Category Archives: LED Manufacturing

By Christian G. Dieseldorff, Industry Research & Statistics Group at SEMI 

Data from SEMI’s recently updated World Fab Forecast report reveal that 62 new Front End facilities will begin operation between 2017 and 2020.  This includes facilities and lines ranging from R&D to high volume fabs, which begin operation before high volume ramp commences.  Most of these newly operating facilities will be volume fabs; only 7 are R&Ds or Pilot facilities.

Between 2017 and 2020, China will see 26 facilities and lines beginning operation, about 42 percent of the worldwide total currently tracked by SEMI.  The majority of the facilities starting operation in 2018 are Chinese-owned companies. The peak for China in 2018 comes mainly from foundry facilities (54 percent). The Americas region follows with 10 facilities, and Taiwan with 9 facilities. See Figure 1.

Figure 1 depicts the regions in which new facilities will begin operation.

Figure 1 depicts the regions in which new facilities will begin operation.

By product type, the forecast for new facilities and lines include: 20 (32 percent) are forecast to be foundries, followed by 13 Memory (21 percent), seven LED (11 percent), six Power (10 percent) and five MEMS (8 percent). See Figure 2

Figure 2: New facilities & lines starting operation by product type from 2017 to 2020

Figure 2: New facilities & lines starting operation by product type from 2017 to 2020

Because the forecast extends several years, it includes facilities and lines of all probabilities, including rumored projects and projects which have been announced, but have a low probability of actually happening.  See Table 1.

FabForecast-table1

 

Probabilities of less than 50 percent are considered unconfirmed, while a probability of 80 to 85 percent means that the facility is currently in construction mode.  Projects with 90 percent probability are currently equipping. As the forecast gets farther out, more of the projects have lower probabilities.

The projects under construction, or soon to be under construction, will be key drivers in equipment spending for this industry over the next several years — with China expected to be the key spending market.

SEMI’s World Fab Forecast provides detailed information about each of these fab projects, such as milestone dates, spending, technology node, products, and capacity information. Since the last publication in August 2016, the research team has made 249 changes on 222 facilities/lines.

The World Fab Forecast Report, in Excel format, tracks spending and capacities for over 1,100 facilities including future facilities across industry segments from Analog, Power, Logic, MPU, Memory, and Foundry to MEMS and LEDs facilities.  Using a bottoms-up approach methodology, the SEMI Fab Forecast provides high-level summaries and graphs, and in-depth analyses of capital expenditures, capacities, technology and products by fab.

The SEMI Worldwide Semiconductor Equipment Market Subscription (WWSEMS) data tracks only new equipment for fabs and test and assembly and packaging houses.  The SEMI World Fab Forecast and its related Fab Database reports track any equipment needed to ramp fabs, upgrade technology nodes, and expand or change wafer size, including new equipment, used equipment, or in-house equipment. Also check out the Opto/LED Fab Forecast.

Learn more about the SEMI fab databases at: www.semi.org/en/MarketInfo/FabDatabase and www.youtube.com/user/SEMImktstats.

In 2015, all economic indicators pointed to continued market growth for both industries, power electronics and LED, especially with IGBT modules boosted by EV/HEV industry and general lighting applications, a killer application for LEDs since 2012. To support this growth and answer the thermal management needs in power electronics and LED, lot of innovative technologies are emerging. According to Yole Développement (Yole), one of the most impressive technical developments is the convergence of thermal management for both sectors, LED and power electronics, particularly the materials used for thermal management. The thermal management convergence is driven by the applications, announces the “More than Moore” market research and strategy consulting company, Yole.

thermal management

Thermal Management Technology & Market perspectives in Power Electronics and LEDs report 
powered by Yole’s Power Electronics & LED teams, reviews insight into synergies between power electronics and LED for thermal management. It describes and analyzes drivers and challenges that are facing industrial companies. This latest report proposes an overview of the market trends and technology evolution including 2015-2021 market figures, technology status and technical roadmap analysis and more. Under this report, Yole’s analysts also offer business model and supply chain analysis across various materials used for thermal management.

A rapid convergence of key technologies is driving unprecedented change. In this dynamic environment, Yole’s goal is to understand their customers’ strengths and guide their success.

