166
Energy &Fuels 1992,6, 166-172
Spectroscopic Analyses of Aromatic Hydrocarbons Extracted from Naturally and Artificially Matured Coals Zouhir Benkhedda and Patrick Landais* CREGU and GDR CNRS-CREGU, BP 23, 54501 Vandoeuvre Cedex, France
Jacky Kister Centre de Spectroscopie MolBculaire, Facult: des Sciences et Techniques St JBrcime, 13397 Marseille Cedex, France
Jean-Marie Dereppe CRIS, UniversitB de Louvain-la-Neuve, 1 Place Louis Pasteur, 1348 Louvain-la-Neuve, Belgium
Marc Monthioux Laboratoire Marcel Mathieu, 2 Avenue du Prgsident Angot, 64000 Pau, France Received August 28, 1991. Revised Manuscript Received December 26, 1991
Experimental simulation of Mahakam delta (Indonesia) coals maturation has been carried out in a confined system under pressure. Chloroformic extracts have been fractionated by liquid chromatography. Aromatic hydrocarbons have been concentrated and analyzed by FTIR spectroscopy (Nicolet 20 SX B), high-resolution proton NMR (Brucker AM 500 MHz), and synchronous excitation-emission UV fluorescence (Perkin-Elmer LS 50). Band assignment was made in order to take into account the evolution of the proton NMR, FTIR, and synchronous W fluorescence spectra with respect to bibliographic data and spectral resolution. Results have been compared to thm of a natural homogeneous series of coals of increasing rank. The corrected organic carbon content (% COC = 90C X loo/(% C + % H + % 0 + % N)) was chosen as a common maturity index for both natural and artificial series. The progressive removal of aliphatic chains, the increase in the proton aromaticity, and the structural rearrangement are similar for both series. However, some significant discrepancies are noticed between the artificial and natural series. They mostly concern the rate of aromatization, the intensity of ring condensation, and the amount of extractable hydrocarbons.
Introduction Detailed analysis of soluble fractions of coal is crucial for understanding their chemistry and for evaluating their modification during thermal maturation. As a matter of fact, the geothermal gradient in natural systems as well as thermal treatment in pyrolysis runs are responsible for the cracking of the kerogen and the genesis of mobile liquid and gaseous hydrocarbons. Numerous models were developed in order to describe the coal structure and its evolution during thermal treatment. The last model refers to a two-phases system constituted by a mobile, liquid, or so-called molecular phase trapped in the rigid phase of the macromolecularstructure.'J The extractable phase of coal can be studied by c h r ~ m a t o g r a p h ychromatography,~~~ mass ~pectrometry,~ thin layer chromatography, or highpressure liquid chromatography.6 'H and 13C nuclear magnetic resonance (NMR)'?*are able to provide reliable average structural parameters such as aromatic and aliphatic content, Fourier transform infrared (FTIR)QJO allows the functionality of the atomic groups to be determined and W synchronous fluorescence (SFS)" provides more global structural data. Such spectroscopic techniques are nondestructive and can be applied to the study of soluble compounds of various average molecular weights as well as solid samples. Furthermore, in most cases they are fast and do not require special sample preparation. On the other hand, simulation of organic matter is an important topic in the field of petroleum science. Various *Author to whom correspondence should be addressed.
experimental systems have been used to pyrolyze organic matter, and it has been shown that confinement was an important parameter in the simulation studieP which favors hydrogen transfers and the stabilization of the free radicals.13 Confiied pyrolysis has proven to reproduce most of the transformations undergone by organic matter during maturation.14J5 However, several discrepancies between natural maturation and confined pyrolysis concerning the amount of total extractable hydrocarbons have been no(1)Kircho, A. A.; Gagarin, s. G. Fuel 1990,69, 885-891. (2) Larsen. J. W.: Mohammadi. M. Enerm Fuels 1990. 4. 107-111. (3)Landais, P.;Zaugg, P.; Monin, J. C.; K h r , J.; Muller, J, F. Bull. SOC.Geol. Fr. 1991, 2, 211-217. (4)Herod, A. A.;Ladner, W. R.; Snape, C. E. Philos. Trans. R. SOC. London A 1981,300, 3-14. (5)Nip, M.; de Leeuw, J. W.; Shenck, P. A. Geochim. Cosmochim. Acta 1988, 52, 637-648. (6)Fox, M. A.; Staley, S.Anal. Chem. 1976,48, 992-998. (7)Dereppe, J. M.; Moreaux, C.; Castex, H. Fuel 1978,57, 435-441. (8)Dickinson, E.M. Fuel 1980, 59, 290-294. Griffiths, P. R. Fuel 1985, 64, 229-236. (9)Wang, S.H.; (10)Sobkowiak, M.; Reisser, E.; Given, P.; Painter, P. Fuel 1984,63, 1245-1252. (11)Lloyd, J. B.F. Nature 1971, 231, 64-65. (12)Landais, P.:Monthioux, M. Fuel Process. Technol. 1988, 20, 123-132. (13)Grint, A.; Mehani, S.; Trewhella, M.; Crook, M. T. Fuel 1985,64, 1355-1361. (14)Monthioux, M.; Landais, P.; Durand, B. Org. Geochem. 1986,10, 299-311. (15)Landais, P.;Michels, R.; Poty, B.; Monthioux, M. J. Anal. Appl. Pyrol. 1989, 16, 103-115.
