1 Introduction

Laser-induced plasmas (LIPs) have acquired great interest in recent years as spectroscopic sources [1]. The generation of high-density and high-temperature plasmas formed by focusing high peak power laser radiation onto a solid target is still an important field in basic science, engineering, and material processing technology. Progress in this field is largely driven by the facts that fundamental aspects of laser–solid interactions and consequent plasma generation, applied techniques in material processing technology, and sample elemental analysis are involved [2]. The interaction of laser light with solid targets is a complicated and not completely understood process. It consists of different stages: laser ablation of the target, plasma generation, laser interaction with the plasma, plasma expansion, and collision with a substrate [3]. The interaction of laser light with solid targets and the resulting plasma have been studied by a growing number of research groups [4–12].

LIPs and pulsed laser ablation (PLA) have found applications in fields such as analytical spectroscopy, thin film deposition, and inertially confined fusion [1]. In particular, irradiating solid targets with visible and ultraviolet nanosecond laser pulses at moderate energy density are widely used for the deposition of thin solid films. The technique has been applied successfully to a wide range of materials including metals, semiconductors, insulators, and superconductors [13, 14]. For the latter, pulsed laser deposition (PLD) is now recognized as one of the most promising techniques for in situ preparation of high-T c superconducting thin films [14, 15]. Moreover, high selectivity and sensitivity elemental mass analysis techniques based on the laser ablation processes have been considered for several years and are currently under further investigation [16]. Recently, several experiments comparing laser ablation of solid targets by femtosecond, picosecond, and nanosecond laser pulses have also been performed [17]. The plasma produced by PLA using short laser pulses has been shown to be an efficient tool to produce small particles in the nanometer size range [18–20].

The interest in the Ti plasma comes also from the wide variety of possible applications ranging from the biomedical field, where this element is used as bulk and biomaterial thin film, to engineering, where titanium is employed as pure or as alloy for its excellent mechanical and chemical properties. In 2005, Tian et al. [21] studied the laser surface modification for fabrication of bioactive films on the surface of titanium alloys to improve their biocompatibility. The effect of some laser processing parameters on the resulting surface properties of titanium alloys was discussed.

The knowledge of the temperature and electron density of the produced plasma is a necessary requirement for better understanding many of the plasma processes. The temperature is often determined from the relative intensities of some suitable optically thin spectral lines with small separation in wavelength but wide separation in the excitation energies. The derivation of the electron density from line profiles requires well-isolated and optically thin spectral lines affected by the plasma electric field which contributes to the width of the line. Fulfillments of these requirements and high accuracy are the major tasks of the plasma diagnostician.

This study uses suitable optically thin isolated Ti II spectral lines for the characterization of the titanium plasma produced by interaction of IR- and visible-pulsed laser beams (i.e., the fundamental wavelength at 1,064 nm and the second harmonic wavelength at 532 nm from a nanosecond Nd:YAG laser) at atmospheric pressure in ambient air employing the time-resolved optical emission spectroscopy. The studies can be divided into three main parts.

The first part considers the evaluation of the behavior of the emission lines and of the continuum emitted from the titanium plasma at different delay times. Relative intensities of optically thin and well-isolated spectral lines are the necessary requirements for the selection of spectral lines not only for evaluation of spectral line emissions from the produced plasma but also a requirement for the selection of spectral lines for characterization of the produced plasma such as temperature and electron density. The experimental ratio of the doublet resonance spectral lines of Ti II at 321.71 and 325.29 nm used for testing the self-absorption of spectral lines emitted from the titanium produced plasma for the above mentioned ambient gas environment.

The second part concerns the temperature measurements of the produced titanium plasma using Boltzmann plots of three Ti II spectral lines at 346.15, 348.36, and 350.49 nm (group 1) employed by Hermann et al. [22]. Now a new group of Ti II spectral lines is suggested and tested for the first time for measurements of plasma temperature; the lines are at 286.23, 321.71, 325.29, 348.36, and 351.08 nm (group 2). Comparisons between the temperatures determined from both groups of spectral lines are presented.

