The products of organic matter (OM) combustion or pyrolysis (B/P), have been commonly applied to environmental and paleoenvironmental characterization and reconstruction or as a basis for the chronology of deposits in sedimentology and archaeology. Wood B/P is thermal decomposition that evolves flammable gases being ignited in the presence of O2 when the temperature exceeds the temperature of ignition. The final burning products, under ideal conditions, are CO2 and H2O. In practice, B/P never proceeds under ideal conditions and, as a result, different products are formed. Material produced from
B/P varies and is described by different names: charcoal, soot, black carbon, elemental carbon, biochar and pyrogenic carbon (Bird, 2006; Preston and Schmidt, 2006; Bird and Ascough, 2012 and Pawlyta and Hercman, 2016). This reflects the complexity of its composition and structure, as well as its importance in a range of scientific areas. These B/P products differ not only in terminology but also in the methodology of analysis and interpretation of the results.
Combustion and pyrolysis of OM produces different types of carbon rich material. During the B/P processes, the relative content of carbon increases and is accompanied by a decrease in O and H, resulting from organic compound degradation. Because of increasing carbon content and structural changes, chemically stable aromatic rings are formed. The products final structure is B/P temperature dependent. At higher temperatures, microcrystalline domains may be formed and chemical
stability is greater, making B/P products resistant to post-depositional alteration (Goldberg, 1985; Vane and Abbott, 1999; Masiello, 2004; Czimczik et al., 2005; Preston and Schmidt, 2006; Eckmeier et al., 2007 and Bird and Ascough, 2012).
The resistant black residue that remains after B/P of OM is responsible for its common occurrence in a wide range of materials and environments. The occurrence of the products has been studied in very old sediments (Cressler, 2001), soils (Schmidt et al., 2002; Brodowski
The carbon isotopic composition of B/P products is widely used as a proxy in paleoclimatic, paleoecological, archaeological and biological studies or as an isotopic fingerprint (Bird and Cali, 1998; Cressler , 2001; Hall
The importance of this information has resulted in extensive methodological and experimental studies of the relation between B/P conditions and the isotopic composition of the B/P products. Most studies focus on the isotopic composition of charcoal, with respect to the B/P conditions (temperature, time, atmosphere; see Bird and Ascough, 2012). In Bird and Ascough (2012), the results of previous work were not only reviewed but a detailed discussion of factors affecting the results was provided, including the choice of material for testing, sample preparation, experimental methodology and analysis methodology.
The black residue that remains after wood combustion is a common material for archaeological and geological investigations. Most published results from these analyses concern carbonaceous microstructural studies, radiocarbon analyses and stable isotopic analyses in aerosols and clastic deposits (Currie et al., 1997; Masiello et al., 2002 and Brodowski et al., 2005). The primary reason for these studies is increasing interest in nanoparticles and their impact on the environments and humans (see review and references there: Nowack and Bucheli, 2007). Investigations of the accumulated wood combustion products in sediments give the chance to determine their provenance: human activity and natural phenomena (like forest or meadow fires). In archaeological and geological investigations it is also very important to distinguish the products from different sources that can be commonly found in the modern environment (e.g. wood burning at the fires, fossil fuel combustion in the engines, industry).
Several studies have examined carbonaceous material from caves. These studies have mostly been for archaeological purposes or to provide information about B/P product precursor (Petranek and Pouba, 1951; Bennington et al., 1962; Watson, 1966; Hill, 1982; Steelman et al., 2002; Gradziński et al., 2002, 2003, 2007; Chang et al., 2008 and Pawlyta and Hercman, 2016).
Products of the complete wood combustion are mostly gaseous, however in the case of incomplete burning, these can form well visible black layer or can be dispersed in the sediment. In the second case, identification of burning wood product is more difficult. One of the common products of incomplete combustion are soot aggregates, which have characteristic structure and can be easily identified even in extremely low concentrations using the transmission electron microscopy (TEM) or scanning electron microscopy (SEM) techniques after relatively simple chemical separation (Pósfai and Molnár, 2000 and Pawlyta and Hercman, 2016).
