Analytical model for low-pressure gas discharges: application to the Hg + Ar discharge.

A general technique for analyzing complicated gas discharges has been developed and applied to the Hg + Ar (fluorescent lamp) discharge. The theoretical model includes electron excitation and deexcitation, two-state ionization through a saturated metastable level, and proper treatment of the self-absorption of the resonance radiation. The analysis yields simple analytic expressions for the electron temperature, the resonance radiation, and the electric field. When applied to Hg + Ar discharges, these analytic expressions yield good quantitative agreement with the available absolute data on the dependence of the electron temperature, the Hg 2537-A radiation, and the electric field on mercury pressure and current.

Direct current cataphoretic effects on Hg 2537-A radiation from Hg + Ar discharges (dc fluorescent lamps).

The Hg 2537-A intensity from Hg + Ar discharges was measured as an independent function of mercury vapor pressure (as influenced by a cold spot temperature) at various constant currents and tube radii with ~4-Torr argon. As a function of mercury pressure, the intensity rises to a peak (which defines an optimum mercury cold spot temperature) and then decreases with further increase in mercury pressure due to the combination of self-absorption and electron deexcitation. The behavior of the optimum mercury cold spot temperature is dependent upon ac or dc conditions. For ac, the optimum mercury pressure is ~7 mTorr (corresponding to a Hg cold spot temperature of ~40 degrees C) and comparatively insensitive to current. By contrast, the optimum mercury cold spot temperature for the dc case is dependent upon whether the anode end or cathode end is cooled. The dc lamp with anode end cooled yields an optimum mercury cold spot temperature less than 40 degrees C and decreases with increasing current, while the optimum with cathode cooled is greater than 40 degrees C and increases with increasing current. We believe that the peak intensity always occurs at the same real mercury density (because that determines the self-absorption), but the Hg cold spot temperature required to achieve this density is affected by dc cataphoretic pumping phenomena.

Effect of Electron Deexcitation and Self-Absorption on the Intensity of the Hg 2537-A Radiation from Hg + Ar Discharges (ac Fluorescent Lamps).

The intensity of the Hg 2573-A radiation from Hg + Ar discharges was measured as an independent function of mercury pressure (0.2-50 mTorr), ac current (50-2100 mA) and tube radius (0.79 cm and 1.27 cm) at a constant Ar pressure of ~4 Torr. For various constant mercury pressures, the Hg 2537-A intensity initially rises linearly with increasing current, but then tends to bend over and approach an asymptotic limit. The nonlinear, asymptotic behavior is due to electron deexcitation of the Hg 6(3)P(1) state at the higher currents in the presence of Hg 2537-A self-absorption. The Hg 2537-A intensity was also measured as a function of mercury pressure at various constant currents. The intensity rises to a peak (which defines an optimum Hg pressure) and then decreases with further increase in mercury pressure due to the combination of self-absorption and electron deexcitation. For high ac currents, the optimum Hg pressure is independent of current but varies inversely with the tube diameter. All this behavior is relevant to the problem of obtaining high efficiency from fluorescent lamps at high powers.

Initial afterglow of the self-absorbed hg 2537-a radiation from hg + ar discharges.

A limitation on the high frequency modulation of gas discharge lamps is the duration of the radiative afterglow which is often dilated by self-absorption. Introducing a foreign gas into the discharge alters the absorption line shape and width by collisions, thus reducing self-absorption and the afterglow decay time. This is a general technique for extending the high frequency modulability. For the experiments, the effect of variable argon pressures on the self-absorbed Hg 2537-A initial radiative decay time (tau) was measured from abruptly terminated discharges as independent functions of the mercury pressure (0.8-70 mTorr) and argon pressure (5-200 Torr). tau increases with the mercury density but is substantially reduced by the argon pressure in quantitative agreement with the theory of Holstein and Walsh and the concept that the initial decay is primarily limited by self-absorption for our range of variables. A detailed theoretical analysis indicates that there are several ways that additional argon reduces the Hg 2537-A self-absorption: (1) the Hg 237-A line gets broader simply because the additional argon atoms increase the Hgndash;Ar collision frequency; (2) adding argon causes the gas temperature to rise, and this drives the Hgndash;Ar collision frequency still higher; (3) the rise in gas temperature also causes an increase in the Hg 2537-A Doppler width. Thus, a general technique for substantially increasing the modulability of a gas discharge lamp emitting self-absorbed radiation has been theoretically and experimentally demonstrated. These results are consistent with our previous analysis performed on measurements of the Hg 2537-A intensity in these discharges relevant to fluorescent lamps. These phenomena may also be relevant in some gas lasers.

Effects of argon atoms on the self-absorption and the intensity of hg 2537-a radiation in hg + ar discharges.

A study was made of the feasibility of increasing the efficiency of fluorescent lamps at high powers by increasing the Hg 2537-A resonance radiation through a reduction of self-absorption. Specifically, we attempted to reduce the Hg 2537-A self-absorption by introducing a higher pressure of a foreign gas (argon) to alter the Hg 2537-A absorption line shape and width by collision broadening. The intensity of the Hg 2537-A line in Hg + Ar discharges was measured as an independent function of mercury pressure (0.7 mTorr to 27 mTorr), argon pressure (5 Torr to 400 Torr), and dc input power (5.5 W to 97 W). A detailed theoretical analysis indicates that there are four ways that additional argon reduces the Hg 2537-A self-absorption: (1) The Hg 2537-A line gets broader simply because the additional argon atoms increase the Hg-Ar collision frequency; (2) adding argon causes the gas temperature to rise and this drives the Hg-Ar collision frequency still higher; (3) the rise in gas temperature also causes an increase in the Hg 2537-A doppler width; (4) the additional argon changes the Hg 2537-A line shape from doppler dominated to a collision dominated profile. The experiments demonstrate, however, that no gain is achieved in the Hg 2537-A intensity with the addition of extra argon in spite of the beneficial effect on the self-absorption escape rate. This advantage is apparently offset by the argon's reduction of the electron energy which leads to fewer mercury atoms excited to the Hg 6(3)P(1) state.

Continuum emission spectra from long linear xenon discharge lamps with metallic halide additives.

The behavior of long linear xenon discharges with metallic halide additives in narrow-bore quartz tubes is described. For various lamp applications, these discharges may have several desirable characteristics: (1) controllable line emission or continuum emission as desired, with excellent color rendition, (2) reasonable efficiency, (3) fast startup (~20 sec), (4) elimination of the need for mercury and its pollution hazard, (5) possible efficient operation in the range between a few watts per linear cm and ~30 W/linear cm. The present lamps are in the laboratory stage, and several problems remain to be solved. It is argued that the continuum emission which sometimes appears in these xenon + metallic halide discharges (and also in commercial mercury arc lamps with certain metallic halide additives) is correlated with a high vapor pressure of the metallic halide additive. Thus, this continuum may be due to either molecular radiation or transitions involving free electrons originating copiously from the ionization of the metallic additive which has a comparatively low ionization potential. If a sufficiently high vapor pressure of any metallic halide can be achieved in the discharge, then a continuum emission seems to be a fairly common phenomenon. A scheme for achieving this objective is described.

Theoretical Analysis of the Behavior of Mercury plus Thallium Iodide Arc Lamps with Variable Mercury Loadings.

A theoretical analysis of the behavior of 400-watt Hg + TlI arc lamps with variable mercury loadings is performed. It is quantitatively shown that, as the Hg loading is doubled, the following effects may be predicted: (a) the effective arc temperature decreases from 5100 K to 4800 K, (b) the effective radius increases by ~10%, (c) the escape factor for the imprisoned Tl 5350 A line increases ~40%, (d) the power radiated in the Tl 5350 A line increases ~10%, and (e) the power radiated in the Hg visible line decreases ~15%. These results are in agreement with the experimental observations of Larson. These effects result in an increase in the lumen-per-watt ratings, which is of technological significance. Self-absorption is an important factor in the radiation of the Tl 5350 A line, and it is shown that the experimental observations are reasonably consistent with the assumption of local thermodynamic equilibrium.

The Effect of Humidity and Temperature Variations on the Behavior of Wire-to-Plane Coronas.

The effect of temperature and humidity on the current-voltage relationship and uniformity of positive and negative air coronas has been studied. Variations in temperature and absolute humidity seem to have a comparatively small effect on the behavior of coronas. By contrast, variations in the relative humidity have readily noticeable effects on the current-voltage relationship. At high voltages (positive or negative) the corona current decreases as the relative humidity increases due to ion and electron hydration effects. By contrast, both positive and negative coronas are initiated at lower voltages at high relative humidities, presumably due to the formation of miniscule water droplets with low ionization potential. The relative humidity also affects the uniformity of negative corona but not positive corona. Presumably the electron emitting properties of a negative wire are altered by moisture adsorbed on the wire surface at high relative humidities. By contrast, positive corona is not grossly affected because it is primarily a gas phase phenomena while negative corona is also sensitive to the electron emitting properties of the wire.

Acoustical resonances in modulated xenon and krypton compact arc lamps.

When the dc current through a compact arc lamp is modulated, are instabilities appear at discrete frequencies. At some of these frequencies, the instabilities are sufficiently severe to extinguish the lamp. The primary frequency at which the arc bows to the side and tends to extinguish has been studied as a function of bulb diameter and current for xenon and krypton lamps. A theoretical treatment of acoustical resonances is presented and it is quantitatively shown that the experimentally determined primary extinction frequency coincides with the primary acoustical resonance. Furthermore, the dependence of the primary extinction frequency on bulb diameter, temperature, pressure, and molecular weight fit the acoustical resonance theory.

The effect of thallium iodide on the arc temperature of Hg discharges.

The addition of metallic iodides to mercury arc lamps has proven to be an effective and practical method of altering the spectral output. If the excitation levels of the metal of the dissociated iodide are lower than the excitation levels of mercury, the mercury arc temperature is considerably reduced. Experimental evidence for this phenomenon is presented.

On the question of equilibrium in mercury + thallium iodide arc discharges.

Two aspects of the concept of local thermodynamic equilibrium are examined in mediumpressure, mercury arc discharges with thallium iodide additives. It is shown theoretically that the experimental intensity of the T 5350 A line is reasonably consistent with a Boltzmann distribution of excited states. In fact, it is theoretically and experimentally shown that self-absorption is a major factor in limiting the intensity of the bright thallium 5350 A line even though Ti is a minor constituent in the discharge and even though this line does not terminate on the ground state. The dramatic spectral inhomogeneity that develops in long, horizontal, ac discharges with excess TII in the presence of a longitudinal temperature gradient is examined in detail. Both emission and absorption studies show that the spectral inhomogeneity is caused by a large gradient in the thallium concentration. This inhomogeneous distribution of TlI vapors is caused by an inhomogeneous and unsymmetrical distribution of condensed THI along the tube walls which results from the temperature gradient. In turn, the temperature gradient is increased by the ionization of thallium which locally decreases the electrical resistivity, electric field, and power input to the arc. Thus, an equilibrium distribution of TlI vapors is attained only if the distribution of condensed TlI is reasonably uniform and the tube geometry and pressure are favorable for rapid diffusion of the vapors. The inhomogeneity also disappears if a sufficiently small quantity of solid TII is added to the lamp so that it is all vaporized.