LETTER
doi:10.1038/nature12927
Resonant Auger decay driving intermolecular Coulombic decay in molecular dimers ¨ffler1,2, H.-K. Kim1, F. P. Sturm1,2, K. Cole1, N. Neumann1, A. Vredenborg1, J. Williams3, I. Bocharova2, F. Trinter1, M. S. Scho 4 ¨cking1, R. Do ¨rner1 & T. Jahnke1 R. Guillemin , M. Simon4, A. Belkacem2, A. L. Landers3, Th. Weber2, H. Schmidt-Bo
In 1997, it was predicted1 that an electronically excited atom or molecule placed in a loosely bound chemical system (such as a hydrogenbonded or van-der-Waals-bonded cluster) could efficiently decay by transferring its excess energy to a neighbouring species that would then emit a low-energy electron. This intermolecular Coulombic decay (ICD) process has since been shown to be a common phenomenon2–12, raising questions about its role in DNA damage induced by ionizing radiation, in which low-energy electrons are known to play an important part13,14. It was recently suggested15 that ICD can be triggered efficiently and site-selectively by resonantly core-exciting a target atom, which then transforms through Auger decay into an ionic species with sufficiently high excitation energy to permit ICD to occur. Here we show experimentally that resonant Auger decay can indeed trigger ICD in dimers of both molecular nitrogen and carbon monoxide. By using ion and electron momentum spectroscopy to measure simultaneously the charged species created in the resonant-Auger-driven ICD cascade, we find that ICD occurs in less time than the 20 femtoseconds it would take for individual molecules to undergo dissociation. Our experimental confirmation of this process and its efficiency may trigger renewed efforts to develop resonant X-ray excitation schemes16,17 for more localized and targeted cancer radiation therapy. The experiment presented here shows that resonant excitation of a K-shell electron to a bound state is followed by Auger decay to an ionic species that can then undergo ICD, as sketched in Fig. 1 and proposed in ref. 15. The initial resonant excitation of the electron occurs as in the experiments that probed resonant interatomic Coulombic decay5,6, but the state undergoing ICD is created after partial de-excitation of the system through a local Auger decay. The Auger decay can lead to the ground state of the molecular ion through ‘participator Auger decay’, although in many cases the excited electron will act as just a ‘spectator’ to an Auger decay in which an electron from the valence or inner valence shell fills the core hole and a second electron from the valence shell is emitted. This spectator pathway produces ionic states which are high enough in excitation energy to allow ICD to occur, and in the case of carbon monoxide accounts for the decay of approximately 75% of coreexcited molecules18. Our experiment explores the overall scenario for two simple model systems—clusters of just two carbon monoxide or two nitrogen molecules—that can be investigated in great detail. This allows us to follow the Auger decay occurring after resonant excitation of an inner-shell electron into the lowest unoccupied molecular orbital (in a P* excitation) and the subsequent ICD: hn 1 N2/N2 R N2*(1 s21 P*)/N2
(local, resonant excitation)
N2*(1 s21 P*)/N2 R N21*/N2 1 eAuger
N21*/N2 R N21 1 N21 1 eICD
(spectator Auger decay)
where hn is the incident radiation, P* is the excited molecular orbital, and eAuger and eICD are the Auger- and ICD-emitted electrons (‘1 s21’ refers to a K-shell electron being removed during excitation). Figure 2a, b shows, for (CO)2 and (N2)2, the correlation between the kinetic energy release of the two molecular ions and the kinetic energy of the electrons measured in coincidence. Unlike in similar plots for ICD in rare-gas dimers11, no discrete structures are observed in Fig. 2. This is a direct consequence of the repulsive nature of the intermediate state populated by the resonant Auger decay and of the vibrational and rotational degrees of freedom of the ionic fragments. The resonant Auger decay onto a repulsive state of the molecule leads to a continuum
a
b
c
(1)
(2)
(ICD 1 two-site Coulomb explosion) (3)
Figure 1 | The overall decay cascade mechanism. Shown is the series of events involved in resonant-Auger-driven ICD (see equations (1)–(3)). a, One molecule (left) of the molecular dimer is core-excited. b, The core-excited state decays by a spectator Auger decay to a highly excited state of the molecular ion. c, ICD transfers the excitation energy to the molecular neighbour (right), where a low-energy ICD electron is emitted.