In our studies of lake and cave deposits we often found traces of combustion products dispersed in the sediments or inside of carbonate speleothems. Separated soot aggregates allow identification of the type of combusted material (Joeng
Three types of modern wood were selected: pine (
Basic data of wood samples.
Wood | Sample | Locality | Cut date | Coordinates | Growth (A.periodD.) a | Nb Mean (‰VPDB) δ13C |
---|---|---|---|---|---|---|
S2 | Siedlce | 2014, March | 52°10′00″N 22°16′30″E | 2013–2001 | 30 –26.5 ± 0.2 | |
D1 | Ojców | 2014, October | 50°12′24″N 19°49′45″E | 2009–2001 | 30 –27.4 ± 0.3 | |
B1 | Błonie | 2014, November | 52°11′48″N 20°37′01″E | 2009–2001 | 30 –24.7 ± 0.2 | |
Wood | Sample | Locality | Radiocarbon age | Calibrated radiocarbon agec | Mean δ13C | |
---|---|---|---|---|---|---|
(yrs) | (cal BP yrs) | (‰) | ||||
CD | Krakow (lab.) | Archivized sample | Poz-79538 | 1265–1184 (68.2%) | 10 –25.8 ± 0.2 | |
1275 ± 30 BP | 1289–1150 (95.4%) |
of the fossil oak was estimated based on radiocarbon analysis performed at the Poznań Radiocarbon Laboratory.
Two sets of wood burning/pyrolysis experiments were performed using burning chamber and furnace. Several types of samples for carbon stable isotope analyses were collected depending on experiment type. In the further part of the work we will use the terms: (1) “charcoal” for charred wood – black residuum collected from burning plate or from sample tube; (2) “ash” for gray powder collected from ash collector; (3) “acid” for transparent liquid sample with pH ~ 1 (“pyroligneous acid” consisting mainly of acetic acid and methanol); (4) “bio-oils” for yellow-brown colour oleic liquid and (5) soot aggregates for separated soot from volatiles condensed at collecting plate.
The first set of experiments involved the simulation of a fireplace in a burning chamber. The wood sample (80– 100 g) was located at the burning plate and an ash collector was located below the burning plate. The volatile burning products,
The next set of experiments was carried out in a furnace. Powdered wood sample (1–2 g) was put in a quartz tube between two quartz wool plugs, which was then placed in the heating chamber pre-heated to a predetermined temperature. The temperature was increased gradually in 50°C intervals in the 200–600°C range. At each temperature increment, the sample was kept for 45 min. A thermocouple was put inside the sample to allow temperature control. Experiments were performed under an atmosphere of N2 or synthetic air (21% O2 and 79% N2; H2O <3ppm, CO<0,1ppm, CO2<0,5ppm, H2<1ppm). At the exit of the sample tube, a water cooler was mounted to facilitate condensation of the volatile products released during the experiment. For further analysis, the “charcoal” and volatile components, condensed at the cooled exit of the sample tube (“acids” and “bio-oils”), were collected.
Wood, “charcoal”, “acids”, “bio-oils” and “ash” were analysed without any pre-treatment. Volatile products condensed at the collecting plate of burning chamber were mixtures of different B/P products, including soot, char and tar. To obtain information about the soot aggregates carbon isotopic composition, several purification steps were applied. Volatile products collected at the plate were divided into three parts in order to check the possible impact of the purification procedure on the isotopic composition. First part was analysed without pretreatment; second one was treated with 2% HCl (3 times, each 24 h in room temperature with washing between the acid steps); and third part was treated with 2% HCl (similarly as in the second step) and 40% HF for 1 h. At the end of the chemical procedure, the samples were repeatedly (
Carbon stable isotopes composition analysis was performed at the Stable Isotope Laboratory of the Institute of Geological Sciences of the Polish Academy of Sciences in Warsaw. Stable isotopic composition of carbon for solid and liquid samples (wood, charred wood, soot aggregates, ash and oil) was determined using a Thermo Flash EA 1112 HT elemental analyser connected to a Thermo Delta V Advantage Isotope Ratio Mass Spectrometer in continuous flow mode. Samples (
The carbon isotopic composition of CO2 was measured by a Picarro G2201-i Analyzer using a cavity ring-down spectroscopy (CRDS) method. At the experiment time, CO2 was continuously pumped at
Isotope ratios are reported as dispersion from the standard (δ value) and expressed in ‰ relative to V-PDB standard.