“Power electronics and LEDs are different industries that today face similar challenges”, explains Dr Pierric Gueguen, Business Unit Manager at Yole. And he adds:”Needs for green energy with lower CO2 emissions have led these industries to develop more efficient and smaller solutions.” At the device level, cost pressure and the need for better performance is pushing designers towards smaller and thinner chips, also leading to increased power density. Such power density targets in both power electronics and LEDs bring a convergence of thermal management requirements, supporting the development of new materials.

Among materials used for thermal management, Yole specifically investigated the market and technology evolution of die attach, substrates, baseplates/PCBs and encapsulants. Overall, the market for these materials was worth US$1.98 billion in 2015 and will grow to US$3.16 billion by 2021 at a CAGR of 6%.

“Their value proposition has the potential to bring business to their suppliers and key differentiating factors to device manufacturers,” commented Pierrick Boulay, Technology & Market Analyst at Yole.

“Power electronic modules represent a healthy market, worth about US$2.9 billion in 2015 and set to reach US$4.5 billion in 2021, growing at 9% CAGR,” explained Pierric Gueguen. In parallel, the LED packaging market reached US$15 billion in 2015, after years of strong growth led by LED TV and general lighting. However, price pressure will moderate growth in coming years, with a 3.4% CAGR leading to a market worth US$18.5 billion in 2021.

Power electronics and LEDs need the right materials to handle thermal management challenges. As those applications are driven by similar technical requirements, one technical solution can be adopted and developed for one industry before being used by another industry. “The 30% of the overall thermal management material market that is common to both LED and power electronics represents US$660 million in 2015”, announces Pierrick Boulay. “According to our estimations, such market segment will reach US$1014 million in 2021”. Moreover, another 30% can be reached by adapting existing technologies used in LED or power for the other application…

From perspectives ranging from manufacturers and material suppliers through to end users, market dynamics, drivers and challenges are presented in this report, for both power electronics and LEDs.
A detailed description of the thermal management report as well as other LED & Power Electronics reports Yole are available on i-micronews.com, reports section.

Cree, Inc. (Nasdaq: CREE) introduces the XLamp XHP50.2 LED, which delivers up to seven percent more lumens and 10 percent higher lumens-per-watt (LPW) than the first generation XHP50 LED in the same 5.0 mm x 5.0 mm package. The new XHP50.2 LED enables lighting manufacturers to quickly improve the performance of existing XHP50 lighting designs. Capable of producing more than 2,500 lumens from its 6mm light emitting surface (LES), the XHP50.2 can reduce the size and cost of new designs and enable innovative solutions to address applications ranging from spot to street lighting.

“Arianna shares Cree’s vision that LEDs should not compromise quality or performance and should provide better lighting experiences in all aspects,” said Lorenzo Trevisanello, R&D manager of Arianna. “Our goals are to achieve the best cost-efficacy and versatility using the most efficient LEDs. Thanks to the XHP50.2 LED’s lumen density and proven reliability, even at high operating temperatures and drive currents, we are able to push the performance and size boundaries of our products even further.”

In addition to light output and efficacy enhancements, the XHP50.2 LED provides improvements to optical uniformity through secondary optics, enabling spot and portable lighting manufacturers to deliver better lighting experiences. The XHP50.2 LED has LM-80 data available immediately, reducing the time required to receive ENERGY STAR® and DesignLights Consortium® qualifications.

“Cree redefined High Power LED performance with the introduction of the industry’s first Extreme High Power LEDs,” said Dave Emerson, senior vice president and general manager for Cree LEDs. “Delivering the industry’s best lumen density and reliability, Cree’s XHP LED family allows our customers to achieve performance levels not possible with other LEDs at the lowest total system cost in a wide range of applications. With the launch of XHP50.2, Cree continues to redefine what is possible with high performance LEDs.”

Featuring Cree’s EasyWhite technology, which provides the industry’s best color consistency, the XLamp XHP50.2 LEDs are available in 2700K-6500K with high CRI options. Product samples are available now, and production quantities are available with standard lead times.

