Slow pyrolysis of walnut shells in nitrogen and carbon dioxide

Fuel 225 (2018) 419–425

Contents lists available at ScienceDirect

Fuel journal homepage: www.elsevier.com/locate/fuel

Full Length Article

Slow pyrolysis of walnut shells in nitrogen and carbon dioxide a,⁎

b

c

O. Senneca , F. Cerciello , S. Heuer , P. Ammendola a b c

T

a

Istituto di Ricerche sulla Combustione, Consiglio Nazionale delle Ricerche, P.le Tecchio 80, 80125 Napoli, Italy Dipartimento di Ingegneria Chimica, dei Materiali e della Produzione Industriale, Università degli Studi di Napoli Federico II, P.le Tecchio 80, 80125 Napoli, Italy Department of Energy Plant Technology, Ruhr-University Bochum, 44780 Bochum, Germany

A R T I C LE I N FO

A B S T R A C T

Keywords: Biomass Slow pyrolysis Char, oxy-combustion Pyrolysis products

Previous studies have shown that increased carbon dioxide concentration upon heat up affects the products of coal pyrolysis and in particular that chars prepared under carbon dioxide rich atmospheres are less reactive than chars prepared in nitrogen, and consistently tars are more aromatic. In the present work, this issue is investigated with reference to a biomass, namely walnut shells (WS), where the lignin component prevails over cellulose and hemicellulose. Preliminary experiments of thermal degradation have been carried out using a thermogravimetric (TG) apparatus, under constant heating rate conditions, in flows of either nitrogen or carbon dioxide. Derivative thermogravimetric (DTG) curves reveal the existence of multiple peaks, which are typically associated with the degradation of different ligno-cellulosic components. A multiple parallel reaction scheme has therefore been used to fit the experimental data and kinetic parameters have been obtained. Walnut shells were also pyrolyzed in a fixed bed reactor at 600 °C in either nitrogen or carbon dioxide so as to collect pyrolysis products in amounts sufficient for further analysis. Char and tar samples have been characterized using different techniques (e.g. GC–MS, elemental analysis, TGA, SEM) revealing limited differences. Combustion rates of the chars have been measured by means of non-isothermal thermogravimetric experiments in air and again small differences have been observed between the samples prepared under carbon dioxide and nitrogen. It has been concluded that under the low heating rate conditions typical of the thermogravimetric apparatus and fixed bed reactor used in the work, the effects of carbon dioxide on liquid and solid products of biomass pyrolysis exist but are less important than for coal. The work is complementary to another paper, which addresses the effect of carbon dioxide on biomass pyrolysis under high temperature and fast heating rate conditions in a drop tube reactor.

1. Introduction Extensive literature exists on pyrolysis of biomass and its main components, such as cellulose and lignin, as well as on combustion and gasification of biomass chars have also been largely investigated [6–20]. Kinetics of char combustion and gasification have been most often measured by thermogravimetric analysis on chars prepared in nitrogen under reference conditions. However, extensive literature has demonstrated that the conditions under which pyrolysis is carried out dramatically influence the properties and in particular the reactivity of the chars towards combustion and gasification. Heat treatment prolonged beyond the completion of pyrolysis has been in fact indicated as responsible of thermal annealing and progressive loss of char reactivity [21–24].

⁎

If during heat up and pyrolysis gaseous reactants do not reach the particles surface, due to severe boundary layer diffusional resistances or to the vigorous efflux of volatiles, thermally activated processes are not affected by the nature of the gaseous environment where they take place, however Senneca et al. [25] pointed out that if oxygen is able to reach the particles surface, the overall thermochemical conversion pattern as well as the reaction kinetics can be heavily affected: oxygen in fact can enhance the abstraction of volatile matter, moreover heterogeneous combustion can take place in parallel with volatile abstraction leading even to heterogenous ignition or complete carbon burnout. This phenomenon was called “oxidative pyrolysis” and documented for a suite of different solid fuels, including biomass, plastics and coals in Refs. [26–28]. For the case of medium rank coals Senneca et al. [25] elaborated maps of heating rate, particle size and temperature when oxidative pyrolysis occurs. Similar maps have not been

Corresponding author. E-mail address: [email protected] (O. Senneca).

https://doi.org/10.1016/j.fuel.2018.03.094 Received 20 December 2017; Received in revised form 4 March 2018; Accepted 13 March 2018 0016-2361/ © 2018 Elsevier Ltd. All rights reserved.