For the description of carbon isotopic composition, we define the “variation factor” (ΔC) as the difference between average of measured δ13C values in the B/P product and the average of the bulk wood sample (listed in
According to this definition, positive values of ΔC indicate enrichment in 13C in the products relatively to the wood. Negative values of ΔC indicate depletion of 13C in the product relatively to the wood. Due to the small number of samples, 95% confidence intervals of ΔC were estimated using Monte Carlo simulation method.
In order to explain relationships between carbon isotopes in the solid, liquid and gaseous combustion products, two series of molecular simulations were performed.
In the first one, models of two the most abundant components of wood, were tested: cellulose and lignin. Cellulose structure was built of 9 glucose monomers. Lignin model containing C160H180O58 was provided by courtesy of Tingting Zhang (Zhang et al., 2016). Reaxff force field optimized for C/H/O combustion (Chenoweth et al., 2008) was used in LAMMPS computer simulation program (Plimpton, 1995). Timestep was set at 0.25 ps, cutoff at 10.0 Å and temperature was increasing continuously 1ºC/ps from 27ºC up to 2227ºC in the NVT ensemble. Simulations for single molecules and with additional oxygen in pressure of ca 10 atm. were additionally performed. Based on the results of simulations, evolution with temperature of pyrolysate masses containing certain number of carbon atoms were calculated. Analysis of the simulation frames was performed recursively for all atoms – atoms belong to chain with certain number of carbon atoms if it was in distance below 1.8 Å to any atom building the molecule (C, O, H). In order to improve statistics, six series of simulations were performed in the same conditions but with slightly modified starting structure.
In the second set of simulations, an isotope exchange reactions between pyrene (a model of residuum) and different gaseous and liquid products were tested:
The aim was to calculate Δ energy, Δ enthalpy and Δ Gibbs free energy of reactions along with zero-point energy (ZPE) and thermal corrections at two significantly different temperatures of 27ºC (300 K) and 427ºC (700 K). After optimization of all the structures, vibration frequencies were computed. Thermal corrections were calculated based on translational, rotational and vibrational contributions (Ochterski, 2000). All the calculations were performed using Gaussian 03 program (Frisch et al., 2004) with Density Functional Theory (DFT) at the B3LYP/DGDZVP level of theory. All the thermodynamic parameters were also provided by Gaussian 03 program.
At the beginning, the temperature range of the experiments had to be determined. The furnace experiments with pine wood were carried out in a N2 and synthetic air (free of CO2) atmosphere. The furnace temperature was increased gradually in 50°C intervals (
emission. For temperature values <200°C, detected traces of CO2 were below recommended operational range of Picarro Analyzer (100 ppm) and its absolute values must be treated with caution. CO2 concentration increased to 1400 ppm when the temperature was increased to 200°C. After a short time, the concentration stabilized. During the next change of temperature, to 250°C, the CO2 concentration increased beyond the Picarro Analyzer limit (4000 ppm).
Second indicator of reaction intensity was the mass loss during combustion (
The results suggest that the B/P reaction started at 200–250°C for oak and beech, and below 200°C for pine. Taking this information into account, further experiments with the furnace were performed in the 200–600°C range.
Two heating experiments in the furnace have been performed, the first one, under a N2 atmosphere and the second one, in synthetic air (free of CO2). Examples of results obtained for pine wood are presented in
The first series of experiments were done using burning chamber as simulation of natural fireplace condition. The carbon isotopic composition values of the combustion products (three independent analyses for each one) from natural fireplace simulation are presented in
“Variation factor” of wood burning products from combustion chamber experiments with estimated 95% confidence intervals*.
