The Tracer 5i
Lee Drake
Hi all,
With the Tracer 5i announced and beginning production, I wanted to send some details on the unit. While I do work for Bruker as an application scientist, I mainly want to provide technical details on the unit. I will also detail software in the second half of this email. Spoiler alerts: Artax can now communicate directly with the Tracer 5i, and the mathematical algorithms underlying S1CalProcess have been pulled from Excel and are now in a stand-alone application.
Tracer 5i: What is different?
Long story short, a higher powered tube, larger detector, and an improved onboard computer. But there are two big things that are worth noting: faster data acquisition, and light element sensitivity.
In our early testing, we noticed that the higher powered tube combined with the larger area detector (30 mm2) meant that we got 3X the counts for light and trace elements for the same energy and current settings. I am attaching results from our early tests here (New Tracer Comparison.pdf). In general, if an assay took 60 seconds on the Tracer IIISD, it will take 20 seconds on the Tracer 5i. And that is before we start increasing the current, which is more easily done. In general, the higher the energy, the lower the current. But we can make much more interesting combinations now - for example, we can run a 6keV assay at 195 μA, targeting light elements.
Traditionally, this would be when we would talk about using the Tracer to see sodium (Na), a difficult element to analyze with EDXRF in general, and handheld units in particular. But this time, we can skip over sodium, and even neon, and instead talk about fluorine (F). One of the tests we run to look at instrument contaminants is analyzing teflon (C2F4). When doing the with a helium purge, I noticed a funny peak at 0.68 keV. I had to change the DPP settings of the instrument to get a full view (we don't display the spectrum in that energy range), but we had a surprisingly large K-alpha peak for F. We did not have a K-beta, but only because K-beta does not exist for elements on the second row of the periodic table. We tested this with geologic materials as well (fluorite and cryolite) and were able to replicate these findings with denser, natural materials. I am attaching a PDF here that gives more details (Fluorine Analysis.pdf). The peak is big enough to cause optimism that even lighter elements might be possible, but here we run into a wall - the 7.5 μm Be window is infinitely thick to the next element, O. We are exploring other materials for a detector window, but for now, F is the limit.
The reason light elements have improved is not so much changes in hardware (though the 30 mmm2 detector helps) but rather in geometry. A physicist at Bruker who some of you may know, Robert Shannon, spent a fair bit of time trying out multiple geometries. His objective was to position the tube and detector radially so that coherent scatter (Bragg peaks) would not appear near elements like silicon, but rather cluster where a less common element like scandium would appear. The result was a massive boon to light element sensitivity.
The most interesting consequence of this is that it changes the rules for when a vacuum is needed. The last gen Tracer needed a vacuum to significantly improve detection of elements lighter than calcium (3.69 keV). There was a slight cost to other elements due to the vacuum grid, which had a secondary effect of absorbing lower energy x-rays, but the benefits of the vacuum far outweighed the costs of the vacuum window. That same calculus doesn't exactly hold with the Tracer 5i. In our testing, we found that the Tracer 5i with a vacuum and vacuum grid had the same sensitivity to Si as it did without a vacuum and with a clear window. That is to say, if Ca was the threshold for whether or not to use a vacuum before, Si is today. Elements like Na, Mg, and Al will benefit from the vacuum still. It is worth noting that the sensitivity to light elements like Si in the Tracer 5i with no vacuum is greater than the Tracer IIISD with vacuum (see New Tracer Improvements.pdf).
With the last gen Tracer, we focused on 4 customizable aspects; energy, current, filter, and atmosphere. Here we add a 5th: collimator. The Tracer 5i can be configured right now with 8 or 3 mm spot. The same component also hosts a manual filter as well. Now, instead of having to use superglue and filter holders, you can simply use a hole punch and make as many as you want. This apparatus sits above an automatic filter wheel with 5 slots. Before, someone had to choose the Tracer IIISD (Manual Filter) and the Tracer IVSD (automatic). Now, you can use both in the same unit. As an added plus, this opens up opportunities for double filtration of the beam. We haven't explored this fully, but there are lots of combinations of tricks to increase visibility of some elements over others.
The other big differences have to do with usability. There is now an onboard touch screen display which can control the instrument, and even facilitate qualitative spectral analysis. Calibrations can be loaded up. The onboard computer can also show the camera image, with an identifier for where the spot is.
There are a couple instances where the last generation Tracer may has some advantages, depending on use. All the above are big steps, but there are a couple compromises that come with them. The first is the size of the nose - it is 60% bigger than the Tracer IIISD - this is a result of the manually changeable collimator and the altered geometry of the tube and detector. However, there may be a silver lining. The design of the manual collimator opens up the possibility of slightly changing the geometry of the beam, which could allow for non-contact measurement. We are still working on this, and cannot guarantee an outcome. But the full implications of a manually changeable collimator have yet to be realized.
We have noticed an Ni artifact in the spectrum, coming from the detector can. We have reduced it to about 20% of what it was using a combination of techniques, and are optimistic we can can reduce it further. But for now, if Ni is a critical element to visually identify in the low ppm range, the Tracer IIISD may be a better fit. We have found we can still quantify Ni to the ppm level - the calibration curves are just offset from 0 (see New Tracer Improvements.pdf).
Software
We have taken many, many, many suggestions on how to handle software for the Tracer platform. These past two years, we finally got to implement them. The old Tracer software package was as follows:
- S1PXRF: for taking data
- S1CalProcess: for quantifying data
- XRayOps: for changing the energy/current of the instrument
- Artax 7: for qualitative and semi quantitative analysis
The new software includes:
- Artax 8: for taking data, qualitative and semi-quantitative analysis, changing instrument parameters
- EasyCal: for building calibrations
- Bruker Instrument Tools: taking data, changing instrument parameters
- CalToolkit: for batch quantification of data
- S1RemoteControl: for controlling the unit from afar
Artax 8
For those of you who used the Bruker Artax system, this is going to look familiar. You can now adjust energy, current, and filter. You can save pdz files, convert to txt files, and build databases. Deconvolution is still an option, and we have added automatic corrections for escape and sum (pile up) peaks. Everything else works the same. We will have legacy versions of Artax (7.4) if you prefer, but Artax 8 can do all the heavy lifting of XRayOps, S1PXRF, and legacy Artax.
EasyCal
This includes the same algorithms that were employed with S1CalProcess, but now in a stand alone application. The interface is quite different, and I will be updating XRF.guru to show how it works, but for now, the changes:
- You can now see the calibration curve
- You can customize normalization for each element within the same cal
- You can build rules (e.g. switch to K-beta if element X is above 100 ppm)
- You can upload the cal to the instrument directly.
Bruker Instrument Tools (BIT)
This will be the equivalent of the doctor's office for the XRF. You can install software updates, upload calibrations, take spectra, and pull data. from the unit. This can be thought of as a modernized S1PXRF, but with more functionality.
S1Remote
This software is just a mirror of the screen on the unit - you can use this to adjust the unit's onboard computer from afar.
CalToolkit
This software is used by our applications team for industrial instruments, but we will be including it with the Tracer5i - it can take a calibration file and apply it to a large data set (similar to the ChemTests sheet in S1CalProcess).
Connectivity
The Tracer's big advantage has always been live spectrum on a computer - that is now possible with either BIT, Artax, or S1Remote. Before, you had to manually identify the com port. That is no longer necessary - you can just select your instrument from a menu. What is different is that you can do this using wifi, in addition to a USB cable. Another change is that the USB cable is no longer proprietary - you can use any microUSB cable.
There is a fair bit more that I could go into - the use of the rail for mounting, the details of the operating system, the tweaks to the software, but for now, let me answer some questions that I expect might be asked - I am happy to answer any that are not addressed in this message.
Does it cost the same?
Yes
I wake up with nightmares of using S1CalProcess. Can I just use EasyCal with my old Tracer?
Yes
Can I use the new version of Artax with the old Tracer?
To analyze data, yes. But it will not connect and run it - you will still need S1PXRF for that.
Are the two versions of Artax comparable? Can I open up existing projects in the new version, and vice-versa?
Yes
I like the old software (S1PXRF, S1CalProcess) - can I use it with the new Tracer?
Yes and no. With the new Tracer we are updating the file format (.pdz) to include some of the new sensors included. This eliminates compatibility with older software. However, you can choose to take data from the Tracer using the older format (v24) of the .pdz files which work with older software.
Do I need a vacuum?
If you are analyzing Na, Mg, or Al, yes. If you are looking at Si or higher, a vacuum won't offer any improvement.
Can I use helium?
Yes - and we will finally send you a kit to use with a helium tank, so it is not a do-it-yourself project anymore.
What Calibrations Come with the Instrument?
We will be preloading every unit with Precious Metals, additional cals will be available when you order. These include obsidian, mudrock, and alloys (including the CHARM set of older bronze alloys). We will also have our industrial calibrations available (alloys, soils) but these can be problematic when publishing data. We are talking about additional calibrations as well, but no details on those yet.
Thats it for now, thanks for reading this far. If you have any questions please let me know.
Thanks,
- Lee
Vanessa Muros
--
Visit the pXRF for Cultural Heritage website at:
https://sites.google.com/site/pxrfgroup/
---
You received this message because you are subscribed to the Google Groups "pXRF for Cultural Heritage" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pxrf+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
Heather Walder
I also appreciate the info on the new Tracer 5i. Is it possible for those of us working with older Tracers to obtain the EasyCal stand-alone application, as well as the updated version of Artax? I'm currently working with a Bruker Tracer III-SD.
Thank you,
Heather Walder
Visiting Assistant Professor
Michigan State University