+44 (0)24 7671 8970
More publications     •     Advertise with us     •     Contact us
 
Technical Insight

Liquid delivery system for high volume NH3 usage in the proliferation of low cost LEDs

The use of LEDs is already widespread in consumer electronics, appliances, and other products. Single LEDs are seemingly ubiquitous; LED assemblies are widely used in mobile electronics, computer displays, and televisions. We can expect that LEDs will spread through industrial lighting and into the general lighting of our homes. By Ryan Clement and Robin Gardiner, at MATHESON.

The growth has occurred quickly. The shift in manufacturing scale, from niche production to high volume /low unit cost processing, is well underway. Along with it, fabricators are engaging in a complete review of all aspects of the supply chain. At the focal point, naturally, is production process technology. Purchase a backlit TV, and the LED unit assembly can be 28% of the costs . An estimated 25 million LED backlit TV’s will be built this year with as many as 500 LEDs per panel. It is expected that this application alone will have a growth rate of 10-15 % through to 2013/2014. Replace an old bulb today with LED technology and as much as 40% of the price you pay is cost of materials . With a potential of 100 billion additional LEDs by 2020, manufacturers are eager for improvements that impact their total cost by reducing production downtime, product variability, and waste. Ammonia delivery and its pivotal role in the LED fabrication process Ammonia (NH3) is one of the primary gases used in high volume for MOCVD growth of gallium nitride (GaN) films used to make LEDs. Consequently, NH3 is being used at increased rates and in increased quantities that parallel the rapid expansion of production. At the outset, and until recently, cylinder-based delivery systems were used to deliver NH3 to the GaN MOCVD reactors. Today, as the manufacturers have been adding reactors to keep up with LED demand, it has become clear that bulk NH3 supplies are required. Weaknesses of conventional bulk delivery Conventional Bulk Specialty Gas System (BSGS) exhibit limitations in three critical areas: 1. Output flow rate is limited, which in turn limits the number of reactors that a BSGS can serve; 2. The bulk liquid supply becomes increasingly contaminated by water (and other impurities) as it is being depleted. This results in variable demands on external purification, and, more importantly, in a significant residual volume of contaminated liquid ammonia that is below spec and must be discarded. 3. Safety and performance considerations involved with heating a large volume of NH3 grow more complicated as the size of the BSGS vessel is increased. These disadvantages are basic and severely limiting. The need for higher productivity in LED production is a realworld requirement. Overcoming any the limits of flow rate scalability, bulk material contamination, waste, performance, and safety would be a positive step. Overcoming them all would be a breakthrough. A newer approach to bulk ammonia supply, Liquid Extraction and Total Vaporization (LETV) is the breakthrough technology that overcomes all of the weaknesses of conventional BSGS systems. First, we will describe the detail of the BSGS limitations; after which we will provide a description of the LETV technology and its advantages. Flow limitations in conventional bulk delivery systems A conventional BSGS is comprised of a bulk container and a gas piping system. Liquid NH3 vapourises within the bulk container and is piped to the point of use as a gas. Limitations associated with heat control of the bulk container (explained below) constrain the size of the bulk container, and impose limitations on liquid ammonia surface area – in turn limiting the flow capacity, and constraining the scalability of the system. So, even with a bulk delivery system in place, fabricators who continue to expand by the addition of multiple GaN reactors place consequent additional burdens on their NH3 delivery systems, and face the need to deploy additional BSGS systems. Impurity issues with BSGS The mechanism employed in a conventional BSGS is, in effect, a single plate distillation. As with all distillations, the composition of what remains inside the container changes over time. As the ammonia is consumed, the concentration of impurities increases. Under process conditions, the overall composition of the bulk supply can change to the extent that it no longer meets the specifications required, and its use must be terminated. Depending upon process requirements and end-user SOP, the unusable supply can be as much as 20% of the original volume; this is an obvious