Ultratech, Inc. (Nasdaq: UTEK), a supplier of lithography, laser-processing and inspection systems used to manufacture semiconductor devices and high-brightness LEDs (HB-LEDs), as well as atomic layer deposition (ALD) systems, today announced that the Laboratory for Emerging and Exploratory Devices (LEED), led by Professor Sayeef Salahuddin, Ph.D. of the Electrical Engineering and Computer Sciences Department at UC Berkeley (EECS UC Berkeley), has chosen the Ultratech-CNT Fiji G2 PEALD system as its instrument of choice for its research activities. Professor Salahuddin was recently honored at the White House by President Barack Obama for his work in developing nano-scale electronic and spintronic devices for low power logic and memory applications.

“ALD provides an exciting way of accessing ferroelectric materials, which play a key role in these types of devices, by providing a means of controlling the film properties through the precise engineering of the composition,” noted Professor Salahuddin. “This has led the way for us to explore the ferroelectric properties of rare earth oxides, such as Hafnium oxide, by adding a variety of dopants, such as silicon (Si), aluminum (Al), and yttrium (Y). Our decision in choosing the Fiji system was motivated not only by the system’s performance, and flexibility but also because of the strong reputation that the Ultratech ALD team has for R&D expertise, coupled with its excellent support.”

Adam Bertuch, senior thin film scientist at Ultratech-CNT, who has played a key role in the development of PEALD oxides at the company, said, “The Fiji is an extremely versatile instrument, which has been at the leading edge of the development of complex materials. Professor Salahuddin’s work in the field of ferroelectric materials speaks for itself, and we are looking forward to having a strong collaborative relationship with him, as well as his scientific group at UC Berkeley.”

Ultratech Fiji G2 ALD System

For advanced thin films, the Fiji series is a modular, high-vacuum ALD system that accommodates a wide range of deposition modes using a flexible architecture and multiple configurations of precursors and plasma gases. The result is a next-generation ALD system capable of performing thermal and plasma-enhanced deposition. Ultratech CNT has applied advanced computational fluid dynamics analyses to optimize the Fiji reactor, heaters, and vapor trap geometries. The system’s intuitive interface makes it easy to monitor and change recipes and processes as required. The Fiji is available in several different configurations, with up to six heated precursor ports that can accommodate solid, liquid or gas precursors, and up to six plasma gas lines. Options include a built-in ozone generator, Load Lock as well as several in-situ analysis tools, which offer significant experimental flexibility in a compact and affordable footprint.

ClassOne Technology, manufacturer of cost-efficient wet processing equipment for ≤200mm substrates, has reported its best-ever sales quarter and is currently doubling its Kalispell manufacturing capacity to meet the demand.

“We’ve been seeing a steady increase in market interest and sales,” said ClassOne Technology President, Kevin Witt. “Most of these users are now focusing on capabilities they couldn’t get before, like wafer-level packaging and More than Moore technologies.”

Witt explained that wafer-level packaging (WLP) has been used for some time with 300mm and larger substrates — but the equipment has not been available for 200mm. “Fortunately, ClassOne focuses specifically on the smaller-wafer markets,” said Witt. “At a very affordable price, we deliver the new technology and advanced 3D features they’re looking for. For example, our Solstice® line of multifunctional electroplating systems enables high-efficiency Cu Through Silicon Via (TSV), Pillar, Bump and Barrier Plating and other capabilities that WLP requires. And that’s one major reason they’re coming to us.”

ClassOne reports that many of the new buyers are keenly interested in More than Moore (MtM) technologies to increase functionality while reducing cost per device. They are producing compound semiconductors, LEDs, MEMS, RF, Wi-Fi and a range of IoT-related sensors and other devices. ClassOne cites the combination of ≤200mm-specific tools, advanced capabilities and affordable pricing as the primary driver behind the current equipment-buying surge in emerging markets.

ClassOne Technology offers a selection of new-technology wet processing tools designed for 75mm to 200mm wafer users. These include three different models of Solstice electroplating systems for production and development as well as the Trident families of Spin-Rinse-Dryers and Spray Solvent Tools. All are priced at less than half of what similarly configured systems from the larger manufacturers would cost — which is why the ClassOne lines are often described as delivering “Advanced Wet Processing for the Rest of Us.”

Within a highly competitive landscape due to a strong price pressure, most of the LED companies are looking for business opportunities and adopt different strategies of development. Vertical integration, product, application and activity diversification. New relays of growth are required for LED players to survive.

