In Situ Variable-Temperature High-Resolution 1H NMR Studies of

856

Energy & Fuels 1997, 11, 856-865

In Situ Variable-Temperature High-Resolution 1H NMR Studies of Molecular Dynamics and Structure in Coal Jincheng Xiong and Gary E. Maciel* Department of Chemistry, Colorado State University, Fort Collins, Colorado 80526 Received December 10, 1996. Revised Manuscript Received April 1, 1997X

We have carried out the first systematic in situ variable-temperature high-resolution 1H NMR study of coal samples between 25 and 230 °C based on the CRAMPS (combined rotation and multiple-pulse spectroscopy) technique. The coal samples studied include two high-volatile and one low-volatile bituminous coals from the Argonne premium coal bank. We found unexpectedly that there are no dramatic changes of 1H CRAMPS resolution over this temperature range. Slight deterioration of resolution at high temperature suggests interference of thermally promoted molecular motion with coherent averaging by the multiple-pulse sequence. Based on a “timesuspension” experiment using the CMG-48 pulse sequence and proton dipolar-dephasing experiments, we estimated the correlation time of the thermally promoted molecular motion as being about 10 µs, which is several orders of magnitude slower than the molecular motion induced with pyridine saturation of coal at room temperature. This result suggests that thermal treatment alone up to 230 °C is not enough to break either the covalent bonds or the noncovalent associative bonds that connect the macromolecular network of coal. With improved dipolar-dephasing experiments based on BR-24 1H CRAMPS detection, we were able to analyze quantitatively the proton dipolar-dephasing behaviors of aliphatic and aromatic protons in coal. We found that the dipolar-dephasing curve of a coal at high temperature can be described as two Gaussian dephasing components, which indicate a coexistence of molecules with two distinctly different mobilities. The promotion of molecular mobility in coal at higher temperature is attributed to an increase of the fraction of molecules in relatively mobile states. This 1H CRAMPS study provides detailed correlations of molecular dynamics with molecular structure for coals at temperatures between 25 and 250 °C.

Introduction Thermally induced changes in coals are of interest from the standpoints of both fundamental and applied research. The study of structural and dynamic changes during thermal activation could shed some light on the nature of the complex structures of coals. Such knowledge could be used to refine current structural models of coals,1-8 which should be useful for improving prospects and efficiencies of coal conversion.9,10 The thermal transformation of coal can be studied by the “equilibrium method” and by the in situ method. In an “equilibrium method”, a coal sample is heated to a certain temperature for a certain period of time, then * To whom correspondence concerning this paper should be addressed. Telephone: (970) 491-6480. Fax: (970) 491-1801. E-mail: [email protected]. X Abstract published in Advance ACS Abstracts, May 1, 1997. (1) Davidson, R. M. Molecular Structure of Coal; International Energy Agency Coal Research: London, 1980. (2) Given, P. H. In Coal Science; Gorbaty, M. L., Larsen J. W., Wender I., Eds.; Academic Press: New York, 1984. (3) Green, T.; Kovac, J.; Brenner, D.; Larsen, J. W. In Coal Structure; Myers, R., Ed.; Academic Press: New York, 1982. (4) Shin, J. H. Fuel 1984, 63, 1187. (5) Given, P. H.; Marzec, A.; Barton, W. A.; Lynch, L. J.; Gerstein, B. C. Fuel 1986, 65, 155. (6) Given, P. H.; Marzec, A. Fuel 1988, 67, 242. (7) Lucht, L. M.; Larsen, J. M.; Peppas, N. A. Energy Fuels 1987, 1, 56. (8) Larsen, J. W. Prepr. Pap.sAm. Chem. Soc., Div. Fuel Chem. 1988, 33, 400. (9) Larsen, J. W. Prepr. Pap.sAm. Chem. Soc., Div. Fuel Chem. 1990, 35, 376. (10) Song, Ch.; Schobert, H. H. Pepr. Pap.sAm. Chem. Soc., Div. Fuel Chem. 1992, 37, 524.

S0887-0624(96)00219-8 CCC: $14.00

quickly cooled to room temperature for analysis. Such a method is useful for determining the permanent structural changes that occur in coals and requires no special high-temperature capability in an analytical instrument. However, an in situ method of observation is required to measure transient, nonequilibrium intermediate states during the thermal transformation and to elucidate thermal motions at the higher temperature. To monitor the dynamical changes in coal with temperature, an in situ method must be used.12 Many thermal analysis techniques have been used for the in situ study of coals. Some of the techniques are based on macroscopic thermodynamic properties, such as heat flow (differential scanning calorimetry and differential thermal analysis), mass (thermogravimetric analysis), and dimension (thermomechanical analysis). Other techniques include the analysis of evolved products using evolved gas analysis (EGA), combined gas thromatography-mass spectrometry (GTMS), and emission and adsorption Fourier transformation infrared spectroscopy (FTIR). Of the various analytical methods available, NMR spectroscopy has proved to be of special importance in coal research.13-15 NMR provides both structural and dynamic information on coal at the molecular level. In (11) Sanada, Y.; Lynch, L. J. In Magnetic Resonance of Carbonaceous Solids; Botto, R. E., Sanada, Y., Eds.; Advances in Chemistry 229; American Chemistry Society, Washington, DC, 1993; p 139. (12) Sakurovs, R.; Lynch, L. J.; Barton, W. A. In Magnetic Resonance of Carbonaceous Solids; Botto, R. E., Sanada, Y., Eds.; Advances in Chemistry 229; American Chemistry Society, Washington, DC, 1993; p 229.

