Thursday, November 28, 2013

The Cloud of Nations
Our world has been transformed by photonics. Our government, businesses and many of our personal activities are assisted or managed by what has been termed a “cloud” based infrastructure of computers, servers and a global communications network enabled by advanced photonic devices, lasers and fiber optics.  (Students: Click on the blue TDJ logo above for more on advanced photonics; become a scientist or engineer and help design our photonic future)  It can be argued that in addition to the physical world, governments and society also maintain a shadow, hyperspace existence in the cloud. The continued research, development and evolution of these technologies is integral to the growth of our economy, society and national defense. In recognition of this significance, SPIE, (The International Society for Optics and Photonics), OSA (The Optical Society), LIA (Laser Institute of America), IEEE Photonics Society and the American Physical Society have teamed to form the National Photonics Initiative (NPI). The NPI has embarked on a campaign to gain visibility with our government, key funding agencies and other industry partners to:

• Drive funding and investment in areas of photonics critical to maintaining US competitiveness and national security.

• Develop federal programs that encourage greater collaboration between US industry, academia, and government labs.

• Increase investment in education and job training programs.

• Expand federal investments supporting university and industry collaborative research.

• Collaborate with US industry to review international trade practices.

The discussion points above were excerpted from the NPI's recent white paper. For a more detailed description of the NPI agenda visit [1]

Photonic technologies have made a major impact on our society, simultaneously enabling advanced capabilities in computing, communications, manufacturing and their disruptive derivatives. Disruptive derivatives can vary in magnitude, yielding both laser guided weapons and unexpected video calls on your smart phone. For the purpose of this article, I will briefly review the history and development of the significant photonic technologies which have enabled what we now call the “cloud” and discuss some of the social-political ramifications.

The Development of the Laser

The fundamental physics behind the laser were theorized by Albert Einstein and Max Plank. Their theories were later confirmed by Rudolph W. Leydenburg, Valentin A. Fabrikant, Willis E. Lam and R.C. Retherford. In later years Charles Hard Townes developed the MASER (Microwave Amplification by Stimulated Emission of Radiation) a microwave amplifier, and later at Bell Laboratories worked with Arthur Leonard Schalow to conceptually define infrared and optical lasers. In 1958 Bell Labs applied for a patent on an optical MASER based on their work. At Columbia University Gordon Gould was conducting graduate work on thallium emissions and in November 1957 coined the acronym LASER to define Light Amplification by Stimulated Emission of Radiation. Conducting concurrent competitive research, Theodore H. Maiman operated the first solid state flash lamp pumped 694 nanometer synthetic ruby LASER at Hughes Research in Malibu California on May 16, 1960. [2]

The Fiber of the Cloud

Ten years later in 1970, Corning, Inc. announced that researchers Robert D. Maurer, Donald Keck, Peter C. Schultz, and Frank Zimar had developed and tested an optical fiber with a low optical attenuation of 17 dB per kilometer by doping silica glass with titanium. After conducting additional research they later produced a fiber with an attenuation of only 4 dB/km, using germanium oxide as the core dopant. This low loss quality over longer distances made fiber optic cable practical for telecommunications and networking. Corning became the world's leading manufacturer of optical fiber. [3]  

Continued improvements in fiber-optic materials have reduced line losses and capacity is being further optimized by utilizing multi-wavelength lasers for data transmission.  In 1990, Bell Laboratories transmitted a 2.5 Gb/s signal over a distance of 7,500 km without repeater regeneration by utilizing a soliton laser and an erbium-doped fiber amplifier.  In 1998, AT&T researchers transmitted 100 simultaneous optical signals, each at a data rate of 10 gigabits per second over a distance of 400 km utilizing dense wavelength-division multiplexing (DWDM) technology, allowing multiple wavelengths to be combined into one optical signal.  This technique increased the total data rate on one fiber to one terabit per second. [3A]

Our Global Photonic Schematic

Today, the backbone of our global communication system is comprised of an interconnected fiber optic cable infrastructure spanning the oceans, continents and civilized peripheries. The first undersea cable networks laid by ships supported telegraphy only, later evolving and expanding to accommodate voice communications via telephone. Network traffic was limited by the number of conductor pairs in the cables and long distance communication was sometimes problematic. Electronic relays were required to maintain call quality over long distances and troubleshooting was challenging as transoceanic cables lying on the sea bottom were difficult to locate, retrieve and repair. Over time, copper conductors were replaced with fiber optic cables and the communications revolution we are witnessing today began in earnest.  The AT&T Tech Channel features an informative video describing the evolution of the undersea fiber optic cable system [4] currently connecting our world.  Today there are over 321 major cables connecting the continents [5] (most under the sea) providing audio, high definition video and data transmission for our global community. Access to this advanced network can be had for the price of a smart phone. Utilizing services like Apple's FaceTime, Google Voice or Microsoft's Skype, it's possible to make real time two way video calls to friends, relatives or businesses almost anywhere on the Earth. This connectivity extends to most of the 3G and 4G cellular networks as well as the Wi-Fi systems we find at the corner coffee shop.

