SFB 616


Pictures SFB616

 Project B3:
The Investigation of Fast Dissipation Processes
Using Femtosecond X-ray Spectroscopy

 During the last several years new methods for the generation of X-rays have become available. These methods take advantage of the extremely high light intensity of femtosecond laser pulses to produce X-ray radiation. The outstanding novel feature of these laser-produced X-rays is their extremely short pulse duration, typically a few hundred femtoseconds. Femtosecond X-ray pulses now make it possible to follow directly the temporal evolution of structural changes associated, for example, with molecular and lattice vibrations, the changing or breaking of chemical bonds, structural phase transitions and the like.

The objective of our project is to study the excitation of lattice vibration resulting from the energy dissipation in electronically excited solids. The excess energy of the electrons is transferred to the lattice and leads to the excitation of lattice vibrations. The atomic motion associated with lattice vibrations can be monitored using X-ray diffraction because in the presence of atomic motion the X-ray diffraction efficiency is reduced by the Debye-Waller factor. Measurements of the Debye-Waller factor therefore reveal the degree of vibrational excitation of the lattice.


Experimental method
Fig 1: Optical pump/X-ray probe scheme for femtosecond time-resolved X-ray diffraction.

 We use an optical pump/ X-ray probe scheme to excite and probe lattice vibrations on a femtosecond time scale (Fig. 1). The X-ray probe pulses are generated by focusing a 100 mJ femtosecond laser pulse on a solid target to produce extremely hot microplasma. The incoherent keV X-rays from the plasma are collected by an X-ray mirror and focused onto the sample to be studied. A fraction of the laser pulse - after suitable frequency conversion to match the energy levels of the sample - is used for the electronic excitation. The time delay between the optical excitation pulse and the X-ray probe pulse can be easily controlled by means of an optical delay line. Thus, it is possible to electronically excite the sample and to record, for example, an X-ray diffraction pattern after some well defined delay time. The build-up of lattice vibration would then manifest itself through the time-dependence of the Debye-Waller factor, i.e. a decrease of the Debye-Waller factor with time.



Fig 2: Ge crystalline layers 150nm thickness hetero-epitaxially grown on Si (111). The angle of incidence of the X-ray probe pulse is chosen for symmetric diffraction from (111) lattice planes. By rotation of the sample about the surface normal asymmetric diffraction from (311) and (400) can also be obtained.

 We have carried out measurements of the time-dependent Debye-Waller effect in Ge. This semiconductor material has a secondary direct band gap near 1.5 eV and can therefore be readily excited using femtosecond laser pulses from a Ti-sapphire laser system at 800 nm.

Due to the strong optical absorption of intense laser light the achievable electronic excitation is often limited to a relatively thin layer near the surface. On the other hand, X-rays tend to penetrate much deeper into the material than laser light. Thus, there is generally the problem of matching the optical and X-ray penetration depths in optical pump/X-ray probe experiments.

We have used hetero-epitaxially crystalline films of Ge of 150 nm thickness grown on Si (111) substrates (Fig. 2). The key point is that in these samples the lattice constants of Ge and Si are not matched. Thus, the X-ray diffraction from the Ge-film and from the Si-substrate can therefore be distinguished because the Bragg angles for diffraction from the surface film and from the substrate are different.


Fig 3: Calculated Debye-Waller factor as a function of the lattice temperature for Ge. Diffraction from (111), (311) and (400) lattice planes.

 The crystal orientation of the Ge-layer is [111]. Thus the symmetric Bragg diffraction monitors lattice planes of type (nnn), for example (111). However, it is also possible to record asymmetric diffraction from (311) and (400) lattice planes by judiciously rotating the sample to fulfill the corresponding Bragg condition. Fig. 3 shows the Debye-Waller factor as a function of the lattice temperature for the three different lattice planes of Ge. Note that for lattice planes with higher (hkl) indices the drop in the diffraction efficiency is greater.


First results

Fig 4: X-ray diffraction signal (integrated reflectivity) from (400)-lattice planes of Ge as a function of delay time between X-ray probe pulse and laser excitation pulse. Results for two different laser pump fluences are shown.

 An example of the experimental results is depicted in Fig. 4. The measured diffraction efficiency of the (400) diffraction is plotted as a function of the delay time between the X-ray pulse and the laser excitation pulse. Negative delay times correspond to the diffraction efficiency of the unperturbed sample. Data for two different laser pump fluences are shown. Following laser-excitation the diffraction efficiency drops down to 96% and 93% for a laser fluence of 29 mJ/cm 2 and 35 mJ/cm 2 , respectively. These values correspond to a rise in the lattice temperature of 450 K and 550 K, respectively. For both laser fluences the experimental data can be fitted with an exponential function with a time constant t = 1.1 ps. Thus we have measured the electron-phonon energy transfer time for Ge to be 1.1 ps.




 We have demonstrated the feasibility of femtosecond time-resolved X-ray diffraction to study the excitation of lattice vibrations associated with the dissipation of electronic energy.

Electronic energy relaxation has previously been studied extensively, for example by measuring time-resolved spectra of the optical absorption, optical reflectivity or photoemission. These experiments primarily provided the transient distribution functions of the electronic energy and revealed how the electronic energy drains off the electronic system. On the other hand, the X-ray diffraction experiments show a new, complementary aspect of the electron-phonon interaction process, that is the appearance of the energy in the vibrational modes of the lattice .

 X-ray diffraction using laser-driven sources is subject to a number of restrictions. For example, the X-ray flux per solid angle is relatively low, and so are the X-ray scattering cross sections. Some of the disadvantages can be overcome by probing the evolution of the system under study with femtosecond electron pulses using electron diffraction. X-ray pulses and electron pulses for performing femtosecond diffraction experiments represent powerful tools for studying ultrafast structural dynamics in atomic matter, in particular energy dissipation phenomena.

 This project is in collaboration with projects B4, B2, B5, A2, and A1.


Internationaler Workshop 2008

Workshop 2008

B3 Ligges et al.
PDF (0.4 MB)

Workshop 2006

Workshop 2006
B3 Nicoul et al.
 PDF (1.6 MB)

Begehung 2005

Begehung 2005
B3 Shymanovich et al.
 PDF (0.6 MB)