N₂O is very stable in the troposphere and its global warming potential is about 300 times higher than that of CO₂.^[1] Even worse, after transported to the stratosphere, N₂O can release active nitrogen oxides (NO) via the de-nitrogenation process induced by the solar radiation or charged particles from the space.^[2] There, NO catalytically destroys the ozone layer. In fact, it is found that the agricultural practices and the use of N-fertilizers have already greatly enhanced emissions of N₂O for the last two decades^[3] which is becoming the largest ozone-depleting emission in the 21st century.^[2] Studies on the de-nitrogenation processes of N₂O in presence of external fields are thus of significance in the air pollution control and attract more and more attentions.^[4,5]

A non-negligible de-nitrogenation channel originating from the Coulomb explosions of N₂O²⁺ has been reported recently.^[5,6] After removal of two electrons from N₂O induced by the solar radiation or charged particles from the space, the molecular dication N₂O²⁺ subsequently dissociates through four different fragmentation channels as follows:^[6–9]

The KER of the de-nitrogenation channel of N₂O²⁺ induced in collisions with electrons,^[7,10–12] highly charged ions (HCIs),^[8,13] and intense infrared laser fields^[5,14–16] have been extensively studied. Generally, the dissociation energy strongly depends on the molecular potential curves populated. Accordingly, the KER distributions are subject to projectile species and energies, and a large variety of spectra have been reported in literature.

In 1991, Eland and Murphy found that the peak of KER is around 7.2 ± 0.7 eV induced by 1300 eV electron impact.^[10] At a much lower collision energy of 50 eV, the KER was found to be at 6.3 ± 1.0 eV.^[11] For higher electron impact energies of 10 keV, two peaks in the KER distribution were identified.^[7] These peaks are located at 4.0 ± 0.5 eV and 5.3 ± 0.6 eV with relative intensities of 58:42, respectively. Later, similar structures were also observed for 5 keV electron impact where the major peak was found at 6.6 ± 0.2 eV.^[12]

After double ionization of N₂O molecules in intense infrared laser fields, the KER has been reported to be around 6.5 ± 2.0 eV,^[14] 6.8± 2.0 eV,^[15] and 6.8 eV^[5] for different laser intensities. After absorptions of the single photons from the He II radiation (He IIα 30.4 nm, He IIβ 25.6 nm, etc.), the peaks were found to be centered around 6.4 ± 0.3 eV^[17] and 6.3 ± 0.3 eV.^[6] However, a different value of 7.2 ± 0.4 eV was reported for the shorter wavelength of 25.6 nm.^[10] The dependence of the total kinetic energy of the de-nitrogenation process on the photon energy of synchrotron radiations was reported by Alagia et al.^[18–20] It was also found that the photodissociation processes of N₂O²⁺ are dominated by the direct Coulomb explosion in the case of photon energies between ∼ 35.5 eV and ∼ 38.5 eV.^[21]

While these studies generally contribute to our understanding of the dissociation dynamics, in the case of the double ionization by electrons, ions, and intense infrared laser fields, various states of N₂O²⁺ can be populated which can hardly be resolved in the KER spectra. Moreover, several electronic dissociation pathways are involved which can interfere, making an interpretation of the observed spectra even more complicated.

Very recently, the dissociation pathways of N₂O²⁺ induced by collisions between 5.7 keV/u Xe¹⁵⁺ and N₂O were distinguished.^[8] In the work, the KER spectra showed some similarities to those of electron/photon impact, and the peak positions were found at 6.8 eV and 9.0 eV, respectively. It was concluded that the measured peak at 6.8 eV originates from two states: (a) the N₂O²⁺ ground state (X³Σ⁻) which dissociates to N⁺(³P) + NO⁺(¹Σ⁺); (b) the metastable state (1³Π) which decays radiatively to X³Σ⁻ and dissociates subsequently. The signature of the metastable (1³Π) state could clearly be recognized from the long tail in the two-fragment time-of-flight coincidence spectrum.^[21,22] A shoulder around 9.0 eV was attributed to N₂O²⁺ (1¹Π, 1¹Σ⁻) dissociating to N⁺ (¹D) + NO⁺(¹Σ⁺) via the potential energy curve crossing of the 1¹Π and 1¹Σ⁻ states of N₂O²⁺

In this paper, we report on the direct Coulomb explosion of N₂O²⁺ induced by the extreme ultraviolet photons at the energy of ∼ 38.5 eV. In our study, the photon energy is selected right below the metastable threshold energy (E_mth = 38.5 eV^[22]), and therefore, the population of the metastable state (1³Π) is effectively suppressed. Meanwhile the molecular ion can only be populated in a few states, which greatly simplifies the analysis of the dissociation pathways. The momenta of the fragment ions are recorded in coincidence with the laser pulse by employing the reaction microscope for each fragmentation event. In the time-of-flight (TOF) coincidence spectra of two fragments, different dissociation channels of N₂O²⁺ are resolved. The KER distributions of the de-nitrogenation channel and the corresponding dissociation pathways are discussed in detail.