The Cloud of Nations

The early 1980s began the evolutionary period in our history in which we saw the rapid expanse of the internet along with the conception and realization of the “Cloud”. E-commerce and the idea of utilizing large data storage centers to strategically manage inventory and facilitate business transactions changed a fundamental business model which continues to evolve. The concept of in-country brick and mortar manufacturing sites and retail store distribution are being rethought every day. The acceleration to a web centric existence has transformed banking, trading, manufacturing, retail markets, personal communications and governments.  How so?

The internet connected cloud enabled by photonics has become the continuum in which multinational corporations and many governments reside. Independent of the necessity for fixed, localized manufacturing and distribution, large multinational corporations and small businesses can participate in an electronic global market place. A company in the US can own factories in Asia which produce and ship products world wide. Similarly, many foreign based companies own large manufacturing facilities in the US simplifying the distribution of their products here. The financial industry has been transformed as the international stock exchanges and banking institutions can now complete transactions in milliseconds or less. The undersea cable system made overnight international wire transfers possible. After business hours, banks in the U.S. can wire transfer large monetary holdings to banks in Australia or Asia where a more favorable overnight interest rate might be obtained (Australian banks typically pay a better overnight interest rate). Having earned significant interest overnight, these large monetary sums can be returned to the originating U.S. bank restoring local liquidity in time for opening business hours the following morning. The evolution in global business and trade among nations has brought our world closer together as an inter-dependent community made possible by the ubiquity of the world wide web, complements of advanced photonics.

Recent history has taught us that idealistic global business models don't always work to resolve disputes.  During the first Persian gulf conflict there was an exodus of Kuwait's displaced population to Europe and the United States. Interestingly, the Kuwaitis took their country with them. With electronic access to their country's wealth and the international banking system, Kuwait continued to function as a country and business entity in the cloud [6] via the web. For the first time in the world's history a nation existed and functioned electronically in hyperspace, untethered from its physical geography. Kuwait's political existence extended to its world wide connectivity with banking institutions, embassies and the United Nations. Similarly, Kuwait's finances and trading activities were all remotely accessible through the stock exchanges and SWIFT computers connecting the international banking system. Many of Kuwait's displaced citizens abroad established temporary residency in Europe and the U.S. aided by the continued financial support of their Kingdom which was fully functional on the web. This impressive logistical feat was managed remotely from hotel suites and conference rooms with laptop computers and direct access to major financial institutions outside the middle east.

Partly Cloudy

We like to think our internet system is fast and efficient, however high speed data crossing the oceans in milliseconds from distant continents at the speed of light are often stalled during local distribution and routing by network hubs which provide last mile connectivity to your home or business. A good example of this phenomenon can be found at fast food restaurants or coffee shops which provide wireless internet service to patrons. Sometimes while having coffee at a local establishment I'll use the wireless service to watch business news often featuring live programming feeds from Asia or Europe. Although this programming reaches the US in a fraction of a second, local network and distribution routers ration the data packets as they time share the program feed among thousands (or millions) of on line recipients. The resulting multiplexing and buffering of the programming can start and stop the video program on your computer every few seconds revealing the latency in the local distribution server as it slows to accommodate everyone, sometimes providing only a few program frames at a time. In these situations some news programming and streaming movies are not viewable and must be downloaded or viewed later on a faster network. Having experienced this phenomenon, I sometimes find myself exclaiming the virtues of 35mm movie film which runs at full program speed without interruption. Much of our wired internet infrastructure has not kept pace with 4G wireless and new 802.11ac distribution speeds and needs to be upgraded.

