We report on the experimental investigation of femtosecond laser filament guided negative coronas. When the coupling between the filament and negative corona was weak, the side fluorescence spectral analysis confirmed the existence of impact ionization although less effect on the filament length was observed. When the coupling was strong so that the negative corona was well connected with the filament, the filament guided coronas at the ends of laser filaments were observed. The newly generated negative coronas were confined around the filament axis, and no streamer-type of coronas guided by the filament was observed under conditions similar to those reported in the work of Wang et al. [Sci. Rep. 5, 18681 (2015)] although both could give rise to an elongation of the filament. A physical picture was proposed to understand the processes of the laser filament guided coronas.
Femtosecond laser filament is regarded as a plasma channel along the femtosecond laser propagation axis. The plasma is produced by multiphoton/tunnel ionization of air molecules.1,2 The peak density of electrons in the filaments is of the order of 1015–1018 cm−3 depending on the focusing conditions.3 The plasma filament generated by acoustic shockwaves results in the formation of a millisecond long low density zone along it.2 Hence, plasma filament has found many potential applications including artificially triggered lightning,2,4 electric power transmission,5 and plasma antenna.6
Toward the applications of using laser filament triggered/guided discharges, the understanding of laser filament interaction with a large electric field or coronas is important. In 2009, the interaction between filaments and the positive/negative electric field was reported using a spherical electrode. The electrode was designed to prevent the streamers from occurring when there was no laser filament passing nearby it.7 Then, the stratified UV emission near the filament for the negative polarity and the leader formation with hot, dense plasma for the positive polarity were observed.8 When propagating a laser filament parallel to the positive arc discharge in air, Schubert et al.9,10 reported that the high voltage discharges could undergo spark-free unloading, discharge suppression, and discharge guiding when the electric field is well below, close to, and above the threshold for discharges, respectively. This effect only appeared at a high laser repetition rate (1 kHz) and has never been observed in single shot experiments because it is due to a cumulative effect.9,10 In 2015, the phenomenon of laser filament guided positive corona discharges was clearly observed and systematically studied in our laboratory.11 Both leader-type and streamer-type of positive coronas could be generated along the laser filaments. The lifetime of the high voltage electric field biased plasma channel was increased by three orders of magnitude into the microsecond regime.11 Through the interaction of the laser filament with high voltage discharge, the lifetime was further increased up to milliseconds by other groups later.5,12
In this work, we report on the experimental investigation on the femtosecond laser filament guided negative coronas. The results show that the ionization along the filament could be enhanced with a negative corona electric field applied to the filament. Leader-type negative coronas were guided and generated at both ends of the filament when the negative corona is strong enough. No streamer-type of coronas guided by the laser filament was observed for negative electric field polarity as compared with the positive one under similar conditions.
The experimental setup is schematically shown in Fig. 1. The ultrashort laser pulse (800 nm/1 kHz/31 fs) with an initial diameter of about 16 mm (FWHM) was focused by a plano–convex lens with a focal length of 30 cm to produce air laser filaments. The air relative humidity in the lab was ∼40%. A copper cylindrical electrode with a 1 mm diameter sharp tip was connected to a DC positive or negative high-voltage power supply with a high voltage varied between 0 and +100 kV or −100 kV. In this experiment, the electrode’s axis and the laser beam were in the same horizontal plane. The distance between the filament and the electrode’s tip was about 1 mm. All the discharge experiments were performed in a home-made Faraday cage (0.5 × 0.5 × 0.4 m3, dashed-dotted line in Fig. 1) that was safely grounded. A digital camera (Nikon D7200) was used to capture the real-color images of the interaction of the laser filament with negative corona discharges from the top. Fluorescence emission produced by the laser filaments with or without negative high-voltage corona discharge was imaged from the side by a pair of lenses to a spectrometer (Princeton Instruments, SP-2560) for spectral analysis.
Schematic of the experimental setup (top view). Lens 1–Lens 3 are UV grade fused silica lenses with focal lengths of 30 cm, 8 cm, and 10 cm, respectively.
Schematic of the experimental setup (top view). Lens 1–Lens 3 are UV grade fused silica lenses with focal lengths of 30 cm, 8 cm, and 10 cm, respectively.