© 1997 American Chemical Society

1H

NMR of Coal

situ high-temperature 1H NMR techniques were used to study thermal transformation of coals by Sanada et al.11 and by Barton and co-workers.12 The proton density and 1H NMR line width were monitored at temperatures between 200 and 450 °C in these published studies. Changes in the line width with temperature were attributed to the thermoplastic properties of coals and were correlated with other coal properties, such as maximum vitrinite reflectance and Hardgrove grindability index (HGI). All these previous studies were carried out at low magnetic field strength (38 and 20 MHz 1H Larmor frequency) using wide-line 1H NMR techniques. Thus, the potentially valuable chemical shift information of NMR was lost in these studies because of line-broadening caused by strong protonproton dipolar interactions. A fundamental question that arose from such studies is whether the multipledephasing components observed are due to different mobilities of the same kind of structural unit or are simply contributions from different structural units with their own characteristic molecular mobilities. An equally important practical question is what kind of structural units in coal can be effectively mobilized and are thus selectively extracted in solvent extraction experiments. These questions can be addressed by a high-resolution solid-state 1H NMR technique called CRAMPS (combined rotation and multiple-pulse spectroscopy), which was introduced by Gerstein and co-workers.16-18 Chemical shift information can be extracted from this technique by means of multiple-pulse averaging of strong proton dipolar interactions in combination with averaging of chemical shift anisotropy (CSA) by magic angle spinning (MAS).16-18 The power and potential of this technique for coal science have been demonstrated in extensive 1H CRAMPS studies of coal samples carried out at room temperature by this research group.19-24 In situ high-temperature 1H CRAMPS techniques, combined with various time-domain experiments, should provide detailed structural and dynamic information on the thermal transformation of coal. This potential is demonstrated in this work, which is the first systematic in situ high-temperature 1H CRAMPS study of coal. One of the most important advances in coal science in the past decade has been the realization that noncovalent associative bonding forces make major contri(13) Maciel, G. E.; Erbatur, O. In Nuclear Magnetic Resonance in Modern Technology; Maciel, G. E., Ed.; NATO ASI Series; Kluwer: Dordrecht, Netherland, 1994; pp 165-224. (14) Botto, R. E., Sanada, Y., Eds. Magnetic Resonance of Carbonaceous Solids; Advances in Chemistry 229; American Chemical Society: Washington, DC, 1993. (15) Gorbaty, M. L. Fuel 1994, 73, 1819. (16) Schnabel, B.; Haubenreisser, U.; Scheler, G.; Muller, R. Proc. Congr. Ampere, 19th 1976, 441. (17) Gerstein, B. C.; Pembleton, R. G.; Wilson, R. C. J. Chem. Phys. 1977, 66, 361. (18) Gerstein, B. C.; Chou, C.; Pembleton, R. G.; Wilson, R. C. J. Phys. Chem. 1977, 66, 361. (19) Frye, J. S.; Bronnimann, C. E.; Maciel, G. E. In NMR of Humic Substances and Coal; Wershaw, R. L., Mikita, M. A., Eds.; Lewis Publishers: Chelsea, MI, 1987; p 33. (20) Bronnimann, C. E.; Maciel, G. E. Org. Geochem. 1989, 14, 189. (21) Davis, M. F.; Quinting, G. R.; Bronimann, C. E.; Maciel, G. E. Fuel 1989, 68, 763. (22) Jurkiewicz, A.; Bronnimann, C. E.; Maciel, G. E. Fuel 1989, 68, 872. (23) Jurkiewicz, A.; Bronnimann, C. E.; Maciel, G. E. Fuel 1990, 69, 804. (24) Jurkiewicz, A.; Bronnimann, C. E.; Maciel, G. E. In Magnetic Resonance of Carbonaceous Solids; Botto, R. E., Sanada, Y., Eds.; Advances in Chemistry 229; American Chemical Society: Washington, DC, 1993; p 401.

Energy & Fuels, Vol. 11, No. 4, 1997 857 Table 1. Elemental Analysis (wt %)a of Argonne Premium Coals Used in This Work name

no.

rank

VMb

C

H

O

N

Sorg

Illinois No. 6 Pocahontas Blind Canyon

301 501 601

HVB LVB HVB

45.7 19.0 47.8

80.7 91.8 81.3

5.2 4.5 5.4

10.1 1.7 10.9

1.4 1.3 1.6

2.4 0.5 0.4

a Weight percentage from refs 28 and 29: C, H, O, N, dmmf (dry mineral matter free basis); Sorg db (dry basis). b Volatile matter in weight percentage.

butions to stabilizing the structure of coal.15 This finding opens the possibility for future coal conversion technology under mild conditions rather than thermolysis. A major hole in previous studies has been a fundamental understanding of temperature effects on noncovalent associative bonds at the molecular level. Since covalent bonds in coal are largely unchanged under mild thermal treatment (i.e., below 250 °C), the NMR study of coal structure and dynamics between 25 and 250 °C should be especially of interest for revealing the nature of noncovalent associative bonds. In this paper, we present experimental results of in situ variable-temperature 1H CRAMPS studies on three Argonne premium coals at temperatures between 25 and 230 °C. In this temperature range, structural changes in coals will be mainly associated with changes in noncovalent associative bonding forces. An understanding of coal over this temperature range is also important to allow us to refine or modify current structural models of coal, such as the molecular/macromolecular model.1-8 Thermal activation below 250 °C should dramatically change the properties of the molecular component of coal but leave the three-dimensional macromolecular framework largely unchanged. Dipolar-dephasing experiments based on 1H CRAMPS detection24,26,27 have been successfully used previously to correlate molecular dynamics with structural features in coal at room temperature.22-24 However, the pulse sequence used in the previous work has also suffered from phase and base line distortions that make quantitation of 1H CRAMPS spectra of samples with relatively broad lines (such as coals) very difficult. In this work, a new pulse sequence was designed and proved to be much better than the original sequence for suppressing phase and base line distortions of 1H CRAMPS spectra of coal. With this new dipolardephasing pulse sequence, we were able to analyze quantitatively the proton dipolar-dephasing behavior of aliphatic and aromatic protons in coal from 25 °C to 230 °C. Experimental Section 1. Sample Preparation. 1H CRAMPS experiments were performed on the following three coal samples obtained from the Premium coal bank of the Argonne National Laboratory: Illinois No. 6 (Premium Coal 301, a high volatile bituminous (HVB) coal), Utah Blind Canyon (premium coal 601, a highvolatile bituminous coal), and Pocahontas No. 3 coal (premium coal 501, a low-volatile bituminous (LVB) coal). The names and the elemental analysis of the three Argonne Premium coals used in this work are listed in Table 1. (25) Berkowitz, N. An Introduction to Coal Technology; Academic Press: San Diego, 1994. (26) Bronnimann, C. E.; Zeigler, K. C.; Maciel, G. E. J. Am. Chem. Soc. 1988, 110, 2023. (27) Ridenour, C. F.; Xiong, J.; Maciel, G. E. Biophys. J. 1996, 70, 511.