0887-0624f 92f 2506-0166$03.00f 0 0 1992 American Chemical Society
Energy & Fuels, Vol. 6, No. 2, 1992 167
Aromatic Hydrocarbons from Coals
Table I. Evolution of the Corrected Organic Carbon Percentage (COC %), CHCIJ Extract and Aromatic (ARO) Hydrocarbons (HC) Yields (mg/g of total organic carbon) for Natural and Artificial Series of Mahakam Coals artificial maturation natural maturation extractable HC, extractable ARO, sample extractable HC, extractable AFtO, temp, "C % COC mg/g of org C mg/g of org C no. % COC mg/g of org C mg/g of org C 250 71.3 10.9 7.0 32362 63.9 7.6 4.4 300 77.4 16.4 12.8 37247 74 11.2 8.5 4.1 350 83.2 41.7 21.4 31590 76.4 6.8 14.8 10.7 375 86.1 63.0 22.8 37249 78.2 400 87.2 26.0 15.6 37251 80.9 n.d.a n.d. 37252 81.9 12.7 8.2 425 89 12.6 11.5 11.4 4.6 450 90.8 6.1 5.5 37254 83.7 37255 84.5 9.6 6.5 "n.d. = not determined.
ticed. They have been related to the occurrence of free but trapped hydrocarbons in the natural series of coal.16 Additional investigations have demonstrated that the composition of the chloroformic extract in terms of saturatee, aromatics, and polar compounds and the distribution of saturated hydrocarbons were similar for both series.14 The analysis of aromatic hydrocarbons by GC has also revealed similarities between natural and artificial maturation." Even if the aromatic hydrocarbon fraction only represents a small part of the total organic matter involved in the maturation process, it offers the opportunity to study aromatization, condensation, and desubstitution processes which cannot always be easily characterized either on the larger molecules such as asphaltenes or on the solid insoluble residue. This is why the present work aims to combine lH NMR, FTIR, and UV synchronous fluorescence spectroscopy (SFS) in order to characterize the soluble aromatic hydrocarbons fractions separated from seven artificially and eight naturally matured type I11 Mahakam (Indonesia) coals and to show how the information obtained from these spectroscopic studies can be used for comparing artificial and natural maturation series. These spectroscopic techniques have been selected as a function of their respective performance but also as a function of the available quantities of aromatic hydrocarbons. As matter of fact, aromatic hydrocarbon yields (1.1-4.6mg) did not allow 13C NMR to be performed. This technique would be helpful to distinguish between primary (CH3), secondary (CH2), tertiary (CH), and quaternary (C) carbons.
Experimental and Analytical Section Confined Pyrolysis in Cold-Seal Autoclaves. Artificial maturation of 200 mg of immature coal has been carried out in a confined pyrolysis system under pressure (100 MPa) for 24 h at temperatures ranging between 250 and 450 O C (isothermal). At the end of each run,pyrolysis residues have been collected from gold tubes and extracted in chloroform. Liquid chromatography of chloroform extracts on microcolumns yields to separate aliphatic and aromatic hydrocarbons fractions from NSO compounds. Detailed information on the experimental and analytical procedures is given in Landais et al.15 Kerogen-corrected organic carbon content (% COC = % C X 100/ (%C + % H % 0 + % N)) has been selected as common maturity parameter for both artificial and natural series of maturation.14J6 Data on COC content, CHC13 extractable hydrocarbons, and aromatic hydrocarbons yields are given in Table I. The deepest sample (3400 m) from natural series of maturation of the Mahakam delta displays a COC content of 8 4 4 5 % . Conversely the artificial maturation series yields samples showing COC content as high as 90.8% (450 OC experiment). Even if such high COC values cannot be recorded in the Mahakam delta coals, it remains interesting to experimentally investigate the behavior
+
(16) Monthioux, M.; Landais, P. Fuel 1987,66, 1703-1708. (17) Monthioux, M.; Landais, P. Chem. Geol. 1889, 77, 71-85.
