Electron Probe Microanalysis (EPMA) requires the acquisition of characteristic X-ray intensities ... more Electron Probe Microanalysis (EPMA) requires the acquisition of characteristic X-ray intensities of elements of interest, which are superimposed upon the background (also called continuum or bremsstrahlung). These background counts must be removed for accurate quantitation, usually by measuring two off-peak channels. An alternative approach, the Mean Atomic Number (MAN) background calculation method, was developed in 1970s. This technique takes advantage of Kramers' law: the generated continuum intensity scales monotonically with the (mean) atomic number of the target. Here, an empirical curve of background intensity versus MAN is generated via measurements at a specific characteristic X-ray peak position for a desired element on a series of standards that do not contain the analyte of interest [1,2]. After a correction for continuum absorption upon the emitted X-rays, we can assume that the MAN curve is a smooth function of the mean atomic number and can easily be used to determine the continuum intensity of a particular material, given its mean atomic number.
<p>The chemistry of erupted minerals and melts are commonly used to determine the p... more <p>The chemistry of erupted minerals and melts are commonly used to determine the pressures, temperatures and H<sub>2</sub>O contents of magma storage regions beneath volcanic centres. In turn, these estimates are vital for hazard assessment, to understand the formation of critical metal deposits, and to inform models of continental crust formation. In the last few decades, more than 100 empirical and thermodynamic expressions have been calibrated using measurements of phases in experimental studies where these intensive parameters are known. By collating these different models into a computationally-efficient, open-source Python3 package, Thermobar, we can critically assess the performance of thermobarometers in igneous systems, and propagate analytical errors. When we apply published models for different mineral equilibrium to a new experimental dataset not used in model calibration, we find that stated errors vastly underestimate the true uncertainty when these workflows are applied to natural systems.</p><p>Specifically, we find that realistic calculation workflows involving Clinopyroxene (Cpx) equilibrium (e.g., iterating pressure and temperature) have uncertainties spanning the entire crust in most tectonic settings. Using Thermobar functions to propagate analytical error using Monte Carlo simulations, we suggest that these large errors result from imprecise analyses of minor elements such as Na in experimental (and natural) Cpx. Common analytical conditions used for Cpx yield highly correlated pressure-temperature arrays spanning the entire crust, which have been incorrectly interpreted as trancrustal storage in natural systems. Insuffucient analyses of each phase in experimental products means that this analytical error is not sufficiently mediated by averaging, so reported mineral compositions deviate from the true phase composition. This impacts thermobarometer calibration, as well as assessment of these methods using test experimental datasets.</p><p>Overall, we demonstrate that the development of Python3 infrastructure for common quantitative workflows in volcanology is vital to allow rigorous error assessment and model intercomparison; such assessments simply aren’t feasible using traditional tools (e.g., Excel workbooks). Specific changes to analytical, experimental and model calibration workflows (e.g., higher beam currents and count times in Na) will be essential to produce a more robust dataset to calibrate and test the next generation of more precise and accurate Cpx-based barometers. In turn, this will enable more rigorous investigation of magma storage geometries in a variety of tectonic settings (e.g., distinguishing true transcrustal storage vs. storage in discrete reservoirs).</p><p> </p>
We show components with eignenvalues ≥ 1.0. The total variance explained by these components if 5... more We show components with eignenvalues ≥ 1.0. The total variance explained by these components if 53% and 46%, respecFvely. See Figure 3 for experimental regions.
