Laser ablated copper-titanium colliding plasma plume expansion at 10−4, 10−2, and 100 mbar of oxygen ambient pressures has been investigated with a view to understanding heterogeneous colliding plasma plume dynamics and the formation of multi-element stagnation layers using optical emission spectroscopy and fast imaging of plume. The observation of bands and their ambient pressure dependent emission intensity opens up the possibility of the stoichiometrically controlled formation of nanoclusters/nanocomposites.
Laser-ablated (LA) plasmas have been extensively studied both theoretically and experimentally for their applications in microelectronics fabrication,1 production of metallic powders with enhanced catalytic properties,2 the production of cluster species for basic research,3,4 pulsed laser deposition for thin film growth,5 etc. Laser ablated plasmas underpin a number of key commercial analytical techniques employed across a wide range of domains, including material science, biopharmaceuticals, environmental monitoring, etc., namely, laser induced breakdown spectroscopy (LIBS),6,7 laser ablation inductively coupled plasma mass spectroscopy (LA‐ICP‐MS),8 and matrix assisted laser desorption ionization time of flight mass spectroscopy (MALDI‐TOF‐MS).9 LIBS can provide a limit of detection (LOD) down to parts per million (ppm) and depends on the signal to background ratio (SBR).10,11 With the aim to improve the LOD or SBR, several experimental configurations have been used, for example, forming the plasma plume in electric or magnetic fields,12–14 allowing the plasma to expand in a non-uniform magnetic field, which confines the plasma plume. Threading and expulsion of magnetic field lines from the plume generates a current, stimulating joule heating resulting in enhanced collisional processes which helps to mitigate the debris and larger material chunks present in the plume.15,16 The enhancement in the SBR has been shown using double pulse LIBS (DP-LIBS).17 In the extension of this idea, a single pulse produced plasma plume can be replaced by a plume derived from multiple colliding plasmas.18–20 At the collision front between two counter-propagating plasma plumes, the behaviour of one of the plasma plumes can be influenced/modified by the presence of the other plasma plume. When two counter‐propagating laser produced plasmas collide, there are two extremes of resulting behaviour; for low density plumes with a high relative velocity, one observes significant interpenetration and collisional mixing via binary collisions, while for high density plasma plumes, moving at low relative velocity, the plumes decelerate rapidly over a very short distance resulting in a narrow interface region of high density called the “stagnation layer.” The seed plasma parameters and the target geometry determine the properties of the stagnation layer and provide the main engineering controls.21 In practice, even quite well defined stagnation layers can vary in size and density so that one speaks of the “degree of stagnation.” In so‐called “hard stagnation,” the resulting layer tends to be tight and possesses well-defined boundaries. On the other hand, “soft stagnation” is characterized by rather larger collisional regions with diffuse boundaries. The stagnation layer properties can be defined by using the collisionality parameter,18–21 , where D is the separation between the two seed plasmas, and is the ion-ion mean free path which is defined by21–23 , where is the free space permittivity, is the ionic mass, is the relative collision velocity, e is the electronic charge, z is the average ionization stage of the plasma, is the average plasma ion density, and is the Coulomb logarithm. It is clear from the above expression that we can have more control on the stagnation layer properties by varying several parameters compared to single plasma. A stagnation layer gives more freedom to have control on temperature, density, composition, degree of ionization, etc., compared to a single plasma plume.
Although research into colliding laser produced plasmas can be traced back to very early experiments in the mid-1970s,24 there have been surprisingly few studies outside high energy density experiments.25,26 Colliding plasmas appear in astrophysics,27 fusion energy generation,28 and pulsed laser deposition29 among others. A more recent work by Harilal and Kunze26 employing time resolved imaging and spectroscopy has shown how that stagnation layer parameters such as density and temperature can be measured by judicious choice of laser, focusing optics, and/or target geometry parameters. To the best of our knowledge, no work has been done on heterogeneous colliding plasmas, in which the two seed plasmas are formed on two different materials rather the same material.