From a packaging point of view, more and more packaged LED manufacturers selected the vertical integration strategy to move towards the module level and add more and more value in their LED components.

Under the new report titled LED Packaging 2016: Market, Technology and Industry Landscape reportYole Développement (Yole) reviews the LED industry and market status. The “More than Moore” market research and strategy consulting company Yole, details process flows and related technologies in LED packaging. Yole proposes also a comprehensive analysis of the cost reduction and its impact at the LED packaging level.

According to Yole’s analysts, the packaged LED market represented a revenue of nearly US$15.7 billion in 2015. This industry should grow to a size of nearly US$18.2 billion by 2020.

led packaging revenue

Following the overcapacity caused by the recent LED TV crisis and the entry of Chinese players, industry consolidation was expected to decrease competition and stabilize price erosion. This eventually happened in China during 2014 and 2015, but with unforeseen effects on the overall industry. Indeed, several smaller players went bankrupt and many midsize players have since been acquired, leading to a situation where dozens of companies are having “going-out-of business” sales. This has triggered strong price decline and, naturally, other LED players had no choice but to match the price trend initiated by the Chinese industry.

ASP for low and mid power LEDs declined 30% – 40% in the second half of 2015. In parallel ASP for high power LEDs, though less affected, still declined 20% – 30%. Globally, 2015 was a rough year for the LED industry, with packaged LED revenue declining for the first time ever: from US$15.1 billion in 2014 to US$15 billion in 2015.

This decrease was emphasized by lower-than anticipated demand in the LED backlight and LED lighting markets. Moreover, strong evolution in currency exchange rates due to the US dollar’s rise contributed to many players’ declining revenue.

2016 has seen the industry begin recovering, and packaged LED ASPs have mostly stabilized for highly-commoditized stock keeping units like the low-power 2835 and mid-power 5630.

Higher power grades for lighting applications are seeing increasing demand, but also stiffer competition, which likely will lead to a significant ASP drop as competition intensifies.

“Thus we expect the packaged LED market to show moderate growth in the coming years, reaching US$18.5 billion in 2021 (CAGR 2016 – 2021: +3.4%)”, explains Pars Mukish, Business Unit Manager at Yole. 

LED packaging market is still a strong opportunity for materials suppliers. Indeed, LED packaging requires specific materials in agreement with application requirements.

Regarding packaging substrates, the high power density of devices induces the use of ceramic substrates, a market that will grow from nearly US$684 million in 2015 to US$813 million in 2021, according to Yole’s LED packaging report.

Encapsulant/optic materials will follow the same trend: Yole’s analysts announce US$400 million in 2015 and US$526 million in 2021. This market segment is driven mostly by the increased use of silicone material offering better reliability/lifetime than traditional epoxy material.

In parallel, with major YAG IP expiring from 2017, the phosphor market will face strong commoditization and price pressure. Consequently, market will only grow from nearly US$339 million in 2015 to US$346 million in 2021.

The LED packaging report (2016 edition) provides a comprehensive overview of all LED packaging aspects. Each step of the packaging process flow including equipment and materials used is described, along with associated trends. Associated technological breakthroughs are also analysed.

For the first time, researchers have created light-emitting diodes (LEDs) on lightweight flexible metal foil.

Engineers at The Ohio State University are developing the foil based LEDs for portable ultraviolet (UV) lights that soldiers and others can use to purify drinking water and sterilize medical equipment.

Nanowires were grown on titanium foil at The Ohio State University. CREDIT: Image by Brelon J. May, courtesy of The Ohio State University.

Nanowires were grown on titanium foil at The Ohio State University. Credit: Image by Brelon J. May, courtesy of The Ohio State University.

In the journal Applied Physics Letters, the researchers describe how they designed the LEDs to shine in the high-energy “deep” end of the UV spectrum. The university will license the technology to industry for further development.

Deep UV light is already used by the military, humanitarian organizations and industry for applications ranging from detection of biological agents to curing plastics, explained Roberto Myers, associate professor of materials science and engineering at Ohio State.

The problem is that conventional deep-UV lamps are too heavy to easily carry around.