and immediate target for cost reduction. Figure 1. Representation of the increasing water contamination of NH3 gas output in standard bulk supply systems and LETV systems. Impurities are easier to remedy when they are stable; LETV impurity level is essentially flat throughout the depletion of the bulk supply Wasted supply is not the only problem. Water is a common impurity for ammonia, and oxygen is the cause of critical defects related to LED brightness. As described above, as the NH3 is consumed, the concentration of impurities increases. Well before the point at which the ammonia supply is rendered completely unusable, the drift in moisture contamination from the BSGS is a process variable that impacts product quality (see Figure 1). Removal of moisture and impurities from the exiting gas is possible, but because the moisture and impurity content changes dynamically as the bulk supply is consumed, purification under these changing conditions represents a challenge to most purification systems. The resulting inconsistency in the quality of the gaseous ammonia output can lead to process and product variability. Heating the bulk container An additional complication is that a conventional BSGS requires that heat be applied to either to the entire bulk NH3 supply or to the liquid surface through microwave technology in order to induce vaporization and to maintain the required flow rates. The application of heat to ammonia presents control problems as well as safety risks. Heating any size vessel poses a control problem, because the amount of heat must be continuously monitored and adjusted as the container is emptied in order to maintain the delivered gas at a constant flow. Larger vessels respond to temperature changes more slowly, and are difficult to control with precision. Larger vessels also pose a larger safety risk when electric heaters and microwave energy are present. Smaller vessels reduce the control problem, but suffer the obvious disadvantage of low capacity. Large or small, vessels of any size present a temperature control problem. In the quest for capacity, there are practical limits to the size of a bulk liquid NH3 container that can be efficiently and safely heated. Phase change dynamics and moisture NH3 has a high latent heat of vaporization and keeping the phase change continuous and controlled in the face of varying demand from the process is non trivial. From a conventional BSGS the concentration of the water impurities increases with exiting gas flow rate; it reaches a maximum; then declines. This behavior is a result of the combined effects of thermodynamics, fluid dynamics and heat transfer. As smooth evaporation occurs, an increase in moisture develops at the NH3 liquid-vapor interface. Surface enrichment increases with an increased flow rate and therefore the moisture concentration increases (see Figure 2). Figure 2. Representation of contamination profiles as a function of gas output flow rate. Flow rate changes are to be expected in a real-world production setting – variable contamination as a function of flow rate is a real-world problem. At a certain flow rate, boiling becomes vigorous enough to disrupt and mix the liquid vapor interface which starts to lower the surface enrichment of moisture. At high flow rate, a film boiling regime develops and heavy mixing is found within the container. As a result, the surface enrichment is not capable of developing; and a lower moisture level is observed. In the real-world setting of LED production, reactors and processes start and stop requiring on-the-fly changes in demand for gaseous NH3. Under this circumstance, levels of moisture will always vary in the output flow of conventional NH3 delivery systems - a result of fundamental physical properties of NH3 with H2O contamination no matter what the purity. In the final analysis, the instability and variability of the vaporization process in the typical BSGS presents barriers for the development of a repeatable, high volume production process. Liquid extraction and total vaporization (LETV) LETV technology circumvents the problems observed with gas phase delivery from a bulk supply. In a critical departure from conventional BSGS technology, LETV technology isolates the storage vessel from the vaporization process. The design concept is to push liquid from a bulk container to an external vaporizer. In doing so, the system is able to achieve total vaporization of the NH3 while maintaining constant temperature, pressure and therefore flow. The vaporizer is isolated from the bulk supply, therefore vaporization conditions are constant, and are easier to control – and the process is, by definition, a safer one.

