Fresenius J Anal Chem (2001) 370 : 60–63
© Springer-Verlag 2001
O R I G I N A L PA P E R
Grażyna Gryglewicz · Stanisław Gryglewicz
Determination of elemental sulfur in coal by gas chromatography – mass spectrometry
Received: 20 November 2000 / Revised: 29 January 2001 / Accepted: 1 February 2001
Abstract A method for the determination of S0 in coal based on the extraction with cyclohexane with subsequent quantitative analysis of elemental sulfur in the extract by GC/MS is described. The quantity of elemental sulfur was determined in four coal samples with different distribution of sulfur forms. The effect of solvent and extraction time on the efficiency of sulfur removal was studied. The elemental sulfur extracted from coal occurred in the form of S6, S7 and S8. Calibration solutions were prepared from freshly recrystalized elemental sulfur. It was found that the injection temperature has a crucial influence on the m/z 64 ion chromatogram.
Introduction The determination of the various sulfur forms in coal is among the most important analytical parameters used to characterize coal. It is generally assumed that sulfur in coal occurs as pyritic sulfur, sulfate sulfur and organic sulfur. The fourth sulfur form which has been detected, i.e., elemental sulfur S0, is considered to be artificial. Duran et al. [1] and Stock et al. [2] showed that pristine coal samples are free from elemental sulfur and concluded that elemental sulfur is not a natural constituent of coal but rather a product of atmospheric oxidation of pyrite or bacteriological action. Elemental sulfur can be formed when coal is exposed to air, for example, during coal storage. Weathered coals can contain a significant amount of elemental sulfur, even up to 0.3 wt% [3]. However, relatively fresh coals are characterized by a few tens to a few hundreds mg/kg S0 [4]. There are some consequences when the coal studied contains elemental sulfur. This is connected to the fact
G. Gryglewicz · S. Gryglewicz () Institute of Chemistry and Technology of Petroleum and Coal, Wrocław University of Technology, ul. Gdańska 7/9, 50-344 Wrocław, Poland e-mail:
[email protected]
that the ISO methods [5] as well as ASTM methods [6] for sulfur forms analysis assume that three sulfur forms are present in coal. Thus, there are only the standard analytical procedures for determining the pyritic, sulfatic and organic sulfur. The latter form is determined in an indirect way, i.e., by subtracting the sum of pyritic and sulfatic sulfur from the total sulfur. If elemental sulfur is present in the coal, it is included in the so-called organic sulfur. As a result, an overestimation of the organic sulfur content takes place. A potential source of interference in the determination of sulfur forms is even more serious for coals which have been chemically altered [7] or microbiologically desulfurized [8]. Moreover, in the case of organic sulfur compounds (OSC) identification when the examined coal is subjected to a procedure involving elevated temperatures, the presence of elemental sulfur generates the uncertainties in the evaluation of the results. For example, it was found that elemental sulfur reacts with methanol [9] and isopropanol [10] under supercritical conditions leading to the formation of artefacts OSC which do not occur in the initial coal. Since the elemental sulfur can not be considered by standards there have been several attempts to determine quantitatively the elemental sulfur in coal and coal derived materials. Most of them involve chromatographic techniques. Generally, the procedures for the determination of elemental sulfur in coal and environmental samples comprise the extraction of elemental sulfur with a solvent and subsequently, its quantification. For the latter purpose Duran et al. [1] used gas chromatography coupled with a Hall electrolytic conductivity detector with a high selectivity for sulfur. They examined the coal samples from the Illinois Basin Coal Sample Program (ICB) and the Argonne Premium Coal Sample Program (APCS). High performance liquid chromatography (HPLC) was applied by Buchanan et al. [11] for determining the elemental sulfur in the tetrachloroethene extract obtained from both ICB and APCS coals. Another technique was used by Louie et al. [12], i.e., supercritical fluid extraction (SFE), using CO2/10% methanol, with quantification by gas chromatography interfaced with atomic emission de-
61 Table 1 Coal samples
Table 2 Sulfur forms analysis, wt% (dry)
Sample
Type
Carbon wt% (daf)a
Total sulfur wt% (dry)b
Sample
Pyritic sulfur
Sulfatic sulfur
Organic sulfur
Bełchatów Siersza 1 Maj Szczygłowice
lignite sub A mvb refuse
61.5 76.1 87.5 85.3
8.50 3.00 4.30 1.81
Bełchatów Siersza 1 Maj Szczygłowice
4.08 2.42 2.36 0.10
0.59 0.14 0.27 1.44
3.83 0.44 1.67 0.27
a dry b dry
ash free basis basis
tection (GC/AED). Comparing to the methods based on conventional liquid solvent extraction SFE gives quantitative elemental sulfur removal in a shorter time. The aim of this work was to elaborate the method for the quantitative determination of elemental sulfur in coal by gas chromatography – mass spectrometry (GC/MS).
