22nd International Symposium on Plasma Chemistry
July 5-10, 2015; Antwerp, Belgium
Conditions for consistent spectral intensity ratio during elemental analysis by
laser-induced breakdown spectroscopy
J.-H. In, J.-H. Choi and S. Jeong
School of Mechatronics, Gwangju Institute of Science and Technology, 1 Oryong-dong Buk-gu, KR-500-712 Gwangju,
South Korea
Abstract: This work reports that the consistency of intensity ratio under varying laser
energy conditions could be achieved for elements with different ionization energy or
spectral lines with different upper levels during elemental analysis by laser-induced
breakdown spectroscopy if ungated detection approach is adopted with spectral lines having
similar self-absorption characteristics. It is demonstrated that the intensity ratio profiles at
different laser energy conditions become nearly the same under normalized time scale.
Keywords: laser-induced breakdown spectroscopy, self -absorption, ungated detection
1. Introduction
The spectral intensity of a plasma produced by
nanosecond laser pulse varies substantially by the change
of laser irradiation parameters such as pulse energy and
spot diameter or the elapsed time from the laser
irradiation to signal collection. However, the ratio of
spectral line intensities varies much weakly with respect
to measurement conditions because intensity ratio
depends primarily on the elemental concentration ratio of
a sample. Thus, the ratio of spectral intensities of
constituent elements instead of absolute intensity is
widely adopted in laser-induced breakdown spectroscopy
(LIBS), an elemental analysis technique to predict the
chemical composition of a target by measuring the
emission spectra of laser-induced micro plasma. By using
a proper reference material, the intensity ratio of LIBS
spectra can be calibrated by the concentration of
constituent elements. To predict the elemental
composition of an unknown sample using LIBS signal
intensity ratio, however, the measurement need to be
made under the same conditions as those for calibration
because intensity ratio itself varies with laser energy and
elapsed time. To ensure the precision of LIBS analysis,
therefore, it is desired to maintain the consistency of LIBS
signal intensity ratio under varying irradiation and
detection conditions. Previously, we reported that LIBS
signal intensity ratio of spectral lines of different elements
becomes nearly independent of plasma conditions
provided that the upper level energies of the spectral lines
and the ionization energies of the two elements are similar
[1].
In this work, it is reported that the consistency of LIBS
Wavelength
range
I
II
P-I-2-79
signal intensity ratio can be achieved for elements with
different ionization energy or spectral lines with
significantly different upper levels. It is demonstrated that
the intensity ratio for properly selected spectral lines
remained nearly invariant under varying plasma
conditions and the normalized intensity ratio profiles
became almost the same.
2. Experiments
For sample, Cu(In,Ga)Se 2 (CIGS) thin film from a
commercial CIGS solar cell module was utilized. The
nominal composition of a CIGS thin film is 25 Cu : 25
(In+Ga) : 50 Se in atomic percent (at.%) but the actual
composition and elemental profile vary with
manufacturer. For experiments, a small size CIGS sample
(5 cm × 5 cm) was cut out of the module and the
transparent conducting oxide and buffer layers on top of
the CIGS layer were removed by etching in a dilute
hydrochloric acid solution in order to directly irradiate the
CIGS thin film for LIBS analysis.
Table 1 shows the spectral lines investigated in this
study. These lines were selected with the consideration of
their spectral characteristics. First, the two spectral lines
in wavelength range I are those with strong selfabsorption. On the other hand, the two spectral lines in
wavelength range II are weakly self-absorbing lines.
For LIBS measurement, a Q-switched Nd:YAG laser
(λ=532 nm, τ=5 ns, top-hat profile) was used to irradiate
the sample. The laser fluence was varied over a wide
range of 8.6~67 J/cm2 so that the influence of laser energy
change on LIBS signal intensity ratio could be clearly
observed. The spot diameter was set to 50 µm. The
Table 1. Spectral properties of the emission lines investigated in this study [2, 3]
Element
Cu I
In I
Cu I
Cu I
Wavelength
(nm)
324.754
325.608
510.554
515.324
E i (eV)
E k (eV)
A ki
gi
gk
A ki g k
0.000
0.274
1.389
3.786
3.817
4.081
3.817
6.191
1.39E8
1.30E8
2.00E6
6.00E7
2
4
6
2
4
6
4
4
5.56E8
7.80E8
8.00E6
2.40E8
First ionization
energy (eV)
7.73
5.76
7.73
7.73
1
emission spectra were measured using an ICCD detector
(PI-MAX3, Princeton Instruments) at either ungated or
3. Results and discussions
Fig. 1 shows the intensity ratios measured at ungated
mode. Note that the spectral lines of Fig. 1(a) are both
strong self-absorption lines having similar upper level
energies but different ionization energies (different
species, Cu & In), whereas those of Fig. 1(b) are both
weak self-absorption lines having different upper level
energies but same ionization energies (same species, Cu).
