AN AMERICAN CHEMICAL SOCIETY JOURNAL VOLUME 6, NUMBER 4
JULY/AUGUST 1992
0 Copyright 1992 by the American Chemical Society
Articles Moisture Release from Argonne Premium Coal Samples. A Quantitative 31PNMR Spectroscopic Study A. E. Wrbblewskif and J. G. Verkade* Department of Chemistry, Iowa State University, Ames, Iowa 50011-3111 Received September 9, 1991. Revised Manuscript Received March 12, 1992
To study the possibility of differentiating types of moisture in coals, four Argonne Premium Coal Samples (APC's) (Beulah-Zap, Illinois No. 6, Blind Canyon, and Pittsburgh No. 8) having a range of moisture contents were extracted at 23 "C with pyridine, dioxane, dimethoxyethane, tetrahydrofuran, ethyl methyl ketone, and acetonitrile. The moisture contents in the extracts were determined after '/.,, 1/2, 1 , 2 , 4 , and 8 h by 31PNMR spectroscopy using the tagging reagent C1P(0)Ph2. From the lowest rank coal almost all of the water was extracted with pyridine within the first 15 min and for the remaining coals studied this process was essentially complete after 8 h. The water extracted from each coal with the least efficient extraction solvent (dioxane or dimethoxyethane) within 15 min is considered as a reasonable estimate of the surface moisture. Higher rank coals (Pittsburgh No. 8 and Blind Canyon) contain about 50% of their total moisture as surface water, while as much as 69 and 88% such water is found for Illinois No. 6 and Beulah-Zap coals, respectively. Discrepancies between moisture contents in APC's determined by drying at 108 "C and by pyridine extraction at room temperature are accounted for by trapping as much as 25 and 15% of the coal water inside micropores of Pittsburgh No. 8 and Blind Canyon coals, respectively. There was no water left in the micropores of the Illinois No. 6 coal after drying. Upon oven drying, the Beulah-Zap coal gave up ca. 4% more water than could be extracted over a period of 24 h by pyridine. This excesa moisture is attributed to decomposition of organic matter under oven-drying conditions. Introduction Water in coal has been classified into four major forms: (i) superficial free water, (ii) capillary condensed water, (iii) sorbed water associated with polar groups and cations, and (iv) water released only by chemical decomposition of organic or inorganic matter.' To date, no methodology has been developed to clearly differentiate among these forms. For most applications, the division of coal moisture into two types is useful, namely, freezable and nonfreezable; the former is bulk or surface water while the latter is represented by pore moisture.2 It is this classification we adopt here. Drying coals at ca. 107 "C is a commonly accepted procedure for moisture estimation in coals for commercial p u q x ~ e a . It ~ has been shown, however, that this approach On leave of absence from Technical University, Ltklai, Poland.
expels bulk surface water and a significant fraction (but not all) of the pore water." Our successful introduction of diphenylphosphinic chloride (ClP(0)Ph2,1) as a reagent for total moisture quantification in various APC's by means of pyridine extraction at ca. 107 "C7 prompted us to extend the use of this technique in studies directed towmd the differentiation of surface from pore water. We reasoned that it might be possible to select a polar, (1) Allardice, D. J.; Evans, D. G. Anal. Meth. Coal, Coal R o d . 1978, I , Chapter 7. (2) Abhari, R.;Isaacs, L. L. Energy Fuels 1990,4 , 448. (3)1988 Annual Book of ASTM Standards, Vol. 05.05, D3302; pp 331-337. (4)Hippo, E. J.; Neavel, R. C.; Smith, S.E.; Lang, R. J.; Miller, R.N. Prepr. Pap.-Am. Chem. Soc., Diu. Fuel Chem. 1987,32,179. (5) Finseth, D.; Prepr. Pap.-Am. Chem. Soc., Diu. Fuel Chem. 1987, 32, 260. (6) Schafer, H. N. S.Fuel 1980,59,295. (7)Wrbblewski,A. E.;Reinartz, K.; Verkade, J. G. Energy Fuels 1991, 5,786.
0887-0624/92/2506-0331$03.00/00 1992 American Chemical Society
Wrdblewski and Verkade
332 Energy & Fuels, Vol. 6, No. 4, 1992 Table I. Argonne Premium Coal Samples Used in This Study identificn sealing coal no. date 801 00321 2/13/87 Beulah-Zap lignite North Dakota 4/18/86 Illinois No. 6 high-volatile 301 00600 bituminous 00601 Blind Canyon high-volatile 601 00285 10/23/86 Seam bituminous 00286 00287 ..-_ .