“Right now, if you want to make deep ultraviolet light, you’ve got to use mercury lamps,” said Myers, who is also an associate professor of electrical and computer engineering. “Mercury is toxic and the lamps are bulky and electrically inefficient. LEDs, on the other hand, are really efficient, so if we could make UV LEDs that are safe and portable and cheap, we could make safe drinking water wherever we need it.”

He noted that other research groups have fabricated deep-UV LEDs at the laboratory scale, but only by using extremely pure, rigid single-crystal semiconductors as substrates–a strategy that imposes an enormous cost barrier for industry.

Foil-based nanotechnology could enable large-scale production of a lighter, cheaper and more environmentally friendly deep-UV LED. But Myers and materials science doctoral student Brelon J. May hope that their technology will do something more: turn a niche research field known as nanophotonics into a viable industry.

“People always said that nanophotonics will never be commercially important, because you can’t scale them up. Well, now we can. We can make a sheet of them if we want,” Myers said. “That means we can consider nanophotonics for large-scale manufacturing.”

In part, this new development relies on a well-established semiconductor growth technique known as molecular beam epitaxy, in which vaporized elemental materials settle on a surface and self-organize into layers or nanostructures. The Ohio State researchers used this technique to grow a carpet of tightly packed aluminum gallium nitride wires on pieces of metal foil such as titanium and tantalum.

The individual wires measure about 200 nanometers tall and about 20-50 nanometers in diameter–thousands of times narrower than a human hair and invisible to the naked eye.

In laboratory tests, the nanowires grown on metal foils lit up nearly as brightly as those manufactured on the more expensive and less flexible single-crystal silicon.

The researchers are working to make the nanowire LEDs even brighter, and will next try to grow the wires on foils made from more common metals, including steel and aluminum.

The latest Research and Markets report, “North America Light Emitting Diode (LED) Market (2016-2022)”,  indicates that the North American LED market is expected to reach $11,702.6 Million by 2022 growing at a CAGR of 9.7% during the forecast period.

The General lighting market dominated the North America LED market in 2015, and that trend is expected to continue until 2022, thereby achieving a market value of $ 4,563.6 Million by 2022 growing at a CAGR of 9.5% during the forecast period. The Automotive market is expected to reach a market size of $1,663.5 Million by 2022.

The U.S market dominated the North America LED market in 2015 and would continue until 2022 thereby achieving a market value of $ 8,683.3 Million by 2022 growing at a CAGR of 8.8% during the forecast period. The Canada application market is expected to reach a market size of $1,755.4 Million by 2022. The Mexico market would witness the growth rate of 14.2% during 2016-2022.

High brightness LEDs (HB-LEDs) are widely used in automotive, signals and signage in the North American region. Major mobile companies such as Apple Corporation have incorporated Organic LEDs (O-LED) in their mobile phones, which will contribute to the growth of the LED market. With widespread adoption in North America, the emerging economies also have started using LEDs in various applications, which should further add to the market growth, offering tremendous opportunities for the LED market players to enter into the LED market.

Using cutting-edge first-principles calculations, researchers at the University of California, Santa Barbara (UCSB) have demonstrated the mechanism by which transition metal impurities – iron in particular – can act as nonradiative recombination centers in nitride semiconductors. The work highlights that such impurities can have a detrimental impact on the efficiency of light-emitting diodes (LEDs) based on gallium nitride or indium gallium nitride.

This is a schematic illustration of Shockley-Read-Hall (SRH) recombination due to iron in GaN. Iron is a deep acceptor with a defect level (black line) close to the GaN conduction band (green). The charge density corresponding to this localized level is illustrated in the middle of the figure. Conventional SRH recombination (left) would proceed via electron capture from the conduction band into the defect level, but the overall rate would be limited by slow capture of holes because the defect level is far from the valence band (blue). The presence of excited states enhances the hole capture rate (right) such that the overall SRH recombination process becomes very efficient. Credit: Sonia Fernandez

This is a schematic illustration of Shockley-Read-Hall (SRH) recombination due to iron in GaN. Iron is a deep acceptor with a defect level (black line) close to the GaN conduction band (green). The charge density corresponding to this localized level is illustrated in the middle of the figure. Conventional SRH recombination (left) would proceed via electron capture from the conduction band into the defect level, but the overall rate would be limited by slow capture of holes because the defect level is far from the valence band (blue). The presence of excited states enhances the hole capture rate (right) such that the overall SRH recombination process becomes very efficient. Credit: Sonia Fernandez