1
Institut fu¨r Kernphysik, Goethe-Universita¨t, Max-von-Laue-Strasse 1, 60438 Frankfurt am Main, Germany. 2Lawrence Berkeley National Laboratory, Chemical Sciences Division, Berkeley, California 94720, USA. 3Department of Physics, Auburn University, Auburn, Alabama 36849, USA. 4UPMC and CNRS, UMR 7614, Laboratoire de Chimie Physique Matie`re et Rayonnement, 75005 Paris, France. 0 0 M O N T H 2 0 1 3 | VO L 0 0 0 | N AT U R E | 1
©2013 Macmillan Publishers Limited. All rights reserved
RESEARCH LETTER a
10 30
9
Electron energy (eV)
8
25
7 20
6 5
15
4 10
3 2
0
b
5
(CO)2
1 0
1
2
3 4 5 6 7 8 Kinetic energy release (eV)
9
10
10
18
9
16
Electron energy (eV)
8
14
7
12
6
10
5
8
4
6
3
4
2
2
(N2)2
1 0
0
0
1
2
3 4 5 6 7 8 Kinetic energy release (eV)
9
10
0
c
values measured strongly support the picture of ICD being the underlying process: the maximum of the observed kinetic energy release distribution is 3.4 eV for (N2)2 and 3.7 eV for (CO)2, which compare fairly well with the values of 3.57 eV and 3.69 eV estimated by calculating the kinetic energy of two singly charged point particles starting a Coulomb explosion at the typical mean intermolecular distance of the CO and N2 dimers, respectively. (This estimate assumes a simple Coulomb potential, which is a good approximation for Van-der-Waals-bound systems, so that the kinetic energy release in atomic units is given by 1/R, with R the distance between the two charges. For R we use the ˚ for (N2)2, whereas dimer intermolecular distance, reported19 to be 4.03 A ˚ .) These the mean of the values reported20,21 for the CO dimer is 3.9 A findings support the scenario of an intermolecular decay mechanism such as ICD. An alternative mechanism that would also lead to two singly charged molecular ions being launched at the mean intermolecular distance of the dimer in its ground state is intermolecular ‘knockout’, as recently observed22 in He2. It occurs if the fast Auger electron is emitted in the direction of the neighbouring molecule and knocks out an electron, thereby ionizing the neighbouring molecule. This process yields the two molecular ions observed: an Auger electron of reduced energy and a low-energy electron. However, this process happens only22 if the neighbouring molecule is located in the direction of emission of the Auger electron. In our experiments, the orientation of the dimer in space at the instant of Coulomb explosion is known from the coincidence momentum measurement, whereas the direction of the fast Auger electron is measured for every event via its recoil on the centre of mass of the two Coulomb-exploding molecular ions. From this we obtain the Auger electron angular emission pattern with respect to the molecular axis of the dimer in Fig. 2c, which is nearly isotropic and thus eliminates the intermolecular knockout scenario (which would lead to an emission pattern strongly directed along the molecular axis). Our experimental data thus reveal ICD to be a prominent decay channel for excited dimers of N2 and CO. ICD might have a substantial (yet poorly studied) effect on the fate of an excited molecular ion: excited molecular ions are rarely stable and dissociate if they are present as isolated species, but might survive as stable entities in solution or in any other chemical environment where ICD can occur. The competition between the ICD (equation (3)) and energy relaxation through dissociation without release of an electron is crucial: N21*/N2 R N1 1 N 1 N2
+
+
Figure 2 | Experimental results. a, Kinetic energy release of (CO)2 versus energy of one of the two electrons created by ICD after resonant excitation and subsequent Auger decay at a photon energy of 287.4 eV (P* excitation of CO). The colour scale shows the intensity in counts. b, Same plot for (N2)2 recorded at a photon energy of 401.9 eV (P* excitation of N2). c, Emission direction of the Auger electron with respect to the molecular axis of the N2 dimer (with statistical error bars). The dimer is oriented horizontally, as depicted by the green icon. The grey circle is a line to guide the eye, corresponding to isotropic emission.
of Auger energies and hence to a continuum of excitation energies of the intermediate N21*/N2 (or CO1*/CO) state. The data in Fig. 2a, b are obtained from the detection of two singly charged molecular ions, revealing the Coulomb explosion of the molecular dimer as the terminal step of ICD. The actual kinetic energy release
(one-site dissociation)
(4)
ICD can thus effectively suppress dissociation if it occurs quickly enough. Alternatively, the inverse might also be true: ICD in a molecular system (equation (3)) might become a rare phenomenon if one-site dissociation (equation (4)) is a very fast competing channel. The CO1* potential energy curves above the single ionization potential of CO are all steeply repulsive23, but the fact that we observe ICD and a breakup of the molecular dimer into CO1/CO1 shows that ICD nonetheless outpaces dissociation. This allows us to use the molecular dissociation as a clock to obtain an estimate of the timescale on which ICD occurs in the present case. The typical slopes of the potential energy curves involved are ˚ 21. From this and the fact that we observe only ICD events 10 eV A where the CO1 does not fragment, we can estimate the maximum time which could have elapsed before the molecular ion has to relax via ICD. For a repulsive state the potential energy of the system decreases as the molecule dissociates. Therefore, for a given potential energy surface we can calculate how long it will take before an internuclear distance has been reached at which the potential energy has dropped below the threshold for ICD. In the case of the molecules and states populated here, this time is less than 20 fs. Accordingly, ICD must occur on a timescale shorter than 20 fs. ICD has been discussed15 in the context of radiation biology and also cancer radiotherapy, which still usually uses broadband irradiation of biological tissue to destroy cancerous cells, with considerable adverse
2 | N AT U R E | VO L 0 0 0 | 0 0 M O N T H 2 0 1 3
©2013 Macmillan Publishers Limited. All rights reserved
LETTER RESEARCH side effects. Tagging cancerous cells with molecular markers containing at least one atom of a high-Z element for resonant excitation by energetically well-defined X-rays, in order to localize the radiation damage to the required site inside a biological system while leaving surrounding tissue unaffected16, is thus an attractive prospect. In principle, because a resonant excitation of the kind described here is typically ten times stronger than the non-resonant ionization used by broadband irradiation, the overall radiation dose can be minimized using monochromatized X-rays tuned to a suitable resonance. ICD offers the added advantage of directly generating low-energy electrons that are known to be genotoxic and thus are effective mediators of the anticancer effects of radiotherapy. The present decay cascade, with ICD occurring efficiently after resonant excitation of a selected atom and its subsequent Auger decay, is particularly attractive, because it is possible to target a specific site in a larger system at which ICD and the emission of genotoxic low-energy electrons should take place. We expect that our experimental validation of this process, and other studies published during review of this contribution that also confirm its existence and even the tunability of the ICD electron energy in raregas clusters24–26, will stimulate further exploration of Auger-electrondriven cancer therapy.
