NMR study, Carbohydr. Res. 262 (1994) 27–47,
http://dx.doi.org/10.1016/0008-6215(94)84003-2.
[32] C.A. Mullen, A.A. Boateng, Catalytic pyrolysis-GC/MS of lignin from several
sources, Fuel Process. Technol. 91 (2010) 1446–1458, http://dx.doi.org/10.1016/j.
fuproc.2010.05.022.
[33] A.V. Marques, H. Pereira, Aliphatic bio-oils from corks: a Py-GC/MS study, J. Anal.
Appl. Pyrolysis 109 (2014) 29–40, http://dx.doi.org/10.1016/j.jaap.2014.07.016.
[34] D. Fabbri, G. Chiavari, S. Prati, I. Vassura, M. Vangelista, Gas chromatography/
mass spectrometric characterisation of pyrolysis/silylation products of glucose and
cellulose, Rapid Commun. Mass Spectrom. 16 (2002) 2349–2355, http://dx.doi.
org/10.1002/rcm.856.
[35] N. Ré-Poppi, M.R. Santiago-Silva, Identification of polycyclic aromatic hydrocar-
bons and methoxylated phenols in wood smoke emitted during production of
charcoal, Chromatographia 55 (2002) 475–481, http://dx.doi.org/10.1007/
BF02492280.
[36] N. Schauer, D. Steinhauser, S. Strelkov, D. Schomburg, G. Allison, T. Moritz,
K. Lundgren, U. Roessner-Tunali, M.G. Forbes, L. Willmitzer, A.R. Fernie, J. Kopka,
GC–MS libraries for the rapid identification of metabolites in complex biological
samples, FEBS Lett. 579 (2005) 1332–1337, http://dx.doi.org/10.1016/j.febslet.
2005.01.029.
[37] N. Strehmel, J. Hummel, A. Erban, K. Strassburg, J. Kopka, Retention index
thresholds for compound matching in GC–MS metabolite profiling, J. Chromatogr.
B Anal. Technol. Biomed. Life Sci. 871 (2008) 182–190, http://dx.doi.org/10.1016/
j.jchromb.2008.04.042.