As compared with much of the world, the US is behind the curve in providing high speed Internet service.  A November 27, 2013 study reported by indicates the U.S. currently ranks 31st place in the world's Internet connection speed rankings. [7]  Current international speed ranking data can be found here [8] as conducted by  In addition we are behind in implementing the new IPv6 internet protocol which will enhance security and network standardization with the rest of the world.  As a technology leader among nations, we can do much better. Some might think this comparison unfair as the United States has a much larger land mass and population making a continual nation wide upgrade of Internet capacities and ISP inter-connectivity a capital intensive proposition. Realistically this cost consideration is not an impediment as our larger U.S. economy and consumer markets will rapidly consume any capacity upgrades made by Internet service providers. An abundant dark fiber infrastructure (previously installed and underutilized fiber optic cable) is readily available for augmentation of existing internet capacity. Rather than reconstruct my thoughts on this subject I will quote myself from recent comments I made while responding to an article appearing on Seeking, an investment newsletter web site.  Thomas D. Jay commentary posted on Seeking September 12, 2013 [9]

“Demands for Internet bandwidth will grow rapidly as new technology comes on line. The new Samsung S4 smart phone and Apple MacBook Air both feature the new 802.11ac wireless standard which operates on both 2.4 and 5 GHz at three times the speed of 802.11n. New 802.11ac routers are now available for commercial and home use. New smart phones and computers will soon get a wireless speed upgrade of 3X. Cable companies will be shifting strategies to accommodate consumers who stream video on the net in addition to those who subscribe to the standard menu of channels on their systems. The move is toward the wide scale implementation (much capacity has already been installed) of 256 bit QAM (Quadrature Amplitude Modulation) providing a greater number of cable channels and capacity. 4K video (4X 1080P resolution) is here now with 16K expected in about 18 months. Kudos to Verizon Fios and Google's 1 Gbit fiber optic system, however Sony is already deploying a 2 Gbit system in Tokyo [10] in anticipation of the need for extra system bandwidth. The US has been far behind in upgrading internet bandwidth and is in need of enhanced server center capacity in addition to fiber-optic distribution networks to homes and businesses. Fios 500 Mbit bandwidth at your home is of no use if the server you're connected to is overwhelmed with the demands of thousands of 20 Mbit subscribers. The price of bandwidth will gradually decrease if network capacity successfully paces growth forecasts. [18] An old programmers' law states that computer programs will grow in size to consume all available processing power and memory. Similarly, streaming internet applications and on line services will grow to consume all available bandwidth. The challenge is to build capacity ahead of the bandwidth demand curve and reduce cost. The principal of Moore's Law should be adapted by Internet Service Providers, by doubling network capacity every two years. Big data and network demand will require it.”  An additional note on QAM:  Where high spectral bandwidth and speed is required, networks usually employ 4096 QAM which has a greater constellation density (see Wikipedia, QAM) [11].

The Need for Speed

Latency in program distribution for both national (the US) and international carriers is problematic. The cumulative wide ranging delays in program arrival time experienced by viewers on the Internet can create an unfair playing field for consumers as well as private investors. On Wall Street, investment firms strategically locate their trading computers and servers as close as possible to the NYSE/EURONEXT and NASDAQ exchanges to gain precious millisecond (or better) advantages in price quote retrieval time and trade execution speed. Investment firms closest to the exchanges have an unfair (but legitimate) trading speed advantage over competitive firms and their investors. While monitoring the media, a slow internet connection can delay the arrival of streamed investment news by as much as twenty seconds or more. As a concerned former trader I've actually confirmed this phenomenon by comparing the arrival time of CNBC's cable broadcast programming with the identical program content streamed on the Internet. Again, the big investment houses win this scenario by employing the use of supercomputers which monitor stocks, market conditions, news events, and other exchanges in real time.  IBM has worked with TD Bank Financial Group to develop high speed computer architectures for use in financial decision making and transactions. [12]  A TD Financial Group computer can monitor markets, quotes, international news, and a wide range of critical investment decision parameters. Programmed with carefully developed algorithms the system makes trading decisions dynamically based on real time data input to the system. Investment firms utilizing supercomputers enable high frequency trading which is frowned upon by some, while many exchanges welcome the liquidity these trading firms provide.