858 Energy & Fuels, Vol. 11, No. 4, 1997

Figure 1. Glass MAS rotor for a sealed sample: (a) exploded view of the rotor; (b) completely assembled rotor with epoxy resin seal. The rotor is made from a thick-wall Pyrex NMR tube (Wilmad). The inner diameter of the rotor is 2.16 mm; the outer diameter is 5.0 mm. To reduce the effect of residual water on the 1H CRAMPS spectra of coals, the coal samples were evacuated to about 10-3 Torr for approximately 24 h at room temperature and stored in a glovebox under a N2 atmosphere. Prior to a 1H CRAMPS experiment, each coal sample was packed in a homemade Pyrex glass spinner designed for the Chemagnetics 5 mm Pencil type magic angle spinning (MAS) module. The glass spinner is made from a thick-wall Pyrex NMR tube (Wilmad). As shown in Figure 1, the glass tube is sealed in the middle with a torch, while both ends of the glass tube are still open. Such a glass spinner is specifically designed for air-sensitive samples and/or samples that need to be sealed so that volatile components are kept in the NMR tube for analysis. The open end of the spinner on the sample-loading side can be easily “sealed” with a Teflon rod that makes a snug fit in the sample tube. If an absolutely airtight seal is required, a glass rod with a diameter slightly smaller than the internal diameter of the spinner can be inserted into the spinner to confine a sample to its position; the spinner can then be sealed either with epoxy resin or with a torch. Epoxy is much easier to use, and its use makes it easier to get a symmetrical seal for magic angle spinning; the epoxy seal does not produce detectable background signal in a 1H CRAMPS experiment, since it is outside the probe coil. 2. BR-24 CRAMPS Experiments. All 1H NMR experiments were performed on a severely modified NT-200 NMR instrument operating at a proton resonance frequency of 187 MHz.30 Modification and redesign on both hardware and software have been made in this work to enhance the performance and stability of the spectrometer. Details of the experimental setup are described elsewhere.31 A Chemagnetics variable-temperature (VT) CRAMPS probe was used in this work. The sample temperature was calibrated on the basis of the 1H chemical shifts of ethylene glycol32 and melting points of known crystalline compounds.31 We estimate the errors of the sample temperatures measured in this work to be within (2 °C. The BR-24 pulse sequence was used in most of the 1H CRAMPS experiments;33 the use of other multiple-pulse sequences will be specifically indicated below, where appropriate. For experiments based on BR-24 detection, a 108 µs cycle time and a 90° pulse width of 1.2-1.3 µs were used. The number of BR-24 cycles used in the experiments was between 128 and 512, depending on the required resolution of the CRAMPS spectra. The recycle delays were 3-10 s. The sample spinning speeds were between 1.4 and 2.0 kHz. In CRAMPS experiments, the probe, transmitter amplifiers, and the matching network between the amplifier and probe (28) Vorres, K. S. Prepr. Pap.sAm. Chem. Soc., Div. Fuel Chem. 1988, 33, 3. (29) Vorres, K. S. Energy Fuels 1990, 4, 420. (30) Maciel, G. E.; Bronnimann, C. E.; Hawkins, B. L. In Advances in Magnetic Resonance: The Waugh Symposium; Warren, W. S., Ed.; Academic: San Diego, 1990; Vol. 14, p 125. (31) Xiong, J. Ph.D. Dissertation, Colorado State University, Fort Collins, CO, 1996. (32) Van Geet, A. L. Anal. Chem. 1970, 42, 679. (33) Burum, D. P.; Rhim, W.-K. J. Chem. Phys. 1979, 71, 944.

