Temperature effect on femtosecond laser-

induced breakdown spectroscopy of glass

sample
Cite as: Phys. Plasmas 24, 013301 (2017); https://doi.org/10.1063/1.4973658
Submitted: 22 September 2016 • Accepted: 22 December 2016 • Published Online: 09 January 2017

Ying Wang, Anmin Chen, Yuanfei Jiang, et al.

ARTICLES YOU MAY BE INTERESTED IN

Effect of sample temperature on time-resolved laser-induced breakdown spectroscopy

AIP Advances 9, 065214 (2019); https://doi.org/10.1063/1.5097301

Enhanced optical emission in laser-induced breakdown spectroscopy by combining

femtosecond and nanosecond laser pulses
Physics of Plasmas 27, 023507 (2020); https://doi.org/10.1063/1.5131772

Influence of distance between focusing lens and target surface on laser-induced Cu plasma
temperature
Physics of Plasmas 25, 033302 (2018); https://doi.org/10.1063/1.5010076

Phys. Plasmas 24, 013301 (2017); https://doi.org/10.1063/1.4973658 24, 013301

© 2017 Author(s).
 PHYSICS OF PLASMAS 24, 013301 (2017)

Temperature effect on femtosecond laser-induced breakdown spectroscopy

of glass sample
Ying Wang,1,2,3 Anmin Chen,1,2,3,a) Yuanfei Jiang,1,3 Laizhi Sui,1,3 Xiaowei Wang,1,3
Dan Zhang,1,3 Dan Tian,1,3 Suyu Li,1,3,b) and Mingxing Jin1,3
1
Institute of Atomic and Molecular Physics, Jilin University, Changchun 130012, China
2
State Key Laboratory of Laser Propulsion and Application, Academy of Equipment, Beijing 101416, China
3
Jilin Provincial Key Laboratory of Applied Atomic and Molecular Spectroscopy (Jilin University),
Changchun 130012, China
(Received 22 September 2016; accepted 22 December 2016; published online 9 January 2017)
In this study, we observed the evolution of the spectral emission intensity of a glass sample with the
increase of sample temperature, laser energy, and delay time in femtosecond laser-induced break-
down spectroscopy (fs-LIBS). In the experiment, the sample was uniformly heated from 22  C to
200  C, the laser energy was changed from 0.3 mJ to 1.8 mJ, and the delay time was adjusted from
0.6 ls to 3.0 ls. The results indicated that increasing the sample temperature could enhance the emis-
sion intensity and reduce the limits of detection, which is attributed to the increase in the ablated
mass and the plasma temperature. And the spectral intensity increases with the increase of the laser
energy and the delay time, however, the spectral line intensity no longer increases when the laser
pulse energy and delay time reach a certain value. This study will lead to a further improvement in
the applications of fs-LIBS. Published by AIP Publishing. [http://dx.doi.org/10.1063/1.4973658]

