Nederlands

Experimental arrangement for the study of plasma breakdown

Measurements of electric field distributions during breakdown are made by means of Stark spectroscopy. To improve the temporal resolution of the Stark spectroscopy technique, a pulse compression system, based on Stimulated Brillouin Scattering can be included. A schematic of the complete experimental arrangement is shown in figure 1.

Figure 1: Experimental arrangement for Stark spectroscopy on plasma breakdown
The experimental arrangement consists of four main parts:
  • Laser system for Stark spectroscopy
  • Vacuum chamber with electrode structure and power supply
  • Fast ICCD camera
  • Stimulated Brillouin scattering (SBS) pulse compression system

Each part of the system will be described separately in following sections.
 

Stark Spectroscopy
The measurement of electric field distributions by Stark spectroscopy is based on measurements of DC Stark effects in the discharge gas. The mixing of energy levels of an atom due to a (large) electrical field is known as the Stark effect. By measuring these mixing effects and comparing them with theoretical calculations, the electric field in the discharge gas can be determined. More information on the Stark effect and Stark spectroscopy can be found in references [1-3].

The DC Stark mixing effects are measured using a 2+1 photon excitation scheme shown in figure 2.

Figure 2: 2+1 photon excitation scheme
This 2+1 photon excitation scheme requires two lasers, one with a constant wavelength and one with a tunable wavelength. The first laser is set up with a constant wavelength, suitable to excite atoms from the ground state to an excited state by two photon excitation. These excited states will de-excite to lower states emitting fluorescence light. This fluorescence is detected outside the plasma by a ICCD camera. The second laser can excite the excited atoms even further to high Rydberg states. If the second laser wavelength matches this excitation, the fluorescence light will decrease, because less atoms are available in the first excited state. Absorption of photons from the second laser can be observed as a 'dip' in the fluorescence signal. Because these Rydberg states have a high principal quantum number, they are very sensitive to electric fields, there is a large effect of Stark mixing. By scanning the wavelength of the second laser and observing the fluorescence light one can make a plot of fluorescence intensity versus laser wavelength. The transitions from the excited states to the Rydberg states show up as dips in the fluorescence spectrum.

On the other hand, the Stark effect can be calculated theoretically. Our calculation method is based on the method introduced by Zimmerman et al. [4] and Saloman and Kelleher [5]. It is based on solving the Schrödinger equation including the electric field perturbation by a numerical diagonalization of the total Hamiltonian, H = H0+HStark, for a well-chosen truncated basis. The matrix H0 consists of the energies of the levels at zero electric field, HStark contains the Stark effects. An example of the results of such a calculation is shown in figure 3.

Figure 3: Stark map showing the (partial) energy level structure of a xenon atom as function of the applied electric field.
These so-called Stark maps show the energies of the different levels as function of electric field. From these Stark maps, fluorescence spectra for different values of the electric field in the discharge are calculated and compared with the measured spectrum. In this way, the electric field distributions of the discharge can be determined.
It is clear that Stark spectroscopy requires narrowband lasers in order to resolve the Stark structure of the atoms. The time resolution of this method is determined by the pulse length of the scanning laser. Normal tunable pulsed systems have a pulse length of typically 7 ns. In our project better time resolution is required to resolve the details of the structure of plasma breakdown. This can be achieved by SBS pulse compression as described in the last section. The SBS pulse compression puts an extra requirement on the laser system. The second tunable laser, which is compressed, has to be a single-mode laser. Normal pulsed tunable lasers do not meet this requirement and a special system had to be developed. In our system a CW ring dye laser (pumped by an Ar+ laser) is used as a seed laser. This laser is then amplified in a three stage pulsed dye amplifier, pumped by a seeded Nd:YAG laser. The results is a single-mode tunable output pulse, which can be compressed by means of SBS.

Breakdown conditions
The choice of the specific breakdown conditions had to meet certain requirements.
  • It has to be operated in noble gas. This is required for the theoretical calculation of the Stark spectra.
  • The breakdown has to be reproducible in time and position. Multiple breakdown cycles are necessary for a complete Stark spectroscopy measurement.
  • The electrodes should have smooth, geometrical shapes in order to make comparison with simulation more straightforward.
  • Lower gas pressures lead to lower velocities of ionisation fronts. This makes the measurements easier to perform.

