Download as pdf or txt
Download as pdf or txt
You are on page 1of 16

Research Article Vol. 30, No.

10 / 9 May 2022 / Optics Express 17038

Characterization of femtosecond laser-induced


grating scattering of a continuous-wave laser
light in air
Y ULAN W U , 1 P ENGJI D ING , 1,2,4 Y UE Z HENG , 1 T ONGXUN Z HAO, 1
Z OUMINGYANG Z HU , 1 X IAOLIANG L IU , 3 S HAOHUA S UN , 1 J IJIN
WANG , 1 Z UOYE L IU , 1,2,5 AND B ITAO H U 1
1 School of Nuclear Physics and Technology, Lanzhou University, No. 222 Tianshui South Road, Lanzhou
730000, Gansu Province, China
2 Frontiers Science Center for Rare Isotopes, Lanzhou University, No. 222 Tianshui South Road, Lanzhou
730000, Gansu Province, China
3 School of Nuclear Science and Engineering, East China University of Technology, Nanchang 330013,
Jiangxi Province, China
4 dingpj@lzu.edu.cn
5 zyl@lzu.edu.cn

Abstract: Nanosecond laser-induced grating scattering/spectroscopy (LIGS) technique has


been widely applied for measuring thermodynamic parameters such as temperature and pressure
in gaseous and liquid media. Recently, femtosecond (fs) laser was demonstrated to induce the
grating and develop the fs-LIGS technique for gas thermometry. In this work, we systematically
investigated the fs-LIGS signal generation using 35 fs, 800 nm laser pulses at 1 kHz repetition
rate in ambient air by varying the pump laser energies, the probe laser powers and the temporal
delays between two pump laser pulses. The stability of single-shot fs-LIGS signal was studied,
from which we observed that the signal intensity exhibits a significant fluctuation while the
oscillation frequency shows a much better stability. A 4.5% precision of the oscillation frequency
was achieved over 100 single-shot signals. By using a previously-developed empirical model, the
fs-LIGS signals were fitted using nonlinear least-squares fitting method, by which crucial time
constants characterizing the signal decay process were extracted and their dependences on the
pump laser energy were studied. From the measured results and theoretical analysis, we found
that the appropriate range of the overall pump laser energy for reliable fs-LIGS measurements is
approximately located within 80 ∼ 300 µJ. The limitations on the accuracy and precision of the
fs-LIGS measurements, the origin of destructive influence of plasma generation on the signal
generation as well as the electrostriction contribution were also discussed. Our investigations
could contribute to a better understanding of the fs-LIGS process and further applications of the
technique in single-shot gas thermometry and pressure measurements in various harsh conditions.

© 2022 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement

1. Introduction
Coherent light scattering from a grating induced by two crossing laser beams has been intensively
studied, from which a nonlinear optical diagnostic technique termed as Laser-Induced Grating
Scattering/Spectroscopy (LIGS) has been developed [1–3]. Fundamentally, LIGS is a four-wave
mixing process, as briefly shown in Fig. 1. Two laser beams with the same polarization and
wavelength intersect at a small angle (θ), and interfere with each other when they overlap
spatiotemporally, leading to the formation of an electric field interference pattern in the crossing
region. Resonant absorption occurs when atoms or molecules absorbing the pump laser are
present in the grating volume, and subsequent collisional quenching releases absorbed energy
into the immediate vicinity and thus the local temperature increases at the regions of high laser

#460257 https://doi.org/10.1364/OE.460257
Journal © 2022 Received 4 Apr 2022; revised 21 Apr 2022; accepted 21 Apr 2022; published 2 May 2022
Research Article Vol. 30, No. 10 / 9 May 2022 / Optics Express 17039

intensity. The change in temperature leads to density oscillation and therefore an oscillation of
the local refractive index, in which case the grating is normally termed as “thermal grating”.
The thermal grating consists of two superimposed components, including a stationary pattern of
density oscillation arising from the change in temperature (termed as “temperature grating”) and
an acoustic wave generation due to sudden density oscillation (termed as “acoustic grating”).
The induced grating pattern leads to two planar acoustic waves starting from its antinodes and
propagating in opposite directions normal to the grating plane. The stationary temperature
grating decays exponentially due to thermal diffusion whereas the acoustic grating also decays
exponentially but as a result of viscous damping effect. Except for the temperature change that
modulates the refractive index, non-resonant pump laser with sufficient intensity can also create
modulations of the refractive index by the electrostrictive effect [4–7]. In an electrostriction
grating, the electric field with an interference structure polarizes the dielectric medium, and the
spatial inhomogeneity of the electric field causes a motion of molecules or atoms towards the
regions of high laser intensity, resulting in an acoustic wave in the transverse direction. The time
evolution of the density or refractive index grating can be probed by scattering a continous-wave
(CW) laser beam incident to the grating plane at the first-order Bragg angle (θ B ). In the direction
that satisfies the phase-matching condition of four-wave mixing process, the scattering results in
a coherent light, i.e. the LIGS signal, which reveals the temporal dynamics of the laser-induced
refractive index grating.

Fig. 1. Schematic illustration of the principle of the LIGS process: (a) Two pump lasers
intersect with a small angle θ and form a time-varying refractive index grating with a spacing
of Λ; (b) A continuous-wave laser is sent to probe the grating region and interact with the
refractive index grating, leading to a coherent signal scattered in the phase-matched direction
of four-wave mixing process, i.e., LIGS signal.

Except for applications in molecular absorption spectroscopy [8–10] and molecular relaxation
dynamics in gaseous media [11–14], LIGS has been widely applied for optical diagnostics of
combustion chemical reactions and gaseous flow by measuring thermodynamic parameters such
as sound speed, thermal diffusion coefficient, temperature, pressure, flow field velocity and
element concentration [15–29]. Buntine et al. observed LIGS signals in four-wave mixing
experiments in gas in 1994 and studied the rotational absorption spectra of OH molecule by
measuring the dependence of its intensity on the wavelength of pump laser [8]. In 1998, Latzel
et al. realized the potential of LIGS for temperature and pressure measurements and obtained the
results with an accuracy of about 4% and 10% in high-pressure methane/air flames, respectively
Research Article Vol. 30, No. 10 / 9 May 2022 / Optics Express 17040

[19]. In 2006, Stevens et al. achieved one-dimensional temperature and pressure measurement
for the first time in NO2 /N2 gas mixture by using a laser sheet [23]. In terms of practical
applications, Williams et al. firstly used the LIGS technique to measure temperature with an
accuracy of 0.4% in a gasoline direct-injection engine in 2014 [27]. In 2016, Sahlberg et al.
accurately measured the absorption spectrum of water molecules and flame temperature using
a 3-µm mid-infrared pulse laser as the pump laser [28], which triggerred the applications of
mid-infrared LIGS technique for species concentration measurements and flame thermometry
[30–33]. In 2017, Froster et al. demonstrated gas thermometry using the LIGS technique at
rates up to 10 kHz in a static cell and 1 kHz in compressed gas flow, extending the technique to
allow time-resolved measurements of gas dynamics [34]. Domenico et al. further increased the
rate to 100 kHz using a burst laser in reacting flows and premixed laminar methane/air flame,
showing the potential of high-repetition-rate LIGS technique for instantaneous detection of sound
speed, temperature, pressure, etc [35]. In 2019, Sahlberg et al. achieved 1-3% accuracy and
4-7% single-shot precision of pressure measurements using the LIGS technique in the range of
0.5-5.0 bar in a gas cell filled with NO-N2 mixture by applying an improved theoretical fitting
method [36]. A more robust way to derive pressure from the LIGS signal was recently proposed
by Willman et al. where the signal decay time was converted to pressure using a calibration
database, based on which a measurement accuracy of about 10% in an internal combustion
engine was achieved [37].
Recently, Ruchkina et al. employed 800 nm femtosecond laser as the pumping source to induce
the grating and develop the so-called fs-LIGS technique [38]. Strong single-shot signal beam was
generated that can be directly observed with naked eyes, showing striking contrast to nanosecond
laser-based LIGS. By using an empirical model to extract temperature from single-shot data, the
fs-LIGS technique was successfully applied for gas thermometry in heated nitrogen gas flows
for up to 750 K with a deviation of approximately 10%. The generation of plasma with high
laser pulse energy was found detrimental to the fs-LIGS signal by smearing out the oscillation
and causing strong fluctuations. Very recently, we applied this technique to measure elevated
gas pressures in the range from 0.2 to 3.0 bar in a static gas cell, and a quasi-linear relationship
with a slope of 0.96 between the derived pressure and the value measured with a transducer was
achieved, showing the feasibility of the fs-LIGS technique for gas-phase pressure measurements
[39].
In this work, we report on thorough characterization of the fs-LIGS signal generated in ambient
air by measuring the influence of laser pulse energy, from which the optimal energies level of both
the pump and probe laser were determined. It was found that the fs-LIGS signal strength depends
on the pump-laser-pulse energy in a highly-nonlinear manner whereas it generally increases with
the probe laser power in a linear manner. We also investigated the dependence of the fs-LIGS
signal strength on relative time delay between two pump laser pulses. The signals were fitted by
using a previously-developed empirical expression and few crucial parameters, including the
oscillation damping time constant τth , the acoustic transit time τtr and the characteristic time
constant of the fast relaxation process τf , were extracted. Their dependence on the pump-laser-
pulse energies were analyzed, from which the suitable energy range for reliable application of the
fs-LIGS technique was eventually determined.

2. Experimental methods
The schematical illustration of the experimental setup is shown in Fig. 2(a). A commercial Ti:
Sapphire femtosecond system (Astrella, Coherent Inc.) operating at a repetition rate of 1 kHz
was used to deliver laser pulses with central wavelength at 800 nm, beam radius of 5 mm (1/e2 ),
pulse duration of 35 fs, and pulse energy up to 5.8 mJ. The laser beam was split into two arms
(50:50) by a beam splitter (BS1). An optical delay line, composed of two high-reflective mirrors
(M4 and M5) and a mechanical translation stage with a minimal time step of 2 fs, was installed to
Research Article Vol. 30, No. 10 / 9 May 2022 / Optics Express 17041

adjust the relative time delay between two pump laser pulses. These two parallel pump beams,
separated by 18 mm, were focused by a spherical lens (L1, focal length of 500 mm, diameter of
50.8 mm) and intersect at a small angle of θ = 2.06◦ . By carefully adjusting the optical delay line
and the mirror M3, two pump beams was managed to overlap in time and space at the intersection
region to produce a grating. To acknowledge its structure and fringe spacing, a CMOS camera
(MV-3000UC, spatial resolution: 2048 × 1536, pixel size: 3.2 µm × 3.2 µm) and a concave
silver mirror were used to perform two-dimensional imaging of the plasma grating with relatively
high pump laser energy. To calibrate the grating spacing, we placed a thin copper wire with a
known diameter (1 mm) in the position of grating and recorded its image by the CMOS camera,
by which the calibration factor of the imaging resolution can be obtained.

Fig. 2. (a) Schematic illustration of the experiment setup. M1-M10: high-reflection mirrors,
BS: beam splitter, DM: dichroic mirror, L1-L3: spherical lenses, PMT: photomultiplier
tube. (b)-(d) Two-dimensional images of the focal volumes in full scale with two pump laser
beams with total energy of 670 µJ, only the probe beam with power of 250 mW and all three
beams, respectively.

A CW laser system (Verdi G8, Coherent Inc.) provides 532 nm laser as the probe, which
was combined with two pump laser beams through a dichroic mirror (DM) and then focused
by L1 into the grating at first-order Bragg angle (ϕ). Full-scale imagings of the focal volumes
produced respectively by two pump laser beams, only the probe laser beam and all three beams
Research Article Vol. 30, No. 10 / 9 May 2022 / Optics Express 17042

were conducted, as shown in Fig. 2(b)-(d), which helps to monitor the intersecting situation of
the probe into the grating. Part of the probe laser was scattered by the refractive index grating, i.e.
fs-LIGS signal light, whose scattering angle was determined by the phase-matching condition.
We placed another lens (L2, focal length of 600 mm, diameter of 50 mm) at the position having
an identical distance to the focal point to guarantee the four beams are parallel. The light trap
blocked both the pump light and probe light so that only the fs-LIGS signal light passed through,
which was then detected by a monochromator (WDG30-z) coupled with a photomultiplier tube
(PMT, R943-02). The usage of the light traps could not remove all the stray light, so a relatively
long (about 1 m) collection path was arranged, and the stray light was largely removed after
several reflections. The central wavelength of the monochromator was set to 532 nm wavelength
for all experiments to filter out light with any other wavelengths. A narrow-band filter (Semrock,
∼15 nm bandwidth) with optimal transmission at 532 nm wavelength was placed before the
monochromator to further purify the incoming signal. The voltage applied for the photomultiplier
tube is adjustable from 0 to 2000 Volts. Since the PMT is very sensitive, it was completely
covered with black cloth. The fs-LIGS signal was finally displayed on an oscilloscope (Tektronix
DPO 4054) with a bandwidth of 500 MHz and maximum sample rate of 2.5 GS/s.
In the laboratory, the room temperature could vary from 17◦ C to 22◦ C. Because of the high
altitude of ∼1600 metres in Lanzhou city, air pressure keeps at 85.12 kPa. The air humidity is
normally 30%.

3. Experimental results and discussions


3.1. Characterization of the fs-LIGS signal
The picture displayed in Fig. 3(a) exhibits the detailed luminescence image of plasma grating
created with pump laser pulse energy of 670 µJ in the air, in which about 4 fringes can be
observed. The grating spacing was determined to be 22.40 µm. It has to be mentioned that the
luminescence imaging of the grating can no longer be acquired if the pump laser pulse energy is
below approximately 500 µJ. Since the grating spacing is only determined by the crossing angle
of two pump laser beams and the pump laser wavelength, it is reasonable to assume that the
grating spacing keeps the same value for low pump laser energies. Based on this assumption, we)︁
can further calculate the grating spacing for varying laser energies with Λ = λpump / 2sin(θ/2)
(︁
where λpump = 800 nm is the pump laser wavelength and θ = 2.06◦ . The calculated result is
22.23 µm, which is in good agreement with the measured value of 22.40 µm.

Fig. 3. (a) Close-up two-dimensional luminescence imaging of the plasma grating, where
the measured fringe spacing Λm was denoted. (b) Photo of the observed fs-LIGS signal
beam (circled with the red dashed line) taken with a mobile phone.

The incidence of the CW probe laser into the grating at first-order Bragg angle θ B =
λprobe /(2Λ) = 0.61◦ results in its own scattering in the phase-matched direction, and the scattered
light is namely the so-called fs-LIGS signal. Compared to nanosecond laser pumping, as an
Research Article Vol. 30, No. 10 / 9 May 2022 / Optics Express 17043

obvious light beam, the fs-LIGS signal can be directly seen with the naked eye. Figure 3(b)
shows a selected photo of the signal beam, which was taken by a mobile phone on a white board.
From the photo, one can also see sideward the strong probe laser beam (partially blocked) and
downward the 800 nm pump laser beam. The spatial positions of each beam could be better
understood by referring to the setup in Fig. 2 and Fig. 1(a) in [38]. The signal beam appears
elliptical and vertically stretched.

Fig. 4. (a)-(b) Selected single-shot fs-LIGS signal and corresponding FFT frequency result
obtained with the pump laser energy of 482 µJ and the probe laser power of 226 mW,
respectively. (c)-(d) Instabilities of single-shot signal intensity and its oscillation frequency
over 100 shots. (e) Instability of the pump laser pulse energy and probe laser power over
140 seconds.
Research Article Vol. 30, No. 10 / 9 May 2022 / Optics Express 17044

Single-shot fs-LIGS signals were routinely generated, and a typical signal is shown in Fig. 4(a).
The experiments were conducted with the pump laser pulse energy of ∼482 µJ and the probe
laser power of 226 mW. One can see that the signal exhibits a fast buildup and much slow decay
superimposed with only few periodic oscillations. The single-shot signal shows a quite good
signal-to-noise ratio (SNR), which is approximately 83. With a view to practical applications
of the fs-LIGS technique for analyzing and time-resolving unsteady phenomena, we recorded
single-shot signals over 100 shots and studied their instability in terms of variation percentage
relative to the mean value. As shown in Fig. 4(c), the single-shot signal intensity exhibits a
maximal instability of approximately 75%, and its standard deviations was determined as 30.7%.
The corresponding Fast Fourier Transform (FFT) result of the signal in (a), which was processed
within the time range from 0 to 500 ns (10% of the signal maximum), is shown in Fig. 4(b).
Unexpectedly, the oscillation frequency exhibits a much better stability, where the maximal
instability is less than 10%. The averaged FFT frequency was determined as 15.09 MHz while
the standard deviation was calculated as 0.63 MHz, 4.5% of the average value. We also recorded
the laser output instabilities of the pump and probe lasers within about 2 minutes, both of which
shows less than 1% output variation as displayed in Fig. 4(e). Therefore, the laser instability
should be a minor factor for causing the drastic variation of the signal intensity. Except for
negligible air flow due to the operation of ventilators in the laboratory, the nonlinear nature of
the fs-LIGS process could be the major contributor to the intensity variation. Nevertheless, the
relatively small change of the oscillation frequency is critical, which would be beneficial for
reliable measurements such as single-shot gas thermometry.

Fig. 5. (a) Averaged fs-LIGS signal for 100 shots and (b) corresponding FFT result.

Figure 5 shows the averaged fs-LIGS signal and corresponding FFT result. Compared to
single-shot signal, the averaged signal has a much higher SNR of ∼500. In addition, the peaks
around 200 ns becomes more obvious, which contributes to a better determination of the FFT
frequency peak with higher amplitude, as shown in Fig. 5(b). The optimal oscillation frequency is
15.27 MHz, which is in consistent with the value obtained from single-shot data. This consistence
approximately meets the requirement that the average of the FFT peaks of the single-shot signals
equals to the FFT peak of the averaged signal.
Further on, we could estimate the air temperature based on the experimental results. Given the
measured grating spacing Λm = 22.40 µm and the oscillation frequency of the averaged fs-LIGS
signal fosc = 15.27 MHz, the sound speed can be then calculated as υair = Λm × fosc = 342.05
m/s. The approximate equation for calculating the sound speed in air can be expressed as
υair = 331.5024 + 0.603055 × Tlab − 0.000528 × Tlab2 by neglecting minor factors such as humidity

and pressure, where Tlab denotes the air temperature in the laboratory [40]. From this equation,
we can calculate the air temperature as 17.7◦ C. It is slightly lower than the temperature of 20◦ C
Research Article Vol. 30, No. 10 / 9 May 2022 / Optics Express 17045

displayed on a thermometer, leading to an accuracy of approximately 13% for the temperature


measurement in ambient air.
Essentially, the accuracy and precision is limited by the finite number of periodic oscillations
so that the transformed frequency has a low amplitude. The diameter of the focused beam at
the focus of the lens is d = 4λf /(πD) ≈ 51 µm, where the focal length f = 500 mm and the
unfocused beam diameter D ≈ 10 mm. Given the grating spacing of 22.40 µm, only two or three
fringes as well as the number of fs-LIGS signal oscillation can be expected. In order to perform
more accurate fs-LIGS measurements, the fringe number needs to be significantly increased.
The effective approach to do it is then two folds: (1) to increase d by reducing the diameter of
unfocused beam while keeping the grating spacing; (2) to reduce the grating spacing with larger
crossing angle between two pump laser beams. For the second approach, however, the reducing
contrast of the grating fringe visibility could be a problem.
The results shown above were obtained when two pump laser pulses were adjusted to reach
spatiotemporal overlap. Optimal overlap was determined by the strongest dynamic interference
of the supercontinuums generated by the pump lasers when they experience filamentation. The
interference pattern is shown in Fig. 6(a), where two brightest spots are filament-forming pump
lasers surrounded by their self-generated supercontinuum light. Optimal overlap is then regarded
as the time zero of relative delays. Figure 6(b) shows the dependence of the first-peak intensity
of the fs-LIGS signal on the relative time delay of two pump lasers. In this measurement, in
order to avoid the influence of plasma generation, the total pump laser energy and the probe laser
power were kept at 150 µJ and 250 mW, respectively. Obviously, the signal intensity quickly
decreases as the delay is adjusted away from the time zero, and the signal completely disappears
when the delay reaches to approximately 75 fs. The full width at half maximum (FWHM) of the
dependence curve with Gaussian fit is then determined as 50.1±1.8 fs, which is in agreement
with the convolution time of two 35 fs laser pulses, i.e. 49.5 fs.

Fig. 6. (a) Photo of the interference pattern of the supercontinuum generated by the pump
lasers when they experience filamentation. (b) The fs-LIGS signal intensity as a function
of the relative delay between two pump laser pulses. Each point was averaged for five
measurements.

In the experiments, there is no apparent electrostriction-induced fs-LIGS signal observed,


which is expected to appear at the frequency twice that of thermal-grating-induced one [4–7].
Normally, electrostrictive LIGS signals require laser energies one order of magnitude higher
than that needed for thermal ones [41]. At high pump laser energies, one negative factor for
studying the electrostriction-induced LIGS is the plasma formation, but this laser-induced plasma
is weakly-ionized in which only about 1% molecules are ionized by the fs laser pulses [42].
Research Article Vol. 30, No. 10 / 9 May 2022 / Optics Express 17046

Therefore, the electrostrictive effect could still play an important role with high energy pump
lasers, and further investigations on its contribution is needed.

3.2. Influence of laser pulse energy on fs-LIGS measurements


In order to study the influence of pump laser energy on the fs-LIGS signal generation and
determine optimal experimental conditions for further applications, we recorded a set of fs-LIGS
signals by varying the pump laser pulse energy from 0.029 to 2.225 mJ, as displayed in Fig. 7.
The experiments were conducted in an environment with the recorded room temperature of 22
degrees. The graphs in left column show the results obtained with low energies from 29 to 136
µJ whereas the graph in middle column shows that obtained with energies from 256 to 1068 µJ.
From the results of low energies, we can see that the signal intensity drastically increases with
the pump laser energy. The minimum laser energy that can generate detectable fs-LIGS signal
was about 29 µJ, as shown by the log-scale plot in Fig. 7(b). Figure 7(c) is the corresponding
FFT results.

Fig. 7. The fs-LIGS signals, signals in log scale and corresponding FFT results recorded
for low pump laser pulse energy from 29 to 136 µJ (a)-(c), for medium energy from 256 to
1068 µJ (d)-(f), and for high energy from 1.229 to 2.225 mJ (g)-(i), respectively.

With the increase of the pump laser energy from 256 to 848 µJ, the fs-LIGS signal intensity
continues to grow, as shown in Fig. 7(d). However, for the result obtained with 1068 µJ shows a
weaker signal strength accompanied with distortions of the oscillation structure if compared to
that of 848 µJ. Note that the corresponding FFT frequency of 1068 µJ result shows a significant
shift compared to other frequencies in Fig. 7(f). This oscillation structure distortion is caused by
the plasma generation in the probe volume when the laser intensity is strong enough to ionize the
air molecules. With higher energies from 1.229 to 2.225 mJ, as shown in Fig. 7(g), the plasma
generation completely smears out the oscillation structures and also brings drastic fluctuation to
the fs-LIGS signal, as we reported previously [38].
Research Article Vol. 30, No. 10 / 9 May 2022 / Optics Express 17047