Table 11. Assignment bands shift ranges, ppm 0.50-1.10 1.10-1.45 1.45-2.00 2.00-2.40 2.40-4.00 6.50-7.24 7.24-8.30 8.30-8.90 a, 8, y, or matic rings.
of the Major 'H NMR Bands" major assignments y+(CH3) y+(CHa CH) and B(CH3) B(CH2, CH) a(CH3) a(CH2, CHI uncondensed aromatic rings (alkyl and naphthnobenzenes) condensed aromatic rings peripheral condensed aromatic rings angular
+ (further) are positions of saturated groups to aro-
of aromatic hydrocarbons for higher levels of maturity which corresponds to realistic geological conditions over a long period of time, i.e., 185 O C during 30 Ma. Proton Magnetic Resonance ('H NMR). 'H NMR spectra have been obtained on a high-resolution Brucker spectrometer AM series. The superconducting magnet system operates a t a field of 11.7 T, providing a 'H resonance of 500.14 MHz. Aromatic fractions have been diluted in CDCIBin 5 mm diameter sample tubes at room temperature. TMS was used as lock signal as well as for an internal standard. Base-line correction was carried out using the DCNMR program. Each spectrum recorded was the result of the accumulation of 128 scans and a sweep width of loo00 Hz has been chmen. Signals detected in the -0.5 to 10 ppm chemical shift range from TMS were attributed to aromatic and aliphatic protons (Table 11). The choice of the different integration domains for 'H NMR spectra has been extensively discussed in the literature. In this study, the signal intensities in the 0.5-4 ppm range are ascribed to aliphatic protons ( H d ) and divided into five bands (a, b, c, d, Aromatic protons absorption (H,)in the 6.5-8.9 ppm spectral range have been divided into three bands (f, g, h).22,23 No signal in the 4-6 ppm region correspondingto olefinic hydrogen absorption has been recorded. The assignments of 'H Nh4R bands are summarized in Table 11. A typical spectrum of the aromatic hydrocarbons fraction is given in Figure la. The integrated intensities of the various resonance areas are normalized to 100. Three parameters have been calculated from the integration of 'H NMR bands: (1)CH, = f + g + h, which represents the percentage of aromatic protons; (2) n = (a + b + c + d + e)/(d + e) = C H d / C H , is the number of aliphatic carbon atoms per saturated substituent;% and (3) Q = (e + d + c)/(a + b) has been (18) Delpuech, 3. J. Adu. Top. Appl. Fossil Energy 1984, 351-376. (19) Prebch, E.; Clerc, T.; Seibal, J.; Simow, W. Tables of Spectral Data For Structure Determination of Organic Compounds; SpringerVerlag: Berlin, 1983; pp H5-H365. (20) Hazlett, R. N.; Dorn, H. C.; Glass, T. E. Adu. Top. Appl. Fossil Energy 1984,351-376. (21) Rongao, L.; Zengming, 2.; Bailing, L. Fuel 1987, 66, 565-571. (22) Lloyd, J. B. F. The Analyst 1974,99, 729-738. (23) Chamberlin, N. F. The Practice of NMR Spectroscopy;Plenum Press: New York, 1974; pp 15-74. (24) Williams, R. B. Symposium Composition of Petroleum Oils, Determination and Evaluation; ASTM Spec. Tech. Publications 224; 1958, 168-194.
et al.
168 Energy & Fuels, Vol. 6, No. 2, 1992 Aromatic protons
Alpihatic protons
1
32362 0
8: 7-
w
I/
6-
0
*
5-
:I
o
2
60
65
70
75
80
85
90
95
%COC 8.b4
4.h
5.h
7.k6
3.h
1.h
0.b6)PPM
(corrected organic carbon 5%) content of the coals. Table IV. Synchronous UV Fluorescence Data for Polycyclic Aromatic Hydrocarbons
6 Alpihatic C-H
+ ,
Figure 2. Plot of the lH NMR parameter n (number of aliphatic carbon atoms per saturated substituent) as a function of the COC
'
1 RING
2RINGS
3"Gs 4 ANGULAR R N G S
5 PERCONDENSED AND 6 U T A CONDENSED RINGS
c-HJk S Aliehatic
19 Aromatic
)
40d0
34i3
2867'
23iO
143
U
"
1d7
U
6O ;
bcm-l
Fmre 1. Typical 'H NMR (a, top) and FTIR (b, bottom) spectra obtained on aromatic hydrocarbon fractions from the Mahakam coals. Table 111. Assignment of the FTIR Bands Used in the Current Study
1364 915-852 780-725
"ar, aromatic; al, aliphatic; Y, stretching vibration; 6, deformation vibration in plane; 7 , deformation vibration out of plane; as, asymmetric; 8 , symmetric. used for estimating the rearrangement of y+ groups and @(CH3) groups versus remaining alkyl groups.25 Fourier Transform I n f r a r e d (FTIR) Spectra. FTIR spectra have been recorded on a Nicolet 20 SXB Fourier transform spectrometer. Each spectrum resulted from the accumulation of 124 scans recorded with a spectral resolution of 2 cm-' in the 600-4OOO-cm~'spectral domain. Valley to valley integration mode has been chosen for bands area determination. Assignment of the different FTIR bands absorption is reported in Table 111. A deconvolution technique using the Bessel program has been applied to increase the apparent spectral resolution of overlapped infrared bands which appear in the 1390-136O-cm-' spectral range. This allows the 6,(CH3) (1376cm-') from 68+,(CH3) (1384 cm-') and total 6,+8+,(CH3) (1364cm-') to be distinguished.% A typical (25)Smith, W.E.; Napier, B.; Home, 0. J. Preeented at the American Chemical Society, Philadelphia Meeting, 1975;pp 369-374. (26)Guiliano, M.; Mille, G.;Kister, J.; Dou, H. Spectra 1985,13,103, 35-39.
I
I
I
I
I
I
1
280
310
340
370
400
430
460
nm
FTIR spectrum of aromatic hydrocarbons fraction is given in Figure lb. The ARO = (C=C)/(w(CH2) + u(CH,) + C=C) absorbance ratio2' was calculated in order to estimate the aromaticity of aromatic rings. Out of plane aromatic 4 - H bands (915-725 cm-') were used to determine the ring substitution pattern. The y(CH), (1 adjacent H)/r(CH,) (four adjacent H) represents the degree of substitution of aromatic rings. Synchronous UV Fluorescence (SFS). Synchronous UV fluorescence spectra have been recorded on a Perkin Elmer LS 50 spectrometer. The excitation and emission wavelengths respectively range between 200 and 800 nm and 200 and 650 nm. THF has been wed as solvent in order to avoid quenching effects. Synchronous W fluorescence analysis corresponds to continuous scan of excitation (AJ and emission (AJ wave8 which are constantly separated by a wavelength differential (AA = A, - A,) which is generally empirically determined. Lloyd'l and Mille et al.% as well as John and Soutarm recommended a AA of 20-25 nm for oil sample analysis, whereas Vo-dinh et a1.30and Der Dick and Kalkreuth31chose a PA of 3 nm for the study of complex aromatic mixtures, humic coals and kerogen extracts. In the present study, an adequate spectral resolution of the aromatic hydrocarbons synchronous fluorescence spectra was obtained with a AA of 23 nm. Assignment of the different bands was determined from the works of Lloyd and Evett,32 Katoh et al.,33 and Mckay and Latham% and are reported in Table IV. (27)Low, M.J. D.; Glass,A. S.Spectrosc. Lett. 1989,22,417-429. (28)Mille, G.;Kister, J.; Guiliano, M.; Dou, H. Spectra 1985,13,103, 27-31. (29)John, Ph.; Soutar, I. Anal. Chem. 1976,48,520-524. (30)Vo-Dinh, T.;Gammage, R. B.; Martinez, R. Anal. Chem. 1981,53, 253-258. (31)Van Der Dick, H.; Kalkreuth, W. Org. Ceochem. 1984, 10, 633-639. (32)Lloyd, J. B. F.; Evett, I. W. Anal. Chem. 1977,49, 1710-1715. (33)Katoh, T.;Yokoyama, S.; Sanada, Y. Fuel 1980,59, 845-850. (34)Mckay, J. F.;Latham, D. R. Anal. Chem. 1972,44,2132-2137.
