Religion & Spirituality

Field portable XRF analysis of environmental samples

Published
of 19
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Description
Journal of Hazardous Materials 83 (2001) Field portable XRF analysis of environmental samples Dennis J. Kalnicky a, Raj Singhvi b, a Lockheed Martin Technology Services Group, Environmental Services/REAC,
Transcript
Journal of Hazardous Materials 83 (2001) Field portable XRF analysis of environmental samples Dennis J. Kalnicky a, Raj Singhvi b, a Lockheed Martin Technology Services Group, Environmental Services/REAC, 2890 Woodbridge Avenue, Edison, NJ 08837, USA b US Environmental Protection Agency, Environmental Response Team Center, 2890 Woodbridge Avenue, Edison, NJ 08837, USA Abstract One of the critical factors for successfully conducting contamination characterization, removal, and remedial operations at hazardous waste sites is rapid and appropriate response to analyze samples in a timely fashion. Turnaround time associated with off-site analysis is often too slow to support efficient utilization of the data. Field portable X-ray fluorescence (FPXRF) techniques provide viable and effective analytical approaches to meet on-site analysis needs for many types of environmental samples. Applications include the in situ analysis of metals in soils and sediments, thin films/particulates, and lead in paint. Published by Elsevier Science B.V. Keywords: XRF; Field portable XRF; Environmental; In situ; Soil contamination; On-site 1. Introduction One of the critical factors for successfully conducting extent of contamination, removal, and remedial operations at hazardous waste sites is rapid and appropriate analytical support to analyze site samples in a timely fashion. Historically, there have been problems obtaining high quality sample analysis results within a time frame necessary to support efficient utilization of the data. Field portable X-ray fluorescence (FPXRF) spectrometry has become a common analytical technique for on-site screening and fast turnaround analysis of contaminant elements in environmental samples. Applications include the in situ analysis of metals in soils and sediments, thin films/particulates, and lead in paint. FPXRF is a non-destructive analytical technique that allows both qualitative and quantitative analysis of the composition of a sample. XRF spectrometry has been utilized in the laboratory for many years. Portable XRF technology has gained widespread acceptance in the environmental community as a viable Corresponding author. Tel.: ; fax: /01/$ see front matter. Published by Elsevier Science B.V. PII: S (00) 94 D.J. Kalnicky, R. Singhvi / Journal of Hazardous Materials 83 (2001) analytical approach for field applications due to the availability of efficient radioisotope source excitation combined with highly sensitive detectors and their associated electronics. While wavelength dispersive XRF has been the mainstay of laboratory instrumentation, energy dispersive XRF (EDXRF) is the technique of choice for field instrumentation primarily due to the ease of use and portability of EDXRF equipment. FPXRF instruments can provide both qualitative and quantitative analysis of environmental samples, and in some cases, without the need for site specific standards. 2. Theory Atoms fluoresce at specific energies when excited by X-rays. Detection of the specific fluorescent photons enables the qualitative and quantitative analysis of most elements in a sample [1 3]. The mechanism for the X-ray fluorescence of an atom is illustrated in Fig. 1. An inner shell vacancy is created (by an incident X-ray photon or other phenomena) leaving an electron hole in the inner shell. An outer shell electron falls to fill the inner shell vacancy as the atom relaxes to the ground state. This process gives off photons with energy in the X-ray region of the electromagnetic spectrum equivalent to the energy difference between the two shells. Each atom has an X-ray line spectrum that consists of a series of discrete energies with intensities related to the probability that a particular transition will occur. The X-rays emitted Fig. 1. Mechanism for X-ray fluorescence of an atom. D.J. Kalnicky, R. Singhvi / Journal of Hazardous Materials 83 (2001) are characteristic of the atom, and provide qualitative identification of the element. The photon energy of a spectral line is the difference in energy, E, between the initial and final levels involved in the electronic transition. Comparing the intensities of the X-rays from an unknown sample to those of suitable standards provides the basis for quantitative analysis of the element. If the shell electron being replaced is a K-shell electron, then the X-ray emission is known as a K X-ray. Similarly, L-shell transitions produce L X-rays. X-ray spectral lines are grouped in series (K, L, M). All lines in a series result from electron transitions from various levels to the same shell. For example, transitions from the L- and M-shells to the K-shell provide spectral lines designated K and K, respectively. A spectrum of X-rays is generated by all the elements in the sample. Each element will have many characteristic lines in the spectrum, since a distinct X-ray will be emitted for each type of orbital transition. 