The third part presents the determination of the electron number density of the produced titanium plasma at different delay times. The Stark width of the Ti II spectral line at 350.49 nm was used for that and the temporal evolution of the electron density is discussed.

2 Experimental set-up

A schematic of the experimental set-up is shown in Fig. 1. The experiment is performed in such a way that a Q-switched Nd:YAG laser (type Brilliant B from Quantel) is used for producing the titanium plasma at the fundamental wavelength (λ = 1,064 nm) and at the second harmonic wavelength (λ = 532 nm). The laser beam is focused onto the target using a quartz lens (f 1 = 50 cm) at 45°; however, the lens position was optimized and readjusted for obtaining the best focus for both laser wavelengths. The calculation of the power density and the focal spot radius were done using the relation [23]:

$$ w_{2} \approx 2.44\,\lambda \frac{f}{{w_{1} }}, $$
(1)

where w 1 is the diameter of the laser beam before focalization (≈1 cm); w 2 is the diameter of laser beam at the focus; f is the focal length of the lens; and λ is the laser wavelength. In this study, the laser energy per pulse is 490 mJ in 6 ns and 200 mJ in 5 ns at the target surface for a laser focal spot of 0.65 and 0.32 mm for the fundamental and second harmonic wavelength, respectively. This corresponds to laser fluence on the target of 0.6 × 1012 W/cm2 for the fundamental wavelength and 1.2 × 1012 W/cm2 for the second harmonic wavelength. The calculated spot size is consistent with the visual observation of the target spots.

Fig. 1
figure 1

Experimental setup

Full size image

A quartz beam reflector is used for monitoring the reproducibility of the incidence laser beam energy on the target, which partially reflects about 6 % of the incident laser energy to an absolutely calibrated energy meter type Ophier model 1Z02165. The target was mounted on a rotating target holder in a stainless steel vacuum chamber, which is slowly rotated to provide a fresh surface after each laser pulse to avoid craters.

The plasma emission is collected in the direction parallel to the target surface (side-on view) with the aid of a quartz lens of focal length (f 2 = 20 cm) in such away to make 1:1 image of the plasma plume at the entrance aperture of an Echelle spectrograph (type Catalina Corp, model SE 200). The spectrograph is coupled to an Intensified Charge Coupled Device (ICCD) camera (type Andor, model iStar DH734-18F). The entrance aperture of the spectrograph is a circular opening 50 μm in diameter and the ICCD is contained in a camera head mounted at the exit port. The system is driven by KestrelSpec software to acquire images from the supported camera, to align and focus the images, and to calibrate and analyze the spectral data. This spectrometer allows for a time-resolved spectral acquisition over the whole UV–NIR (200–1,000 nm) with a constant resolution of λ/Δλ = 4,500 over three points. An Oriel low pressure Hg pin lamp was used for wavelength calibration. The spectra were acquired at different delays after the laser pulse, using constant opening gates of the ICCD (2 μs). The choice of the gate corresponds to a compromise between the need of not having very large variations of the signal during the measurement time, and at the same time having a good signal for calculating line intensities and widths.

The surface of the Ti target was polished to a mirror-like state and the experimental measurements were collected from three different single shots for three different “fresh targets”; this allows us to present an average and a standard deviation of the results.

3 Results and discussions

The emission spectrum of the titanium plasma consists mainly lines of Ti I and Ti II, covering the spectral range of 200–1,000 nm. It has been observed that an intense continuum emission of the titanium plasma is instantaneously produced at the earlier time delays when the laser pulse reaches the target surface this was especially when the fundamental laser applied and less intense in case of second harmonic laser. The continuum emission decreased rapidly with increasing delay time because of the expansion of the plasma in both lasers and as the delay time increases, the isolated lines of the excited Ti neutrals gradually emerge. Moreover, in case of the second harmonic laser, the spectral line at λ = 532 nm for t d = 50 ns was attributed to the laser second harmonic wavelength and did not appear after t d = 50 ns. Figure 2 presents a sample of the collected spectrum emitted from titanium plasma in air.

