United States Industrial Environmental Research EPA-600/7-79-150
Environmental Protection Laboratory July 1979
Agency Research Triangle Park NC 2771 1
Coal Sulfur Measurements
Interagency
Energy/Environment
R&D Program Report
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the INTERAGENCY ENERGY-ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports m this series result from the
effort funded under the 17-agency Federal Energy/Environment Research and
Development Program. These studies relate to EPA's mission to protect the public
health and welfare from adverse effects of pollutants associated with energy sys-
tems. The goal of the Program is to assure the rapid development of domestic
energy supplies in an environmentally-compatible manner by providing the nec-
essary environmental data and control technology. Investigations include analy-
ses of the transport of energy-related pollutants and their health and ecological
effects; assessments of, and development of, control technologies for energy
systems; and integrated assessments of a wideirange of energy-related environ-
mental issues.
EPA REVIEW NOTICE
This report has been reviewed by the participating Federal Agencies, and approved
for publication. Approval does not signify that the contents necessarily reflect
the views and policies of the Government, nor does mention of trade names or
commercial products constitute endorsement or recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/7-79-150
July 1979
Coal Sulfur Measurements
by
J. W. Hamersma and M. L Kraft
TRW Defense and Space Systems Group
One Space Park
Redondo Beach, California 90278
Contract No. 68-02-2165
Task No 203
Program Element No. INE624
EPA Project Officer. Frank E. Briden
Industrial Environmental Research Laboratory
Office of Energy, Minerals, and Industry
Research Triangle Park, NC 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, DC 20460
-------
ABSTRACT
A new technique for sulfur forms analysis based on low temperature
oxygen plasma ashing has been developed. In this method the low tempera-
ture plasma ash is analyzed by modified ASTM techniques after the organic
material has been selectively removed. The proposed procedure has been
tested on 25 coals and compared with ASTM analyses with excellent results.
The data indicate that it is significantly more accurate and precise than
ASTM D2492. A separate set of experiments has shown that it is also fea-
sible to determine organic sulfur directly by trapping SO in the plasma
^
ash effluent. Development of the latter procedure was beyond the scope of
the task.
-------
CONTENTS
Abstract ii
Figures v
Tables vi
Acknowledgment vii
1. Conclusions 1
2. Recommendations 3
3. Introduction 4
4. Results 10
4.1 Coal Sulfur Forms Analysis with indirect
Determination of Organic Sulfur 10
4.1.1 Reactivity of Pyrite, FeS2 and
Other Naturally Occurring Sulfides ... 10
4.1.2 Nonretention of Organic Sulfur by
Alkaline Ash Constituents 13
4.1.3 Optimization of Plasma Ashing
Procedure - Parameter Verification
Studies 14
4.1.4 Optimization of Plasma Ashing
Procedure - Interferences Due
to Iron and Nitrate 16
4.1.5 Preliminary Evaluation of Plasma
Ashing Procedure 17
4.1.6 Comparative Evaluation of Proposed
Procedure with ASTM Methods 19
4.1.7 Comparative Evaluation of
Inorganic Speciation 26
4.1.8 Electron Probe Microanalysis of Coal
Extracted by ASTM Procedure 28
4.2 Sulfur Speciation in Liquified Coal Samples 30
4.3 Direct Organic Sulfur Determination by Plasma
Ashinq and SO Sorption 31
iii
-------
CONTENTS
4.3.1 Introduction 33
4.3.2 SO Sorptlon - Experimental Procedure. . . 34
A
4.3.3 SO Sorption Studies - Results and
Discussions 37
5. Procedures for Coal Sulfur Forms Analysis Via Oxygen
Plasma Ashing 41
5.1 Introduction 41
5.2 Plasma Ashing Procedure 41
5.2.1 Equipment 42
5.2.2 Procedure 42
5.3 Inorganic Sulfur Forms 42
5.3.1 Procedure 42
5.3.2 Pyritic Sulfur 43
5.4 Total Inorganic Sulfur 44
5.4.1 Equipment 44
5.4.2 Reagents 44
5.4.3 Procedure 44
5.5 Total Sulfur 45
5.6 Moisture 45
5.7 Calculations 45
5.7.1 Sulfate Sulfur 45
5.7.2 Pyritic Sulfur - Atomic Absorption .... 45
5.7.3 Organic Sulfur 45
5.7.4 Moisture 46
References 47
Appendix A: Literature Search Coal Sulfur Analysis 50
1v
-------
FIGURES
Number Page
1 Sulfur Forms in Coal 5
2 ASTM Method for Sulfur Forms Determination 5
3 Error Sources in ASTM Method D2492 8
4 IPC Oxygen Research Plasma Instrument 35
5 Experimental Arrangement for 'Evaluation of Solid Sorbents ... 35
6 Solid Sorbent Canister 36
7 Solid Sorbent Canister Placed in Vacuum Line 36
-------
TABLES
Number Page
1 ASTM Precision Requirements for Sulfur
Forms Analysis' 7
2 Reactivity of Sulfide Minerals Under
Ashing Conditions 11
3 Possible Retention of Organic Sulfur in
Plasma Ash 13
4 Total Inorganic Sulfur Analysis-Parametric
Procedure Verification Studies 15
5 Comparison of Gravimetric Sulfate Procedures 18
6 Comparison of Organic Sulfur Determinations
(% w/w by Difference) 20
7 Comparison of Sulfur Forms Analysis Methods 22
8 Total Inorganic Sulfur Comparison 23
9 Organic Sulfur 24
10 Comparison of Inorganic Specification Methods 27
11 Electron Probe Microanalysis of HC1/HN03
Extracted Coals . . 29
12 Sulfur Forms Analysis of Liquified Coal 32
13 Atmospheric Absorption Studies 37
14 Evaluation of Solid Sorbents for S0y Sorption
Under Plasma Conditons 39
vi
-------
ACKNOWLEDGMENT
This document constitutes the final report for the work accomplished
on Task 10, Coal Sulfur Measurements on EPA Contract No. 68-02-2165,
Sampling and Analysis of "Reduced" and Oxidized" Species in Process
Streams. The Chemistry Department of the Chemistry and Chemical Engineer-
ing Laboratory, Applied Technology Division, was responsible for the work
performed on this task. Tne work was initiated under the direction of
Dr. R. M. Statnick and completed under the direction of Mr. F.E. Briden,
EPA-IERL Research Triangle Park, North Carolina. Dr. R.F. Maddalone was
the Program Manager and the Task Order Manager was Dr. J.W. Hamersma. The
laboratory development and testing was performed by M.L. Kraft, W.F. Wright
and R.T. Ikemoto.
-------
1. CONCLUSIONS
A new technique for sulfur forms analysis has been developed in
response to a long-term need by the coal cleaning community for a more
accurate and precise procedure. In this method, low temperature oxygen
plasma ashing is used to selectively remove the organic matrix so that
inorganic sulfur can be extracted and analyzed without interferences
associated with the organic matrix. The inorganic sulfur can be analyzed
for sulfur forms by a modified ASTM technique or total inorganic sulfur
by a highly accurate BaSO. gravimetric procedure. Specific findings are
listed below.
Plasma ashing is greater than 99% specific to the organic
matrix as shown by separate oxidation studies on FeSp(pyrite).
FeS, PbS, and ZnS. The converted sulfide minerals are oxi-
dized to sulfate. Thus, sulfur does not leave the system.
Alkaline components such as CaCO^ or CaO do not retain organic
sulfur even when present at the 6% w/w level.
0 The presence of iron and large amounts of nitrate does not
interfere with the BaS04 procedure for total inorganic sulfur.
The standard procedure to remove these possible interferences
results in reduced recoveries and an increase in the variance.
The pH in the BaS04 precipitation must be adjusted to 1 to
avoid excessive solubilization of BaSO,. A pH greater than 1
can result in the precipitation of iron in certain cases.
0 Sulfur forms analysis can be made on liquified coal which can-
not be analyzed by the ASTM procedure.
A detailed examination of the precision and accuracy of the proposed
method with the ASTM D2492, the proposed ASTM atomic absorption (AA)
modification to D2492, the TRW AA method, the Leco combustion method for
total sulfur, and the new plasma ashing method has shown that:
Total inorganic sulfur can be determined with a precision of
±0.02% as compared to ±0.05-0.08% for the ASTM procedure.
0 The precision for proposed ASTM AA modifications of the ASTM
D2492 for pyritic iron is 0.06-0.07% as compared to 0.01-0.02%
for the TRW-AA procedure. The TRW method uses a less sensitive
iron line and includes a short curve method for instrument cali-
bration.
-------
The Leco combustion method gives results equivalent to the
Eschka method but with a reduction in precision from ±0.02-
0.03% to ±0.06%.
Discrepancies between the actual values obtained by the ASTM
method and the plasma ashing method have been resolved in
favor of the plasma ashing method.
Agreement between the ASTM-AA method and especially the ASTM-
TRW-AA method is generally better than with the ASTM D2492.
The feasibility of a direct organic sulfur analysis by plasma oxida-
tion with subsequent trapping of the organic SOX species on solid sorbents
placed in the exit port of the plasma reaction chamber has also been
demonstrated.
-------
2. RECOMMENDATIONS
The following recommendations are made:
t The oxygen plasma coal sulfur forms analysis procedure should
be further tested under "round robin" conditions to determine
variability between laboratories.
A larger cross section of coal regions and seams should be
analyzed in replicate with results compared to ASTM D2492.
Additional liquified coal samples, and samples obtained from
other types of coal desulfurization systems should be analyzed
by both the plasma and ASTM procedure (where applicable) and
results compared.
A separate program should be initiated to develop the direct
determination of organic sulfur using the plasma technique
and the concept of solid sorbents. The effort should be scoped
to include hardware as well as methods development/optimization
and also investigation of other organically bound species such
as nitrogen and halogens.
-------
3. INTRODUCTION
The need to reduce SO emissions from the burning of coal as mandated
A
by the Clean Air Act has stimulated strong interest in the development of
new or improved methods of physical and chemical coal cleaning. Many of
these methods are specific to either the inorganic or organic form of
sulfur in coal. As these processes are optimized or developed, it becomes
increasingly more important to have an accurate and precise sulfur forms
analysis to delineate the effects of process modifications on sulfur
removal. The present ASTM procedure, D2492, for sulfur forms determines
total sulfate and pyritic sulfurJ Organic sulfur is determined by sub-
traction of the inorganic sulfur forms from the total sulfur. As a
result, the organic sulfur value reflects the errors of all three analyses.
A survey of the existing literature and recognized authorities in
the fields of coal cleaning and analysis indicated a general dissatisfac-
tion with the present procedures. The results of the literature survey
are documented in detail in Appendix A. The responses from workers in
this field ranged from "anything is better" to the need for ±1% relative
precision and accuracy for kinetic studies. Based on this survey, it
was determined that the following would be desirable in a procedure.
Increase the accuracy and precision of the inorganic sulfur
extraction by removing all inorganic sulfur forms by a single
ash extraction. This reduces coal matrix effects and the
number of manipulations.
Increase the accuracy and precision of the inorganic sulfur
determination by using a single BaS04 gravimetric procedure.
This procedure is very well developed and capable of very
high precision and accuracy.
Retain the capability of inorganic sulfur forms analysis.
Develop a method for direct analysis of organic sulfur.
Figure 1 shows the sulfur forms present in coal and Figure 2 shows
the sequence used in the ASTM procedures. Total sulfur is determined by
the Eschka method ASTM D271 and pyritic and sulfate sulfur is determined
by ASTM method D2492. The Eschka procedure is highly accurate and precise
-------
Sulfate (FeSOJ - Present in most coals as a result of oxidation
(weathering of pyrite).
Pyrite (FeS-) Present in minor to major amounts in all coals.
Found both very finely dispersed and in occlusions,
Sulfide (FeS) Found only in those coals that have been treated
in a reducing atmosphere and in certain Illinois
coals.
Organic Sulfur Organsulfur compounds present in all coal
Figure 1. Sulfur Forms in Coal
Determine total sulfur
t Extract FeSO. with HC1 and determine sulfur gravimetrically
t Extract Fe$2 with HN03 and determine iron by dichromate or
permanganate titration
t Determine organic sulfur by difference
Figure 2. ASTM Method for Sulfur Forms Determination
-------
and for these reasons there is a general satisfaction with this procedure.
The actual sulfur forms procedure, however, suffers from several sources
of error which affect accuracy and precision which are tacitly recognized
in the ASTM "precision" requirements listed in Table 1. The cumulative
effect of these requirements for duplicate organic sulfur analyses is that
they can be considered acceptable if they are with 0.01-0.2% when per-
formed in the same laboratory or within 0.3-0.6% when performed in differ-
ent laboratoriesJ TRW's experience with the Meyers Process is consistent
with these results. A pooled standard deviation of ±0.09% for the organic
sulfur determination was calculated for the analysis of 69 sets of coals
in triplicate.^
Of the sources of error in ASTM Method D2492 listed in Figure 3,
matrix effects in the pyritic sulfur determination are a major problem,
which contributes greatly to the limited reproducibility of the method,
as given in Table 1. The nitric acid extraction often removes substantial
amounts of organic material. If not destroyed, this material is oxidized
by dichromate or permanganate titrant in the determination of pyritic iron
yielding a high result. (Pyritic iron must be determined because the
nitric acid often removes some organic sulfur.) The manipulations to do
this result in reduced precision. Often color still remains which results
in a fading and indistinct endpoint requiring a high degree of operator
experience and judgment. This problem is tacitly recognized by greatly
expanded precision limits in Column 3, Table 1, where results between oper-
ations are compared.