In the present work, we report plume dynamics of heterogeneous colliding plasmas of copper and titanium in an oxygen environment. We studied the effect of ambient pressure (at 10−4, 10−2, and 100 mbar) on stagnation layer formation after the collision of copper and titanium seed plasmas and how the plasma parameters, viz., density, electron temperature, and emission line intensities vary compared to single plasma plumes. Band emission of the copper dimmer Cu2 (A-X) and titanium oxide (TiO-γ) from the stagnation layer indicates the possibility of forming nanocomposites in the stagnation layer post collision. A Nd-YAG laser (wavelength 1064 nm) was used for creating the Cu-Ti colliding plasma plumes. The study of dynamics of Cu-Ti colliding plasma plumes has been performed at oxygen ambient pressures of 10−4, 10−2, and 100 mbar. Optical emission spectroscopy (OES) and 2-dimenssional imaging of the expanding plume are used to characterize and study the colliding plasma. The paper is organized as follows. Section II gives the details of the experimental setup, details of the results and discussion are given in Section III, and conclusions are presented in Section IV.
The schematic of the experimental setup used is shown in Fig. 1. In order to study the heterogeneous colliding plasma, we used a Nd:YAG laser of pulse width (full width at half maximum, FWHM) comprising an actively Q-switched oscillator (Quanta Ray INDI) delivering a maximum of 500 mJ at its fundamental wavelength, at a pulse repetition rate (PRF) of 10 Hz. In order to focus the laser beam on the concentric Cu-Ti target disc [Fig. 1], the laser beam was split into two parts using a wedge prism.
The position of the wedge prism was adjusted in order to have equal energy in the beam which was checked with an energy meter placed between the prism and focusing lens. The lens had a focal length . The laser beam was incident on the target at an angle of 45° from target surface, and the resulting focal spot was elliptical in shape having major and minor axes and , respectively. The distance between the two focal spots was 1.3 mm. The laser irradiance at each target surface (Cu and Ti) was kept at for all measurements. The colliding copper and titanium plasma plume expansion study was performed in a vacuum chamber which could be evacuated to a base pressure better than 10−6 mbar and was oxygen gas compatible which was admitted to the chamber via a controlled leak valve. We have used oxygen gas at pressures of 10−4, 10−2, and 100 mbar for this work. To observe optical emission, we have used a collecting lens of focal length 10 cm and an optical fibre jig coupled to the entrance slit of a spectrometer (Shamrock SR 303i, ANDOR Technology, USA) equipped with gated intensified charge-coupled device (ICCD) readout. To observe the dynamics of colliding plasma plumes, the collecting lens and an optical fibre jig coupled to the entrance slit of spectrograph (Shamrock SR 303i, ANDOR Technology, USA) shown in Fig. 1 was replaced with the gated ICCD coupled to a simple imaging optic.
III. RESULTS AND DISCUSSION
We have used optical emission spectroscopy and 2-dimensional images recorded using ICCD for a range of delay times post seed plasma formation to track the stagnation layer formation as well as its physical parameters. The optical fibre array was aligned with the stagnation layer expansion axis in order to collect radiation emitted from whole stagnation layer. The results are presented below as three subsets, viz., imaging of the seed plasmas and layer, spectroscopy of atoms and ions, and finally molecular band emission.
Figures 2(a)–2(c) show the colliding plasma plume expansion at 10−4, 10−2, and 100 mbar oxygen ambient pressures, respectively. All images in the figure are normalized to the maximum intensity in that image. It has been observed from Figures 3(a) and 3(b) that copper and titanium plume front positions “r” vary linearly with time “t” at pressures 10−4 and 10−2 mbar before they start colliding and the formation of the stagnation layer begins to shape. The respective velocities are quite close at 10−4 mbar. It has been found that at 10−4 mbar pressure, the copper plasma plume slowed down 30 ns after the ablating laser pulse, whereas at 10−2 mbar pressure, no change in the velocity of the plume expansion was found before the two plumes collided. This can be explained as follows. For a laser wavelength , the two metals (Cu and Ti) have different optical and thermal properties and hence different ablation rates.
The ablation rate of the material can be derived from thermal depth , where is the thermal diffusivity and is the pulse width of the laser used. is a measure of the localization of heating where the smaller the more localized the heating zone. In our case, the calculated thermal depth of copper while for titanium it has a value .30,31 The lower thermal depth/diffusivity of Ti implies a higher rate of ablation compared to copper. Thus, the titanium plume exhibits greater material expansion than the copper plume. Whereas at higher oxygen ambient pressure 10−2 mbar, the expansion characteristics are attributed to the difference in atomic masses of the two materials used. The fraction of energy lost in each collision in oxygen ambient by a copper atom is , where is the mass of the ablated copper atom and is the mass of the oxygen atom, while for titanium, the fractional loss is given by , where is the mass of titanium atom. Since at higher pressures, titanium atoms lose more energy during collisions with the oxygen atoms than copper atoms, the result will be greater confinement of the titanium plume.32,33 Thus, despite having different ablation rates, both plasma plumes have the same velocity at higher pressures, viz., 10−2 and 100 mbar before colliding with each other [Figs. 3(b) and 3(c)]. Further, Fig. 3(c) shows that the expansion of the each plasma plume follows the drag model , where is the maximum distance to which a plume can expand in an ambient atmosphere, and is the damping coefficient of the drag force.34 The extracted values from Figure 3(c), , confirm that the copper plume experiences more drag than the titanium plume and hence is decelerated more rapidly. This can be understood by noting that in Cu–O collisions, the Cu atom loses 64% of its energy, in Ti – O collisions, the Ti atom loses 75% of its energy, while in Cu – Ti collisions, the Cu atom loses 98% of its energy in each collision. Hence, while Ti loses more energy before colliding with Cu atoms, it gains significantly at the Cu–Ti collision plane. It is also observed from Figs. 2(a) and 2(b) that at 10−4 mbar pressure, the stagnation layer formation starts earlier than at 10−2 mbar pressure , because at higher pressure, the plumes become comparatively more confined and take longer time to interact with each other to form the stagnation layer. Whereas Fig. 2(c) shows that despite there being a collision between the two plumes, no well-defined stagnation layer is formed, and instead, instability at the plume front is observed following the collision.
B. Emission spectroscopy
Optical emission spectra of single and colliding plasma plumes were recorded at varying delays with respect to the ablating pulse. Figs. 4(a) and 4(b) show the line emission spectra of a copper plasma, Figs. 4(c) and 4(d) show the titanium plasma spectra, and Figs. 4(e) and 4(f) show the copper titanium colliding plasma spectra at a delay after the laser pulse and at 10−4 mbar oxygen ambient pressure. In single copper plasmas, only copper atomic line (Cu I) emission features15,21,35 have been observed as shown in Figs. 4(a) and 4(b), e.g., at 450.93 nm, at 453.97 nm, at 465.11 nm, at 510.5 nm, at 515.32 nm, and at 521.82 nm, all at 10−4 mbar oxygen ambient while no singly charged copper ion lines (Cu II) were observed. On the other hand, in single titanium plasma emission shown in Figs. 4(c) and 4(d), titanium ion lines35 (Ti II) at 446.84 nm, at 447.08 nm, at 448.83 nm, at 450.12 nm, at 454.96 nm, at 456.37 nm, at 457.19 nm, at 512.91 nm, at 518.86 nm, and at 521.15 nm, and titanium atomic line (Ti I) emission35 at 453.47 nm are all observed.