8.
9. 10. 11. 12.
13.
14. 15.
16.
17. 18. 19.
METHODS SUMMARY We use cold target recoil ion momentum spectroscopy (COLTRIMS)27,28 to measure in coincidence all charged particles created in a single reaction, using beamline 11.0.2 of the Advanced Light Source at Lawrence Berkeley National Laboratory. The N2 and CO dimers were produced by expanding the gas through a 30-mm nozzle at a stagnation pressure of 10 bar. The nozzle was cooled to approximately 140 K to enhance dimer production. The supersonic beam was collimated by a set of two skimmers and then crossed with the photon beam inside a COLTRIMS spectrometer29. An electric field of 7.4 V cm21 and a parallel magnetic field of 7.0 gauss guided electrons and ions to two position-sensitive micro-channel plate detectors with delay line readout (see http://www.roentdek.com for details on the detectors). The fields were adjusted such that electrons of up to 15 eV kinetic energy could be collected with a 4p solid angle. Owing to the long ion drift arm the spectrometer accepted only N21 or CO1 (in case their kinetic energies were higher than 5 eV), which were emitted within 10u with respect to the spectrometer axis. The light polarization was circular in the case of N2. The data were recorded in list mode. For each ionization event we recorded the positions of impact and times of flight of all registered particles. This allowed us to extract the very weak dimer signal from our data, because the dimer fraction in our beam was only 0.1% to 1%. Thus most recorded ions and electrons resulted from ionization of the monomer. We can identify the ICD channel by selecting only events in which two N21 (or two CO1) ions with equal and opposite momentum occur. This back-to-back emission is a unique signature of the final step of Coulomb explosion following ICD (see equation (3)). Received 8 May 2012; accepted 28 November 2013. Published online 22 December 2013. 1. 2. 3. 4. 5. 6. 7.
Cederbaum, L. S., Zobeley, J. & Tarantelli, F. Giant intermolecular decay and fragmentation of clusters. Phys. Rev. Lett. 79, 4778–4781 (1997). Jahnke, T. et al. Ultrafast energy transfer between water molecules. Nature Phys. 6, 139–142 (2010). Mucke, M. et al. A hitherto unrecognized source of low-energy electrons in water. Nature Phys. 6, 143–146 (2010). Kim, H.-K. et al. Enhanced production of low energy electrons by alpha particle impact. Proc. Natl Acad. Sci. USA 108, 11821–11824 (2011). Barth, S. et al. Observation of resonant interatomic Coulombic decay in Ne clusters. J. Chem. Phys. 122, 241102 (2005). Aoto, T. et al. Properties of resonant interatomic Coulombic decay in Ne dimers. Phys. Rev. Lett. 97, 243401 (2006). Jahnke, T. et al. Experimental separation of virtual photon exchange and electron transfer in interatomic Coulombic decay of neon dimers. Phys. Rev. Lett. 99, 153401 (2007).
20.
21. 22. 23.
24.
25.
26. 27. 28. 29.