I. INTRODUCTION deviation, to enhance the accuracy of quantitative analysis

for the element, etc. With the rapid development of laser
Laser-induced breakdown spectroscopy (LIBS) has
technology and optical detection technology, the develop-
become an established technique that is widely used for
ment of LIBS technology is also booming. Not only that, but
spectrochemical analysis; LIBS is a form of atomic emission
nanosecond LIBS,9 femtosecond LIBS,10 polarization
spectroscopy used to qualitatively and quantitatively detect
resolved LIBS,11,12 dual–pulse LIBS,13,14 time resolved
the elemental composition in any physical state (solid, liquid,
LIBS15,16 and spatial resolved LIBS17,18 have appeared in
gaseous, or aerosol). LIBS is a variety of technology that can
secession; wherein, nanosecond LIBS has been widely
ablate the material by using a high power pulsed laser, which
applied. From the perspective of the application of laser
makes a minute amount of sample instantaneously vaporize,
spectrum detection technology, the nanosecond laser has
resulting in a high temperature and high density laser-
numerous advantages, such as lower cost, more convenient
induced plasma (LIP), and emitting a plasma spectrum with
operation and more simple maintenance, and so on. Making
the distinguishing wavelengths of elements found in the sam-
use of a nanosecond laser as a plasma excitation source of
ple. As is well known, the spectral wavelength and intensity
the LIBS technique not only has many relevant studies, but
shows the composition and content of the elements, respec-
also it has good application prospect. For example, using
tively. In comparison with the other optical methods (such as
various experimental configurations, the spectral and tempo-
laser spectroscopy technology, laser Raman spectroscopy
technology, laser absorption spectroscopy technology, and ral polarization dependencies of nanosecond laser-induced
so on), the experimental method of LIBS is very simple, and plasmas are explored for the analysis of gaseous and solid
it can simultaneously overcome the difficulty of the multi- samples.19,20 Even so, the nanosecond LIBS may cause a
element analysis. Furthermore, it is characterized by a fast, serious fracture in the chemical bonds of material composi-
real-time, in situ, low invasiveness, and multi-elemental tion, and only the LIBS spectra of atoms or ions in the mate-
diagnosis, normally without the need for sample prepara- rial components can be detected. On the contrary, the
tion.1,2 This technology (LIBS) is used extensively in various femtosecond laser has an ultrashort pulse width, it creates
areas such as remote detection, hostile environment monitor- low plasma temperature, and has a smaller intensity of con-
ing, cancerous breast tissues, and cultural heritage conserva- tinuous radiation; not only that, but it also simultaneously
tion as a useful analytical tool.3–8 attenuates, and the excitation threshold decreases with
For the aim of improving the accurate reliability and increasing bandwidth. Moreover, the femtosecond laser has
economic practicality of LIBS, the central issues of the cur- characteristics of low threshold, high stability, and high
rent research are focused on how to enhance the spectrum repeatability. When the laser energy exceeds a specific
signal intensity, to reduce the intensity of continuous spec- threshold, the plasma will be formed. Compared with the
trum, to decrease the basic effect, to reduce the standard nanosecond-laser pulse, which demands several joules per
square centimeter, the femtosecond laser pulse only needs a
a)
E-mail: [email protected] fraction of the joule per square centimeter. Nonetheless, the
b)
E-mail: [email protected] shortcoming of femtosecond LIBS is the much lower

1070-664X/2017/24(1)/013301/8/$30.00 24, 013301-1 Published by AIP Publishing.

013301-2 Wang et al. Phys. Plasmas 24, 013301 (2017)