To have a deeper understanding on how the plasma formation is affecting the fs-LIGS signal,
we have looked into the literatures [43–46] and acknowledged that the lifetime of femtosecond
laser-induced plasma grating is typically on the order of 100 ps ∼ few ns. The decay of plasma is
a process in which ions and electrons recombine. As a result, heat will be released so that the
thermal grating is still in the process of establishing when the plasma has completely decayed.
Naturally, the plasma grating lifetime is much shorter than the lifetime of the LIGS signal. Due to
the unstable nature of the laser-induced plasma, in which initial tiny fluctuation will be amplified
during the plasma generation, the heat release following recombination inherits this instability and
thus leads to unrepeatable single-shot fs-LIGS signal. The rapid heat release also might smear out
the grating structure, resulting in a LIGS signal without oscillation patterns as we see in our results.
The light scattering by laser-induced plasma grating is no longer resolvable within the traditional
thermal-grating-based LIGS theory. The plasma formation, non-equilibrium thermodynamic
equations, electron-ion dissociative recombination, and three-body recombination, etc. have
to be taken into account to better understand the plasma-grating-mediated fs-LIGS results.
Nevertheless, it would be an interesting topic to pursue in the near future since it could be useful
for plasma diagnostics.
To obtain the variation of the fs-LIGS signal strength with the pump laser energy, we selected
the intensities of three peaks to analyze, and the results are shown in Fig. 8(a). P1, P2, P3
successively represent the three peaks and the results are plotted in logarithmic coordinate.
Firstly, all three peak intensities show similar trend that they rapidly increase with the pump laser
pulse energy when it is below ∼500 µJ. Power function fitting to P1 suggests an exponent of
6.38 ± 0.39, implying a high nonlinearity of the scattering process. Secondly, the peak intensities
start to be off the power dependence curve as the pump laser pulse energy increases beyond 500
µJ and eventually saturate. Thirdly, when the pump laser energy is higher than 1.229 mJ, the
fs-LIGS signal becomes so unstable and unmodulated that the peak positions can be no longer
determined, because of which the energy range was limited up to 1.229 mJ.

Fig. 8. (a) The peak intensities of P1, P2 and P3 as a function of total pump laser energy.
The dashed line represents power function fitting to the data. (b) The first peak intensity and
integrated area of the fs-LIGS signal as a function of the probe laser power. The dashed line
represents linear fitting to the data.

We also investigated the dependence of fs-LIGS signal on the CW probe laser power. Figure 8(b)
shows the first-peak intensity and energy in terms of integrated area of the fs-LIGS signal as
a function of the probe laser power. The pump laser energy was fixed at ∼100 µJ. Both the
intensity and integrated area firstly increase linearly with the probe laser power and saturate when
the power reaches to 1.5 W, and then decreases. It was noticed that there is no shift of the peak
position or change of the temporal profile of the signal when the probe laser power increases.
Research Article Vol. 30, No. 10 / 9 May 2022 / Optics Express 17048

Hence, the experiment is best carried out under the condition when the probe laser power is lower
than 1.5 W. However, higher probe laser powers cause stronger stray light spreading around the
signal beam that brings difficulties for detection. In addition, one can note that there appears two
dips in the curve that position around 500 mW and 800 mW, which is not understandable at this
moment and requires further investigations. In the experiments, the probe laser power of 250
mW was applied to avoid the dips and mitigate the stray light.
To determine the pump laser energy for feasible applications, the FFT frequency and amplitude
of the fs-LIGS signals for different pump laser energies were plotted in Fig. 9. It can be noticed
that the frequency appears stable when the total pump laser pulse energy is below 600 µJ. The
optimal pump laser pulse energy with highest FFT amplitude is approximately 150 µJ for the
current configuration of experiments. For pump laser energies exceeding 600 µJ, the FFT
frequency drops with higher laser energies while its amplitude gets close to zero, suggesting
disappearance of the periodic oscillations. The threshold of 600 µJ is approximately in consistent
with the occurrence of weakly-ionized filament plasma in which case shrill sounds can be heard
from the focal volume. From the robustness perspective of the fs-LIGS technique for practical
diagnostics, the small interval of the pump laser energy, within which the signal oscillation
frequency stays approximately constant, is an obvious disadvantage. However, it also suggests
a potential advantage. Since fs laser pulses with only few hundreds of microjoule energy are
capable to generate strong LIGS signals with stable oscillation frequency, high-repetition-rate fs
lasers could be applied for time-resolved combustion and reacting flow diagnostics.

Fig. 9. FFT frequency and amplitude of the fs-LIGS signal as a function of total pump laser
pulse energy.

3.3. Theoretical analysis of the fs-LIGS signal


In order to extract more information from the measured fs-LIGS signal and focus on the impact
of pump laser energy on the signal features, we fitted the temporal evolution of the fs-LIGS
signal by using the empirical expression developed by Kozlov et al. [47,48], and the procedure is
similar to our previous work [38]. From two-dimensional image of the grating in Fig. 3(a), we
could estimate the value of τtr in our experiments. Given that there appears 4 fringes and the
fringe spacing was measured as 22.4 µm, the whole length of the transverse dimension of the
grating is around 67.2 µm. Considering that the sound propagates over half of the transverse
length, i.e., 33.6 µm, one can obtain that τtr ∼ 98 ns. It is considerably smaller than τa which is
on the order of µs [47]. Therefore, the contribution of acoustic damping was neglected in the
expression by simplifying the decay factor from exp(−(t/τtr )2 − t/τa ) to exp(−(t/τtr )2 ).
Research Article Vol. 30, No. 10 / 9 May 2022 / Optics Express 17049

Figure 10(a) shows a typical fs-LIGS signal generated with pump-laser-pulse energy of 136
µJ and corresponding fit curve with the empirical expression using nonlinear least-squares fitting
method and the Levenberg-Marquardt algorithm. In the fitting process, all three contributions
to the generation of fs-LIGS signal, i.e., the instantaneous and fast energy distribution as well
as the electrostriction effect, were considered. The initial values for the fitting parameters were
estimated from the raw data. The initial value for fosc was set as the FFT result of the signal, i.e.,
∼16.10 MHz. From the buildup time of the first peak, the initial value for τf can be estimated as
15 ns. From the decay of overall signal without considering the oscillation, we estimate τth ≈ 250
ns as the signal has decreased to only 1% of its maxima. The initial values for S0 , Mi , Mf and Me
were set to unity. From Fig. 10(a), one can see a fairly good fit except for slight mismatch of a
weak peak appearing in the range from 200 to 250 ns. Nevertheless, the fitting process results in
a residual less than 0.02 and gives a goodness of fit in terms of adjusted R-square as 0.998.

Fig. 10. Fs-LIGS signal obtained with laser pulse energy of (a) 136 µJ and (b) 626 µJ, and
corresponding fitted curves. Low panels show the difference between measured fs-LIGS
signal and the fitted one.

We have analyzed the fs-LIGS signals obtained with different pump laser energies by above
fitting process, and extracted few temporal parameters including τth , τtr and τf . Figure 11 shows
their variation as the pump laser energy increases. The fs-LIGS signal generated with the energy
Research Article Vol. 30, No. 10 / 9 May 2022 / Optics Express 17050

above 700 µJ cannot be fitted since the temporal profile of the signal has been irregularly distorted
due to the plasma generation. In fact, the goodness of fit of the signal has already become
worse once laser filamentation phenomena start to happen, as shown by the result of 626 µJ in
Fig. 10(b). On the other hand, in the lower limit of energy less than 46 µJ, the signal appears
weak and noisy, and it also decays within the oscillation period, which inhibits a reasonable fit.
From Fig. 11(a)-(c), a general trend can be observed that all the extracted parameters becomes
smaller with the increase of pump laser energy. However, in the energy region of approximately
80∼300 µJ, τth and τtr almost keep constant at values of 280 ns and 102 ns, respectively. These
two values are quite close to the initial values we set for τth and τtr before fitting, i.e., 250 ns
and 98 ns. This region, which we hereby name it as the trusted region, is consistent with that in
Fig. 9 where the FFT frequencies of the fs-LIGS signals have an amplitude larger than ∼0.02,
half the maxima. Since τth is related to the thermal diffusivity χ by [47]
(︂ Λ )︂ 2
τth = χ−1 , (1)

where χ depends on local temperature Tlocal , we can conclude that the pump laser does not
introduce additional thermalization to the probed gas and the grating structure remains unchanged
in the trusted region. An important implication of this result is that it provides us a selection region
for choosing the pump laser energy for further fs-LIGS applications for either gas thermometry
or pressure measurements.

Fig. 11. Extracted fitting parameters including (a) τth , (b) τtr and (c) τf as a function of the
pump laser energy up to 700 µJ. (d) The rising edge details of the fs-LIGS signals obtained
with pump laser energies of 136, 256 and 626 µJ, respectively.

The decrease of τtr beyond the trusted region could be attributed to femtosecond laser
filamentation, which originates from the interplay between optical Kerr effect and plasma
Research Article Vol. 30, No. 10 / 9 May 2022 / Optics Express 17051

formation when the peak power of laser pulse exceeds a critical value Pcr ≈ 5.5 × 1013 W/cm2
[42,49]. As a result, the laser focus appears slightly ahead of the geometrical focus and the
laser intensity inside the formed filamentary column is clamped at Pcr while its diameter is
approximately fixed at 130 µm in air [50]. Given the pulse duration of τl = 35 fs and the
filament diameter of d = 130 µm, we could estimate the laser pulse energy required for forming
a filament in our experiments by E = π(d/2)2 τl Pcr ≈ 255 µJ. Thus, the total energy required
for filamentation is approximately 510 µJ by considering two pump lasers, which is consistent
with the observations that conical emission, a typical signature of filamentation, takes place
when the laser energy exceeds about 500 µJ. The occurrence of laser filamentation leads to slight
displacement of the focal point towards the focusing lens, which is approximately proportional
to 1/ Plaser /Pcr where Plaser is the laser power [51]. Correspondingly, the crossing angle θ of
√︁

two pump lasers slightly increases. According to the formula for calculating the grating spacing,
i.e., Λ = λpump / 2sin(θ/2) , Λ would slightly decreases, thus leading to a decrease of τtr in
(︁ )︁
Fig. 11(b). Furthermore, this slight decrease of Λ would directly explain the deduction of the
oscillation damping constant τth since τth ∝ Λ2 . However, the possible slight variation of the
grating spacing using high pump laser energy could bring about additional uncertainties to the
fs-LIGS measurements and thus undermine its feasibility.
Figure 11(c) shows that τf monotonically decreases with the increasing pump laser energy.
Since τf defines the characteristic time constant of the fast relaxation process, a more rapid
relaxation process at higher energies would in principle lead to a faster buildup of the thermal
grating and therefore a quicker rising edge of the fs-LIGS signal. This was indeed observed
in the experiments. Figure 11(d) shows the rising edges of the fs-LIGS signals obtained with
the pump laser energies of 136, 256 and 626 µJ, respectively. One can see the half-width at
half maximum (HWHM) of the rising edges becomes longer with higher energies, which is in
agreement with the decrease of τf .

4. Conclusion
To conclude, we have systematically investigated the fs-LIGS technique using 35 fs infrared laser
pulse to form the grating in ambient air. Single-shot fs-LIGS signals were presented and its
stability in terms of the signal intensity and the oscillation frequency over 100 shots was studied,
which shows that the intensity exhibits significant fluctuation while the oscillation frequency has a
much stability with only 4.5% variation with repsect to the mean value. The signal generation was
also characterized by varying the pump laser energies, the probe laser powers and the temporal
delays between two pump laser pulses. The minimum energy required to generate detectable
fs-LIGS signal was 29 µJ whereas the optimal energy was determined to be around 150 µJ under
current experimental conditions. It was found that the signal strength rapidly increases with the
pump laser energy when it is below ∼ 600 µJ, showing a very strong nonlinearity.
The fs-LIGS signals were fitted using a previously-developed empirical model. From the
fitting, crucial time constants including τth , τtr and τf that respectively characterizes the density
oscillation damping due to thermal diffusion, the acoustic transmission over the grating and
the fast relaxation process, were extracted. Their dependences on the pump laser energy were
obtained and discussed. From the measured results and theoretical analysis, the appropriate
region of the pump laser energy for reliable fs-LIGS measurements was approximately located
within 80 ∼ 300 µJ.
Due to its strong single-shot signal strength with rather low pump laser energy required, the
fs-LIGS technique holds very promising potential for gas thermometry and pressure measurement
in harsh environments such as gas turbine and aero-engine combustor. We believe that our
investigations could contribute to a better understanding of the fs-LIGS process and further
Research Article Vol. 30, No. 10 / 9 May 2022 / Optics Express 17052

applications of the technique in thermometry and pressure measurements in various harsh


conditions.
Funding. National Natural Science Foundation of China (12004147, 12027809, U1932133).
Disclosures. The authors declare no conflicts of interest.
Data availability. Data underlying the results presented in this paper are not publicly available at this time but may
be obtained from the authors upon reasonable request.