The experimental setup is shown in Fig. 1. Briefly, the Ti: sapphire laser system^[23] generates laser pulses with a central wavelength of ∼ 800 nm at 3 kHz repetition rate. For each pulse, the energy is ∼ 5 mJ and the pulse duration is about 25 fs. This intense infrared laser beam is then focused into a 5-cm long, 150-μm-diameter hollow fiber filled with argon gas to generate extreme ultraviolet (EUV) light via high harmonic generation (HHG).^[24–26] After the fiber, a set of grating and pinhole are used to spectrally select the EUV photons and the monochromatic photon energy is determined from the photoelectron spectra in single ionization of helium whose binding energy is about 24.6 eV. Eventually, EUV light of ∼ 38.5 eV (25th harmonic) with a bandwidth of about 0.4 eV can be obtained and the flux converted from the photo diode is about 10⁹–10¹⁰ photons/s. Such pulses are then guided through a vacuum beam line to the reaction chamber. The beam line is pumped down to 10⁻⁶ mbar to reduce the EUV absorption of the background molecules.

Fig. 1.

	Figure Option ViewDownloadNew Window
	Fig. 1. Schematic view of the experimental setup (we note that the electron detector is not activated in the present work).

In the reaction chamber, the monochromatic EUV pulse is focused onto the target. The gas target is prepared with the N₂O gas of about 99.99 % purity using the supersonic gas target method, where the stagnation pressure is set to 12 bars.^[27] Meanwhile, the background pressure in the target chamber of reaction microscope^[28–30] can be kept at a sufficiently low value of 10^–9 mbar. After absorption of the EUV photon, one or two electrons are removed from N₂O molecule which dissociates into charged fragments due to the Coulomb repulsion. A uniform electrostatic field of ∼ 105 V/cm is used to project the positively charged fragments to the recoil ion detector (see Fig. 1). The TOF and position information of each fragment are recorded event by event in list-mode form. The momentum of each fragment thus can be reconstructed in the off-line analysis procedures.

In the following discussion, the coordinate system is defined as following. The EUV light propagates in the Z direction, while the supersonic jet moves in the Y direction which is also the polarization direction of the EUV photons. Accordingly, the X direction is along the TOF spectrometer axis.

Figure 2(a) represents the TOF spectra of all charged particles produced in the reaction chamber. For ions with zero momentum, the TOF T_R follows the relation

Fig. 2.

Figure Option ViewDownloadNew Window

Fig. 2. (a) The TOF spectrum of recoil ions. (b) Two-dimensional TOF coincidence spectrum of charged fragments. Enlarged map is the Coulomb explosion process of N2O2+. The first slash line I centered at (3.90, 5.71) is from the de-nitrogenation photodissociation channel. Here, the x-axis value shows the TOF of N+ fragments while the y-axis value shows the TOF of NO+. The second slash line II at (4.17, 5.52) is the de-oxygenation photodissociation channel in which the x-axis value is the TOF of O+ fragment and the y-axis value is that of N2+ fragment.

Figure 2(b) shows the ion–ion coincidence spectra from the fragments as measured at a photon energy of ∼ 38.5 eV, where the x- and y-axis values represent the TOFs of the first fragment (TOF1) and the second one (TOF2), respectively. Neglecting the relatively small electron and photon momenta, the fragmentation could be treated as a two-body dissociation event, i.e., the two fragment ions fly back-to-back and share an equal amount of momentum. Consequently, distinct slash lines with a slope of about −1 appear in the coincidence spectrum. Whereas, the vertical and horizontal traces in the map represent false and random coincidences which can be easily discriminated in the present analysis and will not affect the quality of our results. In the spectra, two pronounced coincidence slash lines are clearly identified corresponding to the de-nitrogenation channel (I) and the de-oxygenation channel (II) (see the inset in Fig. 2(b)), respectively. The relative branching ratio of the two channels (N⁺ + NO⁺/O⁺ + N₂⁺) is about 7:2, in a good agreement with the previous results.^[21] In the present work, we concentrate on the N₂^O2+→ N⁺ + NO⁺ channel. Comparing with the previous studies (for instance, see Refs. [5,8,10,12,21,31]) which reported a long tail extending from the N⁺ + NO⁺ coincidence trace in the ion–ion coincidence map, in contrast, the tail structure in our data can, if exists, hardly be observed.