Fuel 225 (2018) 419–425

O. Senneca et al.

bubblers is analyzed off-line by a gas chromatograph (GC) coupled with a mass spectrometer (MS). The incondensable gas stream which exits the bubblers passes through a filter to capture residual tar or water and then is sent to gas analyzers (ABB AO2020), including Caldos 27, Magnos 206 and Uras 26 modules, for the measurement of O2, CO, CO2, CH4 and H2. In order to rule out the possibility that the profiles of CO, CO2, CH4 and H2 concentration recorded by the analyzers were affected by the residence time, impulsive experiments were carried out, injecting a pulse of CO2 inside the reactor bed and measuring the temporal profiles of CO2 at the reactor outlet. This allowed to obtain the residence time distribution function of gases E(t) of the reactor. Based on this analysis the gas profiles were corrected by assuming a time shift of 2 min while Fourier deconvolution of the profiles was not considered necessary, on account of the much longer timescale of the pyrolysis reaction (in the order of hours) compared to the residence time of the gas in the reactor (in the order of 100 s). The yield in tar and char was assessed by weighing the collected samples. The overall yield of CH4/CO/H2/CO2 was also measured by integration of the gas concentration curves.

proposed yet for biomass materials. The advent of oxy-combustion processes of coal and biomass, where fuel particles experience large concentrations of carbon dioxide since the very early stages of heat up and pyrolysis, motivate the need of experimental work also on the effects of CO2 on the course of pyrolysis. Recent studies have shown that increased carbon dioxide concentration upon heat up affects the products of coal pyrolysis and in particular that coal chars prepared under carbon dioxide rich atmospheres at different heating conditions are in general less reactive than the corresponding chars prepared in nitrogen, and consistently tars are more aromatic [1–4]. The investigation is currently being extend to biomass. In the present work walnut shells have been exposed to slow heating rate programs in either nitrogen or carbon dioxide using both a thermogravimetric (TG) apparatus and a fixed bed micro-reactor. The first set of experiments allowed to investigate the kinetics of the reactions, the second set of experiments allowed to collect and analyze reaction products (char, tar and gas). The effect of carbon dioxide on walnut shells pyrolysis under high temperature and fast heating rate conditions typical of pulverized fired boilers are investigated in a parallel work [5] using a drop tube reactor (1300 °C).

2.3. Product analysis procedure Char samples produced in the fixed bed reactor in N2 and CO2 were analyzed by several techniques. The elemental composition was determined by a LECO CHN 628. Scanning electron microscopy (SEM) was used in order to get a first impression of the changes in char surface properties. SEM images were taken with a FEI INSPECT S. A Netzsch 409 TG-DSC apparatus has been used to perform combustion experiments of the chars produced in the fixed bed reactor. In combustion experiments, an upward flow of synthetic air of 250 ml/ min (STP) has been used. The heating rate has been set at 5–20 °C/min and the final temperature at 900 °C. Approximately 5 mg of sample have been loaded in the pan in each test to control the depth of the sample layer in the pan and minimize oxygen transfer resistances. The conversion degree f of a char particle is herein defined as:

2. Experimental methodology 2.1. Biomass fuel Walnut shells (WS) used for the current experiments have been sieved to the size 90–106 µm. Proximate and ultimate analysis of the sieved samples are given in Table 1. Notably proximate analysis has been carried out in accordance to the standards EN ISO 18134-3 (moisture at 105 °C), 18122 (ash at 550 °C), 18123 (volatiles at 900 °C), 16948 (ultimate analysis) and 18125 (higher heating value, HHV). 2.2. Procedure of pyrolysis experiments A first screening of the pyrolysis behavior of the materials has been carried out by TGA with a Netzsch 409 TG-DSC apparatus. Approximately 20 mg of sample have been loaded in the pan in each test. An upward flow of gas of 250 ml/min (standard temperature and pressure, STP) has been used. The temperature has been raised from 25 °C to 110 °C and the sample has been held at 110 °C for 5–10 min to release moisture. The sample was then heated up to 900 °C at constant heating rate of 5–20 °C/min and held at this temperature for 30 min. The mass recorded during experiments has been worked out in order to obtain DTG plots of (dm/dt m0-1) vs. T, where m, m0 are the actual and the initial weight of sample (after the dehumidification stage) and T the temperature. Additional pyrolysis experiments were carried out in a fixed bed micro reactor depicted in Fig. 1. The reactor consists of a tubular quartz reactor heated externally (inner diameter 20 mm) by an electric furnace. Approximately 1 g of the sample is placed inside the reactor from the very beginning of the experiment and heated accordingly. A thermocouple is inserted into the bed. N2 or CO2 are fed from the reactor top at flow rate of 200 ml/min (STP) and leave the reactor from the bottom. The reaction products at the reactor outlet are quickly cooled down to 200 °C and then to 0–5 °C as they flow through four bubblers in series. The tar captured by the

f = (m 0−m (t ))/(m 0−m∞)

(1)

where m0 is the sample mass at time t = 0, m(t) the mass at time t and m∞ the residual mass at the end of the TG experiment. The weight loss data measured during combustion experiments have been worked out to obtain instantaneous rate of reaction (df/dt) and Arrhenius plots (ln df/dt/(1 − f) versus 1/T). The tar samples collected downstream the fixed bed reactor were dissolved out in two steps (1-propanol and acetone), following the procedure reported in Refs. [29–32], and analyzed by gas chromatography and online mass spectrometry as detector (GC–MS, AGILENT GC 7890 apparatus coupled with MSD 5975C). In the GC an HP-35 (length 30 m, di 250 μm, film 0.25 μm) column is mounted. Sample injection was done in splitless mode at 300 °C with a gas flow of 1 ml/min (STP). The temperature program consists of four isothermal steps: 50 °C (5 min), constant heating for 30 min to 200 °C (5 min), constant heating for 1.75 min to 270 °C (5 min) and finally, constant heating for 6 min to 300 °C (15 min). The transfer line between the GC and the MS is held at 300 °C. MS spectra are acquired in scan mode in the range 40–500 uma. 3. Results and discussion 3.1. TG pyrolysis

Table 1 Standardized analyses of the examined fuel. Moisture raw (wt%)

Ash dry (wt%)

Volatiles dry (wt%)

C daf (wt%)

H daf (wt%)

N daf (wt%)

S daf (wt%)

O daf (wt%)

HHV daf (MJ/kg)

4

0.42

81.07

52.15

5.77

0.28

0.02

41.78

20.51

Results of TG experiments on biomass are reported in Fig. 2. Pyrolysis in N2 occurs in two stages with peaks at 280 °C and 340 °C followed by a slow mass loss stage above 400 °C. In CO2 the same pattern is observed up to 600 °C. Char gasification starts at around 650 °C with two peaks at 725 and 850 °C. The multiple stages of both pyrolysis and gasification can be attributed to the presence of multiple ligno420

Fuel 225 (2018) 419–425

O. Senneca et al.

Fig. 1. Fixed bed pyrolysis micro reactor.

