Full Text
A New Light Source Using the Phenomenon of Electroluminescence
Fluorescent lamps currently being manufactured are based on the transformation of the resonant ultraviolet radiation of a low-pressure mercury discharge into visible radiation by means of luminophores deposited on the inner surface of the lamp bulb.
Recently, a fundamentally new luminescent light source has been described, in which electrical energy is converted directly into light \(^{1,2}\).
Gudden and Pohl as early as 1921 discovered that strong electric fields are capable of increasing the brightness of the phosphorescence of luminophores that had been previously excited by ultraviolet radiation. This effect has since been studied by a number of investigators \(^{3,4}\).
Destriau\(^5\) was the first to show the possibility of directly exciting luminescence in certain inorganic substances under the action of strong alternating electric fields, and subsequently studied in detail this phenomenon, which he called electroluminescence. A review of the principal works on electroluminescence is given in Destriau’s latest paper\(^1\) and in Garlick’s book\(^4\).
The phenomenon of electroluminescence is observed only in strong alternating electric fields at field strengths of the order of \(10^5\)—\(10^6\ \text{V/cm}\) and at frequencies from 10 to 10,000 cycles.
Sulfides, tungstates, silicates, germanates, and other substances have been investigated as phosphors. At high field strengths all the phosphors investigated exhibit a certain glow; however, the best effect was obtained with specially prepared zinc sulfide samples\(^6\).
A relatively high brightness of the glow of this phosphor is achieved as a result of threefold calcination for one hour at a temperature of \(1200^\circ\text{C}\) and thorough mixing after each calcination. Calcination mainly increases the sensitivity of the surface layers, which may be explained by the oxidizing action of atmospheric oxygen. This phenomenon is confirmed by the fact that an analogous effect can be obtained by calcining in vacuum a mixture of ZnS and ZnO; moreover, in this case, as a result of a single calcination, a homogeneous substance can be obtained that is highly sensitive to the action of the field.
Investigations were carried out of the dependence of the threshold value of the field strength, at which luminescence first becomes noticeable, on the ratio of ZnS and ZnO in the mixture, and also on the activator content (Cu). The optimum ratio was found to be ZnO—75%, ZnS—25%, and Cu—0.2%, for which the threshold value of the field strength at a frequency of 50 cycles is approximately \(3500\ \text{V/cm}\).
The curve of the dependence of the threshold value of the field strength on the activator content (Cu) has a complicated form with two minima, due to the influence of the change in copper content both on the luminescence and on the electrical conductivity of the phosphor.
Destriau’s investigations were carried out mainly at an alternating-current frequency of 50 cycles. Preliminary results of experiments showed a significant increase in the brightness of the luminescence with increasing frequency; moreover, for ZnO the change in brightness is connected with the frequency of the field by a linear dependence. The investigations showed that the intensity of the glow changes periodically with a frequency twice that of the external field. The maximum of the glow does not coincide in phase with the maximum of the field, and the phase shift depends on the phosphor used and on the activator content.
It is easy to show that the maximum strength of the active field inside the phosphor \((E')\) is connected with the maximum of the sinusoidal external field \((E)\) by the relation
\[ E'=\frac{E}{\sqrt{1+\left(\frac{4\pi\cdot 9\cdot 10^{11}}{\omega k\rho}\right)^2}} . \tag{1} \]
where \(\omega\) is the angular frequency;
\(k\) is the dielectric constant;
\(\rho\) is the resistance \((\Omega\cdot\text{cm})\).
Whereas the values of the dielectric constant for various phosphors are known and are equal to 5–8, the exact values of the specific
resistance, which depends on numerous factors, are at present unknown. As relation (1) shows, the maximum of the active field is greatly reduced as the specific resistance of phosphors decreases; moreover, for \(\rho < 10^9\) ohm·cm the internal field is already several times smaller than the external one.
The active field leads the external one by an angle determined by the relation
\[ \tg \varphi = \frac{4\pi\,9\cdot 10^{11}}{\omega k \rho}, \tag{2} \]
Thus, if one assumes that luminescence radiation coincides in phase with the active field, the phase shift between the intensity of luminescence and the applied external-field voltage should increase as the resistance of the phosphor decreases. Such changes in the phase of luminescence were indeed found when the activator (Cu) content in the phosphor was increased. It was found that an increase in the activator content, in addition to increasing the phase shift, leads to noticeable changes in the time course of the radiation curve; moreover, at a concentration of Cu of about 1/100 additional emission maxima are observed.
It is interesting to note that an analogous change in the magnitude of the phase shift is also observed when the field strength is increased, which may be explained by the growth of the conductivity of the phosphor caused by an increase in the number of electrons in the conduction band.
The dependence of the maximum brightness of luminescence \(B\) on the applied voltage \(V\) may be expressed by the empirical relation
\[ B = aV^n e^{-b/V}, \]
where \(a\), \(b\), \(n\) are constant coefficients. The values of the exponent \(n\) for different phosphors vary within the range from 1 to 3.
The coefficient \(b\) depends on the temperature \(T\) according to the relation
\[ b = \frac{\delta}{T - T_1}, \]
where \(\delta\) and \(T_1\) are constants.
Thus, the brightness of luminescence increases with increasing temperature; moreover, extrapolation of the experimental data obtained to the region of low temperatures gives a value of the critical temperature \(T_1\), at which electroluminescence should not be observed, equal to approximately \(225^\circ\) K.
The spectrum of electroluminescence differs markedly from the corresponding spectrum for ultraviolet excitation. Thus, the luminescence band of zinc sulfide shifts toward the red part of the spectrum on transition from ultraviolet excitation to excitation by an alternating electric field.