A fundamental and more serious problem Is many coals contain finely
dispersed (< 5y) pyrlte which Is In the form of single crystals or In
spherical crystal assemblies (frambolds) . These particles may be occluded
in coal particles or enclosed in a kaolin lattice which can retard or pre-
vent their dissolution In the standard / TM add digestion procedure. This
type of pyrite may be as much as 80% ot he total pyrite 1n some coals which
can cause incomplete or marginal extraction efficiency. Thus, experimental
variables such as additional HNO, extraction time and additional or a differ-
ent type of grinding of the coal can significantly affect the results
6
-------
Table 1
ASTM Precision Requirements for Sulfur
Forms AnalysisU)
Sulfur Form
Total
Less than 2%
More than 2%
Sulfate^
Pyritic
Less than 2%
More than 2%
Organic (difference)
Worst Case
Best Case
Repeatability'^)
0.05
0.10
0.02
0.05
0.10
0.22
0.12
Reproducibility(c)
0.10
0.20
0.04
'
0.30
0.40
0.34
0.64
(a)Total sulfur, ASTM D271; sulfate and pyritic sulfur,
ASTM D2492.1
(b)Consecutive determinations on the same sample, in the same
laboratory, with the same operator and apparatus.
(c)Separate determinations carried out on representative samples
in different laboratories.
(^Usually less than 0.2% in freshly mined coal.
-------
HC1 Extraction
Incomplete non-pyritic iron extraction which is extracted
in the next step
HN03 Extraction for Pyrite, Fe$2
Incomplete extraction FeS^ due to matrix effects
X^ j
Incomplete washing of coal to remove Fe
Organic matter extracted from coal can interfere with
end point of titration
Addition of excess SnCl2 to reduce Fe to Fe can
interfere with final titration
If all coloration is not destroyed, the final end point
may fade or be difficult to observe
Figure 3. Error Sources in ASTM Method D2492
by changing the amount of pyrite extracted. Coal ground to 60 mesh x 0
(ASTM specification) has a maximum particle size of 250p and it is very
difficult to grind samples sufficiently to ensure total pyrite exposure.
The net result Is that unextracted pyrite 1s counted as organic sulfur.
The experience of TRW and the Illinois Geological Survey (IGS)* Indicates
that this 1s a common problem with Eastern Interior Basin coals although 1t
also occurs with coals from other regions. The significance of this problem
can be put in perspective when it is realized that all processes designed
to remove organic sulfur also remove inorganic sulfur. Thus, if a
significant portion of the "organic" sulfur is actually pyrite, the
calculated organic sulfur removal in a given process will be signifi-
cantly biased to the high side.
In order to alleviate these problems while maintaining the basic
concepts of the ASTM methods, TRW has proposed a plasma ashing pro-
cedure which is the subject of this report. This method depends on the
selective removal of the organic coal matrix including the organic sulfur
via a low temperature oxygen plasma. This has the following advantages:
*Private communication with H. J. Kuhn, Illinois Geological Survey.
8
-------
Up to 90% of all matrix effects are eliminated.
Very fine encapsulated pyrlte 1s now available for extraction.
A single HN03 extraction will remove all inorganic sulfur
species that can be- determined in a single analysis.
The absence of organic sulfur in the ash allows the use of
the much more accurate and precise BaSCty gravimetric
procedure.
ASTM D2492 can still be applied to the ash for separate sul-
fate and pyrite determinations.
Measuring or trapping the SOX species in the plasma asher exit
gases allows for the possibility of the direct determination
of organic sulfur.
The following sections describe the verification of this method for
the direct determination of inorgan-ic sulfur, inorganic sulfur forms, and
the indirect determination of organic sulfur. Also included is a discus-
sion of the successful proof-of-principle experiments for the direct deter-
mination of organic sulfur, and a text of procedures for sulfur forms
measurement.
Appendix A contains a detailed report of the literature survey.
-------
4. RESULTS
The results discussed in this section are divided into three parts.
\
The first deals with the verification of the oxygen plasma procedure and
the verification of the sulfur forms method, including organic sulfur by
difference. The second part describes the application of the method
described in Section 4.1 to liquified coal samples, and the third, Sec-
tion 4.3, describes the proof-of-principle experimentation for the direct
determination of organic sulfur.
4.1 COAL SULFUR FORMS ANALYSIS WITH INDIRECT DETERMINATION OF ORGANIC
SULFUR
This analysis procedure is based on the selective removal of the
organic matrix including organosulfur species from coal with a low tempera-
ture oxygen plasma. The inorganic sulfur species are quantitatively
retained in the remaining ash. The inorganic sulfur in the ash is then
extracted with dilute nitric acid and determined gravimetrically as BaS04
or speciated by ASTM D2492.
An evaluation of the data in the following sections shows that (1) the
plasma procedure yields more accurate and precise results for inorganic
and organic sulfur than the ASTM procedures; (2) naturally occurring sul-
fides are not affected under plasma conditions; and (3) organic sulfur
species are not retained in the coal ash by reaction with alkaline com-
pounds. The actual procedure has been optimized and is ready for testing
in other laboratories.
4.1.1 Reactivity of Pyrite, FeS2> and Other Naturally Occurring Sulfides
The major requirement of this method is the nonreactivity of pyrite
and other possible naturally occurring sulfides to plasma oxidation. These
can be oxidized to either the oxide or the sulfate. If oxidation to the
oxide occurs, sulfur is lost resulting in high organic sulfur values and
low inorganic sulfur values. Extensive oxidation to the oxide (>5%) could
result in the failure of this procedure because of the inaccuracies that
would be introduced. Oxidation to the sulfate causes the loss of one-half
of the pyritic sulfur, but no loss to the other sulfide minerals. Although
10
-------
this is not as serious, in most cases, the sulfur form is changed and the
viability of the inorganic sulfur forms procedure would be compromised.
Although both reactions are thermodynamically possible, preliminary evi-
dence indicated that the reaction was kinetically controlled in that little
or no conversion occurred. This conversion would be expected to proceed
most rapidly at smaller particle sizes due to the larger surface area and
possible heat buildup. Thus, it was considered essential to test FeS,,,
FeS, PbS, and ZnS under actual plasma conditions using size fractions
representing the lower size limit expected to be found in coal. Two
different size fractions representing average particle sizes of approxi-
mately 3 and 9 microns were prepared by hand grinding under a nitrogen
blanket. The actual average particle sizes shown in Table 2 were deter-
mined using a Fisher sub-siever sizer. Two hundred milligrams of each
particle size were spread in a Petri dish and "ashed" for 3 days at
100 watts power, and 200 cc/min 03, which is identical to the conditions
used to decompose coal. At the end of 3 days, the samples were removed,
Table 2. Reactivity of Sulfide Minerals Under
Ashing Conditions
Compound
PbS
FeS
ZnS
FeS2
Average
Particle Size
y
3.5
9.0
2.5
3.1
3.1
8.2
2.3
10.0
Wt Change
+1.2
+0.6
+4.5
+2.0
+1.0
+1.0
"(c)
%S
Converted to
S04=
+0.06
+0.08
+0.54
+0.20
+0.69
+0.25
+0.1
No change
o/c
/oil
Converted to
oxide(a)
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.(b)
(a)N.D. = Not detected by X-ray diffraction, detection limit 1-2%.
(bhn carbon matrix to allow for even dispersion (see text).
(c)carbon matrix does not allow computation of this value.
11
-------
cooled in a desiccator, weighed (conversion to sulfate would result in
weight gain and conversion to oxide in weight loss), the sulfate content
determined turbidimetrically on a hot water extract of a weighed aliquot
and the oxide formation investigated using an X-ray diffraction technique.
Results are summarized in Table 2.
A small conversion to sulfate was obtained with FeS and ZnS with the
percentage increase inversely proportional to the average particle size.
In no case did the conversion to sulfate exceed 1% of total compound
exposed. As conversion to sulfate does not result in any sulfur loss for
these two compounds, there are no negative or positive analytical biases.
In the case of Fe$2, however, the neat 2.3y sample showed evidence of
ignition to Ffip^s 1n tnose arefls where the dispersed sample formed small
piles. In areas where the Fe$2 was smoothly dispersed no visible reaction
occurred. Dispersal of the 0.2 g pyrite in 1 g of carbon (10% pyritic
sulfur) completely eliminated this problem. This indicates that when
extremely fine pyrite is present, the temperature of the ashing must be
controlled such that autoignition cannot occur and that dispersal in a
carbon matrix is sufficient to prevent this oxidation from occurring.
Conversion to oxides or other compounds was checked by X-ray diffrac-
tion. One hundred milligrams of plasma ashed minerals were dispersed
ultrasonically in a collodian matrix and dispersed as a thin film on a
glass slide. These samples were then analyzed by X-ray diffraction and
the spectra compared to the ASTM powder diffraction file. Operating
parameters were as follows:
GE XRD-5 Operating Parameters
Cu tube - 50 KVP, 20 ma
3° beam
Medium resolution seller slit
0.1° slit, Ni filter
Scan speed 2°/min
Proportional detector
Range setting linear, 500 cps
12
-------
In no case was any change noted either as to shifts or appearance of
additional lines that might be attributed to formation of oxides. An
approximate detection limit of 1-2% for the respective oxides was estab-
lished for each compound by adding a known weight of oxide and determining
response.
4.1.2 Nonretention of Organic Sulfur by Alkaline Ash Constituents
Coal ash often contains considerable amounts of alkaline materials
such as calcium carbonate, CaCCu and calcium oxide, CaO. These materials
are used extensively in SO scrubbing media. Thus, it was considered
A
essential to investigate the possibility of organic sulfur retention by
these materials. In order to do this with the highest degree of precision
and accuracy, demineralized coal containing only organic sulfur was used.
These samples, #C-18572 and K-18844 obtained from the Illinois State
Geological Survey (IGS), had organic sulfur contents of 1.81% and 1.17%,
respectively. They were doped with 3% and 6% w/w CaCCu, or CaO, and ashed
under standard conditions (3 days, 100 watts rf power, and 200 cc/min
oxygen flow). After ashing, they were extracted with 1:1 HC1 and sulfate
was determined turbidimetrically. The results in Table 3 show that in
no case did the sulfate content increase as a result of the ashing pro-
cedure. Thus, SO retention can be eliminated as an interference in this
/\
procedure.
Table 3. Possible Retention of Organic Sulfur in
Plasma Ash
Coal
C-18844B
C-18572B
Initial
Sulfate
Analysis
0.05
0.10
Plasma Ashed Sulfate Analysis
CaC03, % w/w
3%
0.03
0.08
6%
0.05
0.07
CaO, % w/w
3%
0.03
0.08
6%
0.07
0.05
13
-------
4.1.3 Optimization of Plasma Ashing Procedure - Parameter Verification
Studies ~~
The size of the container used in the plasma ashing step and the
gravimetric procedure used for sulfate determination for the initial stud-
ies described in Section 4.1.5 were chosen because they had given satis-
factory results in past studies. Although this procedure appeared satis-
factory, it was essential that further parametric verifications be
conducted so that sensitive areas, if any, in the procedure could be
identified and corrected. This section evaluates (1) the effect of pH on
the precipitation of BaSO^, (2) the effect of container size, and (3) the
possible use of the entire Eschka procedure on the plasma ash for increased
precision and accuracy.
Because BaS04 has a slight solubility in acid solutions, it is neces-
sary to neutralize most of the HN03 present in the acid extract of the ash.
Sufficient acid must be present, however, to prevent the coprecipitation
of iron. Prior experience indicated that a pH of approximately 1 is suffi-
cient to allow quantitative precipitation of BaSO^ and prevent interferences
from iron. It was the purpose of this experimentation to more closely define
the pH adjustment step. This experimentation was performed in parallel
with experimentation on the effect of the size of ashing containers on
sample recovery. Purging and evacuation of the plasma reaction chamber
can disturb ash if care is not taken. A larger ashing container could
alleviate this problem. In addition, a separate series of samples was
analyzed using the Eschka technique. With the latter technique, the
resulting ash was transferred quantitatively from the containers, mixed
with Eschka mixture, and analyzed by ASTM D271.
Quadruplicate samples were used for each condition studies. The
65 x 35 mm ashing containers were fabricated from 250 ml Pyrex beakers.
These containers were also used for the BaSO, precipitation pH studies.
The results of these studies, summarized in Table 4, show that at pH 2
significant amounts of iron oxides can be precipitated and invalidate
the results. Thus, it is important to carefully adjust the pH to 1 in
order to prevent excessive solubilization of BaS04 at a lower pH and inter-
ferences caused by coprecipitation of iron at a higher pH. Comparison
14
-------
Table 4. Total Inorganic Sulfur Analysis-Parametric
Procedure Verification Studies
Coal Sample
Marti nka Mine
Lower
Kittanning
Seam
Delmont Mine,
Upper Freport
Seam
Total Inorganic Sulfur, % w/w'a)
Average
Standard
Deviation
% RSD(b)
Average
Standard
Deviation
% RSD
60x15 mm
Container
pH 1
2.59
2.59
2.57
65x35 mm
Container
pH 1
2.58
2.59
2.55
2.54 2.54
2.57
±0.024
0.93
4.14
4.02
4.07
4.06
4.07
±0.050
1.23
2.56
±0.024
0.94
4.05
4.23
4.09
4.08
4.11
±0.080
1.95
65x35 mm
Container ;
pH 2
Iron precipitate
formed analysis
disconti nued
!
Eschka Procedure^0'
3.96
4.14
4.10
3.98
4.04
±0.090
2.22
except where noted,
All samples extracted with 1:7
'b'RSD = Relative Standard Deviation
^C'Ash analyzed by standard Eschka procedure, ASTM D271 .
15
-------
of results between the two sizes of containers showed no significant dif-
ference In total Inorganic sulfur found. Ashing time was found to Increase
with use of high form containers, hence the use of standard petrl dishes Is
recommended. The Eschka combustion procedure (ASTM D271) was used on a
separate set of plasma ashed Delmont coal samples. These results were com-
pared to the pH 1 data for the same coal 1n Table 4 and found to be signif-
icantly less precise. In addition, the % relative standard deviation for
the Eschka procedure is substantially poorer than with any other data set
in Table 4. The reason for this poor precision may be due to inherent
difficulties in transferring fine, dry ash into the Eschka mixture by
brushing.