References
1. H. J. Eichler, P. Günter, and D. W. Pohl, Laser-induced dynamic gratings (Springer, Berlin, 1986).
2. J. Kiefer and P. Ewart, “Laser diagnostics and minor species detection in combustion using resonant four-wave
mixing,” Prog. Energy Combust. Sci. 37(5), 525–564 (2011).
3. A. Ehn, J. Zhu, X. Li, and J. Kiefer, “Advanced laser-based techniques for gas-phase diagnostics in combustion and
aerospace engineering,” Appl. Spectrosc. 71(3), 341–366 (2017).
4. K. A. Nelson, D. R. Lutz, M. D. Fayer, and L. Madison, “Laser-induced phonon spectroscopy. optical generation of
ultrasonic waves and investigation of electronic excited-state interactions in solids,” Phys. Rev. B 24(6), 3261–3275
(1981).
5. D. Govoni, J. Booze, A. Sinha, and F. Crim, “The non-resonant signal in laser-induced grating spectroscopy of gases,”
Chem. Phys. Lett. 216(3-6), 525–529 (1993).
6. W. Hubschmid, B. Hemmerling, and A. Stampanoni-Panariello, “Rayleigh and brillouin modes in electrostrictive
gratings,” J. Opt. Soc. Am. B 12(10), 1850–1854 (1995).
7. A. Stampanoni-Panariello, B. Hemmerling, and W. Hubschmid, “Electrostrictive generation of nonresonant gratings
in the gas phase by multimode lasers,” Phys. Rev. A 51(1), 655–662 (1995).
8. M. A. Buntine, D. W. Chandler, and C. C. Hayden, “Detection of vibrational-overtone excitation in water via
laser-induced grating spectroscopy,” The Journal of Chemical Physics 102(7), 2718–2726 (1995).
9. J. A. Booze, D. E. Govoni, and F. F. Crim, “Diffraction mechanisms in gas-phase laser induced grating spectroscopy
of vibrational overtone transitions,” J. Chem. Phys. 103(24), 10484–10491 (1995).
10. D. N. Kozlov and P. P. Radi, “Detection of vibrational overtone excitation in methane by laser-induced grating
spectroscopy,” J. Raman Spectrosc. 39(6), 730–738 (2008).
11. W. Hubschmid and B. Hemmerling, “Relaxation processes in singlet o2 analyzed by laser-induced gratings,” Chem.
Phys. 259(1), 109–120 (2000).
12. R. Fantoni, M. Giorgi, L. De Dominicis, and D. Kozlov, “Collisional relaxation and internal energy redistribution in
no2 investigated by means of laser-induced thermal grating technique,” Chem. Phys. Lett. 332(3-4), 375–380 (2000).
13. W. Hubschmid, “Molecular relaxations in mixtures of o2 with co2 observed on laser-induced gratings,” Appl. Phys.
B 94(2), 345–353 (2009).
14. A. Stampanoni-Panariello, D. N. Kozlov, P. P. Radi, and B. Hemmerling, “Gas-phase diagnostics by laser-induced
gratings ii. experiments,” Appl. Phys. B 81(1), 113–129 (2005).
15. M. Gutfleisch, D. I. Shin, T. Dreier, and P. M. Danehy, “Mid-infrared laser-induced grating experiments of c2h4 and
nh3 from 0.1-2 mpa and 300-800 k,” Appl. Phys. B 71(5), 673–680 (2000).
16. E. B. Cummings, “Laser-induced thermal acoustics: simple accurate gas measurements,” Opt. Lett. 19(17), 1361–1363
(1994).
17. S. Williams, L. A. Rahn, P. H. Paul, J. W. Forsman, and R. N. Zare, “Laser-induced thermal grating effects in flames,”
Opt. Lett. 19(21), 1681–1683 (1994).
18. E. B. Cummings, I. A. Leyva, and H. G. Hornung, “Laser-induced thermal acoustics (lita) signals from finite beams,”
Appl. Opt. 34(18), 3290–3302 (1995).
19. H. Latzel, A. Dreizler, T. Dreier, J. Heinze, M. Dillmann, W. Stricker, G. M. Lloyd, and P. Ewart, “Thermal grating
and broadband degenerate four-wave mixing spectroscopy of oh in high-pressure flames,” Appl. Phys. B 67(5),
667–673 (1998).
20. D. J. W. Walker, R. B. Williams, and P. Ewart, “Thermal grating velocimetry,” Opt. Lett. 23(16), 1316–1318 (1998).
21. M. S. Brown and W. L. Roberts, “Single-point thermometry in high-pressure, sooting, premixed combustion
environments,” J. Propul. Power 15(1), 119–127 (1999).
22. R. Stevens and P. Ewart, “Single-shot measurement of temperature and pressure using laser-induced thermal gratings
with a long probe pulse,” Appl. Phys. B 78(1), 111–117 (2004).
23. R. Stevens and P. Ewart, “Simultaneous single-shot measurement of temperature and pressure along a one-dimensional
line by use of laser-induced thermal grating spectroscopy,” Opt. Lett. 31(8), 1055–1057 (2006).
24. R. C. Hart, G. C. Herring, and R. J. Balla, “Pressure measurement in supersonic air flow by differential absorptive
laser-induced thermal acoustics,” Opt. Lett. 32(12), 1689–1691 (2007).
25. J. Kiefer, D. N. Kozlov, T. Seeger, and A. Leipertz, “Local fuel concentration measurements for mixture formation
diagnostics using diffraction by laser-induced gratings in comparison to spontaneous raman scattering,” J. Raman
Spectrosc. 39(6), 711–721 (2008).
26. B. Williams and P. Ewart, “Photophysical effects on laser induced grating spectroscopy of toluene and acetone,”
Chem. Phys. Lett. 546, 40–46 (2012).
Research Article Vol. 30, No. 10 / 9 May 2022 / Optics Express 17053