Abstract While much progress has been made in electron-probe microanalysis (EPMA) to improve the ... more Abstract While much progress has been made in electron-probe microanalysis (EPMA) to improve the accuracy of point analysis, the same level of attention has not always been applied to the quantification of wavelength-dispersive spectrometry (WDS) X-ray intensity maps at the individual pixel level. We demonstrate that the same level of rigor applied in traditional point analysis can also be applied to the quantification of pixels in X-ray intensity maps, along with additional acquisition and quantitative processing procedures to further improve accuracy, precision, and mapping throughput. Accordingly, X-ray map quantification should include pixel-level corrections for WDS detector deadtime, corrections for changes in beam current (beam drift), changes in standard intensities (standard drift), high-accuracy removal of background intensities, quantitative matrix corrections, quantitative correction of spectral interferences, and, if required, time-dependent corrections (for beam and/or contamination sensitive materials). The purpose of quantification at the pixel level is to eliminate misinterpretation of intensity artifacts, inherent in raw X-ray intensity signals, that distort the apparent abundance of an element. Major and minor element X-ray signals can contain significant artifacts due to absorption and fluorescence effects. Trace element X-ray signals can contain significant artifacts where phases with different average atomic numbers produce different X-ray continuum (bremsstrahlung) intensities, or where a spectral interference, even an apparently minor one, can produce a false-positive intensity signal. The methods we propose for rigorous pixel quantification require calibration of X-ray intensities on the instrument using standard reference materials, as we already do for point analysis that is then used to quantify multiple X-ray maps, and thus the relative time overhead associated with such pixel-by-pixel quantification is small. Moreover, the absolute time overhead associated with this method is usually less than that required for quantification using manual calibration curve methods while resulting in significantly better accuracy. Applications to geological, synthetic, or engineering materials are numerous as quantitative maps not only show compositional 2D variation of fine-grained or finely zoned structures but also provide very accurate quantitative analysis, with precision approaching that of a single point analysis, when multiple-pixel averaging in compositionally homogeneous domains is utilized.
This is the peer-reviewed, final accepted version for American Mineralogist, published by the Min... more This is the peer-reviewed, final accepted version for American Mineralogist, published by the Mineralogical Society of America. The published version is subject to change. Cite as Authors (Year) Title. American Mineralogist, in press.
PRSUPR, an integrated program to acqr ple fixed or scanning spectrometer electror le automation i... more PRSUPR, an integrated program to acqr ple fixed or scanning spectrometer electror le automation includes automated standard scribed. X-ray count and coordinate data are ng predigitized standard samples referenced to random-access disk files for on-line or of I marks or user-defined locations from a preusing the clrzuFsample matrix correctior on' Automation for unknown sample speciset ofunknown and standard sample autom lom selected points' line traverses' all.d x-y integrated into the program. Extensive u from user-defined cartesian grids with rectsimple installation program are provided polygon boundaries using mouse' trackball' mer. The program operates on a standard II In addition' the program can operate unatand supports automation, acquisition, and ing any combination of the above standard for a wide variety ofmicroprobe hardware nation procedures' omation includes multiple element analysis
Faced with time constraints and with beam sensitive materials, electron probe micro-analyser (EPM... more Faced with time constraints and with beam sensitive materials, electron probe micro-analyser (EPMA) operators often desire faster analyses to save time and/or to minimize sample damage from exposure to the electron beam. One of the most common approaches to decreasing analytical time is to reduce the amount of time spent counting x-rays both on peak and on background. This results in fewer x-ray counts, however, which translates into reduced analytical precision and higher detection limits. Increasing the beam current can compensate for this reduced count time, but may also result in increased sample damage and an attendant decrease in accuracy.
We present methods combining backscattered-electron (BSE) mosaic imaging, quantitative spot-mode ... more We present methods combining backscattered-electron (BSE) mosaic imaging, quantitative spot-mode electron-probe microanalysis (EPMA), and quantitative compositional mapping by EPMA and micro-xray fluorescence (µXRF) to provide a framework for detailed analysis of terrestrial and lunar samples. BSE imaging provides a base map for the characterization of samples by EPMA. Recent developments in image stitching provide a convenient method of processing BSE mosaic image sets. Characterizing cm-sized samples by EPMA and µXRF provides complementary information about sample chemistry. We acquire BSE mosaic and x-ray stage maps on a JEOL JXA-8200 electron microprobe equipped with five WDS spectrometers and a silicon drift EDS (SDD) that are used to acquire major and minor element maps. Quantitative EPMA maps are acquired using Probe Image, Probe for EPMA, and CalcImage software using conventional EPMA standardization, mean atomic number (MAN) background correction, and a full Φ(pz) correction at each pixel in the map. Maps are acquired with 400-1024 pixels at 0.X-15 µm and 10-100 msec per pixel at analytical conditions of 15kV and 50-100 nA. Large areas are covered by selection of appropriate pixel resolution and step size. An EDAX Orbis µXRF equipped with an SDD is used for spectrum image stage mapping using a 30 µm polycapillary optic, similar pixel resolution, and dwell times of 50-500 msec.