However, in copper-titanium colliding plasmas emission lines shown in Figs. 4(e) and 4(f), an enhancement in the line intensities of the aforesaid lines of Ti I and Ti II and Cu I lines at 510.5 nm, 515.32, and 521.82 nm has been observed, but other Cu I lines, observed in the single plasma case, have been quenched by collisions with the titanium plasma. The greater degree of ionization observed in the titanium plasma is due to localized heating on the titanium surface,30 resulting in faster melting and evaporation of titanium. The fast melting reveals that a smaller fraction of the laser pulse energy was needed to elevate the surface electron and lattice temperature and the formation of melted zone which could then be further heated by the major remaining fraction of the laser pulse energy absorbed by the process of inverse bremstrahlung (IB).20 The absorption coefficient is given by , where , , and are average charge, electron density, and electron temperature, respectively. The quantity is called the coulomb logarithm, is laser frequency, and is the plasma frequency.20,36 Upon the absorption of the laser energy, an enhancement in the kinetic energy of the electrons occurs which increases the collision frequency. The atoms/ions become more and more ionized resulting in enhanced emission intensity of the ion lines. On the other hand, due to the large thermal depth of copper,30 a greater fraction of laser pulse energy is expended in the melting rather than the heating process resulting in the observation of only atomic line emission and with comparatively less intensity. In the colliding plasma case, the greater enhancement of the intensity of the Ti II lines is due to the collision between copper and titanium atoms. When a heavier copper atom collides with a titanium atom, the latter gains more energy due to it lighter mass,32,33 resulting in greater ionization of titanium and suppressed emission of Cu I lines at lower wavelengths along with an enhancement at the longer wavelength range [Figs. 4(e) and 4(f)]. Similar behaviour of the plasma emission as shown in Figs. 5(a)–5(f) at 10−2 mbar shows no prominent effect of ambient oxygen on the emission line intensities which corroborate the similar plume expansion behaviour shown in Fig. 3(b).
Enhanced atomic line intensities at the shorter wavelength range (445 nm–470 nm) and reduced line intensities at higher wavelengths (505 nm–525 nm) along with the presence Cu II ionic line emission at 512.44 nm35 at 100 mbar oxygen ambient pressure, shown in Figs. 6(a) and 6(b), reveal the enhanced collisional processes in the plasma due to its confinement. However, in the case of the titanium plasma, the combined effect of greater ionization due to localized heating and enhanced confinement, resulting in an increased collision rate, is the reason for the enhanced emission intensity of the titanium ionic lines (Ti II) shown in Figs. 6(c) and 6(d). On the other hand, in the colliding plasma case, the enhanced emission lines belonging to the Ti II spectrum along with an underlying continuum shown in Figs. 6(e) and 6(f) are due to the fact that more energy is transferred from Cu species to Ti species during the collision compared to the single titanium plasma plume case.
Spectroscopic methods have been used to calculate the electron temperature and density of the copper, titanium, and copper-titanium colliding plasmas.37 The temporal variation of electron temperature has been determined from time resolved spectra, assuming local thermodynamic equilibrium (LTE) and using relative line intensities of spectral transitions.38 Specifically, we have used the Cu I lines at 510.5 nm, at 515.32 nm, at 521.82 nm, and at 578.21 nm for calculation electron temperature of single copper plasmas, and the Ti II lines at 439.50 nm, at 444.38 nm, at 446.84 nm, at 450.12 nm, at 454.96 nm, at 456.37 nm, and at 457.19 nm lines for calculating the single titanium and copper-titanium colliding plasma temperatures. The slope of the plot of versus yields electron temperature, where is the intensity of the observed transition line, its transition probability, is the transition wavelength, is the statistical weight, is the energy of the upper level, is the Boltzmann constant, and is the electron temperature. For the transition, the upper state is labelled as m and lower state by n; various aforesaid parameters are taken from the literature.15,35,38 In order to calculate the electron density of the single and colliding plasmas, we have used the well isolated stark broadened line of Cu I at 578.21 nm and Ti II at 368.52 nm.35,39,40 The relationship between the electron density and the FWHM of the Stark broadened lines is given by the expression:41 , where is the electron impact parameter and is the electron density. Contributions from other broadening mechanisms, namely, the Doppler broadening and resonance pressure broadening may also contribute to the line width observed in the laser produced plasma. For our experimental conditions, the Doppler half width for Cu I and Ti II lines, given by the expression , where is the Boltzmann constant, T is the thermal temperature, and is the atomic mass), is and , which are very small compared to the observed Stark broadening and consequently can be ignored. Similarly, the dependence of the pressure broadening on the ground state number density and transition oscillator strength of the corresponding transitions are also small and hence can be ignored.35 We also made a check that the main criterion for LTE to pertain holds for our condition,41 namely, that , where is the energy difference between the states which are expected to be in LTE and is the electron temperature. For the transition at and while for the Ti II transition at 368.52 nm, and the lowest density bounds are and , respectively. The experimentally measured values corresponding to these lines are of the order of , and hence, we can state that the LTE approximation for this analysis is valid.