Fig. 2
figure 2

This figure presents a sample of the collected spectrum emitted from Titanium plasma in air

Full size image

It must be noted that continuum radiation results from collisions of electrons with heavy particles, neutrals or ions and also from the recombination of electrons and ions. Thus, it can be inferred that, at the initial stage of the plasma near the ablated surface, there are large numbers of electrons, ions, and neutrals in the excited states. In the laser–metal interaction, the absorbed energy first heats up the target to the melting point and then to the vaporization temperature. In this process, the evaporation occurs from the liquid metal, and the main source of energy loss is the heat conduction into the solid target.

Figures 3 and 4 show the variation of the recorded intensities of some titanium spectral lines generated by 1,064 and 532 nm laser for different delay times. The intensities of the spectral lines were observed to last even up to 20 and 30 μs, respectively. It can be noticed that the emission intensities of the Ti II spectral lines in case of the 532 nm laser are much higher than those induced by the 1,064 nm laser. This higher intensity may be attributed to higher ablation of the sample with the 532 nm laser. Thus, the mass ablation in the case of the 532 nm laser is greater than that of the 1,064 nm laser for the plasma produced in ambient air at atmospheric pressure.

Fig. 3
figure 3

Emission intensity of some Ti II spectral lines emitted from the plasma generated by the 1,064 nm laser versus delay time

Full size image
Fig. 4
figure 4

Emission intensity of some Ti II spectral lines emitted from the plasma generated by the 532 nm laser versus delay time

Full size image

The mass ablation of the metal depends upon the thermodynamic and optical properties of the sample as well as the laser properties (wavelength, pulse duration, and irradiance). Moreover, the ambient atmosphere plays a dominant role in the mass ablation rate [24]. This effect depends also on the physical properties of the ambient atmosphere. In the future, a detailed study should be considered applying equal fluence for the both 1,064 and 532 nm lasers which is not considered in this study.

For evaluating the plasma parameters and extracting quantitative data from the line intensities, it is important to use optically thin spectral lines. The ratio of emission intensities of resonant and non-resonant lines should be verified according to a procedure for the “optically thin” limit described by Cremers and Radziemski [25], Simeonsson and Miziolek [26], and Sabsabi and Cielo [27]. A multiplet is very useful in the assessment of self-absorption [28]. Only a comparison of the measured intensity ratio of both lines with their optically thin limit is required. The optically thin limit in this case is simply given by the branching ratio [29]:

$$ \frac{{I_{1} }}{{I_{2} }} = \frac{{g_{1} A_{1} \lambda_{2} }}{{g_{2} A_{2} \lambda_{1} }}, $$
(2)

where I is the intensity of the spectral line, g is the statistical weight for the upper level of the transition, A is the transition probability, and λ is the wavelength corresponding to the spectral line.

Testing the self-absorption of Ti I spectral emission lines reveals that all Ti I spectral lines are affected by self-absorption while examining Ti II spectral emission lines shows that Ti II spectral lines are free from self-absorption. This is done by comparison of the theoretical optically thin limit defined by the branching ratio in Eq. (2) for the doublet lines of the transition of Ti II at 321.71 and 325.29 nm. The atomic data of the mentioned lines are presented in Table 1 [30]. The experimental measured ratio is in good agreement with the calculated optical thin ratio as defined by the branching ratio, and this is shown in Figs. 5, 6.

Table 1 Atomic data of doublet Ti II spectral lines (321.71 and 325.29 nm) [30]
Full size table
Fig. 5
figure 5

The measured intensity ratio of Ti II spectral lines [I 325.29 nm/I 321.71 nm] at the laser wavelength of 1,064 nm (circles) for different delay times with the calculated optically thin limit (solid line)

Full size image
Fig. 6
figure 6

The measured intensity ratio of Ti II spectral lines [I 325.29 nm/I 321.71 nm] at the laser wavelength of 532 nm (circles) for different delay times with the calculated optically thin limit (solid line)

Full size image

Recalling the fact that transitions arising from high-lying upper levels are hardly affected by any self-absorption [28], it can be concluded that the spectral lines from transitions between the same or higher energy levels are all free of self-absorption for the present experimental conditions.