Martian meteorite Elephant Moraine A79001 (EET 79001) has received considerable attention for the... more Martian meteorite Elephant Moraine A79001 (EET 79001) has received considerable attention for the unusual composition of its shock melt glass, particularly its enrichment in sulfur relative to the host shergottite. It has been hypothesized that Martian regolith was incorporated into the melt or, conversely, that the S-enrichment stems from preferential melting of sulfide minerals in the host rock during shock. We present results from an electron microprobe study of EET 79001 including robust measurements of major and trace elements in the shock melt glass (S, Cl, Ni, Co, V, and Sc) and minerals in the host rock (Ni, Co, and V). We find that both S and major element abundances can be reconciled with previous hypotheses of regolith incorporation and/or excess sulfide melt. However, trace element characteristics of the shock melt glass, particularly Ni and Cl abundances relative to S, cannot be explained either by the incorporation of regolith or sulfide minerals. We therefore propose an alternative hypothesis whereby, prior to shock melting, portions of EET 79001 experienced acid-sulfate leaching of the mesostasis, possibly groundmass feldspar, and olivine, producing Al-sulfates that were later incorporated into the shock melt, which then quenched to glass. Such activity in the Martian near-surface is supported by observations from the Mars Exploration Rovers and laboratory experiments. Our preimpact alteration model, accompanied by the preferential survival of olivine and excess melting of feldspar during impact, explains the measured trace element abundances better than either the regolith incorporation or excess sulfide melting hypothesis does.
Electron microprobe trace element analysis is a significant challenge. Due to the low net intensi... more Electron microprobe trace element analysis is a significant challenge. Due to the low net intensity of peak measurements, the accuracy and precision of such analyses relies critically on background measurements, and on the accuracy of any pertinent peak interference corrections. A linear regression between two points selected at appropriate background positions is a classical approach for electron probe microanalysis (EPMA). However, this approach neglects the accurate assessment of background curvature (exponential or polynomial), and the presence of background interferences, a hole in the background, or an absorption edge can dramatically affect the results if underestimated or ignored. The acquisition of a quantitative wavelength-dispersive spectrometry (WDS) scan over the spectral region of interest remains a reasonable option to determine the background intensity and curvature from a fitted regression of background portions of the scan, but this technique can be time consuming and retains an element of subjectivity, as the analyst has to select areas in the scan which appear to represent background. This paper presents a new multi-point background (MPB) method whereby the background intensity is determined from up to 24 background measurements from wavelength positions on either side of analytical lines. This method improves the accuracy and precision of trace element analysis in a complex matrix through careful regression of the background shape, and can be used to characterize the background over a large spectral region covering several elements to be analyzed. The overall efficiency improves as systematic WDS scanning is not required to assess background interferences. The method is less subjective compared to methods that rely on WDS scanning, including selection of two interpolation points based on WDS scans, because “true” backgrounds are selected through an exclusion method of possible erroneous backgrounds. The first validation of the MPB method involves blank testing to ensure the method can accurately measure the absence of an element. The second validation involves the analysis of U-Th-Pb in several monazite reference materials of known isotopic age. The impetus for the MPB method came from efforts to refine EPMA monazite U-Th-Pb dating, where it was recognized that background errors resulting from interference or strong background curvature could result in errors of several tens of millions of years on the calculated date. Results obtained on monazite reference materials using two different microprobes, a Cameca SX-100 Ultrachron and a JEOL JXA-8230, yield excellent agreement with ages obtained by isotopic methods (Thermal Ionization Mass Spectrometry [TIMS], Sensitive High-Resolution Ion MicroProbe [SHRIMP], or Secondary Ion Mass Spectrometry [SIMS]). Finally, the MPB method can be used to model the background over a large spectrometer range to improve the accuracy of background measurement of minor and trace elements acquired on a same spectrometer, a method called the shared background measurement. This latter significantly improves the accuracy of minor and trace element analysis in complex matrices, as demonstrated by the analysis of Rare Earth Elements (REE) in REE-silicates and phosphates and of trace elements in scheelite.