Xiong and Maciel

Figure 2. Diagram for the pulse sequence of the dipolardephasing experiment based on CRAMPS detection. The dephasing period is 2τ. R ) 35.3° and β ) 45°. are all carefully tuned for best performance. However, the impedance of electronic components in a probe will change with a change in the temperature in a VT experiment. Maintaining the best tuning condition for CRAMPS is critical for VT CRAMPS experiments. In our experience, even just a change of a sample in the probe at the same temperature could result in a significant impedance change in the probe, which may cause the CRAMPS performance to deteriorate. We used a rf sweeper to check and retune the impedance of the probe each time we changed a sample or sample temperature. This method is convenient and works very well. 3. Dipolar-Dephasing Experiments Based on the BR24 CRAMPS Experiments. The 1H-1H dipolar-dephasing experiment is very useful for exploring structural-dynamical details of coals.22-24 In this experiment (Figure 2), a dephasing period (typically several microseconds to a few milliseconds) is inserted before the multiple-pulse sequence. To minimize dephasing effects due to inhomogeneous interactions such as isotropic chemical shifts, a π pulse is inserted into the middle of the dephasing period. During this dephasing period, the magnetization of those protons experiencing strong dipolar coupling becomes dephased and attenuated; the signals of such protons are correspondingly attenuated in the resulting CRAMPS spectrum. Signals from protons that do not participate in strong dipolar couplings are not strongly attenuated during the dephasing period and are relatively more prominent in the observed spectrum. Rapid motion of molecules or portions of molecules can reduce the effects of dipolar interactions and therefore lead to the persistence of some signals in a dipolar-dephasing experiment. A significant improvement of the dipolar-dephasing pulse sequence made in this work is the introduction of a composite pulse, (35.3°)-x(45°)-y , right before the BR-24 detection. Since 1H CRAMPS spectra of coals are still quite broad, even with MAS and multiple-pulse narrowing, the accuracy of quantitation depends very critically on the quality of 1H CRAMPS spectra, especially the flatness of the base line and the absence of interfering rotor lines.21 To avoid base line distortions, well-calibrated composite preparation pulses (vide supra) before the multiple-pulse cycles were designed and used in this work so that proton magnetization that is spin-locked along the effective field of the average Hamiltonian is minimized. Without the composite pulse, the quantitation of CRAMPS spectra of coals is very difficult. The receiver phase was also optimized to minimize the base line distortions due to the “spin-locked” signal. Details on this issue are discussed elsewhere.31 4. CMG-48 CRAMPS Experiments. CMG-48 is one of the most efficient pulse sequences for time-suspension experiments in which both homogeneous and inhomogeneous broadening are refocused.35,36 The time evolution of a spin system in a time-suspension experiment should ideally be suspended if the pulse sequence is perfect and there is no random modulation of the system. Such experiments have been mainly (34) Haeberlen, U. High-Resolution NMR in Solids, Selective Averaging; Academic Press: New York, 1976. (35) Cory, D. G.; Miller, J. B.; Garroway, A. N. J. Magn. Reson. 1990, 90, 205. (36) Butler, L. G.; Cory, D. G.; Dooley, K. M.; Miller, J. B.; Garroway, A. N. J. Am. Chem. Soc. 1992, 114, 125.

1H

NMR of Coal

Energy & Fuels, Vol. 11, No. 4, 1997 859

used for solid-state NMR imaging of nuclei with strong homonuclear dipolar couplings (e.g., 1H, 19F).35-37 In this work, we have, for the first time, exploited the potential of CMG-48 in applications other than imaging. In particular, we have used this pulse sequence in VT CRAMPS experiments to investigate line broadening contributed from random motion, or incoherent sources. A 90° pulse width of 1.3 µs and a cycle time of 216 µs were used for CMG-48. Since the frequency of the NMR signal from a CMG-48 experiment is close to the rf carrier frequency, quadrature detection is preferred. Since we had only a singlechannel phase detector for the 1H NMR signal, we designed and implemented a pseudoquadrature detection scheme to acquire the quadrature data set for CMG-48 experiments. In this scheme, the preparation pulses of CMG-48 multiple pulse cycles are put through a CYCLOPS phase cycling, just as in a normal quadrature detection experiment;38 the data acquired with the phases of the preparation pulses in quadrature are stored and accumulated in different locations of computer memory. The acquired time-domain data are the same as obtained using a normal quadrature detection method, except that there is no signal-to-noise ratio gain of x2 for this detection scheme. There are also some special hardware requirements for implementing this experiment. The technical details are described elsewhere.31

Results 1. Variable Temperature 1H CRAMPS Experiments. Figure 3 shows 1H CRAMPS spectra of three Argonne premium coals at room temperature. One general characteristic of these spectra is that they each consist of two broad contributions. The band that is centered at about 1-3 ppm is associated with protons attached to aliphatic carbons, and the band centered around 6-8 ppm is due to protons attached to aromatic carbons. The two bands can be separated by a homewritten computer deconvolution program, as shown in Figure 3. From the deconvolution, it is possible to obtain quantitative information on the populations of aliphatic and aromatic protons. The apparent resolution of 1H CRAMPS spectra of typical coals is much lower than the resolution of typical rigid crystalline solids. The main line-broadening mechanism is presumably due to anisotropic bulk susceptibility effects and the dispersion of isotropic chemical shifts due to the complex chemical structures of coal. We have demonstrated this point experimentally with a so-called “time-suspension” experiment and a twodimensional spin-exchange experiment.31,39 1H CRAMPS experiments on the three Argonne premium coals 301, 501, and 601 were carried out in situ at temperatures from 25 to 230 °C. The spectra are shown in Figures 4-6, respectively. Such spectra show that there are no dramatic changes over the 25230 °C temperature range, certainly no expected line narrowing at higher temperature. 2. Variable-Temperature Dipolar-Dephasing 1H CRAMPS Experiments. As we have just seen, varing the temperature up to 230 °C did not result in a dramatic change in 1H CRAMPS resolution for coal. The results are not what we had expected on the basis of previous studies from this laboratory in which dramatic (37) Sun, Y.; Xiong, J.; Lock, H.; Buszko, M.; Hasse, J.; Maciel, G. E. J. Magn. Reson., Ser. A 1994, 110, 1. (38) Hoult, D. I.; Richards, R. E. Proc. R. Soc. London 1975, A344, 311. (39) Xiong, J.; Maciel, G. E. Manuscript in preparation.