3. FPXRF analyzers Fig. 2 illustrates a block diagram of a typical XRF spectrometer. An excitation source (X-ray tube, radioisotope, etc.) is used to irradiate a sample which in turn fluoresces. The characteristic X-ray fluorescence is then detected and analyzed. The entire process is interfaced with a computer that provides general instrument control, data generation, and processing. Several different techniques may be used to induce fluorescence in a sample and to detect/analyze the characteristic X-rays given off by the sample. Fig. 2. Block diagram for a typical EDXRF spectrometer. 96 D.J. Kalnicky, R. Singhvi / Journal of Hazardous Materials 83 (2001) Table 1 Commonly used radioactive isotopes for XRF analysis Isotope Half-life Useful radiation Energy (kev) X-rays excited efficiently Fe years Mn K X-rays 5.9 Al Cr Co days Fe K X-rays 6.4 Cf Cd years Ag K X-rays 22.2 Ca Tc 88 W U Am years Np L X-rays Sn Tm 26 Cm years Pu L X-rays Ti Se 3.1. XRF sources Various excitation sources may be used to irradiate a sample [1,3]. In a radioisotope source excited XRF analyzer, characteristic X-rays emitted from a sealed radioisotope source irradiate the sample. Alternately, an X-ray tube may be used to irradiate the sample with characteristic and continuum X-rays. Some of the original application studies reported in the literature for transportable XRF analyzers utilized X-ray tubes as sources [4,5]. Shortly thereafter, radioisotope source FPXRF analyzers were evaluated for environmental applications [6]. Table 1 lists radioisotope sources typically used in FPXRF analyzers. The most commonly used sources include Fe-55, Co-57, Cd-109, and Am-241. Each of these gives off radiation at specific energy levels and, therefore, efficiently excites elements within a specific atomic number range. As a result, no single radioisotope source is sufficient for exciting the entire range of elements of interest in environmental analysis, and many instruments use two or three sources to maximize element range. The half-life of a source is important, especially for Fe-55, Co-57, and Cd-109 sources. With half-lives as short as 270 days, some means (usually electronic) must be provided to compensate for the loss in source intensity over time. These sources may have to be replaced after a few years when their intensity decreases to a level too low to provide adequate sensitivity for the elements of concern. Intensity in X-ray spectrometry is always given in counts per unit time, that is, X-ray photons per unit area per unit time. The unit area is usually the useful area of the detector, which is constant for all measurements and, therefore, is normally not included in the X-ray intensity unit Wavelength versus energy dispersion XRF analyzers are usually classified by wavelength or energy dispersion for X-ray line detection and analysis. Wavelength dispersion involves the separation of X-ray lines on the D.J. Kalnicky, R. Singhvi / Journal of Hazardous Materials 83 (2001) basis of their wavelengths, which may be accomplished with crystals (crystal dispersion), diffraction (diffraction dispersion), or spacial (geometric) dispersion. In energy dispersion, the separation of the X-ray lines is based on photon energies, and is accomplished by electronic dispersion with a pulse height analyzer. FPXRF analyzers typically employ energy dispersion for separation of X-ray lines. Wavelength is inversely proportional to energy and the conversion is [1,3], E ev = 12400/λ, where E ev is the energy in electron volts and λ is the wavelength in angstroms, Å Detectors The X-ray detector converts the energies of the X-ray photons into voltage pulses that can be counted to provide a measurement of the total X-ray flux [2]. X-ray detectors are typically proportional devices where the energy of the incipient X-ray photon determines the size of the output voltage. Voltage discrimination via pulse height selection is used to select a narrow band of voltage pulses to pass to the scaling circuitry. A polychromatic beam of radiation incident upon the detector produces a spectrum of voltage pulses having a height distribution proportional to the energy distribution of the incident polychromatic beam. A multichannel analyzer is used to separate the spectrum of voltage pulses into narrow voltage bands for measurement of individual energies. The three most common types of detectors are: the gas flow proportional detector, the scintillation detector, and the solid-state semiconductor detector. These detectors differ in resolution and analyte sensitivity. Resolution is the ability of the detector to separate X-rays of different energies, and is important for minimizing spectral interferences and overlap. Semiconductor detectors have the best resolution and are preferred for FPXRF instruments. These detectors may require liquid nitrogen as a coolant or employ electronic cooling FPXRF