4.1.4 Optimization of Plasma Ashing Procedure - Interferences Due to Iron
and Nitrate
The procedure used 1n Section 4.1.5 Involves the precipitation of
XA-Ai
BaSO. In the presence of significant amounts of Fe and N03" Ions. These
Ions can be troublesome because of the possibility of coprec1p1tat1on of
Fe2(S04)3 and Fe(N03)3.* On Ignition Fe2(S04)3 1s converted to Fe^,
Equation (2). This results 1n one equivalent weight of SO.* being con-
verted Into a gravimetric product with an equivalent weight of 53.2 Instead
of the equivalent weight of BaSO. of 233.4. This loss of weight results
1n low values. The copreclpltatlon of Fe(N03)3 and conversion to Fe203,
Equation (3), represents a weight gain and yields high values.
BaSO. 925 C^ BaSO.
4 ignition ^ (1)
E.W. 233.4 233.4
1/3 Fe?(SOJ, 925 C^ 1/3 Fe.O.-H 3S0
^ 4 3 ignition
E.W. 133.2 53-2
,
3 ignition (2)
Fe9(NO,), 925°C» Fe90-+ 3NO,
* 3 3 ignition < * * (3)
E.W. 201.7 159-7
16
-------
The simple removal of iron by precipitation with NH.OH and
filtration is not possible because the presence of HN03 promotes the for-
mation of a very fine precipitate which is very difficult to filter. ASTM
E350 is a procedure which uses a HN03 dissolution followed by a BaS04 pre-
cipitation to analyze for sulfur in ferrous metals.5 In this procedure,
these interferences are removed first by evaporation of the solution to
dryness after neutralization with NaC03> and then redissolving the salts.
Fe+++ is reduced to Fe++ with zinc. The pH is adjusted to ^4, with methyl
orange as indicator, followed by the addition of BaC^.
This method was compared to the one which was in use by adding
aliquots of standardized Fe2(SO.)3 (dichromate titration of iron) solu-
tion to 25 ml of 1:7 HNO, and analyzing the resulting solutions by both
methods. Solution concentrations were chosen to simulate inorganic sulfur
concentrations in whole coal of -0.4 and 2.6%. In a separate experiment,
the effect of an additional 0.7% nonpyritic iron was evaluated by doping
the ferric sulfate solutions with ferric chloride. Results, which are
the averages of 10 determinations, are summarized in Table 5. These data
indicate that while both methods appear to be equally precise, recovery
for the simple precipitation at pH 1 method averaged 98 ±3% while the
more complex ASTM E-350 method gave average recoveries of 90 ±4%. For
the simplified method, only in the low sulfur case with added iron was
the recovery less than 90%, while the complex method yielded recoveries
of 83-84%. It is also estimated that the complex method requires three
times the labor of the simplified method. Thus, the simplified procedure
is being recommended for use.
4.1,5 Preliminary Evaluation of Plasma Ashing Procedure
Preliminary evaluation of the procedure was performed on 13 coals.
The coals chosen represent "problem coals" because of sulfur mass balance
problems found during Meyers' processing (TRW technique for removal of
t\
pyritic sulfur by treatment with ferric sulfate). As such, they repre-
sent a worst case test of the method. Organic sulfur values listed in
Table 1 were calculated by difference using three methods for inorganic
sulfur determination. For the whole coal extraction, ASTM D2492 was used
to determine both sulfate and pyritic iron. In addition to the titrimetric
17
-------
Table 5. Comparison of Gravimetric Sulfate Procedures
Test Conditions
Neat Fe2(S04)3
Fe2(S04)3 plus 0.13
mM FeCl3
Average
mM Sulfur*
in Aliquot
0.1202
±0.0006
0.801
±0.0040
0.1230
±0.0006
0.820
±0.0040
Simplified Method
pH/Adjustment
mM Sulfur
0.118
±0.0038
0.805
±0.0088
0.112
±0.0067
0.813
±0.012
5=0.0084
% Recovery
97
±3.2
100
±1.1
91
±5.4
99
±1.5
98 ±3
Complex Method
(ASTM -350)
mM Sulfur
0.100
±0.0078
0.742
±0.0097 .
0.110
±0.0036
0.770
±0.0097
S=0.008
% Recovery
83
±6.5
93
±1.2
89
±2.9
94
±1.2
90 ±4
CO
*As calculated from the iron content which was determined by dichromate titration
-------
determination of iron an atomic absorption method, developed at TRW and
described in Section 5, was also tested. Total inorganic sulfur in the
plasma ashed sample was determined by the BaSCK gravimetric procedure in
Section 5. All organic values were calculated by subtraction of the
appropriate values from a common total sulfur value.
The organic analysis results, Table 6, show that when the results
are evaluated by coal basins, the ASTM-TRW-AA* and plasma ashing proce-
dures give essentially identical results while the standard ASTM proce-
dure is low by 5-6%. In the case of the interior basin coals, the ASTM
procedures are in good agreement while the plasma procedure is low by
]Q%. According to Dr. Anthony of the Old Ben Coal Company, three of
these coals are from the same area and it has been demonstrated that they
contain pyritic sulfur that is unextractable by the ASTM procedures. If
this unextractable inorganic sulfur were extracted from the plasma ashed
sample, this would be the cause of the low organic number. Most of the
Appalachian coals in this table are also "problem" coals from Meyers'
processing. When before and after treatment values are compared, both
the plasma ashing and the ASTM-TRW-AA procedures give consistent results,
while the ASTM procedure gives widely discrepant results that tend to be
significantly low for untreated Appalach an coals. Thus, in reference
to evaluation of coal cleaning processes, it appears that the ASTM pro-
cedure gives results dependent on coal processing while the other two
procedures give results that are virtually independent of the history of
the coal.
4.1.6 Comparative Evaluation of Proposed Procedure with ASTM Methods
The comparative evaluation of the proposed procedure with ASTM methods
was performed on coals in triplicate. These were chosen to represent a
variety of ten seams in the Appalachian and Eastern Interior coal basins.
Each coal was analyzed by the Eschka method, ASTM D271, for total sulfur,
ASTM D2492 for sulfur forms, and the plasma ashing procedure in Appen-
dix B. Analysis of the HNCu solution obtained from extraction of the
*This designation is used to distinguish this method from a similar
method proposed by ASTM.
19
-------
Table 6. Comparison of Organic Sulfur Determinations
(% w/w by Difference)
Mi ne/Seam
E Seam (ROM)
E Seam (Cleaned)
Oelmont
(Upper Freeport)
Bird
(Lower Kittanning)
Egypt Valley
(Pittsburgh)
Fox
(Lower Kittanning)
Williams
(Pittsburgh)
Lucas
(Middle Kittanning)
Brookdale #77
(Lower Kittanning)
Old Ben #2rb'
(Illinois #6)
Old Ben #24(b)
(Illinois #6)
Old Ben #26^ b'
(Illinois #6)
Inland Stee/bJ
(Illinois #6)
Treated(T)
-------
coal was analyzed by the standard titrimetric procedure and a new atomic
absorption procedure which is being circulated for comment by ASTM. This
should be distinguished from the TRW atomic absorption procedure which uses
a different iron absorption line (Section 4.1.7). The results are tabulated
in Table 7. The total inorganic sulfur values are summarized in Tables 8
and 9; the organic sulfur values were calculated by difference.
From Table 7 it is seen that the ASTM titrimetric pyritic sulfur
values exhibit a distinct negative bias when compared to values deter-
mined by AA. The exact reason for this bias is difficult to explain
because the pyritic iron was determined on aliquots of the same extract
for both procedures. In add on, the same iron standard was used for
calibration of both ar: ytic procedures. One explanation is loss of
iron during the iron h^drox1 precipitation, filtration and redissolution
step wh is ecessary for tiie titrimetric procedure but not for the AA
determi: '.ion. This explanation is consistent when the total inorganic
values (i^-Ifate sulfur and pyritic sulfur) determined using the two ASTM
procedures are compared to the TRW plasma-total inorganic values (Table 8);
the results using the AA technique are in excellent agreement with the
values obtained in the plasma ash procedure for the Appalachian coals.
This agreement for total inorganic sulfur is directly reflected in the
organic sulfur values contained in Table 9. The ASTM analyses of Eastern
Interior Coals as represented by Old Ben No. 21, 16 and Inland Peabody
exhibit a positive bias for organic sulfur when compared with the plasma
procedure. This bias probably 1s the result of the presence of very fine
(< 5y) occluded pyrlte present 1n these coals which is not extracted with
nitric acid in the ASTM procedure. This is substantiated by the elec-
tron probe microanalysis studies described in Section 4.1.8 and by private
communications with A. V. Marse of Old Ben Coal Company and H. J. Kuhn
of the Illinois Geological Survey. The presence of very fine occluded
pyrite can also explain the large differences In total Inorganic sulfur
experienced with the tailings sample.
21
-------
Table 7. Comparison of Sulfur Forms Analysis Methods
Mine
Old Ben *21
Old Ben #26
Inland Pea body
Martlnka
E-Seam
C-Prime
Mathies
Lucas
Muskingum
Tailings from
Central Ohio
Coal Co.
Coal Seam
Illinois #6
Illinois *6
Illinois #5
L. Kittanning
U. Freeport
U. Kittanning
Pittsburg
M. Kittanning
Meiogs Creek 19
Meiggs Creek *9
Coal Basin
Eastern Interior
Eastern Interior
Eastern Interior
Appalachian
Appalachian
Appalachian
Appalachian
Appalachian
Appalachian
Appalachian
Pooled Std. Dev.
Total Sulfur
Eschka
1.11 tO. 032
2.26 +0.031
3.24 ±0.012
1.59 .'0.058
1.04 ±0.015
1.52 '0.032
1.60 ±0.020
1.70 ±0.025
5.71 ±0.051
3.28 ±0.035
0.034
Leco
1.28 +0.030
2.21 ±0.031
3.11 ±0.071
1.59 ±0.059
1.06 ±0.025
1.42 +0.040
1.52 ±0.075
1.69 ±0.044
5.62 ±0.061
3.31 +0.096
0.058
Pyritic Sulfura-?ASTM
Titration^
0.32 ±0.045
0.65 +0.036
0.94 ±0.028
1.16 +0.097
0.25 +0.010
0.74 ±0.040
0.57 '0.026
0.67 iO.006
1.87 +0.066
1.87 +0.074
0.051
AAW
0.48 '0.066
0.77 tO. 006
1.03 ±0.006
1.33 ±0.130
0.34 ±0.006
0.90 -0.007
0.66 +0.006
0.82 '0.006
2.42 +0.040
1.92 ±0.170
0.072
Sulfate
Sulfur'0'
ASTM
0.08 iO.OOO
0.16 +0.006
0.38 ±0.012
0.12 +0.006
0.08 +0.006
0.08 ±0.000
0.47 +0.021
0.44 '0.006
1.22 +0.092
0.02 ±0.000
0.030
Total Inorganic
TRW Plasma Ash
0.71 +0.011
1.25 +0.015
1.59 ±0.015
1.49 +0.015
0.40 +0.036
0.96 +0.006
1.20 +0.010
1.19 +0.011
3.33 ±0.040
2.52 ±0.035
0.023
ro
ro
(t)
extraction performed by ASTM D2492; both analysis were performed on same solution.
the proposed ASTM Atomic Absorption procedure was used.
(c)ASTM 02492.
-------
Table 8. Total Inorganic Sulfur Comparison
(a)
Coel
Old Ben #21
Old Ben #26
Inland Peabody
Martinka
E-Seam
C- Prime
Mathies
Lucas
Muskingum
Tailings from
Central Ohio Coal Co.
Pooled Standard
Deviation
ASTM Titration^
0.40
0.81
1.32
1.28
0.33
0.82
1.04
1.11
3.09
1.89
±0.059
} ASTMAA(C:
0.56
0.93
1.41
1.45
0.42
0.98
1.13
1.26
3.64
1.94
±0.078
i
TRW Plasma
0.71
1.25
1.59
1.49
0.40
0.96
1.20
1.19
3', 33
2.52
±0.023
'a'Sum of sulfate and pyritic sulfur for the ASTM methods and
determined directly in the plasma ashing method.
sulfate sulfur plus pyritic sulfur determined by titration
of pyritic iron.
^C'ASTM sulfate sulfur plus pyritic sulfur determined by the
proposed ASTM procedure for determination of iron by atomic
absorption.
23
-------
Table 9. Organic Sulfur
Coal
Old Ben #21
Old Ben #26
Inland Peabody
Martinka
E-Seam
C-Prime
Mathies
Lucas
Muskingum
Tailings from
Central Ohio Coal Co.
Pooled Standard Deviation
ASTM
Titration
0.71
1.45
1.92
0.31
0.71
0.70
0.56
0.59
2.62
1.39
±0.058
AA
0.55
1.33
1.83
0.14
0.62
0.54
0.47
0.44
2.07
1.34
±0. 085
TRW
Plasma
0.40
1.01
1.65
0.10
0.64
0.56
0.40
0.51
2.38
0.76
±0.041
The standard deviation for each set of triplicate analyses was
calculated and then used to calculate the standard deviation for total
inorganic sulfur and organic sulfur as follows:
Inorganic
V-
(1)
for ASTM D2492 and sulfur form on plasma ash,
'Organic
V
total °S04 °FeS2
(2)
24
-------
for ASTM D2492, and
^
+ 2 (3)
'Organic ~W "Total °Inorganic
for the plasma ashed sample with no inorganic sulfur speciation.
In addition, the pooled standard deviation for each set of 10 tri-
plicate samples was calculated. When a sufficient number of data sets
are used, this value gives an estimate of the precision that can be
expected from an analysis method on a long term basis. The formula for
this calculation is given in Equations (4) and (5). For organic sulfur,
the value calculated from either Equation (2) or (3) was used.
A 2 . * «2
(D {T + u) n
22 3 3
o =
where
o = standard deviation
4> = number of values used to calculate a.
Thus, when $ is the same for all o, Equation (4) reduces to
where
n = number of data sets.
