United States Industrial Environmental Research EPA-600/7-79-201
Environmental Protection Laboratory August 1979
Agency Research Triangle Park NC 27711
Pollutants from Synthetic
Fuels Production:
Sampling and Analysis
Methods for Coal
Gasification
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 in 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 wide'range 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-600/7-79-201
August 1979
Pollutants from Synthetic Fuels Production;
Sampling and Analysis Methods
for Coal Gasification
by
S. K. Gangwal, P. M. Grohse, D. E. Wagoner,
D. J. Minick, C. M. Sparacino, and R. A. Zweidinger
Research Triangle Institute
P. 0. Box 12194
Research Triangle Park, North Carolina 27709
Grant No. R804979
Program Element No. EHE623A
EPA Project Officer: N. Dean Smith
Industrial Environmental Research Laboratory
Office of Energy, Minerals, and Industry
Research TriangleL.Park, NC 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, DC 20460
-------
ABSTRACT
A laboratory-scale coal gasification facility has been developed and
is in use to study the generation, sampling, chemical analysis, process
evaluation, and environmental assessment of pollutants from coal gasifi-
cation operations. This report describes sampling and analysis methods
pertaining to this coal gasification environmental study.
Collection and sampling techniques are described for partlculates,
organic condensibles and vapors or gases in the raw product stream of the
gasifier as well as for solid residues.
Gas chromatography (GC) procedures are described for the measurement
of fixed gases, C-|-C5 hydrocarbons, sulfur gases and Cg-Cg aromatics.
Atomic absorption (AA) procedures for measurement of toxic trace elements
include those for arsenic, selenium, lead, cadmium, chromium, and mercury.
Volatile orgam'cs are collected from the gas stream via polymeric
sorbents (Tenax GC and XAD-2), and analyzed by glass capillary GC/mass
spectrometry. The major nonvolatile byproduct (tar) is prefractionated by
solvent partitioning into acid, base, and neutral fractions. Each fraction
is analyzed by capillary GC/MS or high performance liquid chromatography
(HPLC). Typical results are given to illustrate the nature of the compounds
studied, the methodologies and the sensitivities of the methods.
ii
-------
TABLE OF CONTENTS
Section
Abstract
List of Figures
List of Tables
Acknowledgements ....
1.0 INTRODUCTION
2.0 CONCLUSIONS, PROBLEM AREAS AND PLANS
3.0 SAMPLING METHODS
3.1 SOLIDS AND PARTICULATES
3.2 LIQUIDS
3.3 GASES AND VAPORS
4.0 ANALYSIS OF GASES AND VAPORS
4.1 GAS CHROMATOGRAPHY (GC)
4.2 EVALUATION OF POLYMERIC SORBENTS AND GC/MS ANALYSIS
5.0 ANALYSIS OF LIQUIDS
5.1 SEPARATION OF TAR AND CONDENSATE
5.2 TAR PARTITIONING
5.3 GC/MS ANALYSIS
6.0 ATOMIC ABSORPTION METHODS
6.1 SAMPLE PREPARATION AND ANALYSIS
6.1.1 Coals and Residues
6.1.2 Tars
6.1.3 Aqueous Condensate and Scrubber Solutions .
6.1.4 Sorbent Materials for Gaseous Effluents . .
6.2 QUALITY CONTROL PROCEDURES
6.3 TYPICAL RESULTS
7.0 FOURIER TRANSFORM INFRARED SPECTROMETRY
7.1 HIGH RESOLUTION ANALYSIS OF GAS SAMPLES
7.2 GC/FTIR ANALYSIS OF SEMIVOLATILES IN XAD-2 EXTRACT
Page
1
3
5
5
5
7
12
12
21
37
37
38
40
47
47
47
49
. 49
. 49
49
49
54
54
. 57
7.3 LOW RESOLUTION FTIR OF COAL, ASH, AND TAR 59
8.0 SPECIAL TECHNIQUES FOR ANALYSIS OF TARS AND AQUEOUS CON-
DENSATES 65
8.1 PNA ANALYSIS USING GLASS CAPILLARY GAS CHROMATO-
GRAPHY (GC)2 ' 65
8.2 SIMULATED GLC DISTILLATION OF TAR 69
8.3 CLASS CHARACTERIZATION USING ELEMENT SPECIFIC
DETECTORS 69
8.4 ION CHROMATOGRAPHIC ANALYSIS OF AQUEOUS SAMPLES ... 71
111
-------
TABLE OF CONTENTS (continued).
Section Page
References 72
Appendices
I. SELECTION AND PREPARATION OF POLYMERIC SORBENTS 1-1
II. TENAX and XAD-2 ANALYSIS WITH GC/MS COMPUTER II-l
III. QUALITATIVE GC/MS RESULTS FOR ORGANIC VAPORS AND TAR
FRACTIONS III-l
iv
-------
LIST OF FIGURES
Number Page
1 Samples collected during gasification of coal 6
2 Glass sampling system 8
3 Continuous analysis system for H2» 02> C02, CO, and ChL • . 10
4 Schematic of GC analysis system 13
5 Hydrocarbon analysis system 16
6 Comparison of dual and single flames modes for a coal
gasifier product gas 19
7 Performance of Tenax GC cartridges at high hydrocarbon
levels 24
8 Tar partitioning scheme 39
9 Range of selected trace elements in coals 51
10 Percent loss of effluents (Pittsburgh No.8 coal) 53
11 Chemigram of infrared bands (cm~ ) versus time during GC/
FTIR analysis of extract of XAD-2 resin sampled during surge
phase of run 41 58
12 Comparison of coal spectra. Top = Illinois No.6; Bottom =
Zap lignite 61
13 Comparison of tar spectra. Top = Illinois No.6; Bottom =
Zap lignite 62
14 Comparison of ash spectra. Top = Illinois No.6; Bottom =
Zap lignite 63
15 PNA fraction of coal gasification tar by capillary GC-FID:
Determination of selected PNA's 67
16 Analytical system for organic vapors in gas samples .... II-2
17 Gas liquid chromatograph-mass spectrometer computer (GC/MS/
COMP) layout II-3
-------
LIST OF TABLES
Number Page
1 Conditions for GC Analysis 14
2 Cg-Hydrocarbons, Cg-Cg Aromatics and Sulfur Species in Pro-
duct Gas Stream from Semi batch Air-Steam Gasification of
4x8 mesh Illinois No.6 at 200 psia: ppm versus onstream
time 18
3 Integrated Concentrations for Selected Compounds in the
Gasifier Product Gas 20
4 Sample Stability Check 22
5 Performance Evaluation of XAD-2 Traps 25
6 Estimated Total Hydrocarbons Collected on XAD-2 Traps ... 26
7 Breakthrough on XAD Cartridges 27
8 Components of Standard 28
9 Evaluation of Recoveries from XAD-2 Cartridges 30
10 Recovery of Compounds from XAD-2 Resin with Evaporation of
Solvent from 5 ml to 0.5 ml 31
11 Recovery of Compounds from XAD-2 Resin with Evaporation of
Solvent from 195 ml to 5 ml 32
12 Average Percent Recoveries from XAD-2 33
13 Percent Deviation of Response Ratios from Ideality as a
Function of Peak Width 33
14 Average of Deviations (%) of Response 34
15 Quantitative GC/MS Results for Selected Compounds on XAD-2
Resins 36
16 Elemental Analyses of Crude Tar 43
17 Typical Preliminary and Partition Results for Tar and
Condensate 44
18 Quantitative Results for Selected Compounds in Tar
Fractions 45
19 Quantitative Results for Phenols, Cresols, and Xylenols in
Condensate by Reverse Phase HPLC and GC/MS 46
20 Average Material Balance for Three Illinois No.6 Coal
Gasification Runs 50
21 High Resolution FTIR Analysis of Selected Gas Samples ... 56
VI
-------
LIST OF TABLES (continued).
Number Page
24 Selected Compounds in PNA Fraction as Weight % of Coal
Gasification Tar (MS vs. capillary-FID) 68
25 Conditions for Simulated GLC Distillation 70
26 Error in Boiling Points of PNA's as Predicted by the
Procedure . . . . 70
27 Corrected Boiling Point Distribution for Tar Produced from
Gasification of Illinois No.6 Coal (Run 52) 71
28 Operating Parameters for GLC/MS/Comp System 11-4
29 Operating Parameters for LKB 2091 GLC/MS/COMP System .... II-7
30 GC/MS Analysis of Polynuclear Aromatics from Run 21 III-2
31 GC/MS Analysis of Nonpolar Neutrals from Run 21 111-3
32 GC/MS Analysis of Tar Bases from Run 21 III-4
33 GC/MS Analysis of Steady-State XAD-2 Extract from Run 21 . . 111-5
34 GC/MS Analysis of Surge XAD-2 Extract from Run 21 III-6
35 GC/MS Analysis of Glass Fiber Filter Extract from Run 21 . . III-7
36 GC/MS Analysis of Tenax Sample Upstream of Steady-State
XAD-2 Trap for Run 21 II1-8
37 GC/MS Analysis of Tenax Sample Downstream of Steady-State
XAD-2 Trap for Run 21 III-9
vii
-------
ACKNOWLEDGMENTS
The sampling and analysis procedures reported herein were developed and
implemented for the project entitled "Pollutants from Synthetic Fuels Pro-
duction" under sponsorship of the Fuel Process Branch, Industrial Environ-
mental Research Laboratory, U.S. Environmental Protection Agency, Research
Triangle Park, N. C. The guidance of Dr. N. Dean Smith, Project Officer, Mr.
William J. Rhodes, Program Manager, and Mr. T. Kelly Janes, Branch Chief, are
acknowledged.
The collaboration of David Green, John Cleland, Duane Nichols, and
Forest Mixon of the Process Engineering Department at RTI have been central
to the conduct of this project. Substantial contributions to analytical
method development and execution have been provided by Jesse McDaniel, Steven
Frazier, Kenneth Tomer, and Mitchell Erickson. Finally, the authors express
thanks to Dr. Harold McNair, Professor of Chemistry at Virginia Polytechnic
Institute and State University for his valuable suggestions regarding specific
analytical techniques.
viii
-------
1.0 INTRODUCTION
The Research Triangle Institute (RTI), under sponsorship of the U.S.
Environmental Protection Agency (EPA), has undertaken a research project to
study pollution problems associated with synthetic fuels production processes.
The project is currently in the third year of a five-year program. So far,
the major thrust of the program has been the gasification of coal.
Nine U.S. coals have been gasified in a laboratory reactor and some 57
gasification runs have been completed. The reactor and the associated
sampling system were designed, fabricated and put into operation during the
first year of the program. A versatile and flexible sampling system has
resulted from a series of modifications which were made through experience
gained from the earlier runs. This report describes the sampling and
analysis methods for characterization of the gaseous, solid and liquid
effluents from the laboratory coal gasification reactor.
The coal gasification reactor has been operated in a semi batch mode. A
1.5 kg batch of coal is dropped into the hot reactor through which steam and
air are flowing at high temperature (~1000°C) and pressure (~200 psia). A
surge condition prevails for a few minutes following the coal drop until
conditions stabilize as volatiles are driven off due to pyrolysis. Char
gasification then ensues. Provision has been made in the sampling approach
to characterize the two periods as distinct from each other. Emphasis is
placed on the analysis of trace organics, sulfur species, tars and aqueous
effluents with the use of modern instrumental techniques.
Gas samples are collected in glass bulbs and analyzed by gas chromato-
graphy (GC). A variety of detectors including thermal conductivity (TC),
flame ionization (FID) and flame photometric (FPD) are used for measurement
of permanent gases, C-,-C5 hydrocarbons, Cg-CQ aromatics and sulfur gases.
Volatile and semivolatile organics are
-------
XAD-2 resin is extracted using methylene chloride and the extract is analyzed
by GC/MS. Standard atomic absorption (AA) procedures are used for measurement
of toxic trace metals including As, Sb and Hg in coals and residue and
additional AA procedures have been developed for analysis of Se, Cr, Pb, Ni,
Cd and Cu in coals, residues and tars.
The major nonvolatile product is condensed tar which is separated from
the aqueous condensate and solvent partitioned into acidic, basic and three
neutral fractions. Each fraction is then analyzed by capillary GC/MS or high
performance liquid chromatography (HPLC).
Several special techniques are under development for further characteri-
zation of the various effluents. These include Fourier transform infra-red
2
spectroscopy (FTIR), GC-FTIR, glass capillary gas chromatography (GC ),
simulated GC distillation and ion chromatography. Details of the above
procedures and typical results obtained are described in this report.
Related reports are simultaneously being prepared on other aspects of the
screening test runs of the RTI nonisothermal pollutant generation facility. A
report on the technical aspects of the runs presents operating conditions and
test results; and, a second report deals with the health and environmental
significance of the results. Separate reports are also planned on related
topics. One is intended on the transient behavior of the gasification test
runs relative to both reactor operation and the concentration of effluents.
Another is to be a comparison of the results of this project with the available
comparative data on pilot plant and commercial gasification operations.
Future reports will also present information relative to parametric
studies, which examine the generation and control of potential pollutants in
coal gasification under various selected operating conditions. The parameters
under consideration are coal type, coal particle size, reactant flow rates,
chemical additives, and other factors. Information being generated in this
project is intended to provide a basis for the assessment of the potential
health and environmental severity of the effluents from coal gasification
processes and is intended for use in the reduction of such.
-------
2.0 CONCLUSIONS, PROBLEM AREAS AND PLANS
A sampling and analysis protocol has been developed and applied for
collection and measurement of solid, liquid and gaseous effluents from a
laboratory coal gasifier. The type of samples collected include solid
residues and particulates, tars and aqueous condensates, semivolatile and
volatile organics and gaseous species. Analytical chemical measurements are
made using a variety of techniques. These include ultimate and proximate
analysis of coals, ash and tars; continuous analysis of primary gases; gas
chromatographic analysis of gaseous pollutants including C,-Cc hydrocarbons,
Cg-Cg aromatics and sulfur species; atomic absorption analysis of solid and
liquid samples for trace metals; sorption and analysis of volatile trace
organics by GC/MS/computer techniques; and the collection, partition and
analysis of aqueous condensate and semivolatile and nonvolatile tars using
GC/MS and HPLC methods. Several special techniques have been developed or are
under development. These include analysis of polynuclear aromatic species by
glass capillary gas chromatography; Fourier transform infra-red analysis of
gas, tar, coal and residue; simulated GC distillation of coal tars for measure-
ment of boiling point distributions; and ion chromatographic analysis of
cations and anions in aqueous condensates.
Certain problem areas which need further developmental work or alter-
native instrumental techniques have been identified. Of major concern is the
characterization of sulfur, nitrogen and oxygen heterocyclic compounds in the
tar. A study has been initiated using element specific detectors to char-
acterize nitrogen and sulfur compounds by boiling points. Another area of
importance is the characterization and speciation of trace metals in the gas
phase. To date, only XAD-2 resins have been analyzed for this purpose.
However, species like mercury vapor, methylmercury, arsine and hydrogen
selenide are probably not being trapped by XAD-2. An effort is presently
underway to confirm this by installing a backup activated charcoal trap
-------
downstream of the XAD-2. Charcoal is known to trap species like arsine very
2
efficiently. Polynuclear aromatics analysis is another area which requires
further work. This is especially true for five-to-seven ring compounds which
do not yield to easy measurement by gas chromatographic techniques because of
their high boiling points (>500°C). Yet they are known to be highly toxic and
mutagenic. An alternative approach involving high performance liquid chromato-
graphy coupled to a spectrofluorometer will be evaluated. This technique is
especially useful for ultra-trace level PNA's of the order of picograms.
Another area of concern is the spillover of PNA's onto the other tar fractions.
3
Tar bases have been shown to be even more mutagenic than the PNA fractions.
It is felt that PNA spillover may be partly responsible for this. This is
being further investigated. A relatively simple procedure is to inject a
dilute tar sample directly onto a capillary GC-FID and determine selected
PNA's based on retention times. Then by subtraction of the PNA's in the PNA
fraction the degree of spillover can be estimated.
Future plans include continuation of the application of the sampling and
analysis protocol to coal gasification and other synthetic fuel-related
samples. Modifications and improvements will be made as needed to improve
analytical efficiency and reduce analysis uncertainties. Quality control
procedures have been in use all through the project. During parametric tests
and modeling studies, however, there will be a need for making rigorous
quality control checks on the data. Also degrees of accuracy and precision
need to be better defined so that the confidence levels of results can be
ascertained for future test runs.
-------
3.0 SAMPLING METHODS
A versatile and flexible system is in use to sample various effluents
from the coal gasification reactor. This sampling system removes parti -
culates, tars and aqueous condensates, semivolatile organic material, volatile
organics and fixed gases. A simplified schematic of the various samples that
are collected is shown in Figure 1. All liquid and solid samples, unless
otherwise stated, are stored in a freezer in glass bottles with tight teflon
D
caps and Parafilm seals for further analysis.
3.1 SOLIDS AND PARTICIPATES
As shown in Figure 1, these samples are of three types: feed coal,
reactor residue, and particulates. Approximately 50-100 g of a representative
sample of the feed coal and a like amount of the residue following the run are
collected. Portions of these samples are sent to commercial laboratories for
ultimate and proximate analyses by standard ASTM procedures. The product
gases that emerge from the reactor immediately enter in a particulate trap.
This trap contains a stainless steel braided cup which functions as a flow
impinger. The trap is further packed with glass wool as a medium to facilitate
the removal of solid particles from the hot gas stream and is insulated and
heated to prevent tar condensation. Any finer particles which are not captured
in this trap are further removed in the condenser and through the use of glass
fiber filters that will be described in Sections 3.2 and 3.3. Following the
run, the glass wool is carefully removed and refrigerated, pending solvent
extraction and analysis.
3.2 LIQUIDS
1
The gases leaving the particulate trap enter a condenser with an outer
jacket through which liquid freon is circulated for removal of tar and
aqueous condensate. This trap may be tapped periodically during a run for
-------
9
Gasifier
1.
2.
3.
4.
5.
Particulate
Trap
Condenser
Gas Sampling
System
Coal
Residue
Participates
Tar, Water and Insolubles
Entrained Tar and Water
6.