Variation factor ΔC (‰) | ||||
---|---|---|---|---|
Beech | Pinus | Oak | Fossil oak | |
–0.50 ± 0.26 | 0.39 ± 0.28 | 0.80 ± 0.35 | –0.49 ± 0.21 | |
–0.44 ± 0.22 | 0.22 ± 0.28 | 0.58 ± 0.35 | –0.46 ± 0.21 | |
–0.35 ± 0.25 | 0.45 ± 0.28 | 0.89 ± 0.36 | –0.42 ± 0.21 | |
–0.79 ± 0.26 | –1.54 ± 0.28 | –0.13 ± 0.36 | ||
–2.53 ± 0.25 | –0.66 ± 0.28 | 0.14 ± 0.36 | –0.69± 0.23 | |
–0.21 ± 0.28 | 0.90 ± 0.38 | 1.05 ± 0.50 | 3.39 ± 0.21 |
The carbon dioxide for all wood types, except beech, was enriched in 13C relatively to the original wood. Values of ΔC for CO2 achieved 0.90 ± 0.38‰ and 3.39 ± 0.21‰ for pine and fossil oak, respectively. The volatile burning products collected after condensation onto the quartz-glass plate were slightly enriched in 13C for pine and oak (ΔC values up to 0.4‰ and 0.8‰ respectively) and depleted for fossil oak and beech (ΔC values up to –0.5‰). The charred wood (“charcoal”) and “ash” samples were generally depleted in 13C relatively to the original wood, except for oak, where differences were in the range of uncertainty.
Next set of experiments with the furnace allowed for investigation of liquid combustion products (“acids” and “bio-oils”). The results of these experiments are presented in
“Variation factor” of wood burning products from furnace experiments with estimated 95% confidence intervals*.
Variation factor ΔC (‰) | |||
---|---|---|---|
Temperature experiment | |||
Beech: | „Charcoal” | Oil | Acid |
0.16 ± 0.26 | 0.97 ± 0.29 | ||
0.12 ± 0.25 | 0.77 ± 0.23 | –2.38 ± 0.35 | |
–0.54 ± 0.26 | 1.13 ± 0.33 | 3.12 ± 0.54 | |
–0.58 ± 0.26 | 0.08 ± 0.28 | –0.40 ± 0.27 | |
–0.91 ± 0.26 | –1.30 ± 0.28 | ||
0.21 ± 0.18 | |||
–0.07 ± 0.24 | 2.44 ± 0.28 | ||
2.30 ± 0.22 | |||
–0.76 ± 0.19 | 1.81 ± 0.26 | ||
–0.92 ± 0.27 | 1.29 ± 0.27 | ||
–0.86 ± 0.21 | 1.27 ± 0.20 | ||
0.43 ± 0.36 | –0.32 ± 0.39 | ||
0.61 ± 0.36 | 1.76 ± 0.37 | –0.16 ± 0.42 | |
0.19 ± 0.36 | 1.31 ± 0.31 | ||
0.01 ± 0.35 | 1.48 ± 0.32 | ||
0.06 ± 0.36 | 0.64 ± 0.37 | ||
–0.16 ± 0.28 | –2.48 ± 0.22 | 0.40 ± 0.28 | |
–0.43 ± 0.28 | 2.73 ± 0.29 | –0.81 ± 1.73 | |
–0.93 ± 0.28 | 0.94 ± 0.56 | –0.23 ± 0.28 | |
–2.64 ± 0.29 | –2.21 ± 0.28 | ||
–3.88 ± 0.29 | |||
Time experiment | |||
---|---|---|---|
„Charcoal” | Oil | Acid | |
15 minutes | 0.83 ± 0.36 | 0.07 ± 0.36 | |
30 minutes | 1.20 ± 0.36 | ||
45 minutes | 1.16 ± 0.36 | 0.62 ± 0.36 | |
60 minutes | 1.27 ± 0.36 | 0.50 ± 0.36 | |
90 minutes | 0.75 ± 0.36 | 1.45 ± 0.36 | |
150 minutes | 0.97 ± 0.36 | 2.46 ± 0.36 | |
210 minutes | 3.67 ± 0.36 | ||
250 minutes | 1.51 ± 0.37 |
At temperatures >300°C, the “charcoal” samples were depleted in 13C, except oak wood samples. The δ13C depletion was continuous as the temperature increased: it was relatively slow for lower temperatures (even with trends for enrichment at 200°C visibly for oak), and much faster above 400°C.