To obtain high field strengths at low voltages in the circuit, it is necessary to place a thin layer of phosphor between the plates of a flat capacitor. In the electroluminescent capacitor described by Destriau, the phosphor is deposited in a monocrystalline layer on a metal plate and fixed to it with Canada balsam or a very thin rubber film. Over this layer a film of insulating oil is applied, onto which a layer of mica is placed, the lower part of which is made conducting with the aid of a thin film of sea salt with glycerin.
In this way the author was able to obtain capacitors 0.05–0.08 mm thick, which gave a noticeable glow already at a mains voltage of 110 V and a frequency of 50 Hz.
A more perfect design of a luminous capacitor, suitable for use as a light source, is described in a recently published article by Payne et al.^2 The article is predominantly promotional in character and contains very scant data from the scientific point of view. Thus, for example, the article completely lacks
Fig. 1. Diagram of a luminescent capacitor lamp.
information on the phosphor used, the luminescence output, the circuit diagram, etc.
Fig. 2. Time curves of current ($I$), voltage ($V$), power ($W$), and brightness ($B$) of electroluminescence.
The luminous capacitor (Fig. 1) has a layer of phosphor suspended in a suitable solid dielectric and sprayed over the surface of a special “conducting” glass. On one of the surfaces of this glass, by means of special firing, a coating with good electrical conductivity is produced, of thickness on the order of 5 μ, with a light transmittance of about 85%. This layer is distinguished by great hardness and chemical stability.
The surface of the luminescing layer is covered with a thin metallic film, which may be obtained, for example, by sputtering aluminum. The threshold of visible brightness of such a capacitor corresponds to a voltage of 25 V at a frequency of 60 Hz.
With the aid of an oscillograph, curves were taken of the variation with time of the voltage, current, and brightness of the electroluminescence. Characteristic curves for a sinusoidal field at a frequency of 60 Hz are shown in Fig. 2.
As examination of the curves shows, the current leads the voltage in phase by almost \(90^\circ\).
The radiation of electroluminescence coincides in phase with the positive part of the power curve. The dependence of the brightness of luminescence on the mains voltage, the power, and the field frequency is shown in Figs. 3 and 4. The brightness increases considerably with increasing voltage and power; moreover, its limiting value is determined by the breakdown voltage of the phosphor. At constant voltage the brightness rises rapidly when the frequency is increased. Of particular interest is the observed linear relation between the radiation output per unit of supplied power and the field frequency. Thus each charging of the capacitor to a given voltage gives the same brightness of glow, independently of the field frequency.
Fig. 3. Dependence of the brightness \((B)\) of electroluminescence on the voltage \((V)\) and power \((W)\) of the field.
In Fig. 5 a photograph is given of a capacitor luminescent lamp and of objects illuminated by it. It is also possible to manufacture a light source in the form of a glowing cylinder. The brightness of the glow at a frequency of 3500–4000 cps is about 800 apostilbs.
The authors indicate that the new light sources may be used in the form of luminous ceilings, panels, columns, etc. Of particular interest is the application of the new method of exciting luminescence for self-luminous objects: instrument panels of airplanes and automobiles, parts of television sets and radio receivers, watches, etc.
Among the advantages of the new light source should be included: simplicity of manufacture, associated with the absence of vacuum bulbs, the possibility of regulating brightness over wide ranges—
Fig. 4. Dependence of the brightness \((B)\) and output \(\left(\dfrac{B}{W}\right)\) of electroluminescence on the frequency \((\nu)\) and power \((W)\) of the field at constant voltage.
circuits by varying the power, and practically instantaneous switching on and off.
It is interesting to note that, whereas a change in voltage practically does not change the spectrum of electroluminescence, a change in frequency clearly affects the spectral distribution of the radiation energy. Thus, for example, when going from a frequency of 60 cps to a frequency of 300 cps, the color of the glow changes correspondingly from yellow-green to blue-green.
Fig. 5. Luminescent capacitor lamp.
The modern theory of the luminescence of crystal phosphors provides little material for interpreting the process of electroluminescence. Leverenz7 attempted to explain the glow he observed by the breakdown of thin layers of air around particles of the phosphor in organic films, producing the characteristic bluish smoldering glow of atmospheric nitrogen, which can excite luminescence. The untenability of this theory in the light of recent studies of electroluminescence raises no doubts. According to Destriau, the emission of electroluminescence is associated with the transition of electrons, under the action of the field, from the fundamental levels associated with impurities into the conduction band. Since this transition normally requires only a few electron-volts, it remains unclear to what extent this hypothesis is applicable to the fields that produce electroluminescence. Moreover, this theory is contradicted by the absence of noticeable afterglow, as well as by the fact that the glow can be excited in a phosphor that has been in complete darkness for several days, and also by the possibility of continuous glow for several tens of hours.
Thus the effect of electroluminescence, in addition to its undoubted technical significance, is also of considerable interest for the general theory of the luminescence of crystal phosphors.
D. Sh.
References
- G. Destriau, Phil. Mag. 38, 700 (1947).
- E. C. Payne, E. L. Mager, C. W. Serome, Ill. Eng. 45, 688 (1950).
- G. Destriau, Phil. Mag. 38, 774, 880 (1947).
- G. F. Garlick, Luminescent Materials, p. 145 (1949).
- G. Destriau, J. de chim. phys. 33, 587 (1936).
- G. Destriau, S. Soddy, J. de phys. et rad. 6, 12 (1945).
- H. W. Leverentz, An Introduction to Luminescence of Solids, p. 392 (1950).