For LEDs, high-purity material is essential to lighting technology, such as residential and commercial solid-state lighting, adaptive lighting for automobiles, and displays for mobile devices. Imperfections at the atomic scale can limit the performance of LEDs through a process known as Shockley-Read-Hall recombination. The operation of an LED relies on the radiative recombination of electrons and holes, which results in the emission of photons. Defects or impurities can act as a source of nonradiative recombination and prevent the emission of light, lowering the LED efficiency.

The UCSB researchers, in collaboration with researchers from Rutgers University, the University of Vienna, the KTH Royal Institute of Technology in Sweden and the Center for Physical Sciences and Technology in Lithuania, have identified that iron, even at concentrations less than parts-per-million, can be highly detrimental.

Transition metal impurities such as iron have long been known to severely impact devices based on traditional semiconductors such as silicon and gallium arsenide, leading these impurities to be referred to as “killer centers.” It is therefore surprising that little attention has been devoted to understanding the role of transition metals in recombination dynamics in GaN.

“A naïve application of Shockley-Read-Hall theory, based on an inspection of defect levels within the band gap, would lead one to conclude that iron in GaN would be harmless,” explained Dr. Darshana Wickramaratne, lead author on the paper. “However, our work shows that excited states of the impurity play a key role in turning it into a killer center.”

The UCSB scientists identified a recombination pathway by which iron can lead to severe efficiency loss. Sophisticated first-principles calculations were essential to identify and understand the role of the excited states in the recombination process.

“Taking these excited states into account completely changes the picture,” emphasized Dr. Audrius Alkauskas, another member of the research team. “We strongly suspect that such excited states play a key role in other recombination phenomena, opening up new avenues for research.”

The results highlight that strict control over growth and processing is required to prevent the unintentional introduction of transition metal impurities. Sources of iron contamination include the stainless steel reactors that are used in some growth techniques for nitride semiconductors.

“Increasing the efficiency of light emission is a key goal for the solid-state lighting industry,” said UCSB Materials Professor Chris Van de Walle, who led the research team. “Our work focuses attention on the detrimental impact of transition metals and the importance of suppressing their incorporation.”

Gallium nitride (GaN) has emerged as one of the most important and widely used semiconducting materials. Its optoelectronic and mechanical properties make it ideal for a variety of applications, including light-emitting diodes (LEDs), high-temperature transistors, sensors and biocompatible electronic implants in humans.

In 2014, three Japanese scientists won the Nobel Prize in physics for discovering GaN’s critical role in generating blue LED light, which is required, in combination with red and green light, to produce white LED light sources.

Now, four Lehigh engineers have reported a previously unknown property for GaN: Its wear resistance approaches that of diamonds and promises to open up applications in touch screens, space vehicles and radio-frequency microelectromechanical systems (RF MEMS), all of which require high-speed, high-vibration technology.

The researchers reported their findings in August in Applied Physics Letters (APL) in an article titled “Ultralow wear of gallium nitride.” The article’s authors are Guosong Zeng, a Ph.D. candidate in mechanical engineering; Nelson Tansu, Daniel E. ’39 and Patricia M. Smith Endowed Chair Professor in the Electrical and Computer Engineering department, and Director of the Center for Photonics and Nanoelectronics (CPN); Brandon A. Krick, assistant professor of mechanical engineering and mechanics; and Chee-Keong Tan ’16 Ph.D., now assistant professor of electrical and computer engineering at Clarkson University.

GaN’s electronic and optical properties have been studied extensively for several decades, said Zeng, the lead author of the APL article, but virtually no studies have been done of its tribological properties, that is, its resistance to the mechanical wear imposed by reciprocated sliding.

“Our group is the first to investigate the wear performance of GaN,” said Zeng. “We have found that its wear rate approaches that of diamonds, the hardest material known.”