Pittsburgh No. 8 bituminous
401
00380 00381 00382
4/22 186
Table 11. **P NMR Spectroscopic Moisture Determination (in weight percent) in Extracts of Beulah-Zap North Dakota Coal time. h
solvent pyridine dioxane dimethoxyethane tetrahydrofuran ethyl methyl ketone acetonitrile
lid 32.47 29.32 29.30 29.72 30.11 30.51
'I, 32.18 29.29 29.58 29.75 29.69 30.48
1
2
32.26 29.24 29.61 29.66 30.21 30.51
32.44 29.52 29.73 30.17 30.25 30.83
4 32.34 30.03 29.85 30.49 30.22 31.20
8
33.10b 30.14 30.12 30.45 30.63 31.23
"Oven-drying: first day 34.61%; last day 34.46% and 34.66%; vacuum-drying 111 "C 34.56%. bAfter 24 h 33.28%.
water-miscible,non-hydroxylic solvent which would be able to extract all the surface water from coal soon after mixing, while other forms of water held in the coal's interior would be released into the organic phase more slowly. To this end we have undertaken an initial study of moisture release, as measured by a 31PNMR tagging reagent,' from four -20 mesh APC's at room temperature in the presence of a large excess of five polar solvents. For comparison, pyridine was also used.
Table 111. 31PNMR Spectroscopic Moisture Determination (in weight percent) in Extracts of Illinois No. 6 Coals as a Function of Solvent and Time" time, h solvent 114 112 1 2 4 8 pyridine 9.14 9.26 9.26 9.42 9.38 9.39* dioxane 6.48 7.07 7.61 7.82 8.38 8.61 dimethoxyethane 8.22 8.60 8.86 8.75 8.81 9.05 tetrahydrofuran 8.14 8.44 8.70 8.85 9.09 9.05 ethyl methyl ketone 7.14 7.87 8.12 8.54 8.78 8.70 acetonitrile 8.34 8.49 8.68 8.89 8.90 9.06
Experimental Section
"Oven-drying: first day 9.42%; last day 9.50% and 9.53%; vacuum-drying = 9.51%. bAfter 24 h 9.44%.
Reagents. Pyridine, dioxane, dimethoxyethane, tetrahydrofuran, ethyl methyl ketone, and acetonitrile were stored over 4A molecular sieves in flasks stoppered with rubber septa The same batches of solvents were used throughout this study. Chloroform-d was used as received. Diphenylphosphinic chloride (1) was synthesized from chlorodiphenylphosphinein two steps as described in the 1iterat~re.s~ The standard solutions (0.45-0.49 M) of 1 were prepared in an argon-filed glovebag by dissolving 13.615.0 g of the reagent in ca. 175.0 g of chloroform-d. Methyltriphenylphcephonium iodide was prepared according to a literature procedure.1° Coals. The -20 mesh coal samples used in this study were obtained through the hgonne Premium Coal Samples program" and are listed in Table I. The ca. 10-g ampules were opened in an argon-filled glovebag and the coals were placed in tightly capped narrow-mouth vials. Each coalwas used up within 10days. For Beulah-Zap, Illinois No. 6, Blind Canyon, and Pittsburgh No. 8 coals, 10, 20, and 30-g batches were used, respectively. This required that two or three ampules be mixed at the beginning of some of the studies. Total Moisture Determination. For comparison, the total moisture in the coals studied in this report was determined by oven drying at 105-108 "C (ASTM D3302)3and by drying in vacuo (0.5 " H g ) for 24 h in a drying pistol at 111"C, the temperature of boiling toluene. slPNMR Moisture Determination with Reagent 1. In an oven-dried flask containing a magnetic stirrer bar, a sample of a coal was placed (ca. 1g for the North Dakota, ca 2.5 g for the Illinois No. 6 and Blind Canyon, and ca. 4.0 g for the Pittsburgh No. 8 coals). The flask was stoppered with a rubber septum. After addition of solvent (60.0 mL) via syringe, the flask was immersed in a water bath at 23 1 "Cand the SuspensioIlrwasmagnetically stirred. Samples were withdrawn via syringe after 15 and 30 min and after 1,2,4, and 8 h. In addition, a sample was withdrawn after 24 h when pyridine was used as the extraction solvent. Five minutes before a sample was taken, stirring was interrupted to allow settling of the coal particles. Each sample withdrawn consisted of a 0.50-mL aliquot which was added to a septum-capped NMR tube containing methyl-
*
(8) Ocone, L. R.; Schaumann, C. W.; Block,B. P. Znorg. Synth. 1966, 8, 71. (9) Crofts, P. C.; Downie, I. M.; Williamson, K. J. Chem. SOC.1964, 1240. (10) Lightner, D. A.; Crist, B.V.; Kalyamam, N.; May, L. M.; Jack", D. E. J. Org. Chem. 1986,50, 3867. (11) Vorres, K. S. Energy Fuels 1990, 4 , 420.