As can be seen in Table 7, the pooled standard deviations for the
plasma total inorganic sulfur, Eschka total sulfur, and sulfate sulfur
analysis are similar. This is to be expected because all three are similar
gravimetric analyses. The lower value for the plasma inorganic most
25
-------
likely reflects the fewer manipulations and/or lack of coal matrix
effects 1n these samples. Table 8 summarizes the total inorganic sulfur
results calculated by all three methods. For the ASTM procedures, this
is the sum of the pyritic and sulfate sulfur. It is apparent from the
tabulation that the plasma ashing procedure offers a substantial increase
in precision over the other two methods.
Table 9 tabulates the organic sulfur corresponding to pooled stan-
dard deviation values calculated by all three methods. Although these
values are all "diluted" by a common total sulfur value, the plasma ash
number is still better than both ASTM values. The ASTM titration value
is within 50% of the plasma ash value, but it must be remembered that there
are accuracy problems associated with this method. If the pooled stan-
dard deviation for the Eschka technique is set identical to the plasma
ash inorganic value, the corresponding values are ±0.033 and ±0.081 for
the plasma ash and ASTM-AA techniques, respectively. It is felt that
this assumption is valid if extra care is taken with the Eschka technique.
Thus, it can be seen that the plasma ash method offers a substantial
increase in both the precision and accuracy of coal sulfur forms analysis.
4.1.7 Comparative Evaluation of Inorganic Speciation
Addition experimentation was performed on the 10 coals discussed in
Section 4.1.6 to test the application of the ASTM sulfur forms extraction
procedure on the plasma ash and to compare the corresponding analysis of
the whole coal. ASTM D2492 was used for both sulfate extraction and
analysis. The analysis of pyritic iron in the HNO., extract was performed
by atomic absorption in both cases. However, because of the poor pre-
cision of the proposed ASTM-AA method (Section 4.1.6), the TRW-AA method
which has given excellent precision in other work was used.2 The princi-
pal deviations in this method are the use of the less sensitive iron.line
at 244.0 nm, the use of a bracketing technique for instrument standardiza-
tion, and the omission of lanthanum as a background suppressant. These
analyses, performed in triplicate, are presented in Table 10. The whole
coal and direct total inorganic results have been taken from Table 8. The
last three columns of this table compare total inorganic sulfur calculated
using three procedures: (1) ASTM procedure, (2) plasma ash procedure with
26
-------
Table 10. Comparison of Inorganic Speciation Methods
Coal
Old Ben 21
Old Ben 26
Inland Peabody
Marti nka
E-Seam
C-Prime
Mathies
Lacas
Muskingum
Tailings
Pooled Std. Dev.
Sulfate Sulfur
ASTM
0.08 ±0,000
0,16 ±0,006
0,38 ±0.012
0.12 ±0.006
0.08 ±0.006
0.08 ±0.000
0.47 ±0.021
0.44 ±0.006
1.22 ±0.092
0.02 ±0.010
±0.030
Plasma
0.22 ±0.015
0,29 ±0,064
0,53 ±0.074
0.23 ±0.002
0.10 ±0,006
0.16 ±0.009
0.59 ±0.028
0.47 ±0.013
1.25 ±0.040
0.44 ±0.020
±0.036
Pyritic Sulfur (AA)
ASTM (AA)
0.48 ±0.066
0.77 ±0,000
1,03 ±0,006
1.33 ±0.130
0.38 ±0.006
0.90 ±0.007
0.66 ±0.006
0.82 ±0.006
2.42 ±0.040
1.92 ±0,170
±0.068
Plasma
0.41 ±0.013
0.91 ±0,010
1,17 ±0,024
1.12 ±0.002
0,27 ±0,004
0.87 ±0.017
0.57 ±0.011
0.96 ±0.040
2.20 ±0,041
2.04 ±0,000
±0.021
Total Inorganic Sulfur
AST^a)
0.56
0,93
1,41
1.45
0.42
0.98
1.13
1.26
3.64
1,94
±0.078
Plasma^
0.71 ±0.011
1.25 ±0.015
1.59 ±0.015
1.49 ±0.015
0.40 ±0.036
0.96 ±0.006
1.20 ±0.010
1.19 ±0.011
3.33 ±0.040
2.52 ±0.035
±0.023
Plasma^3)
0.63
1.20
1.70
1.35
0.37
1 .03
1.16
1.43
3.45
2.48
±0.042
(a).
Sum of sulfate and pyritic sulfur
(b)
HNOq extraction and gravimetric sulfate
-------
direct gravimetric sulfate analysis, and (3) plasma and ash procedure using
the sum of the sulfate sulfur and pyritic sulfur determined sequentially.
Referring to Table 10, the following observations can be made:
Sulfate sulfur values are comparable between methods for the
Appalachian coals. (The single exception 1s the tailings sample.)
The Eastern Interior coals all exhibit a higher sulfate content
using the plasma procedure. These results Indicate that both
pyritic and sulfate sulfur can occur In very fine occlusions that
are difficult to extract by the ASTM procedure as applied to whole
coal.
Plasma ash pyritic sulfur was determined using a TRW atomic
absorption iron procedure which is substantially different
than the ASTM procedure. The pooled standard deviation of
the TRW procedure was calculated at ±0.021 as compared to
±0.068 for the ASTM procedure. This shows a substantial
improvement in precision over the ASTM procedure.
Comparison of pyritic sulfur values is seen to follow the
postulated trend with the Eastern interior coals (with the
exception of Old Ben #21) exhibiting a higher plasma pyritic
content than the ASTM values. Four of the Appalachian coals
show excellent agreement in pyritic sulfur comprarison. The
Martinka, Lucas and Muskingum coals show some deviation
from expected values. The reason for this poor comparison
is unknown.
The last three columns of Table 10 compare the total inorganic
sulfur values from each of the three techniques used. Com-
parison of these values shows the expected trend, i.e., high
total inorganic values for Eastern Interior coals using the
plasma procedures and agreement in values for Applachian
coals (except the tailings sample). A least squares linear
regression analysis was performed to compare the total
inorganic values obtained using the two plasma procedures.
This results in a correlation coefficient of 0.992, a slope
of 1.03 and an intercept of -0.028 showing the excellent
agreement between the two methods.
t The double analysis required for inorganic speciation results
in a change in precision from ±0.023 to ±0.042. Thus, the
most precise value for organic sulfur is obtained when a
single total inorganic sulfur analysis is used.
4.1.8 Electron Probe Microanalysis of Coal Extracted by ASTM Procedure
The hypothesis that the divergence between procedures is the result
of very fine pyrite, which is unextractable by the ASTM technique, was
investigated by electron probe microanalysis. Samples of Old Ben #21,
28
-------
Old Ben #26, and the tailings sample which was extracted by the ASTM
technique were dried and prepared for analysis. These were selected
because they exhibited the largest deviation in total inorganic sulfur
values (Table 10). E-Seam coal was also analyzed as a baseline coal
because no difference in total inorganic sulfur was observed for this
coal.
Ten individual areas (approximately 30 A in diameter) across the
diameter of the specimen were analyzed simultaneously for Fe and S.
Individual values for each sample were averaged and concentrations cal-
culated using pyrite as a standard. These are summarized in Table 11,
As the electron probe analysis is an X-ray fluorescence technique these
values should be considered as semiquantitatlve only. The organic sul-
fur values were calculated by assuming all iron is present as pyrite and
subtracting an equivalent number of m moles of sulfur from the total
sulfur. These values can be compared to those in Table 9 where the
organic sulfur values obtained by the three different methods are sum-
marized. Old Ben #26 and the tailings sample, both of which exhibited
the largest difference in sulfur, do in fact contain measurable amounts
of Fe. If this iron is considered to be unextractable pyrite, then the
difference in organic sulfur values is easily explained. Neither the E-Seam
or Old Ben #21 coals showed the presence of iron. This was expected for the
E-Seam coal as the ASTM and plasma total inorganic values were in excel-
lent agreement. The Old Ben #21 should have shown the presence of
approximately 0.1% iron. This concentration level is at the approximate
detection limit and explains why none was found.
Table 11. Electron Probe Microanalysis of HC1/HN03 Extracted Coals
Coal
Old Ben #21
Old Ben #26
E-Seam
Tailings Sample
S Analysis
0.5
1.5
0.7
1.4
Fe Analysis
0.06
0.2
0.06
0.8
Calculated Organic S
0.5
1.3
0.7
0.5
29
-------
The microprobe analysis clearly Illustrates the presence of residual
iron in extracted coals, showing deviations in organic sulfur values
between methods. This is a strong indication of the inability of the
ASTM procedure to completely remove what is probably very fine occluded
pyrite. The plasma procedure liberates this pyrite from the coal matrix,
thereby facilitating Its dissolution via nitric acid attack.
4.2 SULFUR SPECIATION IN LIQUIFIED COAL SAMPLES
The present state of the art for coal analysis is seriously deficient
for the sulfur forms analysis of samples in which the coal matrix has been
altered significantly. This includes coke, gasified or partially gasified
coal, and liquified coal samples. The problem lies 1n the fact that the
ASTM procedure relies on the assumption that the Inorganic sulfur species
can be extracted from the surface or by penetration of the acid into the
coal pores. Thus, anything that changes the coal structure significantly
alters the basis for this assumption. Liquified coal 1s the worst case
because coal is changed into a liquid and as a result the inorganic species
are coated with a hydrophobic liquid. The plasma ashing procedure avoids
this problem by removing the organic matrix before any analysis is
attempted. In the experiments listed below no Inorganic speciation was
attempted because the Inorganic sulfur was expected to be very low and the
conditions under which the product was produced were expected to convert
all inorganic sulfur into iron sulfide, FeS.
Four liquified coal samples obtained from Sandla Corp, of Albuquerque,
New Mexico, were analyzed for total, inorganic, and organic (by difference)
sulfur content using the oxygen plasma analytical procedure. The coal
samples were prepared by Sandia using 100 mesh x 0 coal in a creosote oil
solvent at a solvent:coal(daf) ratio of 2.3:1. The initial (cold) hydro-
gen pressure in each case was 1000 psig, and the time at temperature was
30 minutes. Two samples were prepared from Illinois No. 6 coal (Orient No.
4 Mine) and two were from Kentucky No. 11 coal (Fies Mine). The filtered
products were from runs that were carried out at 430 C, whereas the
unfiltered ones were from 410°C runs. The samples are denoted as follows:
G98-57 FLP: Illinois No. 6, filtered product
G98-124 WLP: Illinois No. 6, unfiltered
30
-------
G98-76 FLP: Kentucky No. 11, filtered
G98-106 WLP: Kentucky No. 11, unfiltered
Total sulfur analysis was obtained using the standard ASTM D271
Parr Oxygen Bomb procedure. Total Inorganic sulfur was determined using
the analytical procedure presented in Section 5 with slight modification
to the oxygen plasma ashing step 1n order to prevent the volatile con-
stituents in the liquified coal from frothing. This was done by first
removing the volatile constituents under vacuum at 20 torr. When the
initial frothing subsided, the vacuum was lowered to 0.5 torr and held
for approximately 2 hours before ashing was initiated. The actual decom-
position was performed at 50 watts input power and an oxygen flow of
400 cc/min; these conditions were chosen to minimize possible frothing
losses through localized heating of the samples. The HNOj digestion and
BaSO. precipitation were not changed. In addition to the sulfur specia-
tion, the ash content was calculated. The results summarized 1n Table 12
show that in addition to removal of inorganic sulfur by filtration, sub-
stantial additional sulfur was removed in the 430°C filtered runs. The
Illinois No. 6 results show that this method 1s capable of a high degree
of precision, while the poor precision on the Kentucky No. 11 results
probably indicates a degree of nonhomogeneity in the sample as the result
of settling. It should be noted that this analysis indicates that both
forms of sulfur varied under the experimental conditions employed. The
form of sulfur that changes the most can be very Important from an engi-
neering point of view 1n order to avoid an expensive design modification
that may not be appropriate. This type of Information is not presently
available from ASTM or other procedures.
4.3 DIRECT ORGANIC SULFUR DETERMINATION BY PLASMA ASHING AND SO
SORPTION x
One of the objectives of this task was to Investigate the feasibility
of determining organic sulfur directly. Development of such a system
would be advantageous 1n that (1) a direct sulfur mass balance would be
possible, (2) the necessity of performing an Eschka total sulfur and
inorganic sulfur analysis would be eliminated, (3) direct determination
31
-------
Table 12. Sulfur Forms Analysis of Liquified Coal
Sample
Description
Illinois No. 6
Filtered
Unfil tered
Kentucky No. 11
Filtered
Unfil tered
Sulfur Analysis, % w/w
Total U)
0.52 ±0.010
0.75 ±0.023
0.67 ±0.112
1.11 ±0.136
Total Inorqanic
0.02 ±0.007
0.10 ±0.014
0.02 ±0.014
0.06 ±0.021
Organic (by difference)
0.50 ±0.012
0.65 ±0.027
0.65 ±0.187
1.05 ±0.138
Ash Content
2.9, 2.6
6.2, 9.0
3.6, 5.7
5.4, 7.6
CO
ro
(a)
(b)
Average triplicate determinations
Average of duplicate determinations
-------
would eliminate all the cumulative extraction and analysis errors,
increasing the precision and accuracy of the analysis, and (4) if all
four sulfur analyses (total, sulfate, pyritic, and organic) were per-
formed directly, an error analysis of the individual analysis wouTd be
possible.
4.3.1 Introduction
The experimentation in this section is based on the fact that
organo sulfur compounds contained in the coal matrix are converted into
SO during the oxygen plasma decomposition. As discussed below, the pre-
A
ferred way of measurement is the use of solid sorbent because of the time
for decomposition, vacuum conditions, and presence of oxygen and ozone in
the exit gases. Preliminary experiments indicate excellent recoveries
when molecular sieve or permanganate is used.
