United States
Environmental Protection
Agency
Industrial Environmental Research
Laboratory
Research Triangle Park NC 2771 1
EPA-600/2-79-200
November 1979
Research and Development
&ER&
EPA/IERL-RTP
Procedures Manual:
Level 2 Sampling and
Analysis of Oxidized
Inorganic Compounds
-------
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 ENVIRONMENTAL PROTECTION TECH-
NOLOGY series. This series describes research performed to develop and dem-
onstrate instrumentation, equipment, and methodology to repair or prevent en-
vironmental degradation from point and non-point sources of pollution. This work
provides the new or improved technology required for the control and treatment
of pollution sources to meet environmental quality standards.
EPA REVIEW NOTICE
This report has been reviewed by the U.S. Environmental Protection Agency, and
approved for publication. Approval does not signify that the contents necessarily
reflect the views and policy of the Agency, 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-600/2-79-200
November 1979
EPA/IERL-RTP Procedures Manual
Level 2 Sampling and Analysis of
Oxidized Inorganic Compounds
by
R.F. Maddalone, LE. Ryan, R.G. Delumyea,
and J.A. Wilson
TRW Defense and Space Systems Group
One Space Park
Redondo Beach, California 90278
Contract No. 68-02-2165
Task No. 102
Program Element No. INE624
EPA Project Officer: Frank E. Briden
Industrial Environmental Research Laboratory
Office of Environmental Engineering and Technology
Research Triangle Park, NC 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, DC 20460
-------
INTRODUCTION
With the Increasing awareness of the government and scientific
communities to the possible hazards from the output of various industries,
the Environmental Protection Agency has developed an approach to assess
the environmental impact of any type of industrial process. This approach
consists of a two phase attack. The first phase, Level 1, surveys the site
to detect whether or not a given pollutant class is being emitted. ;.
Level 1 (1) sampling and analysis methods are designed to obtain emission
results accurate to a factor of two to three. A set of criteria is used
to prioritize the streams so that those streams which are found to be a
problem are identified for further study. This next phase, Level 2, is
designed to be specific for a given stream and perhaps for even a given
pollutant. Compared to this phased approach, a direct environmental assess-
ment of a site would use comprehensive sampling and analytical methods to
determine all pollutants that are present with high accuracy. TRW has
studied the two approaches and found that the phased approach (2) is the
more cost effective.
This manual describes an approach to utilizing the data from Level 1
analyses to direct the Level 2 analysis program. The Level 2 analysis,
because it is focused, will use more accurate and sophisticated techniques
to determine elemental concentrations and identify specific compounds. The
analysis of inorganic compounds requires the coordinated use of a variety
of analytical techniques. Some techniques, such as X-Ray Diffraction (XRD),
Transmission Electron Microscopy with Selected area Electron Diffraction
(TEM-SAED) and Electron Spectroscopy for Chemical Analysis (ESCA), have the
potential for direct compound identification, but only for selected com-
pounds or situations. The identification scheme proposed in this manual
consists of:
t Initial Sample Characterization. Elemental and anion composi-
tion, sample stability, and bulk morphological structure are
determined.
• Bulk Composition Characterization. Oxidation state, X-Ray diffrac-
tion information, and functional groups are derived.
• Individual Particle Characterization. Single particle elemental
composition, diffraction pattern, and morphology are measured.
1
-------
The methods were chosen so that as the analyst proceeds, he is applying more
and more sophisticated instrumentation. The degree to which each method
can be applied will vary considerably with the experience of the analyst
and the sample quantity and equipment available. It is recommended that
continuing use of any one method be evaluated in light of the information
derived. In general, it is far better to use a variety of instruments
operated in the most efficient manner rather than a single instrument or
technique to the limit of its capabilities.
In the Level 2 inorganic analysis, the goal is to identify Discharge
Multimedia Environmental Goals (DMEG) compounds and concentrations. (3).
For every element which exceeded its DMEG concentration value, there is a
list of DMEG compounds which contain this element. These compounds are used
to develop a list of potential compounds present in the stream analyzed.
The decision to continue the analysis for DMEG compounds on the list will
depend on a variety of factors:
• Number of DMEG compounds identified exceeding DMEG values.
• Interest in identifying the remaining compounds for those elements
that exceeded DMEG concentrations.
• Cost/availability of necessary equipment.
The analyst must decide which information is necessary, what method will be
applied, and how much more information can be obtained by each further
analysis. In many cases some methods can be bypassed because of results
from previous tests; e.g., quantitative anion analysis may provide suffi-
cient information, so that Fourier Transform Infrared (FTIR) analysis would
be only repetitious. In other cases efforts may direct the analyst to a
specific method since it would be best suited to analyze^ for a given com-
pound. By understanding the information that can be derived from each
technique, the analyst will be better able to select the appropriate
combination of techniques to determine the compounds present in an environ-
mental sample of interest.
The compounds discussed in this manual are primarily in higher oxidiza-
tion states. A companion manual ("EPA/IERL-RTP Procedures Manual for
Level 2 Sampling") discusses organometallic or reduced inorganic species
analysis. Together the two manuals represent a comprehensive approach to
ii
-------
inorganic analysis, and should be considered evolving documents to be
updated as new technologies are developed.
This manual is organized into four sections. The initial section
describes DMEG compounds and their DMEG values and shows how to develop
lists of potential compounds based on Level 1 data, DMEG compounds, and
potential emissions by process. The next section describes the initial
characterization of samples including elemental, anion, thermogravimetric,
and morphological studies. This information is used to help interpret
X-ray diffraction, IR, and surface studies which are described in the next
section. The final chapter describes the use of single particle techniques
to characterize particulate matter.
* • •
in
-------
ABSTRACT
This manual describes the Level 2 analysis procedures for the
identification of oxidized inorganic compounds in environmental samples
from energy and industrial processes. The methods discussed in this manual
are grouped into three major phases: 1) Initial Sample Characterization,
2) Bulk Sample Characterization, and 3) Individual Particle Characterization.
A description of the theory, sensitivity, interferences, sample preparation,
application and information derived is given for each method discussed in
the manual. This manual represents a step in the development of a general
methodology for analysis of process samples. It is intended to define the
concepts of Level 2 analyses and review current procedures available. It
does not define a fixed protocol because the complexity of samples pre-
cludes a prior definition of specific procedures.
iv
-------
CONTENTS
Page
Introduction i
Abstract iv
Contents v
Figures vii
Tables ix
1.0 Evaluation and Use of Level 1 Data 1
1.1 Using DMEG Compounds and DMEG Values as Level 2
Transition Criteria 1
1.2 Oxidized Species Emitted by Various Processes 4
1.3 Level 2 Sampling 10
1.3.1 Solid Sampling 11
1.3.2 Liquid Sampling 16
1.3.3 Gas Sampling 21
2.0 Initial Sample Characterization 25
2.1 Polarized Light Microscopy . 28
2.1.1 Theory 28
2.1.2 Compound Identification - Generalized Procedure . 30
2.1.3 Typical Results from PLM 33
2.2 Thermal Analysis 36
2.2.1 Principles of Operation 36
2.2.2 Applications and Methodology 38
2.3 Anion Analysis 43
2.3.1 Specific Tests 43
2.3.2 Ion Chromatography 43
2.4 Bulk Elemental Analysis 47
2.4.1 Introduction-Atomic Techniques 51
2.4.2 Neutron Activation Analysis 62
2.4.3 X-Ray Fluorescence Analysis-Introduction 65
2.4.4 Sample Dissolution Procedures 73
2.5 Summary of Initial Sample Characterization 73
-------
CONTENTS (Continued)
Page
3.0 Bulk Characterization 79
3.1 Fourier Transform Infrared Analysis 79
3.1.1 Theory 81
3.1.2 Advantages of FTIR 83
3.1.3 Application of FTIR to Environmental Sampling . . 84
3.2 Powder X-Ray Diffraction 88
3.2.1 Theory 94
3.2.2 Powder XRD for Compound Identification in
Environmental Samples 95
3.3 Surface Analysis Using ESCA and SIMS 99
3.3.1 Theory 99
3.3.2 Sampling Handling/Preparation for ESCA and SIMS . Ill
3.3.3 Compound Identification 113
3.3.4 Application of ESCA to Surface Analysis 115
3.4 Summary of Bulk Composition Characterization 119
4.0 Individual Particle Analyses 123
4.1 Introduction to Electron Microscopy 123
4.2 Scanning Electron Microscopy-Energy Dispersive
X-Ray Spectrometry 126
4.3 Electron Probe Microanalysis (EPMA) 128
4.4 Transmission Electron Microscopy-Selected Area
Electron Diffraction . . . , 129
4.5 Summary of Individual Particle Characterization 130
4.5.1 Application of Electron Microscopy to
Experimental Samples 132
5.0 Summary 137
6.0 References 138
Appendix
A. Specific Anion Analysis Procedures A-l
vi
-------
FIGURES
Number Page
1 Decision Procedure for Level 1 to Level 2 Transition . . 3
2 Level 2 Inorganic Sampling Train (Glass) 23
3 Identification of Single Particles Using Polarized
Light Microscopy (PLM) 29
4 Thermogravimetric Evaluation of Calcium Oxalate
Monohydrate; Heating Rate 6°C/min 38
5 DSC of >3 wm Material Collected at the Outlet of an
FGD Run at 10°C/min 40
6 DSC of >3 ym Material Collected at the Outlet of an
FGD Run at 2°C/min and Reduced Sample Size 41
7 Ion Chromatogram of the Dionex Anion Standard Solution . 48
8 Calibration Curve for Chloride Ion Analysis 50
9 Schematic of Atomic Technique 51
10 dM/d(logD5Q) Size Distribution at the Outlet of a
Limestone wet Scrubber at a Coal-Fired Utility 70
11 Trace Element Distribution by Particle Size at the
Outlet of a Limestone Wet Scrubber 71
12 Ratio of Element Concentration to Silicon Concentration
by Particle Size 72
13 Flow Chart of Sample Dissolution Procedure 74
14 Logic Flow for Initial Sample Characterization 75
15 Diagram of Michel son Interferometer 82
16 FTIR of Outlet FBC Material 87
17 FTIR Analysis of Surface Composition of Fly Ash Samples. 89
18 Powder Dispersing Apparatus for Preparing XRD Standards
on Glassfiber Filters 98
19 Survey ESCA Spectra of Particulate Matter from Fluidized
Bed Combustor 101
20 Elemental Inclusions in Inconel Alloy Using IMMA .... 104
21 Depth Profile Analysis of Fly Ash Sampling Using IMMA. . 106
vii
-------
FIGURES (Continued)
Number Page.
22 Block Diagram of McPherson ESCA 36 109
23 Schematic of SIMS Instrument no
24 Sticky Gold Mounting Technique 112
25 High Resolution ESCA Spectrum Showing Chemical
Shift for S"2 and SO^ Compounds 114
26 ESCA Depth Profile Analysis of l3wn Partial!ate Matter
before and after an FGD 116
27 ESCA Depth Profile Analysis of <3ym Particulate Matter
before and after an FGD 117
28 Flowchart for Bulk Sample Characterization 121
29 Interactions of an Electron Beam with a Sample 125
30 Recommended Sequential Application of Individual
Particle Techniques 131
31 SEM-EDX at 3000X of Sample S-2-5 133
32 EPMA Elemental Imaging of Sample S-2-5 . . . 134
33 TEM-SAED of Fibrous Material in S-2-5 136
viii
-------
TABLES
Number Page
1 Application of DMEG Decision Criteria 4
2 Elemental Effluents Cited from the Literature 5
3 Conditions and Typical Inorganic Compounds Emitted
from Various Industrial Processes 6
4 Summary of Initial Sample Characterization 26
5 Refractive Indices of Selected Crystals (Nd25) 30
6 PLM Analyses Results . 35
7 Thermal Analysis Techniques 36
8 Retention Time of Various Ions Found in Wet Scrubbers. . 46
9 Comparison of SOT and Cl"Analysis Methods 49
10 Comparative Detection Limits (ppb) (yg/L) 57
11 Atomic Absorption Analytical Operating Parameters. ... 59
12 Percent Recovery of Spiked Samples by Atomic Absorption. 60
13 ICAP Recovery Results, % . 61
14 Detection Limits Reported for Some Elements by Neutron
Activation Analysis ". 64
15 Summary Bulk Composition Characterization 80
16 Recommended FTIR Sample Preparation Techniques 85
17 General Absorption Regions 90
18 Listing of Assigned Infrared Bands Observed in
Particulate Samples 91
19 Infrared Bands of Some Common Nitrates (cm" ) 92
20 Infrared Bands of Some Common Sulfates (cm ) 93
21 Detection Limits on Commercial XRD Instruments for
Several Compounds 97
22 Principle ESCA Peak Binding Energy for Each Element. . . 102
23 ESCA Data - S/Ca Ratio by Particle Size (ym) and
Depth (A) 119
24 Summary of Methods Used for Individual Particle
Characterization 124
ix
-------
1.0 EVALUATION AND USE OF LEVEL 1 DATA
The analyst will have the results of the Level 1 survey data for a
given site before starting the Level 2 inorganic compound identification
work. The Level 1 environmental assessment methodology (1) has been
designed to determine inorganic elemental composition of the sampled stream
using the combination of Spark Source Mass Spectroscopy (SSMS) and a few
wet chemical techniques. The use of this data as the starting point to focus
the Level 2 studies will significantly reduce the cost (2) and allow the
analyst to select specific methods that are best able to identify the com-
pounds associated with the elements found in the SSMS survey. The follow-
ing sections will suggest an approach to using the Level 1 data to develop
a list of elements for additional study, and will provide information on
potential compounds emitted by different processes.
1.1 USING DMEG COMPOUNDS AND DMEG VALUES AS LEVEL 2 TRANSITION CRITERIA
The focusing of the Level 2 analysis effort is achieved by using a
yardstick to select streams and elements for Level 2 analysis. The yard-
stick employed is the Discharge Multimedia Environmental Goals (DMEG) com-
pounds and their concentrations. The DMEG list includes elements and com-
pounds which have been identified in coal and oil treatment processes and
are regulated by the EPA or on its lists of pollutants of concern (3).
The DMEG concentrations take into consideration a variety of factors,
including toxicity data, half-lives, cumulative tendencies, and relation-
ships between human and animal toxicity data. The toxicity data used to
calculate a DMEG value include threshold limit values (TLV); median lethal
dose (LD50); median lethal concentrations (LC5Q); median toxic dose (TD5Q),
and median tolerance limit (TLm); and carcinogenic, mutagenic and teratogenic
data. The DMEG levels are very approximate concentrations and/or con-
taminants in source emission (air, water or land) that would not result in
significant harmful or irreversible responses in exposed humans or ecology
when these exposures are limited to short durations (acute effects).
The DMEG list supplies a guide for the Level 2 analyst as it provides
a convenient means of organizing a productive approach to inorganic analyses.
The inorganic or organic compounds listed in the DMEG charts are not sought
1
-------
by the Level 1 scheme. However, should an inorganic element or organic
class exceed a/ DMEG concentration guideline, then in the phased approach
to environmental assessment, a Level 2 assessment would be required to
identify and/or quantify the compound forms of the inorganic element of
environmental concern.
The DMEG charts, as originally constructed, contain information on the
concentration levels of interest. Concentration guidelines are necessary
for the decision making process, so that the analyst will know the concentra-
tion at which the Level 1 data trigger the Level 2 activities. Comparison
of the measured Level 1 concentrations of each inorganic element with the
appropriate DMEG decision criterion is used in this manual for proceeding
with and directing the more detailed Level 2 studies. Figure 1 depicts the
decision logic using the DMEG list at the DMEG concentrations. To illus-
trate the approach shown in Figure 1, Table 1 lists some inorganic data
gathered from a stack sample collected with a SASS train at a power plant.
While air data are being used in this example, in other instances the
analyst could use the DMEG values for liquids (yg/L) or solids ( g/g)
depending on the source of the samples.
The analyst will also have a choice of several DMEG values to compare
with the actual concentrations found. As a conservative approach, the most
toxic compound for a given element is selected and its value used for
comparison. The sample concentration is divided by the appropriate DMEG
value. If this ratio is greater than 1.0, then that element in the stream
deserves further Level 2 attention. Based on the results in Table 1, the
elements Li, Be and Pb would be studied further using the Level 2 methodology
described in the following chapters.
The DMEG list also provides possible species and compounds (last column
Table 1) which might result from oil or coal combustion. Independent of
the DMEG compounds, TRW has developed a list of potential compounds by
industrial process. This list discussed in the following section should
be used to supplement the DMEG compounds to form a comprehensive list of
potential compounds emitted from the emission source studied.
-------
*
(ANALYSIS 1
1
C LEVEL 1
SAMPLES
' 1
)
*
EVALUATION
/LEVELI/
/CONCENTRATIONS/
LIST OF /
DMEG CATAGORIES/
LIST VALUES BY
REPORTING POINT>
DMEG
CONCENTRATIONS
BY SOURCE!
AIR. WATER.
SOLID WASTES
ASSIGN DMEG
CATEGORIES TO
LEVEL 1 REPORT-
ING POINT
EACH REPORTING
POINT CAN SPAN
SEVERAL DMEG
CATEGORIES
IS
RATIO
OF SAMPLE
CONCENTRATION
TO APPROPRIATE
JXMEG VALUE LESS
OR GREATER
THAN
1.0
LESS / LIST OF DMEG
" CATEGORIES NOT
REQUIRING LEVEL 2
GREATER
Z
LIST OF DMEG CATEGORIES /
REQUIRING LEVEL 2 /
FOCUS POINT FOR
LEVEL 2 ANALYSIS
Figure 1. Decision Procedure for Level 1 to
Level 2 Transition
-------
Table 1. Application of DMEG Decision Criteria
Compound
Category
Li
Be
B
Pb
Measured Level 1
Source
Concentration
(yg/m3)
2.3 x 102
4.1
1.0 x 103
9.5 x 102
DMEG
(yg/m3)
2.2 x 101
2.0
3.1 x 103
1.5 x 102
Sampl e/
DMEG
(yg/m3)
10.5
2.0
0.3
6.3
DMEG Compounds
Sought in Directed
Level 2 Analysis
Li"1", LiF, Li2C03, and LiH
Be"*"1", BeO, BeO'Al203-Si02
Pb, Pb+2'+4, PbO, PbS04
PbS, PbCO,, Pb,(PO.)0,
Cm
PbCr04, PbMo04, PbHAs04
1.2 OXIDIZED SPECIES EMITTED BY VARIOUS PROCESSES
Level 1 SSMS results and DMEG concentrations are useful to focus addi-
tional effort on specific elements. A major part of the Level 2 effort
will be to determine the compound in which an element is bound. Prior to
starting the Level 2 analysis effort it would be useful to have a list of
potential compounds that might exist in the source stream sampled. This
list can be used to further focus the search for compounds that are known
or postulated to exist in the stream's environment. Unfortunately, the
amount of literature identifying oxidized compounds in effluent streams
from industrial process operations is extremely limited. Where such
effluents are identified, species are generally limited only to common-
place, well known varieties such as the metal oxide forms.
Where such elemental survey data exists (4 through 15) a variety of
oxidized species can be postulated on the basis of process and stream
reaction chemistry. Table 2 presents a listing of environmentally signifi-
cant elements found to exist in effluent streams from various process
technologies. In each case, oxidizing agents exist either as a stream con-
stituent, a process reactant, or both. Table 3 presents a selected listing
of some of the more well known technologies along with typical effluent
characteristics common to each process type. Information from both
Tables 2 and 3 was used to postulate the probable existence of various
-------
Table 2. Elemental Effluents Cited from the Literature
CJ1
Industry Type
Electroplating
Copper Smelting (Primary)
* Roasting
• Reverberating
• Converters
Zn Smelting (Primary)
• Roasting
• Sintering
• Distillation
Copper Smelting (Secondary)
t Sweating Furnace
t Blast Furnace
Coal Fired Boilers
Oil Fired boilers
Sewage Sludge Incineration
Grey Iron Foundry
* Cupola
Glass .Manufacture
Paper Pulp Industry
Ceramic Industray
Municipal Incineration
Lead Smelting (Primary)
• Sintering
* Blast Furnace
• Reverberatory Furnace
Pesticide, Herbicide, Manuf.
Petroleum Refineries
Coking Operations
Nonferrous Alloys Furnaces
Steel Manufacture
t Blast Furnace
• Open Hearth Furnace
• Basic Oxygen Furnace
t Electric Arc Furnace
Cement Manufacture
Aluminum Smelting
0
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Sb
X
X
X
X
X
X
X
X
X
X
X
As •
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
B
X
X
X
X
X
X
Ba
X
X
X
X
X
X
X
X
X
X
Be
X
X
X
X
X
X
X
X
Cd
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Cr
X
X
X
X
X
X
X
Cu
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Mn
X
X
X
X
X
X
X
X
X
X
X
X
X
Mo
X
X
X
X
X
X
X
X
Pb
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Se
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Sn
X
X
X
X
X
X
V
•
X
X
X
X
X
X
X
X
X
X
X
X
X
X-
Zn
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Hg
X
X
X
X
X
X
X
X
X
X
X
Hi
X
X
X
X
X
X
X
X
X
X
X
X
d
X
X
X
. X
X
X
X
F
X
X
X
•x
X
X
X
X
X
x'
X
X
s
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X .
X
X
X
-------
Table 3.
Conditions and Typical Inorganic Compounds Emitted
from Various Industrial Processes
Process Identification
1) Steel Plant
(a) Blast Furnace
(b) Sinter Plant
(1) Windbox:
(2) Discharge end:
(b) Coke Ovens
(1) Oven off gas
(2) Quench Tower
(3) Sour Quench
Water
2) Coal Fired Boilers
Effluent Characteristics from Predominant Process Operations
Particulate Data
Size
Distribution
15-90% <74u
15-45% <40u
9-30% <20u
4-19% <10p
1-10% <5u
80% <100
10 <10
Highly variable
95-97% >47y
Not applicable
25% <10
49% <20
79% <44
gr/SCF
4-30,
7-10 avg
0.2-3.2
1-5
1-15
0.05-0.1
-
2.9 to
3.7 avg
(1)
Flow Rate
(a) 40-140
(b) 60-138
(a) 30-460
(b) 148-230
(a) 0.03-0.2
2.1 M ft3 per
charge
o
900 M fr per
quench
water usually
sluiced; flow
variable and
cyclic
(a) 297-397
(b) 362-434
Temperature
OF
390 at
throat
3000 at
furnace
100-400
100-300
<1832
140-150
-
245-258
Moisture
Vol %
9.6
2-10
«
Variable
depending
on point
in coking
cycle
Effluent
consists
primarily
of steam
-
6.4
Oxidized Species Present in Primary Effluents
Cited from Literature
G, L, S
Fe203> FeO, Cr203
SiO,, Al,0,, CaO,
£ £ *J
MgO, MnO, ZnO, CuO,
NiO, PbO, F, Ba, Cd,
Mo, Sn, V, S02, S03
G
HgO, SeO,
L.
L
A 0 V 0
23' 2s
G, L, S
VQ F 0 Al tl
2* 23* 23*
CaO, MgO, Ni,0, TiO,,
L. L.
v f) Cpfi p n en
INnU) PCU » r n*J[- » JUn )
S02, N0x, 02 (6.2 to
6.8%), chlorides
Probable Based on Stream
Chemi stry
G
The fluoride and sulfide
forms are possible for all
these metals.
L
All indicated forms also
probably exist in liquid
effluent streams.
G
Oxides and sul fides and
halogenated species are
probable for the following
elements (see Table 1):
Sb, A., Be, Cd, Cr, Mn, Pb,
Se, V; Hg, Ni
i
L
The Quench water will con-
tain all of the above as
well as oxidized forms of
most of the trace elements
found in coal.
Oxides, sul fides and halo-
genated species are prob-
able for the following
elements: Sb, As, B, Ba,
Be, Cd, Cr, Cu, Mn, Mo, Pb,
Se, Sn, V, Zn, Hg, Ni. Ash
and aqueous effluents also
contain all species listed.
(1) Flow rate is expressed in: (a) M SCFM or (b) M SCF/ton of product processed.
G = gaseous phase; L • liquid phase; S = solid phase.
-------
Table 3. Conditions and Typical Inorganic Compounds Emitted
from Various Industrial Processes (Continued)
Process Identification
Forest Products Industry
(a) Kraft Pulp Mills
(1) Recovery Furnace
(2) Lime Kiln
(3) Smelt Dissolving
Primary Nonferrous
Metals Industries
(a) Copper
(1) Roasting Furnace
(2) Electrolytic
Oof •« n-i nnit
KeTining't
(b) Lead
(1) Sinter Machine
(2) Blast Furnace
(c) Zinc
(1) Roaster
(2) Sinter Machine
(d) Aluminum
(1) Reduction Cell
Effluent Characteristics from Predominant Process Operations
Particulate Data
Size
Distribution
50-855! <2u
95% <25u
90% <5u
15X <10u
100% <10u
.0.03 to 0.3
14% <5,
31* <10, 70% <20
100% <10
Submicron
particulate
gr/SCF
3-8
avg 3.8
3-20
0.17-1.3
6-24
0.4-4.5
1-11
5-65
0.4-4.5
0.03-2.0
Flow Rate
(a) 20-568
(b) 278-568
(a) 7-50
45 SCF/air
dried ton
(a) 60-131
(a) 140
(b) 130
6-14
25-30
140
2000 to 4000
CFM/cell
Temperature
°F
.270-650
avg 350
400-900
170-200
600-890
250-600
150-250
730-900
320-700
-
Moisture
Vol *
20-40
400-600
Ibs/air
dried ton
670 IDS/
air dried
ton
-
-
-
Dew Point:
122-140
-
Oxidized Species Present in Primary Effluents
Cited from Literature
G, L, S
CO,,, 0 , SO
£. c. £.
Sb, Cr, Zn,
Carbonates, sul fates,
chlorides and oxides
are cited in the
literature as general
groups.
G, L
As?03, Sb-0,, Al?0,,
£ J £. .3 £.6
ST02, S02> S03, Cd,
Pb, Se, Hg, 0
G, L
As203, CdO, CdS03,
CdS04, CdSe, TeO,
PbO, Pb304, ZnO, Sb,
As, Mo, Hg, 0
As, Cd, Pb, Se, Zn,
Hg (from Table 1), 0
Q. -.._-_____
A1203, Si02, Fe203,
Na.O, Fluorides, 0
£
Probable Based on Stream
Chemi stry
G, L, S
Chlorides, hypochloriies,
and oxides are highly
probable for a wide variety
of elements due to bleach-
ing and other oxidizing
operations.
Cd, Pb, Se, and Hg oxides,
sulfites, artd sulfates are
probable.
Compounds of varying oxi-
dation states are probable,
i.e.: Te02, Te03,
(Te02)2 S03 etc. Also
CdTe and other complexes.
Oxidized species in varying
oxidation states are prob-
able -in gaseous and liquid
ef f 1 uents
Significant quantities of
the following have been
found as a result of T-4
field effort: Be,
chlorides, V, Mo, Se, S,
bromides, As, Ni, Ba, Mn.
These elements may exist
as hal ides, oxides, or in
various sulfur forms.
*Process data for electrolytic refining is highly variable; literature does not cite specific flow data.
-------
Table 3. Conditions and Typical Inorganic Compounds Emitted
from Various Industrial Processes (Continued)
00
Process Identification
Refinery Operations
(a) Claus Plant (oil/
gas)
(b) Fixed Bed Catalyst
Regeneration
(c) Moving Bed
Catalyst
Regeneration
(d) Fluid Coker off-gas
Coal Conversion
(a) Gasification and
Liquefaction
Operations
(1) Coal Preparation
(2) Quenching and
Cooling
(3) Fixed Bed
Catalyst
Regeneration:
(4) Sulfur Plant:
(5) Tar Separation
Fertilizer Manufacture
Effluent Characteristics from Predominant Process .Operations
Particulate Data
Size
Distribution
gr/SCF
Flow Rate
•'
Temperature
OF
Fixed and
moving bed
regenerabl e
catalyst
function at
about 850
to 100QOF
at 300 to
700 psig
Moisture
Vol %
Same parameters and species as defined under "Refinery Operations.1
6.3% <5w,
12* <10u
22% <20u
29% <30u
34* <40u
0.7 to 4.0
(a) 16.5
(one unit)
201
_
Oxidized Species Present in Primary Effluents
Cited from Literature
Bromides, chlorides,
fluorides, SOj, 503,
Al, Ba, Be, Br, Ca,
Cd, d0, Fe, K, Li,
Mg, Na, Ni
All elements associ-
ated with coal are
found in G, L, and S
coal conversion
effluent streams.
G, L, S
P205, Si02, A1203,
MgO, CaO, Fe203,
fluorides, 02, S02>
SiF4, HF
Probable Based on Stream
Chemistry
A wide variety of oxide suV
fur form and halogenated
species are possible for
all environmentally signif-
icant elements.
While specific compounds
are not cited in great
detail in the . 1 i terature ,
many analyses identifying
the existence of classes
are available, i.e.,
halides sulfur forms.
oxides etc. All may be in
G, L, and S forms.
Gasifiers are oxygen or air
blown, stream chemistry Is
complex, physical condi-
tions are extreme.
Process conditions are
ideal for the formation of
a wide variety of oxidized
species in both gasifica-
tion and liquefaction
systems.
A number of trace elements.
in addition to those listed.
exist in G, L, S effluents.
Stream chemistry is condu-
cive to the formation of
oxides and fluorides.
Stream parameters are highly variable depending on process design, see text.
-------
Table 3. Conditions and Typical Inorganic Compounds Emitted
from Various Industrial Processes (Continued)
vo
Process Identification
Oil Fired Boilers
Effluent Characteristics from Predominant Process Operations
Participate Data
Size
Distribution
90*
B203, Mn02, SrO, TiO
Probable Based on Stream
Chemistry
Other significant trace
elements found 1n oil par-
ticulates include: Sb, As,
Be, Cd, Se, Sn, Zn, Hg.
These elements also prob-
ably exist as oxides. All
elements 'listed may exist
as chlorides and in vari-
ous sulfur forms.
-------
oxidized species in process effluent streams for the technologies listed.
The existence of such species may be predicted with a high confidence
level because ionization energies and heat of formation energies of numerous
compounds are well within process operating conditions. Also, the exis-
tence of even a few literature-cited oxidized species is enough to indicate
the existence of generic compound types where elements with similar
reactivities have been found.
The assumptions used to postulate the existence of oxidized species
(far right-hand column - Table 3), other than those cited in the literature,
are as follows:
• The existence of oxidizing agents present either in the stream or
during process operations or both.
• The existence of literature-cited oxidized species within the
process stream.
a The existence of temperature and pressure conditions equal to or
above ionization and/or heat of compound formation for literature-
cited elements.
• The composition of feed material and process reactants.
t The elemental analysis of acquired samples such as SSMS (such as
the presence of Cl or F in a solid particulate sample indicating
the presence of chlorides or fluorides).