The observed difference can be ascribed to the collisional population of the metastable 1³Π state in the previous studies,^[21,22] a channel which is energetically not allowed in our experiment. In the present work, we focus on the processes of two electrons removed from the outer-shell Π-orbital of N₂O by absorption of a ∼ 38.5 eV photon. The only accessible electronic states of the N₂O²⁺ dication are the X³Σ⁻, 1¹Δ, and 1¹Σ⁺ states (see Fig. 3(a)). Meanwhile, to a large extent the excitation to the metastable 1³Π state is suppressed due to the limited photon energy. From all these possible electronic states, the N₂O²⁺ dication dissociates into two charged fragments and undergoes direct Coulomb explosions.^[21]

Fig. 3.

	Figure Option ViewDownloadNew Window
	Fig. 3. (a) The potential energy curves of N2O2+ given for the de-nitrogenation channel.[22] Vertical dashed lines indicate the Franck–Condon region. (b) The KER spectra of the de-nitrogenation process. The solid circles are experimental data, and the red solid lines are the fitting lines. The inset highlights the contributions from different pathways to the shoulder formation (see text).

The KER distribution of the N₂^O2+ → N⁺ + NO⁺ dissociation channel is shown in Fig. 3(b), which exhibits a major peak centered at 6.4 eV and a small shoulder located around 8.8 eV at the right wing of the major peak. Table 1 lists the KER values for the de-nitrogenation dissociation channel reported in literatures and results of our work. According to the potential energy curves (PECs) of N₂O^2+[22] shown in Fig. 3(a), N₂O²⁺ in states of X³Ω⁻, 1¹Δ, and 1¹Ω⁺ can dissociate through five different pathways: (i) direct dissociations of the ground state X³Ω⁻ to the first dissociation limit (L1) with the asymptotic potential level of E_L1 = 28.79 eV,^[9] (ii) direct dissociations from the 1¹Δ state to the second dissociation limit (L2), whose asymptotic energy is E_L2 = 30.69 eV.^[9] (iii) predissociation from the 1¹Δ state to 1³Π via their potential curve crossing point (marked as C), (iv) direct dissociation of the 1¹Σ⁺ state to the second dissociation limit (L2), (v) predissociation from the 1¹Σ⁺ states to 1³Π via their potential curve crossing point (marked as D). The corresponding KER peak values of these five channels are listed in Table 2 which are used as the Gaussian peak centers in the fitting procedure. Meanwhile, for all five Gaussian fittings, the same width of 1.3 eV is used in the first order approximation. The corresponding relative contributions from these processes are listed in Table 2.

Table 1.

Table 1.

Table 1. The KER peak values for N2O2+ → N+ + NO+ dissociation channel obtained in the present experiment and earlier reported experiments. . Projectiles Condition KER/eV Electrons 1300 eV 7.2 ± 0.7[10] 50 eV 6.3 ± 1.0[11] 10 keV 4.0 ± 0.5, 5.3 ± 0.6[7] 5 keV 6.6 ± 0.2[12] HCIs 5.9 MeV/u\Xe43+ 7.0[13] 5.7 keV/u\Xe15+ 6.8, 9.0[8] Intense laser 795 nm, 100 fs, ∼ 5 PW/cm2 6.5 ± 2.0[14] 800 nm, 60 fs, ∼ 0.16 PW/cm2 6.8 ± 2.0[15] 750 nm, 4 fs, ∼ 0.3 PW/cm2 6.8[5] Photons 40.8 eV 6.4 ±0.3[17] 12.4–51.7 eV 6.3 ± 0.3[6] 48.4 eV 7.2 ± 0.4[10] ∼ 38.5 eV 6.4, 8.8a a Present work.

Table 1. The KER peak values for N2O2+ → N+ + NO+ dissociation channel obtained in the present experiment and earlier reported experiments. .

Table 2.

Table 2.

Table 2. The KER values of de-nitrogenation channel derived from different dissociation pathways. . Dication state Initial energy/eV Dissociation limit Asymptotic energy/eV KER/eV Relative contribution Label X3Σ− 35.825 N+(3 P) + NO+(1Σ+) (L1) 28.79 7.035 2.6% i 11Δ 36.962 N+(1D) + NO+(1Σ+) (L2) 30.69 6.272 56.3% ii N+(3P) + NO+(1Σ+) (L1) 28.79 8.172 4.4 % iii 11Σ+ 37.546 N+(1D) + NO+(1Σ+) (L2) 30.69 6.856 28.6% iv N+(3P) + NO+(1Σ+) (L1) 28.79 8.756 8.1 % v

Table 2. The KER values of de-nitrogenation channel derived from different dissociation pathways. .