Table IV. 31PNMR Spectroscopic Moisture Determination (in weight percent) in Extracts of Blind Canyon Coals as a Function of Solvent and Time" time, h solvent 114 '12 1 2 4 8 pyridine 5.65 5.93 6.15 6.49 6.51 6.83b dioxane 3.44 3.78 4.03 4.36 4.62 5.05 dimethoxyethane 3.58 4.01 4.72 5.29 5.64 6.06 3.89 4.71 5.41 5.67 5.81 6.01 tetrahydrofuran ethyl methyl ketone 3.51 3.77 4.24 4.37 4.78 5.30 acetonitrile 3.35 3.76 4.23 4.63 5.00 5.63 "Oven-drying: first day 5.85%; last day 5.53% and 5.55%; vacuum-drying = 5.77%. bAfter 24 h 6.72%. triphenylphoephonium iodide as the internal integration standard (ca. 180-250 me), the relaxagent C r ( a ~ a c(10.5-12.0 )~ mg), the solution of the reagent (2.40 mL), and pyridine (0.40mL). After the two layers were mixed together the 31PNMR spectra were taken. For each series a background moisture content was determined on a 0.50-mL aliquot of the solvent using the same procedure. NMR Instrumentation and Integration Procedure. These are the same as described earlier.' Calculations of Percent Moisture. These calculations did not include corrections for the withdrawal of the aliquota since only negligible overestimations are introduced (0.7,0-0.6,0.3, and 0-0.1 relative percent for Pittsburgh No. 8,Blind Canyon, Illinois No. 6, and Beulah Zap coals, respectively). The overestimations increase to their highest values for the 8-h measurementa. Error Estimation. Analytical balances (accuracy iO.OOO1 g and -+0.001g) were employed to weigh by difference the coal and the aliquota, and the total amounts of solvents used, respectively. Thus, the major source of error in the percent moisture determinations by our method originates in the integrations of the NMR signals (i.e., the difference between the analytical signal and the background signal). The total error in the data shown in Tables 11-V is estimated as i 2 % (relative).
Results Following a series of preliminary experiments, the op' I 2 ,1,2,4, and 8 h was timum sampling sequence of 'I4, established. For pyridine extractions, the moisture content was also determined after 24 h in order to determine
Moisture Release from Argonne Premium Coal Samples
NMR Spectroscopic Moisture Determination (in weight percent) in Extracts of Pittsburgh No.8 Coals as a Function of Solvent and Time” time, h solvent ‘14 ‘12 1 2 4 8 2.08 2.22 2.38 2.44 2.44 2.43b pyridine dioxane 1.18 1.31 1.52 1.59 1.60 1.71 1.04 1.23 1.34 1.47 1.53 1.68 dimethoxyethane tetrahydrofuran 1.32 1.43 1.47 1.53 1.67 1.71 ethyl methyl ketone 1.23 1.33 1.44 1.61 1.69 1.64 acetonitrile 1.35 1.47 1.63 1.88 1.86 1.88
Energy & Fuels, Vol. 6, No. 4,1992 333 Illlnois No 6
Table V.
I
C
1
Acetonitrtle, Tetrohydrofuron
Pyridine
Dimethoxyethone
0
% H20
Ohen-drying: first day 1.81%, last day 1.82% and 1.83%; vacuum-drying = 2.05%. bAfter 24 h 2.41%. 7.0
Beulah- Zap
v
+
Pyridine
0 Acelomtrde
A Methyl ethyl kelone Telrohydroluran 0 Dioioru
0 Dimelhoxyelhone
~
I
2
3
4
’
5
6
7
8
Extroction Time (hours)
32.0
Figure 2. Percent moisture removed from -20 mesh Illinois No. 6 coal by various solvents as a function of time.