spectral intensity based on the reduced interaction between

ablation pulse and plasma plume.
According to previously published reports, the initial
sample temperature can be known as an effective factor on
LIBS parameters. Many researchers have been aware of that
the temperature effects are important for LIBS. Tavassoli
and Gragossian reported that increasing the sample tempera-
ture could enhance the spectral emission and reduce the lim-
its of detection for aluminum alloy surface at ambient air
pressure.21 Su et al. showed that heating glass samples could
improve signal-to-noise.22 A study of the potential relevance
of the sample temperature on the emission spectra was car-
ried out for stainless steel samples.23 Eschlbock-Fuchs et al.
investigated the influence of sample temperature on the
dynamics and optical emission of laser-induced plasma FIG. 1. Schematic diagram of the experimental setup. Components include
(LIP) for various solid materials.24 The effect of sample tem- half-wave plate (HWP), Glan laser polarizer (G), and lens (L).
perature on the emission line intensification mechanisms in
orthogonal dual-pulse LIBS was studied by Sangines et al.25 pulsed laser was focused onto the sample surface, and was
Hanson et al. studied the effect of sample temperature and changed to the desired value by means of a combination of a
physical state on the LIBS of molten and solid salts.26 Glan laser polarizer and a half-wave plate (HWP). The
Shoursheini et al. investigated the enhancement of Cu emis- plasma will be produced after the interaction between the
sion lines of a micro-plasma induced by a Nd:YAG laser due laser and sample, and emission spectra is focused by a fiber
to the thermal effect of simultaneous irradiation by a contin- using two lenses (BK7), and was transmitted to an intensified
uous wave (CW) CO2 laser.27 charge coupled device (ICCD) camera (PI-MAX, from
However, to our knowledge, few studies concerning the Princeton Instruments) with 1024  256 pixels through the
effect of sample temperature on femtosecond LIBS have spectrometer (Spectra Pro 500, PI Acton, the grating is 150
been reported to date. Recently our group undertook a study grooves/mm). About the used spectrometer, the spectral
of the glass sample to observe the spectrum of the elements range is 200 nm–900 nm, the slit size is 10 lm, the number
of glass, which is ablated by the femtosecond laser pulse at of greater lines is 150 lines/mm, the number of pixels is
atmospheric pressure; the initial temperature variation of the 1024, as well as the spectral width is 0.5 nm. The ICCD was
sample is discussed. The sample is uniformly heated to triggered via a pulse generator (DG535, from Stanford) that
200  C with intervals of 50  C, the purpose is to investigate can set the size of gate width and synchronize the delay time
the impact on the plasma emission and spectral lines at dif- between the laser pulse and the gate of ICCD, it made the
ferent sample temperatures. Furthermore, plasma is produced experiment avoid the effect of continuous background radia-
by various laser energies. Our results show that the sample tion on the effective signal of spectra and improves the reli-
pre-heated configuration is found to be important for obtain- ability of the measurement. The optical signal was recorded,
ing the optimal femtosecond laser-induced breakdown spec- stored, and analyzed by a computer. To obtain the effect of
troscopy. This study will lead to a further improvement in the sample temperature for spectra, the sample was heated
the applications of the femtosecond LIBS. up to 200  C. When the sample needed to be heated, we
placed the sample on the heater surface, and the heating
II. EXPERIMENTAL SETUP
device is a heater with temperature detection and tempera-
Figure 1 shows the specific experimental setup for mea- ture control, so it can keep the temperature stable. Its accu-
suring and analyzing the femtosecond laser-induced plasma racy is 1  C. Spectral lines were measured at 22  C (the
spectroscopy. A soda-lime float glass (SRM 1830) was cho- ambient temperature), 50  C, 100  C, 150  C and higher
sen as the test sample. The main concentration is SiO2 and
Na2O, wherein, the concentration of SiO2 in percentage by TABLE I. Composition of the glass sample.
weight for SRM 1830 is 73.07%, and the concentration of
Composition Concentration (%)
Na2O in percentage by weight for SRM 1830 is 13.75%, it
also includes other composition (see Table I). The dimension Silicon dioxide (SiO2) 73.07
is 76 mm  26 mm  1.2 mm. The excitation source is Ti: Aluminum oxide (Al2O3) 0.12
Sapphire amplifier (Coherent Libra), which can produce a Iron oxide 0.032
laser beam that the full-width at the half maximum (FWHM) Ferric oxide (Fe2O3) 0.121
is 50 fs, the wavelength is 800 nm, and the repetition rate is Titanium oxide (TiO2) 0.011
Calcium oxide 8.56
1 kHz. The maximum energy of the femtosecond single pulse
Magnesium oxide 3.90
laser for irradiation on the sample is up to 4 mJ. To avoid
Potassium oxide (K2O) 0.04
irradiating to the same location on the sample by femtosec- Sodium oxide (Na2O) 13.75
ond laser pulse, the sample was fixed on an X-Y-Z stage Sulfur trioxide (SO3) 0.26
(Thorlabs, PT3/M-Z8) controlled by a computer. The output
013301-3 Wang et al. Phys. Plasmas 24, 013301 (2017)

temperature 200  C. In this process, the sample reached a sta-

ble temperature for five minutes. Each data point was a result
of an accumulation and calculation, the average of 10 laser
shots. The whole experiment was carried out at atmospheric
pressure.

