Collaboration within Munich Centre for Advanced Photonics* (DFG Cluster of Excellence)

MPQ:	T. Schaetz, S. Kahra, G. Leschhorn, T. Dou, W. Fuss, W. Schmid (Area C2.2) E. Botschafter, A. Schiffrin, R. Ernstorfer, R. Kienberger (Area A.1)
LMU:	R. deVivie-Riedle, M. Kowalewski (Area C2.6)

Time-resolved diffraction of single molecules: structure of short-lived intermediates

We present the preparation of one single molecular ion as a target for single-molecule imaging with atomic resolution. The methods of sympathetic laser cooling, Coulomb crystallization and state preparation allow for focusing anticipated MAP* x-ray pulses of high intensity to considerably better than 1µm on a single molecule, providing very well defined conditions in a nearly background free environment. A launched pump-probe experiment with 4fs UV-pulses investigates the methodology and incorporates all building blocks for target preparation in future single molecule diffraction experiments.

introduction

The investigation of the dynamics within a molecule during a structural change on the relevant timescale (<10-15 fs) gets into reach due to the development of MAP*-light sources that will provide pulses of more than 10¹¹ photons at wavelengths promising a resolution on atomic length scales. We should become capable of recording a diffraction pattern from a single molecule without the need of crystalline periodicity. The three-dimensional, time-resolved structure could be retrieved, if the destroyed molecule can be replaced by identically prepared copies, pounded one after the other.

Besides the work necessary to supply the photons at the required intensities, we have to provide reproducible and precise methods at sufficient repetition rates to prepare and handle the molecular target. Therefore, we try to transfer the unique operational fidelities for handling trapped ions in quantum information processing into the field of attosecond experiments with MAP-light sources and their possibilities of ultimate control of laser fields.

The laser sources of our MAP-collaborators A.1. recently achieved unique specifications (sub 4 fs UV-pulses of 1 µJ at 3 kHz). One could use these pulses, for example to investigate neutral molecules in gas phase. Large Coulomb crystals of molecular ions could provide spectroscopic data. Using the methods of quantum information processing based on trapped ions for the preparation and control of external and internal degrees of freedom[1,2,3] could allow merging the individual experimental advantages to provide single molecular ions as targets with unique operational fidelities.

objectives

We designed our system to enable diffraction pump-probe experiments, e.g. photo-triggered isomerization recorded by x-ray pulses, on single molecular ions to elucidate the structural changes during chemical reactions. We will handle undisturbed single molecular ions (m<104 u) by sympathetically laser cooling the external degrees of freedom in a Coulomb crystal confined in a Paul trap, cooling their ro-vibrational degrees of freedom by buffer-gas- or even direct laser cooling close to their ground states, separate single molecular ions into a diffraction zone by controlling electrical fields on electrodes and light pressure and 3D-orientate them in space by additional light pulses.

We will investigate charged molecules and plan to append charged atoms for trapping, not disturbing the dynamics of the molecule or to protonate molecules, simulating in some cases a natural surrounding. Our generic tool should enable the almost background-free investigation of dynamical effects by x-ray diffraction, driven by the identical but split laser pulse on a sub fs timescale. Originally, we planned to bridge the gap until we get access to the x-ray pulses with electron pulses to investigate structural changes in simple molecular ions in collaboration with group C.2.1. Much to our regret, it was not possible up to now to provide these electron pulses. Since we had reached our anticipated milestones, we searched and found theoretical (C.2.6) and experimental (A.1) support to provide the proof of principle experiment for our originally proposed method of pump-probe experiments on single molecules. In the future, we hope to just have to exchange the laser sources.

Short term objective: proof of method for single molecule imaging

In close collaboration with the MAP-theory group (C.2.6.) we investigated the possibilities to perform the proof of method experiment depicted in Fig. 2. A pump pulse is expected to transfer the vibrational wave packet from the electronic ground (X) into the excited state (A) with a natural life time of ~4 ns. A probe pulse of variable time delay (0-2000 fs) should allow for the transfer from the excited state (A) to the dissociative channel (C). The dissociation of 24MgH+ into neutral 24Mg and H+ should lead to the loss of both particles the disappearance of the dark molecular ion providing the detection of the reaction.

experimental progress

actually: Merging the experimental setups

We currently merge our Ultra-High-Vacuum (UHV) apparatus via a differential pumping section with the MAP* beam line. The unique coatings of our mirrors had been designed and will be fabricated by the MAP-service centre to match the spectral demands for the anticipated experiment.
In parallel: set up of Electro Spray Ionization (ESI) source
In a separate setup, we reanimated a ElectroSpray Ionisation source and achieved sufficient flux of (generic) single molecular ions. Currently we extend the source with a mass filter section and a rf-lens to allow merging the ESI with our ion guide setup.