The data provided by Table 3 should be used in two ways. In most
cases the list of DMEG compounds given for a specific element will include
compounds not present in effluent from the process being studied. Using
Table 3 those compounds can be removed from the potential compound list to
be studied and/or added to by other compounds not listed in the MEG
categories. In addition to this use Table 3 provides information on the
mass loading, which can be used to specify sampling times to obtain enough
sample for later Level 2 chemical analysis. The following section will
provide some guidelines for Level 2 sampling.
1.3 LEVEL 2 SAMPLING
No matter how accurate or precise the analysis procedure, the validity
of the results depends on the accuracy of the sampling procedure. Current
10
-------
Level 1 sampling procedures for liquid and solid samples are probably
sufficient for most Level 2 efforts with the provision that:
• Samples are time-integrated to account for process operational
variances.
• Specialized sampling procedures are employed for unstable species.
The following techniques or procedures have been selected to provide an
overview and guidance for the selection of an appropriate Level 2 sampling
method.
1.3.1 Solid Sampling
Solid Sampling covers a broad spectrum of material sizes ranging from
large lumps to powders and dusts. There is an equally diverse assortment
of potential sample sites including railroad cars, large heaps, plant
hoppers, conveyor belts, and process stream pipes. Obviously no one method
or piece of equipment is suitable for all situations, but the advantages
and disadvantages of each must be weighed in the light of individual field
test conditions. The following discussion provides an overview of common
solid sampling situations and summarizes the sampling approaches and
alternatives available to a field test team.
1.3.1.1 Sampling Methodology and Equipment Survey
The Level 2 sampling of solids may include the use of three manual
grab sampling techniques: shovel sampling; boring techniques, which include
pipe or thief sampling; and auger sampling. Mechanical samplers, both
moving and stationary, can also be used to obtain solid samples. The chief
consideration of solid sampling is the acquiring of representative samples.
Shovel sampling procedures include grab sampling, coning and quarter-
ing, and fractional shoveling. Grab sampling consists of taking small,
equal portions at random or regular intervals, typically from railroad cars,
large heaps, or hoppers. The method is quick and inexpensive. However,
grab sampling makes no allowance for segregation of the sample by particle
size and also tends to give consistently high or consistently low results
depending on the person sampling. As such, grab sampling should be used for
survey sampling.
11
-------
Coning and quartering consists of carefully piling the material into a
conical heap, with subsequent flattening of the cone into a circular cake.
The cake is then marked into quadrants; two opposite quadrants are taken as
the sample and the other two quadrants are discarded. The entire process
is repeated until the desired sample size is obtained. In general, this
method is time-consuming and the symmetry of the intended vertical size
segregation is difficult to achieve in practice.
Fractional shoveling is applicable to materials being loaded, unloaded,
or moved from one place to another by shoveling. In fractional shoveling,
every third, fourth, fifth, or tenth shovelful is taken as the sample. This
method is inexpensive and relatively fast. If performed conscientiously,
fractional shoveling can be more reliable than coning and quartering.
However, its applicability is limited and errors are easily introduced by
carelessness.
Pipe boring techniques represent another class of solid sampling method-
ology applicable to material stored in piles, silos or bins. The usual
method of pipe boring is to insert the pipe into the material to be sampled
at regular intervals. The method is fairly reliable provided that the pipe
is long enough to reach the bottom of the material. However, it is only
applicable to fine or powdered dry materials, because lumps or any stick-
iness will jam or plug the pipe. Small pipe borers can be used to sample
sacks or cans of material. There are primarily two designs of pipe borers
that give best results. One is a simple pipe that is tapered so the end
first inserted is smaller in diameter than the handle end. A more sophis-
ticated design, known as a thief, makes the sample more representative
vertically. It consists of two close-fitting concentric pipes sealed at
the base in a conical point. Longitudinal slots are cut along the side of
each pipe. The thief is inserted with the slots turned away from each other
and then, when the sampler is in position, the outer pipe is rotated, lin-
ing up the slots and allowing the inner pipe to fill the sample. For proper
results with any design of pipe borer, the opening through which the sample
material passes (slots or circular pipe ends) must be large relative to the
maximum particle size.
Auger samplers, a form of drill, pack the sample in the helical groove
of the auger and can be enclosed in a casing if the nature of the sample is
12
-------
such that it will spill when the auger is removed from the hole. Like the
pipe borers, they are simple to use and have the further advantage of being
applicable to a greater variety of materials. For example, augers work well
for materials that are packed too hard for a pipe sampler to be forced in.
For very packed materials, machine-driven augers are available. However, a
thief sampler would be the better choice if sample spillage is a possibility.
Also, both pipe samplers and augers yield poor results if the material
being sampled is poorly mixed.
Mechanical samplers require that the sample material be in motion to
present it to the cutters as a thin ribbon or stream. Design considerations
for feeding these samplers and catching the sample and rejected material
generally necessitate the permanent installation of the sampler into the
flowing sample stream. Numerous mechanical samplers have been designed;
the most popular designs have been variously modified to satisfy specific
applications. However, all mechanical samplers fall into two general
types: those that take part of the stream all of the time (stationary
samplers), and those that take all of the stream part of the time (moving
samplers).
In stationary mechanical samplers, the entire sample stream is fed
continuously through the device and stationary cutting edges divide out
and remove specific fractions. The two best-known designs of this type
are rifflers and whistle-pipes.
Rifflers take several slices of the stream by means of parallel chutes
alternately placed at 90° angles to each other, thereby cutting the stream
in half. Successive rifflers can be arranged in banks to cut the stream
into any desired fraction. The smaller the chute width, the greater the
number of increments in the sample. Therefore, the accuracy of riffler
sampling increases as the ratio of chute width to particle size decreases,
to the limiting condition where the chutes tend to clog. In general,
chutes should be at least three times the diameter of the largest size
particle to avoid clogging. Care must be taken to feed the riffler with a
well-mixed, uniform sheet of material since any compositional variations
due to cross-sectional segregation are multiplied by a bank of rifflers.
13
-------
A whistle-pipe sampler consists of a vertical pipe with notched
openings cut halfway through the pipe, each spaced 90° horizontally from
the one above. Rectangular steel plates are placed in the notches at a
45° angle to the vertical so that the top edges coincide with a diameter
of the pipe. Thus each notch halves the sample and, with a series of five
openings, the sample obtained is 1/32 of the original volume. The same
fraction with improved accuracy can be obtained by using a cutter arrange-
ment that quarters the stream, rejecting opposite quarters, and spaces each
cutter at 45° horizontally from the one above. In either design, a hopper-
shaped liner is placed above each notch to re-center the stream before it
reaches the next cutter.
Both rifflers and whistle pipes have irresolvable design problems
that reduce their reliability. Worn or bent cutting edges distort both the
volume and the particle size distribution of the sample. The housing nec-
essary for these samplers prevents examining them for clogged openings while
in operation. Material streams whose composition varies along the trans-
verse section are even further segregated by either of these samplers.
Moving samplers consist of cutters that move through the free-falling
sample stream taking all the stream for the duration of time they are moving
through it. There are two ways of effecting this. One is with rotating or
oscillating samplers whose cutters are set on the radii of an arc, and the
other is with straight-line samplers whose cutting edges are set parallel
to each other and perpendicular to the line of their path.
Among the well-known designs of rotating arc-path samplers are Vezins,
Synders, and Chas. Synders. They all consist generally of scoops with
vertical sides, set on an axis parallel to the stream flow. The best
oscillating samplers are known as Bruntons. The scoop travels back and
forth across the stream in a pendulum-type motion. The travel path must be
sufficiently long to minimize the bias created by taking more samples from
the sides of the stream than from the middle. All the arc-path samplers
have the advantage over stationary samplers in that they take an accurate
cut, are simply constructed, and are accessible for observation while in
operation. However, damp sample material may tend to clog the scoops and
care must be taken to maintain the cutting edges in good condition and to
keep them completely radial.
14
-------
The straight-line samplers are generally considered to be the most
reliable and accurate of all available types of samplers. The design of
their cutters is such that the sampling scoop spends an equal amount of
time in every portion of the stream. Generally the travel is at right
angles to the stream. Though they provide increased reliability, these
samplers require more maintenance and attention because of their increased
mechanical complexity.
1.3.1.2 Sample Collection and Storage
It is always preferable to sample a moving stream either in pipes or
from conveyor belts, particularly if there is a large particle size range
in the material. Stored containers or heaped beds of material tend to
settle, creating segregation of particles according to size and density,
and it is difficult to compensate for this bias in the sampling. Further-
more, large masses of stored material are extremely difficult to handle.
The interior portions are relatively inaccessible and the amount of time
and space needed to move the material enough to take a representative
sample can quickly become prohibitive. However, such situations can
generally be avoided by a good sampling test plan.
Typically, the solid materials of interest are the feed materials
and the residues from particulate scrubbers such as baghouses, high energy
Venturis, and electrostatic precipitators. Raw feed stock as it passes
through the process stream may pick up other materials as contaminants and,
therefore, differ greatly in composition from what is finally being fed to
the process. Consequently, samples should be taken at the last possible
site before the stream is fed into the process. This means that sampling
will generally be conducted from a feed hopper, if accessible, or from
whatever pipes or conveyors feed the material to the process. Similarly,
scrubber residues can be sampled from whatever collection hopper the device
has or from pipes going to the hopper. Extra handling steps only increase
the chances of the sample becoming contaminated.
As part of their own process control, many plants may have some type
of mechanical sampler already installed into their process stream. When-
ever possible, these devices should be used for taking samples. They are
reliable, take representative samples, and are fast and easy to use. Before
15
-------
being used, however, the samplers' operation and cutting edges should be
checked to ensure accuracy. If reliable automatic samplers are an inte-
grated part of the plant and are available, no sampling equipment will be
needed by the field test personnel.
In cases where it is decided to take samples from moving conveyor
belts, the standard procedure is to stop the conveyor at regular intervals
(e.g., every 10 to 15 minutes) and shovel off a section of the material.
This is continued until the desired sample size is obtained. Flat-nosed
shovels with straight perpendicular sides are best for these sampling
purposes.
Another alternative is to sample process streams as they move through
pipes if there are appropriate ports. A variety of pipe samplers are com-
mercially available. The type most suitable for trace element sampling
is the pneumatic sampler, which eliminates the screw type or scraping action
of other types of samplers which grind the sample and abrade the sampler,
thereby introducing considerable contamination.
An example of a pneumatic sampler is the Model RTA of Quality Control
Equipment Corporation. All parts in contact with the sample can be Teflon
or nylon lined, which is a major advantage. It can be used to sample
solids with a particle size less than 0.64 cm (1/4 in. in diameter), as
well as slurries and liquids from pipes fitted with at least 2-inch
diameter sampling ports. Samples collected by any of the techniques
described must be stored in pre-washed and dried plastic bottles or plastic
lined drums. If plant personnel are to take the sample, they should be
provided with the appropriate bottles or containers.
1.3.2 Liquid Sampling
The factors which must be considered in accurately sampling a fluid
stream for inorganic materials include:
• Stream homogeneity
• Stream flow rate and variations
• Prevention of sample loss
• Sources of contamination
• Sample size.
16
-------
Of these, stream homogeneity is perhaps the most important factor.
Unlike stack effluent streams which are mixed fairly evenly due to higher
thermal agitation and lower fluid viscosities, liquid streams tend to be
more stratified and require more careful sampling. A flow-proportional,
composite sampling technique is required for sampling liquid and slurry
streams for trace materials. By comparison, obtaining a composite repre-
sentative gaseous sample from a stack, a traverse of the pipe or duct is
made. However, this is usually impractical in sampling liquid streams.
In liquid streams, a composite sample can be taken by using several
differently positioned probes, a single multiported probe, or a combination
of these. While either of these approaches is suitable, the single
multiported sampler is usually more convenient.
In the case of slurry sampling, it is also important to avoid segre-
gation of liquid and solid phases. This is similar to the requirement for
isokinetic sampling in particulate-laden gas streams. An EPA internal
study evaluated 60 commercially available models of automatic sewer flow
samples (126). In this study, sampling velocity was determined to be the
most critical factor in sampling sewage slurries. Two units were found to
perform acceptably: Quality Control Equipment Company Model CVE and Testing
Machines, Inc. Fluid Stream Sampler. Both of these units are portable and
completely automated. The QCEC unit also has a built-in ice chamber for
automatic refrigeration of temperature-sensitive samples. Another common
practice for the preservation of liquid samples is to freeze them until they
are ready for analysis. However, this practice is now being questioned for
trace material analysis because of the tendency for metal ions to precipi-
tate upon freezing. The technique should only, therefore, be used for
preserving samples for bacteriological analyses.
It is well known that trace materials in liquid phases may be lost
from a sample through adsorption on sampling line or reservoir surfaces.
Borosilicate glass (Pyrex) surfaces appear to be particularly effective in
removing trace heavy metals, especially under alkaline conditions. How-
ever, plastics such as polyethylene, polypropylene, and Teflon show little
or no tendency to adsorb inorganic materials.
17
-------
In addition to sample loss due to surface adsorption, a sample may also
be contaminated by elements from those surfaces. Surface wall material can
be deposited in a sample either by a chemical extraction of the wall materials
by agents in the sample or by physical abrasion or erosion of the wall by
a sample. The latter case could be a significant problem for slurry systems
because the abrasive nature of the sample could expose unpassivated layers
of the wall to chemical interaction with the sample.
Another important factor which must be considered in sampling liquid
streams for inorganic material constituents is the sample size requirement.
Two principal requirements govern sample size. The first requirement is
that the amount of sample collected must be sufficient for the testing and
analysis procedures to furnish accurate and precise results. The second
requirement is based on the statistical sampling error that can be tolerated.
The minimum sample required for analysis varies between 1 and 1000 yg for
the trace materials of interest using the proposed analysis procedures.
For the lower ppm concentration levels of interest, this translates into
minimum sample volumes ranging between one ml and one liter. This range of
sample volumes is easily within the operating limits of presently available
liquid sampling equipment and presents no special difficulties.
1.3.2.1 Equipment
The sampling of liquids and slurries in the categories defined above
requires equipment suitable for point sampling, as well as for sampling
from ponds, reservoirs, open tanks, open channel flows, and pipes which have
built-in sampling ports.
For pond and tank sampling, and in preliminary or point sampling, point
and depth-integrated samplers are commonly used. The Sirco Uniscoop, which
is made of No. 316 stainless steel, is recommended for point sampling.
The Uniscoop has a handle for collecting be!ow-surface samples and is con-
venient to use. A depth-integrated sampler consists of a weighted bottle
and is easily fabricated.
18
-------
The recommended automatic samplers are Model CVE (Quality Control
Equipment Corporation) and Model 1940 (Instruments Specialty Corporation).
All components in both units which come in contact with the sample are
composed of polypropylene, polyethylene, or Tygon, and the sample never
passes through any valves or pumps. Both units can perform short-term or
long-term sampling at certain time intervals proportional to time or flow
rate. The units also have built-in ice cabinets to preserve the samples
at lower temperatures. The units offer a long-term stability without
mechanical or electronic malfunctions.
While the Model CVE sampler provides composite samples directly in the
field, and the Model 1940 takes sequential samples that are stored in
separate bottles, both models can perform time or flow proportional sampling
depending on the availability of a flow measuring device. The Model CVE
was rated the best unit in a study conducted by the EPA Regional Office at
Kansas City, Missouri.
The equipment discussed in this chapter is capable of handling a wide
variety of process streams found in most industrial applications. For
sampling streams having highly corrosive materials, the Teflon coating of
all metal parts should be considered in order to prevent contamination of
the samples and corrosion of the sampling equipment. However, for most
applications, inherent durability and flexibility of the off-the-shelf
samplers are adequate.
1.3.2.2 Sample Collection Separation and Storage
Prior to sample collection, the liquid sampler must be cleaned in the
field to prevent any contamination. The sampler is flushed out with a liter
(quart) of nitric acid (1:1) followed by several liters of high purity
water to eliminate any particles introduced during shipping and field
storage. The sampler is then placed near the process stream to be sampled.
This site should be free of windblown contamination. The sample probe or
hose is introduced into the stream and the timer set for the proper
sampling period.
All storage bottles that are going to be in contact with liquids must
be cleaned to prevent contamination of the sample by elements leaching from
19
-------
the container material. The following procedure (127) is recommended for
most plastic bottles:
1. Fill with 1 + 1 HC1 (AR grade).
2. Allow to stand one week at room temperature (80°C for Teflon).
3. Empty and rinse, with distilled water.
4- Fill with 1 + 1 HN03 (AR grade).
5. Allow to stand one week at room temperature (80° for Teflon).
6. Empty and rinse with distilled water.
7. Fill with purest available distilled water.
8. Allow to stand several weeks or until needed, changing water
periodically to ensure continued cleaning.
9. Rinse with purest water and allow to dry in a particle- and
fume-free environment.
Following sample collection, the various phases present in a liquid or
slurry must be separated to prevent either gross disruption of the material
content of each phase or the oxidation of selected species due to continued
contact with the liquid phase. For these reasons a preliminary field
phase separation procedure is recommended for all slurry streams.
The equipment necessary for separating the phases of a slurry sample
in the field consists of:
• Filters (Gelman Acupore 1.2 ym)
• Nalgene Buchner funnel and filter flask
• Small vacuum pump
t Acids, bases, 50% Acetone solution, and high purity distilled water.
All the Nalgene equipment used in separation of the sample phases
should be pre-washed using the procedure presented above to prevent con-
tamination. A solution of 0.1 N HN03 (high purity) in a squeeze bottle must
be used as a rinse between samples to prevent cross-contamination. Enough
clean replacements should be available, in the event that the liquids leave
a film on the plasticware.
20
-------
Stabi1i zati on Procedures
For most cases, the standard EPA preservation procedures (128) can be
used to stabilize the liquid samples obtained. The procedures include the
two basic approaches for cations and anions. In both cases the liquid posi-
tion of a sample is divided in half and placed in the pre-cleaned plastic
storage bottles. Into one container enough HN03 (High Purity) is added to
reduce the pH<2. This amount is recorded as well as the original volume
of liquid stabilized. The addition of HN03 stabilizes the trace metals and
prevents absorption on the container walls. With the exception of Hg,
these samples are stable up to 6 months. It is recommended that Hg be
stored no longer than 13 days in plastic containers due to its ability to
permeate the plastic container and the possibility of Hg diffusion from
the laboratory air into the container.
Stabilization of anionic species is slightly different. Most anions can
be stored in the plastic container at 4°C, but only for a limited period of
time. Chloride, F~, and S0^~ can be stored up to 7 days, while Br, I",
NOg, should be analyzed within 24 hours of collection. Reference 128 pro-
vides more detail on these and other anions.
For water-solid streams, the slurry is first filtered using equipment
and filters described above. The filtrate is stabilized like the clear
liquids. The filtered solids are rinsed with a 50% acetone solution to
remove excess moisture, while minimizing dissolution of the solids. This
procedure works very well with Flue Gas Desulfurization (FGD) sludges, and
is applicable to solids obtained from oxidizing processes. The rinsed filter
.cake is then dired in an oven (explosion proof) at 110°C.
1.3.3 Gas Sampling
Level 2 gas sampling is divided into three separate categories:
• Fugitive Gas Emission
• Flue Gas Emission
• Flue Gas Particulate Emissions
The following sections will present some guidelines for Level 2 sampling
methodologies for these categories.
21
-------
1.3.3.1 Fugitive Gas Emissions
Sampling participate matter from fugitive emissions requires a well
thought-out strategy. IERL/RTP has specified a variety of procedures in
Reference 17.
1.3.3.2 Flue Gas Emissions
It is recommended that all gas samples be time integrated to improve
the overall accuracy of the analysis. Most oxidized species can be sampled
with either polymeric (e.g., TEDLAR), or glass containers. The best
approach is to use on-line continuous monitors whenever possible. Con-
tinuous readout will provide valuable information on the cycles or changes
in the emission as the process varies.
1.3.3.3 Flue Gas Particulate Emissions
After the Level 1 analysis has been completed on the particulate
samples taken from the SASS train, the analyst must decide whether to
initiate the Level 2 analysis on the remaining SASS train samples. The
use of samples from a SASS train are attractive from the cost and time
standpoint. There are other points which should be considered prior to
initiating the Level 2 analysis program on SASS train samples:
• Single point. While every effort is made to pick a representative
point, flow fluctuations and particle stratification can lead to
sampling errors on the order of a factor of 2 at stack or con-
trol device inlets and as high as a factor of 3 at the outlet of
control devices.
j
• Contamination of Samples. The SASS train is an all SS 316 train
which means there is a definite potential for Ni, Cr and Fe con-
tamination. Nickel, Cr, and Fe contamination is especially
prevalent in gas streams containing S02 and HC1. The main point
of attack is in the condenser module where dilute solutions of
H2S04 and HC1 will readily attack the condensor surface. Because
of these problems, it is recommende!d that an alternate, all glass
train (Figure 2) be used for Level 2 testing. This train has been
used in field tests and because of its all-glass design Ni, Cr, and
Fe can be monitored. The train itself consists of a particulate
section which has a 3 u cutoff cyclone and a glass fiber filter.
The particulate section is followed by a series of oxidative
impingers which use the same chemistry as the SASS train. The
main drawback with these glass trains is that they currently are
designed to sample at 28 Lpm (1 cfm). This low sampling rate is
partially offset by the high sensitivity of most inorganic
analytical techniques and by its ability to perform a complete
traverse of the flue gas stream.
22
-------
ro
CO
FILTER
HOLDER
HEATED
CONTAINER
(200-C)
CYCLONE
OWFI
THERMOMETER
CHECK
VALVE
It
DWER1TE
THERMOMETERS
BY-PASS
VALVE
0.2M(NH4)2S208
•(0.02M AgNOg
.VACUUM
GAUGE
VACUUM
LINE
MAIN
VALVE
DRY TEST METER
AIR-TIGHT
PUMP
Figure 2. Level 2 Inorganic Sampling Train (Glass)
-------
In addition to obtaining a bulk sample, particle size impactors can be
used to provide sized particle samples for analysis. Section 2.5.3.2
describes the use of an impactor to perform elemental analysis by particle
size.
24
-------
2. INITIAL SAMPLE CHARACTERIZATION
In the first phase of the sample analysis, the goal is to determine
accurately the elemental and anion composition of the sample as well as its
general physical and morphological characteristics. The techniques used
to perform this characterization are summarized in Table 4. Developing a
list of potential compounds has already been discussed in Section 1. The
next step is to view the sample under a Polarized Light Microscope (PLM)
and take a color photomicrograph, so that any changes in the general
appearance of the sample are monitored and recorded during the analytical
work. The PLM can provide both a measure of the complexity of the sample
from the number of different particle types, and also some compound identi-
fication from the optical properties of a particle. In conjunction with
the PLM work, a Thermogravimetric Analysis/Differential Scanning Calorimetry "
(TGA/DSC) scan of the sample is made. This method is used primarily to
determine 1) the stability of the sample, and 2) an appropriate temperature
at which to dry samples to be used in later tests. In some cases, the
compounds present can be determined by weight loss at specific temperatures.
Quantitative anion analysis is performed using ion chromatography (1C).
Besides its excellent sensitivity, 1C provides survey information on the
anions present. Accurate elemental analysis is important for quantitation
of the compounds present. Several techniques Atomic Absorption Spectroscopy
(AAS), X-Ray Fluorescence (XRF), Particle Induced X-Ray Emission (PIXE),
Inductively Coupled Argon Plasma (ICAP), and Neutron Activation Analysis
(NAA) are discussed, and recommendations on their use are given.
The following sections provide information on the techniques sum-
marized in Table 4- The information provided is directed at the know-
ledgeable chemist and is designed to improve his understanding of the uses
and limitations of the methods described.
A section discussing the integration of the above methods into a
coherent analysis pattern is discussed in the last section of this chapter.
25
-------
Table 4. Summary of Initial Sample Characterization
Analysis Method
Principle of Operation
Information Derived
Compound Identification Procedure
Limitations
Level 1 Spark
Source Mass
Spectrometry
(SSMS) Data
ro
CT»
RF potential used to breakdown
sample placed in two electrodes
Resultant ions accelerated out
of source through electrostatic
and electromagnetic analyzers
(similar to organic mass
spectrometry)
Determine mass distribution in
resultant Ion beam. Use detec-
tion system that provides
required sensitivity and
precision:
• Photographic plate system.
Used for total characteriza-
tion of sample since entire
periodic table is examined
and possible interferring
ions are resolved
• Electrical detection system.
Good for single element
determination
Provides elemental concentration
data on elements
Can determine trace elements in
quantities as low as 0.01 ppm
Absolute sensitivities range
from 1 through 400 ng
Need elemental distribution data
to reduce compound choices
Elemental information especially
useful in interpreting IR and
XRD data
Accuracy of analysis typically
100 to 500%. If only Level 1
data is used, could allow a Yes
or No answer to the presence of
jossible compounds
Thermogravimetric
Analysis (TGA)/
Differential
Scanning
Calorimeter (DSC)
TGA records weight loss or gain
as material is heated
DSC measures heat evolved or
absorbed as sample is heated
TGA provides specific information
on thermal stability of sample:
weight loss can sometimes be
correlated with decomposition
of specific compounds
DSC data gives information on
phase transitions or chemical
reactions in sample
TGA/DSC normally cannot deter-
mine compounds present in complex
mixtures without information on
elemental and anion composition
Primary use in this identification
scheme is to provide stable drylnj
temperatures and Identify any
reactive or volatile materials
present
Since small sample sizes are nor-'
roally used, chemical stability of
low concentration materials is
not seen
The maximum temperature of most
Instruments Is 1000°C, which Is
below decomposition point of many
compounds
Polarized Light
Microscope (PLM)
Particles are collected from
various sources, crushed' to
0.05 mm, and examined with
microscope. Of special
importance are observations of:
Refractive index (relief)
Isotropy or anlsotropy
Birefringence
Pleochroism
Fracture
Color
Crystal habtt
At low magnification, general
appearance of sample is noted for
quality control of sample
handling/storage
At higher magnification, crystal
structure, color, refractive/
Index are measured for single
particles
Initial view indicates minimum
number of different particles
and their potential compounds
ATI crystalline compounds have
specific refractive indexes which
can be used to Identify the
compound
Amorphous materials can sometimes
be identified by comparison to
known substances via particle
atlas
Limited to single particle analysis
Trace constituents adsorbed on
particles or extremely small parti-
cles (O.Sv) must be measured with
another technique (SEM-EDX)
Homogeneity Important for correct
identification
Results difficult to quantitate
-------
Table 4. Summary of Initial Sample Characterization (Continued)
Analysis Method
Principle of Operation
Information Derived
Compound Identification Procedure
Limitations
Micro-Solubility
Tests
View small amounts of sample
under PLM while adding cold
Mater, hot water, dilute HC1,
and dilute bicarbonate
solutions
Record information on Individual
particle solubilities
Solubility of particles 1n
specific solvents indicate
the class of compounds present
Use solubility data to verify
later results
Use solubility data in conjunc-
tion with anion micro-spot tests
to reduce number of possible
compound choices
Micro-tests on a microscope stage
require good technique and
extreme care
Results reflect composition of
single particles and not the bulk
of the sample
Micro-Spot
Tests
po
Atomic.Absorption
Spectroscopy
(AAS)
Isolate single particles on
stage of PLM:
• Micro-spot test for specific
an Ions
NOj, NOj, COl, CT)
* Compare to quantitative
standards
Introduce sample into AAS and
decompose with flame or heat
or furnace until gaseous metal-
lic atoms are formed
Make quantitative determination
of amount of metal in sample by
comparing level of sample
absorbance at specific wave-
lengths with that of known
standards
Reveals presence or lack-of
specific anions
Provides concentration data on
metals
With flameless techniques,
detection limits between
0.001 and 1 ng are possible for
various elements
Combination on anion and solubil-
ity information limits the
number of cations present in
sample and aids in single parti-
cle compound identification
Specific cations can be identi-
fied and quantitated
These metals can be correlated
with specific particle types
Cation information can be coupled
with solubility and anion con-
tent information to aid in
compound identification
Micro-tests on microscope stage
require good technique and extreme
care
Results difficult to quantitate
Analyses of non-metals and
metalloids cannot be performed
-------
2.1 POLARIZED LIGHT MICROSCOPY
Polarized light microscopy (PLM) is the first direct compound analysis
method. Many particles can be identified by the determination of such
properties as the refractive index, isotropy or anisotropy, birefringence,
pleochroism, fracture, color, and crystal habit. An excellent survey
article on PLM identification is found in West (18), while a complete
discussion of PLM identification procedures is found in McCone (19)
The following sections will provide additional detail about the
operation of PLM for screening of the sample and preliminary compound
identification.
2.1.1 Theory
Such physical characteristics as color, shape, reflectivity, surface
roughness, clarity, size and size distribution of small particles or
^
crystals can be observed under medium magnification in a conventional
optical microscope. In some cases, these data may be sufficient to iden-
tify a sample compound. If not, the use of polarized light microscopy
(PLM) can provide additional data which may permit direct identification,
particularly with crystalline materials. Crystals can be divided into
three clases (Figure 3) based on the number of refractive indices observed
under polarized light. Crystals with one refractive index (RI) will
remain dark when rotated under crossed polarized filters and are termed
anisotropic. Crystals with two RI's (uniaxial and isotropic) or three
RI's (biaxial anisotropic) appear to flash when rotated under crossed
polarized filters.
Determination of the RI of the crystal can be made via an immersion
technique with an error of ±0.001 units. This is often enough to permit
exact identification of the compound, particularly when a second and
possibly third RI are determined. Table 5 lists the RI's of a variety of
compounds which may be encountered in environmental analysis to illustrate
the range of RI's commonly encountered.
When a tentative identification has been made, chemical verification
can be performed under the microscope. In most cases, there are standard
wet chemical qualitative analysis (20) procedures performed on a single
crystal or small group of particles. For example, solubility in a
28
-------
CRUSH, GRIND
AND SIEVE
SOLID SAMPLE
SOLUBILITY
MELTING POINT
DENSITY HARDNESS
FLUORESCENCE
CHEMICAL
ANALYSIS
PHYSICAL
ANALYSIS
USING MEDIUM
POWER, NOTE-
CRYSTAL SHAPE
CLEAVAGE
INCLUSIONS COLOR
PLEOCHROISM
APPROX. Rl SIZE
Figure 3. Identification of Single Particles Using Polarized
Light Microscopy (PLM)
29
-------
25,
Table 5. Refractive Indices of Selected Crystals (Nd )
Biaxial Anisotropic
Crystal
CaS04-2H20 (gypsum)
CaS04 (anhydrite)
CaO-Al203=2Si02 (anorthite)
A1202 ^SiOg-HpO (pyrophyllite)
3MgO-4Si02-H20 (talc)
BaSO^ (barite)
CaMg(Si03)2 (diopside)
CaC03 (aragonite)
PbC03 (cerussite)
RI alpha
1.521
1.569
1.516
1.552
1.539
1.637
1.664
1.530
1.804
RI beta
1.523
1.575
1.583
1.588
1.589
1.638
1.671
1.681
2.706
RI gamma
1.530
1.613
1.589
1.600
1.589
1.649
1.694
1.685
2.078
particular solvent (e.g., water) may be determined by exposing the crystal
to vapors of the solvent. The rate at which it absorbs the vapor and
liquifies is an indicator of the degree of solubility. Specific tests for
lead, copper, silver, sulfate, chloride, fluoride and several other ions
can be performed by dissolving the crystal in a small droplet of water then
adding a droplet of reagent solution and observing whether or not a pre-
cipitate forms. Sensitivities on the order of 10 - 10 grams are
reported. In general, chemical tests are limited to confirmation of the
presence of a few ions and are not used as a general qualitative analysis
scheme. The three analysis steps, physical, optical and chemical, are
described briefly below.