III. RESULTS AND DISCUSSION

A soda-lime float glass (SRM 1830) is chosen as the test
sample to investigate the initial sample temperature on the
influence of femtosecond laser-induced breakdown spectros-
copy characteristics. As shown in Figure 2, the sample con-
tains multiple elements, making the spectrum very complex.
To avoid the appearance of the interference phenomenon, we
chose a spectral line without overlap for the qualitative anal-
ysis. To collect the spectral signal, we chose a single pulse
excitation, where the delay time of signal collection is con-
trolled by software. The digital delay generator can control
the pulse excitation plasma, and we adjusted the time inter- FIG. 3. Comparison of emission intensity with the same laser energy for five
val of signal to change the background radiation. different sample temperatures (22  C, 50  C, 100  C, 150  C, and 200  C).
Femtosecond laser energy is 0.1 mJ.
Figure 3 shows the spectral emission intensity of Na (I)
in femtosecond laser-induced glass plasma with the same
laser energy for five different sample temperatures (22  C, same laser energy (0.3 mJ) is displayed in Figure 4. Here, we
50  C, 100  C, 150  C, and 200  C). Femtosecond laser find the delay time has significant influence on spectral inten-
energy is 0.1 mJ. The selected spectral line is Na (I) at the sity; spectral intensity decreases with the increase of delay
wavelengths of 589.0 nm and 589.6 nm. Since Na (I) in time. This may be attributable to an optimal delay time mak-
laser-induced glass plasma has the strongest optical emission ing the best spectral intensity, and the intensity of the emis-
compared with other lines. As seen in Figure 3, the emission sion line is determined by comparing a given temperature of
intensity increases with the increase of sample temperature 200  C with the corresponding one at an ambient temperature
in the experimental processes of femtosecond LIBS; this (22  C). The gate width is 50 ns. Comparing these two
increased trend is very evident. At an ambient temperature
(22  C), the emission intensity is weak and the spectral line
of Na (I) is almost unobservable. However, when the sample
temperature is heated to 200  C, the apparent spectral line
appears. This confirms that the increasing sample tempera-
ture can reduce the limit of detection and improve the spec-
tral emission intensity. We will discuss the influence of
sample temperature on femtosecond LIBS in the following
results in order to understand the enhanced effect.
The evolution of Na (I) spectral intensity of the femto-
second laser-induced glass plasma with the delay time for
various initial sample temperatures (22  C, 200  C) at the

FIG. 2. Time-resolved spectroscopy of femtosecond laser-induced glass FIG. 4. Evolution of spectral intensity with the delay time for two sample
plasma at ambient temperature. Femtosecond laser energy is 1.8 mJ. The temperatures (22  C, 200  C) at the same laser energy (0.3 mJ). The gate
gate width is 50 ns. width is 50 ns.
013301-4 Wang et al. Phys. Plasmas 24, 013301 (2017)