Experimental Materials Three Polish coals of different rank from lignite to medium volatile bituminous coal were selected for this investigation. The studied coal samples were not protected from exposure to the air. Additionally, the refuse from beneficiation of Szczygłowice coal was a subject of our interest. The latter sample enriched in pyrite has been remained in a contact with air for many years. The list of the samples is given in Table 1. All solvents used in this work were analyzed by GC/MS and found to be free of elemental sulfur contamination prior to their use. Sample preparation The external standard for the sulfur GC response was prepared using freshly recrystallized sulfur from cyclohexane. To evaluate the reapeatibility, a calibration curve was made for two series of standard solution. The A and B series were prepared from the solution containing 10 mg and 10.4 mg of sulfur in 50 mL cyclohexane, respectively. The initial standard solution was diluted in cyclohexane to desired concentration. Extraction A 2 g sample (particle size < 60 µm) was extracted in a Soxhlet apparatus for 1, 2 and 6 h. 90 mL of a solvent was used for the extraction of elemental sulfur. The obtained extract was transferred to a volumetric flask and diluted to 100 mL using the same solvent. Analysis Sulfur form analysis of coal was performed according to ISO Standards. Pyritic sulfur was determined indirectly via the pyritic iron and the sulfatic sulfur was determined directly in the hydrochloric acid extract obtained prior to the nitric acid extraction for the determination of pyritic sulfur. The Eschka method was used to determine the total sulfur content. The organic sulfur was calculated by subtraction of the inorganic sulfur from the total sulfur. A Hewlett-Packard HP 6890 gas chromatograph coupled with HP 5973 mass detector was used for quantitative determining the elemental sulfur extracted from coal samples. An HP-5 capillary
column (30 m × 0.25 mm I.D., 0.25 µm film thickness, Crosslinked 5% PH ME Siloxane) was used with He of 0.6 mL min–1 as the carrier gas. To get an optimum response from the GC/MS analysis for elemental sulfur, several combination of GC parameter settings were tested. Extracts and standards were analyzed using 5 µm injection at a split of 1 : 10. The mass spectrometer was set at an ionizing voltage of 70 eV.
Results and discussion The results of ISO forms of sulfur analysis are given in Table 2. The coal samples selected for this study contain a total sulfur in the range of 1.81–8.50 wt%. They are characterized by different distribution of sulfur forms. For the refuse from Szczygłowice coal the sulfatic sulfur has the highest contribution which confirms that this sample was exposed to the atmosphere for a long time. The pyritic sulfur predominates in the other coal samples. Solvent selection A number of solvents were tested to choose the solvent with the highest S0 extractability. Table 3 presents the results obtained for Szczygłowice refuse which was extracted for 6 h. The response of the GC/MS to S8 was used to quantify the elemental sulfur. The quantity of sulfur extracted is expressed as a percentage of the mass of coal sample subjected to extraction. Cyclohexane, ethanol and tetrachloroethene were found to give the highest effectiveness in the extraction of elemental sulfur. The extraction with cyclohexane yielded 0.111 wt% of elemental sulfur. The comparable quantity of S0, i.e., 0.105 and 0.098 wt%, was obtained for tetrachloroethene and ethanol, respectively. Cyclohexane and tetrachloroethene are used for the extraction of elemental sulfur in the following part. Table 3 Effect of the kind of solvent on the elemental sulfur content of Szczygłowice refuse measured
Solvent
Elemental sulfur, wt%