After finishing the experiment with fossil oak, the “charcoal” was still reacted at 600°C, and additional two samples were collected (
“Bio-oils” and “acids” exhibited differences in isotopic compositions at low and high combustion temperatures (
To test and to better understand the relationship between isotopic composition and burning time, another experiment was conducted at constant temperature. The oak wood sample was placed in the furnace preheated to 300°C. “Bio-oil” and “acid” samples were collected for carbon isotopic composition analyses. After 150 min, the release of “bio-oils” ended. The release of “acids” ended after 210 min. After 250 min, only small amounts of grey residuum were still left in the quartz tube. The results of carbon isotopic composition of collected burning products are presented in
In the first set of simulations’ results, there is visible a significant difference between temperatures of the onset of pyrolysis in comparison to corresponding temperatures in the experiments. This is related to much faster ramp heating rate of the molecular simulations (~2 ns to reach maximum temperature
The first set of molecular simulations showed that reactions of thermal decomposition of cellulose and lignin are similar without (
In the
Differences in energies and entropy of reactants and products for reactions 3.2–3.5 (in kJ/mol and J/mol·K, respectively) for temperatures of 27°C (300 K) and 427°C (700 K) recalculated for one carbon atom, along with equilibrium constants.
reaction | temperature (°C) | ΔEZPE* | ΔEthermal** | ΔGthermal*** | ΔS**** | K***** |
---|---|---|---|---|---|---|
27 | 0.02998 | 0.03427 | 0.06128 | –0.08033 | 24.87 | |
(3.2 – CO2) | 427 | 0.02998 | 0.08385 | 0.07265 | 0.01423 | 12.56 |
27 | –0.13872 | –0.08699 | –0.02306 | –0.21004 | –9.20 | |
(3.3 – formaldehyde) | 427 | –0.13872 | –0.00577 | 0.01466 | –0.02761 | 2.52 |
27 | –0.10346 | –0.05799 | –0.00923 | –0.15648 | –3.69 | |
(3.4 – glycoaldehyde) | 427 | –0.10346 | 0.00692 | 0.02142 | –0.01799 | 3.69 |
27 | –0.08424 | –0.06414 | –0.03097 | –0.39204 | –12.34 | |
(3.5 – acetone) | 427 | –0.08424 | –0.02422 | –0.00730 | –0.52635 | –1.25 |
where K – equilibrium constant for reaction, in general α = K1/n where n is a number of exchanged atoms (if n = 1 then K = α); R– ideal gas constant; T– temperature.
Positive/negative values of energies indicates that the reactants are more/less stable than the products, respectively. Isotopically heavier gaseous/liquid products of B/P were always considered as reactants (reactions 3.2– 3.5). This means that, if assuming equilibrium in exchange between pyrene residuum and CO2, this gaseous molecule should always be enriched in heavier carbon isotope. Relationships with temperature is different for other organic molecules: formaldehyde (C1 phase) and glycolaldehyde (C2 phase) are depleted in 13C at lower temperatures, while enriched at higher temperatures >427ºC (700 K). Acetone (C3 phase) is expected to be depleted in 13C even at higher pyrolysis temperatures.