waterfalls in Figure 4, the sample temperature has a positive very important compared with others’. Therefore, another rea-
effect on the spectral intensity. It is discernable that the spec- son for increasing the ablation mass is due to the reduced abla-
tral line emission increases with the increasing of sample tion threshold. A hot sample surface appears at the reduced
temperature at all delay times. This experiment demonstrates reflectivity and thermal diffusivity. On one hand, the reduced
that the initial sample temperature can enhance the emission reflectivity can increase the ablation volume and the ablation
intensity. According to the local thermodynamic equilib- mass. On the other hand, the reduced thermal diffusivity will
rium, we can get the intensity of an emission line between maintain the surrounding environment. That is to say, com-
two levels Ei and Ek.28 pared with the ablation mass obtained at an ambient tempera-
    ture, more ablation mass can be obtained at the same incident
Aki gk Ek laser energy when the sample has a higher temperature. The
Ik ¼ F exp N exp  ; (1)
kUðTp Þ kb Tp relationship between the mass removal of ablated material and
the surrounding environment has been described in detail in a
where Ik is the line intensity, Fexp is the experimental coeffi- previously published paper.30 Additionally, the measurement
cient with respect to the efficiency of the optical detection has also already been explained specifically by Sangines and
system, N is the ablated material, Aki is the transition proba- Sobral under this experiment condition.25,31 The results show
bility, gk is the statistical weight of the excited state, Ek is the that the effect of increasing sample temperature can cause the
energy of the excited state, kb is the Boltzmann constant, plasma emission to become much stronger. Such an approach
U(Tp) is the partition function, and Tp is the plasma tempera- can make the coupling between the sample and the laser pulse
ture in Eq. (1). An experiment was carried out by Tavassoli even better, resulting in a higher ablation mass, and enhancing
to determine what causes the spectral intensity enhance- the intensity of the emission line as the sample temperature
ment;21 it showed that an increase in the spectral line inten- increases.
sity of a pre-heated sample was a consequence of both more Another reason for increasing spectral intensity comes
ablation mass and higher plasma temperature. In order to from the plasma temperature. Because of the increase of the
ascertain the influence of these two factors to the spectral sample initial temperature (200  C), it has a higher energy at
emission intensity, Tavassoli conducted a series of experi- the beginning. Therefore, the internal particles have a higher
mental studies.29 It is illustrated that an increase of sample kinetic energy. It leads to particles inside the plasma into
temperature can lead to increasing the ablation mass and the severe activity, showing the enhancement of plasma temper-
plasma temperature, where these two factors eventually ature. Finally, the emitted plasma has a higher plasma tem-
result in the enhancement of the spectral line intensity. perature, and it leads to the enhancement in the interaction
Therefore, the initial sample temperature can impact the between laser and plasma. In conclusion, the contribution of
spectral emission intensity. These two factors should be con- the ablation mass is more significant than the plasma temper-
sidered: an increase in the ablation mass and an increase in ature to the enhancement of spectral emission intensity.
the plasma temperature. For the current experimental condition, the laser wave-
There are also numerous potential reasons to improve length and pulse width are fixed, but the output energy and
the ablated rate and increase the ablation mass, such as the delay time is adjustable. Changing the laser energy can alter
surrounding gas density or the thermal diffusion. The higher the excitation power density focused to the sample surface.
gas density leads to further plasma plume confinement, Figure 5 shows the comparison of spectral intensity of fem-
which will prevent the plasma from being ejected from the tosecond laser-induced glass plasma at different sample tem-
ablation crater, compressing the plasma plume into the cra- peratures (22  C, 100  C, and 200  C). The laser energies are
ter. Therefore, the low density of the surrounding environ- 0.3 mJ (a), 0.6 mJ (b), 0.9 mJ (c), 1.2 mJ (d), 1.5 mJ (e), and
ment is viewed as the main reason for the increase in the 1.8 mJ (f). As can be seen from Figs. 5(a)–5(d), it is evident
ablation rate, so the intensity of the spectral line also that, in the range of laser energy from 0.3 mJ to 1.2 mJ, the
increases. From the perspective of the increasing ablation spectral intensity increases with the increase of sample tem-
mass, when a high power laser pulse irradiates to the sample perature. When we continue to increase the laser energy to
surface, the sample is ablated and is ionized to produce 1.5 mJ, the increased trend of the emission intensity is less
plasma, which begins to cool down and releases the spectral obvious. The spectral intensity of Na (I) first increases and
lines with element characteristics after the laser pulse irradia- then levels off as the laser energy progressively increases.
tion. In the whole processes, the plasma expands, and the When the laser energy is 1.8 mJ, the intensity of Na (I)
plume size is continually enlarged in the low-density envi- reaches the maximum (the saturation value). Moreover, in
ronment owing to the fact that high temperature will lead to the above-mentioned condition, the enhanced maximum of
decrease in ambient gas density as we all know. Compared the spectral line is apparent at 200  C. This is attributable to
to the ambient sample temperature, the higher spectral inten- the increase of the laser pulse energy, which also leads to the
sity is obtained at a higher sample temperature. More abla- increase of the ablated mass since a larger target area is irra-
tion mass is involved in the interaction between the laser diated. The emission intensity no longer changing when the
pulse and the sample when the laser pulse encounters the rar- laser energy ranges from 0.9 mJ to 1.8 mJ is probably caused
efied medium near the sample surface. A greater amount of by the plasma shielding.
material is ejected when the initial sample temperature is The main physical mechanism is closely connected with
increased.30 The total ablation volume depends on many the delay time. Therefore, it is necessary to discuss the influ-
parameters, the local reflectivity of the sample surface is ence of delay time. Figure 6 displays the evolution of
013301-5 Wang et al. Phys. Plasmas 24, 013301 (2017)

