Direct Determination of Oxygen in Coal B. S. Ignasiak,' B. N. Nandi, and D. S. Montgomery Department of Energy, Mines, and Resources, Mines Branch, Fuels Research Centre, Ottawa, Canada Two METHODS are applicable for the direct determination of the total quantity of oxygen in organic matter: Schutze method (I) and modifications (2-8) and neutron activation method (9, IO). The principle of the Schutze method depends on thermal decomposition of the sampIe in an atmosphere of inert gas and conversion of all oxygen-containing products into CO which can be determined directly (5) or indirectly (2). The Schiitze method and its variations require oxygen-free sweeping gas and a relatively long time of analysis. The determination of oxygen using the neutron activation method has been described many times in detail (IO). The great disadvantage of this universal method is the expensive equipment. The recognized methods for determining the oxygen content of coal: Schiitze (I), Unterzaucher (2), and Burns (8), are slow and require considerable skill to achieve satisfactory levels of accuracy. The method described here depends on the pyrolysis of a sample at 1050 O C in a vacuum, catalytic conversion of all oxygen in the pyrolysis products into CO, the determination of the volume of evolved gas, and the measurement of the concentration of carbon monoxide in this gas, using gas chromatography. EXPERIMENTAL
Apparatus. The basic part of the determination, which is the pyrolysis and the measurement of the volume of evolved gas, is conducted in the apparatus indicated in Figure 1. The part of this apparatus indicated by I is referred to as the pyrolyser; it is used for the pyrolysis of the sample, located at a, and the evacuation of the catalyst, situated at b. As it must withstand relatively high temperature, it is made from Vycor glass. The pyrolyser is connected through unit 2 to a 100-ml gas buret 3 by means of heavy rubber tubing. After completion of the pyrolysis, the mercury from reservoir 4 is introduced into the apparatus through the one-way stopcock c. The pyrolysis is conducted in a hinged tubular electric furnace into which the pyrolyser unit is placed as shown in Figure 2. The additional equipment required for the determination of oxygen includes a vacuum pump and 1 Permanent address, Department of Coal Chemistry, Adam Mickiewicz University, Poznan, Poland.
(1) M. Schiitze, 2.Anal. Chem., 118, 245 (1939). (2) J. Unterzaucher, Chem. Ber., 73B, 391 (1940). (3) . . J. A. Kuck, A. J. Andreath, and J. P. Mohns, ANAL.CHEM.. 39, 1249 (1967). (4) M. D. Korszun. Zavod. Lab.. 10.241 (1941). (5) A. P. Tierientiew, A. M. Thkieltaub, E. A. Bondariewska-ja, and L. A. Domoczkina, Dokl. Akad. Nauk SSSR, 148, 1316 (1963). (6) L. W. Kuzniecowa, E. N. Stolarowa, and S . L. Dobyczin, Zh. Analit. Khim., 20, 836 (1965). (7) I. J. Oita and H. S. Conway, ANAL.CHEM.,26,600 (1954). (8) M. S. Burns, R. Macara, and D. J. Swaine, Fuel (London), 43, 354 (1964). (9) L. C . Bate, Nucleonics, 21, 72 (1963). (IO) H. P. Dibbs, Department of Mines and Technical Surveys, Mines Branch, Technical Bulletin TB 5 5 , March, 1964. 1676
ANALYTICAL CHEMISTRY
370 rnm
Smrn
Figure 1. Apparatus for pyrolysis conversion and measurement of volume of gas evolved
gas chromatograph suitable for the measurement of carbon monoxide. In this investigation, a Fisher partitioner was used with a double column : hexamethylphosphoramide on Chromosorb P, followed by a Linde molecular sieve 13 X. By using such a set of columns, it was a simple matter to determine whether the conversion process was quantitative or not, because of the presence of any COS shows that the conversion is incomplete. The sample for analysis (60 mg to 100 mg) was weighed in the quartz tube 5 whose shape and dimension are presented in Figure 1. With coals of high oxygen content such as lignite, the sample weight is kept to 60 mg, while for anthracite the sample weight is increased to 100 mg. Preparation of the Catalyst. Two hundred grams of lampblack were mixed in a beaker containing 630 grams of nickelous formate (very fine powder) and 4000 ml of distilled water. Eight grams of soluble starch were then added to this solution. The contents of the beaker were vigorously mixed and heated to evaporate the water. The heating was stopped when the residue started to solidify. The contents of the beaker were dried for approximately twelve hours at 105 "C in an oven. Then the catalyst was transferred to a Vycor tube swept with hydrogen. The temperature was increased gradually to 1050 O C and maintained under these conditions for three hours. The