
Parallax's new Low Energy X-ray Spectrometer is the first x-ray spectrometer specifically designed for low energy x-ray analysis to provide very high count rates, Peak-to-Background ratios and better than 18 eV energy resolution for x-ray energies below 2.5 KeV (<5eV at Si) -- the region where most spectrometer systems have problems. LEXS uses proprietary x-ray collimator optics in a patented configuration to produce detection sensitivity for low energies better than any other commercially available system. LEXS is intended to be used with electron microscope systems with low beam currents and low acceleration voltages while still achieving very good detection sensitivity for problem elements. Use of low accelerating voltage limits analysis to depths less than .5 micron to avoid confusion with deeper layers. Because of its unusual optical configuration, LEXS can be placed on most electron microscopes including TEM, FESEM, and ESEM systems with port diameters larger than 30mm. There is no prescribed source to optic distance and no preferred orientation allowing it to be placed on almost any EM system. LEXS is sufficiently small and of low weight that it can be placed in a crowded environment. Since it requires no detector cooling, there is no liquid nitrogen filling. LEXS can also be configured for small spot XRF systems. Because LEXS runs under Windows™, graphical output can be easily incorporated into spreadsheets and documents. |
| Mo/B4C 2d = 200 | For very low energy lines such as Be and B |
| Cr/Sc 2d = 80 | For C and N |
| W/Si 2d = 60 | For energies between 500 and 1000 eV |
| TAP 2d = 30 | For energies between 500 and 2000 eV |
| PET 2d = 8.79 | For energies between 1700 and 2400 eV |
| Alternate choices: | |
| 2d = 120 | Specifically for C. This diffractor has limited use other than C but gives
the 5750 cps/na that we quote. Unless you really want C sensitivity, we do not recommend it because you will have to give up sensitivity to other elements and the standard diffractor (2d = 80) works reasonably well for C. |
| 2d = 160 | A new diffractor from Osmic specifically for B that gives
about 50% higher counts for B and also works well for Be. |
| Cr/Ti 2d=80 | Specifically for the Ti (L a) line and it suppresses the N line. |
|
|
LEXS was specifically designed for low energy x-rays where Energy Dispersive Spectrometers have their biggest problems so we did not worry about performance above 2 KeV at the time. However, several users have asked if it could be easily modified to cover the Lead-Molybdenum-Sulfur overlap region (about 2300 eV) and slightly higher with less background. Keeping the entire system optimized for lower energy is more important than achieving higher count rates for this 2300-3300 eV region but getting the background low enough would greatly extend the utility of the system. So, we looked for easy ways to get rid of background at these energies resulting in the improved LEXS. The improved LEXS has the same diffractors to keep optimum performance below 2 KeV but we have managed to greatly reduce background above 1500 eV using an anti-scattering filter so that the high energy PET diffractor can be used over more of its range. Fig. 1a. shows the Mo (L) lines without the background reducing filter (old LEXS) and Fig. 1b. shows them with the background reducing filter (new LEXS). The peak count rates are hardly changed but the background is greatly reduced. LEXS becomes useful for the region between 2200 eV and 2600 eV covering several well known overlaps and it becomes possible to use higher energy lines, such as the U (M) lines for alignment. Without the anti-scatter filter, the P/B for the Si (K) line is about 200. The anti-scatter filter increases the P/B to over 600 greatly improving the sensitivity for this spectral region. |
Fig. 1a. Molybedenum (L) lines without Antiscatter filter.
|
Fig. 1b. Molybedenum (L) lines WITH Anti-scatter filter. Note that the Mo (L1, and Ln) lines are readily visible.
|
Fig. 2a. PbS spectrum from improved LEXS. Peaks left to right are: S(Kα), Pb(Mα) and Pb(Mβ). The S(Kβ) is just to the right of the Pb(Mβ) but is not resolved.
|
Fig. 2b. PbS spectrum with optional Ge diffractor.
|
|
In this spectral region, there are about 8 peak overlaps that could not be resolved with the old LEXS and EDS but with this improvement, all of these can be resolved. Above 2600 eV, there are no significant overlaps that cannot be resolved with LEXS and EDS. |
Some Applications: |
|
Metallurgy
Be in Beryllium Copper Identification of alloys Diffusion of O2 into metals Carbon content of steel Quality control of alloy content Identification of contaminants Cr monolayers on electroform mandrels Detecting surface oxidation on metals |
Semiconductor
Wafer Dopant Analysis Thin Film Analysis Strontium Titanate on Silicon Boron in BPSG glass WSi Stoichiometry Oxide layer Stoichiometry TiN, TaN Stoichiometry Detection of surface oxidation TaSi2 |
|
Coatings
QC of tool coatings (TiN, TiAlN, etc) QC of optical coatings (TiO2) SiO, SiO2) Phosphorus content of electroless Ni Oxidation of coatings |
Other
Dopants in Sol-Gel Optics MgF Analysis of Stardust Comet Dust |
![]() |
![]() |
| Spectrum of GaAs sample showing the Ga and As alpha and beta peaks. |
![]() |
![]() |
![]() |
![]() |
| Manganese Fluoride (not Magnesium Fluoride) showing an O line Mn (L1), Mn (La), and F lines. |
TaSi2 showing peaks only 30 eV apart. |
Physical and Performance Characteristics of LEXS |
|
Size 25 cm. X 25 cm. X 10 cm. The motor drive projects 5 cm. beyond the 10 cm. on part of one side. Detector Gas flow proportional counter using P10 gas (90% argon and 10% methane). No detector cooling is needed. 0-3000 V power supply. Software Runs on a PC under Windows™. (Trademark of Microsoft) |
| Measured LEXS Performance | |||||
|---|---|---|---|---|---|
| Element | Energy(eV) | Counts/sec/na | P/B | Resolution (eV) | Sensitivity(ppm) |
| Be 10KV | 108 | 400 | 40 | 8 | 100 |
| B 10KV | 183 | 6000 | 100 | 18 | <20 |
| C 10KV | 277 | 5750 | >100 | 18.6 | 14 |
| N(BN) 10KV | 392 | 416 | 40 | 16 | 130 |
| O(SiO2) | 525 | 375 | 80 | 17 | 60 |
| Mg 10KV | 1254 | 600 | 400 | 14 | 18 |
| Al 10KV | 1487 | 500 | 300 | 19 | 25 |
| Si 25KV | 1740 | 550 | 500 | 5 | 15 |
Questions and answers about LEXS.Q. I have a light element capable EDS system, Why would I need LEXS? A. There are two answers: resolution and sensitivity. If you want to analyze very thin films or achieve good analytical resolution, you should be using low accelerating voltages (5 KV or less) to avoid exciting the substrate or lower layers. This means you will be using low energy K, L and M lines. There are so many peak overlaps using these lines that EDS is nearly useless while LEXS will easily resolve most of the necessary peaks allowing you to see spectra of films such as TiN, Strontium Titanate on Si, TaSi, and WSi, etc. For all K lines for Z<15, LEXS is much more sensitive than conventional EDS. Using L, and M lines, LEXS is usually more sensitive than conventional EDS would be using all K lines. When conventional EDS is forced to use L lines, LEXS is almost always more sensitive. Q. What is the operational energy range of LEXS? A. Qualitatively: 100 eV to 2600 eV and in some cases 3300 eV. Q. How does LEXS differ from other high resolution x-ray spectrometers? A. There are several Wavelength Dispersive Spectrometers (WDS) available, but LEXS is fundamentally different both in the way it collects x-rays and in its response to low energy x-rays. Conventional WDS uses curved crystals to disperse and focus the x-rays onto a detector. LEXS collects a large solid angle of x-rays diverging from the sample and re-directs them into a parallel beam incident on various flat dispersing elements and then into a detector. Some WDS systems are optimized for only one or two elements for each diffractor while LEXS is optimized over its entire range of operation. LEXS produces far greater count rates for low energy x-rays than any conventional WDS. Conventional WDS systems emphasize the use of high accelerating voltages for higher Z elements which is useless for thin films while LEXS is designed for low accelerating voltages. There has been considerable discussion of high resolution micro-calorimeters for x-ray spectroscopy but due to the requirement of using liquid helium and very small solid acceptance angles, we consider micro-calorimeters to be laboratory curiosities while LEXS is very practical. Q. What options are available? A. Any orientation of the spectrometer, right or left handed versions, gate valve between spectrometer and SEM and specialized software for your types of samples. Q. What about leakage of the argon from the flow proportional counter, especially in FESEM systems. A. We have recently obtained new proportional counter windows with very low leakage rate. We have also replaced the old polypropylene tubing and hose barbs with rigid steel tubing, PEEK tubing and Swage-Lok connectors to nearly eliminate this problem. However, LEXS is available with an optional stand alone pumping system that is used on a system with a gate valve. Q. If the window fails or there is other vacuum failure in the LEXS, does it shut down the SEM? A. With the existing optional gate valve, one would simply withdraw the collimator all the way into the LEXS and close the gate valve to isolate the LEXS from the SEM and continue running the SEM while the LEXS is repaired. Q. What is the best obtainable resolution with LEXS? A. We have obtained a resolution of <4 eV at the Si (K) line using a PET diffractor. We routinely obtain 5 eV at this energy. By using second order lines, the resolution can be approximately halved at the expense of reducing the peak intensities by 10X. |
PO Box 12212
Tallahassee, FL 32317 USA
phone: 850-580-5481
fax: 850-656-5076
Copyright © 1995-2008 Parallax Research, Inc. All rights reserved.