The intensity ratios in Fig. 1(a) and 1(b) remained
consistent for varying laser fluence. On the other hand,
the intensity ratio of spectral lines in Fig. 1(c) having
same upper level energies and same ionization energies
changed significantly with laser fluence. These results
imply that the similarity in self-absorption characteristics
is an important criterion to determine the consistency of
intensity ratio during LIBS analysis at ungated mode.
Fig. 2(a) shows the temporal evolution of plasma
temperature determined from time-gated measurement
data at various laser fluence conditions. The x-axis
represents gate delay. The plasma temperature was
estimated using the measured intensities of spectral lines
in wavelength range II by the following equation [4]
T = (E 2 − E1 ) [k B ln ((I1 g 2 A2 ) (I 2 g1 A1 ))]
(1)
where I 1 and I 2 are spectral line intensities, g 1 and g 2 are
statistical weights, A 1 and A 2 are transition probabilities,
E 1 and E 2 are upper level energies, k B is Boltzmann
constant (1.381×10-23 J/K), T is plasma temperature, and
the subscripts 1 and 2 represent the wavelengths of
510.554 and 515.324 nm, respectively.
Depending on laser fluence, plasma life time changes as
shown in Fig. 2(a). However, when the delay time in xaxis in Fig. 2(a) is normalized by the duration until the
same plasma temperature (8500 K) is reached, the
temperature profiles of plasmas produced at different
laser energy conditions fall nearly onto a single profile as
shown in Fig. 2(b). It was also confirmed that the
intensity ratios of Fig. 1(b) measured at different laser
energy conditions also overlap onto a single profile when
the x-axis of intensity profile was normalized by the same
way. The consistent intensity ratio shown in Figs. 1(a)
and 1(b) at ungated detection is understood due to the
similarity in plasma temperature profiles and intensity
ratio profiles of the spectral lines with similar selfabsorption characteristics.
Fig. 1 Intensity ratios of spectral lines with (a) strong, (b)
weak and (c) different self-absorption characteristics
measured at ungated mode
time-gated mode. For ungated measurement, the gate
width of ICCD was set to a sufficiently long duration of
10 µs over which plasma emission vanished almost
completely. For time-gated measurement, the gate delay
was varied from 150 to 2000 ns and the gate width was
set to 100 ns.
2
4. Conclusion
This work demonstrated that the intensity ratio during
LIBS analysis can be measured consistently even under
fluctuating laser energy condition by measuring the
plasma emission at ungated detection mode using spectral
lines with similar self-absorption characteristics. These
results imply that the reliability and precision of LIBS
analysis can be significantly improved by selecting
spectral lines on the basis of self-absorption
characteristics under fluctuating measurement conditions.
P-I-2-79
Fig. 2 Temporal change of plasma temperature at different laser pulse energies (a) before and (b) after normalization by
the duration until the same plasma temperature (8500 K) is reached.
5. Acknowledgements
This work was supported by National Research
Foundation of Korea (NRF) grant funded by Korea
government (MEST) (No. 2014049289)
6. References
[1] Chan-Kyu Kim, Jung-Hwan In, Seok-Hee Lee,
Sungho Jeong, Opt. Lett., 38, 3032 (2013).
[2]
National
Institute
of
Standards
and
Technology, http://physics.nist.gov/PhysRefData/ASD/lin
es_form.html , (accessed March 2014).
[3]
National
Institute
of
Standards
and
Technology, http://physics.nist.gov/PhysRefData/ASD/le
vels_form.html , (accessed March 2014).
[4] Andrzej W. Miziolek, Vincenzo Palleschi, Israel
Schechter, Laser-Induced Breakdown Spectroscopy
(LIBS): Fundamentals and Applications, Cambridge
University Press, New York, 2006, pp. 122-170.
P-I-2-79
3