Rapid Screening of
Metals Using Portable High-Resolution
X-Ray Fluorescence Spectrometers
ALAN D. HEWITT
ments, Inc., Metorex, Inc.) with silicon (drifted with
INTRODUCTION
lithium) [Si(Li)] or mercuric iodide (HgI2) detec-
Metal pollution over the past decade has come
tors capable of achieving resolutions of 170 and
to be of greater concern due to an increasing aware-
300 eV, respectively. This increased spectral reso-
ness of the pathways leading to chronic human
lution allows for the unambiguous qualitative
toxicity (Spittler and Fender 1979). During this time
identification of metals in complex mixtures.
the use of X-ray fluorescence (XRF) analysis has
Quantitatively, however, the measurement of
increased for establishing the presence of elevated
discrete spectral energies obtained by XRF analy-
levels of metals in environmental samples (e.g.,
sis is often dependent on other metals present in
Furst et al. 1985, Piorek and Rhodes 1988, Grupp
the sample, due to absorption and enhancement
et al. 1989, Watson et al. 1989, Ashe et al. 1991,
effects. For environmental sample analysis, these
Carlson and Alexander 1991, Driscoll et al. 1991,
matrix-specific effects are best addressed by the
Garby 1991, Harding 1991, Puls et al. 1994, etc.).
use of empirical coefficients (Piorek and Rhodes
This instrumental method of analysis establishes
1988). This method of instrumental calibration re-
concentration estimates for several metals over a
quires the acquisition or development of several
concentration range extending from percent lev-
well-characterized materials with a matrix com-
els to around 100 g/g (ppm). The method requires
position similar to that of the samples. Moreover,
little or no sample pretreatment and can be per-
the standards should contain a range of metal con-
formed for a small fraction of the time and cost
c e n t r a t i o n s b r a c k e t i n g t h e d e s i re d l e v e l o f
associated with acid extraction/atomic absorption
quantitation. Due to these requirements, calibra-
or emission analysis. Moreover, XRF analysis is
tion standards are often both material and site spe-
nondestructive, so subsamples can be archived or
cific, and their preparation and/or verification may
analyzed by other procedures. For these reasons,
take one or more weeks.
XRF analysis is a very practical and economical
A more practical solution for an initial site in-
method when screening for high concentrations
vestigation would be a calibration method that is
of metals during a remedial investigation and fea-
insensitive to sample matrix so that only few stan-
sibility study (RI/FS) at a suspected hazardous
dards would be needed to handle a diverse range
waste site.
of samples (e.g., soil composition "sand/slit/clay,"
sediment, sludge, dust, paint chips, etc.). Ideally,
Today several manufacturers offer energy-dis-
this capability would allow samples to be analyzed
persive XRF spectrometers that have been de-
signed to be compatible with field operations. The
independent of any previous site characterization.
most transportable systems have a battery power
With this concept in mind, two quantitation rou-
tines have been proposed for field-transportable
supply option and use one or more radioactive
high-resolution XRF systems. One routine is based
sources as the primary incident radiation for el-
emental excitation. The first generation of these
on fundamental parameters (FP) analysis, which
stand-alone systems was equipped with gas
eliminates the need for several calibration stan-
proportional detectors that had a spectral resolu-
dards by relying on certain physical constants to
estimate metal concentration while theoretically
tion on the order of 8001000 electron volts (1000
eV = 1 keV). More recently, systems have been
correcting for matrix discrepancies (Figura 1987,
marketed (HNU Systems Inc., Spectrace Instru-
1993). Another approach uses response factors (RF)
Previous Page