7.
8.
9.
Gas Bulbs
Polymeric Sorbents
Glass Fiber Filter
Scrubber Solutions
Figure 1. Samples collected during gasification of coal
-------
removal of the accumulated material or at the end of the run for total
collected material. The tar and condensate are separated and the tar is
subdivided into several fractions. An elaborate sample preparation technique,
described in Section 5.1, is used to facilitate partitioning into different
sample classes including aqueous condensate, tar acids, tar bases, nonpolar
neutrals, polar neutrals, and polynuclear aromatics. Solids and particulates
which are insoluble in the neutrals partitioning solvent (cyclohexane) are
also collected. Required quantities of each sample are stored in the freezer
and the rest are disposed of using adequate waste disposal methods.
3.3 GASES AND VAPORS
The gases leave the condenser and at this point the reactor pressure is
reduced through a backpressure regulator. The gases then enter a glass
sampling system shown in Figure 2. Entrained moisture in the gases is further
removed at this point so that its interference with the subsequent sampling of
gases and vapors is reduced. Special emphasis is placed on the sampling
strategy to account for the surge and steady-state conditions characteristic
of the semi batch operation. When a batch of coal is fed to the hot reactor, a
surge condition prevails during which most of the pyrolytic products are
released. Typically the surge period lasts for about 15-30 minutes, during
which a dramatic change in concentration of the pollutant occurs, which could
be as much as three orders of magnitude. Gaseous grab samples are taken at a
high frequency (every 1 to 2 minutes) during the surge condition to efficiently
characterize the concentration changes. The steady-state condition is marked
by a drop in the methane concentration to a relatively stable value of about
2 percent. At this point, gas samples are collected with a time interval of
about 10 minutes or greater. The glass containers used to collect these gas
samples are 500 ml in volume and are equipped with two high vacuum teflon
stopcocks. Approximately 3-5 liters of gas is passed through them prior to
closing the stopcocks. Samples are kept stored in a heated box at about 50°C.
Stability of gas samples and the GC analysis of these samples are described in
Section 4.1.
In addition to direct collection of gases in glass containers, organic
vapors are adsorbed on polymeric sorbents. Tenax-GC and Amberlite XAD-2 are
used for the gas analyses. The sorbents are packed into two types of containers:
-------
All Class
From
Gasifier
00
CT: Corrugated Teflon
T1-T2: Toggle Valves
S1~S6:
Ports with
teflon stopcocks for
sample collection in
small Tenax cartridges
and glass bulbs.
y
Continuous
Gas
Monitors
Meter
Figure 2. Glass sampling system.
-------
large in-line traps, and small cartridges. The in-line traps are cylindrical
tubes with an internal volume of approximately 1 liter. One end contains a
fritted glass disc to retain the sorbent. Approximately 150-200 g of cleaned
XAD-2 (see Appendix I) is packed in each trap. This is followed by a layer of
glass wool, a stainless steel screen and a glass fiber filter (Gelman type A).
End adapters are clamped to each end of each cylinder and sealed with Viton o-
rings. In order to distinguish between the surge and the steady-state con-
ditions the two traps are used in parallel, connected by a three-way teflon
stopcock. This affects the switchover from one condition to another.
Approximately 1000 and 3000 standard liters of gas are produced during the
surge and steady-state conditions respectively. In addition to the in-line
traps, samples are collected in small cartridges (1.5 cm I.D. x 10 cm long).
These are filled with sorbent (either XAD-2 or Tenax-GC) to a length of 6 cm
with glass wool at both ends. They are interfaced to the sampling system both
upstream and downstream of the large XAD-2 traps by means of teflon stopcocks
3
and fittings. When Tenax is used, only 100-150 cm of sample is necessary
2 2
because of its lower surface area (30 m /g). When XAD-2 (300 m /g) is used in
these cartridges, 5-15 liters of sample are required. Efficiency of XAD-2 has
been evaluated using Tenax cartridges upstream and downstream of the large in-
line traps. In addition for selected runs, two XAD-2 or Tenax cartridges have
been used in tandem both during the surge and steady-state periods to determine
breakthrough and displacement.
In addition to the collection of gases in bulbs and sorbents, a IN HgSO^
impinger is used in series with the glass sampling system for measurement of
ammonia. Evaluation of the accuracy of the GC/MS analysis of samples col-
lected on the sorbents are described in Section 4.2.
To ensure successful operation of the gasifier and to detect gross
system malfunctions, it is necessary to monitor the major product gases (H2>
COp, CO, and ChL) continuously. In addition, it is necessary to monitor
oxygen in the product gas which should be absent under normal operation.
However as the coal batch is gasified, 'a stage is reached when oxygen break-
through occurs indicating the completion of the run. An elaborate system has
been connected in parallel to the glass sampling system for continuous analysis
of l^j C02> CO, CH,, and 02 (Figure 3). A sample conditioner removes traces
-------
BPR
From
Gasifier
^Glass
Sampling
System
SV
BP?.: Back Pressure Regulator
TV: Toggle Valve
SV: Solenoid Valve
NO: Normally Open
P: Pump
REF: Refrigeration System
NV, R: Needle Valve
A : NDIR Analyzers
HU: Hi oh Level Alarm
FI: Flow Meter
I
To Dry Test Meter
iout ^ 02
rr
Drain
0-5 V DC
Figure 3. Continuous analysis system for H2, 02, CO, C02 and
-------
of condensibles from the gases via refrigeration (~1°C) prior to their
entering the continuous analyzers. ChL, CO and COp are measured using non-
dispersive infrared analyzers, H2 using a thermal conductivity analyzer and
02 using a paramagnetic analyzer. Pressure and flow in the system are con-
trolled using a backpressure regulator and several rotameters. The analyzers
are zeroed and spanned prior to the run with known concentrations using a
series of solenoid switch valves. Following analysis, the gases are routed
back to the glass sampling system and mixed with the bulk gases for total
volumetric flow measurement using a dry test meter.
11
-------
4.0 ANALYSIS OF GASES AND VAPORS
Collection of gaseous and vapor effluents from the coal gasifier was
described in Section 3.3. Further preparation of the samples and analysis of
selected species is carried out using gas chromatography (GC) and gas chromato-
graphy coupled to mass spectrometry (GC/MS). GC is mainly used for measure-
ment of major gases and volatile pollutants which are directly analyzable from
the glass bulbs. On the other hand, GC/MS is used for the analysis of trace
organics collected on polymeric sorbents and glass fiber filters.
4.1 GAS CHROMATOGRAPHY (GC)
A simple schematic of the GC analysis system is shown in Figure 4. As
stated earlier, the glass bulbs are stored in a heated box at 50°C. The
analysis is carried out without removal of the bulbs from the box by means of
a vacuum sample removal system. The system hardware is essentially similar to
A
the one described by Gangwal and Wagoner for measurement of sulfur compounds
using a dual flame photometric detector (FPD) coupled to a GC. It consists of
a Sargent-Welch vacuum pump for evacuating the sample loops (1 ml) to less
than 1 mm of Hg absolute pressure, a Heise vacuum-pressure gauge (0-3450 mm)
with 2 mm graduations for measurement of sample loop pressure prior to in-
jection and three heated zero volume valves for injection of the sample onto
the GC columns. Table 1 shows the various compounds that are analyzed and the
type of detector, column and conditions that are used.
The sample injection and analysis procedure for each detector consists of
the following. The sample loop (1 ml) and associated valving and tubings are
first evacuated using the vacuum pump (Figure 4). A small amount of sample is
then used to flush the loop and the system is reevaluated. The sample is then
introduced into the loop and the absolute pressure is carefully measured using
the vacuum-gauge. Approximately 300-400 mmHg of sample pressure is used.
Using the zero volume valve, the sample is then injected onto the GC.
12
-------
to
To GC-TC
-.
{
t
(
_y
t
^
\
\
r^
\
S
\
/
I
I
J
)
^
JU
c
r
5
^
>
r^i
•
To GC-FID
^
/•
1 HoatoH R
1 . nca LcO D
Bulbs
2. Sample I
(Zero Vo
3. Carrier
4. Heise Va
_, 5. Vacuum P
_j _£.. Swagel
Valves
^ • To GC-FPD
>S)2
. / /• — -
tr f 1
^L
with Glass
Injection Valve
lume)
Gas Cylinder
Vacuum-Pressure Gauge
Pump
Figure 4. Schematic of GC analysis system.
-------
TABLE 1. CONDITIONS FOR GC ANALYSIS
Detector
Compounds Analyzed
Column and Oven
Temperature
Carrier Gas Flow
(ml/min)
Detector
Temperature (°C)
Sample Loop
Material and Size
Injection Valve
Material
TC
H2, C02, CO, N2, 02
CH^, C2Hg, CgH^
H2S, COS (>100 ppm)
6' x 1/8" SS Porapak N
7' x 1/8" SS Molecular
Sieve 13X
70°C
30
200
SS 1 ml
SS
FID
(a) C-,-Cc hydrocarbons
(b) C6H6, C7Hg, C8H10
(a) 6' x 1/8" Durapak
phenylisocyanate
(18°C)
(b) 8' x 1/8" TRIS-1,2,3
cyano ethoxypropane
(90°C)
20
150
SS 1 ml
SS
FPD
COS, CH3SH, C2H5SH
(CH3)2S, CS2, thiophene
6' x 1/8" Teflon Carbopack
B/XE60/H3P04
50°C for 2 min to 130°C at
32°C/min
30
150
Teflon 1 ml
Hastalloy-C
-------
Standards of known concentration are also analyzed in a similar fashion. Peak
areas of the various components in the sample and standard are measured using
a A/D converter coupled to a computer with a GC-software system. Concentration
of the components of interest are determined in the sampling using equation
(1).
u As Pu '
where: subscript u refers to the sample and S to the standard and C, A,
and P are concentrations, peak areas and loop pressures, respectively.
Equation (1) is only applicable to the TC and FID. The FPD is nonlinear and
a log-log calibration curve of area versus concentration for each sulfur
4
species is prepared by varying the sample loop pressure. These curves are
then used for quantisation of the sulfur compounds. Periodically, calibration
curves are also prepared for the TC and the FID using varying sample loop
pressures in order to confirm their linear range.
The analysis of fixed gases, C-j-C^ hydrocarbons and H2S using the thermal
conductivity detector has been described in detail by Gangwal , et al .
Recently, a further improvement in the system allows analysis of COS in
addition to the above gases. This is especially helpful when H^S and COS are
present in relatively large concentrations (>200 ppm). At these concentrations,
the sulfur specific FPD is saturated because the amount of sulfur injected
falls outside the dynamic range for 1 ml samples.
The flame ionization detector can be used either for the analysis of Cg-
CQ aromatics or C-,-Cg hydrocarbons using two columns in parallel. Backflush
valves are incorporated into the system to reverse the carrier gas flow
through the column. The schematic is shown in Figure 5. This is necessary to
prevent the interference of the heavier materials with analysis of subsequent
samples.
The flame photometric detector (FPD) is used for the analysis of COS
(<200 ppm), CH3SH, C2H5SH, (CH3)2S, CS2 and thiophene. The FPD uses two
flames instead of one used in the conventional Brodey and Chaney design.
15
-------
Sample to^^
Heise Gauge
FID
Zero Volume Gas Sampling
Valve
Backflush Valve
Switch Valves
8' x 1/8" S.S. TRIS (1.2,3)
Cyano Ethoxy Propane Column
6' x 1/8" S.S. Durapak-
Phenyl Isocyanate Column
Carrier
Gas
Figure 5. Hydrocarbon analysis system.
-------
" o
Details of the design have been described by Patterson, et al. The quenching
of the response to sulfur compounds in the presence of a hydrocarbon matrix
eluting simultaneously is a problem common to all FPD designs. This results
in inaccurate measurements. The dual flame design minimizes this problem by
breaking down the complex matrix to simpler compounds in the first flame and
measuring the optical emissions in the second flame. This characteristic of
the dual flame FPD is demonstrated in Figure 6. In the single flame mode, no
C$2 response is seen even though a larger thiophene response is obtained for
the same amount of sample. The FID under the same conditions indicates a
larger background of hydrocarbons at the point of C$2 elution and a smaller
background at the point of thiophene elution. In addition, the dual flame
FPD has been found to be much more reproducible than the single flame FPD.
As much as a 75 percent different response has been observed for the same
sample analyzed consecutively on the single flame FPD. This, in our opinion,
is due to hydrocarbons, still eluting from the previous sample and interfering
with the response of the next sample.
Typical data obtained for various species during gasification run 52 are
shown in Table 2. The pyrolytic products, e.g., hydrocarbons, mercaptans and
thiophene are reduced to very low levels after only 15-30 minutes into the
run. In contrast, the gasification products, H2$, COS, prevail until the
very end of the run (i.e., until oxygen breakthrough). The results in Table
2 become more meaningful when the data are integrated over the duration of
3
the run. Pollutants produced can then be expressed in pg/m of gas. For
three runs carried out with Wyoming, North Dakota and Illinois No.6 coals,
integrated values are presented in Table 3.
A few words are needed here as to the quality control procedures used to
ensure the integrity of the data generated. Since various types of analyses
are carried out on the same bulb, a partial vacuum is created in the bulb
which increases following each analysis. For this reason, the bulbs are
equipped with special high-vacuum glass teflon stopcocks capable of holding
the vacuum down to 1 mm Hg for several days. In addition, since it is not
possible to complete the analysis of all bulbs immediately following
17
-------
00
TABLE 2. C2-HYDROCARBONS, Cg-Cg AROMATICS AND SULFUR SPECIES IN PRODUCT GAS FROM SEMIBATCH
AIR-STEAM GASIFICATION OF 4 x 8 MESH ILLINOIS No.6 AT 200 psia: ppm versus onstream time
Tine
(minutes)
C2H6
C2H4
C2H2
C6H6
C7H8
Ethyl benzene
p + m-Xylene
o-Xylene
H2S
COS
CH3SH
C2H5SH
cs2
Thlophene
1
13300
17000
110
2160
360
9
40
5
11700
280
70
20
90
130
3
25500
20400
130
2230
480
20
70
10
23000
470
140
40
150
150
4
21800
19000
120
2750
750
30
130
30
20300
440
100
30
210
170
6
18300
20100
180
2590
610
20
90
20
19600
430
90
40
240
140
7
16600
18000
160
3060
770
30
140
30
19500
420
60
30
190
160
9
14900
16900
170
3560
820
30
150
30
19300
370
50
20
210
160
11
13800
14100
130
3300
620
20
110
20
19000
380
90
20
180
160
13
12200
12200
150
3570
580
20
110
20
18100
360
80
10
180
160
18 23
9970 4130
8670 2350
100 <50
3520 1770
480 320
20 10
90 70
20 10
15900 9960
330 230
50 20
10 <1
150 70
140 60
33
230
<50
500
160
6
30
6
3400
130
2
2
5
43 54 63 73 107 137 177 226 287 322
<50
260 160 90 50 10 20 69 45
100 70 50 40 10 10 64 34
4 4 3 3 2 1 1 1 0.3 0.2
20 20 20 20 20 8 64 33
44443221 11
2830 2550 2690 3840 2500 4320 1680 270 2900 210
160 150 140 180 160 160 110 170 600 29
0.2
0.7
<5
-------
Column: 6'X 1/8" Teflon
Carbopack B/H3PO4/XE 60
Temperature: 50°C for 2 minutes
programmed to 130°C at
30°C/min
28 ml/min helium carrier gas
1. COS
2. Methyl Mercaptan
3. Thiophene
IN
IN
CS2
IN
-4
LAI
Single Flame Response
No CS2 Response
8 Minutes
3
Dual Flame Response
FID Response
Fiqure 6. Comparison of dual and single flames modes for a coal gasifier product gas.
19
-------
TABLE 3. INTEGRATED CONCENTRATIONS* FOR SELECTED
COMPOUNDS IN THE GASIFIER PRODUCT GAS
(ug/m3)
C2H6
C2H4
C6H6
C7H8
Ethyl benzene
p,m-Xylene
o-Xylene
H2S
COS
CH3SH
^*O ^* C
cs2
Thiophene
Run 51
N. Dakota
NA
NA
NA
9.6E5
6.7E5
9.3E4
2.0E5
8.5E4
1.0E6
1.0E5
3.5E4
5.5E3
1.5E3
1.7E4
Run 47
Wyoming
2.4E6
1.6E6
1.6E4
2.3E6
1.3E6
5.8E4
3.9E5
7.6E4
1.6E6
1.2E5
2.3E4
1.1E4
3.9E3
3.7E3
Run 52
Illinois #6
1.8E6
2.7E6
3.4E4
1.4E6
3.6E5
1.8E4
9.8E4
2.0E4
6.1E6
6.1E5
1.4E4
1.7E4
5.7E4
6.0E4
*Analyzed by gas chromatography.
NA = Not analyzed.
20
-------
sample collection, it is essential that the samples be stable, at least for
a few days in the bulbs. A stability study was made on three bulbs over a
period of seven days. This is shown in Table 4. Only sulfur compounds were
used in this study. As seen from the table, only a five percent (or lower)
change was seen in the concentration of the various sulfur species. This
was considered to be acceptable. Each compound is quantitated by means of
calibration with a standard obtained commercially and certified within ± 2
percent accuracy or better. Periodically, the standards are verified against
a permeation tube calibration system or against NBS standards. Accuracy of
peak integration using the computer system is periodically checked by comparing
the results against alternative methods i.e., peak height, triangulation,
cut-and-weigh or other independent integration systems such as a small
electronic integrator. Finally, since the major gases (Hg. C02» CO and CH^)
are analyzed continuously, only a few samples are periodically verified b5f
the GC and compared with the data from the continuous analyzers. A difference
of less than 10 percent between the two procedures is considered acceptable.
4.2 EVALUATION OF POLYMERIC SORBENTS AND GC/MS ANALYSIS
As described in Section 3.3, organic vapors are collected on two polymeric
sorbents, Tenax-GC and XAD-2. Criteria for selection of these materials are
described in Appendix I. Therein further details are given regarding the
nature of these materials and how they are prepared for vapor collection.