Ueda, K. et al. Interatomic Coulombic decay following the Auger decay: experimental evidence in rare-gas dimers. J. Electron Spectrosc. Relat. Phenom. 166–167, 3–10 (2008). Grieves, G. A. & Orlando, T. M. Intermolecular Coulomb decay at weakly coupled heterogeneous interfaces. Phys. Rev. Lett. 107, 016104 (2011). Marburger, S., Kugeler, O., Hergenhahn, U. & Mo¨ller, T. Experimental evidence for interatomic Coulombic decay in Ne clusters. Phys. Rev. Lett. 90, 203401 (2003). Jahnke, T. et al. Experimental observation of interatomic Coulombic decay in neon dimers. Phys. Rev. Lett. 93, 163401 (2004). O¨hrwall, G. et al. Femtosecond interatomic Coulombic decay in free neon clusters: large lifetime differences between surface and bulk. Phys. Rev. Lett. 93, 173401 (2004). Boudaı¨ffa, B., Cloutier, P., Hunting, D., Huels, M. A. & Sanche, L. Resonant formation of DNA strand breaks by low-energy (3 to 20 eV) electrons. Science 287, 1658–1660 (2000). Hanel, G. et al. Electron attachment to uracil: effective destruction at subexcitation energies. Phys. Rev. Lett. 90, 188104 (2003). Gokhberg, K., Kolorencˇ, P., Kuleff, A. I. & Cederbaum, L. S. Site- and energy-selective slow-electron production through intermolecular Coulombic decay. Nature http:// dx.doi.org/10.1038/nature12936 (this issue). Pradhan, A. K. et al. Resonant X-ray enhancement of the Auger effect in high-Z atoms, molecules, and nanoparticles: potential biomedical applications. J. Phys. Chem. A 113, 12356–12363 (2009). Howell, R. W. Auger processes in the 21st century. Int. J. Radiat. Biol. 84, 959–975 (2008). Botting, S. K. & Lucchese, R. R. Auger decay of the C 1s -. 2p* excitation of CO. Phys. Rev. A 56, 3666–3674 (1997). Aquilanti, V., Bartolomei, M., Cappelletti, D., Carmona-Novillo, E. & Pirani, F. The N2–N2 system: an experimental potential energy surface and calculated rotovibrational levels of the molecular nitrogen dimer. J. Chem. Phys. 117, 615 (2002). Havenith, M., Petri, M., Lubina, C., Hilpert, G. & Urban, W. J. IR spectroscopy of (CO)2 using concentration-frequency double modulation in a supersonic jet expansion. J. Mol. Spectrosc. 167, 248–261 (1994). Meredith, A. W. & Stone, A. J. An ab initio and diffusion Monte Carlo study of the potential energy surface of the CO dimers. J. Phys. Chem. A 102, 434–445 (1998). Havermeier, T. et al. Single photon double ionization of the helium dimer. Phys. Rev. Lett. 104, 153401 (2010). Baltzer, P. et al. Inner-valence states of CO1 between 22 eV and 46 eV studied by high resolution photoelectron spectroscopy and ab initio CI calculations. J. Phys. At. Mol. Opt. Phys. 27, 4915 (1994). Kimura, M. et al. Efficient site-specific low-energy electron production via interatomic Coulombic decay following resonant Auger decay in argon dimers. Phys. Rev. A 87, 043414 (2013). O’Keeffe, P. et al. The role of the partner atom and resonant excitation energy in interatomic Coulombic decay in rare gas dimers. J. Phys. Chem. Lett. 4, 1797–1801 (2013). Kimura, K. et al. Controlling low-energy electron emission via resonant-augerinduced interatomic Coulombic decay. J. Phys. Chem. Lett. 4, 1838–1842 (2013). Do¨rner, R. et al. Cold target ion momentum spectrocopy: a ‘momentum microscope’ to view atomic collisions dynamics. Phys. Rep. 330, 95–192 (2000). Ullrich, J. et al. Recoil-ion and electron momentum spectroscopy: reactionmicroscopes. Rep. Prog. Phys. 66, 1463 (2003). Scho¨ffler, M. S. et al. Matter wave optics perspective at molecular photoionization: K-shell photoionization and Auger decay of N2. New J. Phys. 13, 095013 (2011).
Acknowledgements This work was supported by the Deutsche Forschungsgemeinschaft and the Deutscher Akademischer Austauschdienst. We thank the staff of the Advanced Light Source for excellent support during the beam time. This work was supported by the Director, Office of Science, Office of Basic Energy Sciences, and by the Division of Chemical Sciences, Geosciences, and Biosciences of the US Department of Energy at the Lawrence Berkeley National Laboratory under contract number DE-AC02-05CH11231. The Advanced Light Source is supported by the Director, Office of Science, Office of Basic Energy Sciences, of the US Department of Energy under contract number DE-AC02-05CH11231. We thank K. Gokhberg and L. Cederbaum for suggesting this experiment and for many discussions. M.S.S. thanks the Alexander von Humboldt foundation for financial support. Author Contributions All authors contributed to the experiment. F.T. and T.J. performed the data analysis. All authors contributed to the manuscript. Author Information Reprints and permissions information is available at www.nature.com/reprints. The authors declare no competing financial interests. Readers are welcome to comment on the online version of the paper. Correspondence and requests for materials should be addressed to T.J. ([email protected]) or F.T. ([email protected]).