Lunar and Planetary Science Conference, Mar 1, 2009
Introduction: Advances in instrumentation and analytical techniques using electron-probe microana... more Introduction: Advances in instrumentation and analytical techniques using electron-probe microanalysis (EPMA), and microbeam X-ray fluorescence spectrometry (µXRF) have been applied to the study of lunar samples. These techniques provide compositional information from discrete spots and in digital mapping mode, over a range of spatial resolutions that is appropriate for micro and macroscopic analysis of complex mineral chemistry and textures. We have applied ongoing research and development of EPMA and µXRF algorithms in simulation and analysis software, defocused-beam analysis (DBA), and hyperspectral-imaging techniques to study lunar samples. Improvements in Quantitative X-ray Analysis: Improvements in EPMA and µXRF instrumentation has enabled large-scale X-ray mapping of lunar samples. The advent of silicon-drift (SDD) energy-dispersive (EDS) detectors with high-throughput digital pulse processing electronics has dramatically decreased mapping acquisition times and increased sensitivity. EPMA provides a relatively high spatial resolution of ~1 µm and allows sampling of up to 50-100 µm spot in defocused-beam analysis (DBA), and can analyze a wide range of elements including low energy X-rays, but trace element detection is ultimately limited by continuum Xray production. Analysis by µXRF typically uses a minimum spot size of 50-100 µm, has superior trace element sensitivity and can be used on materials which are not vacuum compatible or exhibit electron-beam damage in the microprobe, but requires thicker samples for analysis due to X-ray penetration and has relatively poor low energy X-ray sensitivity. The two techniques are complimentary for these reasons. EPMA has seen improvements in Φ(ρz) algorithms and more accurate sets of mass absorption coefficients have improved quantitative analysis by both techniques. Quantitative EPMA corrects for atomic number (Z) effects of electron scattering and retardation during X-ray generation, absorption (A) of X-rays in the sample along the path to the detector, and fluorescence (F) of X-rays by both characteristic and continuum X-rays. The k-ratio is the measured X-ray intensity in the sample relative to that in the standard, and the calculated concentration C is determined by iterative calculation from C = k * ZAF. This iterative correction implicitly requires knowledge of the concentration of all elements in the sample volume and their effect on the element of interest. The
The composition of clinopyroxene and clinopyroxene-liquid (Cpx-Liq) pairs are frequently used to ... more The composition of clinopyroxene and clinopyroxene-liquid (Cpx-Liq) pairs are frequently used to calculate crystallization/equilibration pressures in igneous systems. While canonical uncertainties are often assigned to calculated pressures based on fits to calibration or test datasets, the sources of these uncertainties (and thus ways to reduce them) have not been rigorously assessed. We show that considerable uncertainties in calculated pressures arise from analytical error associated with Electron Probe Microanalyser (EPMA) measurements of Cpx. Specifically, low X-ray counts during analysis of elements with concentrations <1 wt% resulting from insufficient count times and/or low beam currents yield highly imprecise measurements (1σ errors of 10-40% for Na 2 O). Low analytical precision propagates into the calculation of pressure-sensitive mineral components such as jadeite. Using Monte Carlo approaches, we demonstrate that elemental variation resulting from analytical precision alone generates pressures spanning ∼4 kbar (∼15 km) for a single Cpx and ∼6 kbar for a single Cpx-Liq pair using popular barometry expressions. In addition, analytical uncertainties in mineral compositions produce highly correlated arrays between pressure and temperature that have been previously attributed to transcrustal magma storage. Before invoking such geological interpretations, a more mundane origin from analytical imprecision must be ruled out. Most importantly, low analytical precision does not just affect the application of barometers to natural systems; it has also affected characterization of Cpx in experimental products used to calibrate and test barometers. The impact of poor precision on each individual measurement is often magnified by the small number of measurements made within experimental charges, meaning that low analytical precision and true variability in mineral compositions have not been sufficiently mediated by averaging multiple EPMA analyses. We compile the number of Cpx measurements performed in N = 307 experiments used to calibrate existing barometers, and N = 490 new experiments, finding ∼45% of experiment charges were characterized by ≤5 individual Cpx analyses. Insufficient characterization of the true composition of experimental phases likely accounts for the fact that all Cpx-based barometers exhibit large errors (± 3 kbar) when tested using global experimental datasets. We suggest specific changes to analytical and experimental protocols, such as increased count times and/or higher beam currents when measuring low concentration elements in relatively beam resistant Cpx in experiments and natural samples. We also advocate for increasing the number of analyses per experimental charge, resolving interlaboratory analytical offsets and improving data reporting. Implementing these changes is essential to produce a more robust dataset to calibrate and test the next generation of more precise and accurate Cpx-based barometers. In turn, this will enable more rigorous investigation of magma storage geometries in a variety of tectonic settings (e.g. distinguishing true transcrustal storage vs. storage in discrete reservoirs).
Electron Probe Microanalysis (EPMA) requires the acquisition of characteristic X-ray intensities ... more Electron Probe Microanalysis (EPMA) requires the acquisition of characteristic X-ray intensities of elements of interest, which are superimposed upon the background (also called continuum or bremsstrahlung). These background counts must be removed for accurate quantitation, usually by measuring two off-peak channels. An alternative approach, the Mean Atomic Number (MAN) background calculation method, was developed in 1970s. This technique takes advantage of Kramers' law: the generated continuum intensity scales monotonically with the (mean) atomic number of the target. Here, an empirical curve of background intensity versus MAN is generated via measurements at a specific characteristic X-ray peak position for a desired element on a series of standards that do not contain the analyte of interest [1,2]. After a correction for continuum absorption upon the emitted X-rays, we can assume that the MAN curve is a smooth function of the mean atomic number and can easily be used to determine the continuum intensity of a particular material, given its mean atomic number.
&lt;p&gt;The chemistry of erupted minerals and melts are commonly used to determine the p... more &lt;p&gt;The chemistry of erupted minerals and melts are commonly used to determine the pressures, temperatures and H&lt;sub&gt;2&lt;/sub&gt;O contents of magma storage regions beneath volcanic centres. In turn, these estimates are vital for hazard assessment, to understand the formation of critical metal deposits, and to inform models of continental crust formation. In the last few decades, more than 100 empirical and thermodynamic expressions have been calibrated using measurements of phases in experimental studies where these intensive parameters are known. By collating these different models into a computationally-efficient, open-source Python3 package, Thermobar, we can critically assess the performance of thermobarometers in igneous systems, and propagate analytical errors. When we apply published models for different mineral equilibrium to a new experimental dataset not used in model calibration, we find that stated errors vastly underestimate the true uncertainty when these workflows are applied to natural systems.&lt;/p&gt;&lt;p&gt;Specifically, we find that realistic calculation workflows involving Clinopyroxene (Cpx) equilibrium (e.g., iterating pressure and temperature) have uncertainties spanning the entire crust in most tectonic settings. Using Thermobar functions to propagate analytical error using Monte Carlo simulations, we suggest that these large errors result from imprecise analyses of minor elements such as Na in experimental (and natural) Cpx. Common analytical conditions used for Cpx yield highly correlated pressure-temperature arrays spanning the entire crust, which have been incorrectly interpreted as trancrustal storage in natural systems. Insuffucient analyses of each phase in experimental products means that this analytical error is not sufficiently mediated by averaging, so reported mineral compositions deviate from the true phase composition. This impacts thermobarometer calibration, as well as assessment of these methods using test experimental datasets.&lt;/p&gt;&lt;p&gt;Overall, we demonstrate that the development of Python3 infrastructure for common quantitative workflows in volcanology is vital to allow rigorous error assessment and model intercomparison; such assessments simply aren&amp;#8217;t feasible using traditional tools (e.g., Excel workbooks). Specific changes to analytical, experimental and model calibration workflows (e.g., higher beam currents and count times in Na) will be essential to produce a more robust dataset to calibrate and test the next generation of more precise and accurate Cpx-based barometers. In turn, this will enable more rigorous investigation of magma storage geometries in a variety of tectonic settings (e.g., distinguishing true transcrustal storage vs. storage in discrete reservoirs).&lt;/p&gt;&lt;p&gt;&amp;#160;&lt;/p&gt;
We show components with eignenvalues ≥ 1.0. The total variance explained by these components if 5... more We show components with eignenvalues ≥ 1.0. The total variance explained by these components if 53% and 46%, respecFvely. See Figure 3 for experimental regions.
Abstract While much progress has been made in electron-probe microanalysis (EPMA) to improve the ... more Abstract While much progress has been made in electron-probe microanalysis (EPMA) to improve the accuracy of point analysis, the same level of attention has not always been applied to the quantification of wavelength-dispersive spectrometry (WDS) X-ray intensity maps at the individual pixel level. We demonstrate that the same level of rigor applied in traditional point analysis can also be applied to the quantification of pixels in X-ray intensity maps, along with additional acquisition and quantitative processing procedures to further improve accuracy, precision, and mapping throughput. Accordingly, X-ray map quantification should include pixel-level corrections for WDS detector deadtime, corrections for changes in beam current (beam drift), changes in standard intensities (standard drift), high-accuracy removal of background intensities, quantitative matrix corrections, quantitative correction of spectral interferences, and, if required, time-dependent corrections (for beam and/or contamination sensitive materials). The purpose of quantification at the pixel level is to eliminate misinterpretation of intensity artifacts, inherent in raw X-ray intensity signals, that distort the apparent abundance of an element. Major and minor element X-ray signals can contain significant artifacts due to absorption and fluorescence effects. Trace element X-ray signals can contain significant artifacts where phases with different average atomic numbers produce different X-ray continuum (bremsstrahlung) intensities, or where a spectral interference, even an apparently minor one, can produce a false-positive intensity signal. The methods we propose for rigorous pixel quantification require calibration of X-ray intensities on the instrument using standard reference materials, as we already do for point analysis that is then used to quantify multiple X-ray maps, and thus the relative time overhead associated with such pixel-by-pixel quantification is small. Moreover, the absolute time overhead associated with this method is usually less than that required for quantification using manual calibration curve methods while resulting in significantly better accuracy. Applications to geological, synthetic, or engineering materials are numerous as quantitative maps not only show compositional 2D variation of fine-grained or finely zoned structures but also provide very accurate quantitative analysis, with precision approaching that of a single point analysis, when multiple-pixel averaging in compositionally homogeneous domains is utilized.
This is the peer-reviewed, final accepted version for American Mineralogist, published by the Min... more This is the peer-reviewed, final accepted version for American Mineralogist, published by the Mineralogical Society of America. The published version is subject to change. Cite as Authors (Year) Title. American Mineralogist, in press.
PRSUPR, an integrated program to acqr ple fixed or scanning spectrometer electror le automation i... more PRSUPR, an integrated program to acqr ple fixed or scanning spectrometer electror le automation includes automated standard scribed. X-ray count and coordinate data are ng predigitized standard samples referenced to random-access disk files for on-line or of I marks or user-defined locations from a preusing the clrzuFsample matrix correctior on' Automation for unknown sample speciset ofunknown and standard sample autom lom selected points' line traverses' all.d x-y integrated into the program. Extensive u from user-defined cartesian grids with rectsimple installation program are provided polygon boundaries using mouse' trackball' mer. The program operates on a standard II In addition' the program can operate unatand supports automation, acquisition, and ing any combination of the above standard for a wide variety ofmicroprobe hardware nation procedures' omation includes multiple element analysis
Faced with time constraints and with beam sensitive materials, electron probe micro-analyser (EPM... more Faced with time constraints and with beam sensitive materials, electron probe micro-analyser (EPMA) operators often desire faster analyses to save time and/or to minimize sample damage from exposure to the electron beam. One of the most common approaches to decreasing analytical time is to reduce the amount of time spent counting x-rays both on peak and on background. This results in fewer x-ray counts, however, which translates into reduced analytical precision and higher detection limits. Increasing the beam current can compensate for this reduced count time, but may also result in increased sample damage and an attendant decrease in accuracy.
We present methods combining backscattered-electron (BSE) mosaic imaging, quantitative spot-mode ... more We present methods combining backscattered-electron (BSE) mosaic imaging, quantitative spot-mode electron-probe microanalysis (EPMA), and quantitative compositional mapping by EPMA and micro-xray fluorescence (µXRF) to provide a framework for detailed analysis of terrestrial and lunar samples. BSE imaging provides a base map for the characterization of samples by EPMA. Recent developments in image stitching provide a convenient method of processing BSE mosaic image sets. Characterizing cm-sized samples by EPMA and µXRF provides complementary information about sample chemistry. We acquire BSE mosaic and x-ray stage maps on a JEOL JXA-8200 electron microprobe equipped with five WDS spectrometers and a silicon drift EDS (SDD) that are used to acquire major and minor element maps. Quantitative EPMA maps are acquired using Probe Image, Probe for EPMA, and CalcImage software using conventional EPMA standardization, mean atomic number (MAN) background correction, and a full Φ(pz) correction at each pixel in the map. Maps are acquired with 400-1024 pixels at 0.X-15 µm and 10-100 msec per pixel at analytical conditions of 15kV and 50-100 nA. Large areas are covered by selection of appropriate pixel resolution and step size. An EDAX Orbis µXRF equipped with an SDD is used for spectrum image stage mapping using a 30 µm polycapillary optic, similar pixel resolution, and dwell times of 50-500 msec.