Figs. 7(a), 7(b), and 7(c) show the temporal evolution of electron temperatures for copper, titanium, and copper-titanium colliding plasmas, respectively. It has been observed that at an ambient pressure of 10−4 mbar, the single plasma plume temperature of copper [Fig. 7(a)] is roughly double than that of titanium [Fig. 7(b)] at a time delay of 100 ns. For the case of the copper-titanium colliding plasmas, the temperature at the collision plane some 100 ns after the laser pulse is similar to Ti and hence displays no enhancement from the interaction with the Cu plasma plume [Fig. 7(c)]. Further if we compare the time evolution of the electron temperatures in all aforesaid cases, we find that single copper plasma temperature remains higher than the corresponding values in the case of single titanium and copper-titanium colliding plasmas.
This is because copper exhibits a great volume of melting and a higher plasma ignition threshold than titanium42 and hence greater absorption of laser energy by the electrons before plasma initiation and a concomitantly higher electron temperature. In the case of copper-titanium colliding plasmas [Fig. 7(c)], the electron temperature gets averaged out due to a high fractional loss of copper energy in collisions with titanium atoms and ions. As the ambient pressure increases (to 10−2 and 100 mbar), the electron temperature difference between single copper plasmas shown in Figs. 7(d) and 7(g) and single titanium plasmas shown in Figs. 7(e) and 7(h) is greater at the initial stage while the copper-titanium colliding plasma temperature remains close to the single titanium plasma due to the averaging [Figs. 7(f) and 7(i)]. The higher electron temperature at higher ambient pressures can be attributed to the increase in the boiling point due to the higher gas pressure at the target surface.43 The longer sustenance of the constant electron temperature at 100 mbar pressure is due to longer emission duration [Figs. 7(g)–7(i)] because of the plasma confinement and hence increased collision rates between the plasma species. Figs. 8(a), 8(b), and 8(c) show the temporal evolution of the electron density of the single copper, single titanium, and copper-titanium colliding plasmas, respectively, at 10−4 mbar oxygen ambient. The electron density is shown to have a maximum in the case of the single copper plasma [Fig. 8(a)] because of its highest electron temperature. The plasma density decreases with time due to a decrease in temperature, so that recombination processes begin to prevail in the plasma. The titanium single plasma electron density [Fig. 8(b)] and colliding plasma electron density [Fig. 8(c)] have similar value of electron densities because of the similar electron temperatures that pertain in each case as noted earlier. It has been observed from Fig. 8(c) that after 200 ns the electron density of the stagnation layer formed by the collision of the copper and titanium plasmas increases up to a time delay of 400 ns and subsequently decreases from this peak. We attribute this peculiar density behaviour to charge separation in the two colliding seed plasmas. Immediately after the seed plasmas are formed, the lighter electrons move faster than the heavier atoms or ions.44,45 Due to this charge polarization, a transient electric field develops, which enhances the speed of titanium ions. Since in the copper seed plasma very little ionization has been observed, the copper atom velocity remains largely unaffected. Further due the heavier mass, copper atoms will be moving slower than that of titanium ions/atoms. Thus, during the initial stage of stagnation layer formation, insufficient copper atom involvement leads to the recombination of titanium ions with the electrons to form the titanium atoms resulting in a decrease in the electron density. At a later stage, more copper atoms infuse the stagnation layer and transfer their energy to the titanium ions during collisions, and hence, the recombination process slows down. As a result, the electron density starts increasing until equilibrium is established.44,45 As the plasma expands, it cools down and recombination again prevails between ions and electrons. At the higher pressure of 10−2 mbar, the time evolution of electron density is shown to be similar to 10−4 mbar (Figs. 8(d)–8(f)). The variation in electron density during the period 200 ns–400 ns in the case of colliding plasmas at 10−2 mbar [Fig. 8(f)] is quite low because of the confinement at higher pressure the stagnation layer formation occurs at a later time and copper atoms get enough time to infuse into the colliding zone when the stagnation layer begins to take shape and hence the electron density equilibrates very quickly. The confinement effect is more prominent at 100 mbar oxygen ambient, and the plasma density remains higher in all cases for longer than in the aforesaid lower pressure cases [Figs. 8(g)–8(i)].