A literature search was undertaken to obtain information on solid
absorbents or solid absorbent systems that had been used successfully
for the sorption of SOX species. These fall Into two general categories:
physical and chemical.6 Physical sorption 1s further divided into
adsorption and absorption. The former is primarily a surface phenomenon
while the latter is characterized by diffusion into the pore structure
of the sorbent. In chemisorption, the sorbed species react with the
substrate to form actual chemical bonds and/or new and stable chemical
species. Because of the generally weak bonds that occur in physical
sorption systems, it was expected that chemisorption would be necessary
for the vacuum system required for plasma ashing. Because this type of
system generally has a low capacity compared to equivalent physical
adsorption systems, the sorbent has to be specific. This is to avoid
depletion of the sorbent by CO, C02, H.,0 and NO which are present 1n
large amounts in the exit gas stream.
The major portion of the literature deals with the sorption of S02
on substrates such as MgO, »° silica gel ,9 cobalt oxide, molecular
11 12
sieves, and Mn02 at atmospheric or elevated pressure. None dealt
specifically with S03 which is a product of the reaction of S02 with
33
-------
13 14
molecular oxygen. ' A single article discussed the determination of
coal sulfur using ascarite and sodium hydroxide as solid sorbents.
However, the major portion of the work utilized a liquid nitrogen trapping
system. Based on these data, Na2C03, NaHC03, NH4N03> KOH, CaO, and
KMnO, deposited on silica gel and molecular sieve 13X were chosen for
further evaluation.
4.3.2 SO Sorption Experimental Procedure
/\
Initial laboratory evaluations of several candidate sorbent systems
were carried out using a nitrogen stream doped with SOp, Sorbents were
prepared by evaporating water solutions of the various species onto 40/80
mesh silica gel. The prepared sorbents were dried thoroughly and about
lOg of each were packed into sorption tubes and held in place with quartz
wool plugs at each end. Molecular sieve 13X was packed directly.
A small single chamber research plasma instrument was received on
loan from International Plasma Corporation (IPC). A picture of the
instrument is presented in Figure 4. This unit was used exclusively for
testing solid sorbents under actual plasma conditions. Testing involved
the introduction of a known volume of SO^ into the oxygen plasma (instru-
ment settings were those used for coal ashing) followed by trapping of
generated species by solid sorbents contained in small canisters placed
in the vacuum line as close to the exit port of the reaction chamber
as possible. Figure 5 shows the assembled unit with the arrow indicating
placement of the solid sorbent canister. Figures 6 and 7 show the con-
figuration of the canisters in the outlet vacuum line.
The starting concentration of S02 in all experiments was obtained
by recording the pressure (torr $02) bled into a known (150 cc) volume
stainless steel pressure vessel. The concentration in mM ('x-O.B for all
runs) was calculated using the PV = nRT relationship. The SO^ 1n the
pressure vessel was then diluted to atmospheric pressure with Np and
attached to an inlet line. The diluted SO- was then bled into the sor-
bents for the initial tests and into the plasma reaction chamber for the
latter tests. The flow rate of the diluted S02 mixture was regulated to
maintain an approximate one hour testing period. At the conclusion of
34
-------
Figure 4. IPC Oxygen Research Plasma Instrument
Figure 5. Experimental Arrangement for Evaluation of Solid Sorbents
35
-------
Figure 6. Solid Sorbent Canister
Figure 7. Solid Sorbent Canister Placed in Vacuum Line
36
-------
the test, the pressure vessel was removed and excess S02 determined by
tltration with dilute base after sparging through a 3 percent H202 solu-
tion. The solid sorbent canisters were removed, extracted with an
HCl/hLOp solution and sulfur determined by precipitation as BaSO^. In
addition to the solid sorbents, the inside of the plasma asher reaction
chambers was wiped as clean as possible with a dampened filter paper and
residual sulfur determined by a BaS04 gravimetric analysis,
4.3.3 SO Sorbtion Studies - Results and Discussions
J\
The results of the initial screening studies are shown 1n Table 13.
Based on these studies, potassium permanganate and ammonium vanadate
were chosen for further study under actual plasma conditions. Calcium
oxide, sodium hydroxide and potassium hydroxide were also evaluated
7 8
based on positive reports 1n the literature. ' L1nde molecular sieve
13X was used because of Us ability to selectively absorb molecules less
than 10$ in diameter (S03^3A*). Because silica gel was used as a
substrate, it was evaluated to establish a baseline.
Table 13. Atmospheric Absorption Studies
Compound % Efficiency
Na2C03 20
NaHC03 20
KMn04 110
NH4V03 100
Silica Gel Neat 20
Besides the effects of varying sorbent concentrations, the effect of
elevated temperature was also evaluated; chemisorption 1s a chemical reac-
tion and will proceed more rapidly as the temperature is Increased.
Because of the physical limitations of the experimental apparatus, 100°C
was the maximum attainable temperature. Although this was probably not
the optimum temperature for some of these systems, it was hoped that any
trends to higher sulfur retention could be observed.
37
-------
The results of 20 experimental runs are shown in Table 14. The
percent recovery of sulfur based on SO- introduced was calculated by
summing the amounts of sulfur species found in the reactor chamber and
the sorbent. Of the sorbents tested, only KMn04 at 25°C and a 10:1
molar ratio and 30/50 mesh 13X mole sieve at 25°C yielded quantitative
recovery. Adding caustic or heat did not increase recovery for the
KMnO. system. Mole sieve exhibited increased sulfur sorption with a
decrease in mesh size which is indicative of a surface area dependency.
KOH, NaOH, CaO, and NH.V03 all showed increased sorption with
elevation of temperature. In all the experiments, 17 to 41 percent of
the added S02 was found retained within the reactor chamber of the
plasma asher. Because of the physical configuration of the instrument,
not all of the internal surfaces could be cleaned; therefore, the repor-
ted total values could be low by an unknown and possibly' varying amount.
The sorption of SO species on the chamber walls was unexpected and the
/\
mechanism is unexplained at this time. It is felt that the problem can
be circumvented. Using a different reactor design 1n smaller removable
chambers (for ease of cleaning) and placing the sorbent canisters either
within or directly behind the reactor chambers would eliminate sorption
sites which cannot be cleaned without completely dismantling the instru-
ment. As redesign and fabrication were outside the scope of the task,
further experimentation was discontinued.
38
-------
Table 14, Evaluation of Solid Sorbents for SO Sorptlon Under Plasma Conditions
Sorbent System (mM)
Test
Number
1
2
3
4
5
6
7
8
9
10
11
12
Temperature
25°C
25°C
25°C
25°C
25°C
100°C
100°C
100°C
100°C
25°C
25°c
25°C
KMn04
-
12
6
6
-
4
4
10
-
MOH
13
-
6
-
26
4
-.
-
9
13X mole sieve
16/30 mesh
13X mole sieve
30/50 mesh
13X mole sieve
30/50 mesh
rerun
% Recovery
Reactor
Chamber
30
13
41
23
23
13
24
28
17
21
17
21
Sorbent
15
32
14
67
40
21
23
20
23
42
71
87
Total
45
45
55
90
63
34
47
48
40
63
88
108
-------
Table 14. Evaluation of S&lid Sorbent for SO Sorption Under Plasma Conditions (continued)
A
Sorbent System mM
Test
Number
13
14
15
16
17
18
19
20
Temperature
25
25
25
100
25
25
100
25
KMn04
6
3
-
-
-
-
-
NaOM
10
10
-
-
10
-
-
NH VO.,
4 3
-
-
-
-
-
5
5
Activated Silica Gel Double
40/80 mesh ,
i 1
CaO
-
148
167
-
-
-
Trap
% Recovery
Reactor
Chamber
34
37
36
23
33
20
41
36*
Sorbent
10
14
18
39
15
6
18
15*
Total
44
51
54
62
48
26
59
51*
-U
o
Total of both traps
-------
5. PROCEDURES FOR COAL SULFUR FORMS ANALYSIS
VIA OXYGEN PLASMA ASHING
5.1 INTRODUCTION
The following methods present the laboratory analysis of coal for
sulfur forms using a low temperature oxygen plasma. The basis of the
analytical procedure is the selective removal, under plasma conditions,
of the organic coal matrix including the organic sulfur species. This is
accomplished with the unaltered retention of all inorganic sulfur compounds.
The unaltered ash is either sequentially extracted with HC1 and HN03 for
sulfate and pyritic sulfur speciation, or extracted with HN03 to obtain
the total inorganic sulfur. Total sulfur is determined by the highly
accurate BaSO, gravimetric analysis of the nitric acid extract. These
values are used in conjunction with an Eschka total sulfur analysis to cal-
culate the organic sulfur content of the coal.
Analytical procedures appear in the order listed below. The standard
ASTM procedures, when followed, are referenced. Reagent blanks must be
obtained in all procedures.
5.2 Plasma ashing procedure
5.3 Inorganic sulfur procedures
a) Sulfate sulfur, ASTM D2492
b) Pyritic sulfur
5.4 Total inorganic sulfur
5.5 Total sulfur - Eschka ASTM D271
5.6 Moisture, ASTM D271
5.2 PLASMA ASHING PROCEDURE
Coal ground to 60 mesh x 0 or finer is ashed at low temperature in
an oxygen plasma instrument.
41
-------
5.2.1 Equipment
Oxygen Plasma Asher - International Plasma Corporation
Model 104B or equivalent.
Balance capable of weighing to ±0.1 mg
t Petri dishes approximately 55 mm in diameter
5.2.2 Procedure
Weigh Ig of coal into a clean tared petri dish and place it in the
reactor chamber of the plasma asher. Secure the chamber doors and initiate
vacuum (Note 1). Adjust the input power to 100 watts and the oxygen flow
to 200 cc/min. Interrupt the ashing at least once per day; remove and
gently shake the sample to expose fresh surfaces (Note 1). The ashing is
complete when a visual examination of the sample reveals the absence of
black coal particles. This takes approximately 72 hours. At the termina-
tion of ashing, remove the samples and store in a clean, covered, draft-
free area prior to analysis.
5.3 INORGANIC SULFUR FORMS
The following methods cover the determination of sulfate sulfur and
pyritic sulfur in the plasma ash generated using the procedure in
Section 5.2. A sequential extraction procedure patterned after ASTM D2492
is used. Sulfate sulfur in the HC1 extract is determined as BaSO.;
pyritic sulfur extracted by the nitric acid is determined by atomic absorp-
tion analysis of pyritic iron (Note 2).
5.3.1 Procedure
Quantitatively transfer the plasma ashed sample into a 250 ml beaker
with successive rinsings of 2/3 HC1; adjust acid volume to approximately
50 ml; boil for 1/2 hour and filter through Whatman #40 filter paper.
Determine sulfate sulfur gravimetrically on the filtrate as per ASTM D2492.
The residue to used for pyritic sulfur analysis.
Note 1. Initiation of vacuum or readjustment to atmospheric
pressure should be performed slowly so as not to disturb the coal or
coal ash. Rapid removal or introduction of air can blow samples out
of the petri dishes.
42
-------
Note 2. Sulfide sulfur, if present in the coal, will not be deter-
mined when this acid sequence is followed. An approximation of the
sulfide concentration can be made by subtracting the total of the
sulfate and pyritic sulfur from the total inorganic sulfur deter-
mined in Section 5.4.
5.3.2 Pyritic Sulfur
Pyritic sulfur 1s determined by an atomic absorption iron analysis
performed on a nitric acid extract of the residue from the sulfate sulfur
determination. Analysis is performed at a less sensitive iron line at
344.0 nm, using a bracketing technique with iron standards.
5.3.2.1 Equipment
Atomic absorption spectrophotometer: Use iron lamp, set 0
analytical wavelength at 344.0 mm with a slit width of 2 A.
Determination made using air-acetylene flame.
Volumetrics: 250 ml
Beakers: 250 ml with covers
5.3.2.2 Reagents
Nitric acid (1/7): Mix 1 volume of concentrated nitric acid
with 7 volumes of water.
Atomic absorption iron standards: Prepare standards from 10
to 150 ppm in increments of 10 ppm by serial dilution of a
1000 ppm standard iron solution. Dilute standards with a 1/1
mixture of 1/7 HNOg and water.
5.3.2.3 Procedure
Transfer the filter paper and residue from the sulfate sulfur
extraction to a 250 ml beaker. Add 50 ml 1/7 HN03, cover with a watch
glass, and boil for 1/2 hour. Cool and filter through Whatman #40 filter
paper directly into a 250 ml volumetric flask. Wash thoroughly with 1/7
HN03 and dilute to volume.
Adjust the atomic absorption spectrophotometer using the manufacturers
directions, the 344.0 nm iron line, a 2 A slit width, and an air-acetylene
flame. Aspirate the sample in order to determine the approximate concen-
tration. Bracket this concentration with iron standards and rerun sample;
record five sample values taken over a 1-minute period of time.
43
-------
5.4 TOTAL INORGANIC SULFUR
The total inorganic sulfur value is obtained by a BaSO, gravimetric
analysis of a single nitric acid extraction of the plasma ashed sample.
All inorganic sulfur compounds, including iron sulfide (FeS), are converted
to sulfate and are reported as total inorganic sulfur.
5.4.1 Equipment
pH meter and electrodes
t Beakers, 250 ml with covers
Gooch crucibles: Prepare with an asbestos mat by filtering a
water suspension of medium texture asbestos fiber. Wash with
approximately 300 ml of hot water, ignite the crucible and
mat at 900°C for 2 hours, remove, cool and weigh.
5.4.2 Reagents
Nitric acid (1/7): Mix 1 volume of concentrated nitric acid
with 7 volumes water
BaCl2 solution (100g/l): Dissolve lOOg barium chloride in
water and dilute to liter
Filter aid: Water suspension of ash free paper pulp
Ammonium hydroxide solution (1/1): Dilute concentrated
ammonium hydroxide with equal volume of water.
5.4.3 Procedure
Quantitatively transfer the plasma ashed sample into a 250 ml beaker
with successive washings of 1/7 nitric acid. Adjust the acid volume to
approximately 50 ml; cover the beaker and boil for 1/2 hour. Filter the
undissolved ash through a Whatman #40 filter paper into a 400 ml beaker.