Acetone Cyclohexane Ethanol Hexane Methanol Tetrachloroethene Tetrahydrofurane Toluene
0.060 0.111 0.098 0.059 0.055 0.105 0.026 0.056
62 Table 4 Effect of extraction time on the elemental sulfur content of Szczygłowice refuse measured Time h
Elemental sulfur wt%
Cyclohexane
1 3 6 10
0.093 0.067 0.111 0.108
1 3 6 10
0.024 0.076 0.105 0.044
Tetrachloroethene
b
3.62
S6
S8
7.32 S7
Response
Solvent
15.12
15.11
a
Extraction conditions The effect of extraction time on the quantity of the S0 extracted using cyclohexane and tetrachloroethene was studied. The area under S8 peak was used to quantify the elemental sulfur extracted. As given in Table 4, the quantity of S0 extracted increases with time reaching a maximum, i.e., 0.111 wt% for cyclohexane and 0.105 wt% for tetrachloroethene, after 6 h of the process. However, a prolongation of extraction process to 10 h results in lowering the S0 content in the tetrachloroehene extract, from 0.105 to 0.044 wt% whereas for cyclohexane, the value remains the same. This indicates that for the S0 extraction, cyclohexane is more suitable than tetrachloroethene. The latter has a higher boiling point, thus, the extraction process is carried out at higher temperature compared with cyclohexane (121.3 vs 80.7 °C). Moreover, Buchanan et al. [9] reported that long Soxhlet extraction times with tetrachloroethene lead to elemental sulfur loss due to reaction between elemental sulfur and the coal matrix. The results obtained here indicate that at the boiling temperature of cyclohexane, an uptake of elemental sulfur by the coal does not occur at reflux for 10 h. No difference in the S6 : S7 : S8 ratio vs extraction time was observed. This suggests that all the allotropic sulfur forms detected are extracted with the same kinetics. Conditions of GC/MS analysis A several parameter sets were tested to optimize the conditions of chromatographic analysis for determining the elemental sulfur. It was found that an optimum GC-MS response is observed when both the injection and column temperatures are the same, 180 °C. Figure 1 a shows an example of a mass chromatogram of m/z 64 for the standard solution of elemental sulfur in cyclohexane. It can be seen that upon injecting the S0 standard, three peaks were detected which are attributed to the allotropic forms of sulfur, i.e., S6, S7 and S8. However, their relative proportion varies with the change of the inlet temperature as shown in Table 5. An increase of injection temperature, from 180°C to 270°C resulted in a change of S8 : S7 : S6 proportion from 6.4 : 1 : 2.5 to 1.6 : 1 : 1.9. Mass spectra
S8
3.61
S6
7.30 S7 Retention time, min →
Fig. 1 Mass chromatogram of m/z 64 of a, standard solution of elemental sulfur in cyclohexane; b, cyclohexane extract from Szczygłowice refuse Table 5 Effect of inlet temperature on the proportion of allotropic forms determined by GC/MS, % Inlet temperature, °C
S6
S7
S8
180 210 240 270
25.6 29.4 36.6 42.0
10.1 15.6 19.4 22.0
64.3 55.0 44.0 36.0
for S6, S7 and S8 indicate that these species undergo fragmentation mainly to m/z 64. Thus, a mass ion of m/z 64 was monitored to quantify the elemental sulfur. The established GC parameters setting are an optimal. Due to low volatility of sulfur, the column temperature below 180°C was too low. When the oven temperature was higher than 180°C, inseparable signals for S6, S7 and S8 were obtained. Probably, this is connected with a dynamic secondary dissociation of sulfur molecules which takes place in the column. A visible effect of this phenomena is the raising background observed. The quantity of elemental sulfur extracted from coal was determined on the base of the calibration curve prepared for standard solution of elemental sulfur in cyclohexane. Figure 2 shows the calibration curve expressed as the S8 peak area of the m/z 64 ion vs sulfur concentration for two standard series indicating high repeatability of MS response. For A and B series, the correlation coefficient of 0.9998 and 0.9987 was obtained, respectively. Moreover, this Figure shows the calibration curves expressed as the S7 and S6 peak areas of the m/z 64 ion vs sulfur concen-
