Comparison of sample temperature effect on femtosecond and nanosecond laser-induced breakdown spectroscopy

Miao LIU(劉淼),Anmin CHEN(陳安民),?,Yutong CHEN(陳雨桐),Xiangyu ZENG(曾祥榆),Qiuyun WANG(王秋云),Dan ZHANG(張丹),Dapeng YANG(楊大鵬)and Mingxing JIN(金明星)

1 Institute of Atomic and Molecular Physics,Jilin University,Changchun 130012,People’s Republic of China

2 College of Instrumentation and Electrical Engineering,Jilin University,Changchun 130061,People’s Republic of China

Abstract In this paper,we investigated the emission spectra of plasmas produced from femtosecond and nanosecond laser ablations at different target temperatures in air.A brass was selected as ablated target of the experiment.The results indicated that spectral emission intensity and plasma temperature showed similar trend for femtosecond and nanosecond lasers,and the two parameters were improved by increasing the sample temperature in both cases.Moreover,the temperature of nanosecond laser-excited plasma was higher compared with that of femtosecond laser-excited plasma,and the increase of the plasma temperature in the case of nanosecond laser was more evident.In addition,there was a significant difference in electron density between femtosecond and nanosecond laser-induced plasmas.The electron density for femtosecond laser decreased with increasing the target temperature,while for nanosecond laser,the electron density was almost unchanged at different sample temperatures.

Keywords:laser-induced breakdown spectroscopy,femtosecond laser,nanosecond laser,target temperature

1.Introduction

Laser-induced breakdown spectroscopy(LIBS)that is a good technique for material identification has been widely accepted as a tool in spectral analysis[1–5].In the technique,a high power laser pulse is focused onto the surface of a sample to produce plasma,emitting spectral lines.Some useful information on the elemental composition from spatially and temporally resolved investigation and detection of spectral emission can be obtained.The technique provides many advantages compared to some classical analysis methods.For example,the sample to be analyzed may be a solid,liquid or gas[6–9].It is also an easy,fast and convenient method to detect material composition,almost without sample preparation[10].There is little damage on the sample.The measurement of LIBS can be performed in vacuum,in atmosphere,in deep sea,or in aerospace exploration[11–13].In the past several years,significant advancements have been carried out on the application of LIBS including fast and remote detection[14],space exploration[15],civilian and military environmental monitoring[16],and so on.

LIBS is based on laser-excited plasma,its characteristics are greatly dependent on laser parameters,such as energy,pulse duration,and laser wavelength.Especially the pulse width affects the mechanism of laser ablation.In the past,most LIBS applications use nanosecond pulse laser as excitation source.Nanosecond laser pulses have the advantages of low operational expenses,high reliability,easy operation,and simple maintenance.Because of these advantages,nanosecond laser pulses are applied for extensive industrial and defense applications[17].Moreover,nanosecond LIBS can cause serious fracture of chemical bonds of material components,and optical emission of atoms and ions can only be detected in most cases[18].However,with the development of femtosecond laser systems,much attention has been paid on femtosecond LIBS.Femtosecond laser has been used in many promising applications,such as high resolution,topographic profiling,biological analysis[19].Since femtosecond laser has shorter pulse duration,the ultrafast laser pulse has many outstanding characteristics.Femtosecond pulse can generate low-plasma temperature compared to nanosecond pulses,and the plasma emission has lower background intensity.It is found that the damage threshold decreases with reducing the pulse duration in femtosecond ablation.It also should be noted that femtosecond laser has low laser ablation threshold,high stability,smaller heat-affected zone,and high repeatability[20].Because of these significant characteristics,nanosecond laser pulse has been replaced by femtosecond pulse in many applications.However,the shortcoming is that the femtosecond pulses are absorbed mainly by the target surface and the duration of the pulse is usually so short that there is no interaction between the excitation pulse and corresponding induced-plasma plume,the emission intensity of spectral line is much lower[21].