Martian meteorite Elephant Moraine A79001 (EET 79001) has received considerable attention for the... more Martian meteorite Elephant Moraine A79001 (EET 79001) has received considerable attention for the unusual composition of its shock melt glass, particularly its enrichment in sulfur relative to the host shergottite. It has been hypothesized that Martian regolith was incorporated into the melt or, conversely, that the S-enrichment stems from preferential melting of sulfide minerals in the host rock during shock. We present results from an electron microprobe study of EET 79001 including robust measurements of major and trace elements in the shock melt glass (S, Cl, Ni, Co, V, and Sc) and minerals in the host rock (Ni, Co, and V). We find that both S and major element abundances can be reconciled with previous hypotheses of regolith incorporation and/or excess sulfide melt. However, trace element characteristics of the shock melt glass, particularly Ni and Cl abundances relative to S, cannot be explained either by the incorporation of regolith or sulfide minerals. We therefore propose an alternative hypothesis whereby, prior to shock melting, portions of EET 79001 experienced acid-sulfate leaching of the mesostasis, possibly groundmass feldspar, and olivine, producing Al-sulfates that were later incorporated into the shock melt, which then quenched to glass. Such activity in the Martian near-surface is supported by observations from the Mars Exploration Rovers and laboratory experiments. Our preimpact alteration model, accompanied by the preferential survival of olivine and excess melting of feldspar during impact, explains the measured trace element abundances better than either the regolith incorporation or excess sulfide melting hypothesis does.
Electron microprobe trace element analysis is a significant challenge. Due to the low net intensi... more Electron microprobe trace element analysis is a significant challenge. Due to the low net intensity of peak measurements, the accuracy and precision of such analyses relies critically on background measurements, and on the accuracy of any pertinent peak interference corrections. A linear regression between two points selected at appropriate background positions is a classical approach for electron probe microanalysis (EPMA). However, this approach neglects the accurate assessment of background curvature (exponential or polynomial), and the presence of background interferences, a hole in the background, or an absorption edge can dramatically affect the results if underestimated or ignored. The acquisition of a quantitative wavelength-dispersive spectrometry (WDS) scan over the spectral region of interest remains a reasonable option to determine the background intensity and curvature from a fitted regression of background portions of the scan, but this technique can be time consuming and retains an element of subjectivity, as the analyst has to select areas in the scan which appear to represent background. This paper presents a new multi-point background (MPB) method whereby the background intensity is determined from up to 24 background measurements from wavelength positions on either side of analytical lines. This method improves the accuracy and precision of trace element analysis in a complex matrix through careful regression of the background shape, and can be used to characterize the background over a large spectral region covering several elements to be analyzed. The overall efficiency improves as systematic WDS scanning is not required to assess background interferences. The method is less subjective compared to methods that rely on WDS scanning, including selection of two interpolation points based on WDS scans, because “true” backgrounds are selected through an exclusion method of possible erroneous backgrounds. The first validation of the MPB method involves blank testing to ensure the method can accurately measure the absence of an element. The second validation involves the analysis of U-Th-Pb in several monazite reference materials of known isotopic age. The impetus for the MPB method came from efforts to refine EPMA monazite U-Th-Pb dating, where it was recognized that background errors resulting from interference or strong background curvature could result in errors of several tens of millions of years on the calculated date. Results obtained on monazite reference materials using two different microprobes, a Cameca SX-100 Ultrachron and a JEOL JXA-8230, yield excellent agreement with ages obtained by isotopic methods (Thermal Ionization Mass Spectrometry [TIMS], Sensitive High-Resolution Ion MicroProbe [SHRIMP], or Secondary Ion Mass Spectrometry [SIMS]). Finally, the MPB method can be used to model the background over a large spectrometer range to improve the accuracy of background measurement of minor and trace elements acquired on a same spectrometer, a method called the shared background measurement. This latter significantly improves the accuracy of minor and trace element analysis in complex matrices, as demonstrated by the analysis of Rare Earth Elements (REE) in REE-silicates and phosphates and of trace elements in scheelite.