Aromatic Hydrocarbons from Coals
Energy & Fuels, Vol. 6, No. 2, 1992 169
Table V. Integrated Intensities of the Various 'H NMR Bands for the Artificial Maturation pyrolysis temperatures, "C initial sample 32362 250 300 350 375 400 32.6 27.8 18.9 15.6 10.6 4.6 23.6 11.9 29.9 26.9 25.8 18.5 21.0 14.4 11.5 19.5 18.9 10.9 7.27 5.8 6.6 5.8 4.9 4.0 6.6 10.9 17.6 21.9 26.8 27.2 2.1 3.6 4.2 6.3 5.7 5.1 11.7 1.7 3.8 6.9 17.8 29.9 0.2 0.7 1.1 1.7 3.6 5.5
Seriesa 425 3.5 10.1 9.5 3.0 23.9 3.3 39.5 6.9
450 1.8 8.4 5.8 2.4 19.5 2.4 51.5 8.3
For band assignments see Table 11.
a b C
d e f g h
Table VI. Integrated Intensities of the Various 'H NMR Bands for the Natural Maturation Series" samples initial sample 32362 37247 31590 37349 37251 37252 37254 32.6 28.0 25.9 19.6 16.7 10.7 8.7 29.9 27.2 26.6 24.7 23.9 18.9 16.9 21.0 19.6 18.5 17.7 15.7 12.7 10.9 5.8 7.8 6.8 6.3 5.6 3.4 2.9 6.6 7.8 11.5 18.1 21.9 23.2 21.5 2.1 3.6 3.9 4.5 4.9 3.5 3.9 1.7 4.6 5.9 7.8 9.6 24.4 31.1 0.2 0.6 0.95 1.3 1.7 3.1 4.1
37255 7.9 15.9 9.8 2.5 22.4 3.1 34.0 4.4
For band assignments see Table 11.
Results The evolution of the aromatic hydrocarbons fractions has been studied by emphasizing the role of four major transformations occurring during thermal maturation: (i) variation of the average aliphatic chain length; (ii) aromatization; (iii) substitution of aromatic rings; (iv) condensation. In each case, the parameters deduced from the different analytical techniques have been crosschecked in order to provide more accurate evolutionary trends. Variation of the Average Aliphatic Chain Length. The parameter n exponentially decreases with increasing values of COC (Figure 2). This is related to the variation of a / C H d band ratio which ranges from 0.34 to 0.047 and from 0.340 to 0.135 during artificial and natural maturation, respectively. This indicates that for both maturation series the average aliphatic chain length drastically decreases during maturation. Furthermore, it can be noticed that natural and artificial maturation trends remain parallel. The parameter n unfortunately gives no indication on the alkyl groups distribution. However, complementary information can be obtained from the 6,(CH,),/6,(CHJd (a(CH3)/(B+ 7)(CH3))ratio deduced from the deconvolution of the 1360-1390-cm-' bands FTIR region which increases from 1.3 for initial coal (32362) to 20 and 37 for natural and artificial maturation series end members, respectively (Figure 3). As far as the absolute quantity of CH,, is not expected to significantly increase during maturation, it is concluded that the thermal maturation affects the parafiiic chains (Tables V and VI) and induces a preferential concentration of cyclic structures attached to aromatic rings. This hypothesis is supported by the evolution of the NMR parameter Q which exemplifies the preferential removal of y hydrogens (Figure 4) and corresponds to the formation of cyclic structures (high values for H, and Hgrelative to H,).25 This also seems to be in agreement w t h the fact that the 'H NMR e band, attributed to a(CH2,CH) ratioed to C H d (e/md), increases from 0.07 (initial sample) to 0.52 and 0.38 for artificial and natural maturation series end members, respectively. Such evolution corresponds to the concentration of methylene bridges and/or naphthenic structures. Furthermore, two different stages can be distinguished in these two last diagrams (Figures 3 and 4): (i) up to 83%
3'
30
.
25
?