Figure 2 shows the real-color images of the experiment results. The negative streamer corona discharge around the tip of the copper electrode was tightly confined in a very small area around the tip [Fig. 2(a)], which was significantly different from the positive high voltage streamer corona [Fig. 2(d)]11,13 under similar conditions. The lengths of negative and positive coronas (electron path length) in Figs. 2(a) and 2(d) are ∼1 mm and ∼10 mm, respectively, which were calculated from the lengths of plasma fluorescence in the imaging figures. These lengths are comparable and even larger than the tip diameter of the electrode (∼1 mm). Figures 2(b) and 2(e) show the laser filament only. When the negative corona interacts [Fig. 2(a)] with the laser filament [Fig. 2(b)], the resultant image is shown in Fig. 2(c). One can see that the two are now connected. As a comparison, bluish-purple fluorescence at the end of laser filaments was observed at both off-axis and on-axis of laser propagation, and the length of laser filaments is extended as shown in Fig. 2(f) when the laser filament interacts with the positive corona under similar conditions.14
Real-color images of the experimental results: negative corona discharge from high voltage electrode (a), laser filament (b), and the interaction between the laser filament and negative corona (c). The exposure time of the digital camera was set to 0.77 s. The negative high voltage and laser pulse energy were set to −50 kV and 5.22 mJ, respectively; the real-color images of the positive high-voltage corona discharge (d), laser filament (e), and the interaction between the laser filament and positive corona (f). The exposure time of the digital camera was 0.5 s. The positive high voltage and laser pulse energy were set to 50 kV and 5.45 mJ, respectively.
Real-color images of the experimental results: negative corona discharge from high voltage electrode (a), laser filament (b), and the interaction between the laser filament and negative corona (c). The exposure time of the digital camera was set to 0.77 s. The negative high voltage and laser pulse energy were set to −50 kV and 5.22 mJ, respectively; the real-color images of the positive high-voltage corona discharge (d), laser filament (e), and the interaction between the laser filament and positive corona (f). The exposure time of the digital camera was 0.5 s. The positive high voltage and laser pulse energy were set to 50 kV and 5.45 mJ, respectively.
Typical fluorescence spectra of filament–negative corona coupling (FIL–CD) detected from the side under the weak coupling condition are shown in Fig. 3(a) together with the fluorescence from the laser filament (FIL) and high voltage negative corona discharge (CD). The high voltage and the laser energy used in the measurements were −60 kV and 6.22 mJ, respectively. Compared with the spectra from CD, the intensities of spectral lines from neutral N2, for instance, 337 nm (N2: 0–0), 358 nm (N2: 0–1), 380 nm (N2: 0–2), 406 nm (N2: 0–3), etc., from the second positive band system of N2 (C3Πu → B3Πg transition)15 in FIL–CD, were increased significantly, which were larger than the sum of CD and FIL by ∼21.2%, 19.1%, 21%, and 19.2%, respectively. The numbers in parentheses (v–v′) indicate the vibrational levels of the upper and lower electronic energy states. The intensities of the spectral lines of , for instance, 391 nm (: 0–0), 428 nm (: 0–1), etc., from FIL and CD were the same as those of FIL. In order to look into the interaction process, the fluorescence intensities of FIL and CD were subtracted from those of FIL–CD. The resultant spectrum is shown in Fig. 3(b). The spectral lines from N2 are clearly observed. These spectra in Fig. 3(b) are due to the enhancement of the interaction process,16 which is similar to the fluorescence spectrum of the negative corona discharge process. The spectral lines from are not observed or negligibly weak. This proves that the spectral lines in Fig. 3(b) are those of neutral N2 enhanced by the combined effect of negative corona and laser filamentation. Fixing the laser pulse energy at 6.22 mJ, the applied voltage dependence of the side neutral and ionic nitrogen fluorescence spectral intensity is depicted in Fig. 3(c). The spectral intensities from neutral N2 increased with the increase in the applied voltage. However, the spectral intensities of 391 nm and 428 nm lines from stay practically constant. Since the emission of N2 fluorescence is due to the electron collision process,15 thus, the fluorescence spectral analysis indicates the enhancement of electron impact excitation during the weak interaction of the filament with negative corona.
