3. Results and Discussion
Monitoring with the ICCD 2 ns Pi-Max 3 camera during pulsed laser deposition (
Figure 2) allowed the calculation of the speed of the ablation plume as being of 6.5 x 10
3 m/s. Based on the same measurements, a periodical plasma layered structure was highlighted at about 1 mm from target, 447 ns delay, with the distance between two layers of particles approximately the same, of 0.06 mm. This phenomenon could be assigned to a perturbation materialized in the electric field of diffusion of the diffusion current generated by the charged carriers in the plasma plume in motion and in the shock wave developed under the very high velocity plasma plume.
The number of species (atoms, ions, crystalline structures, etc.) developed in the ablation plasma during its movement and the phenomena of recombination, crystal growth and cooling were also considered. In this sense, analyzes were carried out with the mass spectrometer connected to the deposition installation. For the analysis of the ablation plume’s composition, an energy of 30 mJ of 532 nm laser beam was used in a pulsed regime of 10 ns and 100 Hz repetition frequency, this being at the same time the source of ionization of the component chemical species. The elements silver, iron, nickel and boron were identified as results according to the spectral lines in the spectra presented in
Figure 3.
Reaching an ionization level of silver atoms is located after a significant time interval, namely 100 µs, as it results from the 3D diagram obtained during the analysis with the mass spectrometer (
Figure 4).
The composition of the ablation plume corresponds to the results of the analysis of Ag target with Fe and Ni impurities, which was treated with boron-based compounds. The EDX analysis of the target was presented in our previous works as being on two analyzed areas of 64.56% Silver, 27.01% Nickel, 8.43% Iron and of 77.20% Silver, 17.56% Nickel, 5.20% Iron, respectively. In addition, the mass spectrometry also highlighted the presence of boron which was expected according to the treatments during the fabrication of the target.
In order to understand the phenomena that contributed to the ionization of the chemical species in the silver target with impurities, used in the laser ablation process, a simulation was conducted in COMSOL 5.6. The mathematical model and its implementation in the Heat Transfer module were made according to [
11,
12,
28], and the parameters in the simulation were the same as the laser parameters used during the analysis with the mass spectrometer. The results obtained by simulation are presented in the form of ionization energy threshold (IET) diagrams in
Figure 5. According to
Table 1 and plots of
Figure 5e,f, the threshold of ionization energies is reached and exceeded even after a single laser pulse for the elements iron (maximum achieved of 77.9 eV compared to IET of 7.9024 eV), nickel (maximum achieved of 48.5 eV compared to IET of 7.98 eV), boron (maximum achieved of 270 eV compared to IET of 8.298 eV). In the case of silver, the simulation indicates thermal energies of 0.62 eV (
Figure 5d) developed under the action of one laser pulse, which is below the ionization energy threshold level of 7.5762 eV. Furthermore, the analysis of the simulation results was performed for a time of 2 μs from the ignition of the laser pulse, as much as the processing memory of the computer allowed in the context where the data acquisition is very large, i.e., a large number of reading nodes given by extra fine mesh required. From the extended diagrams for the duration of 2 μs from the ignition of the laser pulse, it appears that, in the presence of boron, each of the elements silver, iron and nickel, respectively, changes its thermal behavior in the sense of extending the heating effect of the ablation plasma. In this sense, although the thermal energy decreases during the laser pulse, the residual thermal energy will improve the thermal effect of the next laser pulse. This phenomenon leads to the idea that the generation of silver ions under low pulsed laser energy is influenced by the presence of boron or boron compounds present in residual form after introduced during the processing of the target.
Returning to the optical images of layered structures obtained with the ICCD camera, that could be in connection with the analysis of the ablation plume with mass spectrometry, which indicates a large number of neutral and charged chemical species (atoms and ions of silver, iron, nickel, boron, as well as different combinations of Ag-B, Fe-B, Ni-B etc.).
The diffractogram obtained from the XRD analysis of the thin film deposited from the Ag target (Ag-thin film) shows its polycrystalline structure (
Figure 6). The X-Ray diffraction analysis on the Ag-thin film evidenced the crystalline specific sharp peaks and Miller indices as (111), (200), (220) and (311), as in literature [
28,
29,
30,
31,
32]. Since the deposition was performed on a glass substrate at ambient temperature (without heating), it means that the crystalline structure was formed in the ablation plume, by crystal growth. Silver crystalline states in different crystal growth stages could also contribute to the optical images of the plume of ablation as layered structures of
Figure 2.