catalyst was then cooled and stored in an air-tight container. The particles of the catalyst below 16 mesh (Tyler) were used for the oxygen determinations. Method of Analysis. Five grams of catalyst were introduced into tube b of the pyrolyser (see Figure 1). Part a was cleaned carefully to remove all vestiges of catalyst particles and the sample in quartz tube 5 was introduced into part a of the pyrolyser. The apparatus was assembled, set up on a stand, and evacuated initially for five minutes. Then part b of the pyrolyser was inserted into an oven for thirty minutes at 1050 "C. Five minutes before the end of this period, in which the catalyst was degassed, the apparatus was re-evacuated. At the conclusion of this five-minute period, tubing b was withdrawn from the furnace and tubing a was inserted for ten minutes into a thermostaticallycontrolled bath at a temperature of 70 O C to remove traces of moisture from the coal sample. After tubing b was cold, the catalyst was transferred into a by rotating the pyrolyser, and after the pressure dropped to 0.3 mm of Hg, the apparatus was disconnected from the vacuum pump using stopcock c. Then the hinged furnace was opened for some seconds, the
-
HINGED FURNACE
C4TALYST
Figure 2. Location of the pyrolysis tubing in oven at the beginning of experiment
pyrolyser tube I was placed in the furnace as shown in Figure 2 to avoid heating the coal. The furnace was then closed and when the temperature stabilized again at 1050 "C, the furnace was moved very slowly in the direction of the sample. The rate of evolution of the pyrolysis gases from the sample could be observed by eye, and the rate of advance of the furnace was controlled to avoid sudden bursts of gas. The sample was finally advanced to the mid-point of the furnace and retained there for ten minutes. After this time, the pyrolyser was withdrawn from the furnace, and three to four minutes later, the mercury from reservoir 4 was introduced, with vibration, into the apparatus. The converted pyrolysis gas was displaced into the gas buret. The volume of the gas was measured and the temperature and atmospheric pressure was noted. The concentration of carbon monoxide was determined by withdrawing a sample for gas chromatography. The concentration of oxygen in the analyzed sample was calculated from the equation :
850
900
G
TI
1,050
1,000
TEMPERATURE
1,100
OC
Figure 3. Oxygen content by neutron activation of coke from demineralized coal, pyrolyzed for ten minutes at indicated temperature A 0
--
anthracite sub-bituminous coal low volatile bituminous coal oxygen content of pure phenanthrene by neutron activation
from pyrolysis of pure chemical compound not containing oxygen is denoted by K . The quantity K is a blank, which is a function of the conditions of analysis, such as, temperature, time of pyrolysis, method of the preparation of catalyst, time, and temperature of the degasification of catalyst. Consequently, to obtain reproducible results, it is important to adhere strictly to the same analytical procedure.
CO X Pi X VI X 0.003592
%O=-
950
RESULTS AND DISCUSSION
-
where G is the weight of sample (grams), VI is the volume of evolved gas (ml) at temperature TI in O K and P1 atmospheric pressure (mm Hg). The volume of CO (ml) at NTP obtained
The proposed method for the direct determination of oxygen in coal was tested on two standard pure compounds, namely, benzoic acid and acetanilide. Comparative analyses by the neutron activation method were performed on these standards.
Table I. Results from the Determination of Oxygen in Standards Oxygen, determined using Pyrolysis method Mean Neutron-activation method
z
Standards 1) Benzoic acid 2) Acetanilide
26.3,26.7,26.3,26.0, 26.5, 26.7, 26.2, 26.3,26.3 12.2, 11.7, 12.1
Mean
Theoretical,
26.4
26.14, 27.16,27.12, 25.82, 25.54,25.71
26.25
26.20
12.0
12.12, 12.24, 11.76
12.04
11.84
z
Table 11. Oxygen in Coal The Results Expressed on the Dry Coal Basis
Sample 1
Pyrolysis method
Mean
2
3
6.4, 6.9, 6.4, 7.1, 7 . 1 , 6 . 9 6.6, 6.6, 6 . 9 4.2,4.5,4.0,4.0 4.3, 4 . 5 , 4 . 2
6.8 6.7 4.2 4.3
Oxygen, Z Neutron activation method 4
Carbon Mean 5
d.a.f. 6
Ash 7
Bituminous coals Moss 3
Moss 3-demineralized Itmann Itmann-demineralized Lignites Turow-demineralized Smogory-demineralized Sub-bituminouscoal Demineralized Anthracite Fresh Demineralized
24.9, 26.4, 26.4, 26.5 29.9, 31.0, 29.8
26.0 30.2
8.17, 8.58, 8.84, 8.60 6.75, 6.73, 6.13 6.40,6.47,6.68,6.58 4.26,4.24,4.44
82.4 8.55 6.54 6.53 4.31
85.1
6.73 0.29 6.30 0.15
26.13,26.21, 27.77, 25.99, 25.81 33.75, 32.38,29.60, 30.77
26.38
68.8
0.25
31.62
66.3
0.26
20.14
21.1, 20.7
...