20
-
60
75
70
65
85
80
90
%COC Figure 3. Evolution of the aromatic CH,/aliphatic CH3 ratio determined by FTIR spectroscopy during thermal maturation. Artificial maturation
2fo
1
0,O-I 60
I .
' 65
.
' 70
.
' 75
.
' 80
-
'
'
85
90
. '
95
%COC Figure 4. Evolution of the 'H N M R parameter Q (Ha + H,/H,) during thermal maturation.
COC, the variation of aliphatic chain length is limited, and (ii) above 83% COC the average aliphatic chain length drastically decreases. Aromaticity. The aromaticity of the aromatic hydrocarbons fractions increases with maturation for both series. C H , increases from 4% for initial coal to 62% and 41% for artificial and natural series end members respectively (Figure 5). It can be noticed that aromatization is faster
170 Energy & Fuels, Vol. 6, No. 2, 1992 0
Benkhedda et al.
I
Artificial maturation
6-
0
Artificialmaturation Natural maturation
1
0 5-
4-
332362 2-
60
65
70
75
80
85
90
95
%COC
Figure 7. Evolution of the degree of substitution (y(CH), (1 adjacent H)/?(CH), (4 adjacent H)) determined by FTIR durmg thermal maturation. INITIAL SAMPLE
Y
W NATURAL MATURATION
ARTIFICIAL MATURATION
L!L
stage 111
ZCOC = 87.2
Figure 8. Comparison of aromatic hydrocarbon fractions synchronous fluorescence spectra from natural and artificial maturation series. (Tables V and VI). However, this parameter does not allow bulk information about the alkyl groups content to be obtained but facilitates the differentiation between methyl groups and other alkyl substituents attached to aromatic rings. Condensation of Aromatic Rings. Information on the condensation of aromatic rings in coal extracts should permit determination of a suitable analytical procedure for studying the size, the shape, and the composition of the aromatic units. Correlations between aromatic
Aromatic Hydrocarbons from Coals
structure and spectral characteristics have already been deduced from classical fluorescence data35 and several simple rules have been experimentally determined. Baudot et aleMhave found that these rules remain valid for synchronous fluorescence. Selected UV SFS spectra illustrate structural changes undergone by natural and artificial aromatic fractions (Figure 8). Natural and artif~cialmaturation increase from top to bottom. The UV spectrum of the immature coal (32 362) extends from 290 to 340 nm corresponding to oneand two-ring aromatics (Table IV). It exhibita five major peaks (290,307,323,343, and 362 nm) respectively corresponding to the following families of aromatic compounds: biphenyl, fluorene, naphthalene, benzofluorene, and ~ h e n a n t h r e n e .No ~ ~signal between 370 and 440 nm has been recorded. During maturation, three different stages can be noticed according to the spectral characteristics of the SFS spectra: 1. The first stage is mainly characterized by an increase of the UV band located between 370 and 440 nm. This band is usually attributed to six kata-condensed and five pericondensed aromatic rings compounds (Table IV). However, data discussed in the previous paragraphs have shown that, during this first stage, the aromatization process is still limited and the degree of substitution of the aromatic rings reaches a maximum for both maturation series. More probably, this band could be related to highly substituted one- or two-ring aromatics. As a matter of fact, it has been demonstrated that a high degree of substitution and long-chain substituents induce Stokes shifta of the W bands toward higher wavelength^.^^ Aromatic compounds such as lO,l5-dihydro-5H-diinden[ 1-2-a:1',2'-c]-olabsorb in this wavelength range.36 Furthermore, substitution effects are more significant for small molecules than for larger polyaromatic molecules.34 2. During the second stage (86.2% and 80.9% of COC for artificial and natural maturation, respectively), absorptions in the 370-440-nm spectral region are no more represented in the UV spectra. This is mainly related to the important decrease of the average chain length of the substituents noticed during this stage (Figure 4). The spectra are dominated by families of one-, two-, and three-ring aromatic compounds substituted by shorter alkyl chains. 3. The predominance of condensed aromatic rings occurs above 86.2% and 80.9% of COC for artificial and natural maturation, respectively. It can be noticed that the W band between 370 and 440 nm now corresponding to the six kata-condensed and five peri-condensed aromatic rings significantly increases during this phase. Furthermore, the band located between 290 and 325 nm progressively diminishes, thus indicating the condensation of aromatic rings. Information deduced from UV SFS spectra can also be related to 'H NMR data. The angular aromatic ring protons/total aromatic protons ratio exemplifies the condensation phenomenon as far as it drastically increases during the first phase and remains stable in the last stage.