As the prominent differences in laser pulse widths,the ablation mechanisms between nanosecond and femtosecond laser pulses are different evidently.In nanosecond laser ablation,most of laser energy is used to ablate the target and excite the plasma,and then reheat and ionize the plasma species.All these phenomena take place during the nanosecond laser pulse.While femtosecond laser pulse is very short,these phenomena will continue until the end of laser duration.As femtosecond laser pulse is shorter than electronlattice relaxation and heat conduction times,all laser energy is injected directly into the sample before the ablation starts to create high-ionized species and thermal vaporization subsequently[22–24].For the different ablation mechanisms between nanosecond and femtosecond lasers,spectral characteristics can be expected to be very different.Many researchers carried out corresponding experiments.For example,Verhoffet aldiscussed the dynamics of femtosecond and nanosecond laser-excited plasmas by atomic emission spectra[25],and Elhassanet alreported the influence of femtosecond and nanosecond pulses on the spectra of bronze alloys in LIBS[26].

For optical emission from LIBS,sample temperature is also an obvious factor affecting spectral emission intensity.A large number of experiments have been conducted to investigate the effect of the sample temperature on LIBS.Eschlb?ck-Fuchset alstudied the effect of target temperature on the expansion dynamics of laser-excited plasmas[27],and their results showed that the plasma expansion dynamics on solid aluminum alloy,silicon wafer,and metallurgical slag were dependent on the temperature of the ablated targets,and higher sample temperatures produced larger volume for the ablation craters.Zhanget alpresented the effect of sample temperature on laser-induced silicon plasma[28],and found that the intensity of atomic line was higher as the sample temperature was higher.Sambriet alachieved a considerable decrease of environmental gas resistance to plasma expansion when the substrate was heated[29].However,there are few studies comparing the effect of sample temperature on femtosecond and nanosecond LIBS.

In this paper,we studied the effect of the sample temperature on femtosecond and nanosecond LIBS.The emission spectra from femtosecond and nanosecond laser-induced plasmas were measured at different sample temperatures and laser energies.The corresponding plasma temperature and electron density were calculated using Boltzmann plot and Stark broadening,respectively,helping us to better know the characteristics of plasmas produced by femtosecond and nanosecond lasers.

2.2.Experimental setup

The experimental setup for comparing the effect of target temperature on laser-induced plasmas by femtosecond and nanosecond pulses is presented in figure 1.A brass was chosen as the ablated target in air to create plasmas with different target temperatures.In the experiment,the excitation sources included two different laser systems.The first one was a nanosecond Nd:YAG laser(Surelite III,Continuum)with a pulse duration of 10 ns at a fundamental wavelength of 1064 nm.The other one was a femtosecond Ti:Sapphire laser amplifier(Libra,Coherent)with a pulse width of 50 fs at a fundamental wavelength of 800 nm,and the laser pulse was fired by sending a command(‘man:trig’)to the laser system.A half-wave-plate(HWP)and a Glan-laser-polarizer(Glan)were used to change the energy of femtosecond laser.The laser pulses were converged using a focusing lens with a focal distance of 100 mm on the surface of the brass target.The brass was placed onto the surface of a heating table,and they were placed on a 3D stage(PT3/Z8M,Thorlabs)to supply new surface.The spectra from the laser-induced brass plasmas through a plano-convex(7.5 cm focal length and 5.0 cm diameter)were collected and were converged to a fiber.The fiber was connected with a spectrometer(SP500i,1200 lines per mm,PI-Acton,Princeton Instruments).The optical signal was detected using an ICCD camera(PIMAX-4,1024i,Princeton Instruments).To reduce the continuous spectrum emission and increase the collection of spectral signals,the gate delay and width of the ICCD were set to 0.5 μs and 20 μs,and a photodiode was connected to the ICCD to synchronize experimental data.Hence the photodiode signal was defined as the zero-reference time point of the experiment.The experimental measurements were carried out in air with an atmospheric pressure.