pyrolysis as functions of time. In the pyrolysis experiment carried out in N2, the gaseous products mainly consisted of carbon dioxide. CO2 release occurred in two stages starting from 200 °C. It was almost complete below 370 °C and followed by a tail up to 600 °C. For the experiment carried in CO2, it was impossible to distinguish between the CO2 produced by pyrolysis from the baseline. In both pyrolysis experiments, CO release started at about 250 °C. A shallow peak of CO was observed around 300 °C (same peak temperature as carbon dioxide) followed by a pronounced maximum at 360 °C and a tail up to 600 °C. The first peak becomes even shallower in the CO2 experiment. Hydrogen and CH4 release in both N2 and CO2 started at 200 °C and continued up to 600 °C. In CO2, however, two H2 peaks can be well distinguished (at 360 and 600 °C) and three CH4 peaks (two marked peaks at 280 and 420 °C, and a shallow peak at 530 °C), while in N2 the early peaks are not well defined. The ratio of cumulatively evolved H2/CO and CH4/CO were similar in the N2 and CO2 experiment, as can be observed from Table 3. Fig. 2. Walnut shells thermogravimetric pyrolysis in N2 and CO2.

3.4. Tar analysis cellulosic components in the raw biomass.

The tars collected in fixed bed reactor tests were analyzed by GC–MS. Notably, the temperature program of the GC–MS was limited to 300 °C, therefore heavier tar species were not properly identified. The chromatographic analysis of condensed tars highlighted the presence of different compounds. The identified peaks corresponded to benzenes, phenols, furans, oxygenated aliphatic and azo-cyclic compound, characteristic of biomass decomposition. Table 4 reports the yields of different chemical species in the chromatograms of the tars produced in CO2 and in N2. It can be observed that in the N2 tar, the aromatic component is more abundant than in the CO2 tar, conversely the aliphatic component is lower. It is remarked that a detailed quantitative analysis of the individual tar samples has not been reported here, because its accuracy would very much rely on instrument calibration with all the molecules of interest (which are more than 20). However, the ratio of the peak areas provides a reliable indication on whether substitution of N2 with CO2 upon pyrolysis augments or reduces the yield in the given species.

3.2. Pyrolysis in fixed bed reactor The products of pyrolysis in the fixed bed reactor were mainly constituted of gaseous species, condensable hydrocarbons and char. Product yields are reported in Table 2 on biomass dry basis. The char yield is higher than the corresponding values of proximate analysis (Table 1), being the maximum test temperature set at 600 °C, against the 900 °C used for proximate analysis. Notably the mass balance could not be closed to 100% because not all the gaseous species have been analyzed. 3.3. Analysis of gaseous pyrolysis products Fig. 3 shows the profiles of gaseous species evolved during biomass Table 2 Solid, liquid and gas yields (on dry basis) of fixed bed pyrolysis using different atmospheres. Pyrolysis Atmosphere

Char yield (wt%)

Tar (wt%)

Gas yield (by difference) (wt%)

Yield of CO, CO2, CH4, H2 (wt%)

N2 CO2

24.3 24.2

60.3 62.7

15.4 13.1

12.1

3.5. Char analysis

Yield of CO, CH4, H2 (wt%)

Chars collected during the fixed bed experiments have been analyzed for their CHN content by a LECO CHN 628 elemental analyzer. Oxygen content was calculated by difference. The H/C and O/C ratios of the WS and chars produced in N2 (FixB N2) and CO2 (FixB CO2) have been used to locate samples in the van Krevelen diagram of Fig. 4. Cellulose, hemicellulose and lignin values,

5.6

421

Fuel 225 (2018) 419–425

O. Senneca et al.

Fig. 3. Gas evolutions of WS pyrolysis carried out in nitrogen (left) and in carbon dioxide (right).

the two chars. Combustion reactivity of the chars has been measured by means of non-isothermal TGA in air with heating rates between 5 and 20 °C/min. Fig. 6 shows the reaction rate curves obtained from combustion tests of the char samples in the TGA at 5 °C/min. The sample pyrolyzed in N2 exhibits a pronounced and sharp peak at temperature of 450 °C. The char collected in CO2 has a similar peak temperature as the char collected in N2 but exhibits a broader peak. Similar results have been obtained at different heating rates.