Figure 3. 1H CRAMPS spectra of Premium Coals (a) 301, (c) 501, and (e) 601 obtained at 25 °C. The corresponding spectral deconvolutions are shown in (b), (d), and (f). The BR-24 pulse sequence was used with a cycle time of 108 µs and a 90° pulse width of 1.2-1.3 µs. Each spectrum was acquired with 200 scans and a recycle delay of 5 s. The MAS rates were between 1.5 and 1.8 kHz.

line narrowing was observed when coal was saturated with pyridine.22-24 To examine explicitly the dependence of molecular motion on temperature, we carried out proton dipolar-dephasing experiments at temperatures of 25, 120, and 180 °C. Figures 7-9 illustrate the variable-temperature dipolar-dephasing spectra obtained on three Argonne premium coals. We can see very clearly that the dephasing rates were reduced significantly for all three coals when the sample temperatures were increased. This means, as originally anticipated, that thermal treatment indeed promotes internal motion in coal. The increased mobility of molecules in coals at high temperature reduces the effective dipolar coupling, thus prolonging the dephasing time constants. However, the promoted motion does not lead to CRAMPS line narrowing. This suggests that thermal treatment may have an effect on the internal molecular motion of coal that is far different from that produced by solvent saturation. To better quantitate the results on dipolar-dephasing dynamics of coals at high temperature, we have developed C computer programs that can automatically and/or interactively deconvolute NMR spectra, and fit the deconvoluted peak intensities to various dipolardephasing models . A detailed discussion of the programs is included in ref 31. After deconvoluting the variable-temperature (VT) 1H CRAMPS spectra of coals into aliphatic and aromatic peaks, we were able to construct the dipolar-dephasing curves for protons in aliphatic and aromatic moieties, respectively.


Figure 4. 1H CRAMPS spectra of Premium Coal 301 at the temperatures indicated. The BR-24 pulse sequence with a 90° pulse width of 1.2 µs and a cycle time of 108 µs was used to acquire 256 data points, sampled once per cycle. The number of scans was 100. The recycle delays were 5 s. The MAS rates were between 1.6 and 1.8 kHz.

Xiong and Maciel


time t, and Mf and Ms represent the initial intensities of fast and slow dephasing components, respectively. The dipolar-dephasing rates are characterized with dephasing time constants Tf and Ts. Figure 10 shows, as an example, experimental dipolar-dephasing curves from Premium Coal 301 at 180 °C and simulated curves corresponding to the best fit to eq 1. The dipolar-dephasing parameters obtained from the model described above are summarized in Tables 2-4 for the three coal samples, Illinois No. 6 (Premium Coal 301), Pocahontas No. 3 (Premium Coal 501), and Blind Canyon (Premium Coal 601), respectively. Discussion


Each dipolar-dephasing curve can be simulated by a sum of two Gaussian decays, denoted fast and slow, according to the following equation:

M(t) ) Mf exp[-t2/(2Tf2)] + Ms exp[-t2/(2Ts2)] (1) where M(t) represents signal intensity at dephasing

1. Line Width and Molecular Motion. In previous studies by this research group, dramatic CRAMPS line narrowing of 1H CRAMPS spectra of coals was often achieved with pyridine saturation.22-24 Such line narrowing was not well understood. It was proposed to be due mainly to an enhanced mobility of a substantial number of structural moieties in coal. The enhanced motion will partially average different isotropic chemical shifts that would be “locked in” for an untreated coal. Based on these results and the molecular/macromolecular model of coal, one might have expected that molecular components of coal should be significantly mobilized with thermal activation at high temperature and that similar line narrowing should be observed at high temperature. Apparently, this is not the case. The line widths in a CRAMPS spectrum may have a complex dependence on molecular motion. It is wellknown that molecular motion with a correlation time close to the cycle time of a multiple-pulse sequence will interfere with the coherent averaging of the dipolar

1H

NMR of Coal


Figure 7. Stack plot of 1H CRAMPS spectra of Premium Coal 301 obtained in dipolar-dephasing experiments performed at (a) 25, (b) 120, and (c) 180 °C. A 90° pulse width of 1.25 µs and a cycle time of 108 µs were used to obtain the BR-24 spectra with 256 data points. The number of scans was 400. The recycle delay was 3 s. The MAS speed was 1.6 kHz.

Figure 8. Stack plot of 1H CRAMPS spectra of Premium Coal 501 obtained in dipolar-dephasing experiments performed at (a) 25, (b) 120, and (c) 180 °C. A 90° pulse width of 1.25 µs and a cycle time of 108 µs were used to obtain the BR-24 spectra with 256 data points. The number of scans was 500. The recycle delay was 3 s. The MAS speed was 1.6 kHz.

interactions40-42 and may lead to severe line broadening of a CRAMPS spectrum. A CRAMPS peak will start to broaden as the correlation time of motion approaches the cycle time of a multiple-pulse sequence. The line will get very broad when the correlation time matches the multiple-pulse cycle time and will gradually get narrower again as the correlation time is made much shorter than the multiple-pulse cycle time and the inverse of the strength of static dipolar interactions. In the limiting line-narrowing regime, random averaging due to motion dominates completely coherent averaging. In the regime in which line narrowing due to random isotropic motion is dominant, the line width can be estimated to be approximately 3∆2τc , where ∆ is the (40) Waugh, J. S. In Nuclear Magnetic Resonance in Modern Technology; Maciel, G. E., Ed.; NATO ASI Series; Kluwer: Dordrecht, Netherland, 1994, p 359. (41) Burum, D. P. Concepts Magn. Reson. 1990, 2, 213. (42) Kinney, D. R.; Chuang, I.-S.; Maciel, G. E. J. Am. Chem. Soc. 1993, 115, 6785.

line width due to static dipolar interactions and τc is the correlation time of the random motion.40 A typical unaveraged 1H NMR line width of coal is about 35 kHz at room temperature. The line width obtained on coal from CMG-48 in this work is about 30 Hz. To achieve the same line narrowing efficiency as CMG-48, random motion with a correlation time of about 0.01 µs is needed, i.e., to reduce the 1H NMR line width of coal from 35 kHz down to about 30 Hz. This estimate of the required correlation time is based on the assumption of isotropic random motion. In relatively rigid solids such as coal, isotropic random motion is not common; most of the random motion that occurs is expected to be anisotropic in nature. The line-narrowing efficiency from an anisotropic motion could be much lower than that estimated from the above equation. Random motion in coal could also interfere with the coherent averaging of chemical shift anisotropy by MAS in a CRAMPS experiment.40 The relationship between