0 0 M O N T H 2 0 1 3 | VO L 0 0 0 | N AT U R E | 3
©2013 Macmillan Publishers Limited. All rights reserved
doi:10.1038/nature12927
Resonant Auger decay driving intermolecular Coulombic decay in molecular dimers ¨ffler1,2, H.-K. Kim1, F. P. Sturm1,2, K. Cole1, N. Neumann1, A. Vredenborg1, J. Williams3, I. Bocharova2, F. Trinter1, M. S. Scho 4 ¨cking1, R. Do ¨rner1 & T. Jahnke1 R. Guillemin , M. Simon4, A. Belkacem2, A. L. Landers3, Th. Weber2, H. Schmidt-Bo
In 1997, it was predicted1 that an electronically excited atom or molecule placed in a loosely bound chemical system (such as a hydrogenbonded or van-der-Waals-bonded cluster) could efficiently decay by transferring its excess energy to a neighbouring species that would then emit a low-energy electron. This intermolecular Coulombic decay (ICD) process has since been shown to be a common phenomenon2–12, raising questions about its role in DNA damage induced by ionizing radiation, in which low-energy electrons are known to play an important part13,14. It was recently suggested15 that ICD can be triggered efficiently and site-selectively by resonantly core-exciting a target atom, which then transforms through Auger decay into an ionic species with sufficiently high excitation energy to permit ICD to occur. Here we show experimentally that resonant Auger decay can indeed trigger ICD in dimers of both molecular nitrogen and carbon monoxide. By using ion and electron momentum spectroscopy to measure simultaneously the charged species created in the resonant-Auger-driven ICD cascade, we find that ICD occurs in less time than the 20 femtoseconds it would take for individual molecules to undergo dissociation. Our experimental confirmation of this process and its efficiency may trigger renewed efforts to develop resonant X-ray excitation schemes16,17 for more localized and targeted cancer radiation therapy. The experiment presented here shows that resonant excitation of a K-shell electron to a bound state is followed by Auger decay to an ionic species that can then undergo ICD, as sketched in Fig. 1 and proposed in ref. 15. The initial resonant excitation of the electron occurs as in the experiments that probed resonant interatomic Coulombic decay5,6, but the state undergoing ICD is created after partial de-excitation of the system through a local Auger decay. The Auger decay can lead to the ground state of the molecular ion through ‘participator Auger decay’, although in many cases the excited electron will act as just a ‘spectator’ to an Auger decay in which an electron from the valence or inner valence shell fills the core hole and a second electron from the valence shell is emitted. This spectator pathway produces ionic states which are high enough in excitation energy to allow ICD to occur, and in the case of carbon monoxide accounts for the decay of approximately 75% of coreexcited molecules18. Our experiment explores the overall scenario for two simple model systems—clusters of just two carbon monoxide or two nitrogen molecules—that can be investigated in great detail. This allows us to follow the Auger decay occurring after resonant excitation of an inner-shell electron into the lowest unoccupied molecular orbital (in a P* excitation) and the subsequent ICD: hn 1 N2/N2 R N2*(1 s21 P*)/N2
(local, resonant excitation)
N2*(1 s21 P*)/N2 R N21*/N2 1 eAuger
N21*/N2 R N21 1 N21 1 eICD
(spectator Auger decay)
where hn is the incident radiation, P* is the excited molecular orbital, and eAuger and eICD are the Auger- and ICD-emitted electrons (‘1 s21’ refers to a K-shell electron being removed during excitation). Figure 2a, b shows, for (CO)2 and (N2)2, the correlation between the kinetic energy release of the two molecular ions and the kinetic energy of the electrons measured in coincidence. Unlike in similar plots for ICD in rare-gas dimers11, no discrete structures are observed in Fig. 2. This is a direct consequence of the repulsive nature of the intermediate state populated by the resonant Auger decay and of the vibrational and rotational degrees of freedom of the ionic fragments. The resonant Auger decay onto a repulsive state of the molecule leads to a continuum
a
b
c
(1)
(2)
(ICD 1 two-site Coulomb explosion) (3)
Figure 1 | The overall decay cascade mechanism. Shown is the series of events involved in resonant-Auger-driven ICD (see equations (1)–(3)). a, One molecule (left) of the molecular dimer is core-excited. b, The core-excited state decays by a spectator Auger decay to a highly excited state of the molecular ion. c, ICD transfers the excitation energy to the molecular neighbour (right), where a low-energy ICD electron is emitted.
1
Institut fu¨r Kernphysik, Goethe-Universita¨t, Max-von-Laue-Strasse 1, 60438 Frankfurt am Main, Germany. 2Lawrence Berkeley National Laboratory, Chemical Sciences Division, Berkeley, California 94720, USA. 3Department of Physics, Auburn University, Auburn, Alabama 36849, USA. 4UPMC and CNRS, UMR 7614, Laboratoire de Chimie Physique Matie`re et Rayonnement, 75005 Paris, France. 0 0 M O N T H 2 0 1 3 | VO L 0 0 0 | N AT U R E | 1
©2013 Macmillan Publishers Limited. All rights reserved
RESEARCH LETTER a
10 30
9
Electron energy (eV)
8
25
7 20
6 5
15
4 10
3 2
0
b
5
(CO)2
1 0
1
2
3 4 5 6 7 8 Kinetic energy release (eV)
9
10
10
18
9
16
Electron energy (eV)
8
14
7
12
6
10
5
8
4
6
3
4
2
2
(N2)2
1 0
0
0
1
2
3 4 5 6 7 8 Kinetic energy release (eV)
9
10
0
c
values measured strongly support the picture of ICD being the underlying process: the maximum of the observed kinetic energy release distribution is 3.4 eV for (N2)2 and 3.7 eV for (CO)2, which compare fairly well with the values of 3.57 eV and 3.69 eV estimated by calculating the kinetic energy of two singly charged point particles starting a Coulomb explosion at the typical mean intermolecular distance of the CO and N2 dimers, respectively. (This estimate assumes a simple Coulomb potential, which is a good approximation for Van-der-Waals-bound systems, so that the kinetic energy release in atomic units is given by 1/R, with R the distance between the two charges. For R we use the ˚ for (N2)2, whereas dimer intermolecular distance, reported19 to be 4.03 A ˚ .) These the mean of the values reported20,21 for the CO dimer is 3.9 A findings support the scenario of an intermolecular decay mechanism such as ICD. An alternative mechanism that would also lead to two singly charged molecular ions being launched at the mean intermolecular distance of the dimer in its ground state is intermolecular ‘knockout’, as recently observed22 in He2. It occurs if the fast Auger electron is emitted in the direction of the neighbouring molecule and knocks out an electron, thereby ionizing the neighbouring molecule. This process yields the two molecular ions observed: an Auger electron of reduced energy and a low-energy electron. However, this process happens only22 if the neighbouring molecule is located in the direction of emission of the Auger electron. In our experiments, the orientation of the dimer in space at the instant of Coulomb explosion is known from the coincidence momentum measurement, whereas the direction of the fast Auger electron is measured for every event via its recoil on the centre of mass of the two Coulomb-exploding molecular ions. From this we obtain the Auger electron angular emission pattern with respect to the molecular axis of the dimer in Fig. 2c, which is nearly isotropic and thus eliminates the intermolecular knockout scenario (which would lead to an emission pattern strongly directed along the molecular axis). Our experimental data thus reveal ICD to be a prominent decay channel for excited dimers of N2 and CO. ICD might have a substantial (yet poorly studied) effect on the fate of an excited molecular ion: excited molecular ions are rarely stable and dissociate if they are present as isolated species, but might survive as stable entities in solution or in any other chemical environment where ICD can occur. The competition between the ICD (equation (3)) and energy relaxation through dissociation without release of an electron is crucial: N21*/N2 R N1 1 N 1 N2
+
+
Figure 2 | Experimental results. a, Kinetic energy release of (CO)2 versus energy of one of the two electrons created by ICD after resonant excitation and subsequent Auger decay at a photon energy of 287.4 eV (P* excitation of CO). The colour scale shows the intensity in counts. b, Same plot for (N2)2 recorded at a photon energy of 401.9 eV (P* excitation of N2). c, Emission direction of the Auger electron with respect to the molecular axis of the N2 dimer (with statistical error bars). The dimer is oriented horizontally, as depicted by the green icon. The grey circle is a line to guide the eye, corresponding to isotropic emission.
of Auger energies and hence to a continuum of excitation energies of the intermediate N21*/N2 (or CO1*/CO) state. The data in Fig. 2a, b are obtained from the detection of two singly charged molecular ions, revealing the Coulomb explosion of the molecular dimer as the terminal step of ICD. The actual kinetic energy release
(one-site dissociation)
(4)
ICD can thus effectively suppress dissociation if it occurs quickly enough. Alternatively, the inverse might also be true: ICD in a molecular system (equation (3)) might become a rare phenomenon if one-site dissociation (equation (4)) is a very fast competing channel. The CO1* potential energy curves above the single ionization potential of CO are all steeply repulsive23, but the fact that we observe ICD and a breakup of the molecular dimer into CO1/CO1 shows that ICD nonetheless outpaces dissociation. This allows us to use the molecular dissociation as a clock to obtain an estimate of the timescale on which ICD occurs in the present case. The typical slopes of the potential energy curves involved are ˚ 21. From this and the fact that we observe only ICD events 10 eV A where the CO1 does not fragment, we can estimate the maximum time which could have elapsed before the molecular ion has to relax via ICD. For a repulsive state the potential energy of the system decreases as the molecule dissociates. Therefore, for a given potential energy surface we can calculate how long it will take before an internuclear distance has been reached at which the potential energy has dropped below the threshold for ICD. In the case of the molecules and states populated here, this time is less than 20 fs. Accordingly, ICD must occur on a timescale shorter than 20 fs. ICD has been discussed15 in the context of radiation biology and also cancer radiotherapy, which still usually uses broadband irradiation of biological tissue to destroy cancerous cells, with considerable adverse
2 | N AT U R E | VO L 0 0 0 | 0 0 M O N T H 2 0 1 3
©2013 Macmillan Publishers Limited. All rights reserved
LETTER RESEARCH side effects. Tagging cancerous cells with molecular markers containing at least one atom of a high-Z element for resonant excitation by energetically well-defined X-rays, in order to localize the radiation damage to the required site inside a biological system while leaving surrounding tissue unaffected16, is thus an attractive prospect. In principle, because a resonant excitation of the kind described here is typically ten times stronger than the non-resonant ionization used by broadband irradiation, the overall radiation dose can be minimized using monochromatized X-rays tuned to a suitable resonance. ICD offers the added advantage of directly generating low-energy electrons that are known to be genotoxic and thus are effective mediators of the anticancer effects of radiotherapy. The present decay cascade, with ICD occurring efficiently after resonant excitation of a selected atom and its subsequent Auger decay, is particularly attractive, because it is possible to target a specific site in a larger system at which ICD and the emission of genotoxic low-energy electrons should take place. We expect that our experimental validation of this process, and other studies published during review of this contribution that also confirm its existence and even the tunability of the ICD electron energy in raregas clusters24–26, will stimulate further exploration of Auger-electrondriven cancer therapy.