Conditions and procedures used for GC/MS analysis using high resolution
glass capillary columns with a thermal desorption interface for Tenax and
methylene chloride extraction for XAD-2 are described in Appendix II. In
this section, the polymeric sorbents are evaluated for their adsorption
efficiency and typical quantitative GC/MS results for selected organics are
presented.
Tenax-GC was evaluated using a sampling manifold which permitted the
flow through two Tenax cartridges (see Section 3.3). An injection port was
installed upstream of the cartridge and* benzene-doped nitrogen flowed through
both cartridges at 2 1/min. Radiolabelled toluene was injected into the
port and the second cartridge was replaced periodically with a fresh cartridge.
Each cartridge was desorbed thermally onto a scintillation counting fluid and
21
-------
TABLE 4. SAMPLE STABILITY CHECK
ro
ro
Sample No.*
SI
(7 minutes
from coal
drop)
S2
(16 minutes)
S3
(86 minutes)
Time
(hours)
3
28
49
74
98
123
147
3
28
49
74
98
123
147
3
28
49
74
98
123
147
COS
(ppm vol)
550
540
530
530
540
530
510
440
420
420
420
430
410
410
150
140
140
130
140
130
130
CH,SH
(ppm vol)
39
37
37
36
38
36
35
17
16
15
16
16
15
16
<1
C2HgS
(ppm vol)
20
19
19
19
20
19
18
9
9
8
9
7
7
6
<1
CS2
(ppm vol)
230
220
220
220
220
220
210
140
140
130
140
140
130
130
<1
Thiophene
(ppm vol)
250
250
250
240
250
240
230
150
150
150
150
150
140
140
<1
Samples obtained from low-Btu semibatch gasification of W. Kentucky #9 coal.
-------
radioactivity was determined. Three different conditions were examined: (1)
low toluene (0.8 mg), (2) high toluene (22 nig), and (3) low toluene with 3000
ug/1 benzene in the purge gas. These experiments graphically illustrate
(Figure 7), the problem of extrapolating from low to high concentrations.
The amounts used are realistic in terms of actual gasifier process streams.
The change in breakthrough volume for toluene with toluene is large. The
displacement of toluene by benzene is also of importance, since a large
background of hydrocarbons exists in the real sample and only a limited
number of them are sought at low levels. This behavior is not unique to
toluene but the threshold at which the above effects become significant
changes with each sorbate. This experiment indicated that the amount of the
real sample through the Tenax cartridge should be low and 100-150 ml was
found to be appropriate for the gasifier sampling program (see also Section
3.3).
XAD-2 unlike Tenax-GC has a higher sample capacity. This material was
evaluated under gasification conditions by taking small discrete Tenax
samples upstream and downstream of the XAD-2 traps. Further performance
checks were made on XAD-2 by collecting discrete (5-15 liters) samples on two
XAD-2 cartridges in tandem upstream of the large XAD-2 traps. Results for
three gasification runs during which the above approaches were adopted are
shown in Table 5.
It is apparent that breakthrough is occurring at the high loading
present in runs 26 and 32. To compare this with the results from the Tenax
study, a summation of all the compounds quantitated on the XAD traps on a per
gram basis was made. This is shown in Table 6. In the Tenax experiments the
benzene mass was 8 mg/g at 20 percent toluene breakthrough and 16 mg/g at 65
percent toluene breakthrough.
This would indicate a greater capacity for collection of hydrocarbons by
XAD-2 since only 10 percent of such compounds as toluene, C2-benzene, methyl
thiophene, C^-thiophenes and naphthalenes were found downstream relative to
upstream on runs 26 and 33. This is predictable on the basis of greater
surface area of XAD-2 relative to Tenax. Capacity problems are apparent,
however, in the analysis of phenol and cresols. The performance is good in
23
-------
2000 \ig/i benzene in purge
0.08 mg toluene
22 mg toluene
.87 mg toluene
100 150
Purge Volume (J.)
Figure 7. Performance of Tenax GC cartridges at high hydrocarbon levels.
-------
TABLE 5. PERFORMANCE EVALUATION OF XAD-2 TRAPS
Compound
Benzene
Thiophene
Toluene
Methyl thiophene
C2-benzene
C~-thiophene
Benzaldehyde
Phenol
Benzofuran
Cresols
Naphthalene
Biphenyl
Dibenzofuran
Acetic acid
Run 26 Surge Phase -
733 1 of gas in 34 min
Total mass on
XAD trap (ng)
—
—
—
Ill
285
37
Not found
650
225
290
32
11
12
Not found
Concentration found
on Tenax (pg/1)
Upstream Downstream
645 1088
101 116
324 37
67 3.5
243 1.1
16 2.3
12 9
1500 907
57 ND
265 143
510 42
25 4
7.6 ND
ND 39
Run 32 Surge Phase -
1500 1 of qas in 10 min
Total mass on Concentration found
XAD trao fmnl on Tenax l^/U
XAU trap (mgj Upstrean Downstream
124 405
12 133
72 44
6.4 3
39 1.5
2.2 ND
NOT 1.4 2.3
DETERMINED 360 250
18 1.8
22 21
160 45
4.1 4.8
1.6 2.4
3.4 6.2
Run 33 -
Mass on XAD
trap (mg)*
—
--
--
1.9
22
0.8
Not found
83
ND
40
175
0.8
2.8
Not found
Steady State Phase
Concentration found
on Tenax ipg/1)
Upstream Downstream
22 0.2
1.0 ND
15 0.4
0.6 ND
11.5 0.08
0.6 ND
ND 0.45
148 5.0
10 0.05
8.8 0.4
76 2.2
0.8 0.07
0.19 ND
ND 0.85
ro
en
35% of the values found after 100 min on line.
-------
TABLE 6. ESTIMATED TOTAL HYDROCARBONS COLLECTED ON XAD-2 TRAPS
Run No. Hydrocarbons/g XAD (mg)
26 18
33 6
run 33 at 6 mg/g hydrocarbons but in run 26 at 18 mg/g hydrocarbons severe
breakthrough is observed. Run 32 showed the exaggeration of the effects
seen in run 26 due to twice the gas volume and presumably even higher hydro-
carbon accumulations.
The tandem XAD-2 cartridges were analyzed for runs 32 and 33 with the
results given in Table 7. The higher breakthrough of naphthalene on the
cartridge was somewhat surprising until the residence of the sample in the
cartridge (0.75 sec) was compared with that of the trap (1.8 sec). This
residence time is especially important for the less volatile compounds which
may form aerosols.
Recovery studies were performed with 11 to 13 commonly observed organics
ranging from the most volatile, methyl thiophene, to the least volatile,
anthracene. In these experiments a mixture of these compounds in methanol
was applied by direct injection into the bed of three prepared cartridges
(1.5 cm x 6 cm) and permitted to equilibrate. The mixture contained d1Q-
anthracene as an internal standard. After equilibration each cartridge was
extracted separately with methylene chloride for 22 hours in a Soxhlet
extractor. The extracts were evaporated in K-D evaporators to 1 ml and
analyzed by GC/MS/COMPUTER (see Appendix II).
The recovery experiments were repeated using a mixture of compounds
shown in Table 8 with a 20 g portion of XAD-2 to simulate the sample from
the XAD-2 traps. In these recovery experiments three internal standards
were used, including d^-anthracene at the start of the evaporation and
p-chlorotoluene at the final step before analysis. Recoveries were deter-
mined based on each internal standard. In this evaluation only 4 ml of
~150 ml was evaporated to 1 ml.
26
-------
TABLE 7. BREAKTHROUGH ON XAD CARTRIDGES'
ro
Compound
Methyl thiophenes
C2-thiophenes
Cp-benzenes
Benzofuran
Phenol
Cresol
Naphthalene
Biphenyl
Dibenzofuran
Cartridge
1 (yg/D 2
0.8
<0.1
10.7
0.9
25
17
105
1.6
0.5
Run 32
No.
(ng/i)
ND
ND
0.2
ND
ND
ND
24
0.9
ND
Breakthrough
0
—
1.9
0
0
0
19
36
0
Cartridge
1 (yg/1) 2
31
5.6
<236
ND
239
121
263
2.2
0.8
Run 33
No.
(pg/i)
ND
ND
0.4
ND
ND
ND
32
1.2
ND
Breakthrough
0
0
<0.2
0
0
11
35
0
Volume of process stream sampled was 5 liters.
-------
TABLE 8. COMPONENTS OF STANDARD
ng/1
Anthracene-d-jQ
2-Chl oronaphthalene
p-Chlorotoluene
Indole
Phenol
3, 4-Dimethyl phenol
3,5-Dimethylphenol
2,3-Benzofuran
Toluene
1 , 2, 4-Trimethyl benzene
o-Cresol
p-Cresol
Indan
Indene
no
107
36
124
220
114
132
174.6
138.7
140.1
164.4
162.8
154.2
159.4
ng/i
Naphthalene
Anthracene
Phenanthrene
Benzofuran
Biphenyl
Methyl indole
Pyridine
Methyl quinoline
Ethyl benzene
Methyl thiophene
Ethyl thiophene
Quinoline
Methyl pyri dine
126
96
118
170
94
192.6
157.1
173.8
138.7
163.1
158.9
174.9
28
-------
The average and standard deviation of the response from each compound
ratioed against each internal standard response for the same injection were
computed. From these the percent deviation relative to the average was used
as a measure of precision for each compound relative to each internal standard.
The recovery of the compounds selected for quantitation under the para-
metric variation phase of the study were evaluated. The results shown in
Table 9 are for the discrete XAD-2 cartridge sampling procedure.
One set of triplicates had two internal standards including p-chloro-
toluene, added just before the GC/MS/COMPUTER analysis. Overall average
recoveries were 83 to 87 percent. Anomalous results were obtained for indan
on one set of cartridges and for naphthalene on the other.
Similar experiments were conducted for the XAD-2 trap samples. Two sets
of conditions were examined; the usual 10-fold evaporation from 5 ml to 0.5 ml
and a larger 40-fold evaporation from 195 ml to 5 ml. The latter was used in
some earlier work and occasionally when a larger sample mass was required.
The results are given in Tables 10 and 11. Three internal standards were used
for the evaluation.
The analysis of quinolines is very erratic (see below), and when this
class is eliminated, average recoveries are those shown in Table 12. Inter-
pretation of these results is complex since there are most likely competing
effects. There is the possible loss of d-|Q-A or d^-P during the sample
handling process. Possible loss of the sample compounds during handling would
tend to compensate for this. Add to these effects of varying instrumental
precision with compound and internal standard, and an accurate assessment of
the limiting parameters becomes more difficult. Substantial amounts of d^Q-A
are lost in the 40-fold concentration step relative to dg-P and almost no d&-P
was lost relative to p-CT. In the 10-fold concentration of solvent other
factors seem to be involved, perhaps instrumental, and the changes are not
large relative to the variability of the recoveries.
29
-------
TABLE 9. EVALUATION OF RECOVERIES FROM XAD-2 CARTRIDGES
Compounds
Methylthiophene
Ethyl benzene
Ethyl thiophene
p-Ethyl toluene
Benzofuran
Indan
Indene
Phenol
o-Cresol
Dimethyl phenols
(2 isomers)
Naphthalene
Biphenyl
Anthracene
Mass Added
to Cartridge
(vg)
102
86
100
86
109
96
100
106
103
20
181
152
45
Percent Recovered
d Aa'b
Q10A
55 ± 8
63 ± 12
64 ± 10
91 ± 9
—
189 ± 22
42 ± 11
60 ± 2
66 ± 25
ND
127 ± 48
70 ± 42
135 ± 72
p-CTC
60 ± 25
67 ± 25
68 ± 27
83 ± 31
--
155 ± 49
42 ± 10
64 ± 25
63 ± 9
ND
122 ± 18
63 ± 4
124 ± 2.5
Replicate
p-CTC, d
72 ± 13
90 ± 10
73 ± 30
81 ± 28
26 ± 8
35 ± 8
33 ± 4
86 ± 6
100 ± 17
57 ± 7
230 ± 42
107 ± 9
102 ± 3
aAverage and standard deviation of triplicate analysis.
Internal standard was d1Q-anthracene added at the beginning of the extraction.
clnternal standard was p-chlorotoluene added after the evaporation.
Extraction repeated on another time.
30
-------
TABLE 10. RECOVERY OF COMPOUNDS FROM XAD-2 RESIN WITH EVAPORATION
OF SOLVENT FROM 5 ml TO 0.5 ml
Spectrum
Nrv
liU .
—
—
--
--
48
50
80
180
185
188
225
230
256
280
300
360
360
409
422
450
504
510
592
736
741
741
Compound
Pyridine
Methyl pyridine
Toluene
Methyl thiophene
Ethyl benzene
Ethyl thiophene
p_-Xylene
Benzofuran
1, 2, 4-Tri methyl benzene
Phenol
Indan
Indene
o-Cresol
jD-Cresol
2, 6-Dimethyl phenol
3, 5- Dimethyl phenol
Naphthalene
Quinoline
Methylindole
Indole
Biphenyl
Methylquinoline
Dibenzofuran
Phenanthrene
Anthracene
p-Chlorotoluene
d5-Phenol
dig-Anthracene
Percent
a
V
__
—
.
—
88
51
55
22
67
72
44
59
69
65
74
75
57
236
50
80
133
159
77
136
118
61
72
recovery
h
d5pb
—
—
--
—
99
66
72
37
88
93
60
79
116
110
80
127
57
531
60
107
159
190
91
139
142
80
130
relative to
p
p-CTc
—
--
--
--
33
83
90
37
110
116
75
99
116
110
72
122
87
360
74
135
196
234
113
202
175
--
126
162
ad,g- Anthracene
bd,-Phenol
b
Cp-Chlorotoluene
31
-------
TABLE 11. RECOVERY OF COMPOUNDS FROM XAD-2 RESIN WITH EVAPORATION
OF SOLVENT FROM 195 ml to 5
Spectrum
No
1 1 \j •
—
--
—
—
5
8
20
98
101
105
154
160
185
214
240
320
320
280
290
422
480
500
570
712
719
100
719
Compound
Pyridine
Methyl pyridine
Toluene
Methyl thiophene
Ethyl benzene
Ethyl thiophene
o-Xylene
Benzofuran
1 , 2 ,4-Tri methyl benzene
Phenol
Indan
Indene
o-Cresol
f-Cresol
,6-Dimethyl phenol
3,5-Dimethylphenol
Naphthalene
Quinoline
Methyl indole
Indole
Biphenyl
Methylquinoline
Dibenzofuran
Phenanthrene
Anthracene
p- Chi oro toluene
d5-Phenol
dig-Anthracene
Percent recovery relative to
i .a
dioA
_-
—
--
--
323
263
336
345
236
259
178
166
203
219
130
236
131
346
90
146
212
442
86
139
137
263
275
• w
K
d5Pb
__
—
--
—
114
90
111
115
84
89
62
59
72
78
68
101
78
204
26
52
60
125
24
39
39
90
_ _
34
p-CTc
_ _
.—
— _
_-
90
100
124
124
93
99
70
65
80
87
53
88
47
123
32
57
75
155
30
49
48
112
38
^g-Anthracene
bd5-Phenol
Cp-Chlorotoluene
32
-------
TABLE 12. AVERAGE PERCENT RECOVERIES FROM XAD-2
Degree of
Concentration
10-fold
40-fold
d]0-Aa d5-Pb p-CTc
78 ±29 94 ± 32 107 ± 46
201 ±79 72 ± 28 74 ± 28
d,Q-anthracene
Dd5-phenol
"p-chlorotoluene
Based on an idealized chromatographic peak and using a 1.7 sec scan
cycle, one can calculate the limits of error introduced solely by not
acquiring intensity data for an ion at the peak maximum. These data are
shown in Table 13.
TABLE 13. PERCENT DEVIATION OF RESPONSE RATIOS FROM IDEALITY
AS A FUNCTION OF PEAK WIDTH
Peak Width
10 sec
8 sec
6 sec
4 sec
Deviation
1 .7 sec cycle*
18.5
25
33.5
57
1.0 sec cycle
9.5
13.4
18.7
30
*Normal operating mode.
Actual measurements were made and calculated worst case results were not
realized. Results are given in Table 14. Relative retention times for
the compounds are given since, in general, peak widths increase with retention
time. An unexpected trend was observed in these data. The relative deviation
33
-------
TABLE 14. AVERAGE OF DEVIATIONS (%) OF RESPONSE
c*>
Compound
Toluene
Ethylthiophene
Pyrldine
Ethyl benzene
Methylpyridine
Phenol
p-Chlorotoluene
2,3-Benzofuran
1 , 2, 4-Tri methyl benzene
In dan
Indene
p_-Cresol
p_-Cresol
Naphthalene
3,5-Dimethylphenol
3, 4-Dimethyl phenol
Quinoline
Methylindole
Indole
2-Chloronaphthalene
Biphenyl
4-Methylquinoline
Benzofuran
Phenanthrene
d,Q-Anthracene
Anthracene
Relative
Retention
0.16
0.16
0.2
0.44
0.56
0.96
1.28
1.32
1.68
1.74
1.96
2.2
2.88
2.92
3.08
3.28
3.4
3.6
4.0
4.1
4.2
4.8
5.88
5.90
5.92
d,Q-Anthracene
9
11
29
32
39
30
27
31
25
23
29
22
31
38
12
19
10
1
13
6
3
6
11
2-Chloronaphthalene
19
17
14
14
21
11
13
12
16
4
12
7
11
7
5
3
6
7
9
16
12
10
27
p-Chlorotoluene
10
6
8
7
19
9
7
4
5
6
2
14
7
11
17
7
16
20
13
25
24
10
5
-------
was lowest for the internal standard with the retention time nearest to that
of the compound. It is unclear at this time whether this observation is due
to the chromatography or some aspect of the mass spectroscopy since the mass
of the ions used also correlate with retention times and the deviations.
Typical results from GC/MS analysis of the coal gasification process
streams collected on XAD-2 cartridges from three runs are presented in Table
15. In addition qualitative identification of a large matrix of compounds is
presented in Appendix III.
35
-------
TABLE 15. QUANTITATIVE GC/MS RESULTS FOR SELECTED COMPOUNDS ON XAD-2 RESINS
(yg of compound/g of coal)
Benzofuran
Indan
Indene
Phenol
Biphenyl
Naphthalene
Anthracene
Phenanthrene
Di benzof uran
Al kyl benzenes
Cresols
Xylenols
Thiophenes
Pyridines
Qu inclines
Indoles
Run 47
Wyoming
Surge
43
20
89
190
45
94
3
ND
0.7
>300
>110
56
74
ND
59
ND
Steady-State
7
3
36
48
1.5
56
0.4
ND
ND
112
36
21
14
ND
ND
ND
North
Surge
35
9
91
98
2
no
1
2
1
>460
77
160
78
ND
36
ND
Run 51
Dakota Lignite
Steady-State
2
0.5
7
11
0.5
19
0.5
1
0.5
100
5
1
2
ND
ND
ND
Run
Illinoi
Surge
3
1
10
25
1.0
52
0.6
1.0
ND
54
ND
ND
1
ND
ND
ND
52
s No. 6 Coal
Steady-State
2
1
5.5
15
2
100
1
2
0.2
62
ND
ND
3
ND
ND
ND
CO
ND = Not detected.
-------
5.0 ANALYSIS OF LIQUIDS
Comprehensive analysis of liquids generated during coal gasification can
be separated into three categories. Initial activities involve separation of
crude tar and aqueous condensate, determination of pH and total volume of
condensate and total weight and density of the crude tar. The second step in
the analysis involves fractionation of a crude tar sample into chemically
distinct classes of organic compounds by application of a wet-chemical parti-
tion scheme. In the final step, qualitative and quantitative analyses are
performed on specific partition fractions and the amount of phenol, cresols,
and xylenols in the aqueous condensate are ascertained.
5.1 SEPARATION OF TAR AND CONDENSATE
Crude tar and aqueous condensate generated by gasification are contained
in a trap designed to collect liquid byproducts. Prior to comprehensive
analysis, the tar and condensate must be isolated from each other. Separation
is performed by allowing the mixture to drain through a filtration system into
a separatory funnel. While a large portion of the crude tar is adsorbed
during filtering, some residue is carried onto the separatory funnel with the
condensate. This is removed by swirling the funnel gently, allowing the crude
tar to settle, then draining it off. The tar free condensate is poured into
a graduate cylinder where the total volume, pH, and color are recorded. The
sample is then stored pending further analysis.
Isolation of the crude tar begins by dissolving the residue, which was
separated from the condensate, in an organic solvent and transferring this
solution to the separatory funnel. An aliquot of water is added to facilitate
identification of the interface between the organic phase and residual aqueous
condensate. The organic phase is allowed to drain into a roundbottom flask,
and the solvent removed. The walls of the condensate trap and filter system
37
-------
containing the adsorbed tar are then exhaustively extracted, the extracts
combined, the solvent removed, and the total weight of the tar determined.
The tar is redissolved in solvent with aliquots removed for bioassay, ele-
mental analysis, density determination, and partitioning. Specific details
and results are presented in upcoming sections.
The filtration system used to adsorb the tar consists of a powder funnel
containing densely packed glasswool, followed by a funnel containing a fluted
cellulose-acetate filter, (50 cm, grade no.512). A 500 ml or 1000 ml separa-
tory funnel is used to collect the condensate. In general, methylene chloride
(Burdick and Jackson, or Fisher HPLC grade) is used to isolate the crude tar;
although ethyl acetate (Burdick and Jackson) was also used during a mass
balance study for chlorine. Solvent evaporation is performed by a Buchler
rotary evaporator. The total weight of crude tar and the tar density is
determined and the pH of the condensate is measured.
5.2 TAR PARTITIONING
g
Tar samples are exceedingly complex and require fractionation before
direct analysis can be undertaken. Other investigators have utilized either
of two procedures for this process: column chromatography or solvent partition.
Chromatographic methods separate the crude material into fractions of like
polarity and can function as a useful means of reducing a complex sample into
one of more manageable proportions. Solvent partition schemes have been
devised, most notably by researchers from the tobacco industry, in which
group separations are accomplished on the basis of similar chemical pro-
perties, e.g., acids, bases, etc. The latter approach is more practical,
particularly if fractions are to be chromatographed further. A detailed
schematic of the partitioning procedure used is shown in Figure 8 and is a
12
modification of the method utilized by Novotny for air particulate
extracts.
Five fractions are produced by application of the scheme: acids, bases
and three neutral fractions. These three fractions are designated nonpolar
(aliphatics and 1-2 ring aromatics), medium-polar (polynuclear aromatic
hydrocarbons-PNA's) and polar (oxygenated material). Each group is then
either analyzed directly by gas chromatography/mass spectrometry (GC/MS), or
is chromatographed using high performance liquid chromatographic (HPLC)
techniques.
38
-------
[Cruj> Tor (ig.)
WnhJXwHhlNNaOH
CH2CI2 byw , NtOH teywt
WahZXmth
H20
1
1
il
CH^Iiyw
WHh«rithCH2Cl2
NiOt
hy*r
r
CH2CI2liytr H20 hyw
WishZX
with 10%
(pH-10)
Adjust to pH 2 with 6NHC1
Extract «ritkCH2Cl2
HjSOiind I
IX with 2
20X|,
H2SO,
Wash with
cyclohflXMM
| OntoicAridi
AQ
CydohMim layw H20 toytr
(1) Evtpont* to
dfyntss.
(2) Dimln in CH^
Mi«tt*»H2«rith1NHC1
Extract 3X with CH2CI2
With 2X with
H20
OrfwieAeidi
Wish2X
I
AQ
CH2CI2 liyir
H2804lty.r
AdJ8ifto»H12wit(i10NNiOH
Extract 3X with CHi-C^
CH2CI2hy« ---- . H2OUv«
I fph-W
OtftnitBtttt
I
AQ
»"*BXwithCH2CI2
H20toy«
(3)WMb3Xwith
4:1CH3OH/H20
|AdjuittopH12with1NNiOH
Extract 3X with CH2CI2
I
Cydohoxtm Ityir
(DCoMMtnti
I
CH3OH/H20 toyor
With 4X with cyclohtxtM
CH3OH/H20 (tyor
Fnondry
Figure 8. Tar partitioning schematic.
39
-------
5.3 GC/MS ANALYSIS
As with vapor phase samples, qualitative analysis of fractions subjected
to GC/MS is achieved by comparison of the mass cracking pattern of the unknown
mass spectra to an eight major peak index of mass spectra. Quantitative
results are obtained via internal standard procedures using the detector
response of the peak intensity as provided by the mass spectrometer. Internal
standards consist of deuterated materials, phenol -d^, quinoline-dy and
anthracene-d-|0. This process can be applied to single species as well as,
in certain cases, to classes of compounds.
For single species, in order to eliminate the need to construct complete
calibration curves for each compound to be quantitated, the method of relative
molar responses (RMR) was used. The calculations are as follows:
A /moles
RMR - 0)
where A = peak area or other suitable response parameter (u and s designating
unknown and internal standard).
Equation (1) is equivalent to:
RMR •
where g = grams and mw = molecular weight.
Rearranging,
_ Au • mwu • g$
9u ~ AS • mws • RMR
The value of the RMR is determined from at least three independent analyses.
40
-------
The basis of class quantitation by GC/MS lies in the selection of a mass
spectral fragment ion which is common and unique to all members of a class.
If, in addition, that ion is of approximately the same relative intensity for
all, or nearly all, class members, then all compounds can be quantitated on
the basis of that single ion. In practice, no ion is unique to one class,
but for samples which are highly enriched in single compound types, the
presence of a characteristic ion for one type is not unusual. Research of a
mass spectral catalog of fragmentation patterns of alkylated indoles,
pyridines, quinolines, and naphthalenes, revealed an ion that meets the
criteria outlined above for each of these classes. Using the mass spectro-
meter to generate single ion plots (using the characteristic fragment ion)
peaks, are obtained for every class member. Intensity data are then avail-
able for use to quantitate every compound. Summation of the results for
every compound provides a reasonably accurate assessment of the amount of
material associated with individual classes. The acid fraction is analyzed
by either glass capillary GC/MS or HPLC. The latter method has recently been
developed, and owing to its greater speed and reliability, will be routinely
utilized for future analyses.
For GC/MS analysis, an aliquot of the acid fraction (spiked with phenol-
dr- as internal standard) is introduced into the GC/MS system. The intensities
b
of those peaks corresponding to phenol, cresols, xylenols and higher alkylated
phenols are recorded and compared to the intensity of the internal standard.
Using response factors (previously calculated from reference standards),
quantitation is carried out on each compound.
For analysis of tar acids by HPLC the method used is the same as is
described below for gasifier condensate. Aside from speed and ease of
analysis, the HPLC appproach utilizes the aqueous solution of acids instead
of the methylene chloride extracts which are used for the GC/MS procedure.
This is important since the difficulties associated with the extraction of
phenolics from water probably results in an inaccuracy of the GC/MS method.
The tar bases are analyzed directly by glass capillary GC/MS. The
internal standard used here is quinoline-dy, and quantitation is performed as
described for tar acids. Compounds routinely analyzed are: alkyl pyridines,
quinoline, alkylquinolines, alkylbenzoquinolines and acridines.
41
-------
The subfractlonation of the neutral fraction consists of solvent parti-
tion of a cyclohexane solution of the neutrals with methanol/water (1:4) and
nitromethane. The polar neutrals are recovered in the methanol/water fraction
and the medium-polar neutrals are associated with the nitromethane fraction.
Those materials remaining in solution in cyclohexane comprise the nonpolar
neutrals, and are analyzed directly by GC/MS using wall coated open tubular
(WCOT) glass capillary columns.
The medium-polar fraction containing PNA hydrocarbons is obtained by
preferential partition from cyclohexane into nitromethane. Analysis is
carried out by GC/MS using the same column as described earlier for nonpoTar
neutrals.
Those neutral materials that are washed into methanol/water (1:4) from
cyclohexane are termed polar neutrals. The nature of the compounds comprising
this function is such that direct analysis by GC/MS is not practical. Various
modes of high performance liquid chromatography (HPLC) have been carried out
with limited success.
Because of the intimate admixture of the coal tar and condensate in the
tar/water trap, a large quantity of organic matter is associated in the
condensate. The bulk of this material is comprised of the more water soluble
phenols. Although analysis for various hydroxylated aromatic compounds were
initially carried out by GC/MS, due to the well known difficulty of extracting
phenolics into organic phases and the slower turn-around associated with
GC/MS samples, a more direct analytical method was sought.
The use of HPLC for the analysis of phenolics is relatively straight-
forward, and a procedure utilizing this technique, was developed. Because of
their overwhelming abundance in the condensate sample compared to all other
hydroxyaromatics, only phenols, cresols (3 possible isomers) and xylenols (6
possible isomers) were analyzed. Chromatographic conditions were such that
single peaks were obtained for phenol, all cresol isomers, all xylenol isomers
and the internal standard, isopropyl phenol, in that order. The HPLC system
uses a reverse phase column with methanol/water as mobile phase. The UV
transparency of this solvent system permits detection at 215 nm, a wavelength
where all cresols and xylenols have about the same extinction coefficient.
Thus, the quantisation of coexisting species is valid, and only three
42
-------
calibration curves need be prepared instead of one for every isomer. Using
reference standards, calibration curves for the three compound groups are
prepared and the regression equations (least squares) determined. In all
cases the square of the correlation coefficients was greater than 0.99. The
curves were validated through the analysis of known samples.
Condensate samples are pH adjusted (~pH 4.5) to ensure that phenolics are
present in their acid form, a known amount of internal standard is added and
equilibrated and an aliquot is injected onto the hplc. Peak areas are deter-
mined by electronic integration and amounts of phenol, cresols and xylenols
are calculated from the appropriate regression analysis.
Typical results for the analyses of selected compounds in tar and aqueous
condensate from three runs using Wyoming, North Dakota and Illinois No.6 coals
are presented in Tables 16 through 19. In addition, qualitative identifications
of a large matrix of compounds in various tar fractions are reported in Appendix
III.
TABLE 16. ELEMENTAL ANALYSES* OF CRUDE TAR
(expressed in wt %)
Run No.
47
51
52
C
81.4
82.9
85.5
H
7.8
6.9
6.5
N
1.1
1.0
1.0
s
1.5
2.5
1.9
0
7.8
6.5
4.9
*Analyzed by Galbraith Laboratories, Knoxville, TN.
43
-------
TABLE 17. TYPICAL PRELIMINARY AND PARTITION RESULTS FOR TAR AND CONDENSATE
Run
No.
47
51
52
Coal
Wyomi ng
N, Dakota
Lignite
Illinois
No. 6
Total
Tar
(g)
28.8
17.7
49.3
Tar
Density
g/ml
1.14
1.33
1.39
Acids
wt %
31.0
23.5
19.8
Base
wt %
2.6
4.5
2.7
Ins.
wt %
2.0
11.0
9.0
NPN
wt %
27.8
17.1
16.8
PN
wt %
18.1
7.7
7.2
PNA
wt %
18.5
136.5
44.6
Condensate
Volume
pH ml
11 500
9 700
8-9 640
Organic
Extractables
9
4.2
4.2
NA
Ins: Insoluble in cyclohexane.
NPN: Nonpolar neutrals.
PN: Polar neutrals.
PNA: Polynuclear aromatics.
-------
TABLE 18. QUANTITATIVE RESULTS FOR SELECTED COMPOUNDS
IN TAR FRACTIONS
(expressed in grams)
Fraction
PNA
Organic
Bases
Organic
Bases
Naphthalene
Alkylnaphthalene
Fluorene
Dibenzofuran
Anthracene +
Phenanthrene
Fluoranthene
Pyrene
Chrysene
5-Ring Compounds
Alkylpyri dines
Quinoline
Alkylquinolines
Alkylbenzoquinolines
Acridine
Phenol
Cresols
Xylenols
o-isopropyl phenol
trimethyl phenol
Run 47
0.06
0.06
0.03
0.02
0.05
0.02
0.02
0.01
Trace
0.007
<0.0005
0.62
1.13
1.36
0.02
0.12
Run 51
0.57
0.20
0.09
0.08
0.22
0.05
0.04
0.04
0.01
0.04
0.04
0.01
0.01
0.003
0.39
1.19
0.17
0.25
0.05
Run 52
2.44
0.38
0.27
0.31
1.31
0.45
0.31
0.23
0.18
0.03
0.01
0.01
0.18
0.003
0.02
0.05
0.02
<0.0005
0.001
GC/MS Conditions: LKB-2091 GC/MS, 15-20 m WCOT capillary column, 1% Se-30/
BaC03, 100°/2 min/8° min/265°.
45
-------
TABLE 19. QUANTITATIVE RESULTS FOR PHENOLS, CRESOLS, AND XYLENOLS
IN CONDENSATE BY REVERSE PHASE HPLC AND GC/MS
(expressed in grams)
Run No. Phenol Cresols Xylenols Method
47
51
52
0.58
2.03
1.93
0.28
1.04
1.02
0.10
0.16
0.12
GC/MS
HPLC
HPLC
GC/MS:
HPLC:
Finnigan 3300 GC/MS
1000/8°min/250°.
Reverse phase HPLC
600 psi; SF-770 VV
, 24 m SCOT 1% SE-30
analysis, 1:1 MeOH/H-
variable wavelength a
glass capillary column,
0, Water's
etector, 2
Associates, pump
ml/min.
46
-------
6.0 ATOMIC ABSORPTION METHODS
Selected trace elements are analyzed by atomic absorption (AA) following
various digestion procedures. The following sample preparation and analysis
scheme is used.
6.1 SAMPLE PREPARATION AND ANALYSIS
The various trace elements that have been analyzed include Hg, As, Se,
Sb, Cd, Pb, Ni, Cu, Cr, and Be. They can be subdivided into three groups
based on the sample preparation and analysis procedure. Methods for Hg are
unique and it can be classified as Group 1; As, Se, and Sb are in Group 2, and
Cd, Pb, Ni, Cu, and Be are in Group 3. A further subdivision is possible
based on the sample matrix. During the course of this project, five different
types of samples have been analyzed. These include parent coals, reactor
residues, tars, aqueous solutions (condensate or scrubber) and sorbent
materials for gaseous effluents. A mass balance has been attempted from the
results for various tests and is discussed further in Section 6.3. Sample
preparation and analysis procedures for the three groups and the various
matrices are given in the following subsections.
6.1.1 Coals and Residues
For Hg analysis a 1-g sample of coal prepared for gasification (approxi-
mately 6-14 mesh) is placed in a platinum cup. The sample is then placed in a
quartz-lined Parr oxygen bomb calorimeter containing 10 ml of 10 percent
nitric acid. The vessel is pressurized to 24 atm of Q£ and the sample is
combusted. After cooling, the resulting solution is then diluted to 25 ml
with 10 percent nitric acid and submitted for analysis by the cold vapor
1415
technique. ' For Hg in the residue, 1 g of residue material is mixed with
0.5 g benzoic acid (after being ground to approximately 200 mesh), subjected
to Parr 02 bombing and analyzed as for the coal procedure. For Group 2 and
Group 3 elements, the residue is heated and analyzed in exactly the same way
as coals which is described below.
47
-------
For Group 2 elements, the coal (or residue) is crushed to 200 mesh and
subjected to one of two procedures. The first procedure is based on wet acid
ashing; 0.1 g of coal sample (or 0.03 g residue sample) are treated to a
mixture of 3 ml concentrated HLSO. and 5 ml concentrated HNO-, for 1/2 hour on
an aluminum block heated to 200°C. After 1/2 hour, the mixture is cooled and
treated with an additional 2 ml of concentrated perchloric acid and digested
overnight for removal of the nitric and perchloric acids. The mixture is then
cooled and 7 ml distilled water, 2 ml 20 percent ascorbic acid and 1 ml 0.06 M
Nal are added, respectively, mixing after each reagent addition. Finally, the
mixture is extracted with 2 ml of toluene. The toluene phase is then analyzed
for antimony using electrothermal atomic absorption. One ml of 6.0 M Nal is
then added to each mixture and reextracted. The toluene phase is then analyzed
for As and Se by electrothermal AA. Sixty pi of 0.1 percent Ni (N03)2 is
first added to the furnace and dried. Twenty yl of sample is then added and
analyzed. The entire procedure takes, two to three days.
A second procedure may be used for coal (or residue) analysis. The
coal (1.0 g) or residue (0.2 g) is subjected to a dry ashing at 750°C after
mixing with 1 g of Eschka mixture (MgO and Na2CO.J. After four hours the
mixture is cooled, dissolved in concentrated HC1 and diluted to 100 ml with
distilled water. The resulting solution is analyzed for As, Se, and Sb by
hydride generation AA. Here the analyte is reduced to the respective hydride
with sodium borohydride and the resulting gas is swept into an argon-hydrogen
flame for subsequent analysis.
For Group 3 elements, the coal (or residue) is dry-ashed at 450°C for
eight hours. After the ash content is determined, the sample is then sub-
jected to acid digestion and analyses by AA; 0.1 g of ash is placed in a Parr
acid bomb teflon cup and wetted with 1 ml aqua regia. Three ml of concen-
trated HF is added to the mixture. The resulting mixture in the teflon vessel
is placed in the stainless steel bomb and digested in an oven for 150°C for 30
minutes. The vessel is cooled, and the mixture is treated with 2.8 g boric
acid. The resulting solution is diluted to 50 ml with distilled water and
18
analyzed for Cd, Pb, Ni, Cu, Cr, Be and others by electrothermal AA.
48
-------
6.1.2 Tars
The tar material is extracted out of the condensate traps with
methylene chloride. The extract is evaporated at room temperature under a
stream of N2- For mercury, 1 g of tar is Parr 02 bombed and analyzed by the
procedure given in Section 6.1.1. The resulting residue is dissolved in 1 ml
of aqua regia and subjected to electrothermal AA for analysis of lead and
arsenic.
6.1.3 Aqueous Condensate and Scrubber Solutions
The water condensate is extracted with methylene chloride to remove
the organics. The methylene chloride phase is combined with the tar con-
densate materials in Section 6.1.2. The pH of the aqueous phase is adjusted
to 4.0 and extracted with methylene chloride to remove the phenolic material.
The aqueous phase is then analyzed directly by the cold vapor technique for
Hg and electrothermal techniques for the other elements as described in
Section 6.1.1. Acidic scrubbers are analyzed by direct aqueous injection
onto the AA.
6.1.4 Sorbent Materials for Gaseous Effluents
These materials include XAD-2 resins and charcoal traps. Again the
sorbents are subjected to Parr 02 bombing and analyzed for Hg, As, and Pb by
the appropriate AA procedures mentioned in Section 6.1.1.
6.2 QUALITY CONTROL PROCEDURES
All samples are analyzed in duplicate. Wherever possible, the NBS
(National Bureau of Standards) Standard Reference Materials are used in con-
struction of calibration curves. These include SRM 1632a bituminous coal,
SRM 1633 flyash and SRM fuel oil which are analogous in matrix type to the
coal, residue, and tar materials encountered in this study. Wherever NBS
SRM's were unavailable, the method of standard additions was employed to
compensate for matrix mismatching between sample and standard.
49
-------
6.3 TYPICAL RESULTS
Figure 9 shows the AA analysis for seven trace elments in six U. S.
19
coals. For comparison, results from Gluskoter, et al. are presented. Good
qualitative agreement is found. Figure 10 shows the percentages of trace
elements lost to gaseous and liquid effluents for Pittsburgh No.8 coal.
These are obtained by difference between the amounts in the coal and the
20
residue. Again for comparison, results of Attari are presented. Finally,
Table 20 shows overall material balance and the fate of trace elements from
gasification of Illinois No.6 coal. Mercury is almost totally evolved and is
expected to end up in the gas. Gas phase analyses are as yet not available
but can be crudely estimated by difference. Almost twice as much chromium is
recovered than fed. This may be a contribution of the stainless steel reactor.
TABLE 20. AVERAGE MATERIAL BALANCE FOR THREE ILLINOIS NO. 6 COAL
GASIFICATION RUNS
As
Be
Cd
Cr
Hg
Pb
Se
Coal in
5050
1396
183
27. 5K*
188
3716
2407
yg (Average)
Char Tar HgO
3056 176
1265
139 2
52K*
7.1
1639 41
576 27
633
—
7
0.90K*
--
65
668
Vapor
1768+
NA
NA
NA
NA
NA
NA
Material
Balance
76.5% M
dm)
90.6%
80.9%
192.0%
3.8%
45.0%
53.0%
*K = 1000.
**Accounting for vapor.
^Estimated from results obtained using a 100 mg charcoal trap.
NA: Not analyzed.
50
-------
.30 -
2 .20
.10
Hg
Pb
1 • T
I
i ill t i
I P K M W N
9.0 -
•S?
t 6.0
3.0
.
I P K M W N
Se
3.6
2A
1.2
11
I P K M W N
Legend:
I. Illinois #6
P. Pittsburgh #8
K. Kentucky //9
M. Montana Rosebud
W. Wyoming Sub-bit.
N. North Dakota lignite
Note: RTI analysis results
shown relative to
range of Gluskoter,
et.al.
Raf. Gluskoter et a).. 6-77
Figure 9. Range of selected trace elements in coals.
51
-------
As
18
12
6
i t
I P K M W N
1.8
I 1'2
0.6
Be
1 II
I 1
i P K M W N
.30
f -20
.10
LrO
t
-•
•
•
i
i
i
1
*
i
i
<
>
i
30
f 20
10
1
*
, |
•
1 I
ur
>
«
i
i .
A*
\ m
IPKMWN IPKMWN
Reference: H. J. Gluakoter. et al. "Trace Elements in
Coal: Occurence and Distribution"
Illinois State Geological Survey
report to IERL-EPA 6-77
Figure 9. Range of selected trace elements in coals, (continued)
52
-------
100
80
60
40
20
Ref: A. Attari
Institute of
Gas Technology
8-73
A. Attari
RTl
As Be Cd Hg Pb Se
Figure 10. Percent loss of effluents (Pittsburgh No. 8 coal).
53
-------
7.0 FOURIER TRANSFORM INFRARED SPECTROMETRY
Application of developmental work has been initiated for the Fourier
Transform Infrared Spectrometer (FTIR) to the analysis of coal gasification
samples in three distinct areas: (1) high resolution FTIR of gas sample, (2)
gas chromatography/FTIR of semivolatiles, and (3) low resolution FTIR of coal,
tar and ash.
Results from the FTIR analysis at coal gas were confirmed for 13 com-
pounds and five additional compounds were identified. Two compounds were
quantitated. High resolution FTIR, a nearly universal detector, is especially
useful for determining the presence or absence of specific pollutants of
interest which are not easily determined by routine gas chromatographic
methods. The technique also serves as an independent check on the quantita-
tion obtained by GC.
Over 24 individual compounds were identified in an XAD-2 extract from the
GC/FTIR spectra, including benzene, phenol, naphthalene, and some of their
alkyl derivatives. In many cases, positional isomers were distinguished
(e.g., 2,4-xylenol and 2,6-xylenol). Compounds were identified by comparison
with gas phase spectra of standards. Selected compounds were also quantitated.
This technique is superior to GC/MS in its ability to distinguish positional
isomers. The relative utility of class quantitation by GC/FTIR versus GC/MS
is still under evaluation.
The analysis of coal, tar, and ash samples by FTIR indicates that general
information regarding types of inorganics and organics may be obtained. The
qualitative and quantitative capabilities of the technique for these samples
are under further investigation.
7.1 HIGH RESOLUTION ANALYSIS OF GAS SAMPLES
High resolution FTIR is particularly suited for the analysis of low
molecular weight polyatomic gases produced during coal gasification. At high
resolution (0.125 cm"1), the sharp lines produced by rotational energy level
54
-------
transitions are easily distinguished and measured for qualitative and quanti-
tative work. Quantities of gases are determined by comparing the absorption
of a preselected sample spectral line with that of a standard. Thus it is
possible to simultaneously identify and quantitate every gas species present
which absorbs IR radiation and which has had a standard curve generated (H2,
N2, and 02 do not absorb; also quantities of H20 and C02 cannot be determined
due to interference from these species within the optical path of the spectro-
meter).
Gases were received in 500 ml glass gas bulbs fitted with teflon stop-
cocks. A sample was attached to an evacuated (<2 mmHg) Perkin-Elmer Model
186-035 10 m Variable Path Gas Cell and expanded into the cell to equilibrium.
The cell volume was 4.4 liters and the optical path set at 8.76 m. The cell
was fitted with NaCl windows. The pressure was brought to 1 atm with N2 gas
at ambient temperature. The sample spectrum was then collected (100 scans),
Fourier transformed, and ratioed to a previously collected background spectrum
of N2- Identification was made by comparison with standards or reference to
published literature.
Standards of COS, NH3 and HCN were prepared by making appropriate
dilutions of pure gas standards with No- The standards were introduced into
the cell in the same manner as the samples. A series of dilutions for each
gas was made, the spectra collected, and a plot of absorbance versus wave
number for a characteristic absorption line prepared. The absorbance for a
corresponding wave number in the sample gas was then compared to the standard
and the concentration for that species determined.
Four gas samples were analyzed by High Resolution FTIR. Quantisation
was performed on COS and estimated on HCN. Table 21 shows the concentrations
found in the four gas samples. Estimates were made on percent composition
for gases qualitatively identified based on absorbance relative to known
species. It was notable that surge samples showed significantly higher
amounts of COS and C2 hydrocarbon gases than did the steady-state samples.
There was little correlation among steady-state samples, however. Principal
components of all samples were CH4, CO, C02, H^O and COS. A total of 13
confirmed species and 5 suspected components have been identified in the four
samples.
55
-------
Table 21. HIGH RESOLUTION FTIR ANALYSIS OF SELECTED GAS SAMPLES
Species Run 49 Run 49 Run 50 Run 50
Steady State Surge Steady State Surge
ru i i • J.J.A J.J. j-j.
UMyt TTT TTT TT TT
C2Hg - -H- TR -H-
C2H4 " ++ - ++
C2H2 - - - +
C6H6
m 4-4. 4-4- 4-4- XX
LU TT TT TT TT
Pfi i | I i i i XXJ. .l~i~l.
\j\Jn II T TTT TTT TTT
COS TR 120 96 370
C2H40 - TR?
CH3SH - TR
(CH3)2S - TR?
CH3C1 ....
CH3NH2 ....
HC1 -
HCN >148 - - <148
HCHO - TR
NH3 ? -
AsH3 -
Ni(CO)4 ....
(CN), .?.?
CH3OH - T TR
CH2CHCHO - TR
NOX -
coci2 -
HN03 - TR?
Legend:
> 10% Gas Composition
++ 1-10% Gas Composition
+ 10 ppm - 1% Gas Composition
TR Lowest detectable confirmation of presence <10 ppm
— None detected
? Assignment uncertain; interfering absorbance or no reference
standard
Concentrations expressed in PPM for quantitatively analyzed species
56
-------
7.2 GC/FTIR ANALYSIS OF SEMIVOLATILES IN XAD-2 EXTRACT
Because of its speed and sensitivity, FTIR can be used as a detector for
gas chromatographic effluents (GC/FTIR). The technique has been applied to
the analysis of the XAD-2 extracts which have heretofore been analyzed by gas
chromatography/mass spectrometry (GC/MS). Infrared is much better at certain
class distinctions than mass spectrometry, due to the characteristic fre-
quencies for all compounds containing a given functional group. Thus, GC/FTIR
may prove superior to GC/MS for routine quantisation of such classes as "total
aromatics," "total phenolics," etc.
The goal of this study was to demonstrate the feasibility of GC/FTIR for
the analysis of semivolatile coal gas effluents and determine whether it is
suitable as merely a confirmatory technique, or whether it offers advantages
over GC/MS which would make it a primary analytical technique.
A sample of extract from XAD-2 resin of gasifier run 41 was qualitatively
analyzed on a 2 mm ID x 180 cm glass column packed with 3 percent OV-17 on
Chromasorb G (AW) 100/120 mesh. The column was installed in a Varian 3700 GC
equipped with a thermal conductivity detector and interfaced to the Nicolet
7199 FTIR system through the Nicolet GC-7000 GC/FTIR Accessory.
An injection of 1.0 yl of the XAD-2 extract was chromatographed under the
following conditions: temperature program (50°C/3 min to 240°C/10 min at
8°/min); He flow (30 ml/min); T... (222°C); TC detector temperature (280°C).
Infrared spectra consisting of four low resolution scans were collected when
the eluting species exhibited an absorbance greater than 0.005 over any of the
five preselected windows corresponding to specific chemical group absorptions.
The resulting chromatogram of selected infrared bands, called a "chemi-
gram" is shown in Figure 11. The band from 3600-3700 cm is selective for
compounds containing the hydroxyl group (the three large peaks are phenol and
the cresol isomers). The 3025-3100 cm" band is characteristic for aromatics
and/or alkenes. The 1630-1730 cm~ band is characteristic for carbonyl-
containing compounds. Unfortunately some noncarbonyl infrared absorbances
(notably from overtones in the naphthalene spectrum) interfered; no carbonyls
were detected except for an artifact. The 1560-1630 cm" band is charac-
teristic for aliphatic compounds. The 1475-1520 cm band is characteristic
for aromatics and is therefore a cross-check with the 3025-3100 cm band.
57
-------
96DO - 37OO
1"|75 - ISZO
1O90 1200 1320
TIME. SECONDS
Figure 11. Chemigram of infrared bands (cm"1) versus time during GC/FTIR analysis of
extract of XAD-2 resin sampled during surge phase of run 41.
-------
Interpretation of the chemigram indicates that most (but not all) of the
compounds are alkyl aromatics. There are a few phenolic compounds. Table 22
gives the qualitative identifications. Aliphatic compounds do not appear to
be significant.
7.3 LOW RESOLUTION FTIR OF COAL, ASH, AND TAR
The coal, ash, tar and other complex liquids or solids may be analyzed
with minimal sample preparation to give general characteristics of the sample.
The spectra contain information about both inorganic (sulfate, carbonate
silicate) and organic constituents (aromatics, aliphatics, and carboxylates).
Alternatively, the spectra may be used simply as a pattern characteristic of
a specific sample class (e.g., coal rank).
Two coal samples have been thoroughly examined—Illinois No.6 and North
Dakota lignite. The coal and resultant ash were finely milled and pressed
into KBr pellets before their low resolution (8 cm" ) spectra were collected.
The tar was dissolved in CHgCl,, and placed in a liquid solution cell to
obtain its IR spectrum.
The spectra of the coal, tar, and ash from Zap lignite and Illinois No.6
coal are presented in Figures 12 through 14, respectively. A comparison of
the spectra for the two coals is presented in Figure 12. The weak band at
3058 cm in the Illinois No.6 spectra indicates greater aromaticity. The
intensity of the shoulder at 1725 cm" on the major peak at 1610 cm" is
indicative of the carbonyl content. Zap lignite appears to contain more
carbonyl species. The peaks at 1095, 1035, 540, and 473 cm" in Illinois No.6
have been attributed to inorganics (vide infra). This would indicate that
Illinois No.6 contains more inorganic species.
A comparison of the tar spectra for the two coals (Figure 13) indicates
that the tars are remarkably similar. The major difference is the much
higher aromatic content of the Illinois No.6 tar (3058 cm" ).
The two ash spectra (Figure 14) are completely different. In the Zap
lignite ash, the peaks at 1451 (strong), 880 (sharp peak at low end of broad
_1
"hump"), and 710 cm (weak, sharp) have been tentatively assigned to a
carbonate species. The peaks at 1190 (shoulder), 1109 (strong), and 622 cm"
(medium, sharp) have been tentatively assigned to a sulfate species. In the
59
-------
TABLE 22. COMPOUNDS IDENTIFIED IN XAD-2 SURGE SAMPLE FROM RUN 41
File No.* Compound
11 Methyl chloride
15 Benzene
29 Toluene
42 m-Xylene (tent.)
49 o-Xylene (tent.)
56 Ethyl benzene (tent.)
62 1,3,5-Trimethylbenzene (tent.)
66 1,2,4-Trimethylbenzene (tent.)
71 Alkylbenzene
79 Phenol
89 Monosubstituted alkylbenzene (tent.)
94 Indan (tent.)
100 Indene
106 o-Cresol and p-cresol
113 m-Cresol
124 Alkylbenzene
129 2,6-Xylenol
131 2,4-Xylenol
142 Naphthalene
151 1-Methylnapthalene (tent.)
156 2-Methylnaphthalene (tent.)
161 Alkyl aromatic
165 Alkyl artomatic
170 Ester (background)
177 Dimethyl naphthalene isomer (tent.)
183 1,4-Dimethylnaphthalene (tent.)
201 d-jQ-Anthracene (std.)
*File numbers are located along x-axis of chemigram.~~
60
-------
o
o
9
o
o
o
o
in .
u
in
co
o
o
3t
OD Z970
2110 1680 1250
WflVENUMBERS
820
3SD
Figure 12. Comparison of coal spectra. Top = Illinois No.6; Bottom = Zap
lignite.
61
-------
o
o
•
o
o
«— «
o
in .
CJ
~
Jin
in
zo
o
o
atbo ssVo es¥o 2110 isso izso sea 390
WflVENUMBERS
Figure 13. Comparison of tar spectra. Top = Illinois No.6; Bottom = Zap
lignite.
62
-------
3100 2970 25\0
2110 1680
WflVENUMBERS
1250 820
330
Figure 14. Comparison of ash spectra.
lignite.
Top = Illinois No.6; Bottom = Zap
63
-------
Illinois No.6 ash, the broad hump at 1082 cm" and the small doublet at 790
cm have been tentatively assigned to quartz.
In summary, FTIR appears to be a useful technique for solid, liquid and
gaseous samples related to coal gasification. GC/FTIR is currently usable as
a confirmatory technique to GC/MS. Further work is needed to compare GC/MS
and GC/FTIR for class quantitation capability. High resolution FTIR of coal gas
is useful for validation of routine GC analyses and also for analyses of certain
compounds which cannot be determined by conventional GC detectors. Low resolution
FTIR analyses of coal, tar and ash is under further investigation and its utility
for such analyses needs to be fully assessed.
64
-------
8.0 SPECIAL TECHNIQUES FOR ANALYSIS OF TARS
AND AQUEOUS CONDENSATES
Of special importance is the characterization of the complex tar pro-
duced during coal gasification. In Section 5.0, various procedures for
partitioning and analysis of tar and condensate were described. In this
section, additional studies are described which were undertaken for further
characterization of the liquid effluents. These include polynuclear aro-
matics (PNA) analysis using glass capillary gas chromatography, boiling range
measurement using simulated GLC-distillation, class characterization using
element specific detection and ion chromatographic analysis of the aqueous
samples.
8.1 PNA ANALYSIS USING GLASS CAPILLARY GAS CHROMATOGRAPHY (GC)2
o
(GC) has been applied for quantisation of PNA materials in tar. A
chemically bonded temperature stable (300°C) methyl-silicone capillary
column was used. The conditions for analysis are shown in Table 23. The
21
'GROB' split!ess method of sample injection was used and approximately 5-
15 yg were injected for detection of the heavier PNA's, e.g., benzo(g,h,i)
perylene. The splitless technique consists of injecting 2 to 3 yl of the
sample and then 30 seconds later, opening the splitter to remove- the excess
solvent. This prevents a long solvent tail as illustrated in the accompanying
chromatogram (Figure 15) in which 21 PNA's have been identified based on
retention times of standards. Table 24 shows the analysis of several com-
pounds in the PNA fraction of the coal tar for three runs. A comparison is
presented for a few compounds for which GC/MS data are available. At present,
it is planned to use this temperature stable column in a GC/MS for confirmation
of these compounds. In the future, the capillary-GC-FID technique will be
routinely applied to PNA analyses of coal conversion tars and condensates.
65
-------
TABLE 23. CONDITIONS FOR (GCr ANALYSIS
Instrument:
Column:
Injector:
Detector:
Oven:
Carrier Gas:
Makeup Gas:
Hydrogen:
Air:
Varian 3700 with all-glass capillary system.
High temperature OV-101, 25 m x 0.25 mm ID,
WCOT, glass (obtained from Quadrex Corporation,
New Haven, Connecticut).
Splitless (GROB), solvent vented 30s after in-
jection, 250°C
FID 280°C
40°C for 1 min, 4°C/min to 270°C and hold 40 min.
Helium 1.25 ml/min
Helium 29 ml/min
30 ml/min
300 ml/min
66
-------
CTV
I i
1. Naphthalene
2. 2-Meth. Naph.
3. 1-Meth. Naph.
4. Acenaphtylene
5. Acenaphthene
6. Fluorene
7. Phenanthrene
8. Anthracene
9. Fluoranthene
10. Pyrene
11. Benzo(a)fluorene
12. Benzo(b)fluorene
13. Benzo(a)anthracene
14. Chrysene &
Triphenylene
15. Benzo(b)fluoran-
thene
16. Benzo(k)fluoran
thane
17. Benzo(e)pyrene
18. Benzo(a)pyrene
19. Perylene
20. Dibenzo(a, h)an-
thracene
21. Benzo(g. h, Opery-
lene
4
7"TTT
1
---
1
[j •
!
I '
• f i '
r i
1
\jf
I
; '
II
,i
•-
1
,
§--•
OB
O
o""
s
** 111
O III
J
If
III
I
r
'V
<
/
LJ
O
3
3
(
INS
1
TRU
- t
MEN!
1 i '
1MII
I'
:
*!
.i
f:
M
3X
18,
•
1
V
CARRIER GAS: H
MAKE UP: h
DETECTOR: F
HYDROGEN: 3
AIR: 3
INJECTION
MODE: '(
V
COLUMN: 2
TEMPERATURE: 4
SOLVENT: n
SAMPLE SIZE
1
1
'!
AMP FULL SCALE
19-
i 20,
VARIAN 3700
HELIUM 1.25 ml/min
HELIUM 29 ml/min
FID 1280° C)
30 ml/min
300 ml/min
GROB' SPLITLESS (250° C),
VENT IN 30 SECONDS
25 m X 0.25 mm I.D. WCOT.
HIGH TEMP. OV101 (280° +)
C FOR 1 min. TO 270° C
AT 4° C/min, HOLD 40 min
n-C6&CH2C!2(1:1)
' I
94 MIN
Figure 15. PNA fraction of coal gasification tar by capillary GC-FID: Determination of Selected PNA's.
-------
TABLE 24. SELECTED COMPOUNDS IN PNA FRACTION AS WEIGHT % OF
COAL GASIFICATION TAR (MS vs. capillary-FID)
Compound
Naphthalene*
Benzothiophene
2-Methyl naphthal ene
1 -Methyl naphthalene
Biphenyl
C2-Naphthalenes
Acenaphthylene*
Acenaphthene*
Dibenzofuran
Fluorene*
Phenanthrene*
Anthracene*
Benzidine*
Carbazole
Fluoranthene*
Pyrene*
Benzo a)fluorene
Benzo bjfluorene
Benzo a) anthracene*
Chrysene* +
Triphenylene
Benzo(b)fluorenthene*
Benzo( k) f 1 uoranthene*
Benzo (e)pyrene
Benzo(a)pyrene*
Perylene
Indeno(l ,2,3-cd)pyrene*
Dibenzo(a,h)anthracene*
Benzo(g,h,i)perylene
Run 21
MS
5.73
--
1.00
—
--
—
--
--
0.74
0.80
2.30
--
--
—
1.30
0.9
—
—
—
0.80
—
—
—
--
--
—
--
Capillary-
FID
2.91
0.53+
0.38
0.33
0.26+
1.07
1.15
0.38
0.60+
0.63
1.31
0.47
0.05+
0.12+
0.91
0.67
0.26
0.17
0.47
0.46
0.31
0.16
0.21
0.35
0.09,
0.14
0.28
0.18
Run 23
MS
2.70
—
0.93
—
--
—
—
--
0.54
0.48
2.37
0.76
—
—
1.40
1.10
—
—
—
0.87
—
—
—
--
0.38
—
—
Capillary-
FID
1.73
0.36+
0.27
0.25
0.22+
0.74
1.09
0.27
0.54+
0.56
1.42
0.49
0.06+
0.12+
0.79
0.59
0.18
0.09
0.28
0.32
0.18
0.09
0.12
0.20
0.05.
0.08
0.16
0.11
Run 25
MS
1.30
—
1.10
—
--
--
--
—
0.57
0.78
1.90
0.66
--
--
1.10
1.10
--
—
—
0.78
—
--
--
--
0.29
--
—
Capillary-
FID
2.51
0.17+
0.74
0.79
0.29
1.57
1.80
0.41
0.78+
1.22
1.44
0.55
0.02+
0.26+
0.74
0.64
0.41
0.24
0.70
0.62
0.25
0.12
0.19
0.27
0.07.
0.1 7+
0.34
0.27
+Subject to corrections.
*Priority pollutant/consent decree compound.
68
-------
8.2 SIMULATED GLC DISTILLATION OF TAR
The boiling point curves of tars produced from various screening and
parametric tests were measured by a procedure based on ASTM D2887.22 Briefly,
the procedure consists of injecting a tar sample onto a column which separates
the components on the basis of their boiling points. Retention times of
standard hydrocarbons of known boiling points are used to divide the chromato-
gram into boiling point fractions and areas under the chromatogram for each
fraction are measured. Two modifications of the procedure were carried out.
Direct on-column injection was employed so that a low injector temperature
could be used to prevent septum bleed. A problem with the ASTM procedure is
column bleed during temperature programming, resulting in excessive baseline
drift. To alleviate this, two identical columns were used on the sample and
reference sides, respectively, of the thermal conductivity detector. Con-
ditions for the simulated distillation are shown in Table 25.
Another drawback of the ASTM procedure is that the PNA's tend to elute
faster from the column than normal hydrocarbons of similar boiling points.
This tends to give lower results for boiling points of samples with substantial
amounts of PNA materials. To correct for these deviations, it was necessary
to determine deviations for several pure PNA's in the boiling range from 200-
525°C. These are shown in Table 26. Using the data in Table 26, the results
were corrected to give the actual boiling point distribution. Typical results
for a tar from gasification of Illinois No.6 coal are shown in Table 27.
8.3 CLASS CHARACTERIZATION USING ELEMENT SPECIFIC DETECTORS
This study was only recently initiated and work is continuing. It is
planned to characterize the tar and its various fractions using a nitrogen
specific detector (NPD) and a sulfur specific detector (FPD). The objective
is to determine if substantial amounts of heavier (3 ring or greater) nitrogen
and sulfur compounds are present. The procedure will consist of using identi-
cal packed OV-101 columns (all glass system) with a FID, FPD, and an NPD with
identical column conditions and carrier gas flow rate. The FID will indicate
a response to all compounds present. With the use of standards, the chromato-
gram can be separated into two, three, four, five and six ring compounds. Then
69
-------
TABLE 25. CONDITIONS FOR SIMULATED GLC DISTILLATION
Instrument:
Detector:
Injector:
Column:
Flow Rate:
Temperature:
Sample:
Standard:
Varian 3700
TC (350°C)
On-column 230°C
Two 4' x 1/8" S.S. (3% OV-101 on 80/100 Chromosorb
HP)
30 ml/mm through each column (adjusted to produce
a straight baseline during programming)
30°C to 310°C at 5°C/min
Coal tar from gasification
C-C normal hydrocarbons
TABLE 26. ERROR IN BOILING POINTS OF PNA'S AS PREDICTED
BY THE PROCEDURE
Compound
Naphthalene
2-Methyl naphthalene
1 -Methyl naphthalene
Acenaphthylene
Acenaphthene
Fluorene
Phenanthrene
Anthracene
Fluoranthene
Pyrene
Chrysene
Benz(a)pyrene
Perylene
Coronene
Observed Boiling Point
(°C)
207
227
229
251
257
273
300
302
336
342
382
419
422
447
Deviation from Actual
Boiling Point (°C)
11
14
16
19
22
22
40
38
39
51
66
84
78
78
70
-------
TABLE 27. CORRECTED BOILING POINT DISTRIBUTION FOR TAR PRODUCED
FROM GASIFICATION OF ILLINOIS NO.6 COAL (RUN 52)
Observed Corrected
Boiling Range Boiling Range % Recovered
(°C) (°C) (Cumulative)
0-100
100-150
150-200
200-250
250-300
300-350
350-^00
400-450
450-500
500-550
550-600
4-97
97-150
150-207
207-272
272-338
338-404
404-469
469-555
555-599
599-659
659-715
0
1.9
18.9
35.9 (56.7)
15.5 (72.2)
19.6 (91.8)
5.4 (97.2)
1.6 (98.8)
1.0
0.2
0
the same sample will be injected onto the NPD and the FPD systems to give the
amounts of nitrogen- and sulfur-containing compounds, respectively. Preliminary
results for tars with the FPD have shown that greater than 90 percent of the
sulfur compounds contain less than three rings. The information obtained from
such studies is considered extremely useful in the characterization of complex
materials such as tar.
8.4 ION CHROMATOGRAPHIC ANALYSIS OF AQUEOUS SAMPLES
The principle of an ion chromatograph has been discussed in detail
0-5
elsewhere. Using a Dionex ion chromatograph (1C) preliminary work has been
carried out for characterization of various cations and anions in the aqueous
condensate and scrubber samples. It has been found that the 1C will be most
suitable for the analysis of such ions as NH. , Na , K , Cl~, F", and SOT
because of its cost effectiveness and accuracy. Additionally, various modifi-
cations in the standard eluent-column procedure will allow measurement of CN~,
SCN", and S". Typical results from 1C have shown about 4 g/1 of NH-, 2 g/1
of Cl" and 0.02 g/1 of F" for gasification run 23.
71
-------
REFERENCES
1. del and, J. G., et al. Pollutants from Synthetic Fuels Production:
Facility Construction and Preliminary Tests. EPA-600/7-78-171, U. S.
Environmental Protection Agency, Research Triangle Park, N. C., 1978.
2. Grohse, P. M., R. B. Denyszyn, and D. E. Wagoner. Sampling and Atomic
Absorption Spectrometric Determination of Arsine at the 2 ug/nr Level.
Anal. Chem.. 50 (8): 1094-96, 1979.
3. Nichols, D. 6., Pollutants from Synthetic Fuels Production. Quarterly
Report, EPA Grant No. R804979-03, RTI/1700/00-01Q, 1979.
4. Gangwal, S. K., and D. E. Wagoner. Response Correlation of Low Molecular
Weight Sulfur Compounds Using a Novel Flame Photometric Detector. 0^
Chromatogr. Sci., 17 (4): 196-201, 1979.
5. Gangwal, S. K., et al. Analysis of a Semibatch Coal Gasifier Product
Gas Using an Automated Gas Chromatograph. J. Chromatogr. Sci., 16 (8):
368-71, 1978.
6. Gangwal, S. K., et al. A Sampling and Analysis Procedure for Gaseous
Sulfur Compounds from Fossil Fuel Conversion. Paper presented at the
EPA Oil Shale Sampling, Analysis and Quality Assurance Symposium,
Denver, CO, March 26-28, 1979. Proceedings to appear.
7. Brodey, S. S., and J. E. Chaney. Flame Photometric Detector: The
Application of a Specific Detector to Phosphorus and Sulfur Compound
Sensitive to Subnanogram Quantities. J. Gas Chrom.. 4: 42, 1960.
8. Patterson, P. L., R. L. Howe, and A. Abu-Shumays. Dual-Flame Photo-
metric Detector for Sulfur and Phosphorus Compounds in Gas Chromato-
graphic Effluents. Anal. Chem.. 50: 339, 1978.
9. Sharkey, A. G. Carcinogenesis, Vol. 1, Polynuclear Aromatic Hydrocarbon
Chemistry, Metabolism and Carcinogenesis. R. I. Freudent and P. W.
Jones, ed., Raven Press, 1976.
10. Schiller, J. E., and D. Mathiason. Separation Methods for Coal Derived
Solids and Heavy Liquids. Anal. Chem., 49: 1225, 1977.
11. Snook, M. E., W. Chamblain, R. F. Severson, and 0. Chortyk. Chromato-
graphic Concentration of Polynuclear Aromatic Hydrocarbons of Tobacco
Smoke. Anal. Chem.. 47: 1155, 1975.
72
-------
12. Novotny, M., M. L. Lee, and K. Bartle. Methods of Fractionation,
Analytical Separation and Identification of Polynuclear Aromatic Hydro-
carbons in Complex Mixtures. J. Chromatogr. Sci.. 12: 606, 1974.
13. Eight Peak Index of Mass Spectra. Vol.1 (Tables 1 & 2), Vol.11 (Table 3).
Mass Spectrometry Data Center, AWRE, Aldermaston, Reading, U. K., 1970.
14. Parr Bomb Combustion of Sample with High Organic Content, Chapter
7, EPA-600/2-76-160, U. S. Environmental Protection Agency, 1976,
86 pp.
15. Parker, C. R. Water Analysis by Atomic Absorption, Varian Techtron
Pty Ltd., Springvale, Australia, 1972.
16. Aruscauage, P. Determination of Arsenic, Antimony and Selenium in Coal
by Atomic Absorption Spectrometry with a Graphite Tube Atomizer. Journal
Research U.S. Geological Survey, 5 (4): 405, 1977.
17. Bosshart, R. E. Determination of Arsenic, Selenium and Antimony in
Coal, Coal Ash, and Coal Refuse Materials by Hydride Generation-Atomic
Absorption. Bituminous Coal Research, Inc., December 1977.
18. Pradhan, N. K. Coal and Petroleum Analysis by Atomic Absorption.
Varian Techntron Pty Ltd., Springvale, Australia, March 1976.
19. Gluskoter, H. J., et'al. Trace Elements in Coal: Occurrence and
Distribution. EPA-600/6-77-064, U. S. Environmental Protection Agency,
Washington, D. C., 1976.
20. Attari, A. Fate of Trace Constituents of Coal during Gasification.
EPA-650/2-73-004, U. S. Environmental Protection Agency, Washington,
D. C., 1973.
21. Grob, K., and G. Grob. Splitless Injection in Capillary Columns.
J. Chromatogr. Sci.. 7: 584-590, 1969.
22. Standard Test Method for Boiling Range Distribution of Petroleum
Fractions by Gas Chromatography, ASTM D 2887, 1973.
23. Small, H., T. S. Stevens, and W. C. Bauman. Novel Ion Exchange Chromato-
graphic Method Using a Conductometric Determination. Anal. Chem., 17
(11): 1801, 1975.
24. Pellizzari, E. D., et a]I. Collection and Analysis of Trace Organic
Vapor Pollutants in Ambient Atmospheres. I. A Technique for Evaluating
the Concentration of Vapors by Sorbent Media. J. Environ. Sci. Tech..
9: 552, 1975.
73
-------
25. Pellizzari, E. D., et al. Collection and Analysis of Organic Vapor
Pollutants in Ambient Atmospheres. II. Studies on Thermal Desorption
of Organic Vapor from Sorbent Media. J. Environ. Sci. Tech.. 9:
556, 1975.
26. Pellizzari, E. D.., et al. Collection and Analysis of Trace Organic
Vapor Pollutants in Ambient Atmospheres. The Performance of a
Tenax-GC Cartridge Sample for Hazardous Vapors. Anal. Letters, 9:
45, 1976.
27. Pellizzari, E. D. Analysis of Trace Hazardous Organic Vapor Pollutants
in Ambient Atmospheres. Anal. Chem., 48: 803, 1976.
74
-------
APPENDIX I
SELECTION AND PREPARATION OF POLYMERIC SORBENTS
At the initiation of this effort various alternative sampling methods for
vapors were considered and eliminated using two major criteria for evaluation:
(1) the broadest possible range of organic vapors be collected, and (2) a
minimum of artifacts be produced in the sample. Other criteria which were
considered included level of sample recovery and ease of interfacing with the
instrument system. Cryogenic trapping was eliminated on several criteria,
especially on artifact production. If the mechanics of cooling the product
gases were not difficult in themselves, the sample would be subject to extreme
vulnerability to artifact formation—collected as a liquid and/or solid sample
with large quantities of water. The condensed water would serve to collect
both acidic and basic inorganic gases (NHj, HC1, HF, HCN, NOX>..) which might
result in artifact formation. Similar problems occur with impingers and the
range of volatility of the sample collected would be limited. (Among the
sorbents used for collection of organics, polymeric sorbents cause the fewest
artifacts and cover the broadest volatility range.) Two polymeric sorbents,
Tenax-GC and Amberlite XAD-2, were used in this program in a sampling strategy
which allowed them to complement each other.
