1. Introduction

When a high power, ultrashort laser pulse is fired into transparent media, it can generate self-guided filaments [1–5]. A filament stems from a dynamic balance between optical Kerr self-focusing and defocusing effect of the self-generated plasma and/or negative higher order Kerr terms [6–8].This self-guided filament can propagate over a long distance in air up to hundreds of meters [9] and survive even in clouds and atmospheric turbulence [10,11].The clamped intensity in the free propagating laser filament core can reach as high as 5 × 10¹³ Wcm⁻² in air [2–4,12].This intense self-guided filament can therefore ionize chemical molecules, generating electron density of 10¹⁵ – 10¹⁶ cm⁻³ and allow efficient photo-oxidative chemistry of the air molecules [2–4]. Consequently, self-guided filaments can find potential applications in fields such as remote sensing of chemical/biological agents [13,14], and lightning control [15], etc.

In recent years, femtoseconed laser-induced water condensation, generation of snow or ice, as well as formation of mist or cloud have been investigated extensively [16–22]. It was shown that self-guided filaments could induce water condensation in the atmosphere and snow formation in cloud chambers, which would provide a potentially alternative way to the traditional cloud seeding of dispersing small particles like AgI into the clouds. Concerning the mechanisms of water condensation in air, J.-P. Wolf et al. explained first that the binary H₂O-HNO₃ nucleation induced by photo-oxidative chemistry of nitrogen played a key role in the formation of condensation nuclei (CN) when 10-Hz near-infrared (800 nm) ultrashort pulses were used [16–18]. Due to photo-chemically activated processes, ozone and NO₂ molecules were generated with densities up to 10¹⁶ cm⁻³ and 3 × 10¹⁵ cm⁻³ in the filament active volume, respectively [18]. At the same time, HNO₃ was generated with a speculated concentration in the parts per million ranges, which increased the uptake of water in air and finally condensed out as water droplets [18]. The hygroscopic HNO₃ was proposed to be a major ingredient to the laser-induced condensation process [17,18]. Subsequently, the same group extended their work to measure the quantitative chemical composition and the size distribution of aerosols, evidencing the hygroscopic HNO₃ condensation in the form of hygroscopic NH₄NO₃ [23–25]. In addtion, the presence of a few ppb of trace gases like SO₂ and $\alpha$-pinene clearly enhanced the particle yield by number, the latter also by mass [23]. The CN stemmed from hygroscopic NH₄NO₃ and oxidized volatile organics can make the droplets grow at lower relative humidity (RH), and drastically speeds up their growth [24]. While for ultraviolet laser pulses, K. Yoshihara et al. proposed that the water condensation resulted from the photo-oxidation of volatile organic compounds (VOC) [19] and/or hydrogen peroxide (H₂O₂) formed by oxygen (O₂) photodissociation and ozone (O₃) [23]. VOC and/or H₂O₂ as a hygroscopic product could further capture water molecules and form water particles [19,26]. Additionally, our previous experimental and theoretical results showed that laser-filamentation-induced airflow motion around the filament played a significant role in assisting water condensation and precipitation although the generation of HNO₃ in the laser-induced snow was verified [21,22,27]. The violent airflow motion can increase the collision probability between the cloud condensation nuclei (CCN) and water vapor/other particles in a supersaturation environment, resulting in water condensation and precipitation with a higher efficiency.

Here in this paper, we experimentally studied laser-filamentation-induced water condensation, snow formation and airflow in a diffusion cloud chamber filled with air, argon and helium, respectively. The laser-induced airflow, the mass of snow and the energy deposition by the filament were compared for the three ambient gases. The condensation mechanisms in the three gases were also discussed. It was verified that a filament with higher energy deposition accompanied by higher plasma density is much more efficient to facilitate the laser-filamentation-induced condensation and precipitation processes.