FIG. 5. Comparison of spectral intensity for three different sample temperatures (22  C, 100  C, and 200  C). The laser energies are 0.3 mJ (a), 0.6 mJ (b), 0.9
mJ (c), 1.2 mJ (d), 1.5 mJ (e), and 1.8 mJ (f). The delay time is 1.2 ls. The gate width is 50 ns.

spectral intensity of femtosecond laser-induced glass plasma contrast curves in Fig. 6(a), it is noted that increasing the
with a delay time for different sample temperatures (22  C, sample temperature distinctly enhances the spectral inten-
100  C, and 200  C). The gate width is 50 ns. The laser ener- sity; this is approximately 1.5 fold enhancement when the
gies are 0.3 mJ (a), 0.6 mJ (b), 0.9 mJ (c), 1.2 mJ (d), 1.5 mJ sample temperature increases from 100  C to 200  C. This
(e), and 1.8 mJ (f). When the sample temperature is higher, result shows that heating the sample can enhance the spectral
the obtained spectral intensity is stronger. There are three line intensity. As mentioned above, a higher temperature can

FIG. 6. Evolution of spectral intensity with a delay time at different sample temperatures (22  C, 100  C, and 200  C). The gate width is 50 ns. The laser ener-
gies are 0.3 mJ (a), 0.6 mJ (b), 0.9 mJ (c), 1.2 mJ (d), 1.5 mJ (e), and 1.8 mJ (f).
013301-6 Wang et al. Phys. Plasmas 24, 013301 (2017)

reduce the ablation threshold; within the scope of the abla- to 2.0 ls, the sample temperature is varied from 22  C to
tion threshold, the reflectivity is also reduced with the 200  C. The color bar displays the intensity from low to
increase of sample temperature. There also exists another high. The gate width is 50 ns. The relevant laser energy is
phenomenon, it is evident that the spectral intensity gradu- altered from 0.3 mJ to 1.8 mJ every 0.3 mJ. It is noticeable
ally decays with the increase of delay time, and the decline that the spectral line intensity is stronger at a shorter delay
trend of the spectral intensity is roughly the same for differ- time and higher sample temperature, and it is no longer dis-
ent laser energies. cernable for a longer delay time.
It is found that different delay times have an impact on To further understand the effect of the delay time and
plasma spectral characteristics. From these plots in Fig. 6, it sample temperature on the spectral intensity, Figs. 8 and 9
can be observed that, with the increase in delay time, the show the evolution of the emission intensity with laser
spectral intensity first decreases, followed by an increase, energy at three sample temperatures and three delay times.
and then finally decreases again. In the whole process, the In Fig. 8, at the beginning of increasing the laser energy, the
spectral intensity reaches a peak value; this phenomenon emission intensity visibly increases at three different delay
maintains consistent with what is shown in Fig. 4. In other times and at three different sample temperatures. However,
words, there is an optimal delay time for the emission line. as shown in Fig. 8(a), the increasing trend of the spectrum is
The signal is best and the intensity of spectral line is at a no longer apparent after the laser energy is 1.2 mJ. The value
maximum when it reaches this delay time, then the intensity reaches a maximum when the laser energy is 1.8 mJ. At the
begins to gradually decay. The decline trend of the spectral laser energies below 1.2 mJ, the slope of the curve is more
line is roughly the same at different laser energies. significant compared to the trend of the whole graph.
As can be observed from Figures 6(b) and 6(e), the Equally, the evolution of emission intensity with laser energy
emission intensity at 0.6 mJ laser energy and 100  C sample at three different sample temperatures and three different
temperature is nearly equal to that of the emission intensity delay times is shown in Fig. 9. Fig. 9 is a further analysis of
at 1.2 mJ laser energy and 22  C sample temperature in the the information given in Fig. 8, and it shows the same con-
range of delay time from 0.6 ls to 2.0 ls. Therefore, we can clusion as Fig. 8. When the delay time is 0.65 ls, we find
achieve the same or higher emission intensity with a lower that the intensity increases with the increase of sample tem-
laser energy and higher sample temperature. In other words, perature. When the sample temperature is higher, the change
a nearly 100  C increase of the sample temperature can make trend is more evident. The conclusion is the same as dis-
up the difference between the laser energy of 0.6 mJ. cussed above as well as the trend of the curve at 1.3 ls and
Figure 7 shows the distribution of spectral intensity on 2 ls. Therefore, it is demonstrated that increasing the sample
the Na (I) emission lines selected from femtosecond laser- temperature can improve the spectral emission intensity in
induced glass plasma in the range of delay time from 0.65 ls femtosecond LIBS.

