Multi-electron coincidence spectroscopy: double photoionization from molecular inner-shell orbitals

We have studied double photoionization from molecular inner-shell orbitals and investigated the properties of the resultant double core-hole states in molecules, by multi-electron coincidence spectroscopy with a magnetic bottle electron spectrometer. A brief summary of our previous studies is presented.


Introduction
Single photon absorption in the vacuum ultraviolet and soft-X ray ranges often leads to the emission of several electrons from atoms and molecules. The kinetic energy correlation between all the electrons emitted in a single ionization event is very informative to understand the multi-electron emission processes and can be observed with multi-electron coincidence spectroscopy. Such multi-electron coincidence measurements can be performed very efficiently by the use of the magnetic bottle technique [1], thanks to the extremely-high collection efficiency over almost 4π steradian solid angle achievable using this method.
By multi-electron coincidence spectroscopy using a magnetic bottle electron spectrometer, we have examined many different aspects of multi-electron emission processes of rare gases and simple molecules [2][3][4][5][6][7][8][9][10][11][12][13]. One of the highlights in our recent study is the observation of double photoionization from molecular inner-shell orbitals and the resultant formation of double core-hole states in molecules. Here, in addition to ejections of two 1s-electrons from a single atomic site (K -2 double core ionization) [10,11], those from two different atomic sites (K -1 K -1 double core ionization) [12,13] have been identified. We present in this manuscript a brief summary of our previous studies of double photoionization from molecular inner-shell orbitals.