Table 3 Mass ratios of the selected representative light gas species. Pyrolysis Atmosphere

H2/CO

CH4/CO

CO2/CO

N2 CO2

0.09 0.08

0.3 0.4

1.9 –

Table 4 Yields of different chemical species of tar produced in CO2 and in N2. Pyrolysis Atmosphere

Benzenes & phenols (wt%)

Oxygenated aliphatic (wt%)

Furans (wt%)

Azo-cyclic (wt%)

3.6. Kinetic analysis

N2 CO2

74.9 66.6

20.9 24.3

1.1 4.2

3 4.8

The existence of two marked DTG peaks in the temperature range 150–500 °C during WS pyrolysis (Fig. 2), suggested to assume the existence of two components of mass V1 and V2, and adopt for pyrolysis a two-parallel reaction kinetic model:

EaV1 dVi ⎤ V nV 1 = −kV1·V1nV 1 = kV1oexp ⎡− 1 ⎢ dt ⎣ RT ⎥ ⎦

(2)

EaV2 dV2 ⎤ V nV 2 = −kV2·V2nV 2 = kV2o exp ⎡− 2 ⎢ dt ⎣ RT ⎥ ⎦

(3)

to be integrated with initial conditions:

for t = 0: Vi = Xi ·B0·(1−wchar ); i = 1,2

(4)

where Bo is the initial biomass mass, wchar the char yield in the TG pyrolysis experiments and Xi the mass fraction of the i-component in the biomass. kVio, EaVi and nVi are the kinetic parameters (pre-exponential factor, activation energy, reaction order) of pyrolysis of volatile components Vi, while R is the molar gas constant, T the sample temperature, Similarly, the existence of two peaks in the range 600–900 °C with CO2, in Fig. 2, suggested to adopt an analogous two parallel reactions scheme also for char gasification, assuming the existence of two char components for gasification with mass Cg1 and Cg2:

Fig. 4. Van Krevelen diagram for walnut shells and char produced in a fix bed in N2 and CO2 (FixB N2 and FixB CO2) compared with the cellulose, hemicellulose and lignin.

dCg1 dt

as reported in the literature, taken from [33], have also been plotted in the figure for comparison. Notably, WS has H/C and O/C ratios intermediate between those of lignin and cellulose/hemicellulose. Application of the “lever rule” suggests a 2:1 ratio of lignin over cellulose and hemicellulose in the WS sample. The H/C and O/C ratio decreases in the FixB N2 to the FixB CO2, being slightly higher in the N2 compared to the CO2 char. SEM pictures of chars are reported in Fig. 5 A-D. It can be observed that both WS chars have a rather rough surface and that there are no obvious differences in shape and pores between

dCg 2 dt

EaCg1 n ⎤ C nCg1 = −kCg1·Cg1Cg1 = −kCg1o exp ⎡− ⎢ RT ⎥ g1 ⎣ ⎦

(5)

EaCg2 n ⎤ nCg2 = −kCg2·Cg 2Cg2 = −kCg2oexp ⎡ ⎢− RT ⎥ Cg 2 ⎣ ⎦

(6)

to be integrated with initial conditions:

for t = 0: Cgi = Yio·Co (1−wash); i = 1,2

(7)

where, Co is the mass of char after completion of pyrolysis (Co = B0·wchar ) and wash the fractional weight of ash in the char, Yio is the mass fraction of the i-component of the char, kCgio , EaCgi , nCgi the 422

Fuel 225 (2018) 419–425

O. Senneca et al.

Fig. 5. SEM imagines A-B) WS char produced in N2; C-D) WS char produced in CO2.

preexponential factor, activation energy and reaction order of the gasification reaction of the ith char component. A first estimate of preexponential factor and activation energy values was obtained by linear regression of Arrhenius plots. More accurate values of the kinetic parameters were further on obtained by least square fitting over the experimental data of mass loss versus time. Results are reported in Tables 5 and 6. Notably, the pyrolysis model accounts for the loss of 90% of the volatile matter, it does not describe the slow pyrolysis tail that takes place above 500 °C and justifies an additional 10% of weight loss. The agreement between the kinetic model and the experimental curves can be appreciated in Fig. 7. Finally, a similar kinetic model has been applied to char combustion, again assuming the existence of two char components for combustion with mass Cc1 and Cc2.