Energy & Fuels, Vol. 6, No. 2, 1992 171
The study of the aromatic hydrocarbons fraction extracted from natural and artificial series of maturation allows three main phases to be distinguished. These phases can be related to the major changes previously
noticed on the evolution of the CHCl:, extract and solid residue compositions of the same Mahakam ~ a m p l e s . ~ ~ 1 ~ Phase 1. The variations of the average aliphatic chain length as well as the aromatization rate are limited. On the other hand, an important increase in the degree of substitution is observed. This phase corresponds to the diagenetic stage and the first part of the catagenetic stage at the end of which the maximum production of hydrocarbons is reached.14 This phase is characterized by the genesis of highly substituted aromatic hydrocarbons by primary cracking of the coal structure. Phase 2. The average length of the alkyl chains substituted on aromatic rings drastically decreases. Similarly, the aromaticity of the aromatic hydrocarbons increases and the degree of substitution decreases. This phase corresponds to the secondary cracking of the hydrocarbons generated during phase 1. The amount of extractable hydrocarbons decreases and low molecular weight hydrocarbons are formed by progreasive removal of the paraffiic to compounds and alkyl s ~ b s t i t u e n t s . ' It ~ ~is~interesting ~ note that the aromaticity increase of the aromatic hydrocarbons corresponds to the increase of the 13CNMR aromaticity factor of the solid residue.38 This suggests that the main aromatization reactions occur simultaneously in the hydrocarbon fraction and the coal. Phase 3. Even if few samples are available to characterize this third phase, it can be interpreted as a major condensation stage correspondingto the dry gas (methane) formation zone. These three phases can be observed in both natural and artificial series which follow similar general trends. The ring substitution pattern as well as the evolution during maturation of the average aliphatic chain length is similar for both series (Figures 2-4). However, several discrepancies have been noticed between the natural and artificial maturation trends. They concern the rate of aromatization and desubstitution as well as the intensity of ring condensation. The beginning of the aromatization and condensation phase occurs earlier in the natural series (Figures 5 and 8). Similarly, the maximum of ring substitution is slightly shifted toward lower COC values in natural maturation (Figure 7). Such discrepancies could be related to the inadequacy of the COC percentage to be a common maturity parameter for natural and artificial series. As a matter of fact, phases 1, 2, and 3 are noticed for both series but in different ranges of COC. Nevertheless, previous studies12J4have demonstrated that COC exhibited linear correlations with depth of burial and with pyrolysis temperature for natural and artificial series respectively and that no slope variation in these correlations can be evidenced. On the other hand, Monthioux and Landais16 have shown that CHC1, extraction rate as well as the aliphatic CH pattern of infrared spectra in artificially matured coals were significantly different from that of the natural reference series. This has been attributed to the presence of free but trapped hydrocarbons in the structure of natural coals which cannot be extracted by organic solvents such as CHC13. However, the evolutions of the composition of the total extract (% saturates, % aromatics, % polars) as well as the distribution of total alkanes are similar for both artificial and natural series.'* This indicates that the portion of hydrocarbons extracted from naturally matured samples is representative of the whole hydrocarbons pro-
(35) Guibaut, G. C. Practical Fluorescence: Theory, Methods and Techniques, Marcel Dekker: New York, 1973. (36) Baudot, Ph.; Viriot, M. L.; Andr6, J. C.; Jezequel, 3. Y.; Lafontaine, M. Analusis 1991, 19, 85-97.
(37) Monthioux, M.;Landais, P.; Monin, J. C. Org. Geochem. 1985,8, 275-292. (38) Landais, P.; Dereppe, J. M.; Monthioux,M. CJI. Acad. Sci. Park 1988,306,11, 1093-1097.
