Geology Reference
In-Depth Information
One example of this potential application was demonstrated by Reddy et al. (1993) who used SWIR data for the
evaluation of an active volcano (Barren Island volcano), a coal mine fire (Jharia coalfield), and industrial hot spots
(Bokaro Steel Plant). Subsequent applications using SWIR and satellite data are found in Chatterjee (2006),
Chatterjee et al. (2007), Guha et al. (2008), Gangopadhyay et al. (2005, 2006), and Zhang et al. (2004).
Other Analytical Methods
T
here are numerous other methods potentially useful for characterizing minerals nucleated from coal-fire gas. Cost,
analysis time, and accessibility to specialized equipment are limiting factors. Other methods that might be utilized
include (1) infrared spectroscopy, (2) Raman spectroscopy, (3) optical spectroscopy, (4) nuclear magnetic and
electron spin resonance spectroscopy, (5) X-ray fluorescence spectroscopy, and (6) thermal gravimetric
analysis. Calas and Hawthorne (1988) review the details about these and other spectroscopic methods.
To demonstrate how additional information can be used to characterize coal-fire mineral assemblages, an example
of a common method used for analyzing the nitrogen content of a sample is presented. This is the MCDN method
that is useful when nitrogen is present in small quantities in a sample (
100 ppm). Recall from above that EDS and
>
WDS have limited use in detecting nitrogen even at percent levels.
Using MDCN, the total nitrogen content of a mineral assemblage from a gas vent is determined by transforming
solid matter (minerals and organics) to a gas phase via the rapid flash combustion (with pure O
2
) of a powdered
sample (~10mg) encapsulated with tin in a furnace at 1200°C (Schroeder and Ingall, 1994). The method is
described in more detail in Figure 10.1.9. The gases are next separated in a gas chromatograph, where the amount
of nitrogen is measured. Carbon and sulfur can also be detected during the same process because their gaseous
oxides also form and the travel times of these combustion products are different in the gas chromatograph. After
calibration with known standards, very low limits of detection (ppm) and high precision are possible.
Sample dropper
Combustion
(a)
GC
Detection
Chromatograph
(e)
(d)
(g)
(f)
(b)
(c)
© 1996 James P. H. Fuller
Furnace
Sensor
and recorder
Figure 10.1.9. The rotating multisample dropper (a) delivers one sample at a time into the top of the quartz
combustion tube, (b). Each sample is enclosed in an ultrapure tin or silver capsule. The combustion tube contains
granulated chromium III oxide, a combustion catalyst, held at 1200°C. Flash combustion products from (b) into (c)
occur when a pulse of pure O
2
gas is injected into the quartz tube with the sample, generating a temperature as high
as 1700°C. Carbon in the sample is converted to CO
2
. The nitrogen-bearing gas products of combustion include N
2
and nitrous oxides, NO
x
. The gases are swept out of (c) by a constant stream of nonreactive He carrier gas, which
entered at (a). The gases then pass through a reduction column filled with chopped Cu wire (600°C) in which the
NO
x
release oxygen and N
2
gas. Some oxygen is reacted with the Cu and this spent material must be periodically
replaced as new Cu wire. Water vapor is removed by a gas trap at (d) that contains magnesium perchlorate. The
clean sample gases then pass through a gas chromatograph column (e), where they are separated into N
2
and CO
2
.
A stream of the gas sample and a separate reference stream of helium from (f) pass through the detector at (g).
Differences in thermal conductivity between the two streams are displayed as visible peaks and recorded as
numerically integrated areas. Linear regression applied to the combustion of known standard materials yields a
regression line by means of which peak areas from unknowns are converted into total element values for each
sample. Figure modified with permission from Tom Maddox, http://www.uga.edu/~sisbl/ udumas.html.