Figure 1.Experimental setup.M:mirror,Glan:Glan-laser polarizer,HWP:half-wave plate,I:iris,DM:dichroic mirror,Pd:photodiode,L:lens,ICCD:intensified CCD.

3.Results and discussion

Figure 2 shows spectral emission produced by nanosecond and femtosecond pulses under various target temperatures in atmospheric environment.The target temperatures are 22 °C,50 °C,100 °C,150 °C,200 °C,and 250 °C for nanosecond laser at a laser energy of 36 mJ and femtosecond laser at a laser energy of 0.5 mJ.Three spectral lines of Cu(I)at 510.55,515.32,and 521.82 nm were used to investigate the change in the emission intensity of each line with different temperatures.In figure 2,it can be seen that the brass temperatures have a great influence on the emission intensity of LIBS in both cases.The emission intensities of the three spectral lines are improved as the target is heated in the cases of femtosecond LIBS and nanosecond LIBS.In the meanwhile,the continuum emission has little change as the target is heated,thus enhancing spectral signal-to-background ratio.These phenomena indicate that preheating the ablated target may enhance the intensity of spectral emission line.

Figure 2.Distributions of optical emission of nanosecond LIBS(a)and femtosecond LIBS(b)with wavelength and sample temperature.Nanosecond and femtosecond laser energies are 36 mJ and 0.5 mJ,respectively.

To study the sample temperature effect on the intensity of spectral line further,we choose the Cu(I)521.82 nm line for comparison of nanosecond LIBS and femtosecond LIBS.The nanosecond laser energies are 22 and 36 mJ,and the femtosecond laser energies are 0.3 mJ 0.5 mJ.Figure 3 indicates that the brass temperature shows a positive influence on the spectral emission,and both nanosecond and femtosecond laser-induced spectral line intensities grow when the brass target is heated,and the line emission at higher laser energy is stronger than that at lower laser energy in fixed temperature.The reason is that high laser energy will excite the brass target to produce a strong plasma[30].In recent years,many researchers discussed the effect of the ablated target temperature on spectral emission of LIBS.For example,Thorstensenet alinvestigated the damage threshold in silicon by varying the target temperature,and they observed that the heating was a better technique for lowering the damage threshold[31].Tavassoliet alinvestigated the method how to increase spectral line intensity,and the result indicated that the intensity of emission line was attributed to sample mass being ablated and the plasma temperature[32].For higher sample temperature,the ablation threshold will be decreased,and more target material will be ablated.In other words,when the sample temperature is high,the ablation threshold is low and more volume of material can be used to produce plasma,and the interaction between sample and laser is much stronger,the plasma temperature increases[32–34].Therefore,increasing the temperature of the ablated target will increase the spectral emission of LIBS.

Also,we notice the difference increasing trends of spectral emission between nanosecond LIBS and femtosecond LIBS from figure 3.The increase of spectral emission in femtosecond LIBS is more obvious compared to nanosecond LIBS.It is attributed to the difference of ablation mechanisms between nanosecond and femtosecond laser pulses.The widths of femtosecond pulses are so short that the energy is directly injected into the target surface to produce laserinduced plasma[9,22,35].While nanosecond pulses have longer durations compared to femtosecond pulses,nanosecond pulse irradiation of ablated target consists of two stages.The leading of nanosecond pulse excites the target to produce plasma.The produced plasma will absorb the trailing energy of the nanosecond pulse in the processes of the plasma expansion,preventing the energy coupling between the nanosecond pulse and the target,leading to the plasma shielding effect[36–40].Although the increase of the brass temperature was the same and the change of reflectivity was the same in figure 3,due to plasma shielding effect,the energy of nanosecond laser arriving at the brass surface was less than that of femtosecond laser arriving at the brass surface.The coupling between the nanosecond laser and the expanding plasma became stronger when the brass was heated,which may lead to the increase of plasma temperature induced by nanosecond laser compared to the case of femtosecond laser at high brass temperature.Next,we calculated the change in plasma temperature when the brass was heated for femtosecond LIBS and nanosecond LIBS.

Figure 3.Cu(I)line at 521.82 nm of nanosecond LIBS(a)with 22 mJ and 36 mJ and femtosecond LIBS(b)with 0.3 mJ and 0.5 mJ as functions of sample temperature.

The plasma temperature is an important parameter for studying the features of the laser-excited plasma.In order to better understand the characteristics of the plasma temperature,we calculated the evolution of the plasma temperature with the sample temperature.The plasma temperature was obtained by the Boltzmann plot and the plasma was in local thermodynamic equilibrium(LTE).The equation is as follows[41–43]:

For the corresponding calculations,the plasma temperature may be obtained from the slope ofversusEk.Here,irepresents the low level andkrepresents the upper level of a transition,Ikirepresents the line intensity,λ is the line wavelength,gkis the statistic weight,Akiis the transition probability,Ekis the energy,kBis Boltzmann coefficient,andTexis the plasma temperature.In the experiment,the Cu(I)510.55,515.32,and 521.82 nm lines were selected to calculate the plasma temperature.

Figure 4 presents the evolutions of the plasma temperature for both nanosecond and femtosecond laser-excited brass plasmas with the increase of the brass target.An increase in the plasma temperature for nanosecond LIBS and femtosecond LIBS can be seen as the brass temperature increases.The ablation threshold of the sample at high brass temperature is lower than that at low brass temperature,as mentioned above,the ablation threshold decreases with the increase of the brass temperature.Ujihara investigated the reflectivity and other optical constants of metals at high temperatures,finding that the reflectivity in the surface is shown to decrease when the target is heated[44].As sample temperature increases,the reflectivity of sample surface will decrease,and then the coupling of the pulse energy to the target can be enhanced.Therefore,the sample will absorb more pulse energy to produce higher temperature plasma.Also,it can be observed that the increase in the plasma temperature for nanosecond LIBS is more obvious compared to femtosecond LIBS.In figure 4(a),the maximum temperature increase is 1943 K for nanosecond LIBS;in figure 4(b),the maximum temperature increase is 1146 K for femtosecond LIBS.In the case of femtosecond pulses,the laser pulse width was shorter than the electron-lattice relaxation-time.Ionization occurred in terms of delay time shorter than the time at which electrons transferred energy to the lattice,and the absorbed energy was concentrated on the electrons at the target surface.The electrons proceed to rise rapidly while the lattice remains cold[22].The target is ionized directly to plasma and little energy is transferred to heat the surrounding regions.Thus the progress belongs to ‘cold ablation’.In contrast to the ‘cold’ progress,there is a ‘hot’ progress in nanosecond ablation.In terms of nanosecond pulses,the leading edge of the laser was used to excite the brass to generate plasma,while the trailing edge was used to heat the generated plasma.After the plasma generation,instead of interact with material,the rest laser beam interacted with the plasma[36,45].The plasma is initiated in the interaction between the leading edge and the surface of the target,the subsequent laser energy is imparted to it for re-heating.The nanosecond laser ablation is a ‘hot’ progress.With the increase of the brass temperature,the brass surface reflectivity decreases,the target will be coupled with more laser energy,exciting stronger plasma.Also the plasma-shielding effect becomes stronger,in other words,the interaction of the trailing edge laser and the expanding plasma becomes stronger.The increase in the plasma temperature produced in the progress for nanosecond LIBS is higher compared to femtosecond laser ablation.Then,let us see if electron density has the same change trend with the plasma temperature when the brass target is heated.

Figure 4.Plasma temperature as functions of sample temperature for nanosecond LIBS(a)and femtosecond LIBS(b).

The electron density is also a critical factor in laserexcited plasma,and can be obtained from the Stark broadening of spectral line where the contributions of other broadening mechanisms(Doppler broadening and Natural broadening)can be neglected.The electron density is obtained by measuring the width of a spectral line[46,47]:

where,Dλ1/2is the width of emission line,neis the electron density,ω is the electron impact coefficient which can be found in the reference[48].In addition,the spectral width includes instrument broadeningThe relationship ofand(measured width)iscan be obtained by fitting spectral data.The instrument broadening is around 0.04 nm determined by measuring the emission line from a low-pressure mercury lamp.The Cu(I)521.82 nm was chosen to calculate the electron density.

Figure 5 shows the evolutions of the electron density with the brass temperature.The nanosecond pulse energies are 22 and 36 mJ,the femtosecond pulse energies are 0.3 and 0.5 mJ.From figure 5,we can find that the electron densities of femtosecond laser-induced plasmas decrease as the brass was heated,while the change in the electron density for nanosecond laser is a slightly increasing trend.It is also noticed that laser pulse energy has an influence on the electron density for femtosecond and nanosecond laser-induced plasma,the electron density is higher at higher laser energy.As we all know,when the femtosecond laser pulse irradiates a target,laser-excited plasma is created rapidly,then the electron density reaches maximum.As the delay goes on,the plasma expands with a great speed,decreasing the electron density[46,49].When the target temperature is higher,the sample can be coupled with more pulse energy,generating higher-temperature and density plasma[50].However,because the experiment was carried out in atmosphere,the air around the target surface also was heated.Based on the Clapeyron equation,the density of air is inversely proportional to the temperature of air in an atmospheric pressure,so the air density decreases with the increase of the target temperature[51].This leads to a dramatically expansion of the Cu plasma plume.Just for the reasons,the electron density of femtosecond laser-induced plasma is lower,when the sample temperature is higher.In addition,we also observed that the electron density of nanosecond laser-induced plasma plume is almost unchanged as the target temperature increases.A possible explanation is that as the target is heated,more material will be ablated,and the rest of the energy(the tailing edge the pulse)is used for plasma reheating,leading to an increase in the electron density.Based on a stronger interaction of the pulse and the plasma,the role of the decrease in air density balances the increase in the electron density.Thus,the change in the brass temperature results in little change in the electron density of nanosecond laser-induced plasma[39].

Figure 5.Evolutions of electron density with sample temperature for nanosecond LIBS(a)and femtosecond LIBS(b).

4.Conclusions

In this article,we discussed the effect of the target temperature on femtosecond and nanosecond laser-induced plasma spectra,and compared the experimental results between femtosecond and nanosecond.An increase in the brass temperature and laser energy led to an enhancement in the emission intensity of femtosecond and nanosecond laserinduced plasmas,and the increase in the spectral line was more evident for femtosecond laser pulse.The results could be understood by the plasma-shielding effect,the shielding became stronger with the increase in the target temperature for nanosecond laser;while there was almost no the interaction between the femtosecond laser and the plasma.We also compared the plasma temperature in femtosecond and nanosecond laser-excited brass plasmas at different target temperatures by Boltzmann plot method.The same as the emission line intensity,the plasma temperature for the two laser pulses with different pulse widths was improved as the brass was heated.The plasma temperature of nanosecond laser plasma was higher and the range of its plasma temperature increase is bigger as the target temperature increased compared to femtosecond laser.This may be understood by considering the differences in the ablation mechanisms between femtosecond and nanosecond laser pulses,the nanosecond laser with longer pulse duration could excite and re-heat plasma.Lastly,the electron density of the plasma was calculated using the Stark broadening.The electron density dropped for femtosecond laser plume in the processes of heating the brass.When the target was heated,the air near the target surface was also heated,leading to a decrease in the air density and expanding of the Cu plasma.However,for nanosecond laser pulse,with sample temperature increasing,due to a coupling of the laser pulse and the plasma,a decrease in air density balances the increase in the electron density.We hope that the study can provide a better understanding of the target temperature effect on LIBS.

Acknowledgments

We acknowledge the support by the Scientific and Technological Research Project of the Education Department of Jilin Province,China(No.JJKH20200937KJ),and National Natural Science Foundation of China(Nos.11674128,11674124 and 11974138).