t
Blind Canyon
7.c
I
I
I
I
I
I
I
I
I
Acetonltrlle 0
31 0 Methyl ethyl ketone
6.C
-
Tetrohydrofuron
- Y //
Dimethoxyethone
/
% H2O
Dlmethoxyelhone
J
/
Acetonllrde
5.c
+ Pyridine
0 Acelontlrile
A Melhyl ethyl
k@bM Telrohydrofuron
0 Dioiom D Dim~lhoiyelhone
+ Pyridine
1
2
3
4
5
6
7
8
0 Acelonilrile A Malhyl elhyl kelm8
4.c
Telmhydrofutan 0 Dioxane
Exlroctlon Time (hours)
Dinmlhoxyelhon*
Figum 1. Percent moisture removed from -20 mesh Beulah-Zap coal by various solvents as a function of time.
whether more water could be released when the extraction time was extended. Moisture contents in Beulah-Zap North Dakota, Illinois No.6, Blind Canyon and Pittsburgh No. 8 -20 mesh coals extracted with six solvents are collected in Tables 11,111, IV,and V, respectively. For comparison, the moisture in these coals was also determined by oven drying at 105108 OC on the first and the last day (10 days later) of the NMR measurement series with each coal. It was concluded from these experiments that no detectable moisture was lost or gained during handling of samples. Estimates of moisture by drying in vacuo are also included in the tables. Plots of the percentage of water released from APC’s studied in this report va time are shown in Figures 1-4. The curves in these plots were estimated visually. It appeared that in all cases pyridine is a unique solvent compared with the five others used here. Thus within the first 15 min,pyridine generally extracts more water than any of the other solvents even after 8 h. As shown in our earlier work,’ moisture determination by coal extraction with pyridine gives higher values than those obtained wing a standard method? except in the case of North Dakota lignite for which moisture estimates by drying are higher. Furthermore, for North Dakota and Illinois No. 6 coals, almost all of the water was released instantaneously during pyridine extraction. On the other hand, under the same conditions all the water was extracted at room temperature after 2 h from the low-moisture-content coals (Blind
1
2
3
4
5
6
7
8
Extroction Time (hours)
Figure 3. Percent moisture removed from -20 mesh Utah Blind Canyon coal by various solvents as a function of time.
Canyon and Pittsburgh No. 8). No significant differences in water extraction patterns with solvents other than pyridine were found for BeulahZap coal (Figure 1). A steady increaseof released moisture over an 8-h period was observed, and it appears that this process would continue in a like manner over a longer period. On the other hand, only tetrahydrofuran, dimethoxyethane, and acetonitrile have similar patterns of moisture extraction from Illinois No. 6 coal (Figure 2). This pattern is also very similar to that of pyridine. In this case, however, ethyl methyl ketone and dioxane show significant differences. Surprisingly, after 15 min almost the same amount of water (ca.3.5%) was extracted from the Blind Canyon coal with dimethoxyethane, acetonitrile, dioxane, and ethyl methyl ketone, and slightly more (3.89%) with tetrahydrofuran. However, further release of moisture into these solvents follows different patterns (Figure 3). The moisture release patterns from the Pittsburgh No. 8 coal for extraction with dioxane, tetrahydrofuran, dimethoxyethane, or even acetonitrile were rather similar (Figure 4). As expected, pyridine extraction of water from
Wr6blewski and Verkade
334 Energy & Fuels, Vol. 6, No. 4, 1992 Pittsburgh No 8
2
1
1
1
1
Pyridine
I
2
/Methyl
%
ethyl ketone, Tetrohydrofuron
-
HZO I
y
t Pyridine
0 Acelonilrile
A Melhyl ethyl ketone T~lrohydrofuron
0 Omone
0 O~m~lhoa~tlhonc
I I
I
I
I
I
I
1
I
1
2
3
4
5
6
7
8
Extraction Tlme (hours)
Figure 4. Percent moisture removed from -20 mesh Pittsburgh No. 8 coal by various solvents as a function of time. this coal was much higher than those into other solvents. What is striking in this particular case is the almost identical curvatures of the moisture release patterns for the first four measurements made for all solvents.