In the case of single copper plasma plumes [Fig. 8(g)], fluctuations with a period of some hundreds of nanoseconds are evident. Although we do not know the exact origin, we surmise that this observation could be the result of compression (and rarefraction) as the expanding plasma plume expands into the surrounding medium. As the plume expands into the surrounding gas, an increase in the ionization rate results, and hence, the electron density increases. However, over time the average density starts to decrease as recombination overcomes collisional ionization. Once the plasma plume and surrounding compressed gas reach equilibrium, the plume enters a compression phase before rebounding and starting to expand again, albeit with lower energy and hence driving a lower and lower ionization rate as the cycles of compression, rarefraction, and concomitant ionization and recombination evolve. No fluctuation is observed in the case of colliding plasmas at 100 mbar because of the intermixing of copper and titanium plasmas which leads to and averaging out of the electron density [Fig. 8(i)].
C. Band emission
The emission of molecular bands in the copper titanium colliding plasma is due to the rapid drop in the temperature which makes a favorable environment for the formation of molecules. Molecular formation occurs due to three body recombination15 and condensation process. Figs. 9(a)–9(c) show the and bands at 10−4 mbar after delay. Enhanced intensities of molecular bands' emission at delay and emission of band at 800 ns delay at 10−2 mbar are due to the comparatively better confinement of the plasmas which favors the more atoms are get to recombine [Figs. 9(d)–9(f)]. However, at 100 mbar ambient pressure [Figs. 9(g)–9(i)], emission of and bands at much later times reveals that atomic species did not get the favorable conditions to recombine, because plasma takes longer time to get cool down to start the condensation process. But after losing their energies in collisional processes and rapid decrease plasma temperature, they start recombining to form molecules and resulting band emission at a later stage. The frequent collisions between background gas and ablated species, viz., copper and titanium, stabilize the collision complexes between atomic and small molecular species that fosters growth of cluster/nanocomposite species. The formation of clusters/nanocomposites in the stagnation layer could prove to be very useful in the field of LA-ICP-MS, in terms of LOD enhancement, and nanocomposite formation with controlled stoichiometry can also be useful in the medical field and the photocatalytic industry.
Dynamics of laser produced copper-titanium colliding plasmas at different oxygen ambient pressures has been investigated using fast photography and emission spectroscopy. The velocity of seed plasmas (Cu and Ti) before their collision and the formation of stagnation layer have been calculated. The calculation of electron temperature and electron density has been done using emission spectra. The copper plasma is shown to have maximum electron temperature and density. This reveals that the inherent material property does affect the plasma parameters. At 100 mbar ambient and in the case of copper plasma fluctuations in the electron density at an early stage may be due compression (and rarefraction) of expanding plasma plume into the surrounding medium. The abundance of bands in the emission spectra shows the possibility of the formation of nanoclusters/nanocomposites. Current experiments indicate that a degree of control over the stoichiometry of deposited nanocomposites is possible by changing the colliding plasma parameters and represents the next stage in our study of heterogeneous colliding plasmas.
This work was supported by the Science Foundation Ireland under Grant No. 12/IA/1742. We acknowledge EU FP7 Grant Agreement No. 318941 under the project “Ultrafast Photonics-Processes and Interactions (UP-PI)” for travel funds. Pramod Pandey acknowledges support under the EU FP7-PEOPLE-2013-IIF Programme, Grant Agreement No. 628789. This work is associated with the FP7 EU COST Action MP1208 and the U.S. National Science Foundation PIRE Grant No. 1243490.