Electron spin resonance (ESR) spectroscopy is used to study materials with unpaired electrons, such as organic radicals and metal complexes. This method can also be used to follow radical reactions during pyrolysis of carbonaceous materials. However, the temperature dependence of ESR measurement should be considered. To enable reasonable comparisons, results measured at different temperatures must be converted. In this study, we investigated the behavior of free radicals in the process of coal pyrolysis using in situ and ex situ ESR. The ESR data were collected at both pyrolysis and room temperatures, and apparent differences were analyzed. The differences were diminished when our data were converted to the same measurement temperature level based on the Boltzmann distribution law. Furthermore, we investigated the effects of process conditions on the behavior of free radicals in the solid phase of coal. We found that temperature is the most important factor determining the formation and behavior of free radicals in the solid phase, followed by the residence time. Relatively active radicals were quenched by hydrogen-donor solvents to some degree, while stable radicals remained.

The fundamental mechanisms and behaviors in the thermal decomposition of coal have attracted considerable attention in the field of coal conversion processes, such as coal coking, coal gasification, and direct coal liquefaction. Coal pyrolysis is a primary reaction in the above-mentioned processes,1–4 and the final products are determined by the formation and the subsequent reactions of free radicals. Free radicals are generated via bond cleavage in coal at elevated temperatures. However, the measurement of free radicals is difficult owing to their short lifetime.

Coal is complex in terms of its structure, composition, and the products of its pyrolysis reactions. Many research methods have been developed to investigate the mechanism of coal pyrolysis. The overall mass-change is one characteristic, which is used to determine the solid and volatile changes in the solid and the distribution of volatiles, particularly in tars.5–8 Another common used approach is to correlate process variables during coal pyrolysis with the formation of designated products.9–14 Although these studies have contributed to advances in our understanding of the behavior of coal pyrolysis, there have been few reports that have revealed the mechanism of coal pyrolysis from the perspective of free radicals in the reactions. A comprehensive knowledge of free radicals production during coal pyrolysis is important, because the generation of free radicals and their reactions determine the types of products formed. Hence, an effective technique for measuring free radicals is desirable.

Electron spin resonance (ESR) is considered to be a promising technique for measuring free radicals in coal conversion processes, since it was first successfully used for investigations of coal in 1954.15 Many ESR investigations on coal, pitch, kerogen, and their derivatives have been performed using ex situ (conventional) or in situ conditions.16–21 ESR parameters such as spectral intensity, spectral line width, and g value provide information on the spin concentration, molecular environment, and spin type. The signal parameters of naturally formed free radicals have been used to determine the coal rank. Investigations on thermally formed free radicals in coal have also been reported. In ex situ ESR measurements, samples are heat-treated at a designated temperature outside the conventional ESR cavity prior to measurement at room temperature.22–26 Clearly, information on free radicals obtained by ex situ ESR measurement is limited to long-lived (steady) radicals, because active transient radicals with poor stability disappear when the samples are cooled to room temperature. For the case of in situ ESR measurements, samples are directly heated in the ESR cavity at a designated temperature by heated N2, while the measurements are performed. This strategy is used to study free radicals formed during coal pyrolysis, three decades following the invention of ESR.27–30In situ ESR directly measures free radicals formed during coal pyrolysis and offers more detail on the nature of free radicals produced during real coal pyrolysis.

The measurement temperature is the only difference between these two types of ESR measurement strategies. However, there has been very little research on reconciling the differences between the measurement results of these two strategies. Furthermore, to the best of our knowledge there have been no reports of direct comparisons between in situ and ex situ ESR measurement results obtained during coal pyrolysis.

The objective of this work was to investigate the effects of measurement temperature on the ESR signal intensity and to evaluate the comparability of in situ and ex situ ESR measurements. Coal contains a tremendous amount of naturally formed free radicals, and the radical concentration is sensitive to pyrolysis temperature. Thus, we attempted to correlate the radical concentration with other properties of the free radicals and process parameters, including temperature, residence times, solvent, and pressure.

Mataihao bituminous coal (MTH) was selected for this study. The coal sample was ground to <200 mesh and dried in a vacuum oven at 110 °C for 8 h. Ultimate analyses results of the coal sample are listed in Table I.