Wash the solids several times with 1/7 HNO-j. Adjust the pH of the filtrate
to 1.1 ±0.1 using a pH meter and 1/1 NH^OH. Bring the volume to approxi-
mately 300 ml with deionized water and heat to boiling. Add 10 ml of 10%
Bad2 solution, cover and boil for an additional 1/2 hour. Remove the
beakers and allow the BaSO, precipitate to sit overnight. Add a small
amount of filter aid to the prepared Gooch crucibles and filter the
precipitate. Wash it with approximately 200 ml of hot water, then ignite
at 900°C for 2 hours. Cool and weigh.
44
-------
5.5 TOTAL SULFUR
Perform Eschka analysis of separate whole coal sample using ASTM
Method D271.
5.6 MOISTURE
Determine moisture content of whole coal using Method ASTM D271.
5.7 CALCULATIONS
5.7.1 Sulfate Sulfur
Calculate the percentage of sulfate sulfur as follows:
[(A-B) x 13.74]/W = Sulfate Sulfur %
where
A = grams of BaSO. precipitated
B = grams of BaSO. in the blank
W = grams of sample used
5.7.2 Pyritic Sulfur - Atomic Absorption
Calculate the percentage of pyritic sulfur as follows:
C x D
W x
where
x 1.148 = pyritic sulfur %
C = iron concentration in ppm
D = solution volume
W = sample weight in grams
5.7.3 Organic Sulfur
Organic sulfur is calculated as the differences between as Eschka
total sulfur and the total inorganic sulfur obtained on a plasma ashed
sample. Calculate as follows:
Organic Sulfur = (A"B) * 13'74 - * 13-74
45
-------
where
A = weight of BaSCL obtained from Eschka analysis in grams
B = weight of blank from Eschka analysis
W, = weight of sample used for Eschka analysis
C = weight of BaSO. obtained from total inorganic sulfur analysis
D = weight of blank from total inorganic analysis
Wp = weight of sample taken for total inorganic sulfur analysis
5.7.4 Moisture
To calculate all results to a dry basis, multiply each result by
LQQ
(100 - % moisture in sample)
46
-------
REFERENCES
1. 1974 Book of ASTM Standards, Gaseous Fuels; Coal and Coke. Am. Soc.
for Testing and Materials, Part 26. Philadelphia, Pennsylvania, 1974.
2. Hamersma, J.W., and M.L. Kraft. Applicability of the Meyers Process
for Chemical Desulfurization of Coal: Survey of Thirty-Five Coals.
EPNA-650/2-74-025a, for Industrial Environmental Research Laboratory,
U.S. Environmental Protection Agency, Research Triangle Park, No.
Carolina, 1975.
3. (a) Green, R.T. Coal Microstructure and Pyrite Distribution; Coal De-
sulfurization, Chemical and Physical Methods. Am. Chem. Soc., Sym.
Series 64:3. 1977. (b) Kuhn, J.K. The Determination of Forms of
Sulfur in Coal and Related Materials; Coal Desulfurization, Chemical
and Physical Methods. Am. Chem. Soc., Sym. Series 64:16. 1977.
4. Kolthoff, I.M., and E.B. Sandell. Textbook of Quantitative Inorganic
Analysis. The Macmillan Company. 1959. p.322.
5. 1974 Book of ASTM Standards, Chemical Analysis of Metals and Metal
Bearing Ores. Am. Soc. for Testing and Materials, Part 12, E-350,
Philadelphia, Pennsylvania, 1974.
6. (a) Moore, W.J., ed. Physical Chemistry. Prentice Hall, Englewood
Cliffs, New Jersey, 1962. p.744. (b) Weast, R.C., ed. Handbook of
Chemistry and Physics. Chemical Rubber Co., Cleveland, Ohio, 1972.
7. Schoonhoydt, R.A., and J.H. Lunsford. Infrared Spectroscopic Inves-
tigation of Adsorption and Reaction of S09 on MgO. J. Catalysis,
26:261, 1972. *
8. Lunsford, J.H. Structure and Reactivity of Adsorbed Oxides of Sulfur.
U.S. Government Reports Announcement Index, 75(24):58, 1975.
9. Burwell, R.L., and 0. Neal. Modified Silica Gels As Selective
Absorbents for Sulfur Dioxide. Chemical Communications (London),
9:342, 1974.
10. Atiqucr Rahmam, A.K.M., and M.A. Nawab. Adsorption of Sulfur Dioxide
on Cobalt Oxide (III), A Kinetic Approach. Dacca Univ. Studies, 22(1)
79, 1974.
11. Union Carbide. Linde Molecular Sieves Product Information Bull. No.
F-1979-B.
12. Kiang, K. Kinetic Studies of Sulfur Adsorption by Manganese Dioxide.
Ph. D. Thesis, Carnegie-Mellon Univ., College of Engineering and
Science, 1972.
47
-------
13. Matteson, M.J., H.L. Stringer, W.L. Busbee, et al. Corona Discharge
Oxidation of Sulfur Oxide. Environmental Science and Technology,
6(10):895, 1972.
14. Cotton, F.A., and G. Wilkinson. Advanced Inorganic Chemistry. Inter-
science Pub., New York, 1966. p.542.
15. Paris, B. The Direct Determination of Organic Sulfur in Raw Coals;
Coal Desulfurization - Chemical and Physical Methods. Am. Chem. Soc.,
Sym. Series 64. 1977.
16. Littlewood, A.B., ed. Gas Chromotography, Principals, Techniques and
Applications. Academic Press, New York, 1963. p.215.
17. Lowry, H.H., ed. Chemistry of Coal Utilization. Supp. Vol., J. Wiley
and Sons, New York, 1963. p.215.
18. Strambi, E. Calue 16:61, 1943; 24:118, 1954.
19. Ahmed, S.M., B.J. Whalley. A Volumetric Finish for Eschka Analysis
Using Arsenazo (III); Fuel 51:334, 1972.
20. Ahmed, S.M., B.J. Whalley. Analysis of Total Sulfur in Canadian Coals
by a Modified Oxygen Flask Method Using Arsenazo (III); Fuel 51:190,
1972.
21. 1974 Book of ASTM Standards, Gaseous Fuels; Coal and Coke, Atmospheric
Analysis. Am. Soc. for Testing and Materials, Part 26.
22. Hicks, J.E., J.E. Fleenor, and H.R. Smith. Rapid Determination of
Sulfur in Coal; Anal. Chem. Acta 68(2):480, 1974.
23. Mott, R.A. The Rapid Determination of the Forms of Sulfur in Coal;
Fuel 29:53, 1950.
24. Kiss, L.T. X-Ray Fluorescence Determination of Brown Coal Inorganics;
Anal. Chem. 38:1731, 1966.
25. Gluskater, H.J., R.R. Rich, W.G. Miller, R.A. Cahill, G.B. Breker, and
J.K. Kuhn. Trace Elements in Coal Occurence and Distribution. 111.
State Geological Survey, Circular No. 499, Urbana, Illinois, 1977.
26. Kuhn, J.K. X-Ray Fluorescence Analysis of Whole Coal. Trace Elements
in Fuel Symposium; Am. Chem. Soc., Advanced Chem. Series No. 144,
Washington, 1973. p.66.
27. Kuhn, J.K., L.B. Kohlenberger, and N.F. Shimp. Comparison of Oxidation
and Reduction Methods in the Determination of the Forms of Sulfur in
Coal. Illinois Geological Survey, Environmental Geology Notes 66,
1973.
48
-------
28. Berman, M., and S. Ergun. Analysis of Sulfur in Coals by X-Ray Fluor-
escence. U.S. Bureau of Mines Reports of Investigations, No. 7124,
U.S. Department of the Interior, Pittsburgh, Pennsylvania, 1968.
29. Hurley, R.G., and E.W. White. A New Soft X-Ray Method for Determining
the Chemical Forms of Sulfur in Coal. Anal. Chem. 46:2234-2237, 1974.
30. Stewart, R.F., A.M. Hall, J.W. Martin, W.L. Farrior, and A.M. Poston.
Nuclear Meter for Monitoring the Sulfur Content of Coal Streams. U.S.
Bureau of Mines Progress Report TPR 74, U.S. Department of the Interior,
Pittsburgh, Pennsylvania, 1974.
31. Hall, A.M. Precision Tests of a Neutron Sulfur Meter in a Coal Prepara-
tion Plant. U.S. Bureau of Mines Report 8038, U.S. Department of the
Interior, Morgantown, West Virginia, 1975.
32. Cekerich, A., H. Deich, and J.H. Marshall. An Elemental Analyzer for
Coal, Oil and Similar Bulk Streams. MDH Industries, Inc., Internal
Report No. 3, Monrovia, California, 1978.
33. An Examination of Novel Methods for the Determination of Sulfur in
Solid Fuels. Cape Research Report No. 13, British Cape Research Assoc.,
1961.
34. Powell, A.R., and S.W. Parr. A Study of the Forms in Whcih Sulfur
Occurs in Coal. Bull. No. Ill, University of Illinois Engineering
Department, 1919.
35. Koutsoukos, E.P., M.L. Kraft, R.A. Orsini, M.J. Santy, and L.J. VanNice.
Meyers Process Development for Chemical Desulfurization of Coal. EPA-
600/2-76-143a, for Industrial Environmental Research Laboratory, U.S.
Environmental Protection Agency, Research Triangle Park, No. Carolina,
1976.
36. Schehl, R.R., and R.A. Friedel. Computerized Systems for Quantitative
X-Ray Diffraction Analysis of Pyrite in Coal. U.S. Bureau of Mines
Tech. Progress Report TPR 71, U.S. Department of Interior, Pittsburgh,
Pennsylvania, 1973.
37. Sutherland, O.K. Determination of Organic Sulfur in Coal by Microprobe.
Fuel 54:132, 1975.
49
-------
APPENDIX A
LITERATURE SEARCH
COAL SULFUR ANALYSIS
A.I INTRODUCTION
The following section contains a compilation of the literature concern-
inq the analysis of coal for both total sulfur and sulfur forms. Included
with the various sections are the unpublished results of several comparative
analyses performed at TRW in conjunction with EPA contracts dealing with
the Meyers Process (coal desulfurization via ferric sulfate Teaching). Most
of these experiments were designed to answer specific questions arisinn
from the laboratory and bench scale development of this process which could
not be answered using standard ASTM analytical procedures.
Of special interest is a review dealing with the evaluation of the U.S.
Bureau of Mines Nuclear Sulfur Meter. This technique is given special
emphasis as it offers promise for an on-line analytical procedure for total
sulfur, moisture, ash and Btu measurement which has potentially wide appli-
cation in coal cleaning facilities.
A.2 TOTAL SULFUR ANALYSIS
There are four ASTM procedures for the determination of total sulfur in
coal. These are Eschka, bomb combustion, high temperature combustion (Leco)
and peroxide combustion procedures. All four involve the conversion of sul-
fur to its oxides and quantification either by precipitation as BaSO, or
titration by base after conversion to acid (high temperature combustion).
Several modifications of the basic procedures have been tried. The modi-
fications involve a change in the final determination of sulfate and include
volumetric determinations in the Eschka and bomb combustion methods and an
infrared detection system for the high temperature combustion procedure.
Additional procedures used for total sulfur analysis include oxygen
flask combustion followed by either titration or a flame photometric detec-
tion of the sulfur oxides and nondestructive techniques including X-ray
fluorescence and a gamma ray emission nuclear sulfur meter. Of these
procedures only X-ray fluorescence has been used to any extent for Quanti-
tative laboratory analysis.
50
-------
A.2.1 Eschka Procedure - ASTM D271
The Eschka procedure has been used for routine accurate analysis of
coal for approximately 100 years and is accepted as a national standard in
several countries. The procedure involves the controlled combustion of
coal in a mixture of 1 part sodium carbonate and 2 parts of light calcined
magnesium carbonate. The sulfur compounds evolved during combustion react
with the sodium carbonate and under the oxidizinq conditions employed are
retained as sodium sulfate. Sulfate is quantitated by precipitation of
sulfate after work-up of the Eschka mixture. When performed by a competent
analyst, the Eschka procedure is accurate and precise. A slight negative
bias when very high sulfur coals are analyzed has been reported in the
18
literature and has been confirmed by our laboratory using Meyers
processed coal samples doped with elemental sulfur. Table A-l shows the
results of experiments which compared the three most common ASTM procedures.
Note that the bomb combustion procedure showed the highest recovery. The
slightly lower recovery of the Leco and Eschka procedures could be the
result of a smaller amount of sulfur being volatilized before conversion
to sulfur oxides.
A.2.2 Eschka Procedure - Volumetric Finish
In order to circumvent the time needed to complete an accurate gravi-
metric sulfate analysis, several volumetric procedures for the determina-
19 20
tion of sulfate have been proposed. ' These involve the direct titration
of sulfate using either barium chloride or barium perch!orate and a variety
of end point indicators. These methods generally lacked a distinct end
point; however, arsenazo III has given a much more distinct end point than
19 20
even the commonly used thorin. ' These techniques have been applied to
19
the hot water extract from the Eschka fusion product, as well as to an
20
oxygen flask procedure, with excellent results. If the conditions are
properly controlled, the accuracy and precision of these methods are close
to the gravimetric portion of the Eschka procedure. Interference from
cations is both common and serious and hinders sulfate precipitations
and/or causes precipitation of the indicator. As a result, cations should
be removed with an ion exchange column prior to analysis. This step is
time-consuming, thereby nullifying the significant labor savings inherent
in these methods.