A design for the breakdown conditions and geometry has been made, meeting all imposed requirements. The breakdown under study has the following properties.
  • 2 stainless steel electrodes in vacuum vessel
  • Cylindrically symmetric electrodes
  • electrode tip curvature is 4 mm, height ~ 40 mm
  • discharge gas: argon at 4 mbar pressure
  • distance between electrodes is ~3.3 mm (variable)
  • applied voltage: 100 ms square positive pulse of ~500 V at a repetition rate of 500 Hz.
        

Figure 4: Images of plasma breakdown
ICCD camera system
For the detection of the fluorescence light, a fast ICCD camera is used. The most important specifications for our camera are:
  • 1024 x 1024 pixels
  • 13 x 13 mm pixel size
  • 3 ns minimum optical gate width
  • up to 1100 x gain

This camera is fast and sensitive enough to achieve the required spatial and temporal resolution of our measurements.
 

Stimulated Brillouin scattering pulse compression

Stimulated Brillouin scattering is a three-wave parametric mixing process coupling two radiation fields (an incident pump beam and a reflected Stokes beam) with an acoustic wave in a medium. Details of the SBS process can be found in textbooks on non-linear optics, i.e. He [6] and Newell [7]. The SBS process can be used to obtain compression of a high-energy laser pulse [8]. The required experimental arrangement is shown in Figure 5.

Figure 5: Experimental arrangement for SBS pulse compression

The generator-amplifier cell is in our setup a glass tube filled with filtered water (particles < 0.2 mm). The water is the active SBS medium. The beam splitter and Fresnel rhomb are included to separate the compressed beam from the incoming beam.

The fundamental concept is as follows:

  • The incoming pump laser beam passes trough the water-filled tube.

  • The beam is focused back in the water tube by a focusing mirror.

  • In the focus of the high-power laser beam spontaneous Brillouin scattering of the leading edge of the laser beam occurs. A phase- conjugate beam is produced.

  • The back-scattered part of the Brillouin radiation, called Stokes radiation, interacts with the remaining incident beam causing an acoustic wave in the water.

  • The remainder of the incident beam, the acoustic wave and the reflected Stokes radiation interact, resulting in Stimulated Brillouin Scattering (SBS). The incident beam is reflected at the interaction point, the Stokes beam is amplified and temporally compressed traveling back through the water.

Figure 6: SBS pulse compression
Laser pulses from standard pulsed lasers have a pulse duration of typically 5-10 nsec. With the SBS pulse compression system, pulse widths of 200-500 ps can be obtained.

A requirement for SBS pulse compression is a single-mode Gaussian input laser pulse. A multi-mode laser beam will have intensity spikes in the laser pulse, due to mode beating. These spikes disturb the SBS process, resulting in optical breakdown in the water. The necessity of a single-mode input pulse defines specific requirements for the laser system. Special precautions have to be made to ensure that the laser pulses used for sub-nsec Stark spectroscopy are single-mode.

For our pulse compression system we had contact with the Laser Center of the Vrije Universiteit of Amsterdam. They have lots of experience with SBS pulse compression [9-11] and provided us with very useful advise.
References
[1] P.H. Heckman, Introduction to the spectroscopy of atoms, North-Holland, Amsterdam, 1989.
[2] U. Czartnetzki et al., Phys. Rev. Lett., 81(21):4592-4595,1998
[3] V.P. Gavrilenko et al, Phys. Rev. E, 62(5):7201-7208, 2000.
[4] M.L. Zimmerman et al, Phys. Rev. A, 20(6):2251–2275, 1979.
[5] D.E. Kelleher and E.B. Saloman, Phys. Rev. A, 35(8):3327–3338, 1987.
[6] G.S. He, Physics of nonlinear optics, World Scientific, London, 1999.
[7] A.C. Newell and J.V. Moloney, Nonlinear optics, Addison-Wesley, Amsterdam, 1992.
[8] D.T. Hon, Opt. Lett., 5(12):516-518, 1980.
[9] S. Schiemann et al., IEEE J. Quantum Electron., 33(3):358-366, 1997.
[10] S. Schiemann et al., IEEE J. Quantum Electron., 34(3):407-412, 1998.
[11] D. Neshev et al., Appl. Phys. B, 68(4):671-675, 1999.

top