From the fitting results, we conclude that the major peak centered at 6.4 eV can be attributed to two processes: (i) the dominant photodissociation peak centered at 6.3 eV (black solid curve) which represents a direct decay from the 1¹Δ state to the second dissociation limit of N⁺ (¹D) + NO⁺ (¹Σ⁺), (ii) the second peak at 6.8 eV (green solid curve) is assigned to the direct dissociation of the 1¹Σ⁺ state to the same dissociation limit (L2). Meanwhile, the contribution from the ground state X³Σ⁻ (blue solid curve, see Fig. 3(b) and Table 2) is negligibly small. These results indicate that the dication of N₂O²⁺ formed by absorption of one 38.5 eV photon is mainly populated in the 1¹Δ and 1¹Δ⁺ states, and the direct dissociations from these states are the dominant decay processes. These populations are very similar to those observed previously in the double photoionization of N₂O by He II radiations at 30.4 nm^[32] and 25.6 nm.^[22] Whereas, in Refs. [5,8], the major peak centered around 6.8 eV (see Table 1) is attributed to both contributions from the direct dissociation of the ground state X³Σ⁻ and the decay of 1³Π to X³Π⁻ through fluorescence which consequently undergoes the Coulomb explosion.^[20] Moreover, the dissociation of the 1¹Δ and 1¹Σ⁺ states was not addressed.

Accordingly, the shoulder structure around 8.8 eV can be interpreted as follows. As pointed out above, with ∼ 38.5 eV photon energy the electronic states of N₂O²⁺ are mainly populated to the 1¹Δ and 1¹Σ⁺ states. Therefore, there is a good reason to believe that the shoulder structure is ascribed to the decay of the 1¹Δ and 1¹Σ⁺ states of N₂O²⁺ to the first dissociation limit of N⁺ (³P) + NO⁺ (¹Σ⁺). This dissociation takes place through the potential curve crossings with the 1³Π state at C and D shown in Fig. 3(a). These pathways via C and D will produce dissociation energies of 8.2 eV (pink dashed curve) and 8.8 eV (purple dashed curve) as shown in Fig. 3(b), respectively. Besides, the direct decay from the 1¹Δ states to L2 is much stronger than the predissociation to L1. The first reason is that the vibrational states in 1¹Δ are highly populated, and are very close or partly over the barrier. A second reason is that the crossing point (C) is outside the potential well, i.e., beyond the barrier. On the other hand, the predissociation of the 1¹Σ⁺ state via the crossing with 1³Π state (D) is relatively stronger since the crossing is inside the well and bellow the barrier. Taking all the above factors into account, we can conclude that the main contribution to the shoulder structure is from the 1¹Σ⁺ states of N₂O²⁺ to the first dissociation limit of N⁺(³P) + NO⁺(¹Σ⁺) with potential curve crossing of the 1³Π state.

However, our results are different from the earlier results of Chen et al.^[8] In their study, they used 5.7 keV/u Xe¹⁵⁺ beam ions to produce N₂O²⁺, therefore, they observed a shoulder structure situated at 9.0 eV of the de-nitrogenation channel. This structure was explained by the population of the 1¹Π state in N₂O²⁺ which predissociates to the second dissociation limit (L2) of N⁺(¹D) + NO⁺(¹Σ⁺). This predissociation occurs through two crossing points of the potential energy curves, as shown by A and B in Fig. 3(a). The difference between our results and those in Ref. [8] is obvious since, in ion induced collisions all excited states might be populated while in photon impact situation, only few specific states can be populated due to the energy restriction.

The Coulomb explosion of N₂O²⁺ is studied with the reaction microscope double photoionization from N₂O by extreme ultraviolet photons at ∼ 38.5 eV. The de-nitrogenation as well as the de-oxygenation channels are identified from the time-of-flight coincidence spectra. It is found that, in the de-nitrogenation channel, the N₂O²⁺ dications mainly dissociate from the 1¹Δ and 1¹Σ⁺ states. In the KER spectra, a main peak centered at 6.4 eV with a shoulder towards higher energies is observed. Further studies show that the major peak is derived from the direct dissocaition of N₂O²⁺ from the 1¹Δ and 1¹Σ⁺ potential curve to N⁺(¹D) + NO⁺(¹Σ⁺) (the second dissociation limit L2). The shoulder is originated from the dissociation of N₂O²⁺ from the 1¹Δ and 1¹Σ⁺ states to N⁺(³P) + NO⁺(¹Σ⁺) (the first dissociation limit L1) via the crossing of the 1³Π state potential energy curves. Our work shows that the table-top monochromatic EUV beam line provides great advantages in detailed studies of the electronic states and dissociation dynamics involved in photo-ionization of molecular systems.