 

  • The flow rate limitations of existing designs due to heat flux limitations are not present in the LETV. Everything that reaches the vaporizer is vaporized and there is no “concentration effect” on impurities in the bulk container.
  • The flow rate from a single LETV system vaporizer is easily increased to support 50 GaN MOCVD reactors with flow rates up to 5,000 slpm achievable by adjusting the heat flux.
  • The LETV technology overcomes the problem of moisture impurity spikes and drift through the life of the entire bulk container and across a wide range of flows.
  •  

    Because vaporization is isolated as a separate and controlled step, and because all of the liquid NH3 is vaporized and exits the system, the stability of moisture levels in NH3 delivered by the LETV is better than 2% across a widely varied flow rate range, and throughout the depletion life of the bulk supply. Figure 3 – Output from liquid delivery system and NH3 vaporizer at 250 slpm, showing no change in moisture concentration as the bulk supply is consumed LETV: No increase in water contamination during bulk supply depletion The stability of moisture contamination in NH3 is represented in Figure 3. The moisture content was constant as the bulk vessel was depleted. There was no “concentration effect” of impurities as the supply was consumed. The standard deviation in the data was less than 1%. Tests of the LETV technology have shown that the standard deviation was less than 2% even when the flow rate was changed, and regardless of the liquid volume in the bulk container. In fact, LETV is so effective throughout the depletion of the bulk container that the amount of wasted bulk NH3 is limited only by the extent to which the bulk vessel and dip tube enable the emptying of the bulk supply. LETV: Purity of gas supply from UHP liquid bulk supply In conventional BSGS designs the concentration of water contamination will increase as the bulk supply is consumed (to a point where the bulk supply becomes unusable). In an LETV system the purity of the gas output is wholly dependent upon the purity of the bulk supply – and the impurity profiles of both the bulk liquid and the gas output remain essentially flat during use. Maintaining UHP performance from source to delivery is a high demand on any liquid/gas delivery system. Figure 4. Gaseous NH3 purity extracting then vaporizing a UHP liquid source – changes in flow rate over time do not impact purity of output For NH3 with 100 ppb moisture levels in liquid phase the effect of the drying down of the tubing is important in the effort to monitor the LETV performance at such low levels of moisture. The results are given in Figure 4 and for flow rates of 50, 250, and 750 slpm show that the moisture contamination levels can be expected to remain close to constant across a wide range of flow rates, even with a UHP source. Figure 4 clearly demonstrates that LETV technology is capable of delivering gas phase ammonia with consistent moisture content that is equivalent to the moisture impurity level in the liquid phase supply. Notably, this was demonstrated even after abrupt changes in NH3 flow rate, mirroring “real world” conditions used by high volume LED manufacturers. Summary Conventional BSGS systems vaporize NH3 in the bulk supply container, and subsequently transfer the gaseous NH3 to the point of use. Systems of this design have limitations in terms of impurity levels in the gas phase delivery, constancy of impurity levels, capacity, maximum flow rate, and safety. In such systems, water contamination can be such a problem that as much as 20% of the bulk supply is wasted. Liquid Extraction and Total Vaporization (LETV) technology isolates the storage vessel from the vaporization step. This enables better control of the vaporization process – even at high flow rates. LETV technology also removes the limitation on the size of the bulk container. At the same time, LETV technology eliminates the “concentration effect” on impurities, which virtually eliminates waste. The amount of NH3 that can be removed from the bulk container approaches 100% and is limited only by the bulk container design and the access to the liquid phase contents. LETV technology-based systems reduce capital expenditures by being scalable; can increase yields by eliminating variations in source gas; and reduce variable costs by increasing material utilization. Figure 5. Graphic representation of LETV system, illustrating isolation of the vaporizer from the bulk supply Reference 1http://www.displaysearch.com/cps/rde/xchg/displaysearch/hs.xsl/100216_led_backl ght_costs_falling_faster_than_conventional_lcd_back lights.asp 2http://green.venturebeat.com/2010/03/12/despite-advances-led-market-probably volatile-for-two-more-years/ 3 http://www.ledsmagazine.com/news/7/2/18 4 Solid State Lighting II, Ian T. Ferguson, Nadarajah Narendran, Steven P. DenBaars, Yoon-Soo Park, Proceedings of SPIE Vol. 4776 (2002)

    ×
    Search the news archive

    To close this popup you can press escape or click the close icon.
    ×
    Logo
    ×
    Register - Step 1

    You may choose to subscribe to the Compound Semiconductor Magazine, the Compound Semiconductor Newsletter, or both. You may also request additional information if required, before submitting your application.


    Please subscribe me to:

     

    You chose the industry type of "Other"

    Please enter the industry that you work in:
    Please enter the industry that you work in: