Plasma control of the emission spectrum of mercury-noble-gas discharges

For certain applications, it would be convenient to be able to change the colour temperature and the luminous flux of artificial lighting independently. A promising candidate for this purpose is the fluorescent lamp. This thesis describes the investigation of two variable colour fluorescent lamps. The mercury-noble-gas discharge in these lamps produces both mercury and noble-gas radiation, in contrast to the discharge in normal fluorescent lamps that produces only mercury radiation. The additional noble-gas radiation can be used to change the colour of the lamp. The research goal is to obtain a better understanding of the processes involved in the production of noble-gas radiation in mercury-noble-gas discharges. The two most important questions to be answered are: which discharge processes and conditions are important for the production of noble-gas radiation, and how can we produce this noble-gas radiation efficiently? To answer these questions, we performed an experimental study. The investigation comprises another important goal as well: the improvement and the development of diagnostic techniques. The first discharge we studied is a fluorescent lamp with an additional capacitively coupled radio frequency (ccrf) discharge. We performed an optical and electrical characterisation of this hybrid lamp. For the electrical characterisation, we developed a new stray impedance characterisation technique. The results of the electrical and optical measurements show that the additional ccrf discharge is inefficient and that the intensity of the noble-gas spectral lines is low. Another important result is the fact that the mercury density significantly influences the emission spectrum. This result led us to the introduction of a new lamp: the depleted discharge lamp. This is a discharge lamp in which the emission spectrum is controlled by varying the mercury density. A promising way to change the mercury density is using mercury depletion due to radial cataphoresis. The amount of noble-gas radiation produced by this lamp is much higher than the amount produced by the ccrf discharge. Therefore, the research narrowed down to the investigation of the depleted discharge. This discharge has been investigated using the following diagnostic techniques: optical emission spectroscopy, electrical measurements, ultraviolet absorption measurements, spatially- and time-resolved emission spectroscopy and Thomson scattering. The first emission measurements have been performed to find the conditions where the influence of radial cataphoresis on the emission spectrum is significant. It appears that both the discharge current and the mercury pressure can be used to control the emission of noble-gas radiation. The effect of mercury depletion is observed indirectly in the emission spectrum. In order to proof the mercury depletion directly, we measured the mercury ground state density profile using ultraviolet absorption. The results of these measurements are compared to the results of the spatially resolved emission measurements, which show the regions where noble-gas radiation is produced. The noble-gas radiation is produced only when the mercury density is low. The time-resolved emission measurements show the timeevolution of the spectral lines of mercury and neon after the discharge is switched on. The intensity of the mercury lines decreases with time, while the intensity of the neon lines increases. The characteristic time for this decrease and increase is of the same order of magnitude as the characteristic time for mercury diffusion. The picture of the discharge as created by the measurements mentioned above misses the properties of the bulk of the electron gas. These properties can be measured using Thomson scattering. There are two problems when Thomson scattering is performed in fluorescent lamps. At first, the electron density is rather low, which results in extremely low signal levels. The second problem is the stray light in the set-up. The first problem is solved by using a powerful laser and a sensitive detector. The second problem is solved by using a sodium vapour absorption cell in the detection branch of the set-up, which absorbs the stray light. For this option to work, the scattering experiments should be performed at one of the two resonant lines of sodium. This necessitates the use of a dye laser. However, the spectral purity of dye lasers is not sufficient for this experiment. The broadband amplified spontaneous emission (ASE) produced by the laser should be reduced. For this, we designed an ASE filter. This filter consists of two spatial filters and twenty dispersion prisms. The design of the ASE filter is chosen is such a way that the transmission is maximised and the spectral width is minimised. The spectral width of the sodium cell is matched to the width of the ASE filter by adding argon to the cell. The argon atoms broaden the absorption sodium lines. The combination of the sodium absorption cell and the dye laser with ASE filter results in a stray light reduction of more than six orders of magnitude. This reduction is sufficient to perform Thomson scattering in fluorescent lamps. The electron density and temperature have been measured in an argon-mercury and a neonmercury discharge. The results of the Thomson measurements combined with the emission and absorption measurements result in the following picture of the radial cataphoresis process and the production of noble-gas radiation. A high current and a low mercury pressure result in a high ionisation degree of mercury. This high ionisation degree results in significant depletion of mercury. As a result, less inelastic collisions of electrons and mercury atoms will occur. The amount of electrons capable of exciting the noble-gas atoms will increase. Therefore, the discharge will start producing noble-gas radiation. Regarding the application of the studied discharges, we can conclude that both the sheath regions of the ccrf discharge and the radial cataphoresis process can be used to produce noble-gas radiation in mercury-noble-gas discharges. The efficiency of the ccrf discharge for the production of noble-gas radiation is too low. On the other hand, the efficiency of the depleted discharge lamp offers a much better perspective for commercialisation. For this lamp, amplitude modulation offers the opportunity to control the colour and the luminous flux independently