Discussion
172
Energy & Fuels 1992,6,172-175
duced, including those which are still trapped in the macromolecular structure of the coal. Furthermore, it has been shown39 that successive chloroformic extractions could increase the extract yield of natural coals by a factor 3 without modifying ita composition. More detailed investigations have revealed that the isomerization rate of hopanoids was higher in natural serie~.~OSuch retarding effects have also been shown by Eglinton et alaa1They have been generally attributed to thermodynamic effects related to the high temperatures or fast heating rates used in laboratory simulation^.'^ On the other hand, it has been shown that the isomerization rate of free hydrocarbons, even trapped in the coal structure, was higher than the isomerization rate of hydrocarbons still attached to the organic macrom~lecule.~~ Then, the hopanoids generated during the first stage of catagenesis and trapped inside the kerogen will undergo a faster isomerization than those still linked to the coal structure. During the main phase of oil production, these trapped hydrocarbons will be mixed with hydrocarbons generated from the kerogen and the mixture will display an higher isomerization degree than expected. Then, the trapping of hydrocarbons into the macromolecular structure of naturally matured coals induces a delay between their formation and their expulsion43and can be responsible for the enhancement of several chemical reactions which will increase their maturity compared to that of the insoluble residue. Such mechanisms can be inferred in order to explain the differences in the aromatization, desubstitution and condensation rates of aromatic hydrocarbons observed during (39) Monthioux, M. Maturation naturelle et artificielle d'une drie de charbons homoghes. Ph.D., Orleans University, 1986. (40) Monthioux, M.; Landais, P. P. Chem. Geol. 1989, 75, 209-226. (41) Eglinton, T.I.; Rowland, S.J.; Curtis, C. D.; Douglas, A. G. Org. Geochem. 1986,10, 1041-1052. (42) Rowland, S.J.; Aareskjold, K.; Xuemin, G.; Douglas, A. G. Org. Geochem. 1985,10, 1033-1040.
the catagenetic phase. As a matter of fact, naturally and artificially p r o d u d aromatics behave similarly during the diagenetic phase (up to 81% COC) whereas, during catagenesis, similar mechanisms are noticed but are retarded in the artificial maturation series. This specific problem has not been encountered when pyrolyzing type 11kerogens which do not display the so-called two-phases structure of type I11 ~0als.4~
Conclusions Such multidisciplinary spectroscopic analysis facilitates the acquisition of structural data on extractable fractions of coals in order to get a better understanding of the mechanisms of hydrocarbons formation. The crosschecking of the data deduced from the different techniques allows the evolution of the structural parameters to be more accurately interpreted. Especially, the interpretation of the UV SFS spectra can be considerably improved by the knowledge of the ring substitution pattern and of the length of the substituents.% It also provides additional data which allow discrepancies between natural and artificial maturation to be evidenced. Most of these differences are related to the peculiar structure of coals which creates a cage effect and induces the trapping of hydrocarbons. Higher temperatures used in confined pyrolysis experiments provoke the slackening of the coal structure and an easier expulsion of the generated hydrocarbons. Such a process can be responsible for the different rates of aromatization and condensation noticed for natural and artificial maturation series. Acknowledgment. This work has been supported by the COPREP (No. G/10011/90) and by the INSU (No. 91 ATP 645). We gratefully acknowledge C. Moreaux and P. Doumenq for their technical and scientific assistance. (43) Behar, F.;Vanderbroucke,M. Org. Geochem. 1987,13,927-938.
N20Emissions from a Fluidized Bed Catalytic Cracker D.A. Cooper* and A. Emanuelsson Swedish Environmental Research Institute, Box 47086, S-402 58, Gothenburg, Sweden Received September 5, 1991. Revised Manuscript Received December 27, 1991
Measurements of primarily N20 have been carried out at a fluidized bed catalytic cracker employing a zeolite catalyst to degrade desulfurized vacuum-distilled gas oil. In view of the relatively low combustion temperatures (ca. 700 "C) and high NO, levels (340-520 ppm) in the catalyst regenerator section, the N 2 0 levels were surprisingly low, 3-25 ppm. In addition, only trace amounts of hydrocarbons and sulfur-containing species were measured.
Introduction The average atmospheric nitrous oxide concentrations (N20)concentrations are presently at ca. 300 ppb and have been increasing at a rate of 0.1&0,3% per year.12 Since (1) Rasmussen, R. A,; Khalil, M. A. K. Tellus 1983, 35B, 161-169.
0887-0624/92/2506-0172$03.00/0
N 2 0 has an important role in causing global climatic changes, significant worldwide concern has arisen. In the trOPosPhere, N20, being a relatively strong absorber of infrared radiation and relatively long-lived (ca. 150 year (2) Weiss, R. F.J . Geophys. Res. 1981,86,7185-7195.
0 1992 American Chemical Society