(a) The side fluorescence spectrum in the range of 330–430 nm emitted by filament (FIL), traditional negative corona discharge (CD), and filament-negative corona discharge weak coupling (FIL–CD), respectively; (b) the subtracted fluorescence spectrum of the FIL–CD by the sum of CD and FIL; and (c) the intensities of the fluorescence spectral lines at 337 nm, 358 nm, 380 nm, 391 nm, and 428 nm as a function of applying voltage U. The high voltage and laser energy used in the measurements were −60 kV and 6.22 mJ, respectively.
(a) The side fluorescence spectrum in the range of 330–430 nm emitted by filament (FIL), traditional negative corona discharge (CD), and filament-negative corona discharge weak coupling (FIL–CD), respectively; (b) the subtracted fluorescence spectrum of the FIL–CD by the sum of CD and FIL; and (c) the intensities of the fluorescence spectral lines at 337 nm, 358 nm, 380 nm, 391 nm, and 428 nm as a function of applying voltage U. The high voltage and laser energy used in the measurements were −60 kV and 6.22 mJ, respectively.
In order to increase the electric field strength around the tip of the electrode to enhance the interaction between the filament and the negative corona, an electrically grounded metal plate (0.2 × 0.2 m2) was vertically inserted at a distance of 75 mm from laser filaments, as shown in Fig. 1 (dashed line). Although the electric field strength at the location of the laser filament is increased by only ∼1.8 times compared with that without the plate under the same applied voltage (calculated by ,17 where V is the applied voltage, S is the needle-to-plane distance, l is the distance from the needle tip to the axis of the laser plasma channel, and r is the radius of the needle tip), the negative corona zone is significantly enlarged because of the nonlinear avalanche ionization.
As shown in Fig. 4(b), the negative corona at −60 kV from the copper electrode is clearly enhanced by inserting the metal plate as compared with the negative corona without the plate [Fig. 4(a)]. The plasma zone of the corona was significantly increased. When the laser filament [Fig. 4(c)] with a pulse energy of 6.22 mJ couples with weak [Fig. 4(a)] and strong [Fig. 4(b)] coronas individually, the resultant images are shown in Figs. 4(d) and 4(e), respectively. The fluorescence from both ends of the laser filament was slightly enhanced in Fig. 4(d). A clear extension of the fluorescing channel at both ends was observed, as shown in Fig. 4(e). Bluish-purple fluorescence at the extension region observed is an indicator of leader-type coronas. More interestingly, the negative coronas at the ends of the laser filaments are confined around the filament axis and no streamer-type of coronas guided by the filament were observed under similar conditions, as shown in Fig. 2(f).11,14
Real-color images of the experimental results: negative corona discharge from a high voltage electrode without (a) and with (b) the metal plate inserted, laser filament (c), the interactions of (d) and (e) between the laser filament (c) and negative coronas (a) and (b), respectively. The negative high voltage and laser pulse energy were −60 kV and 6.22 mJ, respectively.
Real-color images of the experimental results: negative corona discharge from a high voltage electrode without (a) and with (b) the metal plate inserted, laser filament (c), the interactions of (d) and (e) between the laser filament (c) and negative coronas (a) and (b), respectively. The negative high voltage and laser pulse energy were −60 kV and 6.22 mJ, respectively.
Corona streamer generation in a needle-to-plane geometry for different polarities (positive and negative) has been well studied.18–20 Normally, there are some free electrons, namely, seed electrons, in air created from some events in the natural environment. A high-voltage sharp electrode provides a strong electric field to accelerate the seed electrons near the electrode. When these seed electrons are sufficiently accelerated over a free distance (on average, over the mean free path), they will collide with and ionize a neutral molecule (impact ionization). The secondary electrons from impact ionization are again accelerated and induce further ionization through impact and so on. This results in a chain reaction or avalanche ionization. Both positive and negative coronas rely upon electron avalanche. In a positive corona, the seed electrons at a certain distance away from the positively charged electrode are accelerated along field lines toward the strong field region around the tip of the electrode, resulting in strong impact ionization–avalanche ionization. The resultant electrons will mostly penetrate into the electrode contributing to completing the electrical circuit. The positively charged ions are pushed (accelerated) away from the electrode along the field lines and form a positive charge layer [Fig. 2(d)]. The seed electrons outside the positive ions’ layer will be accelerated by the electric fields of both positive electrode and positive ions’ layer, leading to a larger positive corona. In a negative corona, initially, the seed electrons are repelled (accelerated) by the negative high voltage of the sharp electrode and gain energy leading to avalanche ionization near the cathode with the positive ions’ layer left. The electric field from the negative electrode will be partially canceled by the electric field of the positive charge layer. As a consequence, the seed electrons outside cannot be further accelerated as strong as in the case of the positive corona because of the space charge (or shielding) effect of positive ions, resulting in a smaller negative corona18–20 [Fig. 2(a)].
When a femtosecond laser filament is “injected” near and perpendicular to the axis of a sharp electrode tip at a high negative DC voltage [Figs. 2(e) and 4(e)], a few things happen. Essentially, the filament has deposited a weak plasma channel (plasma density ∼ 1015−18/cm3,3 which can be seen as a source of seeding electrons for avalanche ionization) near the tip of the electrode almost instantaneously. Then, the filament zone heats up and expands, resulting in a low density zone.21,22 This low density zone will facilitate electron impact ionization because the mean free path of an electron is increased. The electrons in the plasma channel will be accelerated toward the ends of the filament and induce avalanche ionization along the way, resulting in the enhancement of the plasma density and the N2 fluorescence intensity (Fig. 3). The mechanism of producing the fluorescence inside the filament zone and in the corona’s streamer is twofold. One is through electron impact excitation of the neutrals when the electrons are accelerated away from the electrode and undergo collisions with the neutral: e + N2 → e + N2(C3Πu), N2(C3Πu) → N2(B3Πg) + hν. The other is through the following recombination mechanism in the plasma:23
When a femtosecond laser filament is “inserted” almost instantaneously into the negative corona zone in the weak coupling case, the electric field in the thin layer of the strong field around the tip of the electrode would only be partially coupled to the filament. Although there will be similar interaction as with the positive corona giving rise to ionization and radiation, the field is not sufficient to extend the filament length as in the case of the positive corona, as observed in Fig. 2(c). However, the increase in neutral N2 fluorescence is an evidence of extra impact ionization (Fig. 3). After inserting a grounded plate, the negative corona zone is expanded further [Fig. 4(b)]. Thus, a good coupling between the negative corona and the filament is established [Fig. 4(e)]. At the two ends of the filament, further avalanche ionization and hence more fluorescence reveal the “invisible” filament tails where there was weak ionization initially.
However, unlike the positive corona where there are streamers radiating out from the two ends and at some irregular points, the extended filament coupled to a negative corona is rather smooth in appearance under similar conditions [Fig. 4(e)]. The basic physics behind the acceleration of electrons by the electric field and the shielding from positive ions for the filament guided coronas is the same as that for the coronas alone. It is the shielding effect from positive ions leading to the weaker acceleration of the electrons along filaments with negative corona. As a consequence, weaker impact ionization along the filament occurs, which explains the weaker corona fluorescence observed in the case of the laser guided negative corona.
In conclusion, driven by the electric field of the negative corona, free electrons inside the filament could be accelerated toward the two ends of the laser filament, resulting in significant impact ionization. When the electric field of the negative corona is high enough (with a grounded plate inserted), impact ionization induced negative corona could even take place at the weakly ionized region (both ends of the filament). The negative coronas guided by the laser filament are confined along the filament axis, and no streamer-type of coronas guided by the filament was observed. A physical picture was proposed to understand the interaction difference of the laser filament with positive and negative coronas.
This work was supported, in part, by the Strategic Priority Research Program of the Chinese Academy of Sciences (Grant No. XDB16000000) and the International Partnership Program of Chinese Academy of Sciences (Grant No. 181231KYSB20160045), China. S.L.C. acknowledges the support from Laval University in Canada.