Quick Start Guide

Summary

1. Turn on the helium and start purging air from the prism. Check to make sure that the helium is bubbling in the outflow bottle.

2. Wait for the air to be fully purged (~8-10 minutes on the old machine, and ~4-5 minutes on the new machine).

3. Analyze the standards and check the results to ensure that the instrument is working properly. Look for the absence of an Ar peak, good separation of Kα and Kβ peaks for Ca and heavier elements, and large counts (>150k cps).

4. Scan your cores. Keep an eye on the data, and make sure that the helium stays bubbling throughout the day.

5. Analyze the standards again at the end of the day, or when you want to check the system (film integrity, contamination, etc.).

Basic Instrument and Lab Safety

The main sources of risk and potential injury in the lab, in order of decreasing likelihood, are as follows:

1. Cutting yourself with a utility blade

2. Cutting yourself with a broken glass slide

3. Cutting yourself with a sharp piece of plastic

4. Hitting your head on the door of the old scanner

5. Pinching your fingers in the door on either scanner

6. Instrument failures, such as broken gas springs, broken radiation safety locks, etc.

You should notice that the top three risks involve cutting yourself. You’ll find yourself using utility blades to cut core endcaps, open d-tubes, and cut plastic off cores. We have numerous pairs of cut-resistant gloves available, and we strongly encourage users to wear them while working with the utility blades. If you feel uncomfortable using the blades, or the blades appear too dull or worn, please ask for help!

While many people are concerned about potential radiation exposure, the risk is extraordinarily low under normal operating conditions. The radiation sources are low power (100 W), and the instruments are designed such that the source cannot be active while the doors are open, even slightly. Both instruments have been inspected by the TAMU Radiation Safety Office and have been deemed to be safe, in compliance, and weak enough to not even warrant monitoring equipment. If you have any questions or concerns, however, please speak to the Lab Manager.

Instrument Controls

There are two XRF Core Scanners in the Gulf Coast Repository (GRC), and they will be referred to from here on out as the “old scanner” and the “new scanner”. Both instruments were manufactured by Avaatech, and both contain nearly identical components. They have the same model of source, same model of detector, and their geometries are as similar as possible, meaning that their analytical results should be very similar. The new scanner, however, has a different door design and most of the manual buttons and switches (as well as the onboard PLC for direct control) have been converted to digital equivalents. Figure 1 shows the controls on the front of the two scanners for comparison.

There are only a few control switches that you as a user need to know about. First, there’s a large red “Emergency Stop” button on each machine. If you feel concerned about the instrument’s behavior (strange noises, portal to another dimension opens inside, etc.), you should push the button and then alert someone in the GCR. It’s fairly easy to recover from an emergency stop, and it’s better to be safe than sorry.

Second, both instruments have buttons that open and close the doors. You need to hold the buttons down for them to operate, and there should be a loud beeping noise while the doors are in operation. The doors cannot be opened if the radiation source is active, so if you are holding the button and nothing is happening, check to see if you left the source on. Sometimes the doors will fail to open if there is a machine error, so if you cannot get the door to operate, please alert someone from the GCR. Please remember that you should always have the door either all the way open or all the way closed. When the doors are partially raised, there are extra stresses on the springs or chains and they introduce addition risks, such as bumping your head.

Finally, the old instrument has a number of additional buttons that users will need to operate. These physical buttons have corresponding digital buttons in the control software for the new instrument and will be discussed in more detail later.

Purging the Prism with Helium

X-rays are readily absorbed by matter. That’s part of what makes them so useful for analytical and diagnostic applications. But, that’s also what makes them difficult to work with at times. The fraction of photons that a material can absorb from an X-ray beam increases exponentially as the energy of the photons decrease. That’s one of the reasons why low energy X-rays are particularly difficult to measure. They tend to be absorbed by materials in the beam path before they can reach the detector.

Unfortunately, many of the important geologic elements give off characteristic X-rays in the low energy region of the spectrum. To help improve the fraction of X-rays that reach the detector, we can do a couple of things: (1) reduce the distance between the sample and the detector, and (2) replace the materials between the sample and the detector with substances that have lower X-ray absorption. Our instrument geometry is fixed (for good reason), so we can’t really do item (1), but we can do item (2). In our case, we want to replace any air in the beam path with helium. Helium has a much lower X-ray absorption than air.

Figure 2 shows the relative peak intensities for a sample when the prism (assembly that holds the detector above the sample) is filled with various ratios of air to helium. You can see that peak intensities from elements such as Mg, Al, and Si are strongly reduced in the presence of even a small amount of air. Heavier elements such as Fe, however, are barely affected. If you (as a user) are only interested in measuring heavy elements, then purging the prism with helium is not necessary. If you are interested in any of the lighter elements (even Ca), then it’s critical to displace all of the air and to keep it out.

To begin purging the prism, you’ll need to turn on the flow of helium. On the old scanner, you’ll push the “Helium” button on the front of the instrument. The light should come on if you have pressed the button long enough. On the new scanner, you’ll use the scanner control software and you’ll click on the “Helium” button in the lower left corner of the window (Figure 3). As with the physical button, you’ll need to hold the button down for a second or so to activate it.

If you’re using the new scanner, you should see the flow rate reported in the panel when the “Helium” button is pressed in the top left (different from the button used to turn the gas on and off at the bottom) (Figure 3). The new scanner has a software-controlled flushing routine that increases the flow rate to 200 mL/min for 50 seconds to speed up the purging process. Afterwards, it drops down to ~30 mL/min for steady-state flow. The older scanner does not have an automated flush routine. Helium flow rates must be adjusted manually. As a result, we usually leave it set for a steady-state flow of ~30 mL/min.

Regardless of which machine you are using, you should get in the habit of routinely checking for helium bubbles in the plastic bottle attached to the source + detector assembly (Figure 4). The bottle serves two purposes: (1) providing back pressure on the gas to help keep a positive pressure in the prism, and (2) providing an easy visual way to make sure that the prism is sealed. The prism itself is kept air-tight by two pieces of very thin and fragile polyethylene film (Ultralene) and the positive pressure of the helium gas. If one of the pieces of film develops a hole, the bubbles will stop being visible and you may start to get air mixing into the prism. If you are not paying attention, your data quality (for light elements) may begin to deteriorate and you may later falsely conclude that element concentrations are changing. Please ask one of the GCR staff to change the film if you stop seeing bubbles.

It takes ~8 minutes at the standard 30mL/min flow rate to purge the air from the prism on the old scanner. It takes ~4 minutes to purge the prism on the new scanner if the default flush routine is used (Figure 5). We commonly recommend that people simply turn the helium on first thing in the morning when they arrive, and by the time they grab a cup of coffee or tea, the prism will be purged.

Setting Up Your First Analyses - Standards

The first and last thing you’ll want to do each day is to analyze the standards. The Lab Manager or one of the Technicians may do this for you, but you’ll want to be aware of this step regardless. We don’t use standards for quantification in the XRF Core Scanner lab. We use standards to monitor instrument performance. We’re looking to see if we’ve successfully purged the air from the prism, if the source is working correctly, if the detector is working correctly, if there is contamination on the film, and if the software is working correctly. We only compare the standard spectra to ones that were previously collected. Each scanner has a set of three standards (plus a material used for orienting the beam in the old instrument) sitting in a rectangular Teflon tray. There is usually a clear plastic cover over the set to keep dust off them when not in use. You’ll want to remove the cover and secure the tray with the first two clamps, making sure the three consecutive samples are oriented towards the front (Figure 6). Make sure that the tray is flush with the beginning of the track. If you have trouble loading the tray, please ask for help. Usage of the clamps is described in more detail in the “Setting Up Core Scans” section.

The standard materials used in the instruments are not identical; only one material is found in both trays. That means you must be careful not to switch the trays between the instruments. Although there is no reason that you can’t analyze the other set of materials, you will confuse people at a later date when they compare spectra over time. The identity of each standard is printed on the side of the cups, if you are interested.

Before you can analyze any materials, you’ll need to turn on the X-ray source. If you’re using the old instrument, press and briefly hold the “X-Ray On” button located on the front of the instrument. If you’re using the new instrument, press and briefly hold the “X-Ray” button located in the bottom left panel of the scanner control software (Figure 7). In both cases, you should see the X-ray indicator lights become illuminated on the top of the instrument. If the source isn’t turning on, please notify someone from the GCR. When you’re finished with your analyses, you’ll want to turn the X-rays off so you can open the door. You simply press the “X-Ray Off” button on the old instrument or the “X-Ray” button in the control software for the new instrument.

Once you’ve purged the air from the prism, loaded the standards, and turned on the X-rays, you’re ready to set up your first analyses of the day. You’ll need to use a piece of software called “XRF Core Scanner Program” (creative name). It will normally be open on the host PC, but if it is not, you will find a shortcut on the desktop. This is the primary data acquisition program for the two scanners, and you can see a screenshot of the main window in Figure 8.

You’ll need to understand all parts of this window at some point, but let’s start with a high level overview. The main window is dominated visually by the spectrum. If data is actively being acquired, then the spectrum will update in real-time, otherwise you’ll see the last successfully acquired spectrum. The default X-axis is channel (explained in the next section), and the Y-axis is total counts. Below the spectrum, you’ll find a number of controls that modify the display. At the very bottom, you’ll find some basic acquisition statistics, such as counts-per-second (cps) and dead time (%). Above the spectrum, in the center, you’ll find a window that will contain either a progress summary (scan time estimation) or messages from the instrument or detector. At the top left, you’ll find a status panel and buttons that open new windows used to setup an analysis (see Figure 9 for a close-up).

The status panel shows you information about the current point (if measurements are ongoing) or the last successful point (if the instrument is idle). Above the status panel, you’ll see five buttons that you’ll need to be familiar with. The scanner communication button on the far left (item 1 from Figure 9) is used to connect the acquisition software to both the instrument and the detector (which has its own separate microcomputer inside the cabinet). If the acquisition software is already open on the host PC, you don’t need to use this button. But, if you’re starting the software from scratch, just click on the button and it should do all of the work for you. You should see a pop-up window identical to the one shown in Figure 10. If the connection is successful, you’ll see two green check marks. Click “OK” to close the pop-up and you’re all set. If you get a red “X”, please ask one of the GCR staff for help. The most common cause of an error here is someone forgetting to return the instrument to “Remote” mode after performing maintenance.

Once you’re sure that the software is connected to the instrument and detector, you’ll want to set up the measurement conditions that you’ll use. The second button above the status panel (item 2 in Figure 9) brings up the “Instrument Settings” window (Figure 11).

You can set the combination of voltage, current, filter, and counting time that you’ll use for measurements in two ways. First, you can set them in this “Instrument Settings” window and use them directly from the measurement window (item 1 from Figure 11). Second, you can add one or more combinations to the pre-defined excitation list, which you can then access graphically in the measurement setup (item 2 in Figure 11). We recommend that you load one of our existing lists, but if you make any modifications, you should save it as a new list (item 3 in Figure 11). The typical conditions that we use to analyze our standards are 9 kV, 0.250 mA, 6 seconds, and no filter. We use those conditions for historical reasons so that we can compare data over a long period of time. Our normal excitation list doesn’t include this combination, so we recommend you just set the 9 kV conditions in this window. Once you’ve set the conditions that you want to use (either in the window directly or in the list), click “OK” to close the window.

Next, you’ll need to set up your measurement plan. Click on the third button from the left above the scanner status panel (item 3 in Figure 9) to open the “Measurement Setup” window (Figure 12).

The “Measurement Setup” window is the window that you will interact with most often when acquiring data with the instrument. The first thing you’ll want to do is make sure you tell the software where you want to write your data files. Click on the “Change” next to “Data Directory” and navigate to the desired location (item 1 in Figure 12). If a folder doesn’t exist yet, go ahead and create one. You can look in the existing data directories to see how files are organized, or ask the Lab Manager or one of the Technicians. Next, give the sample a name (item 2 in Figure 12). For standards, we typically call it “Standard_#” where “#” represents the number of times that we’ve analyzed the standards that day. We’ll discuss core section naming later. Now, you’ll want to decide if you want to run replicates, and if so, how many. Click the check box next to “Replicate measurements” if you want them, and set the number you want in “No of measurements” (item 3 in Figure 12). Ignore “Interval”; just leave that at 1. We typically recommend running 20 replicates first thing in the morning to warm the source up a bit (increases intensity slightly), but it’s optional. We rarely run replicates later during the day.

“Sample Length (mm)” and “Step Size (mm)” are where you define your primary measurement grid (item 4 in Figure 12). The instrument will measure all points on the steps, excluding the first point (0 mm in this case) and including the last point. We’ll talk more about this grid and how to modify it later, but for the standards, the numbers are 150 mm and 50 mm respectively. This means the machine will land at 50 mm, 100 mm, and 150 mm, which corresponds to the center positions of the three standard samples.

Finally, you’ll need to tell the software which excitation conditions you want to use for your analysis. We typically only analyze standards at a single condition. If you want to use the conditions set in the “Instrument Settings” tab, select the first “Single Run” option, and if you want to use one of the conditions in the excitation list, select the second “Single Run” option and highlight the desired list item (item 5 in Figure 12).

When you’re finished, click “OK” to close the window. Now, once you’re ready, click the “Run” button above the status panel (item 4 in Figure 9). If at any point you need to stop a run before it finishes on its own, click the “Stop” button to the right of the “Run” button (item 5 in Figure 9).

Making Sure the Instrument is Performing – Standards QC

We analyze standards so that we can make sure that the source and detector are behaving properly, the prism is free of air, and the film on the bottom of the prism is free of contamination. But, what does “good” look like? How will you know, based on the standards, if the source and detector are behaving properly? How will you know if the prism is free of air and the film is clean?

First, please feel free to ask the Lab Manager and/or Technician. They will be able to quickly glance at the spectra and tell you if everything looks OK. But, if you want to investigate on your own (and we encourage you to do so), here’s what you’ll want to look for.

Look at the spectra in the main window as they’re being collected (Figure 13). Check to make sure that the numbers on the intensity scale (item 1 in Figure 13) are increasing during the measurements. Sometimes, software and/or hardware errors occur and the data does not get recorded on the PC. If that happens, you’ll need to alert someone. You don’t want to waste a lot of time measuring core sections without collecting the data! If the intensities are changing, then the data is being correctly recorded.

Next, look at the area near the rhodium L-line scattering peaks (channels 130-150-ish, or 2.5-3.1 keV) (item 2 in Figure 13). If you have trouble finding the peaks, double click on different features in the spectrum and look at the “Element Finder” window to the right of the status panel. It will help you determine which peaks might be nearby. You want to look for evidence of an argon peak. If present, it will show up on the right shoulder of the Rh L-peaks. See the inset in Figure 13 for examples of how the shoulder grows into an outright peak at higher intensities. If you see an argon peak, it means one of three things: (1) you have air in your prism, (2) you have air between your prism and your sample or in place of your sample – i.e., you didn’t land on the sample correctly, or (3) you have air inside your sample in the form of porosity. We’re only concerned about (1) when we look at our standard results. If you don’t see an argon peak, then our prism is fully purged and our film is intact.

Another thing to look at is the general sharpness of the peaks and the separation of Kα and Kβ peaks for Ca and Fe (item 3 in Figure 13). Kα peaks are larger and occur at lower energy, and Kβ peaks are smaller and occur at higher energy. You’ll want to see relatively sharp peaks, partial separation of the Kα and Kβ for Ca, and full separation of the Kα and Kβ for Fe. Double click on the peaks to help identify them. If we have good peak separation, then this tells us that the detector resolution is reasonable.

Finally, make sure you look at the general measurement statistics during the analyses (item 4 in Figure 13). You want to make sure that the counts-per-second (cps) are in the 150k – 300k range and the dead times are in the 18—35% range. This tells us that the source is reasonably intense and the detector is able to process the data at the expected rates.

If all of the above factors seem fine, there’s a good chance the system is performing well. However, it’s always worth comparing the new spectrum to an older one from the same location on the tray (i.e., only compare 150 mm to 150 mm). You can compare the intensities of your measurements to the intensities measured through time by double-clicking on the "Plot Standards" icon on the desktop. Please contact the Lab Manager if you notice that your measurements are significantly off the trend!

You can also directly compare the spectrum features through time. Click the button below the spectrum window to load another one into the view (item 5 in Figure 13). The overlay will allow you to compare peak height, peak shape, drift, etc. all at once. Small differences should be expected, but large differences should not. Doing an overlay will also help you to see if there is any contamination building up on the bottom of the prism film. Contamination will modify the relative peak heights.

One last thing you’ll need to understand when interpreting the spectrum is that the x-axis is displayed in channels, not energy. Our instruments actually bin photons of different energies into discrete channels, and due to the way that the detectors work, photon collisions are recorded as voltage / current spikes. There is a nominal relationship between channel and photon energy for each detector. If you drag your cursor along the spectrum, it will display both energy and channel in the bottom right of the window. Likewise, when you double click on a peak, the element finder uses this default relationship to estimate the peak energy. However, the relationship between channel and energy can change due to electronic drift! It is not immutable. We’ll talk about this more in other documentation, but suffice it to say that, for the purposes of basic QC, you can pretend that the energy displayed with the cursor in the window is accurate.

Basic Core Preparation

Sample preparation is an important part of most analytical procedures, and it’s an important part of XRF core scanning, too. Our challenge when analyzing core sediments in-situ is that we don’t want to remove the samples from the core itself, and we don’t want to destroy the samples in such a way that they can’t be used for future research. But, it’s critical that you as a user understand what you can and can’t do, and how those choices impact your final peak intensities. I’ll talk a great deal about sample state and peak intensities in another document, so for now, I’ll just cover the basics.

First, it’s important to understand that XRF is largely a surface measurement. If you look at Figure 14, you can see that nearly all incoming X-rays with less than ~10 keV of photon energy are absorbed within the top ~0.5 mm of sediment. The corollary to that is that any X-rays (< 10 keV) generated deeper in the sample will be completely absorbed before they can exit (and reach the detector). Given that the characteristic peaks for our major elements have < 10 keV of energy, we can conclude that we only “see” major elements in the very surface of our sediments. Higher energy X-rays can travel farther into and out of the sample, but the focal geometry of our source and detector set a practical upper limit on what we can measure. Therefore, it’s reasonable to say that the majority of our spectral lines originate with the top 1-2 mm of sediment.

The surface nature of the measurement means that we need to be acutely aware of the quality of surface that we’re measuring. Any irregularities or surface contamination (even a few microns worth) can completely alter and mask the underlying sediment signature.

If you’re working with soft sediments, the easiest way to prepare your core surface is to scrape it with a glass slide (Figure 15, left). The goal of scraping is twofold: (1) create a flat, level surface over most of the section, and (2) expose fresh, uncontaminated sediment for analysis. Ideally, the prism should be able to land flush with the sediment. It’s OK if the edges next to the core liner are a bit rough or uneven. The prism is a bit narrower than the standard 3” core diameter used by IODP, so there’s some room to give. I recommend holding the slide long-ways and tilting it in the direction of travel. That puts a slight downward pressure on the slide and the sediments in the path, creating a smoother surface. You’ll want to make sure to thoroughly clean the slide after each pass to avoid contaminating parts of the core with sediments from another depth. If you’re working with harder sediments, or even hard rocks, you may not be able to scrape the sediments very easily. You’ll want to consult with the Lab Manager and/or Technician to identify the best way (if any) to prepare the core surface.

If you’re interested in measuring the very top or the very bottom of the section, you may need to cut the plastic from the end-caps down below the level of the sediment. If the prism lands on the plastic, it will puncture the film at the bottom and your signal will start to degrade. We recommend using hook blades and cut-resistant gloves when you cut the end caps. The Lab Manager and/or Technician can show you the proper technique, or do the cutting for you if you’re uncomfortable.

The final preparation step is to cover the sediment surface with a layer of Ultralene film (Figure 15, right). Ultralene is 4 micron thick polyethylene, and it provides a barrier between the sediments and the film at the base of the prism. Without the Ultralene on the core, sediment will stick to the prism and will contaminate every analysis. To apply the film, you’ll want to unroll it down the length of the section and tape it in place. Ultralene itself leads to additional signal degradation (as do all materials in the beam path), so if you’re working with very dry sediments or hard rock, it may not be necessary. Consult with the Lab Manager if you’re unsure.

Setting Up Core Scans

When you’re ready to scan your first core section, you’ll follow a similar procedure to the one described above for the standards. First, you’ll want to make sure that the X-rays are turned off (Figure 7). Open the instrument door and load the core into the clamps. Alternatively, if available, you can secure a core holder with the clamps and then lay the section inside of it. Make sure that the top of the core (blue cap on modern IODP cores) is flush with the beginning of the track, and make sure you know which core is actually loaded. If you load the core into the machine backwards, or you load the wrong section, you’ll never figure it out later!

If you loaded the standards yourself, then you should be familiar with how to operate the clamps by now. But, if not, then you’ll need to loosen the clamps by turning the knob counterclockwise and the pushing lever all the way to the left (Figure 6). Place the core or holder in the center of the clamp, and push the lever to the right until the clamps are secure. There are springs on the clamps that will cause compression, and you’ll see a metal pin on each clamp extend outwards as the clamps tighten. Don’t over-tighten! You’ll cause damage to the cores! Once you’ve tightened the clamps, lock them into place by turning the knobs clockwise. Again, don’t over-tighten! Turn them just far enough so that you feel resistance.

Next, close the door, turn the X-rays on, and start setting up your measurement plan (Figure 16). If you ran the standards yourself, then you should be pretty familiar with this workflow. If you didn’t, then I suggest you read through the section above. I’ll only cover on the basics here, and I’ll instead focus on the additional information you’ll need to set when scanning an actual core.

Set the directory where you want to write your data, give the section a meaningful name (typically Expedition-SiteHole-Core-Section if you’re analyzing IODP cores), choose whether you want replicates (most people don’t), enter the length of your core section and the standard point spacing that you want to use, choose your excitation conditions, and choose the down-core slit size that you want to use (most people leave it at the default 10 mm) (items 1-6 in Figure 16). If you’re analyzing IODP cores, you can place your cursor in the sample name field and scan the section cap with one of the bar code scanners. That will enter the full name and section number automatically, preventing you from making typos.

If you want to analyze the section at multiple conditions, you’ll need to select “Multi-Run”. The program will then analyze the core sequentially, using all conditions in the predefined excitation list (in order). The scanner will set the conditions to the first entry in the list, scan all points, return to the start, adjust to the second set of conditions, and repeat until all points have been scanned at all conditions. If you only want to analyze the core at one set of conditions, choose “Single-Run” and select whether to use the conditions from the “Instrument Setup” tab or one set of conditions from the excitation list.

The “Special” tab (item 7 in Figure 16) is where you’ll tell the software to ignore your standard grid and include additional points, exclude points that would ordinarily be analyzed, or set up special areas where you want a different grid spacing (items 1-3 in Figure 17). It’s important to note that the “Special Areas” override all other settings, including the “Special Positions (Adds)” and the “Skip Positions”. Many people use special areas to skip portions of the core by setting the step size equal to the distance between the beginning and end points.

In a lot of cases, you’ll want to skip the very last point of the core, because your point will sit right on the sediment edge. If your section has lots of surface irregularities, such as cracks, pits, drop-stones, etc. then you’ll need to move points to avoid these obstacles. Many people will shift those skipped points to the nearest “safe” area, meaning they’ll have the original location entered under “Skipped Positions” and the new location under “Special Positions”. When choosing which points to skip, you’ll want to be aware of both the size of the analysis spot (typically ~1 cm by 1 cm) and the size of the entire prism assembly. You’ll want to avoid surface features such as cracks and pits in the analysis spot, but you’ll want to avoid sharp features such as rocks and end-caps under the entire footprint of the prism. If any part of the prism lands on one of those sharp objects, you’ll break the film. We have a handy tool available in the lab that will help you identify which points you’ll want to avoid. Ask the Lab Manager and/or Technician for assistance.

Finally, it’s worth pausing for a second to talk a bit more about measurement conditions and resolution. If you’re using the XRF Core Scanner lab, then at some point you’ll have a conversation with the Lab Manager about your scientific objectives. Our standard three-energy excitation program in the lab may or may not be adequate for your purpose, and we can work with you to develop a more personalized measurement plan. Your choice of resolution will depend largely on the heterogeneity of the cores and the preservation state of the material. Table 1 shows the approximate time that it takes to scan a standard 1.5 meter section using different combinations of resolution and excitation conditions. These numbers are based on averages from real scans, but they are just guidelines and may vary on a case-by-case basis. Scanning always involve trade-offs. More excitation conditions means more elements measured. Higher resolution means more data points per section. But, there’s a time cost for analyses, and though you’ll get more data per core, you won’t be able to scan as many sections.

Basic Data QC and Performance Monitoring

You should apply the same basic rules when monitoring your core scans as you do when monitoring the standard scans. If you skipped the section about standards QC, I suggest you go back and read it now. Here are some reminders from that section, as well as some tips that are more applicable when scanning sections.

When you start a new scan, watch the main window to make sure the spectrum starts updating. Keep an eye on the X-position in the status panel to make sure the prism is landing where you expect for the first few spots. People will commonly make mistakes when typing numbers in the measurement setup, and if you catch the errors early enough, you can hit the stop button, make changes to the program, and restart without wasting a lot of time.

During the first few analyses, look through the window of the instrument. Is the prism landing squarely on the sediment? Is the helium still bubbling? Did you leave any objects sitting on the core that might get in the way at some point? Again, this is your chance to stop the run and fix things before wasting a lot of time.

The most important things that you’ll want to do are periodically check the quality of the spectra, and periodically check the helium bottle for bubbles. Keep an eye out for the appearance of an argon peak. If the helium is bubbling nicely and you’re seeing an argon peak, your prism is not landing on your sample correctly. A few bad points here and there is fine, but if all of your points have argon peaks, you may want to stop the run and try to level the surface some more. Keep an eye on the counting statistics. Make sure the counts-per-second are staying relatively consistent. Drastic changes suggest a potential problem developing. Also, a good time to check on the instrument is when it resets to a new excitation condition. You’ll want to make sure that the source and filter successfully adjust and that there are no errors.

If you’re analyzing your sections using the standard lab conditions, you should see spectra that look vaguely like those in Figure 18. Your spectra will of course look different, because your compositions will be different, but there will be certain similarities.

At 10 kV, there will be very little background. You’ll mostly see characteristic peaks. At 30 kV with a thick palladium (Pd) filter, you’ll see a significant background “hump” at higher energies, near the Rh K-line scattering peaks. And, at 50 kV, you’ll likely see very little in the default zoom window, which confuses many people. The reason for the apparent lack of data, however, is that most of the peaks are off the screen. The default x-scale in the window only goes up to around channel 1000, or ~20 keV. The y-scale is auto-adjusted to the highest peak in the entire spectrum, and in the case where those tall peaks are off the screen, it appears as though the rest of the intensities are unusually “low” (Figure 19). You can adjust the x-axis scale to view the full spectrum at any time by clicking the arrow buttons below spectrum in the main window (Figure 19).

Spectra Interpretation and Processing

Now that you understand how to collect your data, you’re probably eager to know how to interpret it and turn it into useful numbers. We’ll cover some of that in the data processing and quality assurance sections. It's an expansive subject, and there's a lot to take into consideration. When you process an XRF spectrum, you’re making an interpretation. You’re leaving behind the realm of data acquisition, and you’re entering a territory where you as the scientist need to take ownership for the assumptions that you make. We can help, and we’re happy to sit with you and walk you through both basic and advanced approaches. We’ll make sure you leave our lab with numbers in-hand, and our Lab Manager will be available for follow-up questions if and when you have them.

Happy scanning!