FIG. 7. Distribution of spectral intensity obtained with respect to variations of delay time from 0.65 ls to 2.0 ls and the sample temperature from 22  C to
200  C. The color bar shows the intensity from low to high, it is a relative value, and its unit is an arbitrary unit. The laser energy is 0.3 mJ, 0.6 mJ, 0.9 mJ, 1.2
mJ, 1.5 mJ, and 1.8 mJ, respectively. The gate width is 50 ns.
013301-7 Wang et al. Phys. Plasmas 24, 013301 (2017)

FIG. 8. Evolution of emission intensity with laser energy at three delay times (0.65 ls, 1.30 ls, and 2.00 ls) for different temperatures (22  C (a), 100  C (b),
and 200  C (c)).

FIG. 9. Evolution of emission intensity with laser energy at three sample temperatures (22  C, 100  C, and 200  C) for different delay times (0.65 ls (a),
1.30 ls (b), and 2.00 ls (c)).

IV. CONCLUSIONS Research Program of China (973 Program, Grant No.

2013CB922200).
To study the effect of the sample temperature in femto-
second LIBS, the experiment of the femtosecond laser-
1
induce glass plasma is carried out in air. First, we compare D. A. Cremers and A. K. Knight, Laser-Induced Breakdown Spectroscopy
the change of the spectral line emission intensity at three typ- (Wiley Online Library, 2006).
2
A. W. Miziolek, V. Palleschi, and I. Schechter, Laser Induced Breakdown
ical sample temperatures (22  C, 100  C, and 200  C). The Spectroscopy (Cambridge University Press, 2006).
experimental results show that the spectral line intensity of 3
D. A. Cremers, L. J. Radziemski, and T. R. Loree, Appl. Spectrosc. 38,
Na (I) increases with the increase of initial sample tempera- 4
721 (1984).
ture; the change trend is very clear, which is attributed to the G. Arca, A. Ciucci, V. Palleschi, S. Rastelli, and E. Tognoni, Appl.
Spectrosc. 51, 1102 (1997).
increase in the ablated mass and the plasma temperature. 5
D. Anglos, K. Melesanaki, V. Zafiropulos, M. Gresalfi, and J. C. Miller,
Second, we compare the spectral emission intensity for dif- Appl. Spectrosc. 56, 423 (2002).
6
ferent laser pulse energies. The spectral line intensity A. K. Knight, N. L. Scherbarth, D. A. Cremers, and M. J. Ferris, Appl.
Spectrosc. 54, 331 (2000).
becomes stronger and stronger under a certain laser energy. 7
T. X. Phuoc and F. P. White, Fuel 81, 1761 (2002).
Finally, the factor is the influence of delay time on the spec- 8
F. Ghasemi, P. Parvin, N. S. Hosseini Motlagh, A. Amjadi, and S. Abachi,
tral emission intensity. It can be observed that the spectral Appl. Opt. 55, 8227 (2016).
9
line intensity gradually increases with the increase of sample H. Nozari, F. Rezaei, and S. H. Tavassoli, Phys. Plasmas 22, 093302
(2015).
temperature for all delay times. In conclusion, for obtaining 10
A. Chen, S. Li, S. Li, Y. Jiang, J. Shao, T. Wang, X. Huang, M. Jin, and D.
the optimal femtosecond laser-induced breakdown spectros- Ding, Phys. Plasmas 20, 103110 (2013).
11
copy, a combination of the femtosecond laser and the sample 12
Y. Liu, J. S. Penczak, and R. J. Gordon, Opt. Lett. 35, 112 (2010).
pre-heated configuration is found to be important: increasing A. E. Majd, A. Arabanian, and R. Massudi, Opt. Lasers Eng. 48, 750
(2010).
a sample temperature in femtosecond LIBS can reduce the 13
G. Cristoforetti, S. Legnaioli, V. Palleschi, A. Salvetti, and E. Tognoni,
limit of detection and improve the spectral emission Spectrochim. Acta Part B 63, 312 (2008).
14
intensity. B. Rashid, R. Ahmed, R. Ali, and M. Baig, Phys. Plasmas 18, 073301
(2011).
15
S. Z. Mortazavi, P. Parvin, M. R. Mousavi Pour, A. Reyhani, A.
ACKNOWLEDGMENTS Moosakhani, and Sh. Moradkhani, Opt. Laser Technol. 62, 32 (2014).
16
A. Chen, Y. Jiang, T. Wang, J. Shao, and M. Jin, Phys. Plasmas 22,
We acknowledge the support from the National Natural
033301 (2015).
Science Foundation of China (Grant Nos. 11674128, 17
V. Bulatov, R. Krasniker, and I. Schechter, Anal. Chem. 72, 2987 (2000).
18
11474129, and 11504129), and the National Basic L. Ahmat, I. Ahmed, and A. Nadeem, Phys. Plasmas 21, 093501 (2014).
013301-8 Wang et al. Phys. Plasmas 24, 013301 (2017)