51
-------
Table A-l. Total Sulfur Analysis of High Sulfur Coals
Sample Doped
With Elemental
Sulfur
ROM Coal
Hexane Refluxed
Heptane Refluxed
Toluene Refluxed
Bomb Wash
Calculated
Sulfur
6.96
6.97
7.20
7.16
Analyzed
Sulfur
6.83
6.82
7.20
7.11
Leco
Calculated
Sulfur
6.80
6.73
7.09
6.83
Analyzed
Sulfur
6.53 ±0.10
6.44 ±0.001
6.70 ±0.01
6.74 ±0.00
Eschka
Calculated
Sulfur
7.15
7.16
7.39
7.35
Analyzed
Sulfur
6.96 ±0.03
6.77
7.10 ±0.06
7.00
%Recovery
Bomb
Wash
98
98
100
99
Leco
96
96
94
99
Ff-hka
97
95
96
95
en
ro
-------
A. 2. 3 Bomb Combustion - Gravimetric Finish ASTM-D271 . D31771*21
In this procedure, approximately Ig of coal is combusted in a stain-
less steel Parr bomb containing 30 atm. oxygen and a small volume of water
or dilute sodium carbonate to dissolve the sulfur oxides. After the
removal of iron as the hydroxide, the pH is adjusted and sulfur is deter-
mined as
At TRW it has been found that the bomb wash procedure using the
gravimetric finish is equally or more precise than the Eschka procedure.
However, some analysts have reported retention of sulfur in the fused
ash and an additional extraction step was necessary for accurate analysis.
The gravimetric portion of this procedure is subject to interference from
iron. Unless carefully removed, the iron will yield low sulfur values
because it can be occluded as iron sulfate in the BaSO^ precipitate and
can be converted to ferric oxide during the ignition step. Additional care
must be taken to avoid occlusion of sulfate when the iron is precipitated.
If large volumes of ferric hydroxide are precipitated, it is mandatory to
redissolve and reprecipitate to obtain a quantitative recovery of sulfate.
A. 2. 4 Bomb Combustion - Volumetric Finish
Only one volumetric technique has been evaluated as a finishing
technique for the bomb c:mbustion procedure. This is a potentiometric
titration procedure using a lead perchlorate titrant and a lead ion-
2?
selective electrode. This procedure has been claimed to be more rapid
and precise than the standard gravimetric technique. However, significant
interferences are noted from copper, mercury and silver with the electrode ,
and nitrate, chloride and bicarbonate with the titrant.
Phosphate must also be absent. In view of these interferences and the fact
that nitrogen, carbonate, phosphate, and chloride can be major constituents
of coal or coal ash, it is difficult to see that this technique can be
applied to a wide variety of coals with dependable accuracy and precision.
A. 2. 5 High Temperature Combustion - Titration ASTM D3177-7522
This procedure entails the controlled combustion of the coal sample
in a specially designed furnace. Combustion products including S02 are
swept out of the reaction zone into a H202 trap where the resulting sul-
furic acid is titrated with dilute base. This technique is less precise
53
-------
but much faster than the Eschka or Parr bomb methods. Typically seven or
eight analyses can be completed per hour. Chlorine is a positive inter-
ferent but a correction can be made by back titrating the NaOH liberated
by the reaction of chloride with mercuric oxycyanide. This reaction can
also be used to determine the chlorine content of the coal.
2n
A.2.6 High Temperature Combustion - Infrared Detection of Sulfur
A new instrument is being marketed by Leco Corporation for the rapid
determination of sulfur in coal and coke. This instrument is based on the
combustion of~0.5g of coal in an induction furnace swept with oxygen. The
SO- combustion products are selectively measured by an infrared detector.
All sulfur compounds are completely decomposed during the high temperature
combustion and a heated delivery system prevents loss of SCL between the
furnace and the detector. The manufacturer claims an accuracy of ±3%;
however, communications with users indicate that ±5% is probably more
realistic.
A.2.7 Peroxide Bomb - ASTM D2711
Although precision and accuracy of this method are reportedly equiv-
alent to the Eschka procedure, this procedure is seldom used.
A.2.8 X-Ray Fluorescence24* 29
Of the non-ASTM procedures, the X-ray fluorescence technique is
probably the most widely used. This technique is based on the detection
and quantisation of sulfur Ka or KB X-ray emissions. When the proper
sample preparation steps are employed and corrections made for interferences
from Ca, Mg, Al, Si and Fe, the precision is equal to that of standard ASTM
procedures. The major difficulty in obtaining an accurate analysis is in
preparing the samples because the sulfur fluorescence yield increases
greatly as the average particle size of the sample decreases. '26 For
optimum analysis, samples are ground to pass through a 325 mesh screen
prior to pelletizing and analyzing. When performed properly, this tech-
nique reduces analysis time and can be used to quantitate other elements.
Variations in relative intensities of the S Ko- and Ke lines have
been noted by Berman^8 in calibration curves developed by comparing doped
mineral matter free coal with high pyrite content coal, The variations
54
-------
were attributed to a shift of the K$ line caused by differences in sulfur
bonding. The bonding shift is the basis of a "soft X-ray" procedure which
29
has been used to directly determine the forms of sulfur in coal. With
this procedure four X-ray intensity measurements are taken of the S, Ka and
3 lines and the sulfur forms are calculated. A comparison of this method
with the ASTM results showed good agreement.29 However, when large amounts
of several forms of sulfur are present, tests by EPA have shown that it is
very difficult to obtain enough resolution for a precise and accurate sul-
fur forms analysis.
^0 32
A.3 U.S. BUREAU OF MINES NUCLEAR SULFUR METER '
r '
This analytical system is based on prompt monitoring of gamma rays
produced by the interaction of fast neutrons with coal. In this method,
fast neutrons are emitted equally in all directions from a small capsule
of californium-252 centered in a bin of coal. The neutrons penetrate the
coal and are slowed by multiple collisions which produce a sphere of ther-
mal neutrons that are captured by the elements in coal, in. proportion to
their neutron cross sections. These neutron capture reactions produce
prompt gamma rays with energies characteristic of each element. A pamma
ray detector then records the gamma ray spectrum. Sulfur atoms in coal,
for example, produce 5.4-MeV gamma rays which appear in a gamma ray spec-
trum as a peak whose height is directly proportional to the sulfur content
of the solid fuel. The detector is a 6- by 7-inch sodium iodide crystal
that provides a reasonable compromise between detection efficiency, reso-
lution, activation, and cost.
Californium-252 was chosen for the analysis of coal because it emits
low-energy neutrons that do not cause appreciable interference from inelas-
tic neutron scattering reactions. The amount of californium-252 requir3d
for sulfur measurement in coal depends on the required measurement response
time. An 80-100 microgram source provides a 2-minute response time with-
out necessitating source replacement for several years. (The half-life of
californium-252 is 2.6 years.) The material beinq measured is not rendered
radioactive. This method is discussed in two major papers30'31 which review
several years of research and development. The first discusses a bench
scale development effort which resulted in a "hard wired" system ready for
installation and checkout at a commercial coal cleaning facility. The
55
-------
second paper discusses checkout activities and actual use of the instrument
in the field.
The main advantage of this system is that it is truly an on-line sys-
tem. The analyses are performed within minutes on a moving stream of
1/4" x 0 coal with each data point representing the analysis of 300 to
800 pounds of coal. This eliminates one of the major problems associated
with coal analysis which is the technique and effort involved in obtaining
a representative 2g sample from very large quantities of coal. With the
use of additional channels, the system could also be used to measure ash,
moisture and heat content, simultaneously, although no attempt was made to
actually perform these analyses. In a carefully calibrated system, pre-
cision was found to be ±0.02% and accuracy to be ±0.05%. These values are
well with the ASTM D271 guidelines of 0.2% reproducibility and 0.1%
repeatability.
A :najor disadvantage in using such an expensive instrument is that
there is no universal agreement that continuous monitoring will facilitate
an increase in productivity of a coal cleaning plant. Also the detector
system appears to be similar to those used for energy dispersive X-rays.
Problems associated with both systems include poor resolution of the sul-
fur escape peak from peaks associated with other elements. Poor resolution
fs compensated for by the use of additional channels in the detector elec-
tronics and a computer to subtract background interferences from the sulfur
signal. At times the background is substantial and reduces the precision
and accuracy of the measurement. Two channels must be used for sulfur
monitoring so that, when the number of pulses overloads the first channel,
the pulses can be collected in the second. This results in better reso-
lution and a correspondingly higher degree of accuracy and precision at high
levels of sulfur. However, because of the nature of counting statistics,
it is expected that the accuracy and precision of the system will deteri-
orate. Although it is not explicitly stated in these papers, it appears
that instrumental adjustments must be empirically set in the field to
eliminate interferences from moisture, iron and ash constituents.
56
-------
Another disadvantage of this system is that it must be calibrated in
the field by comparing instrument response to chemical analysis of the
corresponding samples. Removal and reinsertion of the neutron source sub-
stantially alters these calibrations. In addition, the data presented in
the papers indicate fluctuations of 0.2-0.4% which varies substantially
with the claimed precision of ±0.02%. The reason given for these fluctu-
24
ations is nonuniform coal flow which conflicts with claims that the sys-
tem is relatively unresponsive to bulk density and flow rate. The
fluctuations could become more pronounced when using larger counting
periods for coals that are low in sulfur. This is a serious disadvantaae
because accuracy and precision must be increased as the total sulfur con-
tent of the coal reaches the low levels required by coking coal or air
pollution control requirements.
Additional evaluation of this instrumental system should be performed
using lower sulfur coal (~1.0 percent) and varying quantities of iron
oxide, ash constituents and moisture. These parameters should be evaluated
at one instrument setting for varying coal flow rates and counting periods.
As the system is by design a process monitoring technique for coal cleaning
facilities, it should be proven to be unresponsive to possible variations
in coal ash, iron and water content.
A.4 SULFUR FORMS ANALYSIS
Sulfur forms analysis involves the determination of sulfate, pyritic
and organic sulfur. Although other techniques have been investigated,
ASTM Method D2492 is used almost exclusively. The ASTM method involves
the selective extraction and determination of sulfate and pyritic sulfur;
these values are subtracted from the total sulfur to determine (or estimate)
the organic sulfur. After extraction, several variations on the ASTM
procedures have been attempted with varying degrees of success. Only the
Illinois Geological Survey has developed a unique reductive technique for
the determination of sulfur forms. In addition to the chemical methods,
several X-ray methods have been developed for the determination of total
sulfur and/or sulfur forms directly on the coal. These methods, along
with TRW's oxygen plasma technique are discussed in the following sections.
57
-------
A.4.1 Sulfate Sulfur - ASTM Method1
ASTM method D2492 for sulfate sulfur involves a 30 minute hot 8N HC1
extraction of coal and gravimetric determination of sulfate in the extract.
Dilute HC1 solutions have been shown to attack pyrite upon standing at room
22
temperature. This problem is attributed to reaction with dissolved oxygen
which is more soluble in dilute than in concentrated HC1 solutions. The
ASTM procedure avoids this problem by using hot (boiling) HC1 and lim.iting
contact to 30 minutes. Variations of this basic analytical scheme sub-
33 2
stitute either titrimetric or turbidimetric procedures for the sulfate
determination. Limited success was achieved because interferences from
materials extracted from the coal ash required clean-up procedures that
increased the analysis time to nearly that of the ASTM analysis discussed
below. A direct analysis using the soft X-ray approach (cf. Section A.2.8)
has also been used with very limited success. Other modifications have
been investigated in an attempt to increase the sensitivity and reduce the
time to perform the analysis. The ASTM BaS04 gravimetric determination,
however, remains the method of choice.
33
A.4.2 Sulfate Sulfur - Titrimetric
The procedure described entails the redissolution of precipitated
BaSO. in an alkaline EDTA solution and a back titration of excess EDTA
with a standardized magnesium chloride solution using Eriochrome Black T
as an indicator. Results compared favorably with gravimetric checks.
Better sensitivity was found but no indication of precision or inter-
ferences was reported.
2
A.4.3 Sulfate Sulfur - Turbidimetric
The ASTM BaSO, Gravimetric procedure, as described above, is relatively
long and complicated. In order to avoid this, experiments were performed
at TRW to adapt standard turbidimetric procedures to this analysis. In all
cases, the results of this method were compared to ASTM results on the same
solution. It was found that this method is very sensitive to extraction
conditions and tends to give distinctly low results unless cations (pro-
bably iron) are removed before analysis. The experimental matrix, showing
the effect of digestion time, type of oxidizing agent, and the removal of
cations by ion exchange resin, is shown in Table A-2. The need to remove
58
-------
Table A-2. Sulfate Analysis Comparison
Sample
ROM (Weathered)
L.K. Coal
Meyers Processed Coal #1
Meyers Processed Coal #2
Processed Coal
Doped with Ferric
Sulfate
Digestion
(Reflux)
Time, Hours
0.5
1
2
3
0.5
1
2
3
0.5
1
2
3
1
2
3
Sulfate Sulfur, Ht. %
«w
0.61 ±0.03
0.62 ±0.01
0.64 ±0.01
0.67 ±0.01
0.29 ±0.02
0.31 ±0.01
0.35 ±0.01
0.39 ±0.02
0.60
0.63 ±0.01
0.64 ±0.00
0.65 ±0.01
(0.60)M, 0.61
(0.60JJ° , 0.62
(0.60 (b;, 0.62
Turbidimetric
Peroxide
oxidant^
0.14 ±0.03
0.18 ±0.05
0.24 ±0.08
0.16 ±0.01
<0.01
0.16 ±0.01
0.04 ±0.01
0.27 ±0.00
0.24 ±0.02
0.30 ±0.03
0.19 ±0.02
0.41, 0.61^C'
0.33, 0.67
0.37, 0.61
Bromine
oxidant^3'
0.15 ±0.02
0.14 ±0.05
0.24 ±0.01
<0.01
0.19 ±0.02
0.04
0.27 ±0.04
0.28 ±0.01
0.31 ±0.04
-
en
«o
(a) All analysis performed in triplicate
(b) Calculated sulfate value of sample (ASTM determined sulfate sulfur value and added
ferric sulfate
(c) Cation-exchanged turbidimetric analysis values
-------
cations before final analysis increases the time and labor to the point
where the ASTM procedure is still preferred.