In the images obtained by SEM and AFM analysis (
Figure 7), a morphology consisting of rings and droplets as well as their aggregates is noticeable. Formations with sizes ranging from 2.9 μm to 0.801 μm and even smaller for rings were measured, and in the case of droplets the sizes vary from 2.403 μm to 0.133 μm and even smaller. These indicate the formation of nanoparticles, as well as micrometric structures, some resulting from the aggregation of smaller particles, while others as melt splashes.
This morphology of the thin metallic films indicates, as previously mentioned [
12], the instabilities in the fluid phases during ablation and plasma plume motion from target to support, which consist in the following phenomena and processes that are schematically represented in
Figure 8.
a) One of the instabilities, a RTI type, occurs to liquid - gas interface in the areas on the target with temperatures near the boiling point (
Figure 8 a1), when gas phase starts to develop into the liquid phase in a bubbling phenomenon. The bubbles of lower density than the melted phase will tend to develop in volume, being at the same time compressed by the liquid phase and a pressure gradient will be generated, where p
gas > p
liquid while ρ
gas < ρ
liquid, fulfilling the condition (1):
[
26]. The pressure and density gradients will generate the perturbation responsible for the RTI and the gas bubbles will leave the liquid phase generating also a splash of liquid (
Figure 8 a4). The liquid streams splashed under PRI will break-up into droplets for unstable modes (induced under the laser beam electromagnetic field and the electric field of diffusion current) when
, where k is the wave number. The fastest growing modes occurs for
, when the wavelength of the disturbance is
[
33,
34].
b) The other situation were an instability of RTI kind is developed is on the areas of the target where on the irradiated surface the temperature highly exceeds the boiling point, so that a plasma/gas layer develops on a melted layer (
Figure 8 b1). The plasma/gas will expend the volume in both directions, toward the support, but also toward the target, and, in the latter case, a deformation inside melted material will be induced, continuing to “dig” as a spike shape hole filled with gas (
Figure 8 b2). The side melted material will be under a stress that will be released at the liquid-gas interface in the immediate vicinity, generating again (like situation “a”) a splash of liquid (
Figure 8 b3) that, crossing the plasma, will be under the perturbing effects of the electric field of the diffusion current induced by charged carriers, leading to the liquid streams break-up into droplets (
Figure 8 b4), under the same conditions presented in paragraph (a).
c)
Figure 8c schematically represents the same type of RTI instability as in paragraph b) that occurs in target areas where temperature values are close to boiling points only at the boundary between the liquid phase and the plasma/gas phase developed on the irradiated surface of the target. The pressure in the plasma/gas phase is increased by the thermal effect and phase change, the plasma/gas phase tending to consume its volume and, with the pressure and density gradients under conditions from paragraph (a), will develop an inward peak. The melt will be directed towards the neighboring molten zone, generating on the one hand flows and droplets in the gas/plasma zone (
Figure 8 c1) as a PRI-type instability, and, on the other hand, gas bubbles will form in the molten zone and they will continue to move into the liquid mass in the region of lowest pressure at the interface of the liquid phase with the ambient atmosphere, where in the meantime the ablation plasma "cloud" should have already expanded, lowering the pressure as well and providing the carriers charge to generate the diffusion electric field. The escape of the gas from the bubbles in the liquid mass will be accompanied by a kind of volcanic eruption, in a splash nozzle-less spray when the droplets are generated as PRI under the electric field of the diffusion current disturbance. A fraction of the droplets generated under the conditions described in paragraphs (a), (b) and (c) will be re-deposited, while another fraction will be entrained in a nozzle with less spray effect in the ablation plasma plume, contributing to the following phenomena and processes, including elastic collisions and return on the target or deposition on the walls of the vacuum chamber, but also moving with the plasma plume to the deposition support and becoming seeds for a new nucleation process in droplets in contact with the resulting liquid phase during plasma/gas condensation, before or in the support quarter. The complex phenomena of droplet generation in the PLD process continue with the instabilities developed in coexisting fluids associated with the plasma plume and its movement from the target to the support, also acquiring the effects of the instabilities already presented from the target in its immediate vicinity, in the first nanoseconds after each pulsed laser ignition of the ablation.