Experimental
The experiments were performed at the undulator beamline BL-16A of the Photon Factory, and at the undulator beamline TEMPO of SOLEIL. In each case, the storage ring was operated in single bunch and in top-up modes. For the experiments at the two synchrotron facilities, we used two different, but very similar electron spectrometers based on the magnetic bottle time-of-flight technique. The description of the spectrometers and the data accumulation scheme are given elsewhere [10,[12][13][14].
The energy resolving powers of the apparatuses, E/∆E, were nearly constant at 60 for electrons of E > 1 eV, though ∆E was limited to around 20 meV (FWHM) for E < 1 eV. The measured detection efficiencies decrease slowly with electron kinetic energy from 70 ± 5% (E ≈ 0 eV) to 40 ± 5% (E = 800 eV).
3. Results and discussion 3.1. K -2 double core ionization In the course of our study, observation of K -2 double core ionization was first achieved in the nitrogen molecule [10]. The K -2 double core ionization creates N 2 2+ (K -2 ) states which decay stepwise by emitting two Auger electrons: 2nd step Auger decay Since four electrons (two photoelectrons and two Auger electrons) are ejected from this sequential process, the formation and decay of the N 2 2+ (K -2 ) states can be observed as a four-fold electron coincidence. Figure 1  The spectroscopic information of the N 2 2+ (K -2 ) states formed is revealed on the histogram of the sum of two photoelectrons' energies in Fig. 1(b), which is plotted also as a function of the binding energy of N 2 2+ states. The spectrum is obtained by integrating the coincidence yields on the map in Fig.  1(a) along the direction for constant sum of the two photoelectrons' energies. The main N 2 2+ (K -2 ) state is found at binding energy of 902.55 ± 0.5 eV, in reasonable agreement with a theoretical predication [10,15]. In the higher binding energy side of the main state, satellite structures which are due to the double core ionization accompanied by a simultaneous excitation of a valence electron are clearly visible. The total intensity of the satellite structures is about 25% of the main N 2 which is much larger than the value of the main and satellite intensity ratio for the single core ionization at a comparable excess energy. This is due to the stronger perturbation imparted to the valence electrons in the removal of two core electrons than in that of a single core electron [16].   Fig. 2(b), where the two Auger electrons are plotted in the single energy scale. Two components are resolved in the range for the first-step Auger electrons (375-450 eV), while only a broad structure is seen in the range for the second-step Auger electrons (300-375 eV). The two peak structure of the first-step Auger transition is reasonably reproduced by a calculation without considering the nuclear motion in the process [11]. In contrast, the same calculation cannot fit the broad structure of the second-step Auger transition, and shows a feature with at least two resolved peaks. The calculations including the nuclear motion reach a better qualitative agreement, demonstrating that the effect of the nuclear motion is important in the second-step Auger transition which occurs from the N 2 3+ (K -1 V -2 ) with repulsive potential energy curves.
Following the observation of the K -2 double core ionization in nitrogen molecule, our investigation with multi-electron coincidence spectroscopy has collected information on the spectroscopy and the decay processes of the K -2 double core-hole states of many other molecules including O 2 , CO, CO 2 , and C 2 H 2n (n=1-3) [10,12].
3.2. K -1 K -1 double core ionization While K -1 K -1 double core ionization in the nitrogen molecule was observed later [13], our investigation has firstly identified the process in acetylene [12]: 2nd step Auger The formation and decay of the K -1 K -1 states again should be observed as a four-fold electron coincidence, similar to those of the K -2 states. Figure 3(a) shows the histogram of the sum of the energies of the two photoelectrons detected in coincidence with the two associated Auger electrons, obtained by the same procedure to extract the data shown in Fig. 1(b). The spectrum shows the main K -2 line at 652.5 ± 0.5 eV and satellite states in the higher binding range, similar to that observed for isoelectronic N 2 molecule [11]. On the other hand, any structure associated with the K -1 K -1 signals, which locates around 595.6 eV (as shown later), is hardly discernible in this figure. This is because the two Auger electrons emitted on decay of a K -1 K -1 state are of similar energies in 220-250 eV (as shown later), and the detection dead time prevents the two Auger signals to be distinguished. The K -1 K -1 signal is therefore not present in the four-fold coincidence events, but in three-fold coincidence events consisting of two photoelectrons and one of Auger electrons. The histogram of the sum of the energies of the two photoelectrons in the three-fold coincidence events, shown in Fig. 3(b), exhibits a weak single peak at 595.6 ± 0.5 eV. The peak is assigned to the K -1 K -1 state, whose binding energy is calculated to be 595.86 eV [13].  [12] of the main lines for formation of K -1 K -1 and K -2 are indicated.
We next focus on the Auger decay of the K -1 K -1 state. Figure 4 shows the spectrum of the Auger electron included in the three-fold coincidence events, compared with the two Auger electrons emitted from the main K -2 state. The decay of the K -2 state presents similar characteristics as observed in the isoelectronic N 2 molecule (Fig. 2(b)): the first hypersatellite Auger electron in the kinetic energy range of 270-340 eV, and the second satellite Auger electron in 200-270 eV. Since a single manifold is seen around 220-250 eV in Fig. 4(a), the two Auger electrons emitted in the sequential K -1 K -1 decay are of similar kinetic energies. This kinetic energy range is comparable to that of the second-step Auger electron emitted in the decay of the K -2 state, and shifts, from that of the first-step Auger electron emitted in the decay of the K -2 state, with the difference in binding energy of the K -2 and K -1 K -1 states. These facts suggest that common intermediate C 2 H 2 3+ (K -1 V -2 ) states are created in the Auger cascades from the K -2 and K -1 K -1 states. Figure 5 shows the spectra of the K -2 and K -1 K -1 states observed for the C 2 H 2n (n=1-3) molecules [13], extracted for three-fold coincidence events consisting of two photoelectrons and one of the Auger electrons. Hydrocarbon molecules are known to be species with small chemical shifts in ESCA applications [17]. The chemical shifts for the K -2 states seen in Fig. 5 are somewhat larger than for the K -1 states, although with the experimental resolution (~2 eV) a clear differentiation between C 2 H 4 and C 2 H 6 is not discernible. In contrast the K -1 K -1 states show more larger chemical shifts, where C 2 H 4 and C 2 H 6 are easily distinguished from each other. This observation confirms a theoretical prediction [17]. The observation demonstrates that double core ionization can be a potential new tool for chemical analysis. Figure 4. Auger spectra associated with the Auger decays of (a) the main K -2 and (b) K -1 K -1 states in C 2 H 2 . Figure 5. Spectra of the K -2 and K -1 K -1 states observed for the C 2 H 2n (n=1-3) molecules, extracted for three-fold coincidence events consisting of two photoelectrons and one of Auger electrons emitted at hν = 770 eV.

Conclusions
The results of our multi-electron coincidence spectroscopic studies of double photoionization from molecular inner-shell orbitals have been summarized here. It is demonstrated that the formation and decay of K -2 and K -1 K -1 states are understood in details with the state-of-the-art performance in the spectroscopic method adopted in this study. Our studies are expected to provide useful information in the exploration of double core-hole states formed by two-photon absorption processes using X-ray free electron lasers developed recently.