Lunar and Planetary Science Conference, Mar 1, 2009
Introduction: Advances in instrumentation and analytical techniques using electron-probe microana... more Introduction: Advances in instrumentation and analytical techniques using electron-probe microanalysis (EPMA), and microbeam X-ray fluorescence spectrometry (µXRF) have been applied to the study of lunar samples. These techniques provide compositional information from discrete spots and in digital mapping mode, over a range of spatial resolutions that is appropriate for micro and macroscopic analysis of complex mineral chemistry and textures. We have applied ongoing research and development of EPMA and µXRF algorithms in simulation and analysis software, defocused-beam analysis (DBA), and hyperspectral-imaging techniques to study lunar samples. Improvements in Quantitative X-ray Analysis: Improvements in EPMA and µXRF instrumentation has enabled large-scale X-ray mapping of lunar samples. The advent of silicon-drift (SDD) energy-dispersive (EDS) detectors with high-throughput digital pulse processing electronics has dramatically decreased mapping acquisition times and increased sensitivity. EPMA provides a relatively high spatial resolution of ~1 µm and allows sampling of up to 50-100 µm spot in defocused-beam analysis (DBA), and can analyze a wide range of elements including low energy X-rays, but trace element detection is ultimately limited by continuum Xray production. Analysis by µXRF typically uses a minimum spot size of 50-100 µm, has superior trace element sensitivity and can be used on materials which are not vacuum compatible or exhibit electron-beam damage in the microprobe, but requires thicker samples for analysis due to X-ray penetration and has relatively poor low energy X-ray sensitivity. The two techniques are complimentary for these reasons. EPMA has seen improvements in Φ(ρz) algorithms and more accurate sets of mass absorption coefficients have improved quantitative analysis by both techniques. Quantitative EPMA corrects for atomic number (Z) effects of electron scattering and retardation during X-ray generation, absorption (A) of X-rays in the sample along the path to the detector, and fluorescence (F) of X-rays by both characteristic and continuum X-rays. The k-ratio is the measured X-ray intensity in the sample relative to that in the standard, and the calculated concentration C is determined by iterative calculation from C = k * ZAF. This iterative correction implicitly requires knowledge of the concentration of all elements in the sample volume and their effect on the element of interest. The
The composition of clinopyroxene and clinopyroxene-liquid (Cpx-Liq) pairs are frequently used to ... more The composition of clinopyroxene and clinopyroxene-liquid (Cpx-Liq) pairs are frequently used to calculate crystallization/equilibration pressures in igneous systems. While canonical uncertainties are often assigned to calculated pressures based on fits to calibration or test datasets, the sources of these uncertainties (and thus ways to reduce them) have not been rigorously assessed. We show that considerable uncertainties in calculated pressures arise from analytical error associated with Electron Probe Microanalyser (EPMA) measurements of Cpx. Specifically, low X-ray counts during analysis of elements with concentrations <1 wt% resulting from insufficient count times and/or low beam currents yield highly imprecise measurements (1σ errors of 10-40% for Na 2 O). Low analytical precision propagates into the calculation of pressure-sensitive mineral components such as jadeite. Using Monte Carlo approaches, we demonstrate that elemental variation resulting from analytical precision alone generates pressures spanning ∼4 kbar (∼15 km) for a single Cpx and ∼6 kbar for a single Cpx-Liq pair using popular barometry expressions. In addition, analytical uncertainties in mineral compositions produce highly correlated arrays between pressure and temperature that have been previously attributed to transcrustal magma storage. Before invoking such geological interpretations, a more mundane origin from analytical imprecision must be ruled out. Most importantly, low analytical precision does not just affect the application of barometers to natural systems; it has also affected characterization of Cpx in experimental products used to calibrate and test barometers. The impact of poor precision on each individual measurement is often magnified by the small number of measurements made within experimental charges, meaning that low analytical precision and true variability in mineral compositions have not been sufficiently mediated by averaging multiple EPMA analyses. We compile the number of Cpx measurements performed in N = 307 experiments used to calibrate existing barometers, and N = 490 new experiments, finding ∼45% of experiment charges were characterized by ≤5 individual Cpx analyses. Insufficient characterization of the true composition of experimental phases likely accounts for the fact that all Cpx-based barometers exhibit large errors (± 3 kbar) when tested using global experimental datasets. We suggest specific changes to analytical and experimental protocols, such as increased count times and/or higher beam currents when measuring low concentration elements in relatively beam resistant Cpx in experiments and natural samples. We also advocate for increasing the number of analyses per experimental charge, resolving interlaboratory analytical offsets and improving data reporting. Implementing these changes is essential to produce a more robust dataset to calibrate and test the next generation of more precise and accurate Cpx-based barometers. In turn, this will enable more rigorous investigation of magma storage geometries in a variety of tectonic settings (e.g. distinguishing true transcrustal storage vs. storage in discrete reservoirs).
Uploads
Papers by John Donovan