Xiong and Maciel

Figure 9. Stack plot of 1H CRAMPS spectra of Premium Coal 601 obtained in dipolar-dephasing experiments performed at (a) 25, (b) 120, and (c) 180 °C. A 90° pulse width of 1.25 µs and a cycle time of 108 µs were used to obtain the BR-24 spectra with 256 data points. The number of scans used was 400. The recycle delay was 3 s. The MAS speed was 1.6 kHz. Table 2. Dipolar-Dephasing Data on Illinois No. 6 Coal (Premium Coal 301) fast Gaussian percentagea

Tf (µs)b

slow Gaussian percentagea

Ts (µs)c

aliphatic aromatic

96 90

25 °C 9.7 17

4 10

41 45

aliphatic aromatic

79 55

120 °C 9.2 9.2

21 45

56 51

aliphatic aromatic

76 52

180 °C 9.8 9.3

24 48

57 60

a Estimated standard error for percentage: (3%. b Estimated standard error for Tf: (0.7 µs. c Estimated standard error for Ts: (4 µs.

Table 3. Dipolar-Dephasing Data on Pocahontas No. 3 Coal (Premium Coal 501) fast Gaussian percentagea

Figure 10. Dipolar-dephasing data (amplitudes, open circles) on Premium Coal 301 obtained at 180 °C: (a) aliphatic protons; (b) aromatic protons. The data were fitted with eq 1. Dashed curves represent fast and slow dephasing components. The solid curve is the sum of the two dephasing components.

the line width and the random motion in this case is qualitatively similar to that for dipolar interactions. However, the time scale is different. The interference will be the most severe when the correlation time matches the inverse of the MAS speed, which was between 0.5 and 1 ms in this work. In summary, at least three types of averaging processes should be taken into consideration when one analyzes the dependence of 1H CRAMPS line width on random motion. The isotropic chemical shift dispersion in coal may be partially averaged by random motion, e.g., by averaging conformational differences. Random motion with a correlation time around 10-100 µs could interfere severely with the coherent averaging of dipolar

Tf (µs)b

aliphatic aromatic

100 100

25 °C 15 13

aliphatic aromatic

42 100

120 °C 10 20

aliphatic aromatic

33 75

180 °C 9.5 9.3


Ts (µs)c

58

52

67 25

48 54


interactions by BR-24. The coherent averaging of chemical shift anisotropy by MAS could be destroyed by random motion with a correlation time of about 1 ms. We can see from Figures 4-6 that the line widths of 1H CRAMPS spectra of the three coal samples tend to get larger at higher temperature. This result suggests that there is interference of random molecular motion with the coherent averaging of dipolar interactions and/ or chemical shift anisotropy at high temperatures. This

1H

NMR of Coal


Table 4. Dipolar-Dephasing Data on Blind Canyon Coal (Premium Coal 601) fast Gaussian percentagea

Tf (µs)b


Ts (µs)c

11

33

aliphatic aromatic

89 100

25 °C 10 16

aliphatic aromatic

69 40

120 °C 10 7.0

31 60

59 48

aliphatic aromatic

65 34

180 °C 13 8.5

35 66

62 44


Figure 11. CMG-48 CRAMPS spectra of Premium Coal 601 obtained at (a) 25, (b) 120, and (c) 180 °C. A 90° pulse width of 1.3 µs and a cycle time of 216 µs were used to obtain the CMG-48 spectra with 1024 data points sampled once per cycle. The number of scans was 16. The recycle delay was 3 s. The MAS speed was 1.6 kHz. The central glitches seen in (b) and (c) are due to ineffective “second averaging” around the carrier frequency.35,36

implies that the fraction of protons experiencing random molecular motion with a correlation time of about 101000 µs is increased as the temperature is increased. The relatively slow motion that is promoted at higher temperatures can also be detected via time-suspension experiments on coal. The CMG-48 spectra of Argonne Premium Coal 601 obtained at 25, 120, and 180 °C are shown in Figure 11. The full line width at half height (fwhh) obtained at 25 °C is only 30 Hz. It is increased to 60 and 68 Hz at 120 and 180 °C, respectively. The line width in a CMG-48 spectrum is mainly due to irreversible processes such as spin lattice relaxation and random motions. Although the theoretical treatment of relaxation under multiple-pulse irradiation is very complicated,43,44 the line width of CMG-48 should be determined by T1y, a decay time constant of magnetization in the toggling frame of the multiple pulses.43 The T1y value can be estimated from the value of ∆ν1/2 (fwhh) as 1/(π∆ν1/2) for a Lorentzian line shape, the shape that was observed for all three CMG-48 spectra shown in Figure 11. The 1H T1y of Premium Coal 601 at 25, 120, and 180 °C can then be estimated as 11, 5.3, and 4.7 ms, respectively. According to the theoretical analysis by Vega et al.43 and by Dybowski et al.,44 T1y should be close to the value of T1F in a spin-lock field of (γτ)-1, where τ is the length of the short window in the CMG(43) Vega, A. J.; Vaughan, R. W. J. Chem. Phys. 1978, 68, 1958. (44) Dybowski, C.; Pembleton, R. G. J. Chem. Phys. 1979, 70, 1962.

48 sequence. Since τ was set at 3.0 µs in this work, (γτ)-1 can be estimated as 12.5 G, or 53 kHz. This suggests that the line broadening at high temperature is a result of random motion with a correlation time of about 18 µs, estimated from a (γτ)-1 of 53 kHz. This result is consistent with the observed motional interference with coherent averaging by the multiple-pulse sequence (vide supra). From the observed line width changes of 1H CRAMPS spectra of coals at temperatures between 25 and 230 °C, we can conclude that slow random motions with correlation times ranging from 10 µs to 1 ms are promoted in coal at the higher temperatures. However, it is difficult to get much quantitative information on the molecular dynamics from just these experiments. 2. Dipolar-Dephasing Behavior of Coal at High Temperature. To examine explicitly the dependence of molecular motion on the temperature, we carried out various time-domain experiments based on CRAMPS detection at high temperature. Measurements of 1H1H dipolar-dephasing behavior, 1H rotating-frame spinlattice relaxation times (T1F), and 1H Zeeman spinlattice relaxation times (T1, inversion-recovery) were used to probe molecular dynamics and motion in coals over a very large range of time (frequency) scales, from several hertz to 187 MHz. Since CRAMPS detection is used in all the time-domain experiments, we can obtain dynamical information on different chemical structures in coal. Note that appropriate preparation pulses and/ or composite pulses were used prior to the multiplepulse detection cycles to avoid the base line distortion caused by magnetization spin-locked along the direction of the effective field of the average Hamiltonian of BR24. In this paper, we focus on results from dipolardephasing experiments. The results from spin-lattice relaxation measurements in the laboratory frame and in the rotating frame will be discussed in a subsequent paper.45 To understand better the 1H-1H dipolar-dephasing behaviors of the three coals of this study, distinction between coherent and incoherent processes should be clearly made. The secular part of the spin Hamiltonian is responsible for the coherent evolution of spin systems of a rigid solid. In principle, coherent evolution can be reversed or refocused by pulse sequences or rotations. An incoherent process, usually called relaxation, relies on both secular and nonsecular parts of Hamiltonians. The random modulation of local fields in an incoherent process makes the evolution of the spin system irreversible. Dipolar dephasing is a coherent process driven by the secular part of homonuclear dipolar Hamiltonians. For a pair of spin-1/2 nuclei, the evolution of magnetization under the dipolar interaction produces an oscillatory behavior.46,47 In general, oscillatory behavior can usually be seen in a coherent process in small systems with uniform couplings. However, in a complex system, such oscillations are washed out by destructive superposition of many coherent oscillations of various periods, or couplings due to many-body interactions. This produces a variety of dephasing behaviors over time. Gaussian (45) Xiong, J.; Maciel, G. E. Energy Fuel, in press. (46) Abragam, A. The Principles of Nuclear Magnetism; Clarendon Press: Oxford, 1961. (47) Schmidt-Rohr, K.; Spiess H. W. Multidimensional Solid-State NMR and Polymers; Academic Press: London, 1994.


Xiong and Maciel

decay is usually seen for dipolar dephasing of strongly coupled protons; Lorentzian (or exponential) decay is common for spins that experience weaker dipolar interactions.46 Although the pulse sequence used for measuring the dipolar-dephasing time constant is similar to the one for measuring the spin-spin relaxation time (T2), the real irreversible spin-spin relaxation is usually difficult to detect because of the much faster coherent dipolardephasing process. The dipolar-dephasing results show unambiguously that dephasing rates were reduced significantly for all three coals when the sample temperatures were increased, as a result of molecular motion promoted at higher temperatures. As seen in Tables 2-4, in most cases for both aliphatic and aromatic protons, the dipolar dephasing can be described by a fast-decay Gaussian component with a time constant around 10 µs and a slow-decay Gaussian component with a time constant of 40-60 µs. The fast-decay component corresponds to protons in more rigid moieties, and the slowdecay component corresponds to protons in more mobile moieties for which the mobility partially averages proton-proton dipolar interactions. The fact that the two components can be distinguished from each other implies that proton spin diffusion and/or spin exchange between the rigid and mobile domains of coal are not fast compared with the experimental time scale (10100 µs) of the dephasing periods. As the temperature is increased, the fraction of the fast-decay component decreases and the fraction of slowdecay component increases (Tables 2-4). However, the dipolar-dephasing time constants do not change much with temperature. This suggests that the energy barriers that restrict molecular motions in coals do not change much with temperature. The promotion of molecular mobility in coal at higher temperatures comes mainly from thermal activation of motion-restricted rigid molecules to more mobile states. The results imply that thermal treatment alone up to 230 °C is not enough to break effectively either the covalent bonds or the noncovalent associative bonds that contribute to the macromolecular network of coal. Although the dipolar-dephasing time constants of slow Gaussian components (40-60 µs) are much larger than those of fast-decay components (ca. 10 µs), the slow-decay components are still pretty rigid compared with some “soft” solids such as adamantane. The dipolar-dephasing time constant provides a direct measure of the 1H NMR line width of coal without multiplepulse decoupling. The correlation time τc of the random reorientational motion responsible for the reduced linewidth can thus be estimated. For isotropic random motion of many equivalent dipolar-coupled spins, the reduced line width can be quantitatively analyzed based on the spectral density function of random motion under MAS. In the regime where the NMR line is substantially narrowed from its rigid lattice value ∆, the dependence of 1/Tdd on the correlation time of random motion τc and magic angle spinning speed ωr can be expressed40 as follows:

expression can be simplified to

1 ) 3∆2τc Tdd

(3)

(2)

