Quantitative Determination of Boron in Ulexite with LIBS II
Goal:
Assess the sensitivity of LIBS to quantitatively determine the boron content in Ulexite pellets
Method:
LIBS
Introduction:
Following the promising results obtained for the ulexite/clay pellets, additional pellets prepared from XPI-263 grab samples were prepared with a narrow boron concentration range. The pellets were prepared to assess the sensitivity of LIBS to boron concentration.
Experimental Conditions:
Ulexite pellets were prepared with closely spaced boron concentrations. The pellets were prepared with a new grinding technique. Unlike the previous set of pellets, each of the new pellets was received intact. Each pellet was placed in the LIBS sampling chamber. Ten single shot spectra were acquired at different locations on each pellet.
Data analysis involved calculating the ratio of the boron (249.5422 nm) and silicon (251.40066 nm) lines for each single shot spectrum and then calculating average ratio and standard deviation values for each sample.
Hardware Used:
Single channel LIBS spectrometer (HR2000CG-UV-NIR with LIBS upgrade serial # HR2B338)
BIF600-UV/VIS with LIBS focusing lens at the tip
Big Sky 50 mJ Nd:YAG Laser
LIBS SC Sampling Chamber with 125 mm focal length lens
Radiometrically calibrated spectrometer
Experimental Parameters:
Q-Switch Delay (delay between laser firing and data acquisition): -1.5 microseconds
Single laser shot at a fresh spot on the pellet (10 single shot spectra averaged together for standard curve calculations)
Measurement Mode:
LIBS
Results:
The complete data set acquired previously was reassessed to determine the optimal measurement parameters. Based on the standard curve generated and the standard deviation of the values, the best results were obtained with a radiometrically calibrated spectrometer for results averaged from 10 single shot spectra. Due to changes in the LIBS system configuration (probe was removed and replaced) since the last set of pellet measurements, a new standard curve was generated using the previous set of pellet samples. The new standard curve is shown in Figure 1 below. It is similar to the standard curve reported previously. As shown in Figure 1, when the ratio of these lines was taken at each ulexite concentration, the intensity trend increased with increasing boron concentration.
In Figure 2, the ratio of boron and silicon lines calculated for the new set of XPI-263 pellets is shown. Note that these pellets cover a much narrower range of concentrations (10.15 to 11.70% boron). As anticipated, based on the error associated with the wet analysis technique, the trend is noisy especially in regions where the boron concentrations were very closely spaced (e.g. 10.29 to 10.36% boron). Overall, an increasing trend in intensity is observed with increasing boron concentration. The standard deviations of the values are higher than observed for the previous set of pellets.
Conclusions:
It is difficult to determine the sensitivity of the LIBS technique from this data based on the error associated with the wet assay technique. Sensitivity at the 0.5% boron concentration may be possible based the results shown above.
Figure 1: Ratio of the average intensities (n=10) for boron and silicon lines as a function of boron concentration: Original set of ulexite pellets measured previously
Figure 2: Ratio of the average intensities (n=10) for boron and silicon lines as a function of boron concentration: New set of XPI-263 pellets