20.00, 20.29
1.8,2.1 2.3,2.0,2.2
2.0 2.2
7.59 2.66
7.59 2.66
0.11 92.5
VOL. 41, NO. 12, OCTOBER 1969
11.05 0.91
1677
The results presented in Table I show that the proposed method can be considered as a basically suitable analytical procedure for the direct determination of oxygen in many organic compounds. It is evident that some caution has to be used in those cases where the compound or mixture is known to contain elements that can poison the nickel catalyst, such as, sulfur, selenium, and phosphorous. However, in the experiments described here, the weight of catalyst was at least fifty times greater than the weight of the coal sample. No poisoning effect was observed during coal analysis and the percentage of carbon dioxide (by volume) in converted gases was in the range 0.00 to 0.10%. The establishment of this method for the determination of oxygen in coal required finding the temperature of pyrolysis at which essentially all oxygen from organic matter of coal was evolved from the coke. It was found, from the determination of the concentration of oxygen by neutron activation in cokes resulting from the pyrolysis of some coals of different rank demineralized according to Radmacher ( I I ) , that the temperature of 1050 "C was suitable. This is illustrated in Figure 3. Only in the case of anthracite was the concentration of oxygen, in cokes pyrolyzed at temperatures of 1050 and 1100 "C, relatively high (approximately 0.9%). At the same (11) W. Radmacher and 0. Mohrhauer, Brennsfof-Chem., 37, 353 (1956).
time, the concentration of ash in this particular anthracite, after treatment with hydrochloric and hydrofluoric acids, was surprisingly high (Table 11, Column 7). So, it is assumed that the high concentration of oxygen in cokes from pyrolysis of anthracite at 1050 and 1100 OC is related to the inorganic matter in these cokes. It is worth mentioning that the concentrations of oxygen in both these cokes, though heated to different temperatures-i.e., 1050 and 1100 OC-were the same. Had the oxygen been contained in the organic matter, some decrease in concentration would have been observed. Coals of different rank were analyzed before and after removal of the mineral matter, using the direct method of determining oxygen. The results are shown in Table 11. Here, it is evident that the oxygen contents of the demineralized samples are in good agreement with those found by neutron activation analysis. ACKNOWLEDGMENT
The authors acknowledge the assistance of the Mineral Sciences Division and Mr. C. McMahon, in particular, for determining the oxygen contents of the coal and coke samples by neutron activation analysis, and Mr. M. Pleet for preparing the drawings.
RECEIVED for review May 20, 1969. Accepted July 21, 1969.
Poly-M - phenoxylene A New Liquid Phase J. H. Beeson' and R. E. Pecsar* Varian Aerograph, 2700 Mitchell Drive, Walnut Creek, Calif. 94598
THE TECHNIQUE of gas-liquid chromatography is now an indispensable tool in areas of research ranging from petroleum analysis to clinical diagnosis. The GLC columns are often used at temperatures exceeding 300 OC but the worker who wishes to operate at these temperatures quickly finds that the number of liquid phases available decreases dramatically as a function of increasing operating temperature. Above 300 "C only a few silicone polymers (primarily SE-30 and OV-1) have wide acceptance. The limiting features of most liquid phases at these temperatures are volatility and decomposition which act alone or in combination to give unacceptably high bleed levels at high temperatures. The bleed level is of great importance for several reasons. First and most obvious is depletion of the column. Of equal importance are contamination of collected fractions, saturation of detectors (which often results in operation in the nonlinear concentration regions even when electrical bleed compensation is possible), contamination of tritium and nickel-63 foils in radio-ionization detector types and complication of mass or optical spectra obtained in flow systems which use a gas chromatograph as the inlet. While the silicone polymers are acceptable at temperatures above 300 "C, the chromatographer is still limited by two 1 Present address, Ames Co., Division of Miles Labs., Elkhart, Ind. 46514. 2 To whom reprint requests should be made.
1678
ANALYTICAL CHEMISTRY
facts. These silicones are not acceptable above 350 OC and the silicones represent a very limited selection in an art whose versatility has largely been attributable to the wide selection of phases available. A solution to the chromatographer's problem is to design his own liquid phases. A well designed polymer should have a high pyrolysis temperature, low volatility, a melting point compatible with normal G C operation, and a chemical structure which has features that contribute to chemical specificity. If low volatility and high pyrolysis temperature are taken as the sole prerequisites for the synthetic liquid phase, a survey of the chemical literature indicates that a suitable polymer is poly-rn-phenoxylene (1-3). The 5 and 6 unit oligomers of this polymer series are currently commercially available as 5-ring and 6-ring poly(rn-phenyl) ether (PMPE) and the chromatographic behavior of these phases is well documented. A polymer with these chromatographic properties and improved physical properties at high temperatures has great potential value in many applications-e.g., the (1) J. M.Cox, B. A. Wright, and W. W. Wright, J. Appl. Polym. Sci., 9,513 (1965). (2) G. P. Brown and A. Goldman, Preprints American Chemical Society, Polymer Chemistry Division, 4 (2), 39 (1963). (3) G. P. Brown and A. Goldman, Preprints American Chemical Society, Polymer Chemistry Division, 5 (l), 195 (1964).
