Choose the right tool for the right job

and-held X-ray fluorescence spectrometry (XRF) has been established as the gold standard for in situ elemental determination in metals and alloys in the last 10 years. More than 5,000 analyzers are sold worldwide each year for scrap metals sorting and positive material identification (PMI, a quality assurance tool consisting of the identification of alloy grades in various industries to verify compliance to specification and to avoid mixing up of the materials.)

Hand-held laser induced breakdown spectrometry (LIBS) is an emerging method that shows promising capabilities for alloy analysis and may be able to complement XRF devices in alloy analysis, especially in alloys containing a low atomic number, or light elements like beryllium, lithium, magnesium, aluminum and silicon.

These methods of analysis are based on different principles.
 

XRF analysis

XRF is a method of energy dispersive XRF (EDXRF) in which the radiation produced by a miniature X-ray tube strikes the sample surface and causes ionizations of the inner shell of the atoms constituting the sample. The resulting vacancies in the inner shell of the atom are filled by electrons from higher shells, and thereby photons specific to the element are emitted and detected with a silicon detector. Because XRF involves transition within inner shells of atoms, the spectra obtained will contain a limited number of lines: typically two to six resolved lines per element in EDXRF.

For metal analysis, XRF can work simultaneously for determining elements ranging from titanium to lead within a few seconds.

When needed, a second beam condition is used to determine light elements resulting in longer measurements of typically 10 to 60 seconds.
 

LIBS analysis

LIBS is a method of optical emission spectrometry; but, unlike spark optical emission spectrometry (OES), the emission is subsequent to the generation of a plasma induced by a laser. In LIBS, a laser pulse strikes the surface of the sample and ablates an amount of material in the range of 1 nanogram (ng) and generates a plasma plume (partially ionized gas) in the temperature range of 5,000 to 20,000 Kelvin (K). The energy of the laser is low, but it is focused to a microscopic point on the sample to generate the plasma.

In this plasma, the matter constituting the samples is dissociated into atoms (atomization) and partially ionized. Those atoms and ions will be excited (transition of electrons from lower to higher energy levels of valence shell), and, by returning into their ground state (transition from higher to lower level of valence shell), they will emit characteristic lines for each element. The emitted light is transmitted through optical fibers, and the polychromatic radiation is dispersed in one or more spectrometers by diffraction gratings and detected by CCD (charge-coupled device) chips.

The spectra of LIBS can contain hundreds or even thousands of lines for a single element. The sensitivity of those lines can differ by several orders of magnitude and result in extremely line rich spectra, especially when the sample contains high concentrations of transition metals, as is the case for alloys like stainless steel. In typical hand-held LIBS systems, the dispersion power of the spectrometer is often limited by its size, and some important analytical lines may not be fully resolved from lines emitted by the matrix.

To cover the entire spectral range between 180 and 800 nanometers (nm), multiple spectrometers may be required. Moreover, wavelengths of less than 200 nm (like carbon, 193.09 nm, or sulfur, 180.73 nm) are strongly absorbed by air and require an argon purge of the optical path to be detected.

Almost any element generally contained in metals can be detected with LIBS: The sensitivity for alkaline (lithium, sodium, etc.) and alkaline-earth metals (beryllium, magnesium, etc.) is very high, and the sensitivity for transition metals is good, except for refractory elements (like niobium, molybdenum, tungsten or tantalum) which are difficult to determine. The sensitivity for carbon, phosphorous and sulfur generally is not sufficient to analyze those elements at relevant levels in alloys.
 

On the spot

Typical spot diameter of XRF is 3 millimeters (mm) to 8 mm, whereas the crater generated by the laser in LIBS has a diameter of 50 microns (µm) to 100 µm, typically. Only a 15-µm-to-20-µm diameter fraction of this crater actually will be analyzed. Hence, LIBS may be more sensitive to local heterogeneities. The laser pulse can be moved during the analysis to correct for effects caused by heterogeneities.

On the other hand, much smaller spots and very narrow weld seams can be analyzed using LIBS.

The optical emission induced by the laser is a transient phenomenon, whereas the X-ray beam is constant and well-controlled. Hence, it is expected that XRF delivers more stable, repeatable and reproducible results than LIBS. The quantitative analysis is considered the Achilles’ heel of LIBS—first because of the complex laser-sample interaction process, which depends upon both laser characteristics and material properties, and, second, because of the plasma-particle interaction process, which is time and space dependent.
 

Strengths and weaknesses

When compared with XRF, LIBS offers new possibilities in terms of applications. Lithium can be detected in aluminum alloys used in aerospace, beryllium can be detected in beryllium bronze, and carbon can be detected in carbon steel and cast iron. Moreover, the sensitivity to magnesium and aluminum is much higher in LIBS than in XRF, so sorting of aluminum and titanium alloys can be significantly faster (a few seconds versus 30 seconds to 60 seconds with XRF).

In contrast, LIBS may not be able to detect low concentrations of sulfur and phosphorous, which can be quantified by XRF in stainless steel (SS 303, SS 416) and in phosphorous bronzes.

Generally, for scrap metal sorting, LIBS is expected to be faster for aluminum alloys and more or less equivalent to XRF for stainless steel. LIBS may be able to sort most of the titanium alloys faster than XRF, with few exceptions like titanium Grade 11, containing around 0.15 percent palladium, which can be identified using XRF. Sorting of stainless steel can be done within few seconds using both XRF and LIBS.

For heavier alloys, such as super alloys, copper alloys (except aluminum and beryllium bronze), solders, lead alloys or precious metal alloys, XRF delivers better sensitivity and accuracy for analysis than LIBS. In addition, the measurement of tramp elements in scrap may be difficult for LIBS. For example, the detection of lead and tin in stainless steel alloys at 100 to 500 parts per million (ppm) will be difficult using LIBS.

For fast sorting of aluminum alloys based upon light elements, as well as for separating many magnesium and titanium grades, LIBS could be the best alternative.

When precision and accuracy matter more, as in PMI analysis and quality control, or when price figures are required in scrap trading (nickel, molybdenum, etc.), then HH-XRF could still be the method of choice.

From a regulatory point, the use of XRF requires paperwork, licensing and in some countries, lengthy radiation safety training. In contrast, the use of lasers (class 1 or 3b) does not require any of these. Thus, if your main application is measuring magnesium, aluminum or titanium alloys or red metals containing significant amounts of beryllium, aluminum or silicon, the instrument of choice will be LIBS.

If your application is measuring stainless steel, high-temp alloys or other heavy-metal alloys, the instrument of choice will be XRF. In both cases, the instrument of choice can be used to measure most alloy classes, however, there are compromises in speed, accuracy and precision that must be understood. Make sure that you evaluate the various choices on your unique set of samples and select the tool that meets your needs the best.

John I.H. Patterson is the principal of Portable Analytical Technologies LLC, based in Trenton, New Jersey.