1. Introduction

Cavity-free lasing phenomena of air (including nitrogen, oxygen, argon) by pumping with intense ultrafast laser pulses have attracted many attentions since 2011 [1–11]. These effects are promising to create a virtual laser source in the sky and the coherent emission holds unique potential to improve the detection sensitivity of optical remote sensing and opens new route for remote spectroscopy [1,12–15]. In the traditional schemes of optical remote sensing, the spontaneous fluorescence signal or the elastic/inelastic scattering photons are detected by the ground observer. These signals are emitted to the entire solid angle of 4$\mathrm{\pi }$ and the optical detection is usually incoherent. With a backward coherent optical beam from the sky toward the ground, coherent nonlinear optical techniques such as Stimulated Raman Gain (SRG) can be employed for detection of targeted trace gas or pollutants in the atmosphere, which is expected to improve the detection sensitivity significantly [12,13].

Up to now, various air lasing schemes based on nitrogen, oxygen and argon gases have been demonstrated [1–10]. Among the different schemes, backward lasing emission in atmosphere condition is only achieved by the multiple-photon excited atomic oxygen, atomic nitrogen, and argon [1–2,16,17]. For the achievement of lasing of atomic oxygen or nitrogen, laser induced dissociation of the molecules into atoms is the first essential step and it poses a laser intensity threshold on the UV pump laser [1]. For the 100 ps or nanosecond pump pulses, this threshold effect corresponds to a pulse energy of millijoule, which is not commonly available for UV laser pulses. A. Laurain et al. demonstrated that the laser intensity threshold for the UV pump laser can be almost eliminated by pre-dissociation of the nitrogen or oxygen molecules with a nanosecond infrared pulse [2]. This necessitates the employment of a high-energy IR laser pulse and the synchronization of the IR and UV laser pulses, which complexifies the optical system for achievement of backward lasing emission.

Argon is a monoatomic gas and does not require pre-dissociation, which exists in the atmosphere with a considerable concentration of ∼ 0.8%. A. Dogariu and coworkers employed UV pump pulses at 262 nm as a pump pulse to activate a three-photon resonant excitation of argon atoms [9,16,17]. In the inset of Fig. 1(a), the relevant energy levels of the argon atom are presented. The argon atoms are pumped from the ground state 3s²3p⁶ to the upper lasing state 3s²3p⁵(2P°_1/2)3d through a three-photon resonance excitation. Optical transition between the upper excited 3s²3p⁵(2P°_1/2)3d state and the middle 3s²3p⁵(2P°_1/2)4p state produces radiation at 1327 nm. They observed a coherent emission in both the backward and forward directions at 1327 nm. In these experiments, the energy of the pump laser pulse was just tens of microjoule since no molecular dissociation is involved. Later, with a higher power UV pump laser, coherent 1327 nm emission in the backward direction was observed using ambient air as gain media [16,17], opening the route for the achievement of backward lasing in atmospheric air. However, up to now the nature of this 1327 nm emission is still largely unknown. Moreover, it is still not clear whether an optical gain exists in this UV pump excited argon gas, since only one single pump laser was employed in the previous experiments [9,16,17]. To verify the presence of optical gain or optical amplification, an externally injected seeding pulse at the proper wavelength is necessary.

 

Fig. 1. (a) Schematic diagram of the experimental setup. Inset: relevant energy level structure of argon atom excited by ultraviolet 262 nm pulse through three-photon resonance. (b) The spectrum of the pump laser pulses around 1520, 796, 522, and 262 nm.

Download Full Size | PDF

In this work, we performed a pump-probe experiment of the Ar lasing process by injecting a weak 1330 nm seeding pulse following the UV pump pulses. An obvious energy enhancement of the 1327 nm signal was observed in presence of the seeding pulses, confirming the existence of optical gain and population inversion. With time-resolved measurement, we further characterized the temporal dynamics of the optical gain. It was found to last for ∼ 10 ps and no significant dependence on gas pressure was observed. This suggests that cooperativity behavior does not dominate the lasing process and the nature of this radiation should be identified as amplified spontaneous emission (ASE), not a collaborative radiation process such as superradiance or superfluorescence.