63 Table 7 Sulfur forms analysis including elemental sulfur, wt% (dry)
Fig. 2 Calibration curve determined on the base of: +, S8 for A standard serie; , S8; , S7 and , S6 for B standard serie Table 6 Elemental sulfur content determined on the base of calibration curve for different allotropic forms of sulfur, wt% (dry) Sample
S6
S7
S8
S6+ S7+S8
Bełchatów Siersza 1 Maj Szczygłowice
0.100 0.009 0.126 0.104
0.082 0.017 0.128 0.132
0.071 0.021 0.094 0.111
0.092 0.019 0.121 0.145
Sample
Total sulfur
Pyritic sulfur
Sulfatic Elemental Organic sulfur sulfur sulfur
Bełchatów Siersza 1 Maj Szczygłowice
8.50 3.00 4.30 1.81
4.08 2.42 2.36 0.10
0.59 0.14 0.27 1.44
0.09 0.02 0.12 0.14
3.74 0.42 1.55 0.13
duplicate measurements of elemental sulfur in coal was < 7%. Finally, the results for sulfur forms analysis of the coal studied including the elemental sulfur are summarized in Table 7. Generally, the contribution of elemental sulfur in the total sulfur of coal studied reaches a few percent. This implies an overestimation of organic sulfur when analysis of sulfur forms is performed without determination of elemental sulfur.
Conclusions Extraction with cyclohexane combined with quantitative analysis of the extracted elemental sulfur by GC/MS is proposed as a promising method for the determination of elemental sulfur in coal and related materials. It was found that the sulfur extracted from coal occurs in three allotropic forms, i.e., S8, S7 and S6. To quantify the elemental sulfur, the S6 + S7 + S8 signal of the m/z 64 ion was used. The response of the GC-MS to allotropic forms of sulfur is dependent upon the injection temperature. With increasing injection temperature from 180 °C to 270 °C, a decrease of the S8 signal followed by an increased response to both S6 and S7 was observed.
Fig. 3 Calibration curve determined on the base of the sum (S6 + S7 + S8)
tration for the B standard, which are characterized by very high correlation coefficient, 0.9995 and 1.0000. Determination of elemental sulfur content in coal A typical mass chromatogram of m/z 64 ion for the cyclohexane extract of the coal is given in Fig. 1 b. Table 6 presents the amount of elemental sulfur determined on the base of the m/z 64 peak area, separately for S6, S7 and S8 . It can be seen that there are some differences between the values of elemental sulfur determination for different allotropic forms. Thus, to calculate the elemental sulfur content in coal, the calibration curve expressed as the sum of the m/z 64 ion area for S6, S7 and S8 was taken (Fig. 3). For this calibration curve a correlation coefficient of 0.9995 was obtained. Relative standard deviations of
References 1. Duran JE, Mahasay SR, Stock LM (1986) Fuel 65:1167–1168 2. Stock LM, Wolny R, Bal B (1989) Energy Fuels 3:651–661 3. Buchanan DH, Coombs KJ, Chaven C, Kruse CW, Hackley Kc (1990) In: Markuszewski R, Wheelock TD (eds) Processing and Utilization of High-Sulfur Coals III. Elsevier, Amsterdam, pp 79–87 4. Markuszewski R (1988) Coal Quality 7:1–4 5. ISO 334:1992; ISO 157:1996 6. ASTM D-2492, 1988; ASTM D-3177, 1988 7. Tsai SC (1982) Fundamentals of Coal Beneficiation and Utilization. Elsevier, Amsterdam, pp 353–363 8. Beyer M., Ebner HG, Assenmacher H, Frigge J (1987) Fuel 66:551–555 9. White CM, Rohar PC, Shaw LJ, Collins L (1996) Energy Fuels 10:1187–1188 10. Gryglewicz G, Rutkowski P (2001) Energy Fuels 15:8–13 11. Buchanan DH, Coombs KJ, Murphy PM (1993) Energy Fuels 7:219–221 12. Louie PKK, Timpe RC, Hawthorne SB, Miller DJ (1993) Fuel 72:225–231