Wear rate is expressed in negative cubic millimeters of Newton meters (Nm). The rate for chalk, which has virtually no wear resistance, is on the order of 10 2 mm3/Nm, while that of diamonds is between 10-9 and 10-10, making diamonds eight orders of magnitude more wear resistant than chalk. The rate for GaN ranges from 10¬-7 to 10-9, approaching the wear resistance of diamonds and three to five orders of magnitude more wear resistant than silicon (10-4).

The Lehigh researchers measured the wear rate and friction coefficients of GaN using a custom microtribometer to perform dry sliding wear experiments. They were surprised by the results.

“When performing wear measurements of unknown materials,” they wrote in APL, “we typically slide for 1,000 cycles, then measure the wear scars; [these] experiments had to be increased to 30,000 reciprocating cycles to be measurable with our optical profilometer.

“The large range in wear rates (about two orders of magnitude)…can provide insight into the wear mechanisms of GaN.”

That range in wear resistance, the researchers said, is caused by several factors, including environment, crystallographic direction and, especially, humidity.

“The first time we observed the ultralow wear rate of GaN was in winter,” said Zeng. “These results could not be replicated in summer, when the material’s wear rate increased by two orders of magnitude.”

To determine how the higher summer humidity was affecting GaN’s wear performance, the researchers put their tribometer in a glove box that can be backfilled with either nitrogen or humid air.

“We observed that as we increased the humidity inside the glove box, we also increased the wear rate of GaN,” said Zeng.

Zeng gave a presentation about the Lehigh project in October at the International Workshop on Nitride Semiconductors (IWN 2016) in Orlando, Florida. The session at which he spoke was titled “Wear of Nitride Materials and Properties of GaN-based structures.” Zeng was one of seven presenters at the session and the only one to discuss the wear properties of GaN and other III-Nitride materials.

Tansu, who has studied GaN for more than a decade, and Krick, a tribology expert, became curious about GaN’s wear performance several years ago when they discussed their research projects after a Lehigh faculty meeting.

“Nelson asked me if anyone had ever investigated the friction and wear properties of gallium nitride,” said Krick, “and I said I didn’t know. We checked later and found a wide-open field.”

Tansu said the group’s discovery of GaN’s hardness and wear performance could have a dramatic effect on the electronic and digital device industries. In a device such as a smartphone, he said, the electronic components are housed underneath a protective coating of glass or sapphire. This poses potential compatibility problems which could be avoided by using GaN.

“The wear resistance of GaN,” said Tansu, “gives us the opportunity to replace the multiple layers in a typical semiconductor device with one layer made of a material that has excellent optical and electrical properties and is wear-resistant as well.

“Using GaN, you can build an entire device in a platform without multiple layers of technologies. You can integrate electronics, light sensors and light emitters and still have a mechanically robust device. This will open up a new paradigm for designing devices. And because GaN can be made very thin and still strong, it will accelerate the move to flexible electronics.”

In addition to its unexpectedly good wear performance, said Zeng, GaN also has a favorable radiation hardness, which is an important property for the solar cells that power space vehicles. In outer space, these solar cells encounter large quantities of very fine cosmic dust, along with x-rays and gamma rays, and thus require a wear-resistant coating, which in turn needs to be compatible with the cell’s electronic circuitry. GaN provides the necessary hardness without introducing compatibility issues with the circuitry.

The Lehigh group has begun collaborating with Bruce E. Koel, a surface chemistry expert and professor of chemical and biological engineering at Princeton University, to gain a better understanding of the interaction of GaN and water under contact. Koel was formerly a chemistry professor and vice president for research and graduate studies at Lehigh.

To determine the evolution of wear with GaN, the group has subjected GaN to stresses by running slide tests in which the slide distance and the corresponding number of cycles are varied. The group then uses an x-ray photoelectron spectrometer (XPS), which can identify the elemental composition of the first 12 nanometers of a surface, to scan the unworn surface of the GaN, the scar created by the slide machine, and the wear particles deposited by the slide machine on either side of the scar.

The group plans next to use aberration-corrected transmission electron microscopy to examine the lattice of atoms beneath the scar. Meanwhile, they will simulate a test in which the lattice is strained with water in order to observe the variations caused by deforming energy.

“This is a very new experiment,” said Zeng. “It will enable us to see dynamic surface chemistry by watching the chemical reaction that results when you apply shear, tensile or compressive pressure to the surface of GaN.”