19 25
M. E. Asgill, H. Y. Moon, N. Omenetto, and D. W. Hahn, Spectrochim. R. Sangines, H. Sobral, and E. Alvarez-Zauco, Spectrochim. Acta Part B
Acta Part B 65, 1033 (2010). 68, 40 (2012).
20 26
S. Mahmood, R. S. Rawat, M. S. B. Darby, M. Zakaullah, S. V. C. Hanson, S. Phongikaroon, and J. R. Scott, Spectrochim. Acta Part B 97,
Springham, T. L. Tan, and P. Lee, Phys. Plasmas 17, 103105 (2010). 79 (2014).
21 27
S. Tavassoli and A. Gragossian, Opt. Laser Technol. 41, 481 (2009). S. Z. Shoursheini, P. Parvin, B. Sajad, and M. A. Bassam, Appl. Spectrosc.
22
C. F. Su, S. Feng, J. Singh, F. Y. Yueh, I. Rigsby, D. Monts, and R. Cook, 63, 423 (2009).
28
Glass Technol. 41, 16 (2000). A. M. Chen, Y. Wang, L. Z. Sui, S. Y. Li, S. C. Li, D. L. Liu, Y. F. Jiang,
23
C. Lopez-Moreno, B. Smith, I. Gornushkin, N. Omenetto, S. Palanco, J. and M. X. Jin, Opt. Express 23, 24648 (2015).
29
Laserna, and J. Winefordner, J. Anal. At. Spectrom. 20, 552 (2005). S. H. Tavassoli and M. Khalaji, J. Appl. Phys. 103, 083118 (2008).
24 30
S. Eschlb€ock-Fuchs, M. Haslinger, A. Hinterreiter, P. Kolmhofer, N. L. V. Zhigilei, E. Leveugle, B. J. Garrison, Y. G. Yingling, and M. I.
Huber, R. R€ossler, J. Heitz, and J. Pedarnig, Spectrochim. Acta Part B 87, Zeifman, Chem. Rev. 103, 321 (2003).
31
36 (2013). R. Sangines, H. Sobral, and E. Alvarez-Zauco, Appl. Phys. B 108, 867 (2012).