3.1 Measurements of excitation temperature

The excitation temperature is considered to be one of the most important parameters used to characterize the plasma. Accurate knowledge of the excitation temperature leads to the following understanding of the plasma generation: vaporization, dissociation, excitation, and ionization. For a specific state of plasma, the excitation temperature can be determined from the measurement of the intensity of spectral lines assuming that the population of the energy levels follows the Boltzmann distribution law. The equation for the intensity of a spectral line emitted from plasma at LTE can be given by [31, 32]

$$ I_{ij} = \frac{L}{4\pi }h\upsilon \;g_{i} A_{ij} \frac{{N_{\text{o}} }}{U(T)}\exp \left\{ { - \frac{{E_{\text{exc}} }}{{k_{\text{B}} T_{\text{exc}} }}} \right\}, $$
(3)

where U(T) is the partition function of the atom or ion, L is the plasma length, A ij is the transition probability, g i is the statistical weight of the upper level, N o is the total number density of the atom or ion, T exc is the excitation temperature, and E exc is the excitation energy of the upper level.

However, in case of a relative ratio of two spectral lines from the same ionization stage, the excitation temperature (T exc) can be written as

$$ \frac{{I_{1j} }}{{I_{2j} }} = \frac{{A_{1} g_{1} \lambda_{2} }}{{A_{2} g_{2} \lambda_{1} }}\exp \left\{ { - \frac{{E_{1} - E_{2} }}{{k_{\text{B}} T_{\text{exc}} }}} \right\} $$
(4)

Note that the accuracy of the temperature determination is improved when the difference between the excitation energies of the two lines is larger than 2 eV [33, 34]. Unfortunately, in many cases, the energy difference (E 1 − E2) is below 2 eV. For this reason, it is preferable to use many spectral lines from the same thermometric species. By plotting ln (Iλ/gA) versus E exc, the so called Boltzmann plot is obtained. If the population of the energy levels follows a Boltzmann distribution, the plot will be a straight line with a slope equal—(1/k B T exc). Deviations from this straight line represent deviations from the Boltzmann distribution, which can be interpreted to correspond to an over or under population of the various energy levels.

In this study, the excitation temperature of the generated titanium plasma had been determined initially using the Boltzmann plot of three spectral lines of Ti II at 346.15, 348.36, and 350.49 nm (group 1) which were previously introduced by Hermann et al. [22]. Here, five new Ti II spectral lines were tested and developed to be used for the first time in the determination of the excitation temperature. Their wavelengths are 286.23, 321.71, 325.29, 348.36, 351.08 nm (group 2). The spectroscopic atomic data of the mentioned lines groups are presented in Tables 2 and 3 [30]. Figures 7 and 8 show examples of the emission spectra of these lines for different conditions of measurements.

Table 2 Ti II spectral lines (group 1) and their atomic data [30]
Full size table
Table 3 Ti II spectral lines (group 2) and their atomic data [30]
Full size table
Fig. 7
figure 7

An example for the selected Ti II spectral lines used in determination of the plasma temperature generated by the 1064 nm laser at t d = 2 μs

Full size image
Fig. 8
figure 8

An example for the selected Ti II spectral lines used in determination of the plasma temperature generated by the 532 nm laser at t d = 2 μs

Full size image

As the wavelength interval between the selected spectral lines is not small enough for group {2}, a correction of their intensities is necessary. The necessary correction was performed using the sensitivity calibration curve obtained with a Deuterium–Tungsten Halogen lamp. Figures 9 and 10 show typical Boltzmann plots for different delay times for the plasma generated by 1,064 and 532 nm lasers. The regression was lying between (0.966 and 0.999) for different experimental operating conditions indicating that the selected lines are sensitive to temperature.