From eq 3, we see that a dipolar-dephasing time constant of 40-60 µs corresponds to a random motion with a correlation time of about 16 µs, assuming the line width of protons in a rigid lattice is 35 kHz. This is consistent with what we observed in terms of motional interference with line narrowing in CRAMPS. According to eq 3, the dipolar-dephasing time constant Tdd is inversely proportional to the correlation time τc for isotropic random motion. The dipolar-dephasing time constant is expected to increase with temperature. However, the slow-decay time constants of coal do not change much with temperature up to 180 °C. This suggests that the thermally promoted motion responsible for slow-decay components is anisotropic in nature, even at a temperature as high as 180 °C. Such anisotropic motion cannot average out the dipolar interactions completely. 3. Comparison of Dipolar-Dephasing Behaviors of Coals of Different Rank. For high-volatile bituminous (HVB) coals such as Illinois No. 6 and Blind Canyon coals, the fraction of aliphatic protons that are fast-dephasing components is much larger than for aromatic protons. For the low-volatile bituminous (LVB) Pocahontas No. 3 coal, the situation is reversed. The aliphatic components in HVB coals are generally much richer than in LVB coals. The dipolar-dephasing results summarized in Tables 2 and 4 suggest that a large fraction (65-75%) of the aliphatic structures in HVB coals are in rigid cross-linked macromolecular networks or are smaller molecules that are trapped in very tight voids. Those structures are difficult to mobilize even at 180 °C. In contrast, 66% of the aliphatic protons in the LVB coal are in a relatively mobile state at 180 °C. These results imply that most of the rigid aliphatic structures in HVB coals may be converted to aromatic structures during the coalification conversion process from HVB coals to LVB coals. A large fraction of the protons in aromatic structures of HVB coals (e.g., No. 301 and No. 601) can be mobilized at 180 °C. Only 25% of the aromatic protons in the LVB coal, Pochahontas No. 3 (No. 501), are in a relatively mobile state at 180 °C. This is not surprising, since most of the aromatic structures in a LVB coal are in relatively large aromatic ring systems, which are very rigid and difficult to mobilize.48 For aromatic protons in Pocahontas No. 3 coal, the dipolar-dephasing curves are well fitted as a single Gaussian decay at temperatures of 25 and 120 °C. The single Gaussian decay time constant is increased from 10 to 20 µs as the temperature is increased from 24 to 120 °C. At 180 °C, the dephasing curve is better fitted by two Gaussians with time constants of 9.3 and 52 µs. This suggests that the spin diffusion (and/or spin exchange) rates between rigid and mobile domains decrease with increasing temperature, which is consistent with increased mobility of molecules at higher temperature. At 180 °C, the spin diffusion and/or exchange rates are slow enough to make two Gaussian

In the case in which ωr is much smaller than 1/τc, the

(48) Schobert, H. H. The Chemistry of Hydrocarbon Fuels; Butterworth: London, 1990.

{

}

1 1 2 ) ∆2τc + Tdd 1 + ωr2τc2 1 + 4ωr2τc2

1H

NMR of Coal

components clearly distinguishable. These results suggest that the aromatic protons in Pocahontas No. 3 coal are distributed more uniformly compared to the protons in HVB coals so that spin diffusion and/or exchange between rigid protons and mobile protons is very efficient. For Pocahontas No. 3 coal at 120 °C, two dephasing components can be distinguished for aliphatic protons, while only one component can be detected for aromatic protons. The dephasing time constant (20 µs) of aromatic protons is significantly different from either dephasing constant (10 µs and 52 µs) of the aliphatic protons. This result suggests that the spin diffusion between aromatic and aliphatic protons is not efficient in Pocahontas No. 3 on the experimental time scale (1050 µs), implying that the aliphatic protons and aromatic protons are not intimately close to each other (i.e., at least ca. 5 Å away from each other45). This is reasonable, considering the view that few aliphatic side chains of aromatic structures are present after the coalification process that leads to a carbon composition of around 91% in the resulting coal.48 The above results clearly demonstrate that a detailed correlation of molecular structure with molecular dynamics can be established using the chemical shift information provided by 1H CRAMPS detection. 4. Thermal Treatment versus Solvent Saturation. The dipolar-dephasing experiments clearly show that thermal treatment up to 230 °C can promote molecular motion to only a limited extent. The Gaussian dephasing time constant for a relatively mobile component is still less than 70 µs at 180 °C. However, analogous experiments on pyridine-saturated coals show that a large fraction of molecules are mobilized to show Lorentizian dipolar-dephasing behavior with dephasing time constants as long as about 10 ms even at room temperature.24,31,49 Clearly, pyridine saturation is much more effective in promoting molecular motion than thermal treatment. This result suggests that thermal treatment alone (up to 230 °C) is not enough to break the noncovalent associative bonds that help hold the macromolecular network of coal together. In the absence of solvent saturation, the promotion of molecular motion at high temperature thus relies mainly on (49) Xiong, J.; Maciel, G. E. Manuscript in preparation.


thermal activation, which is apparently not very efficient. A detailed comparison of thermal treatment with solvent saturation of coal is beyond the scope of this paper and will be fully addressed in a separated paper.49 However, we would like to emphasize that the molecular dynamical information obtained from the dipolar-dephasing experiment is consistent with the observed macroscopic properties of coal. For example, the uniaxial optical anisotropy of raw bituminous coal disappears within a few seconds when the coal sample is immersed in pyridine at room temperature; but the anisotropy is not destroyed thermally until the temperature reaches around 350 °C.50 Conclusions In conclusion, 1H CRAMPS dipolar-dephasing experiments can provide detailed information on the molecular dynamics and proton distributions in coals. A “timesuspension” experiment based on CMG-48 is also useful for probing molecular dynamics in coals. With improved dipolar-dephasing experiments based on BR-24 CRAMPS detection, we were able to analyze quantitatively the 1H-1H dipolar-dephasing behaviors of protons in aliphatic and aromatic moieties in coal. We found that thermal treatment up to 230 °C promotes molecular motions in coal with a correlation time of about 10 µs, which is several orders of magnitude slower than the molecular motion induced with pyridine saturation at room temperature. This suggests that thermal treatment alone (up to 230 °C) is not enough to break the noncovalent associative bonds that help hold the macromolecular network of coal together. This in situ variable-temperature 1H CRAMPS study has provided detailed correlation of molecular dynamics with molecular structure for coals thermally treated between 25 and 250 °C. Acknowledgment. This work was supported by the U.S. Department of Energy Grant Number DEFG22-93PC93206. EF960219S (50) Brenner, D. Nature 1983, 306, 772.

In Situ Variable-Temperature High-Resolution 1H NMR Studies of

Recommend Documents