At the initiation of this study a significant body of information on the
performance of Tenax-GC for trace organic analysis especially in ambient air
existed. It was clear from these data that artifacts were at a minimum
for trace analysis but that artifacts could occur at higher concentrations.
The collection efficiency and breakthrough volume needed to be verified at
high concentrations and the implications of displacement chromatography needed
to be defined in order to transfer the sampling technology from trace analysis
to process stream analysis.
1-1
-------
The complementary nature of the two sorbents is a function of the basic
physical properties of the two sorbents. Tenax-GC is a highly temperature
stable polymer of poly(2,6-diphenyl-p-phenylene oxide) with a surface area of
30 m2/g (60-80 mesh). The thermal stability of Tenax allows the collected
sample to be recovered thermally without the introduction of any solvent. In
addition, Tenax has a very low affinity for water and most inorganic gases in
comparison to organic vapors. This results in fewer artifacts being created
as a result of the sample method. The major limitations to the use of Tenax
as a sorbent for sampling process streams is its low surface area which trans-
lates into low capacity and high cost. XAD-2 on the other hand, has a surface
area of 300 m /g and costs approximately 1/50 as much as Tenax. The performance
of XAD-2 is somewhat inferior in two areas, (1) its lack of thermal stability
precludes thermal desorption, and (2) it has a significant affinity for water
and possibly inorganic gases.
Given the properties of the sorbents, the sampling approach taken was to
place XAD-2 in the larger in-line traps with valving to place each on-line
separately. Sampling ports were installed upstream and downstream of these
traps to monitor their performance with small Tenax samples.
Tenax GC is prepared by Soxhlet extraction with methanol and then pentane,
for 22 hours each. The gross amounts of solvent are removed under vacuum at
elevated temperatures. The sorbent is then packed into the glass cartridges
with glass wool plugs in both ends. Within two weeks of sampling the cart-
ridges are thermally desorbed at ca. 275°C under helium purge to remove any
trace organics. The cartridges are sealed in screw cap centrifuge tubes with
teflon cap liners.
XAD-2 is prepared by a multisolvent extraction sequence of water, methanol,
and methylene chloride. First it is soaked in deionized water, the excess
water decanted, and the resulting slurry poured into a 90 x 200 mm cellulose
thimble. A plug of glass wool is then placed on top of the thimble and then
two thimbles stacked in the Soxhlet. The XAD-2 is then extracted with each
solvent in the series for 22 hours and the resin dried in a vacuum oven with
mild heat for 48 hours.
1-2
-------
APPENDIX II
TENAX AND XAD-2 ANALYSIS WITH GC/MS/COMPUTER
Gas chromatography/mass spectrometry/computer (GC/MS/COMP) systems using
high resolution capillary columns is used for the analysis of the Tenax
samples, the XAD-2 extracts and the glass fiber filter extracts. This approach
permits both qualitative identification of species collected and their quan-
titative estimation. Since all of the data acquisition and handling is done
by the computer and the data stored as a permanent file, retrospective searches
are possible. In the early stages of this work complete qualitative interpre-
tations of each sample were performed before selecting specific compounds for
quantisation. As data accumulated, certain compounds and classes of compounds
were selected for routine quantitation. Described below are procedures used
for analysis of Tenax samples and XAD-2 extracts.
J_enax--The organics adsorbed on Tenax were recovered by thermal desorp-
tion which permitted the entire sample to be introduced into a gas chromato-
graphy column. Two internal standards (perfluorobenzene and perfluorotoluene)
were added prior to the desorption step. The inlet-manifold (Figure 16) for
introducing the sample into the analytical instruments consisted of four main
components: a desorption chamber; a six-port, two-position high temperature,
low-volume valve; a gold-plated nickel capillary trap; and a temperature
controller.
The sampling system inlet-manifold is used for recovering vapors trapped
on Tenax-GC cartridges and is interfaced to the GC/MS/COMP system (Figure 17).
The desorbed vapors are subsequently resolved by gas/liquid chromatography and
mass cracking patterns automatically and continuously obtained through the GC
run with a Varian CH-7 mass spectrometer. Operating parameters are those
depicted in Table 28 for the GC/MS/COMP system for the analysis of volatile
organic fractions.
II-l
-------
.PURGE
GAS
i
ro
ION
CURRENT
RECORDER
GLASS
JET
SEPARATOR
HUH
r
CARRIER
GAS
TWO
POSITION
VALVE
CAPILLARY
GAS
CHfiCMATCGHAPH
CAPILLARY
TRAP
ANALYTICAL SYSTEM
THERMAL
OcSOSPTJCN
CHAMecS
HEATED
BLOCKS
EXHAUST
Figure 16. Analytical systems for organic vapors in gas samples.
-------
lalct-
Maalfold-
CLC
Separator
Figure 17. Gas liquid chromatoqraph-mass spectrometer
computer (GC/MS/COMP) layout.
II-3
-------
TABLE 28. OPERATING PARAMETERS FOR 6LC/MS/COMP SYSTEM
Parameter
Setting
Inlet-manifold
desorption chamber
valve
capillary trap-minimum
maximum
thermal desorption time
GC
MS
95M SE30 SCOT
capillary columns
carrier (He) flow
transfer line to ms
scan range
scan rate, automatic-cycle
filament current
multiplier
ion source vacuum
260°C
180°C
-195°C
+175°C
10 min
20°C, 4/C°/min
3 ml/min
210°C
m/e_ 20 - 300
1 sec/decade
300 A
6.0 6
4 x 10 torr
II-4
-------
The mass spectrometer is first set to operate in the repetitive scanning
mode. In this mode the magnet is automatically scanned exponentially upward
from low mass to high mass values. Although the scan range could be varied
depending on the particular sample, values used are from m/z 20 to m/z 300.
The scan is completed in approximately 5 seconds. At this time the instrument
automatically resets itself to the low mass and the run is suitable for
further processing, since it gives some idea of the number of unknowns in the
sample and the resolution obtained using the particular GC conditions.
The next stage of the processing involves the conversion of the spectral
peak times to peak masses. To accomplish this, the spectra are read suc-
cessively into the Varia computer core and then sent at high speed to the
PDP/8 disk by means of the intercomputer interface. Depending on the number
of peaks in each spectrum, the disk capacity allows the storage of 200-500
spectra. Spectra containing time information are stored on the disk, and the
data are returned to an off-line digital computer (IBM 620-i) for mass conversion
by use of the calibration table obtained previously. Normally one set of
calibration data is sufficient for an entire day's data processing since the
characteristics of the Hall probe are such that the variation in calibration
is less than 0.2 atomic mass units per day. Very long runs may require several
cycles of mass conversion using the PDP/8 disk to complete the data processing.
After spectra are obtained in mass-conversion form, full spectra of scans
from the GC run were recorded on the Statos plotter. The total ion current
(TIC) information available at this time are most useful for deciding which
spectra are to be analyzed. At the beginning of the run where peaks are very
sharp, nearly every spectrum is inspected individually. Later in the run when
the peaks are broader, only selected scans are analyzed.
Identification of resolved components is achieved by comparing the mass
cracking patterns of the unknown mass spectra to an eight major peak index of
mass spectra. Particular note is made of the boiling point of the identified
compounds as a function of the elution temperature and the order of elution of
constituents in homologous series, since the SE-30 capillary column separates
primarily on the basis of boiling point.
II-5
-------
XAD-2 -- The organics sorbed on the XAD-2 resin and glass fiber filter
are recovered by solvent extraction with methylene chloride. The extract is
evaported from ca. 150 ml to 0.5 to 1.0 using a Kaderna-Danish (K-D) evaporator
and a micro K-D evaporator. The internal standards, d-jQ-anthracene, dg-
phenol and p-chlorotoluene, are added and the extract is ready for GC/MS/COMP
analysis. Weighed portions of the XAD from the integrated samples are placed
in a thimble and extracted as above. In this case, only a portion (4 ml of
150 ml) of the extract is evaporated to 1 ml and the internal standard added.
The XAD extracts are analyzed on the 1KB 2091 GC/MS/COMP system. The
operating parameters are listed in Table 29.
The mass spectrometer is set to operate in the repetitive scanning mode.
In this mode the magnet is automatically scanned exponentially upward from low
mass to high mass values. Although the scan range can be varied depending on
the particular samples, values used are from approximately m/z 40 to m/z 500.
The scan is completed in approximately 1 second. At this time the instrument
automatically reset itself to the low mass position in preparation for the
next scan, and the information is accumulated by an on-line PDP-11/04 computer
onto disk. The reset period requires approximately 1.4 seconds. Thus, a
continuous cycle of 2.4 seconds is maintained.
The sample is injected and automatic data acquisition is initiated just
before elution of the first component. As each spectrum is acquired by the
computer, each peak that exceeds a preset threshold is recognized and recorded
by mass and intensity. This information is stored on the disk as the scan is
in progress. In addition, TIC values are stored as they are acquired. This
procedure continues until the entire GC run is completed. By this time, there
are from 1500-2000 spectra on disk, which are then used for subsequent or
simultaneous data processing. The data are also recorded on 9-track magnetic
tape for archiving or for subsequent data processing.
The mass spectral data are processed in the following manner. First, the
original spectra are scanned, and TIC information extracted. This is taken
directly from the disk after data acquisition. Then the TIC intensities are
plotted against the spectrum number of a Versatec electrostatic printer/plotter.
The information thus gained generally indicates whether the run is suitable
for further processing, since it gives some idea of the number of unknowns in
II-6
-------
TABLE 29. OPERATING PARAMETERS FOR 1KB 2091 GLC/MS/COMP SYSTEM
Parameter
Setting
Inlet
Split ratio
Injector Temp
Injection Size
GC
15 M SE 30 WCOT
glass capillary columns
carrier (He) flow
programming conditions
separator temperature
MS
scan range
scan rate, automatic-cycle
trap current
multiplier
ion source vacuum
10:1
260°C
U
2.5 ml/min
70°C for 6 min,
8°/min to 250°C
200°C
m/z 40 - 500
2.4 sec/cycle
50 A
350
10~6 torr
II-7
-------
the sample and the resolution obtained using the particular GC conditions.
Full spectra of scans from the GC run are recorded on the Versatec plotter.
The TIC information available at that time is most useful for deciding which
spectra are to be analyzed. At the beginning of the analysis where peaks are
very sharp, nearly every spectrum is inspected individually. Later in the run
when the peaks are broader, only selected scans are analyzed.
Identification of resolved components is achieved by comparing the
fragmentation patterns of the unknown mass spectra to an eight major peak
index of mass spectra. Particular note is made of the boiling point of the
identified compounds with the elution temperature and the order of constituents
in homologous series since the SE-30 WCOT capillary separates primarily on the
basis of boiling point.
II-8
-------
APPENDIX III
QUALITATIVE GC/MS RESULTS FOR ORGANIC VAPORS AND TAR FRACTIONS
Sample results are included in this Appendix for the chemical compounds
which were identified in the XAD cartridges and the tar condensate. These
results are presented for screening test run 21. The tar condensate was
solvent-partitioned prior to analysis, the procedure for which is described in
detail in this report (see Section 5.2). Substantial progress has been made
over the course of this project in the quantitative determination of the
specific compounds which are potentially hazardous, including consent decree
compounds.
III-l
-------
TABLE 30. GC/MS ANALYSIS OF POLYNUCLEAR AROMATICS FROM RUN 21
-QUALITATIVE RESULTS-
Chromato-
grephic
Peak No.
Elution
Te»p.
CC) _
Coapound
Chromato- Elution
graphic Temp.
Peak Ho. CC)
Coapound
1 123 MChyl lnd«M lionets
2 127 naphthalene
2a 129 2,3-benzothlophene
3 132 2-ethyl benzlaldazole (tentative)
4 144 Mthyl naphthalene leoMr
4a 145 5-Mthyl-2,3-benzothiophene
5 146 Mthyl naphthalene leoMr
6 156 blphenyl
7 158 1- or 2-ethyl naphthalene
8 160 dlaethyl naphthalene lamer
9 162 dlaethyl naphthalene iaoner
10 165 1- or 2-athyl naphthalene
11 166 blphenyltne
12 171 acenaphthene
13 17! dlbenzofuran
13a 178 1.3-dihydro-4,6-dlMthylthl»no
(3,4-c)thiopehe (tentative)
14 182 Mthyl acanaphthylene laoMr
15 185 fluorene
16 187 methyl aeenaphthylene iaoMr
17 190 hydroxy fluorene leaner
18 192 hydroxy fluorene iao«*r
19 200 1-Mthyl fluorene
20 200 9-fluor«none
21 205 2-aethoxyfluorene
22 206 dlbenzothlophene leoMr
23 210 anthracene
23a 212 phananthrane
' 24 213 beazidine
24« 214 dlbenzothlophene leoner
25 219 l-nethylbenzo(l,2-b:4,3-b)dithi-
ophene
26 222 3-nethyldibencothlophene
27 223 carbatolc
28 224 Mthyl phenenthrene lioner
29 227 4,5-»«thylene phenanthrene
29a 228 Mthyl pnenanthrene leoner
» 232 2,2'-41cyanoblphenyl
31 238*242 diMthyl phenanthrme laoMre
32 243-246 fluoreatnene end/or
1,2.3,4-tetrahydrofluoianthane
33 248 2-nitro dlaethyl tarephthalate
34 249-250 pyrent and/or
1,2,3,4-tetrahydrofluoranthene
35 259-265 benzofluorenone laoewra and/or
•ethyl pyrene laoawrs
36 265 naphtho(l,2-b)thianaphthene
37 265 2.3-b*nzanthracene
38 265 benzphenaathrene leonere and
trlphenylen*
38a 265 •ethylbenr(a)anthracene laoeurs
38b 265 e»thyl chryeene leoewr
3Bc 265 •ethylbenaophananthrene laoaere
38d 265 dlhydroxy anthraqulnone laonere
39 26S p«ryl»ne
39a 265 benzpyrene
III-2
-------
TABLE 31. GC/MS ANALYSIS OF NONPOLAR NEUTRALS FROM RUN 21
-QUALITATIVE RESULTS-
Chromato-
graphlc
Peak No.
4
5
6
7
8
9
10
11
12
13
14
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
34
35
36
37
39
40
41
Elution
Temp.
(°C)
66
72
73
74
76
76
78
81
82
83
83
86
88
89
91
94
96
99
101
103
105
108
108
110
113
117
119
120
121
122
126
127
131
132
133
Compound
Indene
saturated hydroc«rbons
saturated hydrocarbons
saturated hydrocarbons
C.-cyclohexane or methylcyclohexyl-
cyclohexane
C -thiophene (tentative)
nor-undecane
saturated hydrocarbon
dimethylstyrene or methyllndan
•aturated hydrocarbon
dicyclohexylbutane
n-pentylcyclohexane (tentative)
C12H21 i80""
naphthalene
thlanaphthene, dlnethyllndan
C.-benzene
n-dodecane
cyclohexylbenzene
dimathy linden
dlnethyllndan
dimethyllndan
mathylnaphthalene
nethylthianaphthene
nethylnaphthalene
trims thylindan
trine thy lindan, tetranethy linden or
5,8-dlmethyl-l-n-octyl-l,2,3,4-
tetrahydronaphthalene
l-methyl-4-n-heptyl-l,2,3,4-tetra-
hydronaphthalene or
dlmethyltetrahydronaphthalene or
trinethylindan
blphenyl
dlnethylthlalndene
athylnaphthalena
dime thy Inaphthalene
dlmethylnaphthalane
dime thy Inaphthalena
blphenylene , dlnethy Inaphthalena
acenaphthene
mathylbiphanyl
trimethylthiaindene
Chroma to-
graphic
Peak No.
42
43
44
47
48
50
51
55
58
59
60
61
62
64
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
* 87
88
Elution
Temp.
CO
135
136
137
144
147
149
152
160
163
165
166
169
170
173
178
179
182
184
185
187
190
195
195
199
205
206
214
216
219
228
234
237
253
255
257
261
265
Compound
C.-naphthalene, dlbenzofuran
pentamethyldihydroindene
ar-phenylphenol
fluorene
b iphenylae thane
C.-naphthalene (tentative)
hydroxy fluorene
hydroxy fluorene
1 , 1-diphenylathane
2-nethoxy fluorene
acrldone or phenanthrldone
l,l-dl-(4-aethylphenxl)dodecane
dibenzothiophene
phenanthracane
dj --anthracene
anthracene
dlhydromethylphenylbenzofuran
dibenzoheptafulvene
methyldlbenzothiophene
methylphenanthrene
4 , 5-methy lenephenanthracene
saturated hydrocarbon
phthalata
C16H12 Uooer or
2,2' -dicyanobiphenyl
dimathy Iphenanthrene
saturated hydrocarbon
fluoran thane
pyrene
saturated hydrocarbon
C17H12 i90a*T
C17H12 i80Ber
saturated hydrocarbon
phthalate
saturated hydrocarbon
unaaturatad hydrocarbon (tentative)
saturated hydrocarbons
C18H12 laol"r
C18H12 l80ner (tentative)
C18H12 i80~r
phthalates
III-3
-------
TABLE 32. GC/MS ANALYSIS OF TAR BASES FROM RUN 21
- QUALITATIVE RESULTS-
Chromato- Elation
graphic Temp.
Peak No. (°C)
Compound
Chromato—
graphic
Peak. No.
Elation
Temp.