When ablation plasma starts cooling while getting closer to the substrate, liquid streams are generated in a complex condensation process, from mixed gas and liquid state clouds to crystalline growth (
Figure 8 c2,c3). The droplets are grown on crystalline seeds that originate into the primary droplets developed under RTI on the target during ablation in a nucleation process, in the first nanoseconds. The primary droplets are trained into plasma motion, cooling in time up to the solid state, as monocrystals.
(d) In the deposition chamber, the plasma travels from target to substrate on a path of 2 mm or 6 mm with a very high speed, of 6.3·10
3 m/s. At such value of speed, the plasma pushes the buffer air gases in its front and produces a pressure wave, which remains behind plasma. Plasma plume is a fluid phase that will allow the pressure wave (shock wave) to propagate through its mass leading to a harmonic constrain in plasma (
Figure 8 d1 – d3) leading to RMI [
35], followed by condensation into liquid streams and/or droplets (
Figure 8 d4). Also, the very high speed of the plasma will develop a very high temperature, based on the Mach’s number diagram with stagnation temperature [
10]. The temperature achieved in this way is favorable to plasma phase existence that will cool only when the plasma plume expands enough that arrives to a density value where the pressure wavelength can’t be produced anymore.
The droplets generated under the conditions described in the present paragraphs a), b) and c) will re-deposit and some will be entrained in the spray-less nozzle effect in the ablation plasma plume, contributing to the following phenomena and processes. These phenomena and processes include elastic collisions and return on the target or deposition on the walls of the vacuum chamber. The breaking of the fluid thread is believed to explain the formation of larger (main drops) and smaller (satellites) droplets observed in the SEM images in
Figure 7a. It means that, when the flow of molten metal breaks-in a main drop that will still be attached to the main stream for a period of time by a thinner mass of liquid, it will break-up later, but in drops of smaller sizes, called satellite drops.
The ring – shaped droplets from SEM image (
Figure 7a) are the result of crown splash. The mechanism consists, on the one side in “kinetic crown splash” based on the high kinetic energy of the droplets at the impact with the substrate, and on the other side in the existence of liquid state on the support from the previous layered particles and droplets and, in the latter case, the crown splash is the effect of an instability of RTI kind, or “RTI crown splash”. The “kinetic crown” splash generates rings disposed on the surface and the “RTI crown splash” generates craters in the first layer, which is still in melted state. Heated substrate could enhance the crown splash effect because the film is kept in melted phase on the support during deposition.
The Ag-thin film morphology and deeply embossed topography indicated by the SEM and 3D, 2D and 1D images obtained during AFM analysis indicates a suitable surface for catalytic processes to take place.
In the SEM images on the silver thin film, the uniform dispersion of the RB21 can be noticed (
Figure 9a). In
Figure 9a, the shape of the droplets can be distinguished, just like in the SEM image of the thin silver film in
Figure 7a, but the image is darker and matte because of the organic material that covers the thin silver layer. No brightness differences are observed on the studied surface, compared to the SEM image of the thin silver layer in
Figure 7a, where the droplets are brighter than the background. Higher magnitude of the image evidenced large crystalline structures (
Figure 9b). The crystal structure in
Figure 9b belongs to the components resulting from reaction between RB21 and NaHCO3 in the aqueous solution poured on the silver thin film.
In the FTIR spectra of
Figure 10, the high decomposition of the RB21 dyestuff (chemical structure shown in
Figure 11) in RB21 solution is observed on the Q-Ag and R-Ag samples. The changes of the functional groups are identified in the FTIR spectra (
Figure 10) using previous literature [
36,
37,
38,
39] compared to the initial structure of the RB21. The vibration bands assignments are presented in
Table 2.
