Ocean Optics LIBS2000+ - Manual (Page 12)

1: Introduction
LIBS Emission ID
The reason we at Ocean Optics have designed a broadband high-resolution spectrometer is to be able to
both see and resolve all the lines from all the elements. Our resolution is ~0.1 nm, less than this in the
UV-blue and a bit higher in the red-IR. Because the LIBS-generated lines are stark broadened to about 0.2
nm or so in the UV and about 0.3 nm in the IR, the system can resolve everything that is possible to
observe.
The laser-induced plasma begins life as very hot 15,000 K plasma emitting a large bremsstrahlung
continuum. Depending on the sample matrix, most emission analysis is performed for a few microseconds
for this to decay so as to not mask the line structure. During this time, the higher order transitions decay
away, leaving mostly I and II atomic emissions. These are the ones that are idenified.
Also entering into the emissions from the elements in the interrogated plasma are the collision dynamics
of the plasma. Rate equations can be used to analyze the target, but emission signatures, sample-standards
and correlation techniques are preferred to identify materials with the OOILIBS Software.
Elemental analysis and identification are very important in understanding content, and in some cases are
used for quantitative analysis. To perform analysis and identification, you must use an element catalog to
help determine when a particular element is present.
Currently we use a catalog that was derived from the MIT wavelength tables. It consists of the persistent
lines of the elements from spark spectra and is the closest to a complete catalog of the brightest lines
around the temperature of the decayed plasma. As part of an international standards committee project,
Ocean Optics is in the process of compiling a persistent line emission set for LIBS, using the LIBS
plasma. This will take some time and will be integrated into our base table as time goes on. These studies
are to be done in an Argon gas environment.
How can we be sure that the identified elements are actually present? OOILIBS software contains two
rating mechanisms:
• The first rating mechanism indicates how many lines of those in our persistent line set are
present. Thus, if 5 of 7 possible have been identified, you can be fairly certain that that element is
present. If one of 7 is present, could it be something else? Perhaps a line that is not in our catalog
from another species?
• For the second analysis technique, the emission lines have an appearance value based on
experimental observation. It could be that this single line is the brightest of those from this
element and the others are weak. So, we rate the element using a formula based on these
appearance levels. If only the weakest line of the 7 appears, it is ranked very low and indicates
that this is from one of the unknowns. If it is the strongest line, it will be rated high, allowing you
to be certain that it was present. If you have the highest 5 of the 7 in the table, the rating will be
extremely high.
OOILIBS software allows you to call any number of spectral libraries you wish to use. Our present library
has greater than 2000 lines available. There are no overlaps, but as we advance to more sophisticated
libraries, there certainly will be. The full NIST catalog has so many lines (>100,000) that everything
would be identified many times over, even though the probability of appearance may be small to zero.
That is why we have not included it.
2 161-00000-000-02-0406