Discussion The plots of percent moisture extracted from each of the four coals by the solvents used in this study as a function of time are arranged in order of increasing coal rank in Figures 1-4. The data for these plots appear in Tables 11-V, respectively. During the first 15 min of extraction, pyridine is seen to have removed the most moisture from all four coals. (We call this moisture “very loosely plus loosely bound moisture” in Table VI). While it is tempting to attribute this observation to the superior swelling properties of this solvent, it has recently been shown12that the swelling of these coals (as well as Montana subbituminous, Bruceton high-volatile bituminous and Powhatton high-volatile bituminous) is negligible ( Pittsburgh No. 8 > Illinois No. 6. While this seems reasonable based on the decreasing concentration of water left in the coal in the same order, this order is not the same for these parameters for other solvents and indeed the two orders for the two parameters for a given solvent can be different. For the solvents other than pyridine in Figures 1-4, the initial slopes of the curves become notably larger from Figure 1to Figures 2-4. In Figure 2,these slopes fall into roughly two groups, with those for dioxane = methyl ethyl ~
(12) Otake, Y.; Suuberg, E. M. Fuel 1989, 68, 1609.
ketone (MEK) z acetonitrile (ACN) < dimethoxyethane (DME) = tetrahydrofuran (THF). In Figure 4 the initial slopes of the curves for the solvents other than pyridine are hardly differentiable. Interestingly, the initial percentages of water extracted by these solvents after 15 min decrease in the order Illinois No. 6 > Blind Canyon > Pittsburgh No. 8, suggesting that, with increasing coal rank, these coals decrease their tendency to differentiate these solvents even though the moisture contents are not the same, and in fact decrease with increasing coal rank. From Figures 2-4 it is also seen that the gap between the pyridine curve and the next solvent curve widens, indicating a clustering of the curves toward the lowest curve which is that for dioxane. The exception is the Pittsburgh No. 8 case for which the DME curve falls noticeably below the dioxane plot (Figure 4). These results suggest that the solvent associated with the lowest curve on our plots extracts within 15 min a type of water common to each coal. This very loosely bound moisture in Table VI is likely to be made up mainly of surface water. (For Beulah-Zap, the lowest solvent curve could be either that for dioxane or DME, since they are so close together.) For reasons not presently obvious to us,the percentage of this surface water removed by the lowest-curve solvent correlates linearly with coal rank (i.e., as indicated by the % C) as shown in Figure 5. It should be noted that our estimate of 89% of the moisture extracted from the Beulah-Zap sample by dioxane or DME within 15 min compares well with the observation by Vorres et al. that there is a comparatively fast loss of 85% of the moisture in this coal in isothermal thermogravimetric experiment^.'^ It is attractive to postulate that the moisture extraction percentages at the 15-min mark for dioxane and DME can be quite similar (Figures 1 and 3) because both solvenb are diethers and should therefore have similar bulk water extraction capabilities. It should be borne in mind, however, that Illinois No. 6 clearly differentiates these solvents (Figure 2) for reasons that are presently obscure. Another example of this behavior is that for ACN which is immediately below that of pyridinc except in the case of Blind Canyon where the ACN curve lies considerably below that for THF (Figure 3). Comparison of the total moisture content in the APC’s studied here as measured by room temperature pyridine extraction for 8 h and by oven drying a t 108 “C for 1-h reveals interesting results (Table VI). The Beulah-Zap coal appears to give up more water upon thermal drying than upon 8-h pyridine extraction, while the opposite is true for the remaining coals. We attribute the former result to the thermal instability of lignites which can lead to water elimination as a result of the decomposition of organic matter.14J5 The retention of moisture by the three higher-rank coals upon thermal drying, compared with 8-h pyridine extraction, decreases with coal rank from 25% for Pittsburgh No. 8, to 15% for Blind Canyon, to 0% for (13) Vorres, K. S.; Molenda, D.; Dang, Y.; Malhotra, V. M. Prepr. Pap.-Am. Chem. SOC.,Diu.Fuel Chem. 1991,36,108. (14) Larsen, J. W.; Lee, D. Fuel 1985, 64, 981. (15) Lucht, L. M.; Peppas, N. A. Erddl, Kohle, Erdgas, Petrochem. 1987, 40, 483. (16) Goslar, J.; Kispert, L. D. Energy Fuels 1989, 3, 589. (17) Aida, T.; Fuko, K.; Fujii, M.; Yoshihara, M.; Maeshima, T.; Squires, T. G. Energy Fuels 1991,5,79. (18) Hall,P. J.; March, H.; Thomas, K. M. Fuel 1988,67,863. (19) Larsen, J. W.; Green, T. K.; Kovac, J. J. Org. Chem. 1985,50, 4729. (20) Painter, P. C.; Park, Y.; Sobkowiak, M.; Coleman, M. M. Energy Fuels 1990, 4 , 384. (21) Allardice, D. J.; Evans, D. G. Fuel 1971, 50, 219. (22) Swann, P. D.; Hank, J. A,; Sieman, S.R.; Evans, D. G. Fuel 1973, 52, 154.