27. B. Williams, M. Edwards, R. Stone, J. Williams, and P. Ewart, “High precision in-cylinder gas thermometry using
laser induced gratings: Quantitative measurement of evaporative cooling with gasoline/alcohol blends in a gdi optical
engine,” Combust. Flame 161(1), 270–279 (2014).
28. A.-L. Sahlberg, J. Kiefer, M. Aldén, and Z. Li, “Mid-infrared pumped laser-induced thermal grating spectroscopy for
detection of acetylene in the visible spectral range,” Appl. Spectrosc. 70(6), 1034–1043 (2016).
29. A. Hell, F. J. Förster, and B. Weigand, “Validation of laser-induced thermal acoustics for chemically reacting h2/air
free jets,” J. Raman Spectrosc. 47(9), 1157–1166 (2016).
30. A.-L. Sahlberg, D. Hot, J. Kiefer, M. Aldén, and Z. Li, “Mid-infrared laser-induced thermal grating spectroscopy in
flames,” Proc. Combust. Inst. 36(3), 4515–4523 (2017).
31. A.-L. Sahlberg, D. Hot, R. Lyngbye-Pedersen, J. Zhou, M. Aldén, and Z. Li, “Mid-infrared polarization spectroscopy
measurements of species concentrations and temperature in a low-pressure flame,” Appl. Spectrosc. 73(6), 653–664
(2019).
32. D. Hot, A.-L. Sahlberg, M. Aldén, and Z. Li, “Mid-infrared laser-induced thermal grating spectroscopy of hot water
lines for flame thermometry,” Proc. Combust. Inst. 38(1), 1885–1893 (2021).
33. R. L. Pedersen, A.-L. Sahlberg, D. Hot, and Z. Li, “Saturation dependence of flame thermometry using mid-ir
degenerate four wave mixing,” Appl. Spectrosc. 75(1), 107–114 (2021).
34. F. J. Förster, C. Crua, M. Davy, and P. Ewart, “Time-resolved gas thermometry by laser-induced grating spectroscopy
with a high-repetition rate laser system,” Exp. Fluids 58(7), 87 (2017).
35. F. D. Domenico, T. F. Guiberti, S. Hochgreb, W. L. Roberts, and G. Magnotti, “Tracer-free laser-induced grating
spectroscopy using a pulse burst laser at 100 khz,” Opt. Express 27(22), 31217–31224 (2019).
36. A.-L. Sahlberg, A. Luers, C. Willman, B. A. O. Williams, and P. Ewart, “Pressure measurement in combusting and
non-combusting gases using laser-induced grating spectroscopy,” Appl. Phys. B 125(3), 46 (2019).
37. C. Willman, L. M. L. Page, P. Ewart, and B. A. O. Williams, “Pressure measurement in gas flows using laser-induced
grating lifetime,” Appl. Opt. 60(15), C131–C141 (2021).
38. M. Ruchkina, D. Hot, P. Ding, A. Hosseinnia, P.-E. Bengtsson, Z. Li, J. Bood, and A.-L. Sahlberg, “Laser-induced
thermal grating spectroscopy based on femtosecond laser multi-photon absorption,” Sci. Rep. 11(1), 9829 (2021).
39. Y. Wu, M. Zhuzou, T. Zhao, P. Ding, S. Sun, J. Wang, Z. Liu, and B. Hu, “Gas-phase pressure measurement using
femtosecond laser-induced grating scattering technique,” Opt. Lett. 47(7), 1859–1862 (2022).
40. O. Cramer, “The variation of the specific heat ratio and the speed of sound in air with temperature, pressure, humidity,
and co2 concentration,” J. Acoust. Soc. Am. 93(5), 2510–2516 (1993).
41. A. Stampanoni-Panariello, B. Hemmerling, and W. Hubschmid, “Temperature measurements in gases using
laser-induced electrostrictive gratings,” Appl. Phys. B 67(1), 125–130 (1998).
42. A. Couairon and A. Mysyrowicz, “Femtosecond filamentation in transparent media,” Phys. Rep. 441(2-4), 47–189
(2007).
43. L. Shi, W. Li, Y. Wang, X. Lu, L. Ding, and H. Zeng, “Generation of high-density electrons based on plasma grating
induced bragg diffraction in air,” Phys. Rev. Lett. 107(9), 095004 (2011).
44. Y. Liu, M. Durand, S. Chen, A. Houard, B. Prade, B. Forestier, and A. Mysyrowicz, “Energy exchange between
femtosecond laser filaments in air,” Phys. Rev. Lett. 105(5), 055003 (2010).
45. P. Ding, Z. Guo, X. Wang, Y. Cao, M. Sun, P. Zhao, Y. Shi, S. Sun, X. Liu, and B. Hu, “Energy exchange between
two noncollinear filament-forming laser pulses in air,” Opt. Express 21(23), 27631–27640 (2013).
46. A. Jarnac, M. Durand, Y. Liu, B. Prade, A. Houard, V. Tikhonchuk, and A. Mysyrowicz, “Study of laser induced
plasma grating dynamics in gases,” Opt. Commun. 312, 35–42 (2014).
47. B. Hemmerling and D. Kozlov, “Collisional relaxation of singlet o2 (b1 σg+ ) in neat gas investigated by laser-induced
grating technique,” Chem. Phys. 291(3), 213–242 (2003).
48. D. N. Kozlov, J. Kiefer, T. Seeger, A. P. Fröba, and A. Leipertz, “Determination of physicochemical parameters of
ionic liquids and their mixtures with solvents using laser-induced gratings,” J. Phys. Chem. B 115(26), 8528–8533
(2011).
49. J.-F. Daigle, A. Jaroń-Becker, S. Hosseini, T.-J. Wang, Y. Kamali, G. Roy, A. Becker, and S. L. Chin, “Intensity
clamping measurement of laser filaments in air at 400 and 800 nm,” Phys. Rev. A 82(2), 023405 (2010).
50. S. I. Mitryukovskiy, Y. Liu, A. Houard, and A. Mysyrowicz, “Re-evaluation of the peak intensity inside a femtosecond
laser filament in air,” J. Phys. B: At., Mol. Opt. Phys. 48(9), 094003 (2015).
51. G. Fibich, S. Eisenmann, B. Ilan, Y. Erlich, M. Fraenkel, Z. Henis, A. L. Gaeta, and A. Zigler, “Self-focusing distance
of very high power laser pulses,” Opt. Express 13(15), 5897–5903 (2005).

You might also like