2. Experimental setup

A commercial Ti: Sapphire femtosecond laser system (Coherent Legend DUO) and an Optical Parametric Amplifier (OPA, HE-TOPAS) were used in the experiments. The femtosecond laser system delivers 35 fs pulses at a central wavelength of 796 nm, at a repetition rate of 1 kHz. The output femtosecond laser pulses pump the OPA system to generate tunable mid-infrared laser pulses, with the tuning range from 1100 to 2600 nm. The output signal pulse of the OPA around 1520 nm and the residual 800 nm pulses exit from the OPA in a collinear optical path. The experimental setup is shown in Fig. 1(a). We installed a type I beta barium borate (BBO) crystal with a cutting angle of 44.7° into the optical beam to generate laser pulses at 522 nm through sum-frequency generation (SFG) of the 1520 nm and the 800 nm pulses. The spectrum of the 522 nm pulse is presented in Fig. 1(b). With optimization of the angle of the BBO crystal, the maximum energy of the 522 nm pulse was measured to be about 500 µJ. We used a long pass dichroic mirror (DM 1) to separate the 522 nm pulse from the 800 nm and 1520 nm pump pulses. We found that the OPA system delivers several parasitic spectral components in the NIR range in addition to the signal and idler beams. Therefore, a seeding pulse with a central wavelength of 1330 nm was readily obtained from the parasitic radiation by the installment of a narrow band interference filter at 1330 nm. A mechanical delay line was installed on the optical path of the seed pulse. Consequently, the time delay between the 522 nm pulse and the seed pulse around 1330 nm can be adjusted with an accuracy of 1 µm, corresponding to a delay of 3 fs. The 522 nm pulses and the 1327 nm seed pulses were combined by a short pass dichroic mirror (DM 2). The beam was focused by a fused silica lens with a focal length of 30 cm into a gas chamber filled with argon gas at different pressures. Another 10×10×1 mm BBO crystal with a cutting angle of 22.9° is installed after the focusing lens to generate the required 262 nm ultraviolet pulses from the 522 nm pulses based on second harmonic generation. The spectrum of the 262 nm pulse is also presented in Fig. 1(b). The optimal pulse energy of the 262 nm pump pulse was measured to be ∼10 µJ. The forward optical signal emitted from the argon gas is collected with a quartz lens of f = 10 cm into an optical fiber connected to an infrared spectrometer (Ideaoptics NIR17 + Px). In front of the optical fiber, a bandpass filter at 1330 nm was installed to filter out the background light.

3. Results and discussion

In Fig. 2(a), the spectrum of the forward lasing signal from argon pumped with only the 262 nm pump pulses is shown. We have measured the dependence of this signal intensity as a function of the gas pressure and the results are shown in Fig. 2(b). The lasing signal is found to be linearly proportional to the argon gas pressure. This linear dependence is different from other air lasing systems such as $N_2^ + $. In the case of $N_2^ + $, an optimal nitrogen pressure around 20-100 mbar was usually observed [18–20], which was largely attributed to the pressure-dependent laser intensity since the pump pulse propagation is highly nonlinear and the dense nitrogen plasma plays an important defocusing role [18]. The linear dependence in this study therefore suggests that the pump laser intensity at the zone of focus is independent on the gas pressure, which can be expected since the pump pulse energy was just 10 µJ and no obvious ionization of Ar atoms occurs. This agrees with the fact that no visible plasma to the eyes was produced in the experiments. As a result, with the increase of the gas pressure, the density of the excited Ar atoms grows proportionally. We have measured the polarization state of the 1327 nm signal and the results are presented in Fig. 3. For a linearly polarized pump pulse in the horizontal plane, the forward 1327 nm signal was also found to be linearly polarized in the horizontal plane, which suggests that the forward emission is stimulated emission, instead of fluorescence.

 

Fig. 2. (a) Spectrum of the forward coherent emission around 1327 nm from argon atoms pumped with only the 262 nm pulses. (b) The dependence of the intensity of the forward 1327 nm lasing signal on the pressure of argon gas.

Download Full Size | PDF

 

Fig. 3. Polarization measurement of the 262 nm pump laser and the 1327 nm emission. (a) Polarization of pump pulse. (b), Polarization of the forward 1327 nm emission without seeding pulse.

Download Full Size | PDF

To verify the presence of optical gain inside the laser pumped argon gas, we performed pump-probe experiments with an externally injected seeding pulse around 1330 nm. In Fig. 4(a), we present the results in the spectral domain obtained at a gas pressure of 1 bar. As shown in Fig. 4(a), the weak 1330 nm seed pulse leads to an amplified forward signal compared to that obtained with only the 262 nm pump pulse. Considering the intensity of the seeding pulse at 1327 nm, the ratio of amplification can be estimated to be $\textrm{}{I_f}/{I_0}$ = 7, with I₀ and I_f the intensity of the seeding pulse at 1327 nm and that of the amplified emission. If we take the Rayleigh length of the pump beam as the effective gain length, then the optical gain can be estimated to be $\textrm{g} = \textrm{ln}({{I_f}/{I_0}} )/l = 5.82\textrm{}c{m^{ - 1}}$. The Rayleigh length l of the pump laser at 262 nm has been estimated to be 3.34 mm, by considering the incident 262 nm beam waist ${\mathrm{\omega }_{in}}$ = 3 mm and its beam quality number M² ∼ 2. In Fig. 4(b) and Fig. 4(c), the spatial profile of the 1327 nm emission without/with the seeding pulse are presented. To adapt the transverse size of the 1327 nm beam to the size of the charge coupled device (CCD) chip (bob-cat320, Xenics) used in the experiments, we employed a 10 cm convex lens to reduce the beam size. The injection of the seeding pulse leads to a significant increase in the radiation intensity when Fig. 4(c) is compared to Fig. 4(b). In our experiments, the seeding pulse is too weak and cannot be observed with the CCD. In both cases with and without seeding pulse, the 1327 nm radiations show a ring-shaped spatial pattern, which has been observed in the lasing of atomic oxygen and nitrogen [1,17,18].

 

Fig. 4. Injection of the 1330 nm seeding pulse leads to amplification of radiation intensity. The seeding pulse itself (black line), the signal produced with only the 262 nm pump pulse (red line), and the amplified signal (blue line) are presented. Measured beam profile of the forward emission at 1327 nm without the external seeding pulse (b) and with the external seeding pulse (c). The argon gas pressure was 1 bar.

Download Full Size | PDF

To obtain insight into the temporal evolution of the optical gain, we then performed time-resolved measurement. In this experiment, the time delay τ between the 262 nm pump and the 1330 nm seeding pulses was varied and the amplified signal intensity was recorded as a function of delay τ. The results for three different gas pressures are presented in Fig. 5. Here the positive delay corresponds to the situation where the 262 nm pump precedes the seeding pulse. When the seed and pump pulses overlap in the time domain, the signal intensity experiences a sudden increase, corresponding to optical amplification of the seeding pulses due to optical gain. With a further increment of the delay, the amplified signal shows a gradual decrease up to ∼ 10 ps, indicating the temporal decay of population inversion. The decay processes in Fig. 5(a)–(c) are almost identical, showing no significant dependence on gas pressure. This is different from the situation of superradiance in the case of $N_2^ + $, where the temporal decay of the optical gain grows much faster for higher gas pressure [20,21]. Therefore, we conclude that the cooperative radiation effect, which is essential for superradiance in the case of $N_2^ + $[10,20–25], does not play a significant role in the lasing of argon under our current experiments. In the cases of lasing of atomic species including O, N, Ar and Kr, resonant two-photon or three-photon excitation is responsible for the creation of population inversion [1,9,16,17].

 

Fig. 5. The amplified 1327 nm signal intensity is presented as a function of the temporal delay between the pump and seeding pulse at three different gas pressures. The pressure of argon gas was 1200, 920, and 720 mbar in (a), (b), and (c) respectively. In the three cases, the intensity of the 1330 nm seed pulse is 66 (arb.u.), while the intensity of the 1327 nm signal produced by only the pump pulse is 58, 42, 34 (arb.u.).

Download Full Size | PDF

While the 845 nm emission from atomic oxygen pumped by intense laser field has been identified as superfluorescence enabled by cooperative radiation due to quantum coherence [26], several other works revealed that the atomic species lasing is the result of stimulated emission [9,16,17]. In the current experiments the pressure-independent gain dynamics and the linear dependence of the 1327 nm signal on the gas pressure suggest that the 1327 nm emission is not superfluorescence. How should we understand these facts? In the case of atomic oxygen lasing, the pump field intensity was estimated to be around 6.11×10¹⁰ W/cm²_, which was very high due to the necessity of dissociation of oxygen molecules into atoms [1,26]. Therefore, the corresponding Rabi frequency was found to be 2×10¹¹ Hz, which is higher than the collisional dephasing rate 10¹⁰Hz [26]. As a result, superradiant emission can develop in that case. From another point of view, the optical gain in that situation was measured to be as high as 62 cm^-1, which favors the formation of collaborative emission [1,26]. In the current case, we have estimated the pulse energy of the 1327 nm radiation with the CCD device by comparing the 1327 nm emission with near-infrared pulses at 1327 nm delivered by the optical parametric amplifier. With a precise measurement of the pulse energy delivered by the OPA laser and consideration of the optical density used in the comparison, the pulse energy of the 1327 nm lasing emission can be calibrated. The pulse energy of the 1327 nm radiation was estimated to be ∼ 1 aJ. By assuming the pulse duration of 10 ps and a beam waist of 16.7µm, the corresponding laser intensity is estimated to be ∼ 45 mW/cm². The electric field is then calculated to be $E = \textrm{}\sqrt {I/2n{\epsilon _0}c} $ = 300 V/m. The average Rabi frequency can thus be estimated by $\mathrm{\Omega } = D \times E/h$, where D is the electric dipole moment, E is the electric field amplitude, and h is the Planck constant. The electric dipole moment D for simple atoms and molecules is usually on the order of 1 Debye ∼3.33×10⁻³⁰ C·m. So the average Rabi frequency is calculated to be $\mathrm{\Omega } = {10^3} - {10^4}\textrm{Hz}$. This is 6-7 orders of magnitude less than that in the situation of oxygen lasing pumped with intense laser. Considering similar collisional dephasing frequency since the argon gas pressure is close to ambient air, the Rabi frequency is now much smaller than the collisional dephasing rate. Therefore, one should not expect superradiant emission at 1327 nm in the current experiments. This analysis agrees with the fact that the optical gain in the current experiment is around 5.82 cm^-1, which is one order of magnitude smaller than that in the case of oxygen lasing [1].

4. Conclusion

In conclusion, we have performed a pump-probe study on the cavity-free lasing effect of argon gas pumped by femtosecond UV pulses at 262 nm. With injection of the seeding pulse into the excited argon gas, the forward coherent emission at 1327 nm was significantly amplified and the optical gain was estimated to be 5.82 cm^-1. The optical gain presents a pressure-independent temporal dynamic with a decay time of ∼ 10 ps. In the meantime, the intensity of the 1327 nm radiation is found to be linearly proportional to the argon pressure in the range of 100-1000 mbar. With these above observations, we suggested that the 1327 nm signal is amplified spontaneous emission due to population inversion and superradiance has not been developed in this lasing effect in the current conditions.