Fig. 6. DTG curves in air of FixB char samples produced in N2 and in CO2.

EaCc1 dCc1 ⎤ C nCc1 = −kCc1·Ccn1Cc1 = −kCc10 exp ⎡− c1 ⎢ ⎥ RT dt ⎣ ⎦

(8)

EaCc2 dCc2 ⎤ C nCc2 = −kCc2·Ccn2Cc2 = −kCc20 exp ⎡− c2 ⎢ dt ⎦ ⎣ RT ⎥

(9)

to be integrated with initial conditions: Table 5 Kinetic parameters of WS pyrolysis.

WS Pyrolysis

kV1o 1/min

T

EV1

°C

kJ/mol

3·107

150–500

88

nV 1

1

423

X1o

0.3

EV2

kV2o 1/min

kJ/mol

6·109

125

nV 2

X2o

1

0.6

Fuel 225 (2018) 419–425

O. Senneca et al.

Table 6 Kinetic parameters of WS char gasification.

Char gasification (with CO2)

nCg1

kCg1o

T

ECg1

1/min

°C

kJ/mol

3·105

600–900

138

1

Y1o

0.3

kCg 2o

ECg2

1/min

kJ/mol

2·1021

472

nCg2

Y2o

1

0.7

parameters of the second component are, however, similar for the N2 and CO2 chars.

4. Conclusions Kinetic analysis of the pyrolysis of walnut shells in N2 and CO2 has been performed. Pyrolysis below 600 °C is not affected by the presence of CO2. Two components have been identified which can in first instance be reconducted to the presence of different ligno-cellulosic fractions (lignin, cellulose, hemicellulose) in the material. Such two components pattern is retained also for gasification of the residual char which takes place with CO2 above 700 °C. Walnut shells have been pyrolyzed in a fixed bed reactor at 600 °C in either nitrogen or carbon dioxide. The temperature was low enough to avoid char gasification. The gaseous products seemed to be scarcely affected by the N2 vs. CO2 atmosphere. Tar products resulted to be more aromatic in N2 than in CO2. This result is in contrast to what was obtained for a medium rank coal, as reported in previous papers [1–4]. In the case of coal, in fact, the presence of CO2 was shown to increase the aromaticity of tar and favor formation of soot-like material. Chars produced in N2 have moderately higher O/C and H/C content than chars prepared in CO2. As for the combustion kinetics of the chars in air, two components can still be identified for char produced in nitrogen, but the more reactive component, vanishes for the char produced in carbon dioxide. The kinetic parameters of the second component are similar for the N2 and CO2 chars. It is concluded that under the low heating rate conditions typical of the thermogravimetric apparatus and fixed bed reactor used in the present work, some effects of carbon dioxide can be observed on liquid and solid products of biomass pyrolysis, but they are less important than the effects reported for coal [1–4]. It is concluded, that the presence of CO2 accelerates thermal degradative process both at low and at high heating rates. It is still to clarify if this difference between N2 and CO2 is due to a different timetemperature history of the biomass particles or rather to chemical contribution from CO2. Moreover, it is yet to understand the role of lignin vs. cellulose or hemicellulose components of the parent biomass.

Fig. 7. DTG experimental curve of WS pyrolysis/gasification in CO2 and fitting curves obtained by kinetic parameters. Table 7 Kinetic parameters of char air combustion.

