"It was during my enchanted days of travel that the idea came to me, which, through the years, has come into my thoughts again and again and always happily - the idea that geology is the music of the earth."

Hans Cloos

CU Boulder – Electron Microprobe Laboratory - JEOL JXA-8230

Question, comments? Contact the lab manager (Julien).
You may also consult the electron microprobe database for reservation request and more! Check also our examples of applications!

    General information

    An electron microprobe (EMP) is an electron microscope designed for the non-destructive X-ray microanalysis and imaging of solid materials. It provides precise and accurate chemical composition (for elements Be to U) at the micron-scale of a large variety of solid materials such as minerals, glasses, alloys, and ceramics. The primary advantage of EMP analysis is the non-destructive and in-situ character of the analysis. All you need is a well-polished, flat sample, such as a regular petrographic thin section or an epoxy mount. EMP is the ideal technique for analyzing chemically zoned crystals, for testing a material's homogeneity, for sampling delicately intermixed phases, or for identifying and characterizing phases (chemistry, size, shape). As it is an in-situ technique, information on texture and deformation can be preserved.


    The purchase of the JEOL-8230 would not have been possible without the help many researchers throughout Colorado and beyond that supported our proposal. We are extremely thankful to them. We also acknowledge financial support from NSF-EAR grant #1427626 (PIs Mahan, Allaz and Farmer; 2014-2017) and from the University of Colorado (30% cost-share and laboratory renovation).

    The NEW Electron Microprobe (2016)

    The department of Geological Sciences at the University of Colorado Boulder was awarded a Major Research Instrumentation grant in Summer 2014 to replace the 30-year old JEOL-8600 (*). In Spring 2016, a new 5-spectrometer JEOL JXA-8230 Superprobe equipped with LaB6 electron gun was installed. This new instrument uses only dry pumps (cleaner vacuum), has very good analytical stability, and delivers high quality data. Compared to the older instrument, this new EMP offers:

    - Enhanced spatial resolution (LaB6 electron gun: beam size ca. 0.2-0.7 um);
    - Five wavelength dispersive spectrometers (WDS), four of which are equipped with large-area monochromators for higher sensitivity (2 to 3x gain);
    - Faster acquisition capability, for example 10 elements in a couple of minutes, with detection limits around 50 to 200 ppm;
    - Trace element analysis capability with detection limits around 10 ppm or better;
    - A silicon-drift energy dispersive spectrometer (EDS) from Thermo Scientific with an ultra-thin window to detect light elements (Be to F);
    - A panchromatic cathodoluminescence detector from JEOL;
    - Numerous software improvements for better and faster analyses.

    Additional information is provided in the next tabs. Also check our examples of applications acquired recently.

    As usual, we welcome work from external researchers and private corporations. Contact the lab manager for more information.

    (*) Not retired! The venerable JEOL-8600 has been relocated to Auburn University (Alabama) and will soon be operational again.

    JEOL JXA-8230
    The new JEOL-8230 electron microprobe and its detectors

    Samples Requirements

    Samples must be solids, prepared flat (planar on a micron scale), very well-polished (0.05 um final polish), and must be clean and stable under high vacuum (10-4 Pa). Advise the lab manager if your sample is water- or alcohol-soluble as these liquids are used for fine polishing and cleaning. We strongly recommend samples be either mounted on a petrographic thin-section (27 x 46 mm) or a 1-inch round glass slide, or embedded in a 1 or 1 1/4 inch diameter epoxy disk (see image aside). Samples of irregular or non-conventional size could be mounted, but at the cost of complications and possible inaccurate results.

    Thin-sections or epoxy mounts can be prepared at CU Boulder. Contact Paul Boni (paul.boni [at] colorado.edu) for further information.

    Conventional samples
    Example of perfectly suitable samples for use on our microprobe JEOL-8230.

    Carbon and metal coating

    To ensure conductivity and identical analytical conditions between our standards and your unknowns, samples are usually carbon-coated (ca. 15 nm thin-film). A fully refurbished Edwards Auto306 dual-coater is available in our laboratory, and can coat up to 6 thin sections at once with carbon and/or metal (e.g., Al, Au, Ag). Whereas most applications only require carbon-coating, some specific applications such as monazite dating or beam-sensitive materials will likely require the use of metal coating for enhanced precision and accuracy. We are currently working on developing an adequate protocol for quantitative analysis with metal-coating. Advise the manager of the facility ahead of time if you need to coat your sample in our laboratory.

    Conventional samples
    Fully refurbished Edwards Auto 306 coater: Dual carbon and metal evaporation system, high vacuum with dry pumps only, up to 6 thin sections (or 8 one-inch round mounts) coated at once.

    Wavelength Dispersive Spectrometers

    The JEOL JXA-8230 electron microprobe is fully-automated, and has 5 wavelength-dispersive spectrometers (WDS), one of which is equipped with four monochromators for low-energy X-ray analysis (including light elements, Be to F), and four are equipped with large-area monochromators that reach 2 to 3 times higher count rate. The latter is key for our ability to excel in trace element analysis, and for outperforming many other microprobe laboratories. The spectrometer configuration is as follows:

       - (SP 1) TAP, LDE1, LDEC, LDEB (P-10 counter)
       - (SP 2) TAP-L, PET-L (P-10 counter)
       - (SP 3) to (SP 5) PET-L, LIF-L (sealed Xenon counter)

    The periodic table below lists which element can be detected with which monochromator. If necessary, the lab manager will assist you in choosing the most optimum setup for you work.

    Major KLM X-ray lines
    Periodic table of major X-ray lines (Ka, La and Ma) that can be analyzed with our instrument. Click on the image to see an interactive version of this periodic table (will open in a new page).

    JEOL software & Probe for EPMA

    Users of the JEOL JXA-8230 can choose to use either the JEOL software, or the more advanced “Probe for EPMA” (ProbeSoftware, Inc.) for their quantitative analysis work. Both software has their advantages, and the choice is made depending on the needs.

    JEOL Inc.

    The JEOL software is easier to use and totally appropriate for easy and quick analysis of a few elements. It has some unique features such as automatic peak identification on WDS scans, fast imaging capability, and quantitative mapping using point analysis as a calibration point.

    Probe Software Inc.

    Probe for EPMA offers greater versatility, which comes at the price of greater complexity. It is recommended for experienced users, or for special applications to ensure higher precision and accuracy. It offers unique capabilities, such as:

       - Quick analysis (10 elements in a minute) using the Mean Atomic Background correction;
       - Time-dependent intensity correction to counteract beam damage effects in carbonate, glass, alkali-rich phases, phosphate, etc.;
       - Multipoint background correction for accurate trace element analysis in complex matrix;
       - Full quantitative element mapping;
       - Improved correction routine for peak interferences;
       - Versatility in correction matrix;
       - Combined EDS-WDS analysis: use EDS for major elements, and WDS for minor and traces;

    Energy Dispersive Spectrometer (EDS)

    A new Thermo Scientific UltraDry energy-dispersive spectrometer (EDS) detector was installed with the new EMP. This silicon-drift detector has an ultra-thin window capable of detecting low energy X-rays, and thus light elements (Be to F). The optimum energy resolution at Mn Ka is 125 eV with the longest time constant (6.4 microseconds; see chart below). Major elements (and some minor) can be detected within seconds even at low current. Its ability to reach a million counts per second before saturation allows for high beam current applications. With all that, hyperspectral mapping using our EDS is also possible! Standard-based quantification and EDS hyperspectral mapping for major and minor elements is also available.

    Thermo Scientific

    The latest EDS software “Pathfinder 1.1” from Thermo is used for analysis by EDS only. This advanced software notably offers a point and shoot option for quick phase identification in a sample area and a hyperspectral mapping for quick phase identification. It has never been so easy and fast to identify the unknown!

    Conventional samples
    Full-width at half-maximum (FWHM) of several major X-ray lines measured on our EDS system. Compare to an older SiLi detector, our new EDS can detect light elements (Be to F) and an improved energy resolution (126 eV at Mn Ka versus 160 eV on a SiLi detector).

    BSE, SE and CL

    The EMP is equipped with a Secondary Electron (SE), BackScattered Electron (BSE), and panchromatic CathodoLuminescence (panCL) detectors. SE is used for relief effect, BSE for density contrast and CL to capture visible light photons emitted by some materials. Instead of a tungsten filament, our instrument is equipped with a LaB6 cathode, which offers a higher brightness, a smaller beam size (ca. 0.2-0.7 um), and ultimately better image resolution. Image resolution will of course depend on the beam current, and acceleration voltage (see FIGURE). The maximum field of view at the minimum magnification (40x) is around 3 mm, and mosaic imaging can be used to cover larger areas if necessary.

    The new BSE detector can acquire perfect images even in materials of high density contrast, without problem of under- or over- saturation. The enhanced dynamic range allows for high contrast images to reveal subtle variations in composition of less than 0.1 average atomic number (e.g., between alpha-beta brasses). The panCL detector does not replace a high-quality monochromator CL, yet it provides decent images of zircon, quartz, fluorite, and other cathodoluminescent minerals that can become helpful for quick localization of an area-of-interest. Variations of CL intensity might correspond to crystal defects (e.g. irradiation damage) or to variation of composition, often at the trace element level.

    Conventional samples
    Comparison of images obtained at variable current and voltage in a small area (FOV 12 um). Optimum imaging resolution is achieved at high voltage (25 keV) and low current (< 10 nA). Due to a more limited electron penetration depth, more surface details can be seen at low voltage. Most analytical work is done at 15 keV and 10 to 50 nA. Trace element analyses requires the use of high beam current (> 100 nA) to the cost of spatial resolution (beam size ca. 0.7 um at 200 nA).

    Element mapping

    X-ray element mapping is possible with both the EDS detector (hyperspectral mapping) and the five WDS detectors.

    The EDS hyperspectral mapping option acquires a complete EDS spectrum on each scanned pixel. In minutes, a data-cube is created that can be used to identify chemically-distinct phases. To some extent, major variation of composition within a phase can be identified.

    WDS element mapping is limited to 5 elements per pass. However, such a detector has an unbeatable sensitivity compared to EDS, which makes it ideal for mapping small variations of composition, or for minor or trace elements. WDS element mapping can also be used to map an entire thin section quickly to specifically look for accessory minerals (e.g., monazite, zircon, apatite).

    WDS mapping can be done either in beam scanning mode for a small area (< 50-100 microns), or in stage scanning mode for a larger area. We can now provide the user with fully quantified WDS element maps with incredible quality! As of January 2017, EDS mapping is limited to beam scanning mode only (maximum field-of-view ca. 3 mm), and combined EDS-WDS mapping only works in beam scanning mode. We are working with Thermo, JEOL, and Probe Software to implement combined EDS-WDS stage mapping.

    Mapping time for WDS or EDS depends on the desired spatial and analytical resolution, and on the type of material to be mapped. Quick EDS or small-area WDS maps can be done in a few minutes, but larger area maps or mapping for minor to trace elements will likely require several hours. See the example of applications section for some images.

    Resulting images can be exported as TIF or Bitmap file format. Quantitative maps acquired with Probe Image can be exported in grid files to be treated with third party software such as Surfer from Golden Software, Inc.

    WDS element mapping
    Example of a fully quantitative X-ray element mapping in calcite (shell debris, courtesy K. Chin, CU Boulder).

    You deserve the best

    The quality of analyses performed depends essentially on the quality of sample preparation, the character of the sample material, and the availability of appropriate primary and secondary standards for the desired elements. We are dedicated to providing users with the best results, and if necessary, we will work together to develop new and optimum analysis routines to cover your needs. Regular testing, calibration, and a service contract with JEOL guarantees the instrument is always performing at its optimum.


    The precision of measurements on the electron microprobe is a function of X-ray counting statistics, which depend on the total number of X-ray counts collected on both the standard used for calibration and the sample. The new instrument can attain precision better than 0.3% (relative) on major elements, as determined by numerous replicate analysis of primary and secondary standards over an extended period. Beam current, counting time, overvoltage and the geometry of the system can affect the X-ray count rate, and thus the precision. To reach acceptable precision (ideally better than 2-5%), trace element analysis requires both high current (> 100 nA) and long counting time (several minutes).

    Beam sensitive materials

    The dense package of highly energetic electrons thrown at a solid material can locally produce significant heat, which can potentially damage the material and yield inaccurate results. Hydrous materials, glasses, carbonates, phosphates, and alkali-rich phases (among others) are more prone to devolatilization, diffusion, or even amorphization (melting) or ablation. In such materials, it is often necessary to use a lower current (lower precision) and/or to increase the electron beam size (lower spatial resolution). To improve accuracy, Probe for EPMA can correct for potential beam damage using a time-dependent intensity correction on the first 5 analyzed elements. In the near future, we will be testing the use of metal coating instead of carbon to minimize beam damage effects (e.g., aluminum coating for monazite U-Th-Pb dating, silver or aluminum on carbonate).


    The accuracy of measurements on the electron microprobe depends on accurate knowledge of the composition of primary standards, and the exactitude of the matrix correction used (so-called ZAF or φ(ρz) procedure) and other corrections (e.g., peak interference correction). A general accuracy statement cannot be made, although it is typically better than 2%. Accuracy can be affected by peak interferences if not properly accounted for. When analyzing trace elements, peak interference corrections become crucial, as does the accuracy of the background correction.

    Detection limit

    The detection limit varies depending on the analyzed element and the monochromator used. For most elements, the new EMP can achieve detection limits on the order of 50 to 100 ppm within a couple minutes. The aggregation of multiple spectrometers, use of higher current (> 100 nA), and long counting time (10’ or more) permits the reaching of a higher sensitivity and detection limit around 1 to 10 ppm.

    Analyzed volume

    In most cases, the analytical volume is in the range of a few cubic microns or less (~picograms). The exact volume analyzed chiefly depends on (a) the acceleration voltage used (higher voltage = higher volume), (b) the X-ray energy line targeted (higher energy = smaller volume), (c) the density of the analyzed material (lower density = larger volume), and (d) the electron beam diameter. A Monte-Carlo simulation program such as Casino and Win X-ray can help researchers to evaluate the volume.

    Standards and Reference Materials

    Excellent standards and reference materials are crucial to insure accuracy of microprobe analysis. Our collection of standards and reference materials (synthetic and natural materials) has recently been completed, and now fully covers the need for most applications (silicate, REE-minerals, sulfides, carbonate, metal alloys, etc.). We will continue to improve it further depending on the analytical needs. Check our microprobe database (De-MA) for a list of currently available standards. For more information on standards, you can also visit the Focused Interest Group on Microanalytical Standards (https://figmas.org) created by Julien Allaz (CU Boulder), Anette von der Handt (University Minnesota) and Owen K. Neill (Washington State University) in 2016, under the umbrella of the Microanalytical Society and the Microscopy Society of America.

    Time availability

    There are two time slots available for use of the microprobe:
       - Day: 8 AM to 5 PM
       - Evening: 5 PM to 8 AM the next day (*).

    Anyone using the microprobe must be trained, and must know the theory of X-ray diffraction and electron microprobe in general. See the Teaching tab for more information.

    (*) Only experienced users can use the probe during the evening and overnight, as no support is guaranteed past 6 PM.

    Reservation, Minimum Use, and Cancellation Policy

    Users should sign up for time only if their samples are ready and if they have no other schedule conflicts. If more than one day of analysis is required, it is highly recommended to reserve consecutive days. When sending a reservation request, the researcher is required to specify what materials and what elements will be analyzed to help the lab manager or his assistant to prepare the instrument. When you sign up to use the microprobe, you agree to use the instrument at that time, for a minimum of 1 hour per session. Cancellation is possible if notified a week ahead of time and if a valid excuse is given. With less than a week's notice, cancellation is no longer an option and a two-hour minimum will be charged (per day).

    Refer to the microprobe database De-MA for the online agenda and reservation request. Any reservation request must be approved by the lab manager.


    Priority for use of the probe is as follows:
       1. Maintenance
       2. Lab business and submitted jobs
       3. All other users


    You deserve the best possible results from the instrument and the manager of the facility, within the limits of the techniques involved and human nature. If you have any problems whatsoever, feel free to discuss these issues with the manager of the facility. We all want this laboratory to operate in an efficient and productive fashion!

    Pricing and conditions

    We are open to in-house and outside researchers! The manager of the facility is usually available to assist Monday to Friday, from 8 or 9 AM to 5 or 6 PM, unless otherwise stipulated (official holidays, vacation, illness...).

    Untrained users will receive assistance to obtain their analysis. Only advanced and trained users may work without assistance, and thus may benefit from the unassisted rate. A quick training can be given on the first day of the analysis session (additional charges apply). Contact the facility manager for pricing and conditions.

    Concluding remarks

    For some basic work, a couple hours of training might be sufficient, providing the user knows the theory, the basic principles of an electron microprobe, and the software controlling the instrument (JEOL and Thermo EDS programs, and if required Probe for EPMA and Probe Image).

Julien M. Allaz (University of Colorado Boulder) © update February 1st, 2017
Background picture: View on the southern Alps from Pizzo Forno (Ticino, Switzerland; J. Allaz © 2008).