instrumentation All FPXRF analyzers utilize the basic components illustrated in Fig. 2. Some configurations incorporate a measurement probe connected to an electronics unit via a flexible cable. The probe houses the detector and radioisotope source(s), while the electronics unit contains the microprocessor and data processing electronics. Typically, the probe weighs 3 5 lb and the electronics unit weighs lb. Other FPXRF analyzers are contained in a single unit, and weigh less than 5 lb. Proper radiation shielding is provided by the manufacturer in accordance with applicable regulations governing manufacture and licensing of radioactive devices. The manufacturer also provides training in the safe and proper operation of the analyzer. Table 2 lists representative FPXRF instrumentation. Some instruments provide dedicated element analysis (e.g. Pb in paint), while others provide a variety of elemental analyses depending on source and detector configuration. They generally are readily adaptable to field operations, though they may be limited by the power capacities of their batteries and the availability of liquid nitrogen. All provide a minimum of 8 h of field use with replacement of batteries. 98 D.J. Kalnicky, R. Singhvi / Journal of Hazardous Materials 83 (2001) D.J. Kalnicky, R. Singhvi / Journal of Hazardous Materials 83 (2001) Calibration and quantitation The definition of quantitative XRF analysis depends, to a large extent, on the application and intended use for the data. For environmental applications, FPXRF results are quantitative when measurement precision is within 20%, and results are confirmed by an approved laboratory method [7]. Analysis of reference materials should produce results that are within ±20% of the certified values for target elements that have concentrations more than 10 times the FPXRF detection limit. While this definition is much less stringent than that for classical laboratory XRF analysis, it is a viable approach for most FPXRF environmental applications Factors affecting XRF calibration Quantitative application of XRF methods for environmental applications requires calibration of the XRF analyzer using standards with known compositions [7,8]. The calibration procedure compares X-ray intensity for target elements to known concentrations in standards to develop a quantitation model suitable for analyzing a given type of sample (e.g. soils, liquids, thin films). A number of factors that may affect XRF response must be considered during the calibration process: (1) detector resolution and its relationship to spectral interferences; (2) sample matrix effects; (3) accuracy and suitability of calibration standards; (4) sample morphology (particle size, homogeneity, etc.), and (5) sample measurement geometry. Proportional counter detectors typically have significantly poorer resolution than solidstate semiconductor devices and, therefore, are less able to resolve X-ray spectral overlaps. Therefore, it may be impossible to calibrate certain element combinations solely due to detector limitations, for example, interfering K X-ray lines from neighboring elements. Furthermore, some X-ray line overlaps are so severe that even the best resolution obtained for semiconductor detectors on FPXRF systems is insufficient to separate them (e.g. As K/Pb L), and residual error may persist in the spectral deconvolution techniques used to obtain net intensities for XRF calibration purposes. Matrix effects arise from the impact that variations in concentrations of interfering elements have on the measured X-ray intensity of the target element. These effects produce non-linear intensity response versus target element concentration, and they appear as either X-ray absorption or enhancement phenomena. Most FPXRF analyzers provide means to correct for these effects when the application is calibrated. The severity of matrix effects and calibration method employed generally dictate the number of standards required to calibrate an application. The standards selected to calibrate XRF applications must have accurately known concentrations for the target elements. The accuracy of the standards ultimately defines the best accuracy that can be expected for the XRF calibration model, and the measurement times necessary to achieve it. Calibration standards must also be representative of the matrix and target element concentrations that are to be analyzed. Sample morphology (particle size distribution, uniformity, heterogeneity, and surface condition) must be considered when calibrating environmental XRF applications. Standards should exhibit the same characteristics as the samples to be analyzed to produce a reliable calibration model. Sample placement 100 D.J. Kalnicky, R. Singhvi / Journal of Hazardous Materials 83 (2001) Table 3 Comparison of XRF calibration methods Empirical calibration Site samples must be collected for use as standards and must be certified by independent laboratory methods High costs associated with collection and analysis of site samples and significant time to receive data back from the laboratory