Recently the Singapore Stock Exchange (SGX) Asia's largest bourse operator, expressed interest in attracting a larger number of high speed traders to help bring additional volume and liquidity to their market. SGX has invested over $250 Million in a computer trading platform which can execute transactions in 90 microseconds. Before large scale high frequency trading takes root on Singapore's SGX, trading circuit breakers must be implemented to prevent unforeseen market volatility from adversely impacting the market. Interestingly, high frequency traders have also avoided the SGX as it extracts a transaction fee of 20 basis points (0.2%) for shares traded on the exchange. (Source: Bloomberg News 10/29/2013) [13]

Global Clock Skew

For the purposes of global trading, communications and navigation, it is imperative that clocks are synchronized everywhere on the earth. In addition to navigation assistance, our GPS (Global Positioning System) satellites also provide a time standard beacon, complements of our friends at NIST (National Institute of Standards and Technology). You can program your PC clock to sync directly with NIST instead of your network server. On a global scale we observe that light travels approximately 186,282 miles per second and can circle the earth in 134 milliseconds (about one tenth of a second). If we consider the earth as a large computer, the continents might be compared to microprocessors interconnected by our global internet, milliseconds apart. If we observe our immediate, personal photometric sphere of existence, we find that in one nanosecond light travels one foot. How precise must our world be?

Chip Level Clock Skew

In the metrics of semiconductors nanosecond measurements are woefully imprecise and we must calibrate metrology in picoseconds in order to measure the speed of data traversing millimeter sized computer chips. In the semiconductor industry we refer to differences in data arrival time on a computer chip as clock skew. On a real semiconductor device, electrical signals on copper conductors travel at approximately 66-70% of light speed. An acceptable clock skew range approximating 20 to 200 picoseconds (pico = 10^-12) usually provides acceptable device performance but this specification can vary across device types and design. For each tick of the chip's master system clock, billions of transistor gates must be switched on and off in precise unison. A microprocessor operating at a 2 gigahertz clock speed must have sufficient temporal uniformity across the device so that the arrival time of gate switching signals are all within an acceptable time window. If the switching signal's arrival time falls outside this window, the microprocessor and the program it's running will crash. Careful design considerations go into device fabrication technique and wafer processing to ensure product and performance quality. More information on computer chip clock skew can be found in the paper “Skew Variability in 3-D ICs with Multiple Clock Domains” [14] Hu Xu, Vasilis F. Pavlidis, and Giovanni De Micheli, LSI - EPFL, CH-1015, Switzerland Email: {hu.xu, vasileios.pavlidis, giovanni.demicheli}

As clock speeds increase and device geometries decrease, the <10nm design node will pose additional technical challenges in device timing and data throughput. Recognizing this challenge, IBM designed computer chips that are interconnected with photonics, [15] enabling superior, high speed inter and intra-device communication utilizing beams of light instead of electrical signals carried on copper conductors. The evolution of light speed photonic interconnects for on and off chip communication will minimize concerns with clock skew across chips and networks as device structures shrink to critical dimensions (CD) <10nm and become stacked in dense arrays.

Free Space Optical Communications (FSO)

Free Space Optical Communication refers to the use of optical (visible) and the near optical (near visible) spectrum of light for line of sight transmission of communications and data. FSO is advantageous in areas where traditional radio frequency (RF) traffic is dense precluding the efficient use of available radio spectrum. FSO can also enable precisely directed communications by laser providing undetected, low profile communications for military or other secure communications. An FSO device is also a cool way of describing your infra-red TV remote control.

Wi-Fi and now Li-Fi

Recently there has been much hoopla over new computer networking hardware utilizing Li-Fi, a complementary technology to Wi-Fi (Wireless Fidelity). Li-Fi (Light Fidelity) utilizes LEDs (Light Emitting Diodes) producing visible or near visible infra-red light to propagate data transmissions in place of the typical RF (Radio Frequency) based 2.4 or 5 GHz 802.11n routers that have become ubiquitous. Some time ago I watched a TED presentation in which LED propagation of voice and data was pitched as a visionary concept. The fact is, the semiconductor industry recognized the significance of this idea and put it to work in the wafer fab many years ago.

Wafer Fab Free Space Optical Communications (WFFSO)

In the 1980's a major wafer fab equipment manufacturer provided automated guided vehicles to transport wafer lots and load process tools. The vehicles featured robotic arms and transported wafer cassette boxes. The mobile robots could open an on board cassette box and load/unload a process tool while communicating the wafer lot's status to the fab's Work In Process (WIP) tracking computer. The robots' communications with the fab was achieved with FSO cellular infrared LED clusters which were immune to radio frequency noise and interference from nearby process tools. The high intensity IR LED signals also bounced off the fab walls and ceilings to reach vehicles that were sometimes outside a direct optical path, or in shadowed areas created by fab technicians or equipment. Upload and download of vehicle and WIP programming was easily achieved using