8.
9. 10. 11. 12.
13.
14. 15.
16.
17. 18. 19.
METHODS SUMMARY We use cold target recoil ion momentum spectroscopy (COLTRIMS)27,28 to measure in coincidence all charged particles created in a single reaction, using beamline 11.0.2 of the Advanced Light Source at Lawrence Berkeley National Laboratory. The N2 and CO dimers were produced by expanding the gas through a 30-mm nozzle at a stagnation pressure of 10 bar. The nozzle was cooled to approximately 140 K to enhance dimer production. The supersonic beam was collimated by a set of two skimmers and then crossed with the photon beam inside a COLTRIMS spectrometer29. An electric field of 7.4 V cm21 and a parallel magnetic field of 7.0 gauss guided electrons and ions to two position-sensitive micro-channel plate detectors with delay line readout (see http://www.roentdek.com for details on the detectors). The fields were adjusted such that electrons of up to 15 eV kinetic energy could be collected with a 4p solid angle. Owing to the long ion drift arm the spectrometer accepted only N21 or CO1 (in case their kinetic energies were higher than 5 eV), which were emitted within 10u with respect to the spectrometer axis. The light polarization was circular in the case of N2. The data were recorded in list mode. For each ionization event we recorded the positions of impact and times of flight of all registered particles. This allowed us to extract the very weak dimer signal from our data, because the dimer fraction in our beam was only 0.1% to 1%. Thus most recorded ions and electrons resulted from ionization of the monomer. We can identify the ICD channel by selecting only events in which two N21 (or two CO1) ions with equal and opposite momentum occur. This back-to-back emission is a unique signature of the final step of Coulomb explosion following ICD (see equation (3)). Received 8 May 2012; accepted 28 November 2013. Published online 22 December 2013. 1. 2. 3. 4. 5. 6. 7.
Cederbaum, L. S., Zobeley, J. & Tarantelli, F. Giant intermolecular decay and fragmentation of clusters. Phys. Rev. Lett. 79, 4778–4781 (1997). Jahnke, T. et al. Ultrafast energy transfer between water molecules. Nature Phys. 6, 139–142 (2010). Mucke, M. et al. A hitherto unrecognized source of low-energy electrons in water. Nature Phys. 6, 143–146 (2010). Kim, H.-K. et al. Enhanced production of low energy electrons by alpha particle impact. Proc. Natl Acad. Sci. USA 108, 11821–11824 (2011). Barth, S. et al. Observation of resonant interatomic Coulombic decay in Ne clusters. J. Chem. Phys. 122, 241102 (2005). Aoto, T. et al. Properties of resonant interatomic Coulombic decay in Ne dimers. Phys. Rev. Lett. 97, 243401 (2006). Jahnke, T. et al. Experimental separation of virtual photon exchange and electron transfer in interatomic Coulombic decay of neon dimers. Phys. Rev. Lett. 99, 153401 (2007).
20.
21. 22. 23.
24.
25.
26. 27. 28. 29.