Moisture Release from Argonne Premium Coal Samples
~~
Table VI. Correlation of Moisture Contents with Coal Rank coal Pittsburgh Blind Illinois BeulahNo. 8 Canyon No. 6 Z~D 83.2 80.7 77.7 72.9 % C" 49 51 69 % very loosely 89 bound moisturensb % very loosely plus 86 83 97 98 loosely bound % loosely bound
37
32
28
9
% tightly bound
13
17
3
2
2.4
6.8
9.4
33.1
1.8
5.8
9.5
34.5
-25
-15
0
+4
moisture"*' % "total" moisture by extractionb % "total"by oven dryin@ % "total" moisture differenceh
Energy & Fuels, Vol. 6,No. 4, 1992 336
l\\sh
Pocahontas No 3
90
1
Upper Freeport
.;ti; "0
10
Beuloh- Zap
20
30
40
4
50
% H20
Figure 6. Percent moisture retained after oven drying at 108 "C for 1h as a function of coal rank (% C) and mesh. I
I
1
I
I
I
r 7 4
I
' 210
310
20
50
$0
710
810
1
extracted by the lowest-curve solvent (Le., dioxane or DME) is linear with coal rank for the four coals studied here (Figure 5). Thus, pyridine appears to extract a relatively loosely bound water common to all four coals,which remains after the surface water is removed. This additional water removal may be attributed to the better penetrating (hydrogen bond breaking) ability of pyridine, which may be responsible for removing macroporous moisture. The descent in water extraction rates for the various solvents after the first 15 min reflects solvent penetration into deeper pores from which diffusion of solvent/H20 becomes slower. Conclusions. The time-dependent extraction of water by pyridine at room temperature from the APC coals studied here is essentially complete after 8 h. For the low-rank Beulah-Zap coal, the water extracted by pyridine appears to have no contribution from lignite decomposition as has been suggested to be the case with thermal drying.14Js Pyridine extraction of moisture is also more efficacious in the cases of Pittsburgh No. 8 and Blind Canyon APC coals because of their apparent tendency to trap moisture upon oven drying. With less polar solvents, it becomes possible to identify a very loosely bound (surface) water which is removed within the first 15 min and whose amount correlates linearly with coal rank. Following extraction of the surface water, it appears that a loosely bound moisture is removed whose amount also correlates linearly with coal rank.
zapl
9'0
,Lo
% H20
Figure 5. Plots of percent moisture retained by coal samples after oven drymg at 108 "C for 1h (a), percent moisture released after 15 min of extraction by pyridine minus the percent moisture released after 15 min of extraction by the lowest-curve solvent in Figures 1-4 (b) (i.e., the "loosely bound moisture"), and the percent moisture released after 15 min extraction with the lowest-curve solvent in Figures 1-4 (c) (i.e., the "very loosely bound moisture").
Illinois No. 6 coals. This decrease is a linear function of coal rank as measured by percent carbon analysis for these three coals (Figure 5). As was shown earlier by us,' however, this progression is not linear when additional coals are included. Thus,after a maximum percentage of water retention is observed for the Upper Freeport APC sample, a drop in this percentage is observed for the Pocahontas No. 3 APC (Figure 6). It is possible that moisture-containing micropores are more easily collapsed thermally as coal rank increases, thereby effectively trapping moisture, except in the case of Pocahontas No. 3 which may possess a comparatively more stable pore structure. It is also curious that the relative percentage of the moisture extracted by pyridine after 15 min minus that
Acknowledgment. Ames Laboratory is operated for the U.S.Department of Energy by Iowa State University under Contract No. W-7405-ENG-82. This work was supported, in part, by the Assistant Secretary for Fossil Energy through the Pittsburgh Energy Technology Center. Partial support through DOE Grant No. DEFG2288PC88923 is also acknowledged. Registry No. Pyridine, 110-86-1;dioxane, 123-91-1;dimethoxyethane, 110-71-4; tetrahydrofuran, 109-99-9; ethyl methyl ketone, 78-93-3; acetonitrile, 75-05-8; water, 7732-18-5.