Decomposition starts right after RB21 solution interaction with the silver thin film, proved by the changes in Q-Ag and R-Ag spectra. Carbonate () groups are evidenced by the 1748 cm-1 and 1462 cm-1 bands; the latter also indicates the lattice modes. The crystalline structure of the Q-Ag resulted compound, denoted by the ascending spectrum baseline, is also assigned to the carbonates as Mie scattering effect. Sulfonic (SO3) groups are still intact (1390 cm-1), while the aromatic rings are still present in R-Ag (3075 cm-1), the same in the control test RB21-NaHCO3 (3014 cm-1). Aliphatic C-H is denoted in all samples by the bands at 2930 cm-1, 2855 cm-1 (RB21); 2919 cm-1, 2849 cm-1 (RB21-NaHCO3); 2955 cm-1, 2882 cm-1 (R-Ag and Q-Ag). The heteroaromatic rings with nitrogen and N-H in sulfonamides groups were largely affected, noted by severe transmission attenuation of the 3438 cm-1 (initial dye RB21) and 3449 cm-1 (RB21-NaHCO3) vibration bands, becoming the very weak peak at 3483 cm-1 of Q-Ag dried leakage sample and missing in the R-Ag dried sample. Also, the skeleton vibrations at 1590 cm-1 assigned to pyrrole heteroatomic rings are missing in all treated samples and noticed only in the initial RB21 dye. The same band is assigned to N-H bending in sulfonamide group and its missing in RB21-NHCO3, Q-Ag and R-Ag confirming modifications. Since the vibration bands at 3438 cm-1 (initial dye RB21), 3449 cm-1 (RB21-NaHCO3), 3483 cm-1 (Q-Ag) and missing in the R-Ag also corresponds to the OH functional groups, it turns out that they were also modified compared to the initial molecule of the RB21 dye.
Based on the chemical structure of the RB21 dye (
Figure 11, the changes in the FTIR spectra of the studied samples (R-Ag and Q-Ag) in comparison with the reference samples (RB21 and RB21-NaHCO
3) indicate the following processes and reactions under silver thin film catalytic effect:
1. Sulfonyl group reaction with sodium bicarbonate, with increased rate under catalytic effect of Ag-thin film
The increased reaction rate is observed by the decrease in the intensity of the hydroxyl groups when comparing the FTIR spectra of the aqueous RB21-NaHCO3 solution, dried on a glass plate with the R-Ag (aqueous RB21-NaHCO3 solution dried on Ag thin film) and Q-Ag (leakage of aqueous RB21-NaHCO3 over thin-filmed Ag and dried on glass). In the case of the R-Ag sample, the vibration of the hydroxyl groups is missing from the spectrum. This proves that they have been completely transformed.
2. Deprotonation of the sulfonamide group (SO
2-NH) due to the “nucleophilic attack” of CO
2 by amino group, reaction that requires a catalyst (usually Bronstad catalyst) [
40].
3. Dehydration at the level of the terminal hydroxyl group in the side chain led to the formation of the vinyl group. Usually, the reaction of sodium bicarbonate with alcoholic hydroxyl groups does not occur, because the first is a weak acid in this case (amphoteric character), while the second is a weak base. The lack of specific vibrations for the hydroxyl groups in the case of the R-Ag sample and their reduction in the case of the Q-Ag shows the damage of all OH groups in the dye molecule, resulting in the thin Ag film catalyzing the chemical process.
4. Also, the FTIR spectra indicate changes in the aromatic and heteroaromatic structures when the RB21 and NaHCO3 dye solution comes in contact with the thin silver film.
The silver layer strong roughness due to ring shaped particles is evidenced with profilometry measurements and the profiles are presented in
Figure 12a,b, while the reduced thickness after catalytic interaction is noticed in
Figure 12c. The density of the peaks is lower after RB21 solution was added and there are cavities between peaks, proving that the RB21 solution had an effect of “etching” on the thin layer. The effect shows that some of the silver thin film material was consumed. This etching effect could be caused by entrainment of submicrometer silver particles in the resulting leakage as a result of pouring the aqueous solution of RB21 and NaHCO
3 and/or binding of silver particles to modified and ionized functional groups such as sulfonate and carboxylate.
On the optical microscope image in
Figure 13a, a compact structure is observed, while in the image of
Figure 13b darker areas are noticed, probably due to clogging with reaction products during the interaction with the RB21 solution. The size of the observed area is 1.4 x 1.1 mm, as can be measured on the grid of 0.1 x 0.1 mm attached with the image of the thin film.