2. Experimental setup

The experimental setup is shown in Fig. 1. The femtosecond laser was operated at 1-kHz, with pulse energies up to 8.2 mJ and pulse duration of 30 fs at 800 nm. The pulse energy stability was better than 0.5% rms within eight hours. The laser pulses were focused by a lens of 300-mm focal length, and then launched into a diffusion cloud chamber filled with different ambient gases to generate filaments. A continuous-wave (CW) 532-nm laser beam with 0.5 W was used to probe the airflow motion. The enlarged probe beam through lens 2 and 3 was truncated by slit with 40 mm (height) × 5 mm (width) and co-propagated with the femtosecond laser beam into the chamber. The cloud chamber with a size of 0.5 m (length) × 0.5 m (width) × 0.2 m (height) was made of 304 stainless steel, and its five surfaces except bottom were covered with 1-cm thick insulated foam to isolate the heat transfer from outside during the experiments. The vertical temperature gradient was maintained inside the chamber by using a refrigerating machine to cool the bottom base plate at a temperature of −46 °C, while the top plate of the chamber was kept at room temperature. The humidity inside the chamber was controlled by adjusting the electric current of a heating wire submerged in the water reservoir, which was set at 3 A in the experiments. A water reservoir was mounted at a height of 17 cm relative to the cold bottom base plate inside the chamber. The reservoir was in a 45 cm × 45 cm square frame with a cross-section of 5 cm × 2 cm in a downward-pointing triangle. It held distilled water pumped from a water tank above the chamber. The height of the laser axis relative to the bottom base plate of the cloud chamber was set at 10 mm, where the measured temperature and relative humidity (RH) were about −15 °C and 85%, respectively. One side window of the cloud chamber was used to observe the airflow motion around the filament by using a digital camera (Nikon D7000) to record the 90° sideway Mie scattering. The other side window of the cloud chamber was used to detect the filament fluorescence signals, which was imaged by two lenses onto the slit of a grating spectrometer (Shamrock 303i, Andor).

 

Fig. 1 Schematic experimental setup. Lens 1_, Lens 2 and Lens 3 are 300 mm, 50 mm and 300 mm focal length lens, respectively. The mirror is an 800 nm reflector with a high reflectivity at 45°.

Download Full Size | PDF

Femtosecond laser-filamentation-induced airflow, water condensation and snow formation were investigated respectively for different ambient gases including air, argon and helium. For the experiments in argon and helium, the sealed cloud chamber was firstly evacuated from atmospheric pressure down to ~20 Pa by a mechanical pump, then filled with pure argon or helium (≥ 99.999%) to atmospheric pressure. The pumping and gas-filling processes were repeated three times. Then as the experiments were carried out in air, the distilled water was pumped into the cloud chamber from the water tank, and the heating wire and refrigerating machine began to heat the water and cool the bottom base plate, respectively. After 20 minutes' cooling, femtosecond laser pulses were fired into the cloud chamber to generate femtosecond laser filaments.

3. Results

Figures 2(a), 2(b) and 2(c) show the fluorescence images of filaments captured from the side of the cloud chamber filled with air, argon and helium, respectively. It is found that the longest filament is generated in argon. When the filament is fired in the cloud chamber, vortex pairs with opposite rotation directions below the filament center were observed for air and argon, as shown in Figs. 3(a), 3(b), while vortex pairs could be observed both above and below the filament center in helium, as shown in Fig. 3(c). The temperature above the filament is higher than that below the filament in our cloud chamber, and hence the particles above the filament didn’t grow into larger-sized particles which would produce the intense Mie scattering. Therefore, two large vortices above the filament were not easily observed in air and argon [28,29]. However, in the case of helium, due to its relatively smaller mass compared with the mixed water vapor and residual air, the larger particles formed in the lower temperature zone below the filament might be raised easily with the upcurrent of helium and lead to the intense Mie scattering above the filament [29]. After 60-minutes' irradiation by femtosecond laser pulses, a heap of snow on the cold bottom plate formed below the laser filament center in the three cases, as shown in Figs. 4(a)-4(f). The snow obtained in different gases was carefully shoveled out and weighed, and then analyzed by an ion chromatograph (Dionex500). Table 1 shows the snow masses and $N{O}_3^{-}$ concentration obtained in the environment of the three gases, which are the average values of six independent experimental results. The masses and $N{O}_3^{-}$ concentration in the melted water are 51.2$\pm$18.0 mg / 573.6$\pm$301.0 ppm in air, 101.5$\pm$58.0 mg / 118.4$\pm$56.0 ppm in argon and 33.0$\pm$8.7 mg / 7.8$\pm$7.4 ppm in helium. Additionally, $N{H}_4^{-}$ was also analyzed in the melted water, but not detected due to the very low concentration. It is found that the snow mass produced in air is much lower than that in argon although the $N{O}_3^{-}$ concentration in the melted water in air is much higher than that in argon.

 

Fig. 2 (a)-(c) Side fluorescence images of filaments in pure air, argon and helium, which were captured by a digital camera (Nikon D7000: f number (F) = 5.6, light sensitivity (ISO) = 800, shutter speed (S) = 1/13 s). The dashed line showed the position of the geometric focus. The arrow line indicated the propagation direction of the laser pulses.

Download Full Size | PDF

 

Fig. 3 (a)-(c) Side Mie scattering images in air, argon and helium, which were captured by a digital camera (Nikon D7000: F = 5.6, ISO = 800, S = 1/13 s). The arrow lines indicate the propagation direction of the laser pulses. The dotted curves represent the rotating direction of vortices.