FixB CO2 FixB N2

ECc1

kCc1o 1/min

kJ/mol

2·107 2·107

109 109

nCc1

1 1

Z1o

1 0.65

for t = 0: Cci = Zio·Co (1−wash); i = 1,2

kCc2o

ECc 2

1/min

kJ/mol

—— 1·1017

—— 238

nCc2

Z2o

—— 0.8

—— 0.35

(10)

Where Zio is the mass fraction of the i-component in the char for combustion, kCcio , EaCci , nCci the pre-exponential factor, activation energy and reaction order of combustion of the i char component. The values of the kinetic parameters for char combustion have been reported in Table 7. The agreement between the kinetic model and the experimental curves can be appreciated in Fig. 8. It can be observed that two components are still identified in the char produced in nitrogen. The more reactive component, instead, vanishes in the char produced in carbon dioxide. The kinetic

Fig. 8. DTG experimental curve of FixB Char N2 (left) FixB Char CO2 (right) combustion fitting curves obtained by kinetic parameters. 424

Fuel 225 (2018) 419–425

O. Senneca et al.

Acknowledgments

Quercus Cerris bark. Energy Fuel 2007;21:1078–84. [16] Magnaterra M, Fusco JR, Ochoa J, Cuckierman AL. Kinetic study of the reaction of different hardwood sawdust chars with oxygen, chemical and structural characterization of the samples. In: Bridgwater AV, editor. Advances in thermochemical biomass conversion. Glasgow: Blackie A. & P; 1994. p. 116–30. [17] Adanez J, de Diego LF, Garcia-Labiano F, Abadanes JC. Determination of biomass char combustion reactivities for FBC applications by a combined method. Ind Eng Chem Res 2001;40:4317–23. [18] Gil MV, Riaza J, Álvarez L, Pevida C, Rubiera F. Biomass devolatilization at high temperature under N2 and CO2: Char morphology and reactivity. Energy 2015;91:655–62. [19] Tchapda AH, Krishnamoorthy V, Yeboah YD, Pisupati SV. Analysis of tars formed during co-pyrolysis of coal and biomass at high temperature in carbon dioxide atmosphere. J Anal Appl Pyrolysis 2017;128:379–96. [20] Zellagui S, Schönnenbeck C, Zouaoui-Mahzoul N, Leyssens G, Authier O, Thunin E, et al. Pyrolysis of coal and woody biomass under N2 and CO2 atmospheres using a drop tube furnace – experimental study and kinetic modeling. Fuel Process Technol 2016;148:99–109. [21] Li Q, Zhao C, Chen X, Wu W, Li Y. Comparison of pulverized coal combustion in air and in O2/CO2 mixtures by thermo-gravimetric analysis. J Anal Appl Pyrolysis 2009;85:521–8. [22] Senneca O, Salatino P. Loss of gasification reactivity towards O2 and CO2 upon heat treatment of carbons. P Combust Inst 2002;29:485–93. [23] Salatino P, Senneca O, Masi S. Assessment of thermodeactivation during gasification of a bituminous coal char. Energy Fuel 1999;13:1154–9. [24] Senneca O, Salatino P, Masi S. Microstructural changes and loss of gasification reactivity of chars upon heat treatment. Fuel 1998;77:1483–93. [25] Senneca O, Salatino P. Overlapping of heterogeneous and purely thermally activated solid-state processes in the combustion of a bituminous coal. Combust Flame 2006;144:578–91. [26] Senneca O, Chirone R, Salatino P, Nappi L. Patterns and kinetics of pyrolysis of tobacco under inert and oxidative conditions. J Anal Appl Pyrol 2007;79:227–33. [27] Senneca O, Chirone R, Salatino P. Oxidative pyrolysis of solid fuels. J Anal Appl Pyrol 2004;71:959–70. [28] Senneca O, Chirone R, Masi S, Salatino P. A, Thermogravimetric study of nonfossil solid fuels. 2. Oxidative pyrolysis and char combustion. Energ. Fuel 2002;16:661–8. [29] Ammendola P, Piriou B, Lisi L, Ruoppolo G, Chirone R, Russo G. Dual bed reactor for the study of catalytic biomasses tar conversion. Exp Therm Fluid Sci 2010;34:269–74. [30] Ammendola P, Chirone R, Lisi L, Piriou B, Russo G. Investigation of the catalytic activity of Rh-LaCoO3 catalyst in the conversion of tar from biomass devolatilization products. Appl Catal A-Gen 2010;385:123–9. [31] Ammendola P, Cammisa E, Chirone R, Lisi L, Ruoppolo G. Effect of sulphur on the performance of Rh-LaCoO3 based catalyst for tar conversion to syngas. Appl Catal BEnviron 2012;113–114:11–8. [32] Ammendola P, Cammisa E, Lisi L, Ruoppolo G. Thermo-chemical stability of alumina supported Rh-LaCoO3 catalysts for tar conversion. Ind Eng Chem Res 2012;51:7475–81. [33] Gaur S, Reed T. Thermal Data for Natural and Synthetic Fuels. New York: Marcel Dekker, Inc.; 1998.