Compound
1 100.0 toluene
3 100.0 dimethyl pyridine isomer
4 100.0 dimethyl pyridine isomer
5 103.6 trimetbyl pyridine isomer
6 110.4 2,4-dimethyl-6-ethylpyridine
7 112.4 toluidiae isoaer and/or
benzyl mine
7a 113.2 C2-alkyl pyridine isomer
7b 124.8 methyl ethyl pyridine isomer
7c 125.6 n-methyl-o-toluidine
8 133.6-135.6 quinoline
9 138.0-141.2 isoquinoliae
10 144.4 methyl quinoline isomer
10a 146.0 methyl quinoline isoaer
lOb 148.0 methyl quinoline isomer
11 150.8 methyl quinoline isoaer
12 154.4 methyl quinoline isoaer
12a 155.2 ethyl quinoliae isomer
13 158.4 ethyl quinoline isomer
14 161.2 dimethyl quinoline isomer
15 164.0 diaethylquinoline isomer
16 171.2 dimethyl quinoline isomer
17 174.4 diphenyl amine
18 184.4 alpha naphthyl amine
19 190.4 carbazole
19a 200.0 xanthene
19b 200.0 9-methyl carbazole
19c 206.8 methyl carbazole isomer
19d 207.6 methyl carbazole isoaer
20 210.8 7,8-benzoquinoline
21 215.6 benzoquinoline isomer
22 218.0 3-aethylbenzoquinoline
23 219.6 acridine
23a 223.2 methyl benzoquinoline isomer
23b 224.4 methyl benzoquinoline isomer
24 226.0 phenyl indole isomer
25 230.0 dimethylacridine isoaer
25« 236.8 aethyl phenyl indole isomer
25b 240.8 aethyl phynyl indole isoaer
III-4
-------
TABLE 33. GC/MS ANALYSIS OF STEADY-STATE XAD-2 EXTRACT FROM RUN 21
-QUALITATIVE RESULTS-
Chromato-
graphic
Peak No.
la
2
2a
2b
2c
3
4
5
5a
5b
5c
5d
5e
5f
5g
6
6a
6b
6c
7
8
Elation
Temp.
60°
60°
60°
60°
60°
60°
60°
60°
60°
62°
64°
71°
72°
81°
85°
87°
90°
92°
96°
100°
102°
Compound
unknown
toluene
methylthiopheae isomer
hydrocarbon
C.-benzene isomer ethyl
xylene isomer
CgHg isomer styrene
xylene isomer
hydrocarbon
C, -benzene isomer
C.-benzene isomer
benzofuran
C.-benzene isomer
C.-benzene isomer
indan
indene
diethylbenzene isomer
diethylbenzene isomer
phenol (tent)
C^H.-benzene isomer
C,H, -benzene isomer
Chromato-
graphic
Peak No.
8a
8b
8c
8d
8e
9
9a
9b
10
lOa
lOb
IQc
lOd
lOe
11
lla
lib
12
13
14
14a
Elution
Temp.
105°
108°
111°
116°
118°
122°
124°
127°
145°
147°
148° '
160°
164°
171°
177°
182°
190°
200°
204°
218°
224°
Compound
methylbenzofuran isomer
C.H7 -benzene
cresol isomer
methyl indene isomer
cresol isomer
naphthalene
benzothiophene isomer
C_-phenol
8-methy Inaphtha lene
methyl benzothiophene isomer
a-methylnaphthalene
biphenyl
dimethylnaphthalene isomer
C12H18 isOBler
C^-biphenyl isomer
dibenzofuran (tent)
C.-H.^ isomer
Cj-H.g isomer (tent)
CjgH.g isomer (tent)
D._ anthracene i
C.gH.- isomer
III-5
-------
TABLE 34. GC/MS ANALYSIS OF SURGE XAD-2 EXTRACT FROM RUN 21
-QUALITATIVE RESULTS-
Chromato-
graphic
Peak No.
1
la
2
2a
3
3a
3b
3c
3d
4
4a
4b
4c
4d
4e
4£
4g
4h
4i
4j
4k
41
4m
5
5a
6
6a
6b
7
8
Sa
8b
8c
8d
8e
8f
8g
Sh
8i
Sj
8k
81
8m
8n
So
8p
Elutlon
Temp.
(*C)
60°
60°
60°
60°
60°
60°
60°
60°
60°
60°
60°
60°
60°
60°
60°
60°
60°
60°
60°
60°
60°
60"
60°
60°
60°
60°
60°
60°
60°
60°
60°
60°
60°
60°
60°
60°
60°
60°
60.1°
62.7°
63.5°
63.9°
65.9°
67.1°
68.3°
70.3°
Compound
aethylene chloride (bkg)
1,1, 1-trichloroethane
benzene
thiophene
cyclohexene
C?H14 isomer (tent)
CyH12 isomer
C7Hj0 isomer
C7HJ2 isomer
toluene
methylthiophene isomer
CglL, isomer
C8H16 isomer
CgHj^ isomer
CgH.g isomer
C-Hj, isomer
C.H., isomer
8 14
CgH^ isomer
CgH14 isomer
C8H16 i80IBer
CgH. , isomer
CgHlg isomer (tent)
CgH18 and CgH^ isomers
ethylbenzene
ethylthiophene isomer
dimethylbenzene isomer
dimethylthiophene isomer
C,- thiophene isomer
CgH- isomer styrene
dimethylbenzene isomer
CgH.Q isomer
CgHlg isomer (tent)
C9H16 isomer (tent)
C. -benzene isomer
CgHj, isomer
CgHjg isomer
saturated hydrocarbon
n-propylbenzene
C. -benzene isomer
C. -thiophene isomer
C, -benzene isomer
C3- thiophene isomer
C, -benzene isomer
C-j-thiophene isomer
saturated hydrocarbon
benzofuran
Chromato-
graphic
Peak No.
9
9a
9b
9c
9d
9e
9f
10
lOa
lOb
lOc
lOd
11
lla
lib
lie
lid
He
llf
Hg
llh
iH
12
12a
12b
lie
12d
12e
12f
12g
12h
12i
12j
12k
121
12m
12n
12o
12p
12q
12r
12s
13
13a
13b
13c
Elution
Temp.
71.5°
72.3°
75.1°
78.7°
80.3°
81.1°
83.5°
85.5°
89.1°
91.1°
93.1°
97.5°
98.3°
100.7°
102.7°
103.1°
107.1°
108.3°
111.5°
113.1°
113.5°
114.3°
119.9°
121.9°
123.9°
127.9°
128.7°
121.9°
141.9°
144.7°
146.7°
156.3°
167.5°
171.9°
172.7°
177.5°
186.7°
189.1°
198.7°
199.1°
209.1°
212.3°
213.1°
218.7°
229.9°
245.1°
Compound
C. -benzene isomer
•ethylstyrene isomer
C10H20 isOBer
C.QH_2 isomer
C, -benzene isomer
C^-benzene isomer
indan
indene
diethylbenzene isomer
diethylbenzene isomer
phenol
C . -benzene isomer
4
C/H7 -benzene isomer
C,H7-benzene isomer
methylbenzofuran isomer
methylbenzofuran isomer
C.H.. isomer
4 24
cresol isomer
methyl-(2 ,3-dihydroindene)
isomer
cresol isoner
methyllndene isomer
cresol and methyl indene isomers
naphthalene
benzothiophene
C2 -phenol isomer
2-ethylbenzimidazole (tent)
C12H26 i801D"
saturated hydrocarbon
P-methyl-naphthalene
a -methyl-naphtha lene
C13H2» isomer
biphenyl
C.-H- isomer
acenapbthene
diphenylethane
dibenzofuran
fluorene
C,,H,, isomer (tent)
ID lo
C.,H,0 isomer (tent)
C16H18 isomer
dibenzothiophene (tent)
phenanthrene
DJO anthracene (i)
C18H22 isOBer
methylphenanthrene isomer
C.-H.. isomer
III-6
-------
TABLE 35. GC/MS ANALYSIS OF GLASS FIBER FILTER EXTRACT FROM RUN 21
-QUALITATIVE RESULTS-
Chromato-
graphic
Peak No.
5
Sa
6
7
8
8a
8b
8c
9
10
lOa
11
lla
lib
12
13
13a
13b
13c
13d
13e
13f
13g
Elution
Temp.
134°
135°
151°
153°
163°
167°
169°
172°
173°
177°
178°
183°
185°
189°
191°
193°
197°
199°
• 200°
204°
207°
212°
213°
Compound
naphthalene
benzothiophene
B-raethy Inaphtha lene
a-methylnaphthalene
biphenyl
dimethylnaphthalene isomer
dimethylnaphthalene isomer
dimethylnaphthalene isomer
acenaphthylene
acenaphthene
alkyl-biphenyl isomer
dibenzofuran
trimethylnaphthalene isomer
isopropylnaphthaleae
fluorene
2,2,4-trimethyl penta 1,3-diol
di-isobutyrate (bkg)
C,,H,nO isomer (tent)
lj lu
C,,H,nO isomer (tent)
1J 1U
C16H18 isomer
C16H18 isomer
methyl-fluorene isomer
C.CH., isomer (tent)
15 lo
dibenzothiophene
Chromato-
graphic
Peak No.
13h
14
I4a
14b
14c
14d
14e
14f
15
15a
15b
16
16a
16b
16c
16d
17
17a
18
20
21
21a
21b
Elution
Temp.
CO
217°
218°
219°
229°
231°
233°
234°
239°
250°
250°
250°
250°
250°
250°
250°
250°
250°
250°
250°
250°
250°
250°
250°
Compound
phenanthrene
DIO anthracene (I)
anthracene
unknown
methyl-phenanthrene isomer
methylene-phenanthrene isomer
C15^12 isomer
unknowns
C16H10 isoner
C16H10 i80mer
2-benzyl benzimidazole
C16H10 isoioer
diphenyl disulfide or unknown
methyl pyrene isomer
methyl pyrene isomer
unknown
phthalate compound (bkg)
C18H12 isOBer
C18^12 isoiner
phthalate compound (bkg)
phthalate compound (bkg)
C20H12 i80aer (tent)« (5
C-..H. , isomer (tent), (5
20 12
rings)
rings)
III-7
-------
TABLE 36. GC/MS ANALYSIS OF TENAX SAMPLE UPSTREAM OF STEADY-STATE
XAD-2 TRAP FOR RUN 21
-QUALITATIVE RESULTS-
Chromato-
graphic
Peak No.
1
2
3
4
4a
5
6
6a
7
8
8a
9
10
lOa
lOb
11
12
12a
12b
15
16
16a
17
17a
18
18a
19
20
20a
21
21a
22
22a
23
23a
23b
23c
24
24a
Elution
Temp.
CO
57°
58°
60°
70°
72°
74°
79°
82°
85°
86°
87°
90°
96°
97°
102°
103°
116°
118°
122°
134°
136°
137°
140°
140°
140°
143°
144°
146°
151°
152°
154°
154°
155°
156°
156°
157°
158°
159°
164°
Compound
co2
carbonyl sulfide + sulfur
dioxide
hydrocarbon
acetone
CC.H.- isomer
methylene chloride (bkg)
carbon disulfide
C^HgO isomer
perfluorobenzene (I)
n-hexane
t r i chlo rome thane
perfluorotoluene (I)
benzene
thiophene + carbon tetrachloride
trichloroethylene
acetic acid
toluene
methylthiophene isomer
CgHjg isomer
xylene isomer
xylene + C.-thiophene isomers
C.-thiophene isomer
styrene
hydrocarbon
xylene isomer
C.-thiophene isomer
C.H20 isomer
C10H22 i80Der
saturated hydrocarbon
benzaldehyde
C.-benzene isomer
phenol (tent)
C.-benzene isomer
saturated hydrocarbon
C.-benzene isomer
saturated hydrocarbon
benzofuran
C.-benzene isomer
C^-benzene isomer
Chroma to-
graphic
Peak No.
25
25a
25b
26
27
27a
27b
28
28a
28b
29
30
30a
31
31a
31b
32
32a
33
33a
34
34a
35
35a
36
38
39
39a
39b
40
40a
41
42
42a
42b
42c
43
Elution
Temp.
CO
164°
166°
167°
167°
168°
169°
169°
170°
173°
174°
176°
177°
178°
180°
181°
182°
184°
185°
187°
188°
192°
193°
208°
210°
211°
216°
219°
221°
223°
224°
225°
226°
230°
233°
240°
240°
240°
Compound
C.-benzene isomer
saturated hydrocarbon
cresol isomer
C.H.-benzene isomer
indene
saturated hydrocarbon +
C,-benzene isomers
acetophenone
cresol isomer
C.-benzene isomer
C, -benzene isomer
methylbenzofuran + C.-phenol
isomers
methylbenzofuran isomer
saturated hydrocarbon
C.-phenol + C^-benzene isomers
C.-phenol isomer
C,H^ -benzene isomer
C.-phenol + methylindene
isomers
C^H. -benzene isomer
C.-phenol isomer
benzole acid (tent)
naphthalene
benzothiophene
p-methylnaphthalene
methylbenzothiopheue isomer
or-methylnaphthalene
biphenyl
saturated hydrocarbon
C^-naphthalene isomer
C.-naphthalene isomer
unknown
C.-naphthalene isomer
C..H. isomer + hydrocarbon
if, O
C..,Hj_ isomer
dibenzofuran
C..H.- isomer
tributyl phosphate (tent)
diphenyl oxazole isomer
III-8
-------
TABLE 37. GC/MS ANALYSIS OF TENAX SAMPLE DOWNSTREAM OF STEADY-STATE
XAD-2 TRAP FOR RUN 21
-QUALITATIVE RESULTS-
Chromato-
graphic
Peak No.
1
2
3
3a
3b
4
5
6
6a
6b
6c
7
7a
7b
8
9
9a
10
lOa
lOb
11
lla
12
12a
12b
12c
12d
13
13a
14
15
15a
15b
Elution
Temp.
CO
56°
57°
58°
66°
67°
68°
70°
71°
73°
74°
76°
76°
78°
78°
87°
92°
93°
95°
97°
100°
113°
116°
125°
130°
136°
138°
146°
148°
151°
152°
153°
156°
159°
Compound
co2
sulfur dioxide
carbonyl sulfide
butene isomer + freon 11
acetone
n-pentane
methylene chloride (bkg)
carbon disulfide
C^H. isomer
C5HJO isomer
C/HgO isomer
hexafluorobenzene (I)
n-hexane
chloroform (tent)
perfluorotoluene (I)
benzene
carbon tetrachloride
acetic acid
C,H.0 isomer
trichloroethylene
toluene
4-methyl-2-ethyl-l,3-dioxolane
(tent)
silane compound (bkg)
xylene isomer
hydrocarbon
hydrocarbon
hydrocarbon
benzaldehyde
phenol
C,H,N isomer
hydrocarbon
hydrocarbon
hydrocarbon
Chroma to-
graphic
Peak No.
17
18
18a
19
19a
20
20a
20b
20c
21
21a
22
24
25
*
it
*
27
*
28
*
31
32
33
34
35
36
37
*
Elution
Temp.
CO
164°
166°
167° *
168°
173°
174°
176°
178°
179°
180°
182°
183°
186°
189°
190°
197°
204°
206°
207°
208°
210°
217°
220°
228°
229°
232°
236°
240°
240°
Compound
cresol isomer
indene
acetophenone
cresol isomer
C, -benzene isomer
C. -phenol isomer
methyl benzofuran isomer
C. -phenol isomer
C, -benzene isomer
C. -phenol isomer
C,H, -benzene isomer
C.-phenol isomer
benzoic acid
naphthalene
benzothiophene
C-H.-benzene
oethylbenzothiophene
(tent)
B-methy Inaphtha 1 ene
methyl benzothiophene
cr-methylnaphthalene
isomer
isomer
thiobenzoic acid or unknown
biphenyl
hydrocarbon
dimethyl naphthalene
C.-naphthalene isomer
acenaphthylene
acenaphthene
dibenzofuran (tent)
C13H10 isOBier
isomer
*Minor unnumbered peak.
III-9
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/7-79-201
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Pollutants from Synthetic Fuels Production: Sampling
and Analysis Methods for Coal Gasification
5. REPORT DATE
August 1979
6. PERFORMING ORGANIZATION CODE
AUTHORIS) S.K.Gangwal, P.M.Grohse, D.E.Wagoner,
D. J. Minick, C. M. Sparacino, and R. A. Zweidinger
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Research Triangle Institute
P.O. Box 12194
Research Triangle Park, North Carolina 27709
10. PROGRAM ELEMENT NO.
EHE623A
11. CONTRACT/GRANT NO.
Grant No. R804979
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 COVERED
Task Final; 8/78 - 7/79
14. SPONSORING AGENCY CODE
EPA/600/13
is. SUPPLEMENTARY NOTES ffiRL-RTP project officer is N. Dean Smith, Mail Drop 61, 919/
541-2708. Grant-related reports include EPA-600/7-78-171, EPA-600/7-79-200, and
EPA-600/7-79-202.
16. ABSTRACT
The report describes sampling and analysis methods involving a laboratory-
scale coal gasification facility used to study the generation, sampling, chemical
analysis, process evaluation, and environmental assessment of pollutants from coal
gasification. It describes methods for particulates, organic condensibles, and va-
pors or gases in the raw product stream of the gasifier as well as for solid residues.
It describes gas chromatography (GC) procedures for measuring fixed gases, C1-C5
hydrocarbons, sulfur gases, and C6-C8 aromatics. Atomic adsorption (AA) proce-
dures for measuring toxic trace elements include those for arsenic, selenium, lead,
cadmium, chromium, and mercury. Volatile organics are collected from the gas
stream using polymeric sorbents (Tenax GC and XAD-2), and analyzed by glass ca-
pillary GC/mass spectrometry (MS). The major nonvolatile byproduct (tar) is pre-
fractionated by solvent partitioning into acid, base, and neutral fractions. Each
fraction is analyzed by capillary GC/MS or high-performance liquid chromatography
(HPLC). Typical results are given to illustrate the nature of the compounds stu-
died, the methodologies, and their sensitivities.
7.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b. IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Pollution
Coal Gasification
Sampling
Analyzing
Dust
Aerosols
Organic Compounds
Vapors
Gases
Residues
Pollution Control
Stationary Sources
Synthetic Fuels
Particulates
136
13H
14B
11G
07D
07C
8. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS {This Report)
Unclassified
21. NO. OF PAGES
102
20. SECURITY CLASS (TMspagc)
Unclassified
22. PRICE
EPA Form 2220-1 (t-73)
ra-io
------- |