Ueda, K. et al. Interatomic Coulombic decay following the Auger decay: experimental evidence in rare-gas dimers. J. Electron Spectrosc. Relat. Phenom. 166–167, 3–10 (2008). Grieves, G. A. & Orlando, T. M. Intermolecular Coulomb decay at weakly coupled heterogeneous interfaces. Phys. Rev. Lett. 107, 016104 (2011). Marburger, S., Kugeler, O., Hergenhahn, U. & Mo¨ller, T. Experimental evidence for interatomic Coulombic decay in Ne clusters. Phys. Rev. Lett. 90, 203401 (2003). Jahnke, T. et al. Experimental observation of interatomic Coulombic decay in neon dimers. Phys. Rev. Lett. 93, 163401 (2004). O¨hrwall, G. et al. Femtosecond interatomic Coulombic decay in free neon clusters: large lifetime differences between surface and bulk. Phys. Rev. Lett. 93, 173401 (2004). Boudaı¨ffa, B., Cloutier, P., Hunting, D., Huels, M. A. & Sanche, L. Resonant formation of DNA strand breaks by low-energy (3 to 20 eV) electrons. Science 287, 1658–1660 (2000). Hanel, G. et al. Electron attachment to uracil: effective destruction at subexcitation energies. Phys. Rev. Lett. 90, 188104 (2003). Gokhberg, K., Kolorencˇ, P., Kuleff, A. I. & Cederbaum, L. S. Site- and energy-selective slow-electron production through intermolecular Coulombic decay. Nature http:// dx.doi.org/10.1038/nature12936 (this issue). Pradhan, A. K. et al. Resonant X-ray enhancement of the Auger effect in high-Z atoms, molecules, and nanoparticles: potential biomedical applications. J. Phys. Chem. A 113, 12356–12363 (2009). Howell, R. W. Auger processes in the 21st century. Int. J. Radiat. Biol. 84, 959–975 (2008). Botting, S. K. & Lucchese, R. R. Auger decay of the C 1s -. 2p* excitation of CO. Phys. Rev. A 56, 3666–3674 (1997). Aquilanti, V., Bartolomei, M., Cappelletti, D., Carmona-Novillo, E. & Pirani, F. The N2–N2 system: an experimental potential energy surface and calculated rotovibrational levels of the molecular nitrogen dimer. J. Chem. Phys. 117, 615 (2002). Havenith, M., Petri, M., Lubina, C., Hilpert, G. & Urban, W. J. IR spectroscopy of (CO)2 using concentration-frequency double modulation in a supersonic jet expansion. J. Mol. Spectrosc. 167, 248–261 (1994). Meredith, A. W. & Stone, A. J. An ab initio and diffusion Monte Carlo study of the potential energy surface of the CO dimers. J. Phys. Chem. A 102, 434–445 (1998). Havermeier, T. et al. Single photon double ionization of the helium dimer. Phys. Rev. Lett. 104, 153401 (2010). Baltzer, P. et al. Inner-valence states of CO1 between 22 eV and 46 eV studied by high resolution photoelectron spectroscopy and ab initio CI calculations. J. Phys. At. Mol. Opt. Phys. 27, 4915 (1994). Kimura, M. et al. Efficient site-specific low-energy electron production via interatomic Coulombic decay following resonant Auger decay in argon dimers. Phys. Rev. A 87, 043414 (2013). O’Keeffe, P. et al. The role of the partner atom and resonant excitation energy in interatomic Coulombic decay in rare gas dimers. J. Phys. Chem. Lett. 4, 1797–1801 (2013). Kimura, K. et al. Controlling low-energy electron emission via resonant-augerinduced interatomic Coulombic decay. J. Phys. Chem. Lett. 4, 1838–1842 (2013). Do¨rner, R. et al. Cold target ion momentum spectrocopy: a ‘momentum microscope’ to view atomic collisions dynamics. Phys. Rep. 330, 95–192 (2000). Ullrich, J. et al. Recoil-ion and electron momentum spectroscopy: reactionmicroscopes. Rep. Prog. Phys. 66, 1463 (2003). Scho¨ffler, M. S. et al. Matter wave optics perspective at molecular photoionization: K-shell photoionization and Auger decay of N2. New J. Phys. 13, 095013 (2011).
Acknowledgements This work was supported by the Deutsche Forschungsgemeinschaft and the Deutscher Akademischer Austauschdienst. We thank the staff of the Advanced Light Source for excellent support during the beam time. This work was supported by the Director, Office of Science, Office of Basic Energy Sciences, and by the Division of Chemical Sciences, Geosciences, and Biosciences of the US Department of Energy at the Lawrence Berkeley National Laboratory under contract number DE-AC02-05CH11231. The Advanced Light Source is supported by the Director, Office of Science, Office of Basic Energy Sciences, of the US Department of Energy under contract number DE-AC02-05CH11231. We thank K. Gokhberg and L. Cederbaum for suggesting this experiment and for many discussions. M.S.S. thanks the Alexander von Humboldt foundation for financial support. Author Contributions All authors contributed to the experiment. F.T. and T.J. performed the data analysis. All authors contributed to the manuscript. Author Information Reprints and permissions information is available at www.nature.com/reprints. The authors declare no competing financial interests. Readers are welcome to comment on the online version of the paper. Correspondence and requests for materials should be addressed to T.J. ([email protected]) or F.T. ([email protected]).
0 0 M O N T H 2 0 1 3 | VO L 0 0 0 | N AT U R E | 3
©2013 Macmillan Publishers Limited. All rights reserved