Download Full Size | PDF

 

Fig. 4 (a), (c) and (e) Captured images of snow piles respectively in air, argon and helium after 60-minutes' irradiation by the femtosecond laser (Nikon D7000 digital camera with a shooting oblique angle of 45°). The arrow lines indicate the propagation direction of the laser pulses. When blocking femtosecond laser and 532-nm green light, the corresponding captured images of snow piles (Nikon D7000 digital camera with a shooting angle of 90°) are shown in (b), (d) and (f), respectively.

Download Full Size | PDF

Table 1. Snow Masses and${\text{NO}}_{\text{3}}^{\text{-}}$ Concentration of the Melted Water in Air, Argon and Helium

View Table | View all tables in this article

The HNO₃ resulted from the residual nitrogen in the cloud chamber. To check and verify the residual nitrogen in the chamber when filled with argon or helium, a spectroscopic analysis was carried out by detecting the fluorescence of nitrogen molecules and ions near the filament in different gases. The corresponding side fluorescence spectra were measured in the wavelength range of 300–480 nm using a spectrometer, as shown in Figs. 5(a)-5(c), for pure and humid air, argon and helium, respectively. One can see that the obtained spectra (after subtracting background) consist mainly of spectra from molecular nitrogen and ionic nitrogen, lying between 300 nm and 480 nm [30,31]. No emission line from nitrogen or oxygen atoms was detected. The fluorescence intensity of certain neutral molecular nitrogen and ionic nitrogen lines, for example, at 337.1 nm (C³∏_u(0)-B³∏_g(0)), 357.7 nm (C³∏_u(0)-B³∏_g(1)), 380.5 nm (C³∏_u(0)-B³∏_g(2)) and 427.8 nm (B²∑⁺_u(0)-X²∑⁺_g(1)), is much stronger in air than those when the chamber is filled with helium, but weaker than those in argon. Only four kinds of weaker fluorescence emissions at 310.4 nm (C³∏_u(4)-B³∏_g(3)), 337.1 nm (C³∏_u(0)-B³∏_g(0)), 357.7 nm (C³∏_u(0)-B³∏_g(1)) and 427.8 nm (B²∑⁺_u(0)-X²∑⁺_g(1)) were detected in the chamber filled with helium, indicating very low content of residual nitrogen and hence $N{O}_3^{-}$. The obviously enhanced fluorescence signals of molecule nitrogen in argon case is attributed to electronic energy transfer between a metastable argon atom and a nitrogen molecule in their collisions, which leads to the population enhancement of some excited states of molecular nitrogen and ionic nitrogen [32,33]. This is also confirmed by the result that the concentration of $N{O}_3^{-}$ in the melted water for argon gas case is significantly more than that for helium gas under the same experimental conditions (Table 1). The collisions between the metastable argon atoms and nitrogen molecules probably provide more chemical reaction paths for nitrogen. Similar experiments were carried out in another chamber of higher vacuum degree filled with argon (without water reservoir in this chamber) and the obviously enhancement of fluorescence signals of typical nitrogen lines was also observed. An addition of water vapor resulted in the decrease of the corresponding fluorescence intensity in the whole spectral range, especially for humid and argon cases. This probably results from the decrease in the clamped intensity of filament with the addition of water vapor.

 

Fig. 5 Sideway fluorescence spectra of laser filaments in the range of 300-480 nm in air (a), argon (b), and helium (c), respectively. The identification of nitrogen and helium lines is from [34].

Download Full Size | PDF

4. Discussion and implications

In order to interpret the difference in the mass of laser-induced snow in a cloud chamber filled with different ambient gases, we analyzed the mechanisms of the snow formation in different gases.When a femtosecond laser pulse propagates in the humid gas, the filament forms, as well as generates high concentration of electrons, ions, etc. Lots of new compounds would form with the assistance of charged particles and photo-oxidation reactions. These compounds or their by-products would become the important source of cloud condensation nuclei (CCN). Most CCN would form initially around the filaments where charged particles with a high density are generated. Meanwhile, intense airflow motion would be generated after the firing of laser filaments into the cloud chamber because of heat deposition by the filament [22].The initial CCN would move with airflow motion and grow rapidly owing to the frequent collision induced by vortices as well as the created supersaturation environment [28,37]. Under the same ambient gas, a stronger laser filament would generate more ions, electrons, and molecules or other compounds around filament which would be the initial source of CCN. Meanwhile, a stronger laser filament would induce more heat release through plasma recombination [38], and thus stronger airflow motion would be generated around the filament, which accelerates the growing of CCN and induces more snow precipitation. More snow formation also increases the release of latent heat of condensation that would further accelerate the airflow motion [39].

The characters of filaments mainly depend on the focusing geometry [40,41], the laser power, and the second-order Kerr refractive index coefficients n₂ of the ambient gas. The filament with the laser intensity at the level of 10¹³-10¹⁴ Wcm⁻² is able to ionize the atoms and molecules via a multiphoton process, by which a few percent of the laser energy is absorbed to generate charged particles such as electrons, ions and charged molecules. For a mature laser filament, the clamped intensity I in the laser filament can be estimated by the equation: n₂ × I ≈N_e (I) /2 N_crit [3,42,43], where n₂ (shown in Table 2) is the Kerr nonlinear index of refraction of gas, N_e (I) is the electron density from multiphoton/tunnel ionization, N_e (I) from multiphoton ionization can be estimated roughly by the following relation: N_e (I) ~[(0.76 n₂N_crit)^κ/σ_κt_p N₀]^1/(κ-1) [3], where σ_κ is the cross section for multiphoton ionization, N₀ is neutral atom density, κ< I_p / $\hslash {\omega }_0$ > + 1 denotes the number of photons at the frequency ${\omega }_0$ necessary to liberate an electron, t_p is the laser pulse duration, and N_crit = ε₀mω² / e² is the critical plasma density, with m being the mass of electron, ω the laser angular frequency, and e the elementary charge. As can be seen from Table 2, n₂ is the largest in air, the second in argon and the smallest in helium. Thus the clamped intensity in the laser filament should be the highest in helium, the second in argon and the lowest in air [42–45]. Under our experimental conditions, the input power of the laser pulse is ~264.5 GW, which is orders of magnitude higher than the critical power in air and argon, while close to the P_cr in helium (~282.6 GW). It indicates that mature filaments with clamped intensity can be obtained for air and argon, but not for helium. For helium gas, the geometrical focusing and the plasma-induced defocusing processes should be dominant owing to the low input power. In order to reach a balance between the focusing and plasma-induced defocusing effects, the higher ionization potential in helium would result in the smaller diameter of focusing filament. Therefore, for the same input laser energy, the intensity of laser filament in helium will be much higher than that in argon or air (~5 × 10¹³ Wcm⁻², 1.710¹⁴ Wcm⁻² [38] and 6.810¹⁶ Wcm⁻² in air, argon and helium, respectively). However, owing to much higher the residual nitrogen content in air than that in helium, and hence the intensity of fluorescence signals of nitrogen in helium was only a little smaller than those in air. We measured the input and output power before and after the filament using a laser power meter (PM 30, Coherent Inc), and obtained the energy deposited into the filament in air, argon and helium (shown in Table 2), respectively. The energy absorption efficiency η by the filament in air (18.7%) is significantly higher than that in helium (2.0%), but less than that in argon (22.7%). Additionally, due to very high ionization potential of helium, it is difficult to ionize helium atoms and so the plasma density of laser filament in helium should the lowest in the three gases although the intensity of filament in helium is highest [43–45]. Therefore, the plasma density inside the filament in argon is higher than that in air because it is mainly determined by the clamped intensity of filament considering the clamped intensity of argon is larger than that of air while the ionization potentials of argon and air are almost the same [42–44]. We also estimated the velocities of airflow by dividing the moving distance of some visible particles in the side scattering video by an exposure time of 1/25 s between two adjacent frames. The velocity of airflow is 4.41$\pm$0.74 cm/s in air, 11.42$\pm$0.72 cm/s in argon and 3.34$\pm$0.23 cm/s in helium, respectively. These experimental results confirmed further our analysis, that is, higher plasma density owing to more laser energy deposition inside the filament induces more heat release. This corresponds to more initial CCN, and hence stronger airflow (airflow with the higher velocity) would be generated. The stronger airflow results in the faster growing of CCN, consequently, the mass of snowfall increases. This can explain why the highest snow amount is obtained in argon, the second in air, and the least in helium.

Table 2. Ionization Potentials I _p, Second-order Kerr Refractive Index Coefficients n₂, the Critical Power P_cr, the Number of Required Photons κ for Ionization, the Cross Section for Multiphoton Ionization σ_κ and the Percentage of Energy Deposited into the Filament η Measured for Air, Argon and Helium, respectively at 1 atm and 800 nm.

View Table | View all tables in this article

5. Conclusion

In conclusion, we experimentally investigated femtosecond laser-filamentation-induced airflow, water condensation and snow formation in air, argon and helium ambient gases, respectively. The snow masses are 51.2$\pm$18.0 mg in air, 101.5$\pm$58.0 mg in argon and 33.0$\pm$8.7 mg in helium. The energy absorption efficiency by the filament in air is significantly higher than that in helium, but less than that in argon. We analyzed the condensation mechanisms in the three gases and found a filament with higher energy deposition followed by higher plasma density would generate stronger airflow in a supersaturation environment and thus more water condensation and snow precipitation.