ANALYSIS & OPINION
New strategies in solid-state lighting
9 December 2013Tweet
Visitors to the Strategies in Light Europe trade fair. Credit: EPIC
The cost of producing LEDs remains one of the main barriers to implementing solid-state lighting in homes and industry. Dr Calogero Sciascia, senior scientist at SAES Group, reports from the European Photonics Industry Consortium’s technology workshop at the Strategies in Light Europe conference, which took place from 19 to 21 November in Munich, Germany
Reducing the cost of LED production is one of the more sensitive themes in the LED market. According to Alexander Loesing, co-founder and chief marketing officer at Azzurro Semiconductors, the LED market is characterised by strong price competition. The dollar per lumen cost of LEDs is still too high, and this situation partially hinders mass consumption and preserves a worldwide overcapacity.
Loesing was speaking at the European Photonics Industry Consortium’s (EPIC) technology workshop at the Strategies in Light Europe conference, which took place from 19 to 21 November in Munich, Germany.
The configuration of the market tends also to promote large companies, which become even bigger enlarging and integrating their production capabilities throughout the value chain. Loesing gave a number of examples: Samsung, Epistar, Toshiba, Cree, etc. Similar points were made by Professor Wang N Wang, chief science officer at semiconductor manufacturer IQE.
The higher control of crystal growth parameters allows manufacturers to scale up their production and cost savings through large yields. Both Loesing and Wang indicated that the silicon wafer is a valid alternative to sapphire on gallium nitride (GaN) growth with regards to cost savings. The economical advantage is not just related to the cost of the material, but there is the advantage of a different and more consolidated (and amortised) production platform. On the other hand, the technology surrounding GaN on silicon is less mature and, to some extent, more complex. Thus careful consideration of the pluses and minuses of the technology has to be taken into account.
One of the trends in the semiconductor industry is for larger growth reactors, noted Professor Michael Heuken, vice president of corporate research and development at semiconductor company Aixtron SE. He said larger growth reactors are required in order to accommodate larger wafers or, alternatively, more substrates, with the first option preferred for GaN on silicon and the second for GaN on sapphire.
Professor Heuken opened the EPIC session with a historical introduction and, surprisingly, the story of gallium nitride devices dates back to 1951 with a patent for substituting silicon with III-V semiconductors. Since then, a lot of progress has been made and Professor Heuken described the basic concept behind a modern LED device and the metal-organic vapour phase epitaxy (MOPVE), which is the preferred growth technique for achieving such fine control on the structure.
Sticking with semiconductor growth, Dr Kolja Haberland, chief technological officer of LayTech, which supplies metrology equipment for the semiconductor industry, went into details of in-situ wafer characterisation. Physical homogeneity results in a high quality product, and several parameters (substrate temperature, layer thickness, roughness, number of defects and wafer bowing) affect the final wafer quality and consequently the emission properties of the LED. Thus a continuous online check of such parameters provides the information for closed-loop control and consequent feedback for the reaction.
The apparently simple task of wiring the LED to the mains is actually a complex activity driven by the same constrains of energy efficiency, lifetime, integration, cheapness and modularity acting on the LED itself, according to Claudio Adragna, director of application and system architecture at STMicroelectronics. The safety of LEDs is also partially concerned with the electronic circuitry, and in ultimate analysis many of the ‘smart’ properties attributed to LEDs are actually implemented at the driver level.
SAES Group, based in Milan, Italy, specialises in metallurgy and material science. The company develops shape memory alloys (nitinol materials) and getters (reactive materials) for MEMS, microelectronic hermetically packaged devices, Organic Light Emitting Diodes (OLED) and for the photovoltaic industry. At the EPIC workshop, Dr Sciascia presented the role of active barrier and the getter materials for preserving the unique optical properties of both organic and inorganic light emitting diodes. He also presented the shape memory alloys as possible candidates for miniaturised actuators for LEDs.