There are three main sources of isotopic effect for the studied reactions 3.2–3.5:
Zero-point energy (ZPE) correction, which arises from the fact that the lowest possible energy of the molecule at temperature of 0K is not the electronic energy but is higher and is related to molecular vibrations. Influence of this correction is different for CO2 (positive) than for other product molecules (negative) in comparison to pyrene (
Thermal corrections, which are result of occupation of certain vibrational and rotational energy levels by molecules at considered temperatures (supplementary materials contains table with listed all thermal corrections). In the case of all the studied reactions 3.2–3.5 this effect leads to enrichment of gaseous/liquid products in 13C, comparing to energy levels that takes into account only ZPE (
Differences in entropy. Generally, higher entropy is observed for isotopically heavier molecules (supplementary materials contains table with translational, rotational and vibrational contributions to entropy). Based on the received results it is visible that in the case of all reactions 3.2–3.5 there is a decrease of entropy (
p – pressure; ΔV – difference in volume between substrates and products. Because there is no difference in volume of the molecules substituted with different carbon isotopes: ΔV = 0 and therefore in all calculations ΔH = ΔE.
Summarizing all the three contributions to isotopic effect (
The isotopic composition of the B/P products depended first on the isotopic composition of the initial material. The wood samples studied are typical terrestrial plants that use a C3 photosynthesis pathway. The typical δ13C value for C3 plants is in the range from –25‰ to –32‰ (mean 28‰; Wickman, 1952; Baertschi, 1953 and Craig, 1953, 1954) and the carbon isotopic composition of the wood samples used in this study agrees with these data.
Wood is a complex material and consists of several phases with different carbon isotopic composition. The major wood components, cellulose and lignin, differ by 4–7‰. Cellulose is enriched by 1–2‰, and lignin is depleted by 2–6‰ relatively to the whole wood composition (Benner et al., 1987). These main wood components also have different B/P characteristics (Rowell and LeVan-Green, 2005). Whole wood thermal degradation starts below 250°C. At 300–375°C, most of the carbohydrate polymers, except lignin, are degraded. Hemicellu-lose decomposition starts at
The experiments performed in this study show that the B/P initiation temperature is dependent on the type of wood. For the beech and oak, B/P decomposition starts at 200–250°, while for pine at lower temperatures (< 200°C). This agrees with results of Shafizadeh (1984) and Rowell and LeVan-Green (2005).
General trends of carbon isotopic changes in B/P products obtained in experiments agree with results published earlier by different authors. In general, it is agreed that the isotopic composition of charcoal differs from wood by up to 2‰ (Ferrio et al., 2006; Turney et al., 2006; Ascough et al., 2008; Das et al., 2010 and Bird and Ascough, 2012). It is also agreed that it is temperature dependent. Below 300°C, there is an increase in δ13C in charcoal, which is interpreted as a loss of isotopically lighter phases and/or more resistance of isotopically heavier cellulose (Jones et al., 1993; Czimczik et al., 2002; Poole et al., 2002; Hakkou et al., 2006 and Bird and Ascought, 2012). Above 300°C, charcoal is depleted by 1–2‰ relatively to wood (Jones and Chaloner, 1991; Jones et al., 1993; Bird and Grocke, 1997; Czimczik et al., 2002; Turney et al., 2006; Ascought et al., 2008 and Bird and Ascought, 2012). This is explained by the preferential decomposition of isotopically heavier cellulose. Above 400°C, further 13C depletion is interpreted as caused by preferential loss of 13C during C=C bond formation in polyaromatization (Qian et al., 1992 and Krull et al., 2003). Effect of temperature relation between charcoal (residuum) δ13C and temperature were visible in the results of presented studies (experiments with furnace). Initially, at temperatures < 300°C, we observed an increase in δ13C, with maximum values at 300°C (200°C for pine) but isotopic variation factor (ΔC) values were significant only for oak and pine (ΔC200 were 0.43 ± 0.36‰ and 0.21 ± 0.18‰ and t-values for Student tests were 2.479 and 2.453, respectively; critical value is 2.032). At higher temperatures, the δ13C values decreased until 500°C, and then stabilized for pine and modern oak. For all wood types, except oak, δ13C depletion was significant at the temperature 400°C and above (
In the furnace experiments, the samples of “acids”/”bio-oils” were collected from the material condensed on the cool surface of the quartz sample tube. Like for “charcoal”, the carbon isotopic composition of these products is temperature dependent (