For ex situ ESR measurements, the coal pyrolysis experiments were performed in a sealed glass tube reactor (2 mm outer diameter and 1.8 mm inner diameter) under N2 atmosphere. The glass tube reactor was inserted into the bottom of a sealed stainless-steel tube. The stainless-steel tube was inserted into a preheated sand bath and reached the desired temperature in approximately 30 s. The holding time at the desired temperature was 120 s. At the end of the holding time, the stainless-steel tube was removed and rapidly quenched in a water bath. The glass tube reactor was immediately removed from the stainless-steel tube and stored in a liquid nitrogen Dewar to preserve the sample properties for subsequent ESR measurements.

For in situ ESR measurements, the coal pyrolysis experiments were also performed in a sealed glass tube reactor (2 mm outer diameter and 1.8 mm inner diameter) under a N2 atmosphere. However, the glass tube reactor was inserted into the ESR cavity and heated to the desired temperature within 30 s under a heated N2 flow. The holding time at the desired temperature was also 120 s. The ESR measurements were then performed immediately at this temperature. All measurements were repeated at least three times.

Temperature programmed pyrolysis of coal was performed on AutoChem II 2920 equipment (Micromeritics, USA). A 30-mg sample was loaded into a U-type quartz tube reactor and ramped from room temperature to 800 °C in high-purity Ar (99.999 %). The heating rate was maintained at 10 °C/min and the flow rate at 50 ml/min. The production of CO2 (m/e = 44), O (m/e = 16), CO (m/e = 28), CH3 (m/e = 15), C2H3 (m/e = 25), C3H3 (m/e = 39), C4H7 (m/e = 55) and C5H9 (m/e = 69) was monitored by an on-line quadruple mass spectrometer (Omnistar, Pfeiffer).

The ESR measurements were performed with an EMXplus-10/12 ESR spectrometer from Bruker equipped with a variable temperature cavity with a range from room temperature to 1200 °C. The ESR spectrometer was operated at 9.85 GHz and 0.1 mW. The central magnetic field was 320 mT, the sweep width was 600 mT, the sweep time was 10 s, and the time constant was 0.01 s. For the ex situ ESR measurements, the samples were measured at 20 °C, for in situ ESR measurements, the samples were measured at the pyrolysis temperature and the results were scaled to 20 °C-basis for direct comparison with the results of ex situ measurements.

Figure 1 compares the radical characteristics measured from the ex situ and in situ experiments. We attempted to ensure that the only difference between the ex situ and in situ experiments was the measurement temperature.

During the ex situ experiments, the coal sample was heated in a sand bath at the designated pyrolysis temperature for 120 s, cooled in a liquid nitrogen Dewar and measured at room temperature (20 °C). We assumed that radicals preserved after the 120-s holding time could be effectively frozen in liquid nitrogen. During the in situ experiments, the coal sample was heated in a spectrometer-equipped heater at the designated pyrolysis temperature for 120 s and ESR measurements were performed at the pyrolysis temperature.

The ordinate axis in Fig. 1 represents the normalized absorption peak area of the ESR spectrum, which is proportional to the normalized sum of radicals. These results showed differences between the ex situ and in situ experiments for the coal sample.

Generally, the normalized absorption peak areas from in situ experiments are lower than those from ex situ experiments, because high measurement temperatures tend to give smaller values. The higher the measuring temperature the greater the difference observed.

However, the number of radicals generated should be the same if the radicals studied were frozen in liquid nitrogen effectively. Hence, the temperature dependence of the ESR measurement must be the main factor contributing to this the difference.

1. Temperature conversion based on Boltzmann distribution law

Theoretically, the signal measured by ESR depends on the distribution of un-paired electrons in high and low energy states (α and β) according to the Boltzmann distribution law, as given by Eq. 1:

nαnβ=eΔEkT=ehνkT,
(1)

where nα is the number of free electrons in the high energy state; nβ is the number of free electron in a low energy state; ΔE is the energy difference between the two spin states; h is Planck’s constant; ν is the wave frequency; T is temperature; and k is the rate constant.

Clearly, the ratio of un-paired electrons (nα/nβ) in the high and low energy states varies with temperature. The ESR signal intensity also varies with the nα/nβ ratio for the same total number of un-paired electrons, which also vary with the measurement temperature. A high measurement temperature corresponds to a high nα/nβ ratio, which results in a lower ESR signal intensity. Therefore, to make reasonable comparisons, results have to be converted according to the measurement temperature.

Based on the Boltzmann distribution law, results measured at one temperature can be converted to another temperature with multiplication by a conversion factor. To enable a reasonable comparison between the results obtained from the ex situ and in situ experiments shown in Fig. 1, the in situ results were converted according to Eq. 2:

AT293=ATfT,
(2)

where AT(293) is defined as the ESR absorption peak area at 20 °C; AT is the peak area measured at the in situ temperature, and fT is the conversion factor, which can be calculated according to Eq. 3, derived from Eq. 1 as:

fT=ehνk12931T
(3)

The results after conversion are shown in Fig. 2. The filled symbols represent data scaled to a room temperature measurement temperature. The unfilled symbols represent data from the ex situ experiments, for which the ESR data was collected at the room temperature. As expected, the differences between the data sets were diminished when the data were adjusted for the temperature level. In addition, the data were calibrated against pure MnCl2, a commonly used standard for calibrating the signal of ESR instruments and our temperature conversion process was verified.

These results further confirmed that the temperature conversion performed in this study was effective and that data from in situ and ex situ ESR shows the presence of steady radicals formed during coal pyrolysis. Therefore, the results from only the ex situ ESR measurement are adequate for studying the effects of process variables on free radicals in the solid phase during coal pyrolysis.

Fig. 3(a) shows that the peak amplitude of the integrated curve varied with temperature. The central magnetic field of the ESR spectra deviated with increasing temperature, which indicates that different types of free radicals formed as the temperature was changed. This suggestion is supported by the change of g values, as shown in Fig. 5. Because the normalized peak area is proportional to the radical concentration the variation of the radical concentration (CS) shown in Fig. 3 can be separated into three distinct regimes as temperature increased. First, CS increases with increasing temperature and features two turning points at 250 and 380 °C. The rapid increase in CS above 380 °C has received considerable attention in coal conversion studies.

From Fig. 3(b), three distinct temperature regimes were observed. The apparent activation energy Ea for the formation of free radicals has been calculated based on the equation of CS = CS0+ Aexp(−Ea/RT) where R is the gas constant,31 A is a constant, and CS0 is the baseline concentration at each stage. The Ea values for stage 1 and 3, determined from the slopes of log(CS − CS0) versus 1/T plots as listed in Fig. 4.

Petrakis and Grandy16 have listed the dissociation energies of various radicals and bonds relevant to coal pyrolysis. These energies vary from 10 to 100 kcal/mol. Fig. 4 shows that the observed magnitudes for Ea in regime 3 are within this range and provides direct quantitative support for the hypothesis that coals are depolymerized into new free radicals at temperatures above 380 °C. However, the Ea values for regime 1 were too low to be bond dissociation energies and the Ea values for regime 2 were negative. The three stages of all coals are different, and controlled by different reaction mechanisms.

One explanation for the three-stage feature of Cs has been proposed32 based on progressive release of CO, CO2, and H2O as temperature increases when coal is heated to near 250 °C. Fig. 6 shows that the g values of free radicals in regime 1 decreased from 2.00379 to 2.00352 with increasing temperature, relative to the g-values of O-containing radicals (2.0035–2.00398). The release of H2O, O, CO2 was also observed in Fig. 6, which is associated with breaking of weak bonds to generate new radicals that cause an increase of Cs in regime 1. Various types of free radical and their typical g value data are listed in Table II.33 

When coal is heated at temperatures between 250 and 380 °C, certain coals swell and decompose to release hydrogen-rich molecules. This phenomenon is partially supported by the BET data in Table III. We found that the surface area decreased as the temperature was increased from 250 to 350 °C. This result indicates that the space in the coal structure became restricted as the structural units became closer to each other, inducing stronger interactions of free radicals in the different coal structure units. In turn, the free radicals were more likely quenched by reactions between themselves.

In regime 3, the apparent activation energy was 28.750 Kcal/mol, which is in the range of the dissociation energies of covalent bonds in coals. This suggests that free radicals were generated from covalent bond cleavage in the coal. The g values in regime 3, decreases from 2.00295 to 2.00281 with increasing temperature, which is within the g-value range of aromatic radicals (2.0025–2.00291). The formation of aromatic radicals is also implied by the release of the small-molecule alkanes above 380 °C, as shown in Fig. 6.

Radical concentrations of the samples used in this study were calibrated against DPPH from their sample peak areas and the calibration line is shown in Fig. 7. Fig. 8 shows a comparison of the free radical concentration with residence time at various pyrolysis temperatures. We found that the residence time had no marked effect on free radical concentration at low temperature, which indicates that a negligible amount of covalent bonds were broken under these conditions. However, the free radical concentration clearly increased with longer residence time at high temperature owing to extensive bond cracking of coal. The free radical concentration increased more rapidly as time proceeded at higher temperatures.

In non-hydrogen donor coal pyrolysis processes, or when the free radicals have no access to an external hydrogen donor, free radicals may abstract hydrogen from the hydrogen-poor coal structure or combine with other large free radical fragments to form a high molecular weight char. When a hydrogen-donor solvent is added, the free radical may abstract hydrogen from the donor and form a relatively low molecular weight product. This is a typical process in coal liquefaction, and the hydrogen donor solvent plays a key role. Performing our ESR measurement on coal mixed with solvent at high pressure allowed observation of the properties of free radicals similar to those formed under actual liquefaction conditions.

Fig. 9 shows the free radical concentration of coals treated with naphthalene and tetralin with increasing temperature at 5 MPa under a N2 atmosphere. Naphthalene, an aromatic compound, is a poor solvent for quenching free radicals; however, tetralin, a hydroaromatic, can provide hydrogen and form stable aromatic compounds, making it a more effective hydrogen donor. We found that the free radical concentration of coal treated with naphthalene was higher than that of coal treated with tetralin. This difference became more notable at higher temperatures and for longer residence times. Thus, a certain fraction of the free radicals were quenched by the hydrogen-donor solvent, and the amount of hydrogen in tetralin was sufficient to quench newly-generated free radicals and to maintain the residual free radicals at a low concentration. Some of more stable radicals could not be quenched by tetralin, and were likely protected by a p-π conjugate system in the coal tar or in polycyclic aromatic hydrocarbons. The concentration of this type of radical is a useful index for evaluating the hydrogen-donor ability of a solvent. Fig. 9 shows that pressure had no notable effects on free radical concentration.

We investigated the behavior of free radicals in coal pyrolysis processes using in situ and ex situ ESR. The measurement temperature was found to markedly influence the ESR signals associated with the quantitative analysis of free radicals. We compared the results of these two of measurement techniques. We observed disagreement between the results of in situ and ex situ ESR; however, the data could be scaled to the same temperature basis for comparison by multiplication with temperature calibration factors.

Although in situ ESR is considered to be necessary for identifying radicals generated in coal pyrolysis, our research showed similarities between the ex situ and the in situ ESR measurement results after scaling. This result suggests that steady (long-lived) radicals in coal pyrolysis processes, can survive for a period of time at ambient temperature.

Temperature and residence time are important process variables that affect the generation and behavior of free radicals in the solid phase. Three distinct stages of radical concentration vs temperature were observed. The apparent activation energies for the regimes were different, which implied that the three stages were controlled by different reaction mechanisms.

A hydrogen-donor solvent can partially quench active radicals, but stable radicals remained in the solid phase. In the presence of a sufficient quantity of a hydrogen donor, the concentration of free radicals was maintained at low level. The concentration of these radicals may be used as an index to evaluate the hydrogen-donor ability of the solvent.

We thank Synfuels China Technology Co. Ltd. for financial support. We thank Andrew Jackson, PhD, from Liwen Bianji, Edanz Group China (www.liwenbianji.cn/ac), for editing the English text of a draft of this manuscript.