April 2009: Transfer of experimental setup to the MAP* beam line

We transferred (one floor, approx. 150m) our fully functional UHV apparatus including the optical- and laser setup to the MAP-beam line (AS4b at MPQ) to get access to 4 fs pulses at a wavelength of 290 nm.

Archiv of technical/experimental progress (project started in April 2006)

End of 2006: We are able to show the separation of individual molecular ions of the two-component Coulomb crystal into a separate interaction zone within an additionally segmented linear setup a potentially important requisite for future higher repetition rates.
Already early 2007: We set up a laser system for Doppler cooling of 138Ba+ for efficient sympathetic laser cooling to investigate molecular ions of larger mass/charge ratios (e.g. like the envisioned ESI-protonized azobezene).
Until December 2007: We set up a 500 mm long radio frequency ion guide including additional control electrodes to allow for axial confinement within three trapping regions as depicted in Fig.1B. Until June 2008: We photo-ionize 24Mg atoms in the loading zone of our guide and breed in a photochemical reaction 24MgH+ molecules via controlled input of H2.
Until September 2008: We transfer the (molecular) ion along the guide through two differential pumping stages into a UHV chamber (210^-10 mbar) where we sympathetically laser cool the external degrees of freedom close to the Doppler limit of 1mK. We show that it is possible to reload the Coulomb crystal with (molecular) ions to compensate future loss due to x-ray photon interaction.
Until Dezember 2008: The molecular ions get imbedded in the crystalline structure of the directly laser cooled 24Mg+ and shifted to their intended localization via light pressure. (The internal vibrational degrees of freedom are supposed to be in the ground state).
Until February 2009: Our detection system allows for a spatial resolution below 1µm and a close to 100 % detection efficiency for the (molecular) ion (see Fig. 1A).

Outlook

We envision the investigation, whether the bandwidth (approximately 30 nm) of the 4 fs pulses allows to provide a coupling between the bound molecular states to realize scattering rates sufficient for direct detection of the fluorescence light of a single molecular ion. Theoretical predictions suggest shifting the spectrum sufficiently red to allow for a related direct laser cooling of the internal degrees of freedom of the molecule. In addition, we might be able to transfer the population of the ground state (X) directly via a two-photon transition to the dissociative state (C) and to apply the second (probe) pulse to transfer the population to another dissociative channel (B). The regained 24Mg+ might still be trapped therefore providing a new detection channel where molecular (dark) ions do not disappear put turn into bright ions again. For the farer future, this might allow to think of experiments, where the charged product of a reaction could be kept and further investigated.

As long-term goal with x-ray pulses of sufficient intensities, we want to adapt the proof of principle experiment to the specific needs. An important prerequisite is the envisioned field free orientation of the molecular ions via laser pulses.

A) Single molecular (dark) ion in our current ion-trap setup (indicated by the arrow). The bright dots represent directly laser cooled 24Mg+, sympathetically cooling the molecular ion, doped into the Coulomb crystalline structure after being shuttled along 500 mm of our ion guide. This generic method provides molecular ions as targets with spatial resolution better than 1µm. B) Our setup, with a) ion guide b) electrodes providing axial confinement c) atomic ovens d) electron gun e) differential pumping sections f) CCD-camera system g) (still separate) separation trap h) gas source i) (still separate) ESI-source.

A) Potential energy curves for the electronic states of 24MgH+, computed by our collaborators in C.2.6. and in good agreement with those of the group of M. Drewsen (Aarhus). For time resolved pump-probe experiments on single molecular ions, we plan to excite the vibrational wave packet from the ground (X) to the excited state (A) via a pump (4 fs) UV pulse. A second pulse with adjustable delay will probe the related probability to further excite the oscillating wave packet into the dissociative channel (C). B) The probability to reach state (C) for the spectrum of the pulse is expected to depend on the inter-nuclear distance. Thus, we expect the oscillation of the wave packet to be mapped on the probability for dissociation and therefore on the probability for the disappearance of the dark spot in Fig. 1A.

[1] H. Schmitz et al., The "arch" of simulating quantum spin systems with trapped ions, Appl. Phys. B 95, 195 (2009).
[2] A. Friedenauer et al., Simulating a quantum magnet with trapped ions, Nature Physics 4, 757 (2008).
[3] R.Schuetzhold et al., Analogue of cosmological particle creation in an ion trap, Phys. Rev. Lett. 99, 201301 (2007).