Fig. 9
figure 9

Boltzmann plots for different delay times for the plasma generated by the 1,064 nm laser obtained from group {1} and group {2} spectral lines

Full size image
Fig. 10
figure 10

Boltzmann plots for different delay times for the plasma generated by the 532 nm laser obtained from group {1} and group {2} spectral lines

Full size image

Figure 11 shows the evolution of the temperature of the Ti plasma generated by the 1,064 nm laser at atmospheric pressure of ambient air determined from group {1} and group {2} spectral lines. The temperature was highest at the initial stage of the plasma, decreasing rapidly with delay time. A dip in the excitation temperature can be noticed at early times. The temperature ranged from about 1.40 to about 1.12 eV in the delay time range (0–7 μs) for group {1}, and from about 1.47 to about 1.09 eV for group {2} in the same time delay range. Figure 12 shows the temporal variation of the temperature of the Ti plasma generated by the 532 nm laser at atmospheric pressure of ambient air determined from group {1} and group {2} spectral lines. The temperature is higher at its initial stage, decreasing rapidly with delay time. A slightly dip followed by a peak in the excitation temperature can be noticed at the early times. The excitation temperature ranged from about 1.41 to about 1.11 eV in the delay time range (0–7 μs) for group {1}, and from about 1.39 to about 1.10 eV for group {2} in the same time delay range. These results reveal that the behaviors of the plasma temperatures determined from both groups of spectral lines are in agreement.

Fig. 11
figure 11

Temporal variation of the excitation temperature of Ti plasma generated by the 1,064 nm laser from group {1} and group {2} spectral lines

Full size image
Fig. 12
figure 12

Temporal variation of the excitation temperature of Ti plasma generated by the 532 nm laser from group {1} and group {2} spectral lines

Full size image

3.2 Measurements of the plasma electron density

Stark broadening measurements of spectral lines are considered to be the most powerful spectroscopic technique for determining the electron density of dense plasmas. This method assumes the Stark effect to be the dominant broadening mechanism (Doppler broadening, pressure broadening mechanisms, collisions with neutral atoms, i.e., resonance and Van der Waals broadenings being negligible). The validity of this assumption was generally admitted in works on LIP and justified in various studies [35–37]. For well-isolated lines in neutral and singly ionized atoms, Stark broadening is predominantly driven electron collisions. As Stark broadening is rather weakly independent of the temperature, the electron density can be determined with a good accuracy.

The full width half maximum (FWHM) of Stark broadened lines is related to the electron number density [38, 39] and can be given by

$$ \Updelta \lambda_{{{1 \mathord{\left/ {\vphantom {1 2}} \right. \kern-0pt} 2}}} = 2W\frac{{N_{\text{e}} }}{{10^{16} }}\left[ {1 + 1.75A\left( {\frac{{N_{\text{e}} }}{{10^{16} }}} \right)^{\frac{1}{4}} \left( {1 - \frac{3}{4}N_{\text{D}}^{{ - \frac{1}{3}}} } \right)} \right], $$
(7)

where W is the electron impact parameter or Stark width parameter, N e is the electron density in cm−3, A is the ion impact parameter, and N D is the number of particles in the Debye sphere which can be estimated from the relation

$$ N_{\text{D}} = 1.72 \times 10^{9} \frac{{T_{\text{e}}^{\frac{3}{2}} }}{{N_{\text{e}}^{\frac{1}{2}} }} $$
(8)

W and A are functions of temperature and can be obtained approximated by second order polynomials from Ref. [39].

In this study, the Ti II spectral line at 350.49 nm was used for determining the electron density of the produced plasma. In LIP where low temperature and high densities are present, the contribution of Doppler, Van der Waals, and resonance broadenings are very weak so the second term of Eq. (7) can be neglected. Then

$$ \Updelta \lambda_{{{1 \mathord{\left/ {\vphantom {1 2}} \right. \kern-0pt} 2}}} = 2W\frac{{N_{\text{e}} }}{{10^{16} }} $$
(9)