1.
P. R.
Solomon
,
T. H.
Fletcher
, and
R. J.
Pugmire
, “
Progress in coal pyrolysis
,”
Fuel
72
,
587
597
(
1993
).
2.
J.
Gibbins-Matham
and
R.
Kandiyoti
, “
Coal pyrolysis yields from fast and slow heating in a wire-mesh apparatus with a gas sweep
,”
Energy & Fuels
2
,
505
511
(
1988
).
3.
R. J.
Tyler
, “
Flash pyrolysis of coals. 1. Devolatilization of a Victorian brown coal in a small fluidized-bed reactor
,”
Fuel
58
,
680
686
(
1979
).
4.
L.
Shi
,
Q.
Liu
,
X.
Guo
,
W.
Wu
, and
Z.
Liu
, “
Pyrolysis behavior and bonding information of coal—A TGA study
,”
Fuel Processing Technology
108
,
125
132
(
2013
).
5.
D. B.
Anthony
,
J. B.
Howard
,
H. C.
Hottel
, and
H. P.
Meissner
, “
Rapid devolatilization and hydrogasification of bituminous coal
,”
Fuel
55
,
121
128
(
1976
).
6.
A. H.
Al-Abbas
and
J.
Naser
, “
Computational fluid dynamic modelling of a 550 MW tangentially-fired furnace under different operating conditions
,”
Procedia Engineering
56
,
387
392
(
2013
).
7.
H. N.
Stone
,
J. D.
Batchelor
, and
H. F.
Johnstone
, “
Low temperature carbonization rates in a fluidized bed
,”
Industrial & Engineering Chemistry
46
,
274
278
(
1954
).
8.
D. W.
Van Krevelen
,
C.
Van Heerden
, and
F. J.
Huntjens
, “
Physicochemical aspects of the pyrolysis of coal and related organic compounds
,”
Fuel
30
,
253
259
(
1951
).
9.
K.
Miura
,
K.
Mae
,
M.
Shimada
, and
H.
Minami
, “
Analysis of formation rates of sulfur-containing gases during the pyrolysis of carious coals
,”
Energy & Fuels
15
,
629
636
(
2001
).
10.
J.
Machnikowski
,
H.
Machnikowska
,
M. A.
Díez
,
R.
Alvarez
, and
J.
Bermejo
, “
Characterization of coal-derived pitches as precursors for advanced carbon materials by chromatographic and related techniques
,”
Journal of Chromatography A
778
,
403
413
(
1997
).
11.
J.
Machnikowski
,
H.
Kaczmarska
,
A.
Leszczyńska
,
P.
Rutkowski
,
M. A.
Díez
,
R.
Álvarez
, and
R.
García
, “
Hydrogen-transfer ability of extrographic fractions of coal-tar pitch
,”
Fuel Processing Technology
69
,
107
126
(
2001
).
12.
D. L.
Zoller
,
M. V.
Johnston
,
J.
Tomic
,
X.
Wang
, and
W. H.
Calkins
, “
Thermogravimetry−photoionization mass spectrometry of different rank coals
,”
Energy & Fuels
13
,
1097
1104
(
1999
).
13.
M.
Sun
,
X.-X.
Ma
,
Q.-X.
Yao
,
R.-C.
Wang
,
Y.-X.
Ma
,
G.
Feng
,
J.-X.
Shang
,
L.
Xu
, and
Y.-H.
Yang
, “
GC-MS and TG-FTIR study of petroleum ether extract and residue from low temperature coal tar
,”
Energy & Fuels
25
,
1140
1145
(
2011
).
14.
C.
Lievens
,
D.
Ci
,
Y.
Bai
,
L.
Ma
,
R.
Zhang
,
J. Y.
Chen
,
Q.
Gai
,
Y.
Long
, and
X.
Guo
, “
A study of slow pyrolysis of one low rank coal via pyrolysis–GC/MS
,”
Fuel Processing Technology
116
,
85
93
(
2013
).
15.
J.
Ubersfeld
,
A.
Etienne
, and
J.
Combrisson
, “
Paramagnetic resonance, a new property of coal-like materials
,”
Nature
174
(
1954
).
16.
L.
Petrakis
and
D. W.
Grandy
,
Free radical in coal and synthetic fuels [M]
(
Elsevier
,
Amsterdam
,
1983
).
17.
T. G.
Fowler
,
K. D.
Bartle
, and
R.