2.1.2 Compound Identification - Generalized Procedure^
Sample Preparation
Solutions should be evaporated and the solid material crushed. Solid
samples can be crushed or milled to obtain particles of approximately
50pm diameter. Samples are mounted on microscope slides and held by any
conventional powder support (tacky Canada balsam, immersion oil, etc.).
30
-------
Liquid samples for chemical analysis should be placed in a slotted
microscope slide and covered with a cover slide.
Physical Analysis
Using medium power, note the crystal shape and number and orientation
of cleavage planes. Using these observations, assign a crystal habit if
possible (cubic, columnar, platy, spherulitic, needle-like, fibrous, etc.).
Note any inclusions of gas or solid in the particle or crystal. Using
ordinary white light, observe the color of the sample. Few common inorganic
compounds are colored and this is therefore an important observation. In
anisotropic substances, rotation of a beam of polarized light may cause the
color to vary with the orientation of the light plane, a phenomenon termed
pleochroism. Note whether the material is pleochroic. If so, more than
one refractive index will be observed. If the material has not been
crushed, observe and record the particle size by comparison with standard
grids or an ocular scale. If the particle or crystal has not been identi-
fied from these observations, proceed to optical analysis.
Optical Analysis
To determine whether the substance is isotropic or anisotropic,
polarizing light filters are placed between the light and sample and between
the sample and ocular. If the crystal (e.g., NaCl, cubic) remains dark when
rotated in the field of polarized light, it is isotropic, possessing only
one refractive index. Completion of the characterization of isotropic
substances requires only the determination of the refractive index. Aniso-
tropic crystals will change color or flash when rotated in the cross-
polarized beam. Anisotropic or birefringent crystals can be further divided
into two classes, which are analyzed separately (Figure 3).
One of the basic optical characteristics of birefringent crystals is
the so-called interference figure, which serves to subdivide the aniso-
tropic group into two classes, uniaxial and biaxial. One method for locat-
ing interference figures is to focus on the field, cross the prisms, remove
the ocular and move the slide until an interference figure appears. This
is usually the most rapid method when the particles are randomly oriented.
The interference figure for a uniaxial crystal appears as a dark Maltese
cross extending across the field. Concentric rings of colors (isochromatic
31
-------
curves) may also be present. The interference figure for a biaxial crystal
consists of hyperbolic curves which move in and out as the stage is rotated.
Light passing through uniaxial anisotropic (or doubly refractive)
crystals is broken into two rays traveling at different velocities. For
uniaxial crystals, these rays are known as the ordinary (o>) and extraor-
dinary (e) rays. The sign is determined by whichever ray has the greater
index, such that, e - u = birefringence, the sign of which is called the
"optic sign."
The determination of the refractive indices of a uniaxial crystal
requires that the crystal be properly oriented in polarized light; the
individual indices are then resolved and are determined in the same manner
as for an isotropic crystal. In general, uniaxial crystals show two
refractive indices when placed in positions of extinction between cross
prisms. The indices are measured focusing on a field of a few crystal
fragments using low intensity parallel light and a 4-rran or 8-mm objective.
The analyzer is inserted, a crystal showing maximum birefringence is selec-
ted and centered, and the stage is rotated until the crystal is dark
(extinction). The analyzer is then removed and the index is estimated by
comparison with the known index of the mounting medium using the Becke
line method. The analyzer is again inserted, the stage rotated to the
second position of extinction, and the other index is estimated. This
procedure is repeated using appropriate mounting media until at least two
or three checks are obtained for the higher and lower indices. If the
crystal is uniaxially positive, the lower index value will be that of the
ordinary ray, while if the crystal is negative, the higher value will be
that of the ordinary ray.
The identification of biaxial crystals by conoscopic methods
resembles closely the procedure described above for uniaxial crystals, the
main difference being that three refractive indices are determined. In
addition, biaxial crystals have another property, the optic angle, which can
be measured readily and used as an^aid in identification. When an inter-
ference figure is obtained which indicates that a crystal is biaxial, it
is used to determine the optical character as was done in the case of
uniaxial crystals. Further, it may be used for the estimation of the optic
32
-------
angle and the preliminary estimation of the intermediate index of refraction
(3). The final measurement to be made is the determination of the three
refractive indices.
Refractive Index
The determination of refractive indices for biaxial crystals is
complicated by the fact that there are three different indices. These
indices are represented by Np, Nm, and Ng, (petty, medium, great), or more
commonly by a, 3, and y. To determine these indices, their order of mag-
nitude is first estimated by the degree of relief in Shillaber's oil or
Canada balsam. The a and y indices are then determined by locating a crys-
tal of optimum birefringence, rotating to extinction, and determining the
index by means of the Becke line method (18). The crystal is then rotated
to the second position of maximum extinction and the second index determined.
By repeating this procedure with appropriate immersion liquids until the
highest and lowest indices have been established, the values for y and a
will be obtained. For a biaxial positive crystal, 3 will have a value
nearer the index. If it is biaxial negative, 3 will be nearer the a index.
It is of value to note that crystals having optic angles approaching zero
degrees will have a value approaching either a or y values, and crystals
with optic angles near 90° will have a 3 index almost half-way between the
other indices. The value of 3 is estimated from the values determined for
a and y, the optic sign, and the optic angle.
Chemical Analysis
Confirmation of the chemical composition of a sample particle or
crystal may be obtained by microchemical analysis. The crystal is dissolved.
by exposure to a saturated vapor of the solvent, usually water. These tests,
called Chamot-type precipitation tests can be found in detail in refer-
ence 20.
2.1.3 Typical Results from PLM
As part of a Comprehensive Assessment of an industrial boiler equipped
with a wet scrubber and capable of being oil- or coal-fired, PLM analyses
were performed on samples taken at the inlet and outlet of the wet scrubber.
33
-------
The samples from both an oil firing and coal firing boiler consisted of an
EPA M-5 train with a cyclone and filter at the inlet and an EPA M-5 train
with only a filter at the outlet. During coal fired tests, flyash samples
from a mechanical precipitator were also taken and analyzed.
Table 6 shows the type of results possible from PLM. Information on
weight, modal diameter and size range was obtained from most of the samples.
Only outlet filters were not completely analyzed because of the low amount
of material and their hydroscopic nature.
Particles found in most samples were flyash, partially fused flyash,
oil soot, and iron oxide (hematite and magnetite). Traces of quartz,
shards, coke, and calcite were found in many samples. Both scrubber cake
samples contained calcium sulfite hemi-hydrate (CaS03«l/2H20) which was the
principal component of sample 202-4-scrubber cake. The presence of an
unknown sulfate in the outlet filter during the coal fired tests was a
strong indication that the scrubber was adding scrubber liquor reactive
products (NaHS03) to the outlet mass loading, and this was later confirmed
from the elemental FTIR and XRD studies.
Regardless of whether coal (series 201) or oil (series 202) was used
to fire the boiler, oil soot was found in the samples collected. The oil
soot found during the oil tests was largely in the form of complete ceno-
spheres with smooth, unbroken walls. Oil soot from sample 201-1-flyash
which is representative of the oil soot in all the 201 series samples
appears to be broken, abraded, and has a grainy surface texture. The more
worn appearance of the 201 series oil soot indicates that it was probably
oil soot retained in the ducts from some earlier oil combustion. Knowing
this information, chemical analysis of the coal samples would be corrected
for the affect of the oil soot.
34
-------
Table 6. PLM Analyses Results
COAL FIRED BOILER SAMPLES
Components
Partially fused flyash
Flyash
Oil soot
Magnetite
Iron oxide
Coke
Quartz
Calcite
CaS03-l/2HzO
Unknown sulfate
201-1-Flyash
A
65-80%
1-5%
5-15%
10-20%
2-5%
1-5%
<2%
B
30
4
25
12
15
40
12
C
5-140
1-20
2-100
3-45
1-40
5-160
5-60
201-1-1-Cyclone
A
55-70%
10-20%
10-20%
10-20%
1-5%
<2%
<2.%
B
25
4
15
12
7
60
20
C
5-65
1-16
2-80
3-25
<1-50
6-100
5-40
201-1-1-Filter
A
40-55%
35-50%
10-25%
<2%
<2%
B
12
2
8
5
5
C
5-40
<1-13
1-40
2-14
1-18
201-1-Scrubber Cake
A
55-70%
10-20%
10-20%
1-5%
<2%
<1%
10-20%
-
B
20
5
15
10
3
6
6
(length)
C
4-60
1-15
1-60
5-45
1-21
1-10
3-21
(length)
201-1-0-
Filter
A
25-40%
15-25%
.
1-5%
50-65%
OIL FIRED BOILER SAMPLES
Components
Partially fused flyash
Flyash
Oil soot
Magnetite
Iron oxides
Unknown sulfate
CaS03-l/2H20
Calcite
Water droplets
201-4- I -Cyclone
A
25-35%
1-5%
50-75%
<2%
<1%
1-5%
B
20
3
20
12
8
C
6-40
<1-15
<1-140
3-50
3-35
202-4-I-Filter
A
10-20%
15-30%
45-60%
1-5%
15-30%
202-4-Scrubber Cake
A
1-5%
95%+
B
30
30
C
1-80
2-80
202-4-0-
Filter
A
5-10%
50-65%
30-45%
CO
in
Key: A - Estimated weight percent
B - Estimated modal diameter (\an)
C - Size range
-------
2.2 THERMAL ANALYSIS
This section describes a number of methods which measure some property
of the material as a function of temperature. The primary methods discussed
will be Thermogravimetric Analysis (TGA) and Differential Scanning Calo-
rimetry (22-29)/(DSC). These tests will primarily be used to determine the
thermal stability of the sample so that a safe drying temperature can be
selected.
2.2.1 Principles of Operation
Thermal analysis is a general term for techniques which measure
physical changes in material occurring upon a change in temperature.
Common physical changes which can be monitored include changes in weight,
length or specific heat of a sample.
The most frequently used measurements can be categorized into the
six techniques listed in Table 7. Of these, the first three are the
principal techniques to be used in a Level 2 analyses scheme.
Table 7. Thermal Analysis Techniques
Method
Function
Temperature
Range(°C)
Differential Scanning
Calorimetry (DSC)
Thermogravimetri c
Analysis (TGA)
Differential Thermal
Analysis (DTA)
Thermomechanical
Analysis (TMA)
Evolved Gas Analysis
(EGA)
Dynamic Mechanical
Analysis (DMA)
Measures heat flow into or out
of a sample
Measures sample weight changes
Measures temperature excursions
of a sample
Measures sample dimensional
changes
Measures specific evolved sample
decomposition products
Measures sample modulus and
damping changes
-180-725
Ambient to
1200
-180-1600
-160-1200
Variable by
technique
-150-500
36
-------
TGA is concerned with the weight change of a sample as the temperature
is increased in a prescribed manner. Differential thermal analysis involves
the measurement of changes in heat content as a function of the temperature
difference between the sample and a thermally inert reference material as
both substances are heated, or cooled, at identical, well-controlled rates.
The results of such an analysis allow a measurement of melting points,
vaporization points, crystal phase transitions, chemical changes and other
enthalpic changes. The weight changes accompanying any of these changes
are determined by TGA.
The instrumentation for a TGA consists of a precision balance, a
furnace which can be programmed for reproducible temperature changes over a
specified time period and a recorder. There is generally a reaction chamber
in the furnace which enables an analysis to be performed in atmospheres other
than ambient air, e.g., inert, or under vacuum. Either a null-point or a
deflection type of balance is acceptable. The recording system should pro-
vide a continuous record of weight and temperature to ensure a complete
description of thermogram features. Furnace design and controls must be
capable of providing a satisfactory, smooth input so that either a linear
heating program (typically 10° - 600°C/hr.) or a fixed temperature can be
maintained over an operating range up to at least 1100°C.
A DTA contains four basic elements: 1) a sample holder with a dual
thermocouple assembly, 2) a furnace and its controller, 3) a flow control
system and 4) a recorder. As with the TGA, the furnace must be able to
be controlled to provide a reproducible, constant heating or cooling rate,
over a range of about 0° to 30°C/min. Again a desirable maximum operating
temperature is about 1000°C.
Differential scanning calorimetry differs from DTA in the respect that
the difference in power required to maintain the sample and the inert
reference material at the same temperatures as they are heated or cooled is
measured. The DSC contains two separately heated sample holders—for the
sample and the reference. A power difference will be observed whenever
physical or chemical changes occur in the sample: this power difference
represents the thermal energy absorbed, or released, as a consequence of the
changes in the sample. The temperature of each is measured with resistance
37
-------
thermometers. Here too a means of assuring a steady, reproducible,
well-controlled temperature change is essential.
/
2.2.2 Applications and Methodology
In the Level 2 analysis thermogravimetry will be most often used to
determine the drying ranges of materials and the most suitable drying
conditions to be used before weighing samples and proceeding with other
analyses. The most useful information will be obtained by using either a
slow heating rate or several different heating rates in an inert
atmosphere.
A sample thermogram of the changes calcium oxalate monohydrate under-
goes during heating is shown in Figure A".
The plateaus on the TGA graph indicate the sample is maintaining a
constant weight; that is, the sample is stable in that particular tempera-
ture range. Inflections represent such physical or chemical changes as
water or solvent loss, formation of another compound, or sorption of
volatile materials in the sample.
I
•5
11.0 mg
900* 1000"
Temperature, °C
Figure 4. Thermogravimetric Evaluation of Calcium Oxalate
Monohydrate; Heating Rate 6°C/min.
38
-------
In interpreting data from a T6A or DTA analysis, consideration must
be given to the apparatus and to the procedure used. Variations due to
the dynamic nature of the method are very possible. In particular, there
is often little correlation between results from isothermal (run at single
temperature) runs and non-isothermal (temperature scan) runs.
The shape of the T6A or DTA curve is affected by the heating rate,
the nature of the sample and its container and the atmosphere in which the
analysis was done. At a given temperature, the extent of change in the
sample varies inversely with the heating rate. A slow heating rate further
helps to differentiate rapid, successive changes. A small, finely divided
sample will behave more uniformly and reproducibly.
This result is demonstrated in Figures 5 and 6, which show the DSC of
>3 ym material collected at the outlet of a limestone wet scrubber.
The Figure 5 run at 10°C/minute shows three large endothermic peaks
at 87°C, 120°C, and 190°C with several smaller shoulders at 140°C and
230°C. The major peaks roughly correspond to the dehydration of CaS03«
1/2H20 (100°C), CaS04'2H20 (to the half hydrate - 128°C) and CaS04-l/2H20
(163°C).
An attempt was made to resolve these shoulders by reducing the amount
of sample and greatly reducing the heating rate. A heating rate which is
too high or too large a sample would prevent rapid equilibration of the
sample and would cause the response to changing temperatures to lag. This
fact is illustrated by the improved resolution obtained in Figure 6. The
scan is quite different, showing peaks at 47°C, 94°C, 160°C, and shoulders
at 115°C and 175°C. The 47°C peak is probably due to surface water and
baseline drift. Clearly, the 94°C and the 160°C peaks correspond to the
CaS03. 1/2H20 -> CaS03 and the CaS04 • 1/2H20 + CaS04 dehydrations. The
shoulder at 115°C might be due to the dehydration of CaS04 '2H2Q. The
175°C shoulder is unidentified, but may represent a phase change of CaS04
between a and 3 forms or a reaction of an unidentified material.
Once a compound is tentatively identified, it is possible to use the
peak area under a DSC curve to calculate the AH of the reaction and use it
to identify the compound present. Conversely, if the compound is known,
then the AH can be used to quantitate the compound. Assuming that the
39
-------
-p.
o
U 1UU T U <=!UU
TEMPERATURE. °C CCHROMEL/ALUMELJ
Figure 5. DSC of >3 urn Material Collected at the Outlet of an F6D Run at 10°C/min.
-------
25O 3QO 35O 4OO
TFMPFP1ATI IRF Tt C HHBOMFI /AU JMFI 1
5OO
Figure 6. DSC of >3 ym Material Collected at the Outlet of an F6D Run
at 2°C/min and Reduced Sample Size.
-------
94°C and 160°C peaks are due to dehydration of CaSOg-l/ZHgO and CaS04*
1/2H20, respectively, it is possible to calculate the material present if
the AH of the dehydration is known. The formula for this calculation is:
A (60 BE Aqs)
m = AH
where
A = Peak area, (sq. in.)
m = Sample mass, (rug.)
B = TIME BASE setting, (min/in.)
E = Cell calibration coefficient at the temperature of the
experiment (dimensionless).
Aqs = Y-axis RANGE, ((meal/sec.)/in.)
AH = mcal/mg
If the AH for the identified reaction is available, an estimate of the
amount of material present from the DSC will provide a valuable piece of
information to aid in the selection of a compound identification method.
42
-------
2.3 ANION ANALYSIS
The analysis of samples for anions is a necessary first step to
quantitate the compounds that might be present. The anion data and the
elemental analysis data can be compared on a mole basis to determine if
the elements and anions make a match.
Anion data will also expand or modify the potential compound list
developed in Chapter 1. The recommended method of analysis for most
samples is ion chromatography on the basis of sensitivity, accuracy, and
flexibility. Included in this manual are specific anion tests for all the
anions expected in environmental samples. The following sections will
describe the use of 1C for anion analysis.
2.3.1 Specific Tests
Anions are commonly analyzed individually using wet chemical proced-
ures. Sample preparation procedures must be selected which do not
contaminate the sample prior to analysis. Samples dissolved in aqua regia,
for example, cannot be analyzed for nitrate or chloride. In general,
aqueous samples may be analyzed directly and most anionic species are water
soluble in their common forms (except oxides and most phosphates). In
Appendix A, procedures are given for the individual analysis of aqueous
samples, bromide, iodide, carbonate, bicarbonate, cynaide, fluoride,
nitrate, nitrite, orthophosphate, sulfide, sulfite, sulfate, and ammonia.
Most of these quantitative tests have been taken from the EPA "Manual of
Methods for Chemical Analysis of Water and Wastes". A procedure for the
analysis of samples for ammonia is included in this section although this
species is usually present as a cation in aqueous media. Sample collection
and preparation procedures, detection limits, and potential interferences
are given.
2.3.2 Ion Chromatography
The chromatographic separation of anionic (and cationic) components
of a sample has been refined recently (30,31). Ion chromatography (1C)
instruments are commercially available which can analyze for the following
anionic species: acetate, arsenate, bromide, carbonate, chloride, fluoride,
iodide, nitrate, nitrite, orthophosphate, sulfate, sulfite (and ammonium).
Typical parameters for analysis of aqueous sample for these species by 1C
are given in the following section.
43
-------
1C Theory
Conventional ion exchange chromatography (32) involves the separation
of ions based on the differences in their rates of exchange with similarly
charged ions bonded to stationary ions of the opposite charge on a resin.
Sample ions of interest are placed at the head of the column and eluted
with a mobile phase, or eluent, which contains similarly charged ions. The
function group, at the active site of a resin, is always opposite in charge
to the exchanging ions. The rate at which the ions exchange positions is
determined by the ion's attraction to the functional group. The physical
characteristics of size and charge determine the ion's attractions or
interaction with the functional group or resin. Ions which have a charge
that can be polarized toward the functional group will have a stronger
interaction with the resin, thus a slower rate of exchange.
When ions elute from the resin, they are in a background of the
mobile phase ions. Modern ion chromatography is a process which employs
an anion or cation ion-exchange column followed by a second acid or base
ion-exchange column. The first column is called the separator and is
usually a low capacity column. The second column, called the suppressor,
neutralizes the mobile phase and is usually a high capacity column. The
neutralization of the mobile phase ions allows the eluting sample ions to
be detected by simple conductivity. Ion chromatography can be applied to
the analysis of samples for acetate, bromide, chloride, fluoride, iodide,
nitrate, nitrite, orthophosphate, sulfate, sulfite, (and ammonia, as well
as other cations).
Sample Preparation
The sample is extracted with hot distilled water extraction to remove
all of the soluble anions. The same sample is then extracted with a 0.1N
HNOo solution to solubilize the remaining anions. This approach will
= _ _ _ _3
solubilize most of the common anions (Cl , SO^ , NOj , Br , I , PO. ), but
the analyst must be aware of potential problems with species like SCL~ and
C0g~. Species like CaSOg would be insoluble in water and would oxidize or
volatilize in HNOg. If C03 or S03~ are suspected to be present in the solid
then gas evoluation by HC1 addition, followed by a dilute NaOH trap is
recommended. In both cases air must be excluded to prevent $03" oxidation
44
-------
and C02 absorption. Both anions can be run in the NaOH solution using
the columns shown in Table 8.
Application of Ion Chromatography to Environmental Analysis
Samples from the venturi/spray tower (limestone) scrubber at the
EPA/TVA Shawnee Test Facility in Paducah, Kentucky and the Arthur D. Little
Alkali Pilot Scrubber were analyzed by ion chromatography.
The anions of interest for both dual alkali and lime/limestone wet
scrubbers are: sulfite, sulfate, carbonate, chloride and nitrite. The
anion analysis procedure was developed by TRW based on the Dionex instruction
manual but modified (33) to optimize the separation and response in the
matrices encountered at lime/limestone wet scrubbers. Table 8 summarizes
the analytical conditions needed for analysis of the cations and anions
found in wet scrubbers.
The Detection Limit (DL) for sulfite, sulfate and nitrite is one ppm;
carbonate is 5 ppm whereas chloride could vary between 0.1 and 10 ppm
depending on conditions. In the chloride analysis, the DL improves if
eluent is added to the sample before analysis. The addition of eluent to
the sample suppresses the water dip on the ion chromatogram which occurs
immediately before the chloride peak. Carbonate usually has a high DL due
to carbonate found in the blanks.
*
The analytical scheme is based on the retention time for sulfate which
is the last species of interest to elute from the columns. The conditions
stated below will cause sulfate to elute in 15 to 17 minutes, but the
sulfite and nitrate peak will be only partially resolved. Conditions can
be changed so that sulfite and nitrate will be completely resolved, but the
retention time of sulfate increases to about 45 minutes. Wet scrubber
samples have little nitrate, however, and the incomplete sulfite and nitrate
resolution is not a problem. Because sulfite is constantly oxidizing to
sulfate, the shorter analysis time is preferred.
Quantitative analysis may be performed using either peak height or
peak area; however, a series of experiments have shown that peak height is
better than peak area due to variations in peak shape and detector response
45
-------
Table 8. Retention Time of Various Ions Found in Wet Scrubbers
Species
Cl"
S03
NO-
S0~
C03
Ca+2
Mg+2
Na+
K+
Cone.
10 ppm
10 ppm
10 ppm
10 ppm
10 ppm
10 ppm
10 ppm
10 ppm
10 ppm
Retention
Time
4.5 min
10
8
16
6
12
8
7
12
Dionex
Separator
Col umn
An ion
3 x 500
An ion
3 x 500
An ion
3 x 500
An ion
3 x 500
Bio Rad
6 x 500
Cation
6 x 250
*
Cation
6 x 250
Cation
6 x 250
Cation
6 x 250
Dionex
Suppressor
Col umn
An ion
6 x 250
Anion
3 x 500
Anion
3 x 500
Anion
3 x 500
None
Cation
9 x 250
Cation
9 x 250
Cation
9 x 250
Cation
9 x 250
Flow Rate
30%
30%
30%
30%
30%
40%
40%
40%
40%
Elluent
0.003M NaHCOs/
0.015M Na2C03
0.003M NaHCOs/
0.025M Na2C03
0.003M NaHCO-/
0.025M Na2C03
0.003M NaHCO,/
0.025M Na2C03
Water
for both
cations
0.001 M
p-phenylene-
diamine
dihydrochloride
0.005N HN03
0.005N HN03
Temp.
24°C
24°C
24°C
24°C
24°C
24°C
24°C
24°C
24°C
a*
-------
characteristics. Figure 7 is an ion chromatogram of a solution which
contains F', Cl~, NO^, P04'3, Br", NO^, and SO^2. Each ion has its own
basic shape/ F" and Cl~ are sharp spikes where as S04~2 is a broad Gaussian.
Calibration curves are constructed by running standard solutions of known
ion concentration on the ion chromatograph and determining peak heights.
A plot of sample concentration vs peak height is constructed. Figure 8
shows that the linear range for Cl" ions is from less than 1 ppm to 10 ppm.
The calculation of the slope of the curve is determined by a "least
squares" fit of the data points.
The accuracy and precision of the analytical scheme was tested using
samples from the lime/limestone wet scrubber at the EPA/TVA Shawnee Test
Facility in Paducah, Kentucky. Table 9 shows a direct comparison between
TVA and TRW sample analyses of lime/limestone wet scrubber samples for
-2
Cl and SO^ . The relative error for each set of values is listed as well
as results of an alternate analysis employed as a check. The chloride ion
values were very consistent for both analytical techniques employed.
Because of most chlorides solubility, those compounds would be less
susceptible to changes in ph, and as a result would not be expected to vary.
The data confirmed that as the relative differences between the reported
values ranged from 1.8% to 4.5%.
2.4 BULK ELEMENTAL ANALYSIS
The results from the Level 1 SSMS will provide specific information
on the number of elements present in the sample and approximate (factor
of 2-3) concentrations. Elements exceeding their MATE values will be
analyzed more accurately. The choice of the final Level 2 elemental analy-
sis method will be left to the individual analyst. This decision will be
influenced primarily by the in-house or local vendor capabilities. The
following sections will provide an overview of elemental analysis procedures
with the emphasis on understanding their use and limitations. The methods
discussed are: Atomic Absorption Spectroscopy, Atomic Fluorescence
Spectroscopy, Neutron Activation Analysis, X-Ray Fluorescence and Proton
Induced X-Ray Emission.
47
-------
COLUMNS
ELUENT
INJECTIO
POINT
DIONEX AN ION
DIONEX ANION
SEPARATOR
SUPPRESSOR
3 X 500 mm
6 X 250 mm
0.003 M NaHCO3/0.0024 M Na2CO3 AT
30% FLOW
Figure 7. Ion Chromatogram of the Dionex Anion Standard Solution
48
-------
Table 9. Comparison of SO. and Cl Analysis Methods
Species
Sulfate
Chloride
Method of Analysis
Ion Exchange (TVA)
Ion Chromatography (TRW)
Potenti ometric (TVA)
Ion Chromatography (TRW)
#5253
Cone.
32994
34000
2676
2726
Rel.
Error
3%
2%
#5254
Cone.
17584
16209
1418
1454
Rel.
Error
8%
3%
#5255
Cone.
17700
16660
1418
1485
Rel.
Error
6%
5%
#5256
Cone.
31442
34404
2614
2675
Rel.
Error
9%
2%
10
-------
26
24
22
20
18
16
14
12
10
8
6
4
2
0
C1~. SLOPE-2.59
y-INTERCEPT - -0.15 CM
CONDUCTIVITY SCALE - 10X
6
8
9
10
Figure 8. Calibration Curve for Chloride Ion Analysis
50
-------
2.4.1 Introduction-Atomic Techniques
The electronic structure of each element is unique, and only specific
energies will cause excitation of an electron from the atomic ground state
to a higher energy level. On return to the ground state, the excess energy
is given off at an energy which again is specific for that element. Sev-
eral analytical techniques have been developed which exploit this elemental
specificity. These are schematically shown in Figure 9. Atomic absorp-
tion spectroscopy (AAS) uses the technique of measuring the absorption of
light from a monochromatic source (hollow cathode lamp) by a particular
element or measuring the absorption at a given wavelength from the spectrum
of a continuum source (Xe lamp or Hg hollow cathode lamp). Atomic
fluorescence spectroscopy measures the light emitted from an element return-
ing to its ground state after excitation by a hollow cathode lamp or
continuum source. Atomic emission spectroscopy uses the emission produced
as electrons return to the ground state following thermal excitation by a
flame or plasma. Each of the techniques is applicable to approximately
60 elements. Their advantages and limitations will be discussed below
and the methods compared as to their sensitivity and applicability to
multi-element analysis. Methods of preparing samples for analysis by
these techniques are presented in Section 2.5.4.
2.4.1.1 Atomic Absorption (AAS)
The analysis of a sample by AAS requires the following steps: intro-
duction of the sample, evaporation of the solvent, atomization (conversion
to the atomic ground state) and measurement of the absorption of light.
Two techniques exist for the first three steps. Conventional AAS aspirates
1st Excited
State
f Ground _
State
Atomic Atomic Atomic
Absorption Fluorescence Emission
Figure 9. Schematic of Atomic Techniques
51
-------
the aqueous sample into a flame (FAAS) where thermal energy evaporates the
solvent and decomposes compounds to atoms. In the so-called "flameless"
techniques (NFAAS), an aliquot of the sample is placed in a furnace, the
solvent removed by heating at low temperature, and the sample subsequently
atomized by rapid heating of the furnace. The light source for either sys-
tem is commonly a hollow cathode lamp whose cathode is constructed of the
element(s) to be analyzed and filled with an inert gas (Ne or He) at low
pressure. As current is passed through the cathode, atomic emission lines
are produced, characteristic of the composition of the cathode, further
enhancing specificity. The absorption of this light by the element in
the sample follows Beer's Law over a given concentration range. Mono-
chromators are used to select the analytical wavelength for measurement by
a photomultipiier/amplification system. The signal from the hollow cathode
is usually chopped or modulated and the detector synchronized to the chopped
signal. In this way the DC signal due to other emissions (the flame itself
or molecular emission from compounds formed in the heating chamber) can be
electronically suppressed.
Potential Problems in NFAAS and FAAS
FAAS requires larger sample volumes per analysis (1 to 5 ml) than NFAAS
(0.1 to 0.5 ml). Furthermore, factors affecting viscosity of the sample
(and thus the flow rate of the sample into the flame) such as the presence
of organic solvents or high solid content require special attention when
standards are prepared for FAAS analysis. The method of standard addi-
tions is recommended for most analyses of complex matrices by both FAAS
and NFAAS. The presence of condensed vapors or smoke from residual organic
material drastically reduces the precision of NFAAS analysis. Therefore,
following evaporation of the solvent, samples must be ashed in the furnace
prior to atomization. Most NFAAS instruments are equipped with program-
mable furnaces which include an ashing step. Care must be taken when NFAAS
is used on samples which contain the volatile elements (As, Se, Sb, Hg, Cd,
and Pb). Hydride evolution techniques (34,35,36) are available for As,
Se, and Sb as well as a cold vapor technique for Hg. NFAAS of As, Se, and
Se has been accomplished using Ni(N03)2 addition (34) to form the arsenides,
selem'des and antimonides, which are non-volatile at ashing temperatures.
52
-------
The efficiency of atomization is reduced in some cases by ionization
of the analyte to a state other than its atomic ground state. This can be
overcome by the addition of a substance with preferentially ionizes. For-
mation of compounds (oxides, phosphates) which do not decompose to atoms in
the flame or furnace also decreases sensitivity. Use of higher tempera-
tures can often overcome the problem (nitrous oxide/acetylene flames in
FAAS) as will the addition to the sample of a substance which will inhibit
oxide or phosphate formation (lanthanum, for example). In some cases,
spectral interferences can occur. For example, analysis for small amounts
of the transition metals in an iron matrix would require pretreatment of
the sample. In other cases, using the method of standard addition may be
sufficient.
Application to Multielement Analysis
The spatial requirement that the light source and detector be placed
at 180° reduces the applicability of single element hollow cathode lamps
in both flame and flameless AAS. True simultaneous analysis exists for
only a limited number of elements (usually two) where multiple source/
detector pairs are focused on the sample chamber. Multi-element (2 to 4)
hollow cathode lamps can be used to analyze samples sequentially by use of
a scanning monochromator and a single detector. The intensity and lifetime
of these lamps is usually less than that of single element lamps.
Use of continuum sources to excite several elements which are analyzed
by programmed scanning has been successful to a limited extent. The power
output of these lamps (Deuterium hollow cathode lamps and to a lesser
extent Xenon arcs) is significantly lower than the elemental hollow cathode
lamps, thus their sensitivity is higher. Using a continuum source for
FAAS or NFAA, excitation parameters must be set at the lowest common
denominator for the elements being observed, further reducing the sensitiv-
ity for the other elements. For the above reasons, it is recommended that
AAS not be used for multi-element analysis. AAS can best be used for
selected elemental analysis requiring high accuracy (±10%) or when a limited
number of elements is analyzed.
53
-------
2.4.1.2 Atomic Fluorescence (AFS)
The instrumental requirements for AFS are essentially those of AAS,
usually flame, with the difference that the fluorescence signal is measured
at right angles to the excitation path. For spatial reasons, AFS is more
conducive to multi-element analysis since detectors can be placed anywhere
along a plane perpendicular to the incident beam. Thus use of multiple
monochromator/detector systems permits true simultaneous analysis. Further-
more as with all fluorescence techniques the intensity of the emission is
proportional to the intensity of the incident beam. Use of electrodeless
discharge lamps (EDL's), which have greater output than hollow cathode
lamps, increases sensitivity in fluorescence systems. The applicability
to multi-element analysis is restricted by the same considerations as
discussed for AAS, namely, the need for an excitation source for each
element or a lower power continuum source. Potential problems in analysis
by AFS are the same as those discussed for AAS, and, in addition, commercial
units are not readily available.
2.4.1.3 Flame Atomic Emission Spectroscopy
In atomic emission spectroscopy the excitation source is the flame
itself. Thus, the principal drawback of source/detector orientation in
AAS and AFS is eliminated. Since the emission occurs in all
directions, a theoretically unlimited number of single element monochromator/
detector systems could be configured around the flame. In practice, how-
ever, it is more common to use a single detector and scan the emission
spectrum with a monochromator. Programmable scanners allow preselection
of given wavelengths for study and the monochromator is rapidly moved
between the chosen "windows." Thus, selectivity and multiple element
capability exist in AES. Another advantage of emission techniques is that
they are linear over several orders of magnitude in concentration (up to
the limit where self absorption occurs) whereas fluorescence techniques
are linear over about three and absorption methods to a limit of about two
orders of magnitude concentration change.
Its principal drawback, however, has been the lack of sensitivity
due to the high background emission of the flame itself. Broadband
emissions make the flame spectrum complex and variations in the flame
54
-------
introduce severe noise problems not present in AAS. Use of flames which
have lower backgrounds improves the sensitivity of the technique. The
most promising approach to solving this problem, however, has been the use
of non-combustion flames, or plasmas.
Interferences in AES
As with other atomic techniques, sensitivity of the method is affected
by formation of refractory oxides, electron transfer reactions due to the
presence of easily ionizable elements (the alkali metals), factors which
affect viscosity of the sample, changes in flame composition or temperature.
The method of standard additions is recommended for most analyses.
2.4.1.4 Inductively Coupled Argon Plasma Spectroscopy (ICAP)
Inductively coupled argon plasma spectroscopy (37,38,39,40,41,42) is
essentially a "flame" emission technique where the excitation source is a
plasma torch whose temperature is on the order of 7000°K. An inductively
coupled plasma is maintained by a high-frequency, axial magnetic field in
a laminar flow of argon at atmospheric pressure. The discharge gas in the
induction-coupled plasma is pure argon, which reduces background signal
significantly. In order to maintain very high temperatures without wall
contamination, a laminar flow of cold argon surrounds the plasma.
The atomization system usually consists of a pneumatic nebulizer which
allows the introduction of solutions of various concentrations and of any
degree of acidity or viscosity. Organic solvents may be employed if the
sample container is made of an appropriate material. The argon flow rate
is the primary factor controlling the rate of addition of aerosol to the
plasma. Droplet size of the aerosol will be affected by viscosity, surface
tension, and sample depth.
Signals are monitored with dual channel monochromators, one set to
monitor atomic emissions and the other to monitor plasma emissions to cor-
rect for background. A commercial instrument using a single light disper-
sion device (grating) and a mask which focuses set emission lines onto
40 dedicated photomultipliers is available.
55
-------
Interferences in ICAP
Inter-element effects can be minimized by proper selection of plasma
geometry. Since the elements in the sample are in an essentially inert
atmosphere, interferences from chemical reactions should be minimal.
Refractory oxides or phosphates can still form and reduce sensitivity as
with all techniques requiring atomic ground states. Electron transfer
reactions can occur in the presence of easily ionizable substances (the
alkali metals), reducing sensitivity for some metals. Matrix effects are
minimal when proper background correction, mentioned above, is employed.
Memory effects are problems inherent in the instrument. Because of the
high temperature of the plasma torch, an analyte may adsorb on the wall
of the sample tube tip when the plasma gas is turned off. This memory
effect can be reduced by cleaning the sample tube tip with HNCL solution
at the beginning of a new start-up. A second memory effect, located in
the nebulizer, is caused by the analyte solution creeping down both the
gas and sample needle and away from the tip. The analyte solution cling-
ing to the outside of the needles may be gradually dislodged as it is
replaced by the succeeding blank solution causing a positive interference.
Another problem is that many elements have complicated emission spectra.
For example, Fe has an extremely rich emission spectra under the conditions
that exist in the plasma. It is quite possible that the emission lines
chosen for the other elements lie very close to an Fe emission line. If
the monochromator cannot distinguish between the two lines, a spectral
interference results and the instrument gives a reading that is too high
because of the presence of the iron emission line. Spectra inferences can
be further compounded by the fact that most commercial ICAP units use fixed
photomultipliers which limits the ability to select alternate lines to
avoid spectral overlap.
2.4.1.5 Comparison of the Methods
Minimum detectable quantities (detection limits) can be used as an
indication of the applicability of a method to a particular analysis.
Table 10 lists order-of-magnitude detection limits for FAAS, NFAAS, and
ICAP. As a screening method ICAP offers the best combination of sensitivity
and multielement capability. For the exact determination of a particular
element, atomic absorption is the method of choice owing to the availability
56
-------
Table 10. Comparative Detection Limits (ppb)
Element
Al
As
B
Cd
Co
Cr
Cu
Fe
Mn
Mo
Ni
P
Pb
Pt
Se
Si
Ti
U
V
Zn
ICP10
10
15
4^
1
2
^\
2
I
0.5
5
5
30
15
20
15
10
1
75
2
1
Flame AA
20
100
1000
1
5
3
2
5
3
10
8
c
10b
10
50
100
60
50
7000
20
0.6
Furnace AA
0.004
0.06
•
0.008
0.03
0.005
0.008
0.003
0.004
0.06
0.02
3.0
0.03
0.45
0.10
0.10
0.30
0.15
0.0007
57
-------
of instruments, their specificity and sensitivity. The methods are simlar
in sample preparation requirements and potential interferences. Due to the
greater amount of work in the AAS area, this method is the method of choice
for single element analysis.
2.4.1.6 Application of AAS and ICAP to Environmental Analysis
As part of an environmental assessment (43) conducted on samples from-
a Fluidized Bed Combustion (FBC), a comparison of AAS and ICAP recoveries
was conducted. Weighed particulate samples from the fluidized bed combustor
were transferred to Teflon-lined Parr digestion bombs and digested overnight
at 130°C with 5 ml of aqua regia. The resulting solutions were filtered
through Whatman No. 41 filter paper. The collected residue was ignited,
fused with 1 g of high purity sodium carbonate Na2C03> dissolved, combined
with the original filtrate and diluted to 50 ml. Leachate samples were
prepared through 24 hour refluxing extractions in a solution of hydrochloric
acid at pH 4 and a basic colution of ammonium hydroxide at pH 9. Undis-
solved particulates were removed by filtration and were not recovered for
further analyses. The AAS instrument employed for this analysis was a
Jarrell-Ash 810 equipped with an FLA-10 Graphite Tube Furnace. The pro-
cedures listed in Table 11 were chosen to meet the Level 2 accuracy
requirements of ±15%.
ICAP analyses were conducted using an Applied Research Laboratories
(ARL) prototype instrument, Model QA-137, on all the leachate solutions.
In general, the detection limits are on the order of 0.01 vg/mL, although
they vary from element to element and from matrix to matrix. Standard
additions of Al, Ca, Cd, Cr, Co, Cu, Fe, Pb, Mg, Mn, Mo, Ni, K and Ba
were made to each leachate to assess matrix effects and recovery.
Calibration of the instrument was accomplished using reagent salts diluted
with deionized water.
Tables 12 and 13 summarize the spike recovery data for AAS and ICAP
respectively. Both methods showed excellent recoveries from both the
acidic and basic leachate solutions. ICAP has the added benfit of analyz-
ing all the elements reported at a cost less than one twentieth of the AAS.
58
-------
Table 11. Atomic Absorption Analytical Operating Parameters
Analytical ,
Element Wavelength. A
Ag
Cu
N1
•
L1
Mg
Pb
Zn
Hg
Cd
Be
Cr
Pt
V
SI
Sr
3281
*
3247
2320
6708
2852
2833
2139
2536
2288
2349
3579
2659
3185
2516
4607
Background , Silt .
Wavelength, A Width, A
3235 (Ne)
None
2316
None
None
2825
None
None
None
None
None
Continuum
3196 '
None
None
10
10
4
2
4
4
4
10
4
4
4
4
2
2
2
Atomization
Source
G.F.*
G.F.*
6.F.*
A1r-C2H2
A1r-C2H2
6.F.*
A1r-C2H2
Flameless
A1r-C2H2 .
N20-C2H2
A1r-C2H2
6.F.*
6.F.*
N20-C2H2
A1r-C2H2
Analytical Procedure
Dry - 30 sec. 0 200°C
Ash - 60 sec. 9 400°C
Atomize - 7 sec. 9 2300°C -
Argon Gas
Dry - 30 sec. 0 200°C
Ash - 60 sec. 0 600<>C
Atomize - 6 sec. 0 2500° C -
Argon Gas
Dry - 20 sec. § 200°C
Ash - 30 sec. 0 800°C
Atomize - 6 sec. 0 2800°C -
Argon Gas
Atomic Emission
Atomic Absorption
Dry - 40 sec. 0 200°C
Ash - 40 sec. 0 600°C „
Atomize - 6 sec. 0 1900°C
10/al 1.0% HN03 & 0.2%
(NH4J6Mo7024 added In furnace
Atomic Absorption
SnCl2 reduction N2 Sparge 1 1/min
Atomic Absorption
Atomic Absorption
Atomic Absorption
Dry - 30 sec. 0 300°C
Ash - 30 sec. 0 1150°C „
Atomize - 5 sec. 0 2700°C -
Argon Gas
Dry - 30 sec. 0 200°C
Ash -40 sec. 0 1000°C .
Atomize - 5 sec. 0 2800°C -
Argon Gas
Atomic Absorption
Atomic Emission
-
Graphite furnace.
59
-------
Table 12. Percent Recovery of Spiked Samples
by Atomic Absorption
Element
Zn
Mg
Ag
Ni
Li
Cu
Cd
Be
Cr
Pb
Pt
V
Si
Hg
Sr
Solutions
Leachates
Acid
104
-
80
60
104
80
-
100
-
105, 97
100
113
96
100
75
Average Recovery 93.4
Basic
105
-
80
110
103
80
-
100
-
93
64
107
122
100
80
95.3
Participate
-
95
78
-
-
.
110
100
84
130, 84
-
116
107
102
106
101
60
-------
Table 13. ICAP Recovery Results, %
Acidic
Leachates
Basic
Leachates
Fe
94.05
74.1*
K
98.4
102.0
Mg
94.0
95.2
Mn
93.0
94.3
Al
94.9
95.1
B
100.3
94.0
Ba
103.0
112.0
Ca
96.5
104.3
Cu
94.5
97.4
Cd
91.1
96.7
Co
96.6
96.3
Cr
97.7
92.7
Mo
94.5
94.2
Ni
97.2
105.1
Average
Recovery
96.1
96.7
*Fe(OH)3 was observed as a precipitate
-------
2.4.2 Neutron Activation Analysis
Neutron activation is another alternative to provide accurate multi-
element analyses of environmental samples (45 through 54). The method is
based on the radiochemical reactions possible when an element is exposed
to a flux of neutrons. When exposed to a flux of thermal neutrons, many
elements will undergo a nuclear reaction (neutron capture) creating another
nuclide according to the reaction:
N A_(n)JMB + Y
with the immediate emission of electromagnetic radiation. Subsequent
decay of the product (B) produces gamma ray emission at discrete wave-
lengths. Monitoring of these characteristic gamma rays provides a selective
and sensitive method for determining the composition of an unknown sample.
Quantitative analysis via NAA requires the knowledge of several constants
and experimental values including: the activation cross-section for A, the
half life of B, the neutron flux, the irradiation time and the delay time
following removal from the flux and prior to analysis.
The most commonly used and most intense neutron sources are provided
by neutron chain reactors utilizing the fission reaction. Fast neutrons
(>1 Mev) produced by the fission of uranium are moderated to epithermal
(~0.4 ev) and further to thermal (<0.4 ev) energies. These thermal neutrons
assure continuation of the fission reaction but can also be used to pro-
duce artificial radioactivity, induced when a target material is exposed to
thermal energies. In order to determine accurately the flux to which the
sample is exposed, a standard is irradiated simultaneously with, and as near
as possible to, the unknown sample. A relative counting of the standard
and sample is performed and the unknown weight is calculated as follows:
(g)u/(g)S = (A)u/(A)S
where (g)u and (g)S are the weights of unknown and standard, and (A)u and
(A)S are count rates of unknown and standard, respectively. Many poten-
tial sources of errors can be eliminated by irradiating the sample with a
standard of similar composition. When the composition of a substance is
unknown, a preliminary irradiation should be performed and can be used as
62
-------
a qualitative analysis in order to fabricate a suitable standard. Care
should be taken that the sample and standard are approximately the same
weight, shape and thickness.
Analysis of the gamma ray spectrum from a sample is usually performed
using a high resolution lithium-drifted germanium (GeLi) detector connected
to a multi-channel analyzer. Most facilities provide for direct computer
analysis of the spectrum. Using this method, multi-element analysis can be
performed on a single sample.
Detection Limit
The sensitivity of NAA for a given element depends on the isotopic
abundance of the stable isotope (which becomes the radioactive isotope),
the cross section for neutron capture of the stable isotope, the available
neutron flux, the length of irradiation, the half-life of the radioactive
isotope produced, the decay scheme of the radioactive isotope, the decay
period before counting, the efficiency of the radiation detector for the
type of radiation being measured, and interference from other elements in
the sample. The latter can be eliminated by pre- or post-irradiation chem-
ical separation of the interferent. Some optimized detection limits are
given in Table 14.
For multi-element analysis, however, the irradiation scheme can be
far from optimum for some elements since half-lives can vary from a few
minutes to several years. For complete analysis two irradiations are
required. First, a short irradiation followed by immediate analysis for
the short-lived components (Al, Ca, Cu, S, Ti, Cl, Br, Mg, Mn and In) is
performed. A subsequent analysis for elements with longer half-lives
(K, Zn, Ga, W, Sb, La, Sm, Eu, Au, Cr, Se, Co, Fe, Se, Ag, Ce, Hg, Th and
Ni) requires a longer irradiation time (several hours) followed by a cool-
ing period (to eliminate short-lived emissions) of 20 hours and analysis
of elements with half-lives of 8-50 hours. A further cooling period of
20-30 days is required prior to analysis for elements whose half-lives are
longer than 10 days.
Interferences:
As with other multielement techniques, matrix problems can be
encountered. Preparation of standards whose matrix is similar to the
63
-------
Table 14. Detection Limits Reported For Some Elements
By Neutron Activation Analysis (55)
Atomic
Number
11
12
13
14
16
17
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
37
38
39
40
42
44
45
46
Element
Na
Mg
Al
Si
S
Cl
K
Ca
Sc
Ti
V
Cr
Mn
Fe
Co
Ni
Cu
Zn
Ga.
Ge
As
Se
Br
Rb
Sr
Y
Zr
Mo
Ru
Rh
Pd
Detection
Limit, M-g
0.0003
0.01
0.0005
3.0
7.0
0.003
0.01
0.1
0.001
0.004
0.00004'
0.05
0.000004
10.0
0. 0003
0.02
0.0001
0.008
0.0001
0.003
0.0002
0.002
0.0001
0.02
0.0001
0.02
0.1
0.005
0.001
0.0001
0.002
Atomic
Number
47
48
49
50
51
52
53
55
56
57
58
59
60
62
63
64
65
66
67
68
69
70
71
72
73
74
75
77
78
79
80
El ement
Ag
Cd
In
Sn
Sb
Te
I
Cs
Ba
La
Ce
Pr
Nd
Sm
Eu
Gd
Tb
Dy
Ho
Er
Tm
Yb
Lu
Hf
Ta
W
Re
Ir
Pt
Au
Hg
Detection
Limit, (Jig
0.0003
0.0004
0.000003
0.001
0.0002
0.002
0.00001
0.00005
0.0003
0.0004
0.01
0.002
0.0006
0.00003
0.000003
0.0006
0.003
0.000002
0.00007
0.0001
0.03
0.0005
0.00002
0.001
0.05
0.0002
0.00006
0.00002
0.01
0.00001
0.01
64
-------
sample is important. Interelement interferences can be eliminated by pre-
or post-irradiation chemical treatment but is not usually required.
Sample Preparation
Solids usually require little or no sample preparation. A known
quantity is sealed in a polyethylene or quartz container and irradiated.
Liquid samples may be run directly or evaporated onto a surface prior to
being sealed in a container. Some care must be taken that the container
can withstand the temperature of the irradiation area and that expansion
of the sample on heating does not cause it to rupture. In samples where
the matrix contains large amounts of a potential interferent, pretreatment
to remove the element (ion exchange, precipitation, etc.) is advised.
2.4.3 X-Ray Fluorescence Analysis-Introduction
When exposed to the proper amount of energy, many elements will undergo
excitation of an inner core electron. Upon cascading of the remaining
electrons to fill the hole created, X-rays are emitted, the wavelengths of
which are characteristic of the element. This forms the basis for the
techniques of Energy Dispersive X-ray Fluorescence (EDXRF or XRF) (Refer-
ences 56-65) and Particle Induced X-ray Emission (PIXE) (Reference 66-73).
The two methods are distinguished by the excitation technique and are
discussed separately below.
2.4.3.1 XRF (Theory)
In conventional X-ray fluorescence, X-rays are used as the excitation
source. An X-ray tube or a radioactive source of X-rays is used as a
primary (continuum) source of radiation. X-ray tubes have the advantage of
greater power and therefore lower sensitivity. The broadband radiation
from this source is usually focused onto a target of a pure metal (Cu, Mo,
Ti or Sm) to produce a coherent X-ray beam which is then focused on the
sample. Use of such secondary sources reduces background due to coherent
and incoherent scattering of the continuum source radiation. A disadvantage
of the selective secondary source excitation technique is that the number
of elements excited by the secondary X-ray beam is reduced since excitation
efficiency is highest when the exciting energy just exceeds the binding
energy of the electron in a given atomic shell. For this reason, multiple
analyses are usually performed using, for example, titanium secondary
65
-------
emitters for elements of atomic numbers 13-20, molybdenum for elements
20-38 (plus lead and mercury) and samarium for elements of atomic numbers
38-56. One approach to monitoring X-rays from the samples is to use a solid
state lithium-drifted silicon detector coupled to a multichannel analyzer.
Another system common employed is wavelength dispersion. In this system the
secondary X-rays from the sample are directed onto a crystal. These X-rays
are diffracted by the crystal according to the Bragg equation. The sepa-
rated X-rays are detected using a proportional or scintillation counter
which is rotated through 180°.
Sample Preparation
Proper sampling procedures can eliminate the need for sample prepara-
tion. Collection on a thin film (e.g., Teflon, Mylar, cellulose) allows
direct sample analysis. Sealing a known quantity of a solid material
between two thin films or evaporation of an aliquot of a liquid sample onto
a thin film or filter is usually sufficient.
Interferences
Absorption of the X-rays from the sample by the support medium (filter
or thin film) may be significant for elements below 10 in atomic number when
they are imbedded in a substrate. Use of Mylar or membrane filters, where
the sample remains on the surface can significantly reduce this problem.
Further, these substrates have low mass and thickness, reducing background
count. Self-absorption is only a problem for elements above potassium and
then only when present in large amounts (e.g., >200-400 ug/cm2). Since
particle size and sample thickness may be significant in elements below
atomic number 20 (calcium), the accuracy of analysis for these elements
varies widely, depending on the size, thickness, and composition of the
sample. For elements below calcium, wavelength dispersive XRF, with its
higher resolution and lower background, is recommended.
Interferences due to the overlap of X-rays from one element with
another may be significant when one is present in large excess. Typically,
multiple lines result from a single element, corresponding to filling of
holes in the K and, for the heavy metals, L electron shells. The lead La-,
for example will overlap with the arsenic Ka-j. Such interferences may be
eliminated by placing filters between sample and detector, use of a
66
-------
secondary source which does not excite both elements, or mathematical
subtraction of the interfering line following analysis for the interfering
element from its characteristic X-ray.
Detection Limits
Detection limits for the elements above calcium in atomic number are
in the range of 1-10 ng/cm of irradiated surface. The detection limit
depends on the background count rate, however, and therefore varies with
substrate and sample composition. Detection limits for the elements of
atomic number in the range of 21-50 are typically within a factor of two
of each other. Use of wavelength dispersive XRF can increase sensitivity
a factor of two for most elements, but is usually employed when sample
composition is known. For a comprehensive evaluation of detection limits
by element and instrument, see reference 65.
2.4.3.2 Particle Induced X-Ray Emission (PIXE)
An alternative method for the excitation of samples for X-ray analysis
is the use of accelerated particles, usually protons. Proton beams of 2-5
MeV at a current of 10-50 amps produce the same effect as observed in X-ray
fluorescence. The beam may be supplied by an accelerator or cyclotron and
detection and amplification is similar to that discussed above. This
technique is limited to thin samples or small amounts of sample mounted on
a thin support (usually Mylar, Teflon or cellulose acetate). With thick-
nesses of more than approximately 1 mg/cm , appreciable background con-
tinuum causes reduced sensitivity. Furthermore, energy losses as the beam
passes through the sample and support become significant when the sample is
greater than about 10 urn diameter. The high sensitivity of the method
compensates for the need to use small samples, and has been used to advan-
tage in analysis of size fractional samples of aerosols from volumes as
small as 30 liters of air.
I
Interferences
The single most important interference is due to energy loss and
scattering in thick samples. Particle size becomes important, since self
absorption of the X-rays may occur. As with XRF, overlap of L lines from
the heavier elements with k lines from the lighter ones may be reduced by
67
-------
filtering or other techniques. Sophisticated computer techniques are
usually employed for data analysis.
Detection Limits
The detection limits for the approximately 15 elements which can be
detected quantitatively by PIXE are in the microgram region; however, the
value depends on the element and the matrix, since the detection limit
depends on the cross section of the element and the magnitude of background
continuum.
Application of PIXE to Sized Particle Samples
In recent years the amount of trace elements emitted from industrial
processes has begun to be a major concern of government sponsored research.
Because there is a relationship between size of the particle and its pene-
tration into the lower respiratory system, research has been directed at
finding methods to measure the trace element content of sized particles.
Some investigators have simply collected gross quantities of sample,
and used sieving techniques to fractionate it (74). This approach always
runs the risk of modifying the sample through the sampling or fractionation
process, consequently leaving in doubt the true gas phase particle composi-
tion. By sampling with an impactor and analyzing the individual stages
for trace element content, both mass and elemental concentrations can be
calculated (75).
A variety of impactors is available for sampling flue gas streams.
In a recent test conducted by TRW an MRI 1502 impactor was used to collect
the particles on a greased (Apiezon L) substrate (Kapton film) placed on
each collection stage. Only a portion (typically one impaction spot) was
mounted in a plastic photographic slide mount for PIXE analysis by Crocker
Nuclear Laboratory of the University of California at Davis. In practice
two or more impaction spots are sent for analysis. These impactor studies
were performed at a coal-fired power plant at the outlet of a limestone
scrubber with a venturi/absorber combination using Chevron type mist
eliminators.
The impactor data was reduced using the procedure (76,77) developed
by Southern Research Institute (SoRI) for IERL/RTP. In this approach
LdM/d(logDgQ)J the data are normalized so that a smooth curve can be drawn
68
-------
through the limited number of data points obtained from the impactor. The
normalized data mass placed in the dM/d (log D5Q) format is shown in
Figure 10. The data for Run 135 was corrected for a negative filter weight
using the weight percentage of the filter from Run 136. The filter data
point for 135 is shown in brackets. Even without this correction the data
agree rather well. The slight offset in the cut points is due to the
difference in flowrate between the two runs. In both cases the data show
a bimodal distribution with an apparent maximum around 0.4 to 0.5 pm.
In contrast to the mass data, the elemental data obtained from PIXE
analysis show maxima at slightly different particle sizes (Figure 11).
.Other elements were detected, but only the elements with PIXE data for all
stages were presented. Calcium, Pb, Zn and Si had a maximum at 20y and
0.9y while S was the only element to show a maximum at 0.5y. Iron showed
its concentration maximum over the mass minimum. The lower particle size
maximum for the elements corresponds closely to the total mass value
(0,4p to 0.5y).
It was found that the total elemental weight (even corrected for oxygen)
was well below the mass weight found in the stages. If significant amounts
of organic material were present, then the total elemental weight would
be low due to the insensitivity of PIXE to carbon. A Thermogravimetric
Analysis (T6A) in air of outlet fly ash produced no significant weight
loss which would occur if organic carbon were present. Additional work
with sized standards and mounting procedures is required to improve the
absolute accuracy of PIXE in this application.
In order to avoid this apparent problem, the elemental data can be
ratioed to Si (a major constituent of fly ash) to attain a relative concen-
tration per stage. The assumption here is that any analysis error would
affect all elements in the same fashion. Figure 12 summarizes impactor/
PIXE data handled in this fashion. Sulfur showed a consistent enrichment
through all the stages and back-up filter. The apparent enrichment on the
filter might be due to H2S04 being collected by the final filter, since
the impactor in these tests was run at stack temperatures, which would not
prevent the collection of H2S04 aerosols. Calcium shows a weak positive
trend while Pb and Fe showed a moderately .strong trend to increasing con-
centration at smaller particle size. Zinc showed the strongest trend of
69
-------
1000
O
o
1.0 10.0
GEOMETRIC AERODYNAMIC DIAMETER,
100.0
Figure 10. dM/d(logDso) Size Distribution at the Outlet of a Limestone
Wet Scrubber at a Coal-Fired Utility
70
-------
1,000
o
2
T>
100
10
a
0
A
a
0
a
A
•
• -Co
0«S
A.pb
D*Zn
0-F.
©-Si
8
1.0
10.0
100.0
GEOMETRIC AERODYNAMIC DIAMETER, I*
Ftgure 11, Trace Element Distribution by Particle Size at
the Outlet of a Limestone Wet Scrubber
71
-------
1000
100
3
o
g
• = Co/Si
0»%/Si
A * Pb/S!
Q * Zrv/Si
0 - Fe/Si
o
I
2
t-
10
1.0 10
GEOMETRIC AERODYNAMIC DIAMETER,
100
Figure 12. Ratio of Element Concentration to Silicon
Concentration by Particle Size
72
-------
the metals paralleling sulfur throughout tmpactor stages, hut dropping off
on the filter.
Results like these have been used to explain the reason for an apparent
enrichment of certain elements after passage through a wet scrubber. Wet
scrubbers as shown in Figure 9 emit mainly small particles and remove the
large ones. If an element has a concentration increase with decreasing
particle size, then preferential removal of the large particles would
appear to make that element more concentrated in the outlet particulate
matter. This hypothesis is illustrated by Zn which exhibited an enrichment
across the scrubber and was found to have higher concentrations in the
smaller particles.
2.4.4 Sample Dissolutibn Procedures
For many of the elemental procedures, the sample must be in the form
of a solution. To accomplish this, a dissolution scheme is required that
will dissolve all of the material, since aqua regia or other strong acid
extractions cannot release all the trace elements in solutions. This is
due in part to the fact that most environmental samples from oxidizing
processes produce an ash which is mostly composed of alumino-silicates. In
many combustion processes the alumino-silicates take the form of glassy
spheres which might be hollow or solid, but generally impervious to acid
attack.
The approach shown in Figure 13 represents an exhaustive approach to
.dissolve almost all of the matrix in order to release any trapped trace
elements. Fly ash samples treated in this fashion are normally completely
dissolved in the HN03/HF solution which is suitable for most analytical
methods. ICAP cannot tolerate HF due to the quartz capillary in the
nebulizer, but the HF can be removed by heating with excess HC104. For
additional information on dissolution schemes, see references 16 and 78.
2.5 SUMMARY OF INITIAL SAMPLE CHARACTERIZATION
Figure 14 presents the analysis flow for the initial sample charac-
terization portion of the Level 2 inorganic compound identification scheme.
As it was discussed in Section 1.0, the initial step is to evaluate the
Level 1 SSMS data to determine which elements exceed their most toxic DMEG
value. This step focuses the search on compounds which are potentially
73
-------
RESIDUE
OXYGEN**
PARR-BOMB
OVER HNOo
RESIDUE
IGNITE FILTER tt
FUSE TO Na2CO3
DISCARD
^RESIDUE I DISSOLVE IN
^— I mi i ITF HMO
DILUTE HNO,
COMBUST**
IN OXYGEN
PLASMA
HNO3 +HF°IN t
TEFLON PARR-BOMB
FILTER
FILTRATE
FOR
ANALYSIS
**
tt
CATION ANALYSIS ONLY
SILICON ANALYSIS NOT APPLICABLE
PROBABLE LOSS OF SOME As, Se, Sb, Hg
SODIUM ANALYSIS NOT APPLICABLE IF THIS STEP
PERFORMED (NOT NECESSARY IF ONLY
LOOKING FOR No)
REMOVE HF PRIOR TO ICAP ANALYSIS BY
HEATING WITH HCIO4 TO NEAR DRYNESS
(NOTE: LOSS OF As, Se, Sb, Hg POSSIBLE)
Figure T3. Flow Chart of Sample Dissolution Procedure
74
-------
11
t
^\ OMEG /
\^ VALUES /
J »I>MEG VALUES
/LIST ELEMENTS 7
EXCEEDING DMEG /
VALUES _/
4
/^S^ PROCESS
*\*Ji DATA
T
LIST POTENTIAL
COMPOUNDS
PRESENT
XSE1E\
yeOMPOUNDS\
. / STABLE \
"— VUNDER PROCESSOR/
O \ SAMPLING /
N^ONDIT-/
\ IONS/
/UP-DATE POTENTIAL /
COMPPUND LIST /
*
STUDY GENERAL
CHARACTERISTICS
OF PARTICLES
f
t
1TGA/DSC
(N2 OR AIR)
*
CHARACTERIZE
SAMPLE FOR
WEIGHT GAIN/ LOSS,
— RF ACTION TFMPFDATI IBPC
•PHASE CHANGES
1
MICROSOLUBILITY MICRO-SPOT TEST
JKTS ON SPECIFIC
NEUTRAL^' ANIONS/CATIONS
*
/SOLUBILITY OF /
SPECIFIC GROUPS /
OF PARTICLES /
^
\2_j^
/~lisTS OF 7
/ ANION VS /
1 SOLUBILITY /
|
/ STABLE DRYING /
/ TEMPERATURES, VAPOR-/
/IZATION TEMPERATURES, /
/DECOMPOSITION POINTS/
/AND AIR STABILITY /
i >
t
',
UP-DATE LIST OF
POTENTIAL
COMPOUNDS
| '
SELECTS SPECIFIC ANION/
CATION TESTS FOR
POTENTIAL ELEMENTS
-------
harmful, and elimlnates those which probably are not harmful at the
concentrations emitted. At point 2, a list of potential compounds is !
developed drawing upon the DME6 compounds (3) and the process specific
compounds found in Section 1,2. This list of potential compounds is
scrutinized for compounds which obviously could not be stable under the
sampling conditions (Table 2). Using standard references like the CRC
Handbook of Chemistry and Physics, Merck Index or Critical Tables,
unstable compounds are culled from the list. This list will be used as
both a checkpoint for closure and a focus for the research.
The first test run on the sample will be a PLM analysis and a TGA/DSC
scan of the sample. As soon as the sample is in the laboratory, it should
be viewed with a PLM and a photomicrograph taken, in color, to serve as
quality control. If any changes in the general appearance of the sample
occur during the duration of the analytical activities, these should be
noted. The PLM can also provide a measure of the complexity of the sample
simply by noting the number of different types of particles.
Polarized light microscopy is the first direct compound analysis
method. Particles can be identified by the determination of such proper-
ties as the refractive index, isotropy or anisotropy, birefringence,
pheochroism, fracture, color, and crystal habit. Microspot tests for com-
mon anions and tests of the solubility of the particles in water, acid,
and base can be performed directly on the sample as it is being examined
under the microscope. These microtests will alert the analyst to perform
quantitative analyses for the anions detected, provide information about
the potential success of full scale dissolution, and possibly confirm
identification of a given particle. In addition to these data the PLM
analysis can provide information on particle size.
In conjunction with the PLM work a TGA/DSC scan of the sample should
be made. This test is used primarily to study the stability of the sample
and determine the appropriate temperature at which to dry samples to be used
in later tests. In a few cases it is possible to determine the compounds
present by the weight loss at specific temperatures. At this point several
of the major compounds will have been identified and the information on
stability and morphology can be used to update the potential compound list.
76
-------
By step 4 the analyst is ready to perform bulk elemental analysis and
quantitative anion analysis on the samples. The elemental composition of
the bulk sample"may be determined by several techniques. Each has basic
advantages and disadvantages which should be carefully weighed before
chosing a particular method for analysis. ICAP is recommended for Level 2
analysis due to its low cost, acceptable sensitivity and its simultaneous
multi-element analysis capability Cup to 60 elements). Availability of
instrumentation is a critical factor in chosing a method and therefore
such methods as AAS, XRF and NAA might be used to supplement ICAP or be
used in place of ICAP on the basis of availability. Whatever the choice,
the instrument should be run in the most accurate fashion employing
standards made up in the same matrix as the sample or using standard addi-
tion procedures to overcome any interferences due to the matrix. The goal
is to achieve ±15% or better accuracy.
While an elemental analysis is being performed, anion analysis using
an ion chromatograph (1C) will be performed. 1C is recommended for anion
analysis because of its ability to survey the sample qualitatively while
quantitating most anions at the 1 ppm level. At the completion of anion
and cation analysis the ionic charges (assuming the highest oxidation
state) are compared and the degree of closure is assessed. Unless the PLM
or TGA provided direct compound identification, it is possible only to say
that some fraction of the cations is present as sulfates, chlorides, etc.
In most combustion processes the remaining fraction can be assumed to be
present as oxides.
At the end of the initial sample characterization, information will
have been obtained in the following areas:
1. General appearance of a sample
2. Particle size distribution
3. Index of refraction and crystal structure
4. Weight loss with respect to temperature
5. Bulk elemental concentration
6. Bulk anion concentration
77
-------
If this information is sufficient to narrow the possible choices of
compounds to an acceptable level (based on the end use of the data), then
the analyst may wish to stop. If there is a need for more specific com-
pound identification, then this initial sample characterization will provide
a strong starting point to interpret the data from the compound identifica-
tion procedures discussed in the following section.
78
-------
3.0 BULK CHARACTERIZATION
After initial sample characterization, bulk characterization begins.
In this portion of the Level 2 analysis scheme the following techniques
are applied:
• X-ray diffraction (XRD)
• Fourier transform infrared (FTIR)
• Electron spectroscopy for chemical analysis (ESCA) and Secondary
Ion Mass Spectrometry (SIMS).
These techniques are described in Table 15 in summary form. The first
phase of this Level 2 analysis scheme has taken a list of prospective com-
pounds and compared them to the results from the elemental and anion tests.
In addition some information from PLM on the compounds present will be
available. With this as the starting point, an FTIR spectrum of the sample
will be run. The IR spectrum will provide specific information on the
functional groups present and possibly identify specific compounds. With
this information in hand, interpretation of the XRD spectra will be greatly
simplified. XRD provides direct compound information, but unfortunately
it is normally sensitive only to materials present above *1% by weight.
Finally depth profile analysis by ESCA or SIMS will be used to study the
surface composition of the samples. Important information on the formation
of trace element compounds can be gained as well as knowledge of surface
reactivity or catalytic activity. This data is extremely important in
predicting the toxicity of particles as they interact with lung tissue.
3.1 FOURIER TRANSFORM INFRARED ANALYSIS
Infrared spectroscopy has been widely employed in commercial and
synthetic organic industries. The major strength of this technique is its
applicability to both qualitative and quantitative analysis of most com-
pounds in all phases (solids, liquids and gases). In general, infrared
spectroscopy is fast, requires small quantities of sample, can differentiate
between subtle compound structural differences, e.g., isomers, conformers,
crystalline forms, and it can be applied to surfaces. Interpretation and
quantisation of infrared spectra requires access to reference spectra,
79
-------
Table 15. Summary Bulk Composition Characterization
Analysis Method
Principle of.Operation
Information Derived
Compound Identification Procedure
Limitations
Infrared and Far
Infrared
Many inorganic anions have
specific absorption bands in
infrared. These bands can be
used to identify and quantify
the anions present
Either wet chemical or quanti-
tative IR techniques directed
toward specific anions
Used to determine presence of
specific anions such as
MnO^, P0|, or CrO^
Confirmation and quantitation
of specific anions
Only small shifts are seen in the
spectra with different cations.
Anion information is essential for|
interpreting XRD data to elimina
potential compounds
Ratio's of anion/cations used to
predict potential compounds
Inorganic halogens have no bands in
the IR
Spectra can change depending on
moisture content of sample
Time consuming, since directed
toward specific anion
Electron Spectro-
scopy for Chemical
Analysis (ESCA)
i 00
o
Sample is irradiated with
X-rays, causing inner shell
electrons to be ejected:
• Energy of these ejected'
electrons is a measure of the
binding energy of electrons
as modified by the chemical
surroundings of the emitting
atom
• Energy shifts in the binding
energy of electrons emitted
from same element indicate
different chemical environ-
ments
Elemental characterization
determines oxidation state of
elements present in sample
Can determine bulk concentrations
of homogeneous samples at or
above 0.1%
Though ESCA is extremely 'surface
limited since electrons have
shallow (3 to 20A°) escape depth,
this makes the ESCA a very use-
ful tool for studying absorption
phenomena such as SOj on soot
or flyash
Most commercial instruments have
ion (Ar+) beam for sequential
removal of atomic layers for
depth profile analysis
Direct compound identification not Interpretation and quantitation
normally possible since there are of data is difficult and requires
usually only small shifts in |standards matching the matrix
binding energy of elements in the
same oxidation state associated
with anions or cations
X-Ray Powder
Diffraction (XRD)
Powder sample diffracts a
primary X-ray beam into a
series of diffraction lines
characteristic of a given
crystalline substance
Quantitative compound deter-
minations are made, commonly
using an internal standard with
subsequent quantification by
comparison to standard curves
Interpretation of diffraction
pattern provides qualitative
information on crystalline
materials present. Diffraction
lines are matched with spectra
of pure compounds in the ASTM
powder diffraction tables
The diffraction lines are studied I State of the art sensitivity is
and potential compound diffractionIlimited to -0.05% depending on
spectra are compared to lines in (compound and matrix. Routine
sample spectra Jsensiture is closer to 0.5%
Potential compounds are eliminatedIOnly crystalline materials can
or proposed based on information |be seen
from SSMS or AAS (elemental dis-
tribution), ESCA (oxidation state),
and IR (anions present)
-------
calibration with pure components and, most importantly, a know!edgable
analyst.
Fourier Transform Infrared Spectroscopy (FTIR) obtains
spectral information identical to that from conventional IR with the major
advantage of greater sensitivity. The following sections will discuss the
advantages of FTIR over conventional IR, sensitivities, principle of opera-
tion, available equipment, limitations, and implementation of FTIR to
environmental samples.
3.1.1 Theory
Although the identical spectral information can be obtained from both
conventional and Fourier Transformed Infrared Spectroscopy, it is obtained
in quite a different manner. Conventional dispersive Spectroscopy employs
a prism or grating, which disperses the polychromatic infrared radiation
into a spectrum of frequencies, and then scans the energy in each individual
frequency interval sequentially. To obtain quality spectra with a dis-
persive instrument, narrow slits are required. The slits ensure that the
frequency intervals, detected at any one time, are sufficiently narrow so
that the desired resolution is obtained.
Interferometric Spectroscopy, commonly called Fourier Transform
Spectroscopy uses an interferometer instead of a grating or prism monochro-
mator. The spectral information for all the frequencies is obtained at
the same time during one scan whose duration is on the order of 1 sec.
The actual spectrum is obtained by taking the inverse Fourier transform of
the interferogram formed.
The. most common interferometer used in Fourier Transform systems is
the Michelson interferometer. (79) A simplified diagram of such an inter-
ferometer is shown in Figure 15. The interferometer consists of two
mirrors at right angles to each other. One mirror is stationary while the
other moves in a direction perpendicular to its front surface. A beam-
splitter is positioned at an angle of 45 deg to the two mirrors, which
divides the incident beam. Ideally, the beamsplitter should transmit
50% and reflect 50% of the light. The two beams pass to the stationary and
moving mirrors and reflect to the beamsplitter. The beams are then
recombined at the beamsplitter and exit the interferometer to the detector.
81
-------
MIRROR
DRIVE
STATIONARY
MIRROR
s.
n A
n
M
1 '^cmAtt
' '
1 1
II
II
1 1
^
\
\
\
\
*
MOVING
MIRROR
>
UNMODULATED
INCIDENT
\ BEAM
N
\
\
N BEAM SPLITTER
\
M AIM ii k.<^en
EXIT
BEAM
Figure 15. Diagram of Michel son Interferometer
Various degrees of interference (from totally constructive to totally
destructive interference) are produced for each frequency by the optical
path difference in the two arms. Consider a monochromatic source. When
the optical path lengths of each arm of the two light beams are identical,
there will be constructive interference of the two light beams when they
recombine at the beam splitter. If the movable mirror is moved 1/4 of a
wavelength, the two beams will be 180 deg out of phase when they recombine
at the beam splitter; they will destructively interfere. Each individual
incident frequency will produce an output oscillation with a cosine varia-
tion whose frequency is dependent on the incident frequency. Since each
frequency can only interfere with itself, the output of the interferometer,
the interferogram, for a polychromatic source is the sum of the oscilla-
tions at each individual frequency. The point of maximum intensity in the
interferogram occurs at the position where the optical path lengths for
the two arms of the interferometer are identical. It is only at this
mirror position, that every frequency constructively interferes. Thus, it
can be seen that the interferometer establishes specific phase relation-
ships for each frequency as a function of mirror displacement. The resul-
tant interferogram is related to the intensity as a function of frequency
spectrum by a Fourier transformation. Fourier analysis of the inter-
ferogram picks out the pattern for each frequency and determines the
82
-------
magnitude of the oscillation at that frequency, the Fourier coefficient.
The Fourier Transform analysis is a tedious process and must be handled by
high speed computers.
Commercial Fourier spectrometers are based mostly on the Michel son
interferometer. They differ little in principle, but vary greatly as far
as the optical, mechanical and electronic components are concerned. The
spectral ranges covered depend on the nature of the beam splitter and
detector. Most instruments cover the range from 40,000 cm"1 to 10 cnf1
(0.25 to 1000 microns) with scan times from about 1/1Oth second to
several hours. Computer capabilities are present on all marketed
instruments.
3.1.2 Advantages of FTIR
Fourier Transform spectroscopy is a less direct way of obtaining a
spectrum than conventional dispersive spectroscopy; however, there are
certain advantages that make the interferometric method favored over the
dispersive method. The most important points are:
§ The increased energy throughput (Jacquinot's advantage).- There
are no slits and throughput is limited by the size of the mirrors,
100 to 200 times better than a dispersive instrument. (80,81)
• The multiplex advantage (Fellgett's advantage) - The interferometer
receives information from all frequencies in the spectrum simul-
taneously and therefore, the signal for each resolution element
is proportional to T, where T is the total scan time. For equiva-
lent scanning times and optical throughput, the rapid interferometer
has a higher signal-to-noise ratio than conventional IR. The
signal-to-noise ratio can be improved proportionately to the
number of scans. (81,82)
t The improved accuracy of the frequency scale - The position of the
movable mirror is controlled by a reference interferometer that has
a monochromatic laser as a light source. This internally cali-
brates the frequency scale and is not sensitive to the system's
temperature or humidity. (83)
• The computer capabilities - The inverse Fourier transform of the
interferogram must be computed to obtain the intensity as a function
of frequency spectrum. The utilization of the fast Fourier trans-
form algorithm of Coley and Tuckey (84) by a dedicated minicomputer
makes on-line Fourier analysis a reality. Digitized spectral data
are produced by this process and once a spectrum is recorded and
stored, the computer can aid in data presentation, enhancement and
interpretation,
83
-------
Simply, the FTIR has increased sensitivity, increased speed, and better
wavelength accuracy than conventional IR with computer capabilities for
data reduction.
3.1.3 Application of FTIR to Environmental Samples
Environmental particulate samples undergoing infrared analysis truly
require the advantages attained from both Fellgett's and Jacquinot's
principles. These samples generally are poor infrared energy transmitters
and the compounds of interest to the Level 2 analysts can be present at
concentrations less than 1%.
When applying FTIR to characterize environmental samples, the analyst
should be aware that:
t The identification of inorganic compounds by infrared spectroscopy
is nonroutine. There is not a wealth of reference spectra to aid
the analyst in interpretation.
• Although FTIR is a rapid analytical method, the complexity of
environmental samples warrants careful and labor intensive sample
preparation. Representative samples must be acquired once optimum
sample quantities and drying temperatures are determined and com-
pound verification and quantitation is best accomplished by
standard addition techniques.
• Many simple inorganic compounds such as borides, silicides, nitrides
and oxides, do not absorb radiation in the region between 4000 and
600 cnH. Therefore, the far-infrared region, 200 to 10 cm-1, must
be scanned. This necessitates that many solids be prepared twice
for FTIR using two dispersing agents for optimum spectra definition.
• The identification of polyatomic anions, e.g. 003, $04, NO^, etc.,
is straightforward. However, compounds such as KNOs can be dis-
tinguished from NaNOa or Ca(N03)2 only when standard spectra are
available. In this case elemental data obtained from the initial
sample characterization can be used to decide which compound is
present.
t Inorganic compounds can react with many of the standard IR window
and support materials (cation exchange). The analyst must be
aware of this and decide on the most stable paraparative medium.
Inorganic compound identification of environmental samples has been
successfully employed (85,86) in characterizing samples from a Fluidized Bed
Combustor (FBC).
84
-------
The sample types considered in this discussion are loose particulates
(cyclone catches, bulk solids, e.g. feed materials, slurry solids, over-
flow bed materials, ash, waste solids, etc.), filters (SASS train, water
filtrates, etc.), and MRI impactor stages (Kapton lines and neat). FTIR
can be successfully employed in the analysis of gaseous grab samples for
many inorganic gases, e.g., CO, C02, NH3, N02, S02, N1(CO)4, AsH., or
PH3. (87,88)
The moisture content of a sample affects the spectral quality and in
many cases drying is necessary to produce well resolved spectra. Informa-
tion from TGA analysis should be used by the analyst to establish a drying
temperature which allows water evaporation without sample decomposition.
FTIR scans should be run, if time and sample quantities are sufficient, both
with and without drying.
Recommended sample preparation methods for the various sample types
considered in this manual are given in Table 16. Although other mulling
agents and window materials are available, those recommended are free from
ion exchange reactions. (89) In the mulling technique, finely ground
particulates (<40 micron grain size) are suspended in the mulling agent and
the resultant slurry is supported between two infrared transmitting windows.
Table 16. Recommended FTIR Sample Preparation Techniques
Mulling Window
Sample Type Spectral Region Agent Material Comments
1 nnco
L.UU3C
Rarticul ates
and MRT
ttllU rll\l
Stages*
_i
3800-1333 cm '
1333- 400 cm
600- 45 cm"1
Fluorolube BaF9
Nujol
Nujol
£»
AgCl
Polyethylene
/These window
1 combinations
1 provides good
< quality spectra
I without ion
1 exchange
\ reactions
Filters
3800- 45 cm
-1
None
KRS-5
Attenuated Total
Reflectance
(ATR)
After removal with distilled in glass or spectro-grade hexane and room
temperature evaporation.
85
-------
Most samples do not require prolonged grinding and adequate dispersions are
-obtained from a 10% (e.g. 25 mg sample/250 mg mulling agent) sample load-
ing. Hand grinding in an agate mortar and pestle produces well ground
samples for most environmental particulates. Additional sample or mulling
agent can then be added until a spectrum of desired intensity is obtained.
Teflon, aluminum, or polyethylene standard spacers can be used to control
path length for samples undergoing species quantisation or to obtain more
intense scans. The computer capabilities available in FTIR do not necessi-
tate strict control of path length for quantitation as data can be ratioed
to the intensities of the mulling agent absorption peaks. However, control
of path length is sometimes necessary to obtain high quality spectra.
An example of an acceptable quality spectrum is shown in Figure 16 (44),
This spectrum was taken of particulate matter emitted from an FBC. The
expected material (CaS04, CaCOg, Si02) is clearly visible. These par-
ticulate samples were hand ground, mulled with Nujol (mineral oil) and
supported on AgCl windows. Forty scans were averaged in this test,
although more scans could have been made if higher sensitivity were
warranted. In most cases, the performing analyst's judgement determines
the parameters for instrumental operation. The analyst observes the
obtained spectrum and determines: which spectral regions require more
scans for improved sensitivity, which wavelength regions require better
point-to-point resolution for ease of interpretation, and what computer
data manipulations (e.g., subtraction of mulling agent and window back-
grounds, comparison with standards, comparison with other samples, quanti-
tation ) are essential.
The application of Level 2 inorganic analytical techniques to samples
before and after aqueous and acid extractions (leachate studies) provides
important environmental toxicity information:
• Surface compounds are removed and interior ones exposed for
identification.
• Waste storage and ground water pollutant potential is evaluated.
• The impact on throat and lung tissue, which readily absorb water
extractable compounds, is characterized.
86
-------
oo
I—-mwt-—i—I—i-1"-
i CiQMRM F. ;S-?t-fi I
m.
It 00
1300
1260 u bo lobo i5o abo 755 ebo
WRVENUMBERS
o
Figure 16. FTIR of Outlet FBC Material
-------
FTIR's computer capability for digital wavelength subcontraction is techni-
cally suited to provide the most complete documentation of inorganic corn-
pounds' behavior, concentration changes, and stability in leaching
environments. Figure 17 provides a scheme for conducting surface and
leachate analysis of fly ash and emission particulates by FTIR.
Interpretation of the infrared spectra on the basis of characteristic
frequencies can provide the identity of specific anions and some individual
compounds. General absorption regions for several anions are given in
Table 17.
Several investigators have done extensive work with inorganic com-
pounds (9) and have been able to produce specific correlations between
observed spectra and individual compounds. (89,90,91) Table 18 gives a
general overview of the inorganic IR bands. Tables 19 and 20 list the
characteristic bands for several nitrate and sulfate compounds which could
be present in environmental samples. There are definite analytical fre-
quencies which can be used to identify compounds, particularly when sup-
porting elemental analysis information is available.
The computer equipped FTIR gives the analyst the advantage of generat-
ing an extensive library of inorganic compounds for automated spectral
interpretation. It is feasible for all the DME6 listed compounds to be
prepared, scanned and retained on this computer. In this technical manual's
integrated analytical approach, FTIR supplies major information concerning
the identify and quantity of inorganic compounds present. Results obtained
using FTIR are complementary and are supplementary to those achieved from
the implementation of ESCA and XRD. This is particularly important for
materials which are non-crystalline and thus not seen in XRD.
3.2 Powder X-ray Diffraction
The diffraction pattern given by a crystalline compound when exposed to
a collimated beam of monochromatic X-rays is unique and, therefore, an
excellent method for identifying compounds. Recent advances in XRD
instrumentation and procedures have improved compound detection sensitivity
to 0.05% (92,93). Compound selectivity and improved sensitivity make XRD
a necessary analytical technique for Level 2 inorganic compound identifica-
tion. This section discusses the principle of operation, available
-------
cx>
vo
INITIAL f
SAMPLE
t
HjO EXTRA
INITIAL SA
SOLID
ACID EXTI
OF
EXTHACTEI
SOLID
CTIONOF .
MPU
PORTION
(ACTION
K.O
> SOLIDS
PORTION
FTIR ANALYSIS
FTIR ANALYSIS
SPECTRUM A
LIQUID
PORTION
4
EVAPORATE |
\
FTIR ANALYSIS
SPECTRUM 1
PTIR ANALYSIS
SPECTRUM C
-
LIQUID PORTION
FTIR ANALYSIS
SPECTRUM E
SUPPLEMENTARY
INSTRUMENTAL
ANALYSES
r — L
f^~ CNROMATOBRAPHY 1
i 1 **f5sS?' 1
l_ I AAS J
1
PERFORM:
.. . SSMS
9 . ICP
• AAS
DATA REDUCTION
QUANTITATE
OUANTITATE
,.„,._. OITAIN PREUMIN
— INTERPRET uw cyt frywaur
- SPECTRUM 1 COMPOUNDS
/
S?!?*:T S^jCTRUM C iww»i««T S SFECH
^v '
, fc QUANTITA1
' ELEMENTS |i
A F«oSVpfrf«ui?^£ k INTERPKT fc OITAIN PKUM
SPEOKUM F COMPOUNDS
ARY
S ^
IUMD
LENT!
UM 1
y
NO
Vg
INARM
IE
H
^
y
»:
CON FIRM IV
FTIRSlHTRACnONOF
STANDARDS FROM
COMPUTER FILE
USE SPECTRUM D FOI
SUITRACnON
CONFIRM «Y FTIH 1
suntAcnoN OF l_
STANDARDS FROM 1 1
COMPUTE! PILE 1
REMARKS
USTOFAMONS PRESENT
IN MjO EXTRACT
UST Of ELEMENTS
REDUCES POSSIKE COMPOUNDS
IN EVAPORATED HjO EXTRACT
•
CONRRMS WATER EXTRACT
SPEOES EQUIVALENT TO
ORIGINAL SURFACE SPECIES
DIRECT ANALYSTS OF WATER
EXTRACTED SAMHE IS POSSIILE
HIT COMPLETE SUBTRACTION
OF FLY ASH RACKGROUND
IS DIPnCULT TO DETERMINE
»
LIST OF ELEMENTS REDUCES
POSSIRU COMPOUNDS IN
ACID EXTRACT! ON
ACID SOLUUE SURFACE SPECIES
Figure 17. FTIR Analysis of Surface Composition of Fly Ash Samples
-------
Table 17. General Absorption Regions
An ion
Absorption Bands (cnr1)
S0
C0
610
1130
- 690 (m)
- TT8C (s)
610 - 640 (m, sp)
1350 - 1370 (s)
650 - 680 (m)
1430 - 1450 (s)
- 1100 (vs)
m = medium, s = strong, sp = sharp, v = very
90
-------
Table 18. Listing of Assigned Infrared Bands.Observed
in Participate Samples
Frequency, cm
3140
3020
2920
2860
2800
1768
1720
1620
1435
1400
1384
1360
1190
.
1140
Species
NH4+
NH4+
HYDROCARBON (C-H)
HYDROCARBON (C-H)
4.
NH4
N02"(BULK)
NH4* (HALIDE)
H20
co32"
NH4+
N03" (SURFACE)
N03" (BULK)
po43-
;-.
P043'
Frequency, cm"
1120
1110
1035
980
880
840
800
I
780
728
670
627
620
600
470
Species
P043"
so42"
ST044"
so42"
2_
CO,/
v J
N03~(BULK)
Si044"
Si044"
co32-
P043"
po43"
2-
so42
3_
P04
4-
Si044
91
-------
Table 19. Infrared Bands of Some Common Nitrates (cm"1)
Compound
NaN03
KN03
Ca(N03)2-XH20
Fe(N03)3-9H20
Ca(N03)2-3H20
Pb(N03)2
Band Category^3'
VW
2428
•
1044
-
2431
W
807
M
836 sp
824 sp
820 sp
835 sp
836 sp
726
836 sp
S
^1430
^1640
1615
1587
VS
1358
1790
1380
1767
"-1350
1361
^1785
1378
1790
1373
ro
W = Weak, M = Medium, S = Strong, V = Very, SP = Sharp, B = Broad
-------
Table 20. Infrared Bands of Some Common Sulfates (cm" )
Compound
Na2S04
K^
CaS04-2H20
MnS04-2H20
FeS04-7H20
CuS04
PbS04
Band Category^
VW
•
990
W
645
1010 (sh)
1670
510 (vb)
607
1150 (sh)
1020 (sp)
1600 (sh)
M
318
2200 (b)
660
1025
1625
680
805
860
592 (sp)
623 (sp)
S
620
603
667
1630 (sp)
3410 (b)
825
3225 (b)
611 (vb)
3330 (b)
1200
•v3300 (b)
VS
mo
1110
620
1130 (vb)
1135 (vb)
1090 (vb)
1090 (vb)
VO
(a) V = Very, W = Weak, M = Medium, S = Strong, SH = shoulder, B = Broad, SP = sharp
-------
equipment, advantages, limitations, and application of powder X-ray
diffraction analysis to environmental samples.
3.2.1 Theory
The atoms that make up a crystal or crystalline material occupy a
three-dimensional periodic array. The arrays of atoms in the crystal make
up planes which diffract the X-rays in much the same way as a grating
diffracts ordinary light. The relationship between the X-ray wavelength,
X, the angle between the incident and diffracted beam, e, and the inter-
planar distance in the crystal, d, is found in the Bragg equation, shown
below:
nX = 2d sin 9 0)
where n is the order of the diffraction. If the x-ray beam is mono-
chromatic, there will only be a limited number of angles at which con-
structive diffraction of the beam will occur/ '
In the case of a single crystal, the diffraction will consist of
a series of individual diffracted beams arranged systematically according
to the symmetry of the crystal and the manner in which it is oriented.
If very many small, randomly oriented crystals are placed in the X-ray
beam, each Bragg reflection will consist of a continuous cone of diffracted
rays. Any specific crystalline material will produce a series of cones
of fixed angle and fixed relative intensity. The intensity of the
diffracted ray is dependent on the kind of atoms and their arrangement
in the "unit cell," which is the smallest repeating unit in the crystal.
Thus, any one particular crystalline compound can be identified from
its X-ray pattern which is distinguished by a unique set Of d spacings
and intensities.
Two basic types of instrumentations are used,to detect diffracted
X-rays: 1) film, and 2) Geiger or scintillation counters. Camera units
are advantageously used when: (1) the quantity of material available for
analysis is small, as determinations can be made on as little as 2 ng, or
(2) when the sample is reactive or volatile. In both cases the sample is
sealed within a glass capillary tube filled to <1 cm in to depth.
94
-------
By knowing the wavelength of the incident X-ray, the d spacings of the
planes in the crystal can be determined by measuring the film, and calcu-
lating the Bragg angle. Suitable tables are available to simplify this
procedure. (95) This method of examining powders is called the Debye-
Scherrer-Hull method, or the powder diffraction method. X-ray diffraction
units employing Geiger or Scintillation counters require larger sample
sizes of 50 to 100 mg of material. For environmental samples, the Geiger
or scintillation detectors are generally used, since more accurate line
resolution and intensities are obtained from the strip chart recorder
or the computer printout. An ancillary piece of equipment which has
been effectively used with Geiger or scintillation detectors is a graphite
monochromator. This attachment, mounted between the sample compartment
and the detector assembly, isolates the incident X-ray wavelength of
interest from background radiation. This has the effect of improving
the sensitivity from ^ 1-2% of material present to 0.1 - 0.3%. (93)
The application of XRD to environmental samples has two major limita-
tions: (1) small amounts of the pure compounds of environmental interest,
<0.1%, are usually mixed with an amorphous matrix and (2) the glass fiber
filters used in particulate sampling systems produces a background scatter-
ing which can mask the diffraction pattern due to crystalline species.
These, application difficulties can be minimized through the use of the
instrument operating and sample handling procedures given in the following
section.
3,2.2 Powder XRD For Compound Identification in Environmental Samples
An XRD unit employing a Geiger or scintillation counter and a graphite
monochromator is recommended for the analysis of environmental samples of
loose particulate matter and particles impacted on glass fiber filters.
Sample mounting (93) to take advantage of strong diffractions at low
V
angles for some compounds requires that the sample be spread in a thin film
over the illuminated area. Filter samples can simply be cut into
2.5 cm x 7.5 cm strips and mounted on glass microscope slides using double
sided adhesive'tape.
Loose particulate matter must first be ground to <40y in an agate
mortar and pestle. Care should be taken to avoid excessive force which
95
-------
might cause sufficient localized heating to alter the sample's chemical
make-up. Fifty to one hundred mg of the ground sample is then ultrasoni-
cally dispersed in 1 to 2 ml of a 1:4 mixture of collodion in alcohol.
The mounting plate is a microscope slide which has been covered with
double-sided adhesive tape. An area of 15 mm x 35 mm is masked off with
masking tape to form a shallow trough. The collodion suspension is
poured into this trough and carefully spread until an even, thin film
is formed in the trough. After the slide dries, it is ready for analysis.
This mounting scheme using the double side adhesive type has reduced the
problems (96) of cracking and buckling of prepared samples. Care still
should be taken to avoid undue handling and excessive humidity or tempera-
ture changes.
Another mounting technique adds the ground particulate matter to amyl r
acetate to form a slurry. This slurry can then be poured into the masked
off area on the microscope slide. Once the amyl acetate is allowed to
evaporate, the sample is ready for analysis. Finally, a thin layer of
petroleum jelly could be spread over the surface of the microscope slide.
The loose particulate matter can then be sprinkled over the greased area.
After obtaining the XRD spectrum, the resultant data is converted
o
into d-spacings in Angstrom (A) units and relative intensities. This
data is then used in searching the ASTM Powder Diffraction file to
identify compounds present. Due to the complex nature of the majority
of environmental diffraction spectra, it is also advantageous to have
available the results of preliminary survey analysis for major elements
and anions to reduce the number of possible compounds. In addition the
potential compound list can be used as a starting point to search the
ASTM file. Identification of a compound is considered valid if the
O
primary and secondary diffraction lines are found with less than ±0.0.5 A
differences.
Quantitative analysis is best accomplished using standards prepared
in silicic acid matrix, to produce the type of amorphous background present
in fly ash. Standards can also be produced from compound additions to
NBS fly ash or a standard addition technique to the sample undergoing analy-
sis. Glassfiber filters presents challenge to developing standards,
96
-------
since particles can penetrate the filter forming a non-homogeneous
dispersion in the glassfilm. One way to surmount this problem is to
place the standards on a glassfiber filter. The apparatus to do that
is shown in Figure 18. It consists of a large (ML) wide mouth jar,
wide enough to accept 7.5 cm open face filterholder and two or more
tapered tubes arranged around the filter. A ground (<40ym) sample of the
standard (or standards) is placed on the bottom of the jar at the focus
of the nozzles.
After the filterholder is loaded with a pre-weighed glassfiber filter,
and connected to the vacuum pump, the pump is turned on and sufficient
flow is maintained to suspend the sample. After a fixed period of time the
filter is removed, weighed, and the surface loading calculated. Addi-
tional standards at different surface loadings are made by varying the
pumping time.
Using these techniques TRW has found a variety of inorganic com-
pounds. In all the environmental samples analyzed numerous diffraction
lines were obtained and XRD pattern interpretation was complicated but
rewarding. Major compounds (>1%) were readily identified, many minor ones,
<1%, identified and quantitated, and detection limits were established for
inorganics of environmental interest in the process streams. Table 21
tabulates the detection limits for several compounds of environmental
Interest (found by experiment).
The application of powder XRD to environmental assessment samples
will continue to produce a usable data base for future inorganic compound
identification efforts.
Table 21. Detection Limits on Commercial XRD Instruments
for Several Compounds
Compound
Estimated Detection Limit
CaS04
As2°3
Hgso4
PbO
Cr203
Mn02
^3-5
97
-------
TO
LAB AIR VACUUM LAB AIR
PUMP
RUBBER
STOPPER
FILTER
OPEN FACE
FILTER HOLDER
WIDE MOUTH
JAR
STANDARD
Figure 18. Powder Dispersing Apparatus
For Preparing XRD Standards
on Glassfiber Filters
98
-------
3.3 SURFACE ANALYSIS USING ESCA AND SIMS
One facet of the bulk characterization of environmental samples is
the study of the surface composition. Obviously those materials that are
present on the surface will be the first to be released into the environ-
ment. The surface can be involved in catalysis and chemical reactions
in the atmosphere, toxic affects in the lung, or leaching of trace
elements in a landfill. For these reasons an understanding of the sur-
face composition of a particle will be important to assess the environmen-
tal impact of particulate matter emissions. The following sections
describe the use of X-ray Photoelectron Spectroscopy (XPS or its more
common name ESCA for Electron Spectroscopy for Chemical Analysis) and
Secondary Ion Mass Spectrometry (SIMS) for surface analysis.
3.3.1 Theory
This section will relate some of the key concepts in ESCA and SIMS
so that an environmental chemist can understand and use the data obtain-
able from ESCA or SIMS. Several good survey articles on ESCA are avail-
able, (98, 99, 100, 101) but for an extensive review of the theory of
ESCA the reader is directed to Siegbahn's work. (102) A good series
of articles on SIMS is contained in Heinrich. (103) For a general survey
on SIMS and surface analysis, the two articles by Evans (104, 105) are
recommended.
Fundamentals - ESCA - An ESCA spectrum results when a beam of mono-
energic X-ray photons (hv) is directed at a sample. The X-rays are
absorbed by the sample's atoms and result in the emission of an electron.
The kinetic energy (Ek) of the emitted electron plus the energy required
to remove it from the atom (binding energy) must equal the energy of the
X-ray photon. In practice an additional energy correction is required for
the work function (()>) of the spectrometer material. Thus, the binding
energy (Eb) can be determined from:
While all electrons with binding energies less than the X-ray photon can
be ejected, not all electrons have the same probability of being ejected.
Consequently, the intensity of an ESCA signal for a given atom will vary
99
-------
depending on the orbital of the electron's origin. Figure 19 shows an
example of the survey spectrum obtained from an FBC particulate matter
sample. Any element above H can be detected; in this example, Ca, 0,
C, Si, Al, and Fe were found. Table 22 compiled from Siegbahn (102) and
Beardon(106) lists the principle ESCA lines by element. The reason that
ESCA is a surface technique is that the mean free path (MFP) for an elec-
tron in a solid is limited. The source X-rays penetrate deeply into the
sample, but only those electrons ejected from near the surface escape to
be analyzed. The remainder suffer energy losses and appear as background
radiation. The MFP of an electron is dependent on its kinetic energy and
the matrix. At the present time some rule of thumb estimates for MFP
are: 5-20A in metals, 15-40A in oxides, and 40-1OOA in polymers. (107, 108)
Depth profile analysis is possible by employing an Ar+ beam to remove
layers of the sample. The sputter rate will depend greatly on the matrix,
so that for highly accurate results standards are essential. Standards
should be made so that an Ar beam operating at standard conditions can
be calibrated for its sputter rate in a given matrix. For ash samples
no procedure has been developed to calibrate beam sputter rates, but
most instrument manufacturers have information on other silicate matrices
which can be applied to fly ash. Unless quantitative measurements on
surface concentration are going to be made, qualitative estimates of the
sputter depth will be sufficient to characterize the sample.
Fundamentals - SIMS - Secondary Ion Mass Spectrometry is the general
name applied to techniques which use a beam of primary ions to produce ions
from the surface for mass analysis. The diameter of the beam used to sputter
the ions from the surface can be spread over several millimeters of
sample, or focused into a narrow beam (2-300 vim). The latter systems are
normally called Ion Microprobe Mass Analyzer (IMMA) to discriminate their
ability to provide depth and high lateral resolution of the elemental
composition of the surface analyzed. For the most part, this discussion
will be directed at IMMA units, since they can be operated in either a
focused (<1 ym resolution) or rastered mode to provide survey composition
information over a wide area.
100
-------
1000.00 900.00 800.00 700.00 600.00 500.00 400.00 300.00 200.00 100.00
BINDING ENERGY, EV
Figure 19. Survey ESCA Spectra of Particulate Matter from
Fluidized Bed Combustor
-------
Table 22. Principle ESCA Peak Binding Energy
for Each Element
Element
H
He
LI
Be
B
C
N
0
F
Ne
Na
Mg
Al
Si
P
S
Cl
A
K
Ca
Sc
Ti
V
Cr
Mn
Fe
Co
Ni
Cu
Zn
Ga
Ge
As
Se
Br
Kr
Rb
Sr
Y
Zr
Nb
Mo
Tc
Ru
Rh
Pd
Energy (eV)
^^H^W4lflV^VWVgVHBIHII^MIBIMHMIBIIBVVBHiaMBBg^hMMHBaMMH0M^^^^^KM
14
25
55
in
188
284
399
532
686
867
1072
1305
118
99
135
164
200
245
294
347
402
455
513
575
641
710
779
855
931
1021
1116
1217
1323
1436
69
89
111
133
158
180
205
227
253
279
307
335
Element
Ag
Cd
In
Sn
Sb
Te
I
Xe
Cs
Ba
La
Ce
Pr
Nd
Pm
Sm
Eu
Gd
Tb
Dy
Ho
Er
Tm
Yb
Lu
Hf
Ta
W
Re
Os
Ir
Pt
Au
Hg
Tl
Pb
Bi
Po
Rt
Rn
Fr
Ra
Ac
Th
Pa
U
Energy (eV)
367
404
443
485
528
572
620
672
726
781
832
884
931
978
1027
1081
1131
1186
1242
1295
1351
1409
1468
487
506
18
25
34
45
50
60
70
83
99
118
138
158
184
210
238
268
299
319
335
360
381
102
-------
In practice a beam of inert (typically Ar) or reactive gas ions (02)
generated in a duoplasmatron are accelerated to 15-25 keV. Using electro-
static lenses, a beam of 2-10 pm at <_ 1 x 10"9 A ion current is produced.
These primary ions as they impact cause the upper atomic layers to be
sputtered or stripped off. The material that leaves is mostly composed
of neutral atoms or molecules, but a small fraction is ejected as positive
or negative ions. These secondary ions are extracted into a double
focusing mass spectrometer with a typical resolution on the order of
300 to 5000 for mass/charge separation. Ion detection is accomplished
using electrical or photographic devices. For surface analysis work
electrical detectors using discrete and continuous dynode multipliers and
the Daly-type detectors(109) are most commonly employed. Signal process-
ing and data manipulation are performed with the aid of dedicated
mini-computers.
An IMMA beam can be used in a stationary mode for localized analysis
or rastered over the surface for survey analysis. When the beam is
rastered over the surface, the mass spectrum is taken in the analogue
recording mode as the sample atoms are sputtered. The ion intensities
are quantitatively determined by counting the mass or masses (elements)
of interest for a preset time. In this mode elemental mapping can be
performed by using the intensity of a specific ion to modulate the inten-
sity axis of an oscilloscope, which is synchronized with the rastering
beam. Figure 20 taken from McHugh (110) shows the Ti and B rich inclusions
near a grain boundary in an inconel alloy sample obtained using this
elemental mapping procedure.
As a natural result of the secondary ion production, successively
deeper layers of the sample are exposed as the beam is rastered over a
given area. Depth profile measurements are performed by monitoring the
mass (or masses) of interest while rastering the primary beam on an area
of at least 5 or 6 beam diameters on a side. In this manner a flat bottom
crater is produced so that by proper gating of the detector counting
electronics, only those ions generated from the flat bottoms are collected
and counted.
103
-------
56
Fe+
52r,+
Figure 20. Elemental Inclusions in Inconel Alloy Using INMA
-------
The depth resolution will depend directly on the beam energy: the
higher the kinetic energy of the primary ions, the more the ions penetrate
and stir the upper atomic layers. Additional factors such as 1) mean
escape depth of secondary ions, 2) recoil implantation, 3) molecular ion
interferences, 4) primary beam induced diffusion of matrix species,
5) non-uniform sputter removal of matrix layers, and 6) implanted primary
ion chemical and lattice damage effects will tend to reduce the depth
resolution of IMMA. While it is theoretically possible to see a monolayer,
practical considerations reduce the resolution to ^25-50 A. For a more
detailed discussion of these problems, see References (111, 112, 113).
Figure 21 is a sample depth profile by IMMA on a sample of fly ash taken
from a coal fired power plant. Only Si, Ca, Fe, Cu and Au are plotted.
Each element monitored was ratioed to Al to normalize the data and to
enhance the response to surface layers. Since a surface layer would
shield the fly ash, the main source of Al, the element to Al intensity
ratio would be increased and produce clearly defined inflection points in
the depth profile curve. The Au was vapor deposited as part of the
sample preparation and shows what a surface layer would look like on SIMS
elemental depth profiles. Fe, Cu, and Si were fairly constant at all
depths, but Ca exhibited an increasing concentration with depth.
Sensitivity - ESCA
The observed intensity of an ESCA signal is a function of the element
and its concentration on the surface of the material studied, The most
important factor in being able to estimate the quantity of an element
present is the efficiency with which the X-rays are absorbed by the
atoms in the sample. If the cross section for X-ray absorption for an
atomic subshell is low, the resulting ESCA sensitivity for those electrons
(element) will be low. Scofield (114) has published a list of elemental
sensitivities and Swingle (115) has normalized them to relative to C Is. In
practice the observed peak intensities are corrected to the relative
intensity ratios in Scofield's table before the abundance or normalized
percentage composition of the surface is calculated. Each element's peak
height (PH ) for a selected electron is measured, then multiplied by the
intensity factor (IF) and a scale correction factor (SF) to obtain a
105
-------
135-OUT-CY-SIMS
CM
O
H-
Q
UJ
O
20 -
0 2760
© = 28/27x35
D = 40/27 x 56
5520 8290
TIME IN SECONDS
11040 13800
0=56/27x9
0 = 64/27x159
A= 197/27x358
Figure 21. Depth Profile Analysis
of Fly Ash Sample Using IMMA
106
-------
normalized peak height (PHn). Thus: PHn = (PHe) (IF) (SF). All the
normalized peak heights are summed and the individual normalized percentage
distribution can be determined:
PH
% Distribution Element X = — — ^ — x 100%
x=l
For the spectra in Figure 19 the normalized surface ( 15-40 A) composition
was: Fe-1%, 0-53%, Ca-10%, Na-1%, S-5%, Si-16% and Al-15%. Most elements
can be detected at 0.1% abundance with a 10-15 minute data acquisition
time, but under favorable circumstances detection limits have gone as low
as 0.001%.
Sensitivity - SIMS - The question of sensitivity is quite complex and
is imminently bound with the processes which generate the secondary ions.
In the simplest sense the generation of secondary ions can be thought to
occur via two ionization processes: 1) kinetic ionization and 2) chemical
ionization (110, 116, 117, 118).
The kinetic process occurs when inert ions are used to bombard a
surface. The transfer of energy during the beam penetration causes
lattice bonds to be broken and the ejection of some atomic electrons into
the conduction band of the material. Most of the ions produced this way
are neutralized before they leave the surface. However, these neutral
atoms retain a great deal of energy and can lose an electron via Auger
or quantum deexcitation processes to produce an ion for mass analysis.
Chemical ionization processes require that a chemically reactive
species be present to reduce the number of conduction electrons available
for the neutralization of ions. Oxygen typically is a good example, since
it forms many compounds which would render the region surrounding the
beam non-conductive. If oxygen is not present in the sample, then it
can be used as the bombarding ion to increase the number of ions emitted.
107
-------
In most materials the chemical ionization mechanism produces the majority
of ions. For this reason oxygen bombardment is most commonly used to
reduce the differences in ion production from a specific element caused
by matrix changes.
The sensitivity of a SIMS instrument will also depend on good col-
lection and detection efficiency. Assuming an average atomic weight of
100, an ionization efficiency of 20% (ions/element), a mass spectrometer
efficiency of 10%, and 6 ions over a few seconds period required for
1Q (104)
detection, a detection limit of 10 grams can be expectedv '. Ioni-
zation efficiencies of the order postulated can occur for positive ions
of alkali metals and negative ions of halogens. For other elements
ion yields of down to 10 can be expected, with the expected reduction
in sensitivity. Correction factors for ionization efficiency, isotope
abundance and matrix are normally included in some form in the software
of the instrument.
Instrumentation - ESCA - Figure 22 (taken from a McRherson ESCA 36
manual) diagrams the key components of an ESCA system. Essentially an
ESCA spectrometer consists of sample chamber with an X-ray source (usually
Mg) which is kept below 10" torr. The electrons ejected from the sample
are analyzed using an electrostatic energy analyzer and are detected by an
electron multiplier. Scan and readout systems used in ESCA Instrumenta-
tion are either of a continuous or incremental type. Most often the
incremental scanning mode is employed so that signal averaging can be
achieved by performing repetitive scans over the energy region of interest.
Most commercial systems employ a dedicated minicomputer to store the data
for later manipulation and display,
Instrumentation - SIMS - Figure 23 shows the key components in an
ARL IMMA. SIMS units use a duoplasmatron for its ion source. This
source uses both electrostatic and magnetic constriction to produce
a high brightness source. The beam is focused using electrostatic lenses
and can be rastered for imaging or depth profile work. Vacuums of 10"8 to
10" torr are commonly found in the sample chamber. The mass spectro-
meter for IMMA units is normally a double focusing (electrostatic and
magnetic) device with a resolution of between 300 to 3000 depending on the
108
-------
ENERGY DISPLAY
SPHERE POWER SUPPLY
(SCAN)
X-RAY
SOURCE
PAPER .TAPE PUNCH
OTHER AVAILABLE
READ-OUT DEVICES
X-Y RECORDER
BULK STORAGE
CARD PUNCH
TO LARGE COMPUTER
Figure 22. Block Diagram of McPherson ESCA 36
-------
ION MICROPROBE MASS ANALYZER
COUNTER
RECORDER
-15 KV
(+15 KV)
TARGET-
+ 20KV
ALIGN DUOPLASMATRON
— BEAM SWEEP PLATES
GAS
SEC. MAGNET
ELECTRIC SECTOR '
SAMPLE
Figure 23. Schematic of SIMS Instrument
-------
instrument. Electrical detectors employing either discrete or continuous
dynode electron multipliers or Daly-type detectors are used. Signal
processing and data manipulation is accomplished using a dedicated
mini-computer.
3.3.2 Sampling Handling/Preparation for ESCA and SIMS
ESCA can analyze gases, liquids and solids, but since it is a vacuum
technique low vapor pressure solids are most easily run. This discussion
will only cover the mounting of loose particles or solids for ESCA
analysis.
A variety of methods for mounting particles is available. The most
common method used is to spread a thin layer of sample over the surface of
a double-sided adhesive tape. Unless the surface is completely covered, the
tape will provide a background signal and under the combination vacuum,
X-ray, and etching beam, it may decompose or deposit an organic film on
the particles. This is not important unless carbon is to be analyzed in
the sample. The main problem is that the tape insulates the sample and
surface charges can be built up as the ESCA electrons are ejected. The
increased surface charge (positive) retards th& ejection-of the electrons,
causing a shift in the spectrum, while the non-uniformity of surface
charge causes line broadening.
There are two recommended approaches based on TRW's use of ESCA and
SIMS. The first is a variant of the double-sided adhesive tape approach
and it is called the "sticky gold" mounting approach. A piece of double-
backed adhesive tape is placed on a sample stage, and then a light film of
gold is vacuum deposited over the tape. A small drop of carbon or silver
paste is applied to an edge of the now gold-coated tape to ensure electri-
cal contact with the stage. Enough sample to completely cover a 4 x 4 mm
area is then placed onto this conductive sticky surface and lightly pressed
with a glass microscope slide to attach it securely. For added conduction
an additional layer of gold or perhaps carbon can be placed on the now
mounted sample as shown in Figure 24.
Another approach (115) is to put an excess of the sample in a folded
indium strip and hand press it into the soft foil. The main advantage
111
-------
SAMPLE
CARBON PASTE
GOLD COVER LAYER
"STICKY GOLD" LAYER
DOUBLE SIDED
SCOTCH TAPE
MOUNTING STAGE
Figure 24. Sticky Gold Mounting Technique
-------
is that a clean unworked surface is presented for analysis. Usually (as
with the sticky gold) the surface is completely covered, so that only the
sample is analyzed.
SIMS like ESCA can have surface charging problems which will affect
the analysis of non-conductive materials. The same mounting procedures
for ESCA can be used for SIMS. It is especially recommended that the
"sticky gold" procedure be employed for SIMS. This procedure was actually
developed for SIMS work by TRW and has been found to eliminate any mount-
ing or charging problems.
3.3.3 Compound Identification
The primary role of ESCA and SIMS in this analysis scheme is to pro-
vide Information on surface concentrations of trace elements. Inherent
in ESCA and SIMS is the ability to perform compound identification opera-
tions. The following sections briefly discuss the compound identification
abilities.
Compound Identification - ESCA
A unique trait of ESCA is the ability to determine the chemical
environment of an atom. ESCA is able to do this because the binding
energies of the core electrons are influenced by the valence electrons
and thus the chemical environment of the atom. If one of these valence
electrons is removed, the amount of shielding of the core electrons from
the nucleus is diminished. The effective nuclear charge experienced by
the core electrons increases, and thus the Eb is increased. In the
simplest case a shift in an ESCA line reflects a change in the oxidation
state of an element. In general any parameter (i.e., oxidation state,
ligand electronegativity, coordination) that affects the electron density
about the atom is expected to result in a chemical shift in electron bind-
ing energy. The shift is frequently on the order of 1 eV per unit change
in oxidation state. Consequently high resolution ESCA can tell the dif-
rt -O —O _2
ference between species like C or C03" and S" , S03" , and S04 . An
example of sulfur chemical shifts is shown in Figure 25, which is a high
resolution ESCA spectrum at 100 A depth for >27 pm particles collected
from an FBC unit. The surface high resolution spectrum showed only S ,
113
-------
170 168 166
BINDING ENERGY, EV
Figure 25. High Resolution ESCA Spectrum Showing Chemical Shift
for S"2 and SO^ Compounds
-------
but after removing the top layer with an Ar+ beam a layer or particle of
S" was uncovered. The surface S+6 was probably due to air oxidation of
this sulfide particle. The same result was found by Werner (119) at
another FBC.
Compound Identification - SIMS - Unlike ESCA, chemical compound
information is not directly obtainable from IMMA. An IMMA, however, can
obtain elemental information from a localized region, to determine the
exact elemental ratio at the surface, in an inclusion, or of a discrete
particle. In many cases the software provided by the manufacturer will
include correction factors applicable to many matrices. If these cor-
rection factors are not applicable to the environmental samples studied,
then an empirical procedure must be employed.
The basic ingredients in this type of procedure include:
• A set of homogeneous standards representing a wide variety of
matrix types.
. **•'
• Relative elemental sensitivity factors derived from standards.
0 Methods to extend these sensitivity factors to other matrices.
• Standardized instrument operating conditions.
A detailed discussion of this approach is found in McHugh (110).
3.3.4 Application of ESCA to Surface Analysis
To illustrate the depth profile capability of ESCA, Figures 26 and 27
show the ESCA analysis performed on particulate samples from a coal-fired
power plant test. The samples were obtained from the inlet and outlet of
an F6D using an EPA Method 5 train with a cyclone and filter. The data
for Ca, C, and S are presented as atom percentages, normalized to 100%,
and ratioed to the Al concentration at 500 A. These are relative values,
not absolute concentrations, since not all elements present in the
particulate matter are included. Ratioing the elemental concentrations
to the Al concentration at 500 A accentuates any surface concentration
and tends to normalize any analysis errors.
115
-------
ESCA RESULTS FROM CYCLONE PARTICULATE
• Go (OUT)
O Co (IN)
• S (OUT)
D S (IN)
C (OUT)
A C (IN)
69% OF INLET PART
17% OF OUTLET PART
97.4% REMOVAL OF >3u
100
200
o
A
300
500
Figure 26. ESCA Depth Profile Analysis of x3ym
Particulate Matter before and after
an FGD.
116
-------
8
z
UJ
01
UJ
Co (OUT)
Co (IN)
S (OUT)
S(IN)
C (OUT)
C(IN)
31% OF INLET PART
83% OF OUTLET PART
75.5% REMOVAL OF < 3* PART
100
400
500
o
A
Figure 27. ESCA Depth Profile Analysis of <3ym Particulate Matter
before and after an F6D.
-------
Relative concentrations for the Method 5 cyclone (particulates >3 ym)
show that the sulfur/aluminum ratios for both the inlet and outlet particles
are high at the particle surface and decrease with depth. This type of
data would be expected if a volatile compound of S were adsorbed or con-
densed on the particles. Note that the sulfur concentration tended to
level out above zero, indicating that the coated particle contained sulfur.
Subsequent XRD analysis confirmed the presence of CaSO^ in the inlet and
outlet cyclone catches. The sulfur rich surface layer was probably formed
by H2$04 condensing on the particles. An opposite effect was seen for the
carbon, where the C/A1 ratio increased with depth, which indicates that
unburned coal particles are present in the inlet cyclone sample.
The relative carbon concentration of the inlet filter (<3 ym) particles
is much lower than that of the cyclone fraction and shows a sharp surface
o
dependence. Interestingly, the C/A1 ratio goes to zero after 400 A for the
inlet samples but in the outlet filter samples remains fairly constant ,
after the initial drop. Once again a substrate containing small amounts
of carbon might be present. In this case <3 ym particles of unburned coal
(concentrated in the outlet filter by removal of the large particles by
the FGD) or CaC03 (a scrubber reaction product) could be present. The
S/A1 ratio indicates a surface layer of sulfur rich material similar to that
found in the cyclone catch.
The Ca/Al ratios for the inlet and outlet cyclone and inlet filter
samples are very similar, all showing a slight depletion at the surface and
roughly equivalent concentrations. The outlet filter shows a higher con-
centration of Ca, which, after 100 A, corresponds nicely with the S
concentration.
Catches from impactor plates and filter also were analyzed by ESCA to
determine relative concentrations at the particle surface as a function of
particle size. An interesting relationship was found when the Ca/S ratio
was determined by selected particle sizes and at various depths, Table 23.
The first two stages exhibited mixed trends that might be due to a mixture
of fly ash and a Ca-S compound coated with H2S04. The particles in the
8.5 and 3.5 ym size fraction appear to be primarily a Ca-S compound with
118
-------
Table 23. ESCA Data - S/Ca Ratio by Particle Size (pm) and Depth (A)
Depth (A)
Surface
100-500
500
53.9
2.78
4.55
••
Average Geometric Aerodynamic Diameter (vim)
20.2
3.23
2.50
8.5
0.83
1.16
3.5
1.35
1.28
1.00
1.8
5.56
1.52
0.93
0.9
7.69
7.14
4.76
0.51
6.25
5.00
0.29
11.11
2.94
(-) ESCA spectra not run
little or no coating of H2S04. The particles in the 1.8 ym fraction
exhibit a strong coating of H2S04 on a Ca-S particle. Finally the last
three stages show a thick coating of H2S04 and may represent H2S04 drop-
lets formed when the H2S04 vapor entering the F6D was rapidly cooled in
a stream of fine particles.
Identification of these Ca-S compounds might have been made using the
ESCA system in the high resolution mode to identify sulfur's oxidation
state. While this approach was not taken, additional XRD identification
work on the bulk samples identified CaS04 and CaSO.,.
3.4 Summary of Bulk Composition Characterization
The methods presented in this section are the primary tools used for
direct inorganic compound identification. Of the three discussed in
Section 3.0, XRD is the method most specific for compound identification;
however, it has several weaknesses. XRD cannot identify non-crystalline
compounds, and its sensitivity is normally limited to compounds present
at or above 0.1%. In order to offset these fundamental weaknesses, FTIR
and ESCA (SIMS) have been grouped with XRD in this phase of the analysis
scheme. FTIR provides functional group (anion) identification, some com-
pound identification (crystalline and non-crystalline), and specialized
analysis techniques (subtraction/addition of spectra). ESCA provides
specific information on the oxidation state of the elements, but its
119
-------
primary use is to provide elemental and compositional information about
the surface of the material studied. Figure 28 diagrams the relationship
and the application of these three techniques.
One must always compile and interpret the data of previous analyses
and use it to direct later analyses. In the Initial Sample Characteriza-
tion phase, the Level 1 data were used to develop a list of compounds for
analysis. In the Bulk Composition Characterization phase, the accurate
quantitative elemental and anion data determined in the initial sample
characterization phase were used to estimate the concentration of potential
compounds. These concentration estimates determine whether a compound
can be seen by XRD or FTIR. For example, to confirm the presence of a
suspected compound, the region around its primary diffraction lines could
be step scanned for a longer period of time to improve sensitivity.
Similarly, the number of FTIR scans taken can be increased to improve the
signal to noise ratio and the sensitivity. This is not to imply that
survey scans are not run. Survey scans of the sample are necessary to
provide information not found by other methods, and to insure the complete-
ness of the overall analysis. Whenever compounds are known or suspected,
however, a focused analysis is used to confirm or deny its presence.
It is suggested that an FTIR scan be run prior to the XRD analysis,-,
so that the compound information derived is available to aid the inter-
pretation of the XRD spectra. Figure 28 shows two pathways: a sample
scanned as received with an emphasis on far IR spectral data to identify
transition metal anions and a surface study using acid (or water) extrac-
tion to remove surface compounds. Subtraction of the latter spectra from
the former will emphasize those bands (compounds) associated with the
surface which can be compared with the ESCA results. Further uses of the
subtraction capabilities of FTIR were noted in Section 3.1.
After the FTIR spectra are run and interpreted, the XRD work can
begin. The goals of the XRD studies are threefold: (1) study the overall
make-up of the sample, (2) identify compounds from the list refined by
the Initial Sample Characterization, and (3) confirm the presence of any
compounds deduced from FTIR or from the previous analyses. As mentioned
120
-------
STUDY BULK
CHEMICAL
COMPOSITION
ASSIGN FftOIAMUTYTOI
SEE POTENTIAL COM-
POUNDS WITH SPECIFIC
METHOO
ALLOWS
MATCH UP OF
METHOD WITH COMPOUNDS
BASED ON CONCENTRATION
ESCA
EXTRACT ALIQUOT
OF SAMPLE WITH
AOUAREGIA
PERFORM FAR IR SCAN FOR
TRANSITION ELEMENT
ANIONS IN BULK OF SAMPLE
RECORD
SPECTRA
IN FAR IR
SUBTRACT INSOLUBLES
SPECTRA FROM
ORIGINAL SAMPLE
STUDY SURFACE TRACE
ELEMENT COMPOSITION
OXIDATION STATES,
CHEMICAL
ENVIRONMENT
/PRESENCE OF 7 /
/ TRANSITION / /
I ELEMENT ANIONS / /
PRESENCE OF
QUANTITAR
SPECIFIC
ANIONS
WET CHEMICAL OR
INSTRUMENTAL
ANION TESTS
/LISt POSSIBLE NEW /
COMPOUNDS /
FOUND /
L
LIST IDENTIFIED
COMPOUNDS WITH
ESTIMATED .,
CONCENTRATION
HAVE
ALL DMEG
COMPOUNDS
EXCEEDING MA
VALUES BEEN
FOUND
7
5 FURTHER
YSIS COST
JUSTIFIED FOR
UNASSIGNED
ELEMENTS
Figure 28.
Flowchart for
Bulk Sample
Characterization
121
-------
earlier, these goals are best accomplished by using both a survey and spot
scan to provide overall and high sensitivity information. Once a compound
has been identified, XRD can be used to quantitate the material.
Surface studies using ESCA or SIMS are performed once the FTIR and
XRD studies are completed. Surface composition data is important to:
(1) determine the catalytic activity of the particle, (2) study the forma-
tion mechanism, or (3) determine the ultimate toxicity of a particle.
Surface studies are highly recommended primarily for the last reason,
since it has been shown that many trace elements exhibit a surface
dependence. The reasons postulated for this result involve the volatility
of some compounds and their later condensation on the surface of the
o
already cool fly ash particles. The end result is a surface (25-50 A)
concentration as high as several percent when the bulk analysis
indicate ppm concentrations. ESCA and SIMS should be used to study this
and other surface-related phenomena.
Once the FTIR, XRD and ESCA/SIMS data have been interepted, the list
of DME6 compounds is reviewed and the identified compounds removed. By this
time the majority of compounds sought should have been identified or ruled
out at some level of sensitivity. It is possible that some compounds of
high interest were not found or that a question about the structure of the
individual particles was raised. If a decision is made to continue the
study of the sample, the next level of research is at the single particle
level. Section 4,0 describes three techniques for the study of individual
particles.
122
-------
4. INDIVIDUAL PARTICLE ANALYSES
Once the analyst has completed both the initial sample characterization
and the bulk sample characterization, information will be available rang-
ing from gross physical characteristics through the identification of
selected compounds. In the event that not all of the MATE compounds
flagged from the SSMS results have been identified or ruled out, additional
identification work can be performed using single particle analysis
techniques. Analytical techniques (summarized in Table 24) which are
suggested for identification of individual particles include:
• Scanning Electron Microscopy with Energy Dispersive X-Ray
Spectrometry (SEM-EDX)
t Electron Probe Microanalysis (EPMA)
• Transmission Electron Microscopy with Selected Area Electron
Diffraction (TEM-SAED)
Extensive work (121 to 125) has been done in this area. The follow-
ing sections will summarize the use of SEM-EDX, EPMA and TEM-SAED for
single particle analysis.
4.1 Introduction to Electron Microscopy
When a collimated electron beam is focused on a sample, several
phenomena, shown in Figure 29, can occur. As the beam impacts the sample,
low energy (1-50 eV) secondary electrons are given off at the point of
impact. Variations in the secondary electron emission intensity across
the surface can be used to form an "image" of the surface, which is the
technique of Scanning Electron Microscopy (SEM). The beam may cause
ionization of an atom in the sample by knocking one of the inner shell
electrons out of its orbital. In a manner similar to XRF, the inner-
shell rearrangement may cause emission of an X-ray photon. Using an
energy dispersive X-ray (EDX) or wavelength dispersive fluorescence
spectrometer to detect these X-rays will permit the identification of the
elemental composition of the area under the beam. If the sample is suf-
ficiently thin, the electron beam will pass through the sample. Varia-
tions in the scattering cross-section of the sample cause the beam to
penetrate at different intensities which is the basis for Transmission
123
-------
Table 24. Summary of Methods Used for Individual Particle Characterization
Analysis Method
ro
Scanning Electron
Microscopy (SEH)/
Energy Dispersive
X-Ray Spectrometer
(EDX)
Principle of Operation
Electron Probe
Microanalysis
(EPMA)
Transmission
Electron Micro-
scopy (TEM) -
Selected Area
Electron Diffrac-
tion (SAED)
Analysis
In SEM:
- The specimen is swept by electron beam
- Secondary electron emission intensity
is recorded
- The signal modulates brightness of
oscilloscope beam, producing an
image
- Morphological characteristics of the
specimen are determined from the
image
Using SEM in conjunction with EDX:
- The secondary X-rays produced are
monitored and individual elements
present in the sample are identified
and quantified
EPMA is used for elements above>atomic
Number 6
Small energetic electron beam impinges
on surface of specimen, causing charac-
teristic X-ray emissions which are
analyzed by wavelength dispersion
techniques
For qualitative analysis: Wavelength
positions are used
For quantitative analysis: Peak heights
(intensity ratios) are measures on both
the unknown and on a standard of known
composition
In TEM, electron beam is Impinged on a
thin film of sample and the resultant
transmitted electron beam is observed
and recorded
quantitative analysis using TEM is .
superior to SEM because:
• Smaller samples can be observed and
Identified
• Chemical species such as asbestos are
more reliably identified, (TEM's
selected area electron diffraction
analysis is more dependable than SEM's
elemental analysis)
Information Derived
SEH system provides high resolution morpho-
logical information on individual particles
The EDX attachment allows Identification of
individual elements in the particle
Specific X-ray fluorescent wavelengths can
be monitored to produce elemental distri-
bution of the element (NOTE: These plots
are especially useful for particles com-
posed of various occluded materials)
SEM information is a valuable adjunct to
the PLM, especially for particles <0.5 v
(SEM magnifications are routinely in excess
of 50.000X)
Obtains single participate elemental
composition of elements from carbon and
above
Many Instruments use wavelength dispersive
X-ray spectrometers and can resolve
elements S through N1
Provides high resolution photographs
Produces single particle X-ray diffraction
pattern
Compound Identification Procedure
SEM's high resolution images often
allow particle identification
The EDX information can be used
to determine elemental ratios and
the exact composition of the
particle
Compound Identified by elemental
ratios
EPMA essential when elements C
through Na are present since
SEM-EDX does not see those
elements
Identifies crystalline compounds
by their characteristic diffrac-
tion patterns
Limitations
Relatively long counting times are
required for trace elements, but the
EDX instrument stability limits
counting time to 10 or 15 minutes
At high count rates, peaks may broaden
Particles in close proximity may
interfere and preclude unambiguous
analysis
EDX does not resolve elements from
S to N1 very well
Quantitative work depends on having
suitable standards
Identification possible only for
particles containing discrete, com-
pounds rather than a homogeneous
mixture
Better quantitative results when
standards are used whose composition
closely matches the specimen
Only crystalline material can be
Identified
Magnetic Density
Gradient Selective
Dissolution
Separations
Magnetic separation is used to remove
any magnetic material from rest of
sample
In density gradient separations, parti-
cles are floated in solvents of known
density. Particles are separate by
differences in their density
Sample is extracted using selective
dissolution (with different solvents)
Determines particles specific density mag-
netic characteristics, and solubility in
solvents
Main use is to reduce complex systems
Separating complex mixture into
simpler fractions aids compound
identification
Can use Information on particle
density, solubility, and magnetic
properties to identify compounds
Density gradient will only separate
discrete particles; occluded material
will have an average density
Many compounds have solubilities in
organic solvents used in density
column
Selective dissolution scrambles the
compounds unless specific compound
solvent systems can be found
-------
BACKGROUND
ELECTRON x_^y
BACKSCATTERED **** PHOTONS (EDS, EPMA)
ELECTRONS
N
BREMSSTRAHLUNG
VISIBLE
LIGHT
ro
01
ELASTICALLY
SCATTERED
ELECTRONS
SECONDARY
ELECTRONS (SEM)
AUGER ELECTRONS (AES)
INELASTICALLY
SCATTERED AND
TRANSMITTED (TEM)
J ELECTRONS,
* DIFFRACTED
ELECTRONS (SAED)
Figure 29. Interactions of an Electron Beam with a Sample
-------
Electron Microscopy (TEM). Some of the incident electrons are
backscattered causing a background signal. If the material is crystal-
line, a diffraction pattern may be formed similar to X-ray power dif-
fraction except that the incident electrons are diffracted. Comparison
of this pattern with patterns of known compounds may permit identification
of the crystalline material. This technique of selected area electron
diffraction (SAED) has for example been applied to the identification of
asbestos (120) particles in air and water samples. The above techniques,
in various combinations, have been successfully applied to the analysis
of a wide variety of environmental samples (121, 122). The two most common
of these combinations, TEM-SAED and SEM-EDX, will be discussed below, as
well as the technique of electron probe microanalysis. Use of electron
microprobe analysis, complemented by scanning electron microscopy, trans-
mission electron microscopy, and optical microscopy, will greatly increase
the analyst's knowledge of the particles which make a sample.
' * *
4.2 Scanning Electron Microscopy-Energy Dispersive X-Ray Spectrometry
An electron optical system produces a collimated electron beam
focused on an area of up to one micron in diameter and penetrates to a
depth of 2-5 nm. The beam strength and density of the sample affect
resolution, since increased beam strength or decreased density permits the
primary electron beam to penetrate deeper into the sample (increasing the
volume range). Similarly, the secondary electron intensity is-affected by
sample composition and density, and such instrumental constants as the
angles of the incident beam and the detector relative to the sample
(i nstrument geometry).
Secondary electrons emitted when the beam strikes a particle are
monitored and displayed on an X-ray cathode ray tube (CRT). During a
scan of the sample, the beam is deflected and the CRT display synchronously
moved by a coordinated magnetic deflection system so that the rasters of
the CRT screen correspond to the rasters of the beam. In this way a visual
image is produced which can be displayed, photographed, and stored on
magnetic tape. The output of an SEM is a visual image of the surface
topography of the sample. Resolutions of over 10 nm are available with
126
-------
typical magnifications available from 7X-240.000X. In addition to the
enhanced magnification of SEM systems relative to optical microscopy,
the depth of field of an SEM is 500 times that of an optical microscope.
The information available from such an image is similar to that of an
optical microscope-shape, roughness, size and size distribution of particles,
and homogenity of the crystal or particle. McCrone (19), besides the optical
micrographs, has a complete SEM-EDX library, so that morphology can be
used to identify particles below the range (
-------
complicated by its non-linear energy distribution and the presence of
discontinuities in the background. By using the techniques of frequency
filtering, the background can be successfully removed without significant
increase in data reduction time. Removal of spectral interferences can be
performed during computer reduction of the data from knowledge of the
intensity ratios of major and minor peaks of each element identified.
This is facilitated by the Gaussian shape of the peaks and the fact that
typical interferences result from Kg overlap with a Ka, so that the
interference will be minor except when the elements are present in widely
different concentrations.
Sample Preparation, SEM-EDX
Particulate matter for SEM-EDX analysis can be mounted on a number
of types of material. Typically beryllium and carbon stages are
used, since they produce no X-ray background. Other common mounting
stage materials are aluminum and gold. Regardless of the stage material,
a layer of Au, C or Pt is vapor deposited to ensure that electrical charge
is not built up during the analysis. Charge build-up will cause the
image contrast to be too white and might cause the particles to be
repelled from the surface of the mounting stage. The latter possibility
occurs when highly non-conductive particles are analyzed, and can be
prevented by using the "sticky gold" mounting technique.
4.3 Electron Probe Microanalysis (EPMA)
Electron microprobe analysis, is very similar to SEM-EDX. Whereas
the primary objective of SEM is the determination of surface morphology
and has been modified to permit EDX, the objective of EPMA is the chemical
analysis of a sample using EDX. As a consequence, several parameters
vary significantly, although the hardware (electron gun, magnetic focusing,
Si(Li) detector, multi channel analyzer/computer) is the same.
The sensitivity of
-------
in EPMA. Electron microprobes are equipped with electron optical systems
to permit location of the electron beam. Much lower magnification (approxi-
mately 3000X maximum) is attained than with the SEM due to image broadening
at the higher beam currents employed in EPMA, and therefore much lower
resolution is possible.
Quantitative electron microprobe analysis is best performed on flat,
polished surfaces mounted in fixed geometries where emission of secondary
X-rays (due to backscattered incident electrons), reabsorption of X-rays
by the sample and secondary X-ray emission from primary X-ray excitation
are minimized. Using a lithium-drifted silicon solid state detector,
multichannel analyzer and dedicated minicomputer, major constituents
(greater than 0.1% of the sample) may be analyzed individually or simul-
taneously in a typical time of 100 seconds with times of 1000 seconds
typical for trace constituent analysis. The elements from oxygen to
uranium can be determined. With concentrations greater than 10% of the
sample, relative errors of ±5% can be expected (124) for most elements and
higher relative errors, usually positive, for components less than 10% by
weight. Detection limits depend on a number of parameters, including the
orientation of beam; sample and detector and other geometric factors; data
reduction procedure employed; and the morphology and composition of the
sample itself. Typically, at longer count rates (1000 seconds) samples
may be analyzed for components of several parts per million or more in a
bulk sample. The method has been applied to analysis of ores and
minerals (125), fly ash, atmospheric particulate aerosols, particles
deposited in lung tissue and a large variety of other environmental samples.
Sample preparation is similar to that for SEM-EDX.
4.4 Transmission Electron Microscopy-Selected Area Electron Diffraction
Transmission electron microscopy provides greater imaging capability
than SEM and better detection of small particles due to the ability to
control brightness and contrast. TEM also provides a field of view 2-3
times that of SEM. It also has the requirement, however, that samples
must be thin enough to permit transmission of the electron beam. For such
samples, TEM has the advantage that the internal as well as surface
features can be observed since variations in the TEM image result from
129
-------
deficiencies in the scattering ability of the material sectioned.
Biological or organic samples, thin metallic or nonmetallic films, for
example, may be investigated. For samples opaque to the electron beam,
a carbon film image of the surface may be made and the film subjected
to TEM for observation of surface features of the object. In this manner
the surface of samples too large or thick to be analyzed by SEM or normal
TEM may be observed. Liquid solutions or suspensions may be evaporated
onto a TEM grid for analysis.
As the electron beam passes through a crystalline sample, the
electrons may be diffracted by the lattice. The diffraction pattern of a
single crystal may be sufficient to identify the compound by comparison
with the patterns of known compounds. The size of the crystal is critical,
however. Typically, crystals below approximately 0.3 vim diameter give
scattering intensities too low to permit accurate identification. At
sizes of approximately 0.8 ym or greater, the crystal is opaque to the
electron beam.
4.5 Summary of Individual Particle Characterization
Individual particle characterization is conducted to determine the
specific morphological and chemical composition of the individual particles
in a sample. It is also possible to use these techniques for compound
identification of surface features on the particles themselves. SEM-EDX,
EPMA and TEM-SAED should be applied sequentially (Figure 30) to allow the
assessment of the results in light of the added cost and the benefits
derived. The general approach for individual particle characterization
using SEM-EDX and EPMA is to use accurate trace elemental mapping of a
single particle to determine the empirical formula. SEM-EDX can also add
high resolution morphology information on small particles (<0.5 ym) or on
surface features. EPMA produces similar results for slightly larger
particles (>0.2 pm), but has the added advantage of increased sensitivity
and wider elemental range (> atomic number 6).
If a particle appears crystalline and cannot be identified by elemental
mapping, then TEM-SAED can be employed. TEM-SAED should only be applied
to the small particles or thin fibers. If surface features on a large
particle are oriented properly (in relief), TEM-SAED is an excellent
method to identify them.
130
-------
OBTAIN DETAILED MORPHOLOGICAL
INFORMATION, SINGLE
PARTICLE ELEMENTAL
SCAN AND ELEMENTAL RATIOS
/ESTABLISHED ELEMENTAL/
f RATIOS FOR SINGLE 7
PARTICLE
LIST COMPOUNDS
WITH CONCENTRATION
ESTIMATE BASED ON
ELEMENTAL VALUES
IS
FURTHER
ANALYSIS
JUSTIFIED FOR
UNASSIGNED
ELEMENTS
ELEMENTAL
RATIOS ESTABLISHED
FOR SINGLE PARTICLES
FOR ELEMENTS SC
LIST COMPOUNDS
WITH CONCENTRATION
ESTIMATE BASED ON ;
ELEMENTAL VALUES / YES
CAN
UNAS-
SIGNED DMEG
IPOUNDS BE
DENTIFIED
THESE
RATIOS
7
YES
/ LIST UNASSIGNED
/ FRACTION OF
! KNOWN ELEMENTAL
COMPOSITION
ELEMENTAL
DATA USED TO
QUANTIFY
BASED ON SOLUBILITY
AND ELEMENTAL DATA
SELECT SEPARATION
SCHEME
METIC
ATI ON
1
DENSITY
GRADIENT
SEPARATION
SELECTIVE
DISSOLUTION
SEPARATION
1
Figure 30. Recommended Sequential Application of Individual
Particle Techniques.
131
-------
4.5.1 Application of Electron Microscopy to Experimental Samples
To further illustrate single particle analysis, overflow bed materials,
sludge, and particulates from a fluidized bed combustor (FBC) were analyzed
by SEM-EDX, EPMA and TEM-SAED (44). Uncoated samples were mounted on gold
plates for SEM-EDX and subsequent EPMA. Samples for TEM-SAED were mounted
on copper mesh and coated with carbon. The instrument used for SEM-EDX
was a JEOL SEM equipped with the Princeton GammaTech EDX unit.
The samples were fairly uniform in composition with the major elements
identified as calcium and silicon with traces of elements such as titanium
and potassium. Iron was present in the instrument background and could not
be detected in the sample.
Figure 31 is a SEM-EDX micrograph of sludge material from the FBC.
These particles were very crystalline and irregular in nature. There are
some spheres present with highly irregular porous surfaces. The spheres
(2a) range in size from 10 to 40 ym. The dimensions of the thick appearing
fused material (3b) range up to 100 x 60 ym, thin flakes (Ic) up to 40 x 90
ym. The spheres (2a) are composed of calcium. The fused crystalline material
in area Ic is predominately silicon with iron present, and in area 3c silicon
is a major component, with calcium also present (fused in the silicon matrix).
The EPMA was conducted on the same samples mounted and imaged for the
SEM-EDX. The instrument employed was an Applied Research Laboratories Model
EMX. The backscattered X-rays produced when the sample was irradiated were
imaged. This was done to verify areas which were higher in a particular
element and to establish the association between those elements (Ca and S)
found in prior analyses. Figure 32 shows the portion of the sample rastered
and the backscatter image for Ca, 0, Si, S, Fe, and K. The elements identi-
fied include calcium, oxygen, silicon and sulfur as major (>10%) components
iron and potassium as medium (5-10%) components, and aluminum, magnesium
chlorine, and titanium as low (1-5%) or trace (<1%) components. The EPMA
already showed that Ca and Si exist in discrete areas, and that S showed no
definite particle preference. Although in this application an area was
rastered rather than a single particle, the technique of elemental mapping
can be applied to single particles just as well.
132
-------
Figure 31. SEM-EDX at 3000X of Sample S-2-5
133
-------
FLYASH
S-2-5
Calcium,
red Image
neon, Ka, Backscattered Image
Oxygen, Ka, Backscattered Image
Sulfur, Ka, Backscattered Image
Iron, Ka, Backscattered Image
Potassium, Ka, B*ckscattered Image
Figure 32. EPMA Elemental Imaging of
Sample S-2-5
134
-------
Transmission electron microscope analysis with selective'area, electron
diffraction was employed to assess the uniqueness of the individual particles
observed in the SEM-EDX. Combining the information derived from SEM-EDX and
EPMA can aid'the analyst in describing the total number of the various species
present. Many substances which appear essentially identical in elemental
composition as measured with the electron probe will be determined by TEM-
SAED to have unique diffraction pattern, allowing the identification of the
material present.
The TEM-SAED system used in this portion of the inorganic work was an
RCA Model EMU-3H. It had a resolution of 10 A and a 100 kV beam. The
accuracy of the determined d-spacing is +Q.5. Samples were mounted on a
-..
copper mesh coated with carbon. This provides for electron transmission
and conductance. The TEM images are not of good quality due to the sample
o
thickness, > 1500 A. In most cases thin areas of the sample had to be
chosen for electron diffraction analysis rather than uniquely appearing
crystalline material. Nevertheless sample S-2-5 (Figure 33) is a good
example of TEM/SAED application. In this case a definite fine crystalline
material was imaged, isolated, and identified as Si02- These fine "partic-
ulates were less than one micron in size.
135
-------
TEM Image, Sample S-2-5.
Thin area analyzed using
SAED.
Electron Diffraction Pattern
Area Analyzed
Determined
d-spacings
4.41
3.14
2.64
2.28
1.93
1.48
Compound
Identified
Si02
Literature
d-spacings
4.26
3.34
2.46
2.28
1.82
1.54
Figure 33. TEM-SAED of Fibrous Material in S-2-5
136
-------
5.0 SUMMARY
The Level 2 analysis procedures listed in this manual were selected
and integrated into a scheme of analysis in order to be able to charac-
terize completely most inorganic environmental samples. This analysis
scheme will determine:
• Morphology of particles
• Thermal stability data
• Quantitative elemental data
• Quantitative anion data
• Direct identification of specific compounds
0 Identification of compounds present on the surface of a particle
• Elemental depth profile data
• Valence state information on selected elements
• Elemental composition of individual particles
• Identification of surface features.
Even with these successes, the analytical paths presented in this
document must be considered evolving ones. The main difficulty with
inorganic Level 2 analysis lies in the direct measurement of species
present at concentrations less than *1% but greater than trace H ppm).
This measurement gap is not filled by any one compound isolation and
quantitation technique. Future work by IERL should be applied in that
measurement area and toward improving the specific procedures used in
each method application.
137
-------
6.0 REFERENCES
1. Lentzen, D.E., Wagoner, D.E., Estes, E.D. and Gutkneeht, "IERL/RTP
Procedures Manual: Level 1 Environmental Assessment" (Second
Edition), EPA-600/7-78-201, January 1979.
2. Hamersma, J.W. and Reynolds, S.L., "Field Test Sampling/Analytical
Strategies and Implementation Cost Estimates: Coal Gasification
and Flue Gas Desulfurization," EPA-600/2-76-093b, April 1976.
3. Cleland, J.G. and Kingsbury, G.L., "Multimedia Environmental Goals
for Environmental Assessment: Volume II MCG Charts and Background
Information," EPA-600/7-77-136b, November 1977.
4. Vandegrift et al., ParticUlate Pollutant System Study, Volumes I, II
and III. Midwest Research Institute, PB-203-521, May 1971.
5. A. Stern, ed. Air Pollution, Volumes I, II, and III, Academic Press,
New York, 1968. ,,
6. J. Daniel son, ed. Air Pollution Engineering Manual, U.S. Environmental
Protection Agency, R.T.P., May 1973.
7. A.J. Forney et al., "Trace Element and Major Component Balances
Around the Synthane P.D.V. Gasifier," P.E.R.C., Pittsburgh,
Pennsylvania, PERC/TPR 75/1, August 1975.
8. C. Schmidt, et al., "Mass Spectrometric Analysis of Product Water
from Coal Gasification," U.S. Department of the Interior BOM/TPR/86,
December 1974.
9. C. Burklin, et al., "Control of Hydrocarbon Emissions from Petroleum
Liquids," Radian Corporation, EPA Contract No. 68-02-1312 T1Z,
September 1975.
10. W. Fulkerson ed. et al., "Ecology and Analysis of Trace Contaminants,"
ORNL Progress Report, October 1973 to September 1974.
11. T. Hutchinson (Coordinator) "International Conference on Heavy Metals
in the Environment," Toronto, Ontario, Canada, October 1975.
12. H. Anderson and W.Wu, "Properties of Compounds in Coal Carbonization
Products," B.O.M. Bulletin 606, U.S. GPO, Washington D.C.: 1963.
13. J. Dealy and A. Kill in, "Engineering and Cost Study of the Ferroalloy
Industry," Office of Air and Waste Management, EPA No. 450/2-74-008
(PB-2360762), May 1974.
14. W, Beers "Characterization of Claus Plant Emissions," Process
Research, Inc., Cincinnati, Ohio, EPA Contract No. 68-07-0242T,
April 1973.
138
-------
15. H. Lebowitz, et al., "Potentially Hazardous Emissions from the
Extraction and Processing of Coal and Oil," Battelle Columbus Labora-
tories, NERC, PB No. 241-803, April 1975.
16. Maddalone, R.F. and Quinlivan, S.C., "Technical Manual for Inorganic
Sampling and Analysis," EPA-600/2-77-024, January 1977.
17. Kolnsberg, H., "Technical Manual for the Measurement and Fugitive
Emissions," EPA-600/2-76-089a, 1976.
18. West, P.W., "Polarized Light Microscopy," The Chemist Analyst, 35, 28
(1945) .
19. McCrone, W.C. (ed.), "The Particle Atlas," Ann Arbor Science
Publishers, Inc., Ann Arbor, Michigan, 1967.
20. Chamot, E.M., and C.W. Mason, "Handbook of Chemical Microscopy,"
New York, John Wiley and Sons Inc., 2nd ed., 1939, 2 volumes.
21. Leavitt, C., Arledge, K., Shih, C., Orsini, R., Hamersma, W.,
Maddalone, R., Beimer, R., Richard, G., and Yamada, M., "Environmental
Assessment of Coal- and Oil-Firing in a Controlled Industrial Boiler,"
Volumes I, II and III, EPA-600/2-78a, b, c, August 1978.
22. Wendlandt, W.W., "The Development of Differential Thermal Analysis
in the United States, 1887-1976," Amer. Lab. 9(1). 59 (1977).
23. Seyler, R.J. and Kalbfleisch, E., "Instrumentation for Multiple
Digital Differential Thermal Analysis," Amer. Lab., 10(1), 15,
(1978).
24. Pella, E. and Mebriloni, M., "Temperature Measurements with a
Differential Calorimeter," J. Ther. Anal., 3_, 343 (1971).
25. Flynn, J.H., "An Analytical Evaluation of Differential Scanning
Calorimetry (DSC)," NBS (U.S.). Spec. Pub!., 388, 119 (1970).
26. Morie, G.P., Powers, T.A. and Glover, C.A., "Evaluation of Thermal
Analysis Equipment for the Determination of Vapor Pressure and Heat
of Vaporization," Thermal Acta, 3_, 259 (1972).
27. Ortiz, L.W. and Rogers, R.N., "Simplification of Differential
Temperature Calibration and Emittance Measurements in Scanning
Calorimetry," Thermo. Acta, 3_, 383 (1972).
28. Brennan, W.P., Miller, B. and Whitwell, J.C., "Rate of Change of
Reference Temperature in DSC and DTA," Thermo. Acta, 2_, 354 (1971).
29. Mita, I., Imar, I. and Kambe, H., "Determination of Heat of Mixing and
Heat of Vaporization with a Differential Scanning Calorimeter,"
Thermo. Acta, 2_, 337 (1971).
139
-------
30. Mulik, J.D., Todd, 6. Estes, E., Puckett, R., Sawicki, E., and
Williams, D., "Ion Chromatographic Determination of Atmospheric
Sulfur Dioxide," Ion Chromatographic Analysis of Environmental
Pollutants, Sawicki, E., Mulik, J.B., and Wittgenstein E., Eds.
Ann Arbor Science, Ann Arbor, Mich., p. 23-40, (1978).
31. Small, H., Stevens, T.S., and Bauman, W.C., "Novel Ion Exchange
Chromatographic Method Using Conductimetric Detection." Aiial. Chem.
47, 1801 (1975).
32. Snyder, L.R. and Kirkland, J.J., Introduction to Modern Liquid
Chromatography. Wiley-Interscience, New York, p. 283, 1974.
33. Acciani, T.R., and Maddalone, R.F., "Chemical Analysis of Wet Scrubbers
Utilizing Ion Chroma tography," EPA-600/7-79-151, July 1979.
34. U.S. EPA, "Methods for Chemical Analysis of Water and Wastes,"
EPA-600/4-79-020, March 1979.
35. Smith, R.C., Van Loom, O.C., Knechtel, J.R., Fraser, J.L., Pitts, A.E.,
and Hodges, A.E., "A Simple and Rapid Hydride Generation-Atomic
Absorption Method for the Determination of Arsenic in Biological,
Environmental and Geological Samples," Anal. Chem. Acta, 93, 61 (1977).
36. Kneip, T.J., "Tentative Method of Analysis is for Elemental Mercury
in the Working Environment by Collection on Silver Wool and Atomic
Absorption Spectroscopy," Health Lab. Sci., 1£, 158 (1975).
37. Dickinson, G.W., and Fassel, V.A., "Emission Spectrographic Detection
of the Elements at the Nanogram per Milliliter Level Using Induction-
Coupled Plasma Excitation," Anal. Chem., 41_ (8), 1021 (1969).
38. Scott, R.A., Fassel, V.A., Kniseley, R.N., and Nixon, D.E.,
"Inductively Coupled Plasma - Optical Emission Analytical
Spectrometry," Anal. Chem. 46(1), 75 (1974).
39. Larson, G.F., Fassel, V.A., Scott, R.H., and Kniseley, R.N.,
"Inductively Coupled Plasma - Optical Emission Analytical
Spectrometery. A Study of Some Interelement Effects," Anal. Chem.,
47(2), 238 (1975).
40. Olsen, K.W., Haas, W.J., and Fassel, V.A., "Multielement Detection
Limits and Sample Nebulization Efficiencies of an Improved Ultrasonic
Nebulizer and Conventional Pneumatic Nebulizer in Inductively Coupled
Plasma - Atomic Emission Spectrometry," Anal. Chem.. 49.(4), 632 (1977).
41. Fassel, V.A., and Kniseley, R.N., "Inductively Coupled Plasma - Optical
Emission Spectroscopy," Anal. Chem.. 46_, 1110A (1974).
42. Kahn, H.L., Smith, S.B., and Schleecher, R.G., "Background and
Development in Plasma Emission Spectroscopy," Amer. Lab.. 11 (8), 65
(1979).
140
-------
43. Ronan, R., "Simultaneous Analysis of Liquid Samples for Metals by
inductively Coupled Argon Plasma Atomic Emission Sper-1-
\ES), U.S. E.P.Ao, Region 5, Central Reaional
), Illinois. y
44- iu1dli- Bed combustor sampies'H
45. GallorimYM., Greenberg, R.R., and Gills, I.E., "Simultaneous
Determination of Arsenic, Antimony, Cadmium, Chromium, Copper and
Selenium in Environmental Material by Radiochemical Neutron Activation
Analysis," Anal. Chem. 50_ (11): 1479 (1978).
46. Gills, I.E., and McClendon, L.T., "Role of Neutron Activation Analysis
in the Evaluation of Sampling, Storage and Analysis of Samples for the
o|P0??LSvlronmental Bank1n9 System," J. Radioanal. Chem. 39 (1),
cob, I 1 977). — — ^— — — — — .
47. Gladney, E.S., Hensley, W.K., and Minor, M.M., "Comparison of Three
Techniques for the Measurement of Depleted Uranium in Soils," Anal.