Performed molecular simulations offers another alternative explanation of isotopic variation of B/P products. Results of simulations of isotope exchange between gaseous/liquid products and model of residuum show significant differences of equilibrium constant for exchange reaction (
Significant differences between δ13C of charcoal produced by pyrolysis and incomplete combustion reported in published studies were explained by the acceleration of cellulose decomposition by pyrolysis in comparison to decomposition in the presence of O2 (Turney et al., 2006; Ascough et al., 2008; Das et al., 2010 and Bird and Ascough, 2012). The experiments with pine wood (
Published B/P experimental results for δ13C values of charcoal suggested no time dependence of the carbon isotopic composition (Turney et al., 2006; Ascough et al., 2008; Das et al., 2010 and Bird and Ascough, 2012). The B/P experiment with oak at 300°C showed significant changes in isotopic composition of incomplete burning products with time (
In the simulation of natural fireplace experiment the carbon isotopic composition of CO2 indicated enrichment in 13C isotope relatively to “ash” (
In all the experiments, difference between fossil and modern oak wood is visible in the mass loss, carbon isotopic composition of the B/P products and its temperature dependence. For fossil oak, the δ13C value of the condensed volatiles was depleted (ΔC around –0.4‰; see
This study shows differences in the carbon isotopic composition of wood burning/pyrolysis products. The extreme values of the variation factor were around –4‰ for charcoal from fossil oak at 600°C and around 3‰ for “acids” from beech wood at 400°C. The typical range of variation was between –2‰ and 2‰.
The carbon isotopic composition differentiation in B/P products is temperature and time dependent. Variation factor values are wood type dependent, but general trends are similar for all types of modern wood examined (
The observed temperature dependence for the carbon isotopic composition is a result of two effects. The first one is the different thermal stability of the main wood components, which are also characterized by different carbon isotopic composition. Thus, the less thermally stable and isotopically lighter lignin decomposition controls the isotopic composition of the B/P products at the lower temperatures, while heavier cellulose is the dominant component at temperatures above 300°C. Superimposed on this effect there can also be differences in isotopic composition of gaseous/liquid products that have significantly different equilibrium constants for reaction of isotopic exchange with residuum.
The experiment with B/P decomposition of oak wood at 300°C shows the time dependence of the carbon isotopic composition for all B/P products. It confirms that B/P should be treated as a continuous process, with the results depending on the degree of process development.
This is especially important given the number of studies using the variation in the isotopic composition of combustion products as a record of paleoenvironmental change. Natural burning processes are dynamic and temperature, O2 availability and other conditions change rapidly. Collected burning products from geological or archaeological sites contain no information about temperature or other conditions of the burning process. It is especially important when studies are performed using samples of burning products separated from deposits as a dispersed organic material. The application of chemical pre-treatment methods necessary for burning products (e.g. soot aggregates) separation may introduce additional isotopic fractionation (see
The differences in the carbon isotopic composition of the combustion products and its dependence on temperature or reaction time make it necessary to take great care in studies that use the isotopic composition of combustion products as a paleoenvironmental proxy or as an isotopic metric for the identification of source material.
The differences in the isotopic composition of carbon in soot aggregates have become a motivation for expanding experimental research. The studies carried out do not provide definitive and sufficient answers to the questions posed, but they indicate the importance of the complex study of combustion products instead of focusing only on selected products. The most important seems to be the extension of the research to chemical analysis of B/P products, especially volatile components condensing on collecting plates and gaseous products. Molecular simulations seem to be a good tool that can be particularly useful in explaining the differentiation of the carbon isotopic composition during B/P processes. Finally, it is advisable to collect a larger number of research results so that it is possible to analyze the significance of the differences found, as well as the variability of the obtained results for a larger number of wood samples.