XRF must be calibrated with site-specific standards prior to project initiation A large number of standards may be required to model and correct for matrix effects Results based on a good calibration model will be accurate and directly comparable to laboratory analysis Fundamental parameters calibration Must know or estimate 100% of sample composition including unmeasured balance Pure elements and/or readily available certified reference materials may be used as standards No site-specific calibration is required; should be applicable to any site with same sample type All elements are included in the FP quantitation algorithm; concentrations in standards need not bracket the levels at the site FP model may require initial fine-tuning using certified reference materials is a potential source of error, since the X-ray signal is sensitive to measurement geometry and decreases as the distance from the excitation source increases. This error is minimized by maintaining the same source/sample geometry for all calibration standard and sample measurements Calibration methods There are two major approaches for calibrating FPXRF applications. The empirical approach relies on a suite of site (or typical ) standards and regression mathematics to generate a site-specific calibration for elemental response and matrix effects. The fundamental parameters (FP) approach utilizes X-ray theory to mathematically pre-determine interelement matrix effects combined with pure element or known standard intensity responses to develop a quantitative algorithm for a specific sample type. FP methods provide multi-site capabilities by eliminating the requirement for site-specific standards. A comparison of site-specific and FP calibration methods is given in Table Empirical calibrations Empirical calibrations are typically based on a set of previously collected site-specific calibration standards (SSCS) that have been analyzed by reliable independent laboratory methods [8 10]. They must be representative of the matrix and target element concentration ranges at the site. Standards must bracket the full range of target element and interfering matrix element concentrations, and must reflect variations in element ratios to produce a representative calibration model. The highest and lowest concentrations in the SSCS set define the calibration range. Samples used to generate the calibration must be prepared in the same way as samples that will be analyzed at the site. The SSCS set should include several samples with concentrations near the critical concentration of concern, i.e. the action level, to improve the accuracy of the empirical calibration model. The greater the knowledge about the sample matrix (how it varies at the site), the more representative the calibration model is and, therefore, the more accurate the results. D.J. Kalnicky, R. Singhvi / Journal of Hazardous Materials 83 (2001) Typical models used for empirical calibration are described elsewhere [3,9,10]. A minimum of 5 10 samples are needed to generate a simple linear model for a single analyte when interelement matrix effects are not significant. As the number of elements analyzed increases, more calibration samples are required to adequately characterize target element concentration ranges and correct for interelement matrix effects. For some applications, it may be necessary to produce more than one calibration model to maintain linearity over the concentration ranges in question. If the sample matrix varies significantly, a calibration model should be generated for each matrix type present at the site to provide better characterization. In some cases, taking out the ratio of the analyte intensity to the scattered X-rays from the source (backscatter) may be useful to correct for matrix effects, because backscatter intensity is proportional to the average composition of the sample. The ratio technique may also be useful for generating non-site-specific empirical models provided a sufficient number of standards typical of the sample matrix are available. For example, analysis of metal contaminants in soils where backscatter may provide information on the average composition of the soil sample Fundamental parameters calibrations FP techniques have been understood and commonly utilized on laboratory XRF systems for many years to analyze a wide variety of materials [1 3,11,12]. Historically, FPXRF instruments that have been used for environmental applications have relied upon site-specific calibration methods that have not been useful for more than one site and/or sample matrix. With the availability of field portable computing power, the FP approach is valid for FPXRF analyzers and provides multi-site capabilities by eliminating the requirement for site-specific standards. However, uncertainties in the data used to generate theoretical coefficients may lead to errors and biases in FP analytical models based on them. Therefore, adjustments based on certified reference materials may be necessary to produce reliable result
Search
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks