The Atomizer is an accessory, based on the LINA-Spark technique that prepares “on line” an aerosol from any solid sample for quantitative bulk analysis by ICP-OE or ICP-MS. The LINA-Spark Atomizer eliminates the need to put solid samples in solution, resulting in:
simplified sample preparation
no contamination
no loss of elements
no expensive chemicals
considerable gain of time
How it works: The parallel beam of a pulsed Nd-YAG Laser is concentrated with a lens in such a way that at the surface of the sample a “LINA-Spark”, that is a short-lived plasma, is created.
During its lifetime of about one microsecond, the plasma evaporates some material from the surface. These vapors condense immediately and agglomerate in small clusters with a dimension of about 5 – 10 nm. They are carried away in the argon stream and brought into the ICP torch.
Because of their fineness only few particles are lost in transport and they are efficiently evaporated, ionized and excited in the ICP.
The flow rate of the aerosol carrier gas can be chosen to suit best the ICP. For ICP-OES this is usually around 0.5 l/min. whereas for ICP-MS it is around 1.1 l/min.
The aerosol is basically dry. This is favorable for reducing O/H interferences in the ICP-MS. The aerosol can be artificially humidified for certain applications to improve the ionization efficiency in the ICP.
Each single LINA-Spark creates a very shallow crater with a diameter of about 1.2 mm. If a bigger surface needs to be examined for analysis the lens is moved after each pulse to different programmed X-Y coordinates. It is possible to examine a surface of up to 15 mm2 in one analysis.
What can be analyzed? The LINA-Spark Atomizer can produce an aerosol from any solid material, such as
Metals: Pure and precious metals, high temperature alloys
Ceramics: High-tech ceramics, refractories
Glass: Quartz-glass, fusion pearls
Slags: Blast furnace slags, steelmaking slags
Minerals: Geological mapping, sediments
Plastics: PVC, PE
Semiconductors: Silicon wafers, Germanium
Powders must be pressed solidly, if necessary with a binder, so that they do not disintegrate under the hard attack of the plasma explosion.
Brittle minerals and heterogeneous material must be ground and pressed into a solid pellet.
The sample can be small and thin, because the LINA-Spark adds very little heat to the sample.
Each plasma pulse evaporates material from a circular surface of 1.2 mm diameter. By moving the plasma during an.alysis, the analyzed surface can be increased to about 15 mm2. The removal rate in depth depends on the material, the sampled surface and the repetition rate of the Laser pulses: for steel it is about 4 µm/min for a 2 mm diameter spot (3 mm2 and 10 Hz repetition rate.)
It is not necessary to clean the sample before a bulk analysis: during the “prespark” period the LINA-Spark removes the undesirable material from the surface.
If an analysis of surface elements is desired, the prespark period phase is eliminated and measurements are taken from the first sparks. Depth profiles can be done by measuring repeatedly during several minutes.
Two different techniques are used to examine the aerosol. They can be characterized as follows:
ICP-OES: - Limits of detection in the solid sample are typically 1 ppm, but depend on the sensitivity of the spectral lines. - Precision of measurement is typically 2% rel. for medium and high concentrations. Precision and accuracy are improved by use of an “Internal Standard”. - Spectrometers with simultaneous measurement (multichannel- or “chip”-instruments) are preferred. “Internal Standard” and other elements at high concentrations must be measured with weak (and not sensitive!) spectral lines, which are often not accessible. A full wavelength coverage, as e.g. provided by the VARIAN VISTA is then an asset. For more information www.varianinc.com
ICP-MS: - Limits of detection in the solid sample are typically 10 ppb for most isotopes. - Precision typically better than 5% rel. for traces and low concentrations. - Measuring matrix element (and use it as Internal Standard) may be difficult. - Quadrupol and HR-MS are both suitable. In Germany, three LINA-Spark Atomizers are currently used in connection with the HP4500 ICP-MS with excellent results. www.agilent.com
Accurate bulk analysis Why the LINA-Spark overcomes the problem of fractionated evaporation, a problem which cannot be solved properly with an ordinary Laser Ablation.
If a heat pulse is applied to a very limited area of the surface (or surface layer) of the sample, this heat will preferentially evaporate those elements which are easily evaporable. The exposed zone will be impoverished of those elements. The border of the zone will be little affected and will stay in its original condition. The next pulse will find less of the EEE (easily evaporable elements) in the surface layer and therefore evaporates a mixture which is closer to the original composition. After a number of pulses, the composition of the evaporated elements corresponds exactly to the original composition of the material.
This is however only true if the surface of the exposed area is much bigger than the surface of the circular border zone! The diameter of this circular border zone will increase slightly with each pulse, and new EEE will be evaporated.
In other words, if the removal of material is strictly in one direction (depth), the stationary evaporation conditions will correspond to the composition of the sample. But with three-dimensional removal of material, stationary conditions are not achieved, since each pulse evaporates a little virgin material.
Experts in Laser Ablation say that the problem of EEE becomes manageable if the “aspect ratio of the crater comes down to about unity”, which is rarely the case in Laser Ablation. (Aspect ratio: depth of crater by diameter of crater).
The aspect ratio of the LINA-Spark “crater” is in the worst case 0.1, but is normally rather 0.001. The removal of material with the LINA-Spark is strictly unidirectional.
The great benefit of this unidirectional removal is that after a certain “prespark period” (during which the EEE are still dominant) the composition of the aerosol corresponds strictly to the composition of the sample. Only under this condition can an accurate analysis be obtained!
Surface Analysis: When performing bulk analysis, it is not necessary to clean the sample surface before submitting it to the LINA-Spark Atomizer: the first few LINA-Sparks remove all loose dirt on the surface and start to remove the surface layers. This period of time to remove contaminations and prepare the surface is called “prespark time”. For bulk analysis, the aerosol composition is not measured during this period.
If a measurement of the surface composition is desired, the prespark time is skipped and measurements of the aerosol composition are taken right from the beginning.
In order to assess the sensitivity of the method, we have connected the LINA-Spark Atomizer to a HP 4500 ICP-MS. We have contaminated the Silicon wafers in various degrees with platinum group elements, which are not naturally contaminating elements like Fe, Ca, Na etc.
With a total measuring time of 100 s we found for the 8 elements Ru, Rh, Pd, Re, Os, Ir, Pt and Au an average limit of detection of 109 atoms / cm2. The analyzed area was 15 mm2. With refinements this LOD can be lowered by another factor of 0.1!
This method is about as sensitive as the CVD method used in the wafer industry. However, the CVD measures only the average contamination over the whole surface, whereas with the LINA-Spark method and many spot analysis, it is possible to see how this contamination is distributed over the surface.