A.4.4 Pyritic Sulfur - ASTM Method1
This procedure was first recommended by Powell in 1919 and still
24
remains the method of choice. Pyritic sulfur is extracted from coal by
the oxidative attack of boiling dilute nitric acid. Because varying
amounts of organic sulfur are also extracted, pyritic sulfur is quantified
by the determination of pyritic iron. Therefore, removal of non-pyritic
iron is necessary either directly or as a blank with HC1. This correction
or pre-extraction 1s performed 1n the sulfate extractions step. In the
ASTM procedure, the pyritic iron is titrated directly with either dichro-
mate or permanganate solution. This procedure is long and involved
because of the need to first destroy any organic compounds that have been
extracted. In spite of these precautions, organics often remain, causing
indistinct and fading end points. In addition, very fine (<5p) occluded
pyritic sulfur often is not extracted. The reproducibility between opera-
tors also is often surprisingly different for certain coals. For this rea-
son ASTM has set the reproducibility for duplicate determinations carried
out by different laboratories to be 0.3% and 0.4% for coals having less than
and greater than 2% pyritic sulfur, respectively. These limits are often
not satisfactory for modern coal cleaning engineering calculations.
A.4.5 Pyritic Sulfur - ASTM Extraction - Determination of Iron by
Atomic AbsorptioTT
In order-to overcome the problems inherent in the ASTM procedure for
work on the TRW-EPA Meyers Process, an atomic absorption procedure was
2
developed to replace the ASTM titrimetric procedure. This technique
eliminates the complex work-up and the interferences associated with the
titration procedure by using the specific Iron absorption at 344.0 nm.
Absorption readings are taken using a short curve approach which increases
the normal AA precision. Table A-3 lists a series of comparison analyses
made on 46 samples where the pyritic sulfur ranged from less than 0.1% to
greater than 5.0%.
60
-------
Table A-3. Pyritic Sulfur Analysis^'^'^ Atomic Absorption (AAS) vs. ASTM Procedures
Coal Mine
Sample
Musk ing urn
Powhattan No. 4
Isabella
Mathies
Williams
Robinson Run
Shoemaker
Delmont
Marion
Lucas
Bird No. 3
Martinka
Meigs
Dean
Kopperston No. 2
Harris No. 1 and 2
North River
Homestead
Ken
Star
Eagle No. 2
Lower Kittanning
Lucas
% w/w Pyritic Sulfur
AAS
0.22 ±0.028
0.46 ±0.064
0.06 ±0.007
0.08 ±0.000
0.28 ±0.049
0.08 ±0.014
0.44 ±0.148
0.22 ±0.078
0.04 ±0.007
0.22 ±0.049
0.11 ±0.014
0.12 ±0.007
0.18 ±0.035
0.20 ±0.007
0.02 ±0.000
0.02 ±0.000
0.17 ±0.028
0.22 ±0.028
0.24 ±0.050
0.04 ±0.021/Hi
0.25 ±0.004jy.
0.48 ±0.038
0.12 ±0.007
ASTM
0.26 ±0.007
0.43 ±0.057
0.07 ±0.007
0.02 ±0.000
0.30 ±0.035
0.08 ±0.014
0.46 ±0.120
0.20 ±0.134
0.05 ±0.014
0.20 ±0.007
0.16 ±0.035
0.12 ±0.007
0.16 ±0.035
0.16 ±0.035
0.06 ±0.035
0.07 ±0.042
0.12 ±0.021
0.22 ±0.092
0.30 ±0.050
0.08 ±0.028
0.19
0.33 ±0.035
0.21 ±0.034
Sample
Marion
Mathies
Meigs
Powhattan
Eagle No. 2
Jane
Fox . .
Meigs^6' , \
Powhattan No. 4*e'
Muskinqum^
Mathies^J
Marion
Powhattan No. 4
Robinson Run
Lucas
Williams
Isabella
Shoemaker
Meigs
Bird No. 3
Delmont
Eagle No. 2
Egypt Valley
% w/w Pyritic Sulfur
AAS
0.06 ±0.021
0.02 ±0.000
0.18 ±0.035
0.46 ±0.064
,0.18
0.62
0.50
0.43
0.64
0.60
0.98 ±0.007
0.84 ±0.007
2.53 ±0.000
2.72 ±0.014
1.24 ±0.007
1.94 ±0.000
1.05 ±0.042
2.18 ±0.007
1.88 ±0.191
2.64 ±0.021
4.27 ±0.014
2.66 ±0.03d. ..
4.70 ±0.004^°'
ASTM
0.05 ±0.022
0.08 ±0.000
0.16 ±0.035
-
0.11
0.63
0.47
0.43
0.54
0.48
1.05 ±0.065
0.90 ±0.017
2.57 ±0.060
2.89 ±0.190
1.42 ±0.082
2.23 ±0.062
1.07 ±0.070
2.19 ±0.100
2.19 ±0.030
2.87 ±0.062
4.56 +0.044
2.67 ±0.15°
5.07 +0.02d
'^Unless otherwise noted, all analysis have been performed on two samples of treated coal.
'Values without standard deviation are single determinations.
All values greater then 1% are untreated coal.
(d)
(e)
Average of 3 determinations.
Analysis from trial runs.
-------
Precision of the analysis was found to be better using the AA procedure
as the pooled standard deviation is 0.032 for the AA and 0.060 for the ASTM
procedure. Agreement between the two procedures is excellent with few out-
liers. There is currently an ASTM committee evaluating a similar AA proce-
dure for pyritic iron and it is expected that this procedure will be adopted
by the ASTM within a year.
A.4.6 Pyritic Sulfur - X-Ray Fluorescence10'27'35
In support of the Meyers Process bench scale coal desulfurization
program, initial development and evaluation of an X-ray fluorescence deter-
35
mination of pyntic iron in coal was investigated. The procedure as
applied is similar to the total iron procedure used by the Illinois State
Geological Survey. ' In this procedure, nonpyritic iron is first
extracted with dilute HC1, then the residue is dried, pelletized and ana-
lyzed by X-ray fluorescence spectroscopy using the Fe Ka line. The major
problem with the procedure is the preparation of standards with which the
analysis of coal can be made because the calibration curve must be generated
using coal with known amounts of naturally occurring pyrite.* The problem
is attributed to the presence of finely divided pyrite (1-15 microns) which
increases the fluorescence yield. Reasonably accurate (based on ASTM com-
parative analysis) analysis of several coals were obtained when the above
procedure was followed. The results are tabulated in Table A-4.
Experience at IGS has shown that the precision and accuracy of this
type of procedure is usually best in the iron range of 1-3%.* Beyond 3%,
the calibration curve has a tendency to become nonlinear and below 1%, the
precision and accuracy suffers somewhat from matrix effects. Typical pre-
cision for a total iron analysis is 0.04 RSD for the X-ray procedure and
0.05 for the ASTM procedure.
A.4.7 Pyritic Sulfur - X-Ray Diffraction
Iron pyrite is a member of the cubic system which gives rise to stronq,
well defined X-ray diffraction lines. The monitoring of the 311 reflection
and comparison to an internal standard reflection from Ni (200 line) is the
basis of a computerized X-ray diffraction procedure developed by the
U.S. Bureau of Mines. Sensitivity of X-ray diffraction is normally limited
*Persona1 communication, Dr. J. J. Kubn, Illinois State Geological Survey.
62
-------
Table A-4. Pyritic Sulfur Analysis - ASTM Vs. X-Ray
Fluorescence Procedures % w/w
Coal Mine
Harris
Isabella
Lucas
Shoemaker
Bird
Delmont
ASTM
0.66
1.42
1.43
2.28
2.47
3.94
X-Ray
0.98
1.30
1.58
2.10
2.68
3.94
to compounds present at S% or greater. The Bureau of Mines procedure
makes use of a scanning capability with instrumental output stored in a
computer. The scan may be automatically repeated for any number of times
with the results for each scan accumulated. Comparative analysis with
ASTM values shows reasonable agreement. Standard deviation is approxi-
mately equal to that attainable from the ASTM procedure although the time
of analysis is much shorter. High mineral matter coals give rise to back-
ground reflections which interfere with both the pyrite and nickel peaks.
When this condition exists, the reproducibility of the analysis deteriorates
A.4.8 Pyritic Sulfur - Reduction Technique Using Lithium
it 11 i i j * <» *
Aluminum Hydrije
The Illinois State Geological Survey (IGS) has developed a procedure
for pyritic sulfur analysis based on the reaction of pyrite with lithium
aluminum hydride in tetrahydrofuran which converts the pyrite (FeS2) to
iron sulfide (FeS). Subsequent acidification of the sulfide liberates
H2$ which is trapped in a gas scrubber containing a CdSO. solution. Reac-
tion of H2S with CdS04 liberates H2S04 which is then titrated with a stan-
dardized base. Pyritic sulfur values compare favorably with ASTM values.
Precision is ±0.05 which is the same as values generated using the ASTM
procedure. Problems encountered involving the mesh size are compensated
for by grinding samples to 400 mesh x 0. The procedure is not readily
adaptable to routine analysis as LAH .is very hazardous to handle and must
be kept away from moisture.
63
-------
A.5 DIRECT DETERMINATION OF ORGANIC SULFUR
Experimentation to directly determine organic sulfur in coal was
initiated by Powell and Parr in 1919; a series of solvents was used to
34
solubilize the organic sulfur. Although they found that phenol dissolved
the largest amount of the coal organic matter, less than one-half of the
organic sulfur was removed. In a different approach, inorganic sulfur was
first removed by nitric acid extraction, then the residue was extracted
with hot alkali. This extract contained what they termed humic acids;
"humic" sulfur contained little ash and the major portion of organic sul-
fur (total sulfur less the pyritic and sulfate sulfur). Because of the
variability and difficulty in obtaining exact balances, it was decided that
humic sulfur could not be determined directly. Since this time, organic
sulfur has been determined by subtraction of the pyritic and sulfate sulfur
from the total sulfur. The major difficulty with this approach is that
inaccuracies and lack of precision in the total, sulfate and pyritic sul-
fur determination are cumulative and result in lack of accuracy and pre-
cision in the calculated organic sulfur value. The pooled standard devi-
ation for the several hundred samples analyzed for TRW's Meyers Process
o
was ±0.1%. It was also found, with coals containing significant amounts
of finely divided pyritic sulfur, that the nitric ac;d extraction is often
var 'le and incomplete; this resulted in significant differences (0.1-0.5%)
in -, determined organic sulfur when performed at different times or in
diff- nt laboratories. Precision in a given set of analyses was usually
within the above limits. To avoid these problems, several alternative
procedures have recently been tested to determine organic sulfur directly.
These approaches include a soft X-ray method, microprobe analysis, and an
approach using oxygen plasma.
2Q
A.5.1 Organic Sulfur - Soft X-Ray Method
This method was discussed in Section A.4. Accuracy and precision are
in good agreement with ASTM values in selected coals where one form of
sulfur predominate-. However, serious resolution problems occur when
several forms of sulfur are present in large amounts. In these cases,
precision and accuracy degenerate rapidly and the results are of little
use.
64
-------
37
A.5.2 Organic Sulfur - Microprobe Analysis
In this method, coal samples are pretreated with HC1 to remove soluble
iron and sulfate sulfur and then further cleaned by float-sink procedures
at a specific gravity of 1.2. The resulting float fraction, which contains
about 80% of the total organic sulfur, is dried, pelletized and a polished
section is analyzed. This is accomplished by a scanning technique which
analyzes at least 100 points for both iron and sulfur. The organic sulfur
is calculated after a correction factor for unremoved pyrite is applied.
Only one analysis was reported and the precision of analysis was ±0.13%.
Presently, United Technologies Research Corporation is working on a tech-
nique to analyze both pyrite and organic sulfur without extensive
preparation procedures.
A.5.3 Organic Sulfur - Oxygen Plasma
An oxygen plasma technique for the direct determination of organic
sulfur is presently being investigated. This is the subject of Section 4
of this report and the reader is referred there for details.
65
-------
TECHNICAL REPORT DATA
(Please read InUrucnons on the reverse before completing'
1 REPORT NO
EPA-600/7-79-150
3. RECIPIENT'S ACCESSION NO.
4 TITLE AND SUBTITLE
Coal Sulfur Measurements
5. REPORT DATE
July 1979
6. PERFORMING ORGANIZATION CODt
7. AUTHOH(S)
J.W. Hamersma andM.L. Kraft
8. PERFORMING ORGANIZATION REPORT NC
9 PERFORMING ORGANIZATION NAME AND ADDRESS
TRW Defense and Space Systems Group
One Space Park
Redondo Beach, California 90278
10. PROGRAM ELEMENT NO.
INE624
11. CONTRACT/GRANT NO.
68-02-2165, Task 203
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD CO
Task Final; 1/76 - 12/78
COVERED
14. SPONSORING AGENCY CODE
EPA/600/13
is SUPPLEMENTARY NOTES jERj_,-RTP prolect officer is Frank E. Briden, MD-62, 919/541-
ABSTRACT The report describes s. new technique for sulfur forms analysis based on
low-temperature oxygen plasma ashing. The technique involves analyzing the low-
temperature plasma ash by modified ASTM techniques after selectively removing
the organic material. The procedure has been tested on 25 coals and compared with
ASTM analyses with excellent results. The data indicate that it is significantly more
accurate and precise than ASTM D2492. A separate set of experiments showed that
it is also feasible to determine organic sulfur directly by trapping SOx in the plasma
ash effluent. Development of the latter procedure was beyond the scope of the task.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
Pollution
Coal
Sulfur
Measurement
Organic Compounds
Oxygen
Ashes
Plasma Devices
b.IDENTIFIERS/OPEN ENDED TERMS
Pollution Control
Stationary Sources
Oxygen Plasma Ashing
Organic Sulfur
c. COSATi Field'Grour
~2TB~
201
13 B
08G
07B
14 B
07 C
s. DISTRIBUTION STATEMENT
Release to Public
19 SECURITY CLASS (ThisReport)
Unclassified
20 SECURITY CLASS (This 'page/
Unclassified
21. NO. OF PAGES
73
22. PRICE
EPA Form 2220-1 (»-73)
66
------- |