Kandiyoti
, “
Limitations of electron spin resonance spectroscopy in assessing the role of free radicals in the thermal reactions of coal
,”
Energy & Fuels
3
,
515
522
(
1989
).
18.
T. G.
Fowler
,
K. D.
Bartle
, and
R.
Kandiyoti
, “
Low temperature processes in a bituminous coal studied by in situ electron spin resonance spectroscopy
,”
Fuel
66
,
1407
1412
(
1987
).
19.
T. G.
Fowler
,
K. D.
Bartle
, and
R.
Kandiyoti
, “
Role of evolved volatiles during pyrolysis of a bituminous coal as deduced from in situ electron spin resonance spectroscopy
,”
Fuel
67
,
173
176
(
1988
).
20.
L.
Petrakis
and
D. W.
Grandy
, “
Formation and behaviour of coal free radicals in pyrolysis and liquefaction conditions
,”
Nature
289
,
476
477
(
1981
).
21.
T.
Yokono
,
K.
Murakami
, and
Y.
Sanada
, “
The effect of heating rate on radical concentration during coal pyrolysis
,”
Fuel Processing Technology
17
,
7
11
(
1987
).
22.
C.
Bolton
,
C. E.
Snape
,
R. J.
O’Brien
, and
R.
Kandiyoti
, “
Influence of carrier gas flow and heating rates in fixed bed hydropyrolysis of coal
,”
Fuel
66
,
1413
1417
(
1987
).
23.
N.
Qiu
,
H.
Li
,
Z.
Jin
, and
Y.
Zhu
, “
Temperature and time effect on the concentrations of free radicals in coal: Evidence from laboratory pyrolysis experiments
,”
International Journal of Coal Geology
69
,
220
228
(
2007
).
24.
B.
Pilawa
,
A. B.
Wiȩckowski
, and
M.
Lewandowski
, “
E.p.r. studies of thermal decomposition of exinite and inertinite
,”
Fuel
75
,
1181
1185
(
1996
).
25.
I. C.
Lewis
and
L. S.
Singer
, “
Electron spin resonance of stable aromatic radical intermediates in pyrolysis
,”
Carbon
7
,
93
99
(
1969
).
26.
L. R.
Rudnick
and
D.
Tueting
, “
Investigation of free radicals produced during coal liquefaction using ESR
,”
Fuel
63
,
153
157
(
1984
).
27.
M. S.
Seehra
,
B.
Ghosh
, and
S. E.
Mullins
, “
Evidence for different temperature stages in coal pyrolysis from in situ e.s.r. spectroscopy
,”
Fuel
65
,
1315
1316
(
1986
).
28.
T.
Yokono
,
S.
Iyama
,
Y.
Sanada
,
S.
Shimokawa
, and
E.
Yamada
, “
High temperature and high pressure 1H n.m.r. and e.s.r. studies on coal liquefaction reactions
,”
Fuel
65
,
1701
1704
(
1986
).
29.
Z. S.
Gönenc
,
T. G.
Fowler
,
R.
Kandiyoti
, and
K. D.
Bartle
, “
Correlation of tar yields from a fixed bed coal pyrolysis reactor with in-situ e.s.r.-spin population measurements
,”
Fuel
67
,
848
852
(
1988
).
30.
T. G.
Fowler
,
K. D.
Bartle
, and
R.
Kandiyoti
, “
Role of evolved volatiles during pyrolysis of a bituminous coal as deduced from in situ electron spin resonance spectroscopy
,”
Fuel
67
,
173
176
(
1988
).
31.
M. S.
Seehra
and
B.
Ghosh
, “
Free radicals, kinetics and phase changes in the pyrolysis of eight American coals
,”
Journal of Analytical and Applied Pyrolysis
13
,
209
220
(
1988
).
32.
M. L.
Gorbaty
,
J. W.
Larsen
, and
I.
Wender
,
Coal Science [M]
(
Academic Press
,
1982
), p.
1
.
33.
L.
Petrakis
and
D. W.
Grandy
,
Free radical in coal and synthetic fuels [M]
(
Elsevier
,
Amsterdam
,
1983
).