Abstract: In this paper we describe the design of a strategy based on a number of selection criteria that will lead us to identify the materials that will replace mercury as the active medium in low-pressure plasma lamps. The selection criteria focus on the thermodynamic characteristics and radiation emission of the materials as well as on safety regulations. The strategy is in no way a complete solution but a rational approach to one, based on existing sources and databases. Experimental studies and spectral analyses of the selected materials confirm the theoretical predictions of the strategy, revealing emission spectra suitable for general lighting lamps.
Invention
1. Introduction
Humanity has two major energy issues to deal with. The first is the availability of non-renewable energy sources and whether global energy demand can be met. The problem lies in their depletion in certain parts of the planet and in geopolitical factors that make their availability on the global market a very important matter. The second issue concerns the planet’s environmental changes and the effects of those changes on our lives. The burning of fossil fuels, the most common means of generating energy, results in the production and emission of carbon dioxide as a by-product, one of the gases responsible for the greenhouse effect.
Considering that humans use roughly one fifth of the world’s energy output for lighting, the importance of light source technology, both economically and environmentally, is easy to appreciate. Light sources and lighting not only represent a market worth billions of dollars, but the energy consumed for lighting is also responsible for billions of tonnes of carbon dioxide every year.
The dominant technologies in the light source field are incandescence, electric discharge (plasma) and light-emitting diodes. The advantage of incandescent products is the high colour rendering index of the white light they emit, while that of diode-based solid-state sources is their long average life and durability. The fact remains, however, that light sources in both of these categories fall short of electric discharges when it comes to energy efficiency. Incandescent lamps and diode sources produce less than 20 lumens per watt, whereas products based on electric discharges produce more than 40 lumens per watt, reaching as much as 200 in the case of the low-pressure sodium vapour discharge lamp. Undoubtedly, mercury plays the central role in the light source market, as an element found in the majority of discharge lamps, from low-pressure mercury fluorescent lamps to high-pressure high-intensity discharge (HID) lamps and ultra-high-pressure projection lamps. Even in high-pressure sodium vapour lamps and high-pressure metal halide lamps, mercury is added and acts as one of the main active media.
Light source scientists and researchers today face the challenge not only of raising lamp efficiency but also of eliminating the use of mercury by identifying other materials to serve as efficient radiators. Replacing mercury is a requirement born of a long-standing environmental concern about mercury, particularly its organic form, which is hazardous and poses a threat to human life. Today this concern is expressed within an official framework of government directives (the well-known European RoHS directive, the acronym standing for Restriction of Hazardous Substances). Besides mercury, this directive bans the use in electronic products, lamps included, of lead, cadmium and hexavalent chromium.
For this reason, the researchers and authors of this report examined existing technologies and decided to focus on the development, if not the reinvention, of electric discharges as light sources, since this is the technology in urgent need of new materials to replace the existing active media. The efficiency target for all scientists is, of course, to match or exceed the values offered by mercury in discharge tubes (in low-pressure mercury tubes we find a 65% conversion of electrical energy into radiation, with emitted visible light reaching 120 lm/W for some lamps, although there is usually a trade-off between efficiency and colour rendering index, depending on the lamp’s fluorescent phosphor, which may differ for different applications).
In this paper the authors drew up a set of selection criteria based on their own views and, by scanning a number of databases and going through stages of elimination, arrived at a list of candidate materials that are believed capable of acting as active media in new light source products.
2. Selection criteria
2.1. Candidate materials
The fact is that the use of pure chemical elements is limited by their particularly low vapour pressures, with a few exceptions that have already been the subject of research for decades. The noble gases, sodium and mercury have already formed the basis of a large number of lighting products. For other elements with suitable emission lines, some molecular form has been used, in high-pressure lamps where temperatures are high enough for the compound to vaporise sufficiently, after which the atoms dissociate and some of them act as the active medium and source of radiation.
Electron temperature and gas or vapour temperature of the plasma as a function of pressure.
In our research we will focus on substances whose emissions match a predefined spectral range, and we will include not only atoms but also diatomic molecules or radicals (so as to solve, at least to some extent, the problem of very low vapour pressure found in most elements), provided at least one of the constituent atoms exhibits useful emission lines. The reason for evaluating diatomic compounds is that in the gas phase, during the electric discharge, the atoms dissociate to a large extent and the atomic emissions contribute to the final emission spectrum. Finally, parent compounds will be chosen if a diatomic molecule of interest is not available or not physically stable.
2.2. The spectral range
Under the operating conditions of a low-pressure mercury vapour discharge lamp (alternating current at 50 Hz where electromagnetic ballasts are used, up to a few tens of kHz with electronic control gear), the most intense mercury emission line is the one at a wavelength of 254 nm. This emission line is also known as the resonance line, since the quantum (electronic) transition that produces it is from the first energy level to the ground state of the atom. To produce useful light for illumination, a phosphor powder is applied to the inner walls of the tube. It absorbs the photons of ultraviolet energy and, once de-excited, releases photons with energy in the visible part of the spectrum. Energy is of course lost in this process, since an ultraviolet photon is exchanged for a visible one. This energy loss is known as the Stokes loss. It is not limited to mercury but occurs with any active medium that emits at energies outside the visible spectrum and therefore requires wavelength conversion using a phosphor.
The efficiency of a light source is defined as the radiated power divided by the total electrical power supplied to the source, whereas luminous efficacy is defined as the radiant power emitted in the visible part of the spectrum divided by the total electrical power. A further factor that must of course be taken into account when calculating luminous efficacy is the sensitivity of the human eye to the various visible wavelengths, which peaks at around 555 nm, as shown in Diagram 1.
The sodium resonance line lies close to the peak of the human eye’s photopic sensitivity curve (589 nm).
One of the reasons the low-pressure sodium vapour discharge lamp has such high luminous efficacy is that its resonance line lies close to the peak of the human eye’s sensitivity curve (589 nm). The authors therefore set as the first selection criterion that the new candidate materials should be efficient emitters of radiation at wavelengths in the 380 nm–580 nm range. The reason for this criterion is to eliminate any Stokes losses associated with photons the human eye cannot detect, which have to be converted into visible ones by phosphors. At short wavelengths the eye’s sensitivity falls away, and the visible spectrum therefore starts at about 380 nm. At the long-wavelength end, the authors consider there is no reason to include wavelengths above 580 nm: sodium is already a known emitter at about that wavelength, the eye’s sensitivity begins to decline, and in any case radiation at that wavelength or longer cannot be converted into radiation of another wavelength by phosphors. On the other hand, radiation in the blue part of the spectrum can be converted to cover the whole of the visible spectrum.
2.3. Atomic emissions
The entire periodic table was scanned in search of elements with strong emission lines in the predefined region of the spectrum. A number of elements are excluded, not because their emissions are unsuitable but because of safety regulations and bans on their use in products, such as the whole actinide series apart from thorium, which exhibit radioactive characteristics.
Other elements excluded under the European Union directive on hazardous substances (Restriction of Hazardous Substances, RoHS) are mercury, lead, cadmium and chromium.
2.4. Molecular emissions
Once a list of elements from the periodic table had been drawn up that meet the selection criterion requiring the strongest emission lines to lie in the 380–580 nm range, the next step was to find diatomic molecules containing at least one of the selected elements and meeting the same selection criterion.
2.5. Vapour pressures
As already mentioned, in high-pressure discharge lamps, where the gas temperature is several thousand degrees Celsius, almost any compound can be vaporised sufficiently to act as the active medium. In this paper we concentrate on low-pressure cases, where the temperature does not exceed a few hundred degrees Celsius and the coldest point of the lamp is at an even lower temperature (40 degrees Celsius for mercury fluorescent lamps). We therefore set a further selection criterion, which requires that the candidate material, or a parent compound of it, have a vapour pressure of at least 0.1 pascal (0.75 milliTorr) at 300 degrees Celsius or lower. These values are indicative and were chosen with the operating conditions of the two best-known low-pressure discharge lamps, mercury and sodium, in mind.
Where the diatomic molecules that pass all the selection stages do not exist in a stable form, a parent compound will be chosen. The selection rule we set in this case is that the stable molecules of the parent compound must contain the atom with the suitable emissions in the lowest possible valence, so as to avoid what we call stoichiometric condensation. This can be explained with an example. Consider a metal M that vaporises sufficiently when in the form MX4, where X is a halogen or oxygen atom (or a combination of the two, as in some oxyhalides), but condenses and does not vaporise sufficiently in the form MX2, which is a stable compound. In the plasma phase of a lamp that starts with MX4 molecules, all the metal will then condense and be sequestered once MX2 and X2 are formed in the gas phase (as molecular oxygen or halogen gas). On the other hand, if the products of the reactions in the plasma do not form stable compounds with low vapour pressures, then under these conditions the problem of condensation can be avoided.
Taking into account the final list of diatomic molecules that radiate in the spectral region of interest (both the diatomic molecule and at least one of its constituent atoms), the parent compounds of the diatomics that meet the vapour pressure criterion, and the final stoichiometric condensation criterion, we arrive at a very short list of candidate materials. Some of the compounds on the list have already been mentioned in publications and patents.
3. Experimental results
The spectrograph used to record some initial spectra was an Ocean Optics USB2000 with a 1.3 mm bandwidth, a grating with 600 grooves/mm and a 10 mm slit. The spectra were recorded after an exposure of about 100 milliseconds. The detector inside the spectrograph was a CCD camera with 2048 elements in a row. The temperature on the outer surface was measured at two points (directly above the electrode and at the centre of the tube) using a K-type thermocouple instrument. Finally, luminance (L) was measured with a Topcon BM-7 placed 50 cm from the lamp, with a field-of-view angle of 0.2 degrees (measurable area: ~Æ1.7 mm). The luminance meter consists of a focusing lens, a field stop that defines the field-of-view angle, and three silicon photodiodes with filters whose transmittance matches the sensitivity of each of the three groups of the eye’s sensors. Luminance and chromaticity coordinates are shown on the instrument’s display and are calculated as (specific silicon photodiode response) / [(field-of-view angle) * (measurable area)] * (calibration factor).
4. Conclusions
The spectroscopic analyses reveal that the design of the selection criteria did indeed lead to some of the goals that were set. These included the emission of atomic radiation in the visible part of the spectrum from the selected compounds, as well as molecular emissions covering as much of the visible spectrum as possible.
Acknowledgements
The researchers would like to thank the Greek Ministry of Development and the General Secretariat for Research and Technology for funding the research through the ENTER programme, as well as Professor Georges Zissis of the Plasma and Energy Conversion Laboratory (LAPLACE) at the University of Toulouse for co-funding the project.
The project is co-funded by:
* 75% of public expenditure from the European Union – European Social Fund
* 25% of public expenditure from the Greek State – Ministry of Development – General Secretariat for Research and Technology, and from the private sector, under Measure 8.3 of the Operational Programme Competitiveness – 3rd Community Support Framework.
Dr Spyros Kitsinelis
PhD in physical chemistry and spectroscopy,
Light source research scientist and technology transfer consultant