The Stark width parameter of the Ti II spectral line of 350.49 nm has been experimentally determined by Hermann et al. [22, 40]. The analysis of the line profile proceeded using a matlab routine in the way that the convolution of the measured instrumental profile with a Lorentzian profile represents the Stark profile of variable width, and comparison the resultant Voigt profile with the measured spectral line using the least square fit method. This leads to an accurate value of the Stark full width at half maximum of the interested spectral line. The measured instrumental width was 0.09 nm (determined from the FWHM of the Hg line at λ = 280.35 nm emitted by a standard low pressure Hg lamp. Figures 13 and 14 present an example for the line profile of Ti II at λ = 350.49 nm with its best fit for the plasma generated by 1,064 and 532 nm lasers at t d = 2 μs.

Fig. 13
figure 13

Example of the line profile of the Ti II at λ = 350.49 nm line (dotted line) with its best fit (solid line) for the plasma generated by the 1,064 nm laser

Full size image
Fig. 14
figure 14

Example of the line profile of Ti II at λ = 350.49 nm line (dotted line) with its best fit (solid line) for the plasma generated by the 532 nm laser

Full size image

Figures 15 and 16 show the temporal variations of the electron density for the plasma generated by 1,064 and 532 nm laser, respectively. The electron density is high at the initial stage of the plasma, decreasing rapidly. Therefore, the electron density ranges from about 2.45 × 1018 to about 5.41 × 1017 cm−3 in the delay time range (0–20 μs) for the plasma generated by the 1,064 nm laser. The electron density ranges from about 2.04 × 1018 to about 1.94 × 1017 cm−3 in the delay time range (0–30 μs) for the plasma generated by the 532 nm laser.

Fig. 15
figure 15

The temporal variation of the electron density of Ti plasma generated by the 1,064 nm laser

Full size image
Fig. 16
figure 16

The temporal variation of the electron density of Ti plasma generated by the 532 nm laser

Full size image

4 Conclusions

A new set of Ti II spectral lines as good candidates for measurements of the excitation temperature for LIPs generated by irradiation of titanium targets in ambient air is developed. Studying the emission spectrum of titanium plasma generated by laser wavelengths of 1,064 and 532 nm for different delay times at atmospheric ambient air pressure in the spectral range (200–1,000 nm) reveals that an intense continuum emission of the titanium plasma is instantaneously produced when the laser pulse reaches the target surface. The continuum rapidly decreases with the delay time. In general, the second harmonic laser-induced spectra are characterized by a lower continuum contribution, shorter duration, and better resolution of the spectral lines with respect to the spectra from the plasmas produced by the fundamental laser wavelength. The present experiment has demonstrated that the second harmonic LIP decays for a longer time than the fundamental LIP as the recombination process proceeds slower.

It must be noted that both the temperature and the electron number density are higher close to the target surface. Indeed, the leading edge of the laser pulse rapidly heats, melts, and vaporizes material just above the surface. Some of the laser energy then heats the evaporated material. In an ambient gas, the expansion of the laser-produced plasma of the target material can be quite complex and strongly dependent on a number of properties like atomic mass and background atoms, initial plume energy, density, etc. The plume expansion dynamics is determined by the interaction of the plasma species with the background gas, which results in a change in the properties of the plasma emission.

Moreover, the temperature from the Boltzmann plot gives the effective excitation temperature and this is equal to the electron temperature by theoretical definition, and this implies indeed a Maxwell electron distribution. This is of course conditioned by the existence of Maxwell distribution and the existence of a Boltzmann distribution between the upper levels during the laser pulse duration. At time t = 0, the laser pulse of few ns duration hits the target and as the ICCD gating is 2 μs, so at all later times the data are averages over this 2 μs. The time scale of establishing a Maxwell distribution is given the electron–electron collision time and that for LTE between the upper levels by electron collision between the upper levels. An estimate indicates that the electron–electron collision times are less than 1 ps for the initial high densities, hence assuming a Maxwell distribution and thus a temperature is justified. At these electron densities, collision between the upper levels are also fast enough to establish PLTE between high-lying levels. Estimates indicate respective relation times less than the laser pulse duration of 1 ns.