Chem. 50(4). 652 (1978). -
48. Heeoner, J., "The Reliability of Radiochemical and Chemical Trace
Analyses in Environmental Materials," Electr. Nucl . Technol . , 22, 1
(1977). ' —
49. Kuskaya, Y., Tsuji, H., Tamori, Y., Saqawa, T., Ohmori, S., Imai, S.,
and Ozaki, T., "Neutron Activation Analysis of Biologically Essential
Trace Elements in Environmental Speciments Using Pyrrol idinedithio-
carbamate Extraction," J. Radio-Anal. Ghem. 37(2), 823 (1977).
50. Nakahara, H., Yanokung, M., and Murakami, Y., "Environmental Effects
of Geothermal Waste Water on the Nearby River System," J. Radioanal.
Chem. 45(1), 25 (1978).
51. Nakanishi, T., and Samson, B., "Low Energy Photon Spectrometry in
Nondestructive Neutron Activation Analysis of Environmental Samples,"
J. Radioanal. Ghem., 37.(2), 945 (1977).
52. Robertson, D.E., and Carpenter, R., Activation Analysis in Strategies
of Marine Pollution Monitoring, E.D. Goldberg, ed. Wiley, New York,
N.Y., pp 93-195, 1976.
53. Spyrou, N.M., and Kerr, S.A., "Cyclic Activation: The Measurement of
Short-Lived Isotopes in the Analysis of Biological and Environmental
Samples," J. Radioanal. Chem. 48(1), 169 (1979).
54. Valente, I., Minski, M.J., and Bowen, H.J.N., "Rapid Determination of
Antimony in Biological and Environmental Samples Using Instrumental
Neutron Activation Analysis," Anal . Chem. 50(11), 1479 (1978).
55. Kay, M.A., McKnown, D.M., Gray, D.H., Eichoi, M.E., and Vogt, J.R.,
"Neutron Activation Analysis in Environmental Chemistry," Amer. Lab.
5.C7), 39 (1973).
141
-------
56. Currie, L.A., "Detection and Quantisation in X-Ray Fluorescence
Spectrometry," in X-Rav Fluorescence Analysis of Environmental
Samples,T.G. Dzubay, ed., Ann Arbor Sci., Ann Arbor, Mich., pp 289-306.
57. Florkowski, T., Holynska, B., and Piorek, S., "X-Ray Fluorescence
Techniques in Analysis of Environmental Samples," in Proceedings
of the International Symposium on Measurement, Detection and Control
of Environmental Pollution, pp 213-31, 1976.
58. Florkowski, T., Kuc, T., and Piorek, S., "Determination of Trace
Elements in Plants by the X-Ray Fluorescence Analysis for Environmental
Pollution Investigations," Int. J. Appl. Radiat. Isot^ 28_(8), 679
0977).
59. Hulett, L.D., Dunn, H.W., Dale, J.M., Emery, J.F., Lyons, W.S., and
Murty, PoS., "The Characterization of Solid Specimens From Environ-
mental Pollution Studies Using Electron, X-Ray and Nuclear Methods,"
in Proceedings of the International Symposium on Management,
Detection and Control of Environmental Pollution, pp 29-42, 1976.
60. Maskowicz, A., "A Method of Correction for Absorption Matrix Effects
in Samples of 'Intermediate1 Thickness in EDXRF Analysis," X-Ray
Spectrum, 9_(1), 14 (1979).
61. Meier, H., and Cluger, E., "On the Application of Radioisotope X-Ray
Fluorescence Analysis for the Solution of Environmental and Industrial
Problems," J. Radioanal, Chem., 32_(2), 413 (1976).
62. Pella, P.A., Lorbes, K., and Heinrich, K.F.J., "Quantitative Multi-
element Analysis of Environmental Samples by X-Ray Fluorescence
Spectrometry." NBS Spec. Pub!.» 464, 125 (1977).
63. Van Espen, P., Mullens, H., and Adams, F.C., "Automated Energy-
Dispersive X-Ray Fluorescence Analysis of Environmental Samples,"
Fresenius' Z. Anal. Chem., 285(3-4), 215 (1977).
64. Qagman, J., Miller, J.L., Griffin, R.J., and Kellogg, R.B., "Verifica-
tion of a Particle Size Correction Method for X-Ray Fluorescence
Spectrometric Analysis of Environmental Samples," Anal. Chem., 50(1),
37 (1978). ~~
65. Gilfrich, J.V., Burkhalter, P.G., and Birks, L.S., "X-Ray Spectrometry
for Particulate Air Pollution - A Quantitative Comparison of Techniques,"
Anal. Chem., 45_(12), 2002 (1973).
66. Chu, T.C., Navarette, V.R., Kaji, H., Izawa, G., Shiokawa, T.,
Ishil, K., and Morita, S., "A Study on Proton Induced X-Ray Analysis
and its Application to Environmental Samples, J. Radioanal. Chem.,
36(1), 195 (1977).
67. Finston, H.L., Williams, E.T., Baumas, S.E., Jacobson, S.H., and
Bond, Jr., A.M., "Chemical Analysis with a Low Energy Particle
Accelerator," J. Radioahal. Chem., 43(2), 523 (1978).
142
-------
68. Malmquist, K., Akselsson, R.A., and Johansson, G., "Determination of
Heavy Metals and Fluorine in Airborne Participate Matter in an
Outdoor Environment by Simultaneous Use of PIXE and Ganma^Ray
Detection," Proe. Anal. Div. Chem. So.. 15(1). 13 (1978).
69. Shiokawa, T., Chu, T.C., Navasette, V.R., Kaji, H.,-Izawa, G.,
Ishii, K., Morita, S., and Tawara, H., "A Study on Proton Induced
X-Ray Analysis and its Application to Environmental Samples" Nucl.
Instrum. Methods. 142(1-2), 199 (1977).
70. Walter, R.L., Willison, R.D., Gutknecht, W.F., and Shay, Jr., R.W.,
"The Application of Proton-Induced X-Ray Emission to Bioenvironmental
Analyses," Nucl. Instrum. Methods, 142(1-2). 181 (1977).
71. Desaedeleer, G., Winchester, J.W., Pilotte, J.O., Nelson, J.W., and
Moffitt, H.A., "Proton-Induced X-Ray Emission Analysis of Roadway
Aerosol Time-Sequence Filter Samples for Pollution Control Strategy,"
Meas. Detect. Control Environ. Pollut., Proc. Int. Symp., 1976,
IATA1Vienna, Austria, pp 233-44.
72. Pilotte, J.O., Nelson, J.W., and Winchester, O.W., "Application of
Multi-Station Time Sequence Aerosol Sampling and Proton Induced
X-Ray Emission Analysis Techniques to The St. Louis Regional Air
Pollution Study for Investigating Sulfur-Trace Metal Relationship,"
PrOc. ERDA Symp. X-Gamna-Ray Sources Appl., H.C. Griffin, Ed.,
TTTT5: Springfield, Va., pp 161-3, 1976.
73. Malmquist, K., Akselsson, R.A., and Johansson, G., "Determination of
Heavy Metals and Fluorine in Airborne Particulate Matter in An
Indoor Environment by Simultaneous Use of PIXE and Gamma-Ray
Detection." Proc. Anal. Div. Ghem. Soc., 15(1), 13 (1978).
74. Smith, R.D., Campbell, J.A., and Nielson, K.K., "Concentration
Dependence Upon Particle Size of Volatilized Elements in Flyash,"
Environ. Sci. Tech., 13(5), 553 (1979).
75. Ensor, D.S., et al, "Evaluation of a Particle Scrubber on a Coal-Fired
Utility," EPA-600/2-75-074, November 1975.
76. Harris, D.B., "Procedures for Cascade Impactor Calibration and
Operation in Process Streams," EPA-600/2-77-004, January 1977.
77. Johnson, J.W., Clinard, G.I., Felix, L.G., and McCain, J.D., "A
Computer Based Cascade Impactor Data Reduction System,. EPA-600/7-78-
042, March 1978.
78. Starkovich, J.A., Maddalone, R.F., Kraft, M.L., Zee, C.A., Lin, C..
and Flegal, C.A., "Measurement Techniques for Inorganic Trace Materials
in Control System Streams," EPA-600/2-76-283, December 1976.
143
-------
79. Tolansfry, S., "An Introduction to Interferometry," Longmans,
Green and Co., London, 1966.
80. Jacquinot, P., Rep. Prog. Phys., J3S, 267 (1960).
81. Bell, R.J., "Introductory Fourier Transform Spectroscopy," Academic
Press, New York, 1972.
82. Fellgett, P.B., J. Phys. Radium, 19, 187 (1958).
83. Connes, J., and Connes, P., J. Opt. Soc. An., j>6, 896 (1966).
84. Coley, J.W., and Tuckley, J.W., Math Comput., 1£, 297 (1965).
85. Ryan, L.E., Beimer, R.G., Maddalone, R.F., EPA-600/7-79-063b,
Feb. (1979).
86. Cunningham, P.T., Johnson, S.A., Prepr. Pap. Nat'l Meet. Div.
Environ. Chem., An. Chem. Soc. 14, 42 (1974) (Eng.).
87. Belorgrot, C., Kachmarshy, J., Moeler, D., Pluchina, A., Appl. Opt.
15_(3), 708-13, (1975) (Eng.)
88. Gillespie, C.R.E., Smith, B.T., Anal. Instrum.. 15, 91 (1977).
89. Kagel, R.D., and Nyquist, R.A., "Infrared Spectra of Inorganic
Compounds," Academic Press, New York, 1973.
90. Miller, F.A., and Williams, C.H., "Infrared Spectra and Characteris-
tic Frequencies of Inorganic Ions," Anal. Chem., 24, 1253 (1952).
91. Aremow, L.C., and Vanderberg, J.T., "High Resolution Spectra of
Inorganic Pigments and Exteriors in the Mid-Infrared Region from
1500 cm-' to 200 cm'1," Beckman Reprint R-6236.
92. Cole, H. (Ed.) Trans. Am. Crystal!og. Assoc.. Vol. 12 (1976).
93. "Sampling and Analysis of Reduced and Oxidized Species in Process
Streams," EPA Contract 69-02-2165, TRW DSSG Technical Progress
Narrative No. 8, February 1977 and Narrative No. 12, June 1977.
94. "Encyclopedia of Industrial Chemical Analysis Vol. No. 3 General
Techniques," Ed. F.D. Snell, Interscience Publishers, New York (1966).
95. "Tables for Conversion of X-Ray Diffraction Angles to Interplanar
Spacing," Applied Mathematics Series 10, National Bureau of Standards,
U.S. Govt. Printing Office Washington, O.C., 1950.
96. Klug, H.P., and Alexander, L.E., "X-Ray Difraction Procedures for
Polycrystalline and Amorphous Materials," 2nd Edition, John Wiley
& Sons, New York.
144
-------
97. Estep, P.A., Kovach, J.J., Waldstien, P. and Karr, C., "Infrared and
Raman Spectroscopic Studies of Structural Variations in Minerals
Acta 14> and 15 Samp1es'" Geochimica Et Cosmochimica
98' ISS^Mn!!;^' "X"fay photoelectron Spectroscopy,11 Anal. Chem.. 45,
788A (1973).
99. Betteri dge, D. "Analytical Aspects of Photocell ecti on Spectroscopy,"
Intern. J. Environ. Anal. Chem. .*!. 243 (1972).
100. Riggs, W.M. and Parker, M.J., "Surface Analysis by X-Ray Photo-
electron Spectroscopy."
101. Shirley, D.A. , "Electron Spectroscopy," North Holland Publishing
Company, Amsterdam-London, 1972.
102. Sieghbahn, K. , et al, "Electron Spectroscopy for Chemical Analysis-
Atomic, Molecular, and Solid State Structure Studies by Means of
Electron Spectroscopy," Air Force Materials Laboratory, Wright
Patterson Air Force Base, Ohio, Technical Report AFML-TR-68-189,
October, 1968.
103. Henrich, K.F.J. and Newbury, D.E., Secondary Ion Mass Spectrometry;
Proceedings of a Workshop on Secondary Ion Mass Spectrometry and lo
MTcroprobe Mass Analysis. NBS Special Publication 427. October, 197
104, Evans, C.A., "Secondary Ion Mass Analysis: A Technique for Three
Dimensional Characterization," Anal. Chem.. 44, 67 A, 1972.
105. Evans, C.A., "Surface and Thin Film Analysis," Anal. Chem.. 47,
855 A (1975).
106. Bearden, J.A. and Burr, A.F., Rev. Mod. Phys.. 39, 175 (1967).
107. Steinhart, R.G., Hudis, J. and Perlman, M.L., Phys. Rev. B. j>,
1016 (1972).
108. Baer, Y., Heden, P.F., Hedman, J., Klasson, M., and Nordling, C.,
Solid State Communication. 8, 1479 (1970).
109. Daly, N.R., Rev. Sci. Instrum.. .31, 264, (1960).
110. McHugh, J.A., "Empirical Quantisation Procedures in SIMS," from
Proceedings of a Workshop on SIMS and IMMA, NBS Special Publication
427, October 1975.
111. McHugh, J.A., "Factors that Influence an Elemental Depth Concentra-
tion Profile," from Proceedings of a Workshop on SIMS and IMMA, NBS
Special Publication 427, October, 1975.
145
-------
112. Evans, C.A. and Pemslen, J.P., Anal. Chem.. 42, 1060 (1970).
113. Evans, C.A., Jr. "Advances in Mass Spectrometry," p. 436, A. Quayle,
Ed., Institute of Petroleum, London, England, 1971.
114. Scofield, J.H., Lawrence Livermore Laboratory Report, UCRL - 51326,
1973.
115. Swingle, R.S. and Riggs, W.M., "ESCA>" Critical Reviews in
Analytical Chemistry, Volume 5, Issue 3, pages 267-321, 1975.
116. Castaing, R. and Hennequin, J.R., "Advances in Mass Spectrometry,"
Vol V, p 419, A. Quayle, Ed., Institute of Petroleum, London,
England, 1971.
117. Andersen, C.A., Int. J. Mass Spectrom. Ion Phys. 2, 61 (1969).
118. Andersen, C.A., Int. J. Mass Spectrom. Ion Phys.. 3, 413 (1970)
119. Werner, A.S., "Investigation of Particulate Sulfur by ESCA," in
the "Workshop Proceedings in Primary Sulfate Emissions from
Combustion Sources," EPA-600/9-78-020a, b, August, 1978.
120. Cutright, D.E. Huget, E.F. and J.M. Brady. Asbestos: A Subtle
Carcinogen in the Dental Laboratory, NTIS Document No. AD-A045
027/OST, 14 p. 1978.
121. Bradshaw, R.W., Stoltz, R.E. and D.R. Adolphson. Formation of
Protective Layers on Alloys Developed for Use in Coal Gassification
Environments. Quarterly Progress Report, July 1, 1977-September 30,
1977. Contract No. EX-76-C-040789, NTIS Document No. SAND-78-8209,
37 p.
122. Finnegan, S.A., Coso Geothermal Corrosion Study. Report No. NWC-
TP-5974. NTIS document No. AD-A045 511/35 T, 89 p., 1977.
123. Beaman, D.R. and J.A. Isasi, "A Critical Examination of Computer
Programs Used in Quantitative Electron Microprobe Analysis," Anal.
Chem. 42, 1540 (1970).
124. Gehrke, R.J. and Davies, R.C., "Spectrum Fitting Technique for
Energy Dispersive X-ray Analysis of Oxides and Silicates with
Electron Microbeam Excitation," Anal. Chem. 47., 1537 (1975).
125. P.W. Wright, "Procedure for Quantitative Electron Probe Micro-
analysis." Anal. Chem. 45_, 101 (1973).
146
-------
126. Shelley, P. E. and G. A. Kirkpatrick, "Assessment of Automatic Flow
Samples," EPA-R2-73-261, National Environmental Research Center,
Cincinnati, Ohio, June 1973.
127. Moody, J. R. and Lindstrom, R. M., "Selection of Cleaning of Plastic
Containers for Storage of Trace Element Samples," Anal. Chem.,
49 (14)» 2285 (1977).
128. U.S. E.P.A., "Methods for Chemical Analysis of Water and Waste,"
EPA-600-4-79-020, March 1979.
147
-------
APPENDIX A
SPECIFIC ANION ANALYSIS PROCEDURES
-------
APPENDIX A - SPECIFIC ANION ANALYSIS PROCEDURES
Summaries of specific anion tests for Br", I", COZ, HCOl, Cl", CN", F",
_ _ — —
NOZ, NO,, HP07; S~, SOZ, SOT are contained in the following sections. In
•3 c. *r <3 *r •
addition to these anions, a specific test for NH^ is also included. Brief
summaries of the procedures, inferences, sampling, and limits of detection
are given. Those methods taken from the EPA Water and Waste Water Manual
have their Storet number listed, while ASTM procedures are referenced by
their number.
A-l AMMONIA (STORET NO. 00610)
Summary of Method
The sample is buffered at a pH of 9.5,and distilled into a solution of
boric acid. The ammonia in the distillate is determined colorimetrically
by Nesslerization (Hglg, KI, NaOH).
Interferences
Glycine, urea, glutamic acid, cyanates and acetamide hydrolyze very
slowly in solution on standing. At a pH of 9.5, urea hydrolysis amounts to
about 7.0% and cycanates amount to about 5.0%.
Glycine, hydrazine, and some amines will react with Nessler's reagent
to given the characteristic yellow color in the time required for the test.
Organic compound such as ketones, aldehydes, alcohols, and some amines
may cause an off color on Nesslerization. Some organic compounds like for-
maldehyde may be eliminated by boiling at a low pH prior to Nesslerization.
Residual chlorine must be removed prior to the ammonia determination
by pretreatment of the sample.
Sampling
Collect samples in accordance with American Society for Testing and
Materials methods D510, D860, D1066, D1192, and D1496. If sample cannot
be immediately analyzed, it should be placed in a plastic bottle, pre-
served by the addition of 1 ml/liter of concentrated sulfuric acid (sp gr
1.84), and stored in a frozen condition. The ammonia content in the
sample will remain for up to 30 days.
A-2
-------
Limits of Detection
The colorimetric determination of ammonia has a detection limit of
0.50 mg/liter with an uncertainty of ±0.03 mg/L.
A-2 BROMIDE AND IODIDE (STORE! NO. 71870)
Summary of Method
The determination of bromide and iodide consists of two separate
experiments. The iodide is first determined in the sample, and then a
second experiment determines the combined iodide and bromide. The bromide
content of the sample is calculated from the difference between the
iodide and the combined iodide and bromide determination.
The iodide in the sample is oxidized to iodate with saturated bromine
water in an acid buffer solution. The excess bromine is destroyed by the
addition of sodium formate. Potassium iodide is added to the sample solu-
tion with the resulting liberated iodine being equivalent to the iodate
initially formed in the oxidiation step. The liberated iodine is deter-
mined by titration with sodium thiosulfate.
In a second sample, iodide and bromide are oxidized to iodate and
bromate with calcium hypochlorite. The iodine liberated by the combined
reaction products is measured after destruction of the excess
hypochlorite and addition of potassium iodide.
Interferences
Iron, manganese, and organic matter interfere with the above
methods. Treatment of the initial samples with calcium oxide removes
the interferents.
Sampling
Collect samples in accordance with American Society for Testing and
Materials methods D1192 and D3370.
Limits of Detection
The titrimetric determination of iodide and bromide has a detection
limit in the mg/liter range.
A-3
-------
A-3 CARBONATE AND BICARBONATE CASTM D513-71}
Summary of Method
Carbon dioxide is liberated by acidifying and heating the sample in a
closed system which includes a condenser, a gas scrubber, a C02 absorber,
an expansion bladder, and a gas-circulating pump. The C02 gas is absorbed
in a barium hydroxide solution and the excess hydroxide is titrated with
standard acid. The concentrations of the carbonate species are determined
from the pH and total C02 values.
Interferences
Carbon dioxide can be lost from solution during transit and storage
of samples.
Any volatile acid, base, or barium precipitate not removed by the
scrubbing solution will interfere.
Hydrogen sulfide is an interferent but is removed with the iodine
scrubbing solution.
Sampling
Collect samples in accordance with American Society for Testing and
Materials methods D1066, D1192 and D3370.
Limits Of Detection
The titrimetric determination of total C02 has a detection limit in the
mg/liter range with a precision of 0.25 mg/liter.
A-4 \ CHLORIDE (STORET NO. 00940)
Summary of Method
The sample solution is adjusted to pH 8.3 and then titrated with
mercuric nitrate solution in the presence of diphenyl carbazone-bromophenol
blue indicator. The persistence (10 sec) of the blue-violet mercury
complex color indicates the end point for the titration.
Interferences
Anions and cations normally in surface waters do not interfere.
Sulfite must be eliminated (addition of H202).
A-4
-------
Sampl ing
Collect samples in accordance with American Society for Testing and
Materials methods D1066 and D3370.
Limits of Detection
The titrimetric determination of chloride has a detection limit in the
mg/liter range.
A-5 CYANIDE (STORET NO. 00720)
Summary of Method
The sample is refluxed with a solution consisting of H2S04 and CUpClg.
Hydrogen cyanide gas is liberated and absorbed in NaOH solution. Either a
colorimetric or titrimetric procedure may be used for the determination of
cyanide.
The colorimetric procedure calls for the neutralization of the
absorption solution with acetic acid to pH 6.5 - 8.0 and the addition of
0.2 ml of chloramine-T solution. The solution absorbance is measured at
620 nm after twenty minutes. This procedure is recommended for solutions
where the cyanide concentration is 1 mg/liter or less.
For the titrimetric procedure, the absorption solution is titrated
with AgN03 using Rhodamine B as the indicator.
Interferences
Interferences include substances which contribute to color or turbidity
changes. The presence of cyanate or thiocyanate or the presence of
organic nitrogen compounds also interferes.
Some organic cyanide compounds like nitriles decompose under distilla-
tion. Aldehydes convert cyanide to nitrile.
Sampling
Collect the sample in accordance with American Society for Testing and
Materials methods D1192 and D3370. If the sample cannot be analyzed
immediately, stabilize it by the addition of NaOH solution to a pH of 12 or
more and store it in a closed plastic bottle. If chlorine is present, add
ascorbic acid as soon as the sample is collected.
A-5
-------
Limit of Detection
Titrimetric and spectrometric determinations of cyanide have detection
limits of 1 mg/L and 0.02 mg/L, respectively.
A-6 FLUORIDE (STORE! NO. 00951)
Summary of Method
The pH of the sample is adjusted to 5.2 - 5.5 with 0,5 N H2S04. Carbon
dioxide is removed from solution by heating on a hot water bath. The pH of
the sample is adjusted to 6.3 with a buffer solution of 1M sodium citrate -
citric acid - 0.2M KNOg. Fluoride is determined using a specific ion
electrode and standard addition methods.
Interferences
Metals such as Fe and Al can form complexes with F~. These inter-
ferences are minimized by the addition of the citrate buffer solution, which
preferentially complexes the metals, and by controlling the pH.
Sampling
Collect samples in accordance with American Society for Testing and
Materials methods D510, D1192 and D1496. Sample are preserved in a plastic
container and stored at 4°C.
Limits of Detection
The specific ion electrode determination of fluoride has a detection
limit of 1 mg/liter with an uncertainty of ±0.1 mg/L. ^
A-7 NITRATE (STORET NO. 00620)
Summary of Method
Nitrate ion reacts with brucine in a strong sulfuric acid solution to
develop a yellow color. The color is measured at 410 run but the Beer-
Lambert relationship does not hold for this system. A plot of absorbance
versus concentration produces a smooth curve which replaces the Beer-
Lambert relationship. The rate of color development varies directly with
temperature and the intensity varies inversely with the temperature,
Samples containing 1 to 50 mg/liter of NO^ ion can be analyzed by this
method. Samples with large concentrations of NOg ion must first be diluted.
A-6
-------
Interferences
Chlorine can have an Interference of up to 5 rag/liter but may be
eliminated by the addition of sodium arsenite.
Strong oxidizing or reducing agents will cause interferences.
Nitrite can have an interference of up to 7 mg/liter but is eliminated
by the sulfanilic acid in the brucine reagent solution.
Organic matter in the sample may interfere by increasing the
absorbance readings at 410 nm. The problem comes about because concentrated
sulfuric acid will char the organic molecules. This problem can be
eliminated by pretreatment with aluminum hydroxide and specially treated
activated carbon.
Sampling
Collect samples in accordance with American Society for Testing and
Materials method D1192 and D3370.
Limit of Detection
The spectrometric determination of nitrate has a detection limit in
the mg/liter range.
A-8 NITRITE (STORET NO. 00630)
Summary of Method
Nitrite is measured spectrophotometrically following diazotization with
sulfanilamide and coupling with N-(l-naphthyl)-ethylenediamine dihydro-
chloride. The resulting azo dye is measured at 540 nm. The method is
applicable to samples containing 0.01 to 1.0 mg/L. Nitrate is determined
in a separate aliquot of the sample following reduction to nitrite using
a column of granular copper/cadmium mixture. Nitrate is calculated from
the difference in nitrite concentrations of the two solutions.
Interferences
Interferences due to high concentrations of metals (e.g., Fe and Cu)
may be eliminated by complexing with EDTA. Turbid samples should be
filtered. Samples containing oil or grease must be pre-extracted if this
nitrite/nitrate procedure is employed.
A-7
-------
Sampling
Collect the samples in accordance with American Society for Testing and
Materials method D1066, D1192, and D3370.
Limit of Detection
The spectrometric determination of nitrite has a detection limit of
O.OT mg/liter.
A-9 ORTHOPHOSPHATE (STORET NO. 00665)
Summary of Method
A solution of ammonium molybdate-vanadate is added to the sample
producing the yellow color of molybdovanadophosphoric acid. The color
intensity is proportional to the orthophosphate concentration of the
sample. The yellow color is measured at 400 nm. This method is recom-
mended for samples where the concentration of othophosphates is 1 to
10 mg/liter. For more dilute samples the colorimetric ascorbic acid
reduction method should be applied.
Interferences
High concentrations of the ferric ion will interfere with this-method
as will other highly colored species.
Sampling
Collect samples in accordance with American Society for Testing and
Material methods D1066, D1192 and D3370. If analysis cannot be performed
immediately, preserve by the addition of 40 rag of mercuric chloride per
liter of sample and store at 4°C.
Limits of Detection
The spectrometric determination of othophosphates has a detection
limit in the mg/liter range.
A-10 SULFIDE (STORET NO. TOTAL 00745)
Summary of Method
After the sample is acidified, sulfide is stripped from the sample with
an inert gas and collected in a zinc acetate solution. The stripped sulfide
forms a zinc sulfide suspension in the collection solution. The addition
A-8
-------
of excess iodine to the suspension causes the sulfide to react under
acidic conditions. Thiosulfate is used to measure the unreacted iodine
to determine the quantity of iodine consumed by sulfide.
Interferences
Reduced sulfur compounds, such as sulfite, thiosulfate, and hydrosulfite
decompose in acid and will cause erratic results.
Volatile Iodine-consuming substances give high results.
Samples must be taken with a minimum of aeration to avoid oxidation
of sulfide.
Sampling
Collect samples in accordance with American Society for Testing and
Materials methods D1066, D1192 and D3370. If analysis cannot be performed
immediately, preserve with zinc acetate.
Limits of Detection
Precision and accuracy for this method have not been determined.
A-ll SULFITE (STORET NO. 00740}
r
Summary of Method
The sample is taken in a tube of special design which excludes air
until the sulfite has been reacted with a solution containing HC1, KI,
and KI03. The excess iodine chloride is determined by titration with
thiosulfate using a deadstop endpoinMndicating apparatus. This method
is recommended for sample concentrations of 0.1 to 6 mg/liter.
Interferences
Reducing agents such as sulfides and certain heavy-metal ions react
with iodine chloride.
If nitrite ion is present, it will oxidize sulfite when the solution
is acidified.
Sampling
Collect samples in accordance with American Society for Testing and
Materials method D1192 and D3370.
A-9
-------
Limits of Detection
The titrimetric determination of sulfite has a detection limit in the
0.1 mg/liter range.
A-12 SULFATE (STORE! NO. 00945)
Summary of Method
There are three basic methods for the determination of sulfates:
gravimetry, turbidimetry and titrimetry. Gravimetry is a primary measure
of sulfate ion in all types of water but it is time consuming and has a
lower sensitivity. Turbidimetry is the most sensitive and titrimetry
is the fastest.
For the gravimetric method, sulfate ion is precipatated and weighed as
barium sulfate after removal of silica and other insoluble matter by
filtration.
For the turidimetric method, sulfate ion is converted to a barium
sulfate suspension. Glycerin and sodium chloride solutions are added to
stabilize the suspension and minimize interference. The turbidity is
determined on a spectrophotometer and compared to standard sulfate
solutions.
For the volumetric method, sulfate ion is titrated in an alcoholic
solution under controlled acid conditions with a standard barium chloride
solution, Thorin is employed as the indicator with a color change of
yellow to pink.
Interferences
Gravimetric interferences are caused by other substances being occluded
or adsorbed on the barium sulfate. Sulfites and sulfides may oxidize and
precipitate as sulfate. Turbidimetric interferences are caused by
insoluble suspended matter. Volumetric interferences are usually caused
by cations and anions which may coprecipitate. Most interfering cations
are removed by ion exchange columns.
A-10
-------
Sampling
Collect samples tn accordance with. American Society for Testing and
Materials methods D1066, Oil92 and D3370.
Limits of Detection
The gravimetric method has a detection limit of 10 mg/liter.
The turbidimetric method has a detection limit of 2 mg/liter.
The volumetric method has a detection limit of 1.5 mg/liter if ion-
exchange chromatography is employed, 5 to 10 mg/liter otherwise.
A-11
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/2-79-200
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
EPA/IERL-RTP Procedures Manual: Level 2 Sampling
and Analysis of Oxidized Inorganic Compounds
S. REPORT DATE
November 1979
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
R.F.Maddalone, L.E.Ryan, R.G.Delumyea, and
J.A.Wilson
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
TRW Defense and Space Systems Group
One Space Park
Redondo Beach, California 90278
10. PROGRAM ELEMENT NO.
INE624
11. CONTRACT/GRANT NO.
68-02-2165, Task 102
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 PERIODCOVERED
13. TYPE OF REPORT AND PE
Task Final; 7/76 -
14. SPONSORING AGENCY CODE
EPA/600/13
15. SUPPLEMENTARY NOTES IERL_RTP project officer is Frank E. Briden, Mail Drop 62, 919/
541-2557.
16. ABSTRACT
The report describes Level 2 analysis procedures for identifying oxidized
inorganic compounds in environmental samples from energy and industrial processes
The procedures include: (1) initial sample characterization, (2) bulk sample charac-
terization, and (3) invidual particle characterization. The theory, sensitivity, inter-
ferences, sample preparation, application, and information derived are described
for each procedure. The report is a step in the development of a general methodology
for analysis of process samples. It defines the concepts of Level 2 analyses and
reviews currently available procedures. It does not define a fixed protocol because
the complexity of samples precludes definition of specific procedures without exam-
ination or analysis.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Pollution
Inorganic Compounds
Oxidation
Sampling
Analyzing
Pollution Control
Stationary Sources
13B
07B
14B
8. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PAGES
167
20. SECURITY CLASS (Thispage}
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
------- |