This work has been financed by the German Research Foundation (DFG) within the framework of the SFB/Transregio 129 “Oxyflame”. We acknowledge Ruhr-University Bochum (RUB) Research School for awarding O. Senneca with an International Visiting Professor Grant (VIP_4_2016_02). Help in experimental work by Luciano Cortese and Fernando Stanzione is gratefully acknowledged. References [1] Pielsticker S, Heuer S, Senneca O, Cerciello F, Salatino P, Cortese L, et al. Comparison of pyrolysis test rigs for oxy-fuel conditions. Fuel 2017;156:461–72. [2] Senneca O, Apicella B, Heuer S, Schiemann M, Scherer V, Stanzione F, et al. Effects of CO2 on submicronic carbon particulate (soot) formed during coal pyrolysis in a drop tube reactor. Combust. Flame 2016;172:302–8. [3] Apicella B, Senneca O, Russo C, Heuer S, Cortese L, Cerciello F, et al. Separation and characterization of carbonaceous particulate (soot and char) produced from fast pyrolysis of coal in inert and CO2 atmospheres. Fuel 2017;201:118–23. [4] Heuer S, Senneca O, Wütscher A, Düdder H, Schiemann M, Muhler M, et al. Effects of oxy-fuel conditions on the products of pyrolysis in a drop tube reactor. Fuel Process Technol 2016;150:41–9. [5] Senneca O, Cerciello F, Cortese L, Heuer S, Schiemann M, Scherer V. Effects of oxyfuel conditions on walnut shells pyrolysis in a drop tube reactor. Tenth Mediterr. Combust. Symp., Naples, Italy:2017. Submitted to the MCS10 Special Issue of Fuel. [6] Evans RJ, Milne TA. Chemistry of tar formation and maturation in the thermochemical conversion of biomass. In: Bridgwater AV, Boocock DGB, editors. Developments in thermochemical biomass conversion. London: Blackie Academic & Professional; 1997. p. 803–16. [7] Morf P, Hasler P, Nussbaumer T. Mechanisms and kinetics of homogeneous secondary reactions of tar from continuous pyrolysis of wood chips. Fuel 2001;81:843–53. [8] Senneca O. Kinetics of pyrolysis, combustion and gasification of three biomass fuels. Fuel Process Technol 2007;88:87–97. [9] Antal MJ. Effects of reactor severity on the gas-phase pyrolysis of cellulose- and kraft lignin-derived volatile matter. Ind Eng Chem Prod RD 1983;22:366–75. [10] Antal MJ. A review of the vapor phase pyrolysis of biomass derived volatile matter. In: Overend RP, Milne TA, Mudge LK, editors. Fundamentals of biomass thermochemical conversion. London: Elsevier; 1985. p. 511–37. [11] Boroson ML, Howard JB, Longwell JP, Peters AW. Products yields and kinetics from the vapor phase cracking of wood pyrolysis tars. AIChE J 1989;35:120–8. [12] Di Blasi C, Buonanno F, Branca C. Reactivities of some biomass chars in air. Carbon 1999;37:1227–38. [13] Luo M, Stanmore B. The combustion characteristics of char from pulverized bagasse. Fuel 1992;71:1074–6. [14] Branca C, Di Blasi C. Devolatilization and combustion kinetics of wood chars. Energy Fuel 2003;17:1609–15. [15] Branca C, Iannace A, Di Blasi C. Devolatilization and combustion kinetics of

425