Visualisation of the matrix-effect
Friday, March 20, 2020

Science
Projects
A brief excerpt of a LIBS-homogeneity test performed on our pellets used to illustrate the importance of matrix-matched reference materials. Using laser-scanning confocal microscopy we were able to highlight the differences regarding how different materials react to a laser-beam.

Within the framework of a cooperation with Applied Photonics Ltd., UK (APL) a total of 13 Nano Pellets were measured using a benchtop modular LIBS system comprised of a motorised XYZ stage, a Quantel Q-smart 450 laser emitting at 1064 nm (pulse width 6 ns) and a SpectroModule 6 multi channel spectrometer. In order to visualise the matrix-effect two samples of contrasting matrices were selected. OREAS 24-NP, a granodiorite and IF-G-NP an iron formation sample. The crater characteristics were investigated using a Keyence VK-250 laser-scanning confocal microscope at the Nano-Laboratory at the technical faculty of Kiel University, Germany. All samples were pressed using a programmable hydraulic press, to ensure the Nano Pellets were produced using the same conditions. An overview of the craters is shown in Fig. 1.

Fig.1: Overview of images taken with the laser-scanning confocal microscope. Images A & B show an overlay of optical and laser image for OREAS 24b-NP (A) and IF-G-NP (B) respectively. Images C & D show a false-colour map representing the surface morphology. Warm colours represent high areas whereas cold colours represent greater depth. The scale bar indicates 500 µm and is applicable for all images.
Fig.1: Overview of images taken with the laser-scanning confocal microscope. Images A & B show an overlay of optical and laser image for OREAS 24b-NP (A) and IF-G-NP (B) respectively. Images C & D show a false-colour map representing the surface morphology. Warm colours represent high areas whereas cold colours represent greater depth. The scale bar indicates 500 µm and is applicable for all images.

In Fig. 1 the difference between the craters’ appearances becomes evident. The crater in OREAS 24b-NP is more regularly shaped than the crater in IF-G-NP.

The same is true for the affected area surrounding the crater. The affected area in OREAS 24b-NP, despite exhibiting cracks and minute differences in height, is fairly level, whereas the affected area surrounding the crater in IF-G-NP is more chaotic and shows heightened domains within a generally shallower area. A tilted view of the craters is shown in Fig. 2.

Fig.2: False-colour images of OREAS 24b-NP (A) and IF-G-NP (B) at a tilted angle. The depth in OREAS 24b-NP is greater than the crater in IF-G-NP.
Fig.2: False-colour images of OREAS 24b-NP (A) and IF-G-NP (B) at a tilted angle. The depth in OREAS 24b-NP is greater than the crater in IF-G-NP.

Two profiles averaged over 100 lines (width 1 µm with a step of 1,5 µm between each line) across the crater are shown in Fig. 3. Evidently, the crater in OREAS 24b-NP (79 µm) is deeper than the crater in IF-G-NP (35 µm), while the diameter for both is roughly the same at ~250 µm. The profile also shows the rough morphology of the affected area of IF-G-NP in contrast to the smooth affected area of OREAS 24b.

Fig.3: Profiles across the craters and surface of OREAS 24b-NP (top) and IF-G-NP (bottom) averaged using 100 lines (line-width of 1 µm, step of 1,5 µm between each line). Total width of the transect is approx. 2,5 mm for each sample
Fig.3: Profiles across the craters and surface of OREAS 24b-NP (top) and IF-G-NP (bottom) averaged using 100 lines (line-width of 1 µm, step of 1,5 µm between each line). Total width of the transect is approx. 2,5 mm for each sample

Interestingly, the crater of IF-G-NP is more cylindrical than in OREAS 24b-NP. This could possibly be due to the existence of a cut-off-depth with respect to laser focus, where the crater-depths affects the ablation due to lack of focus. Apparently, this depth lies between 35 and 79 µm, for the APL-system and laser-settings. Recently, automated stages countering the ablation rate and thereby re-focussing the laser have been developed. The Keyence Multifile-Analyser software enables the measurement of volumes. The results of this measurement are shown in Fig.4 as well as Tab.1.

Fig.4: Volume measurement of OREAS 24b-NP as well as IF-G-NP. The orange horizontal lines indicate the current transect (1µm width) shown at the bottom of each respective image. Therefore, they are not representative of the entire crater morphology. The general crater morphology is characterised in Fig.2. The purple highlighted areas show, which part was considered for volume calculation.
Fig.4: Volume measurement of OREAS 24b-NP as well as IF-G-NP. The orange horizontal lines indicate the current transect (1µm width) shown at the bottom of each respective image. Therefore, they are not representative of the entire crater morphology. The general crater morphology is characterised in Fig.2. The purple highlighted areas show, which part was considered for volume calculation.

Summarising the acquired data (Tab.1) and correlating it with each sample’s density allows the calculation of an ablation rate and the ablated mass, for each pulse as well as in total. This assumes, that each pulse ablates exactly the same amount, which may very well not be true. However, for the purpose of demonstrating, that there is a noticeable difference in general and not in detail this approach was deemed sufficiently accurate.

Tab.1: Comparison of crater characteristics as well as ablation yield between OREAS 24b-NP and IF-G-NP during ablation using the same laser settings: 100 pulses, 1064 nm, 100 mJ and ~300 µm spot-size.
Tab.1: Comparison of crater characteristics as well as ablation yield between OREAS 24b-NP and IF-G-NP during ablation using the same laser settings: 100 pulses, 1064 nm, 100 mJ and ~300 µm spot-size.

Conclusions

There is an obvious and expected difference in crater morphology between OREAS 24b-NP and IF-G-NP. The differences can be attributed to variations in laser/matter interaction, due to significantly different sample mineralogy and thus composition. These differences can be minimised by using lasers emitting close to or within the UV-spectrum (i.e. 213, 193 nm) and decreasing the pulse width from the nanosecond to femtosecond range. Further, it shows, that the nano-pellets, despite being binder-free, remain stable during ablation and are an important step toward improved matrix-matched quantification.

Acknowledgement & Disclaimer

We would like to thank the entire APL-Team for their willingness to cooperate and the effort that went into data evaluation. OREAS 24b and IF-G are commercially available reference materials. We took the original powder, reduced the particle-size down to the nanometer range and pressed pellets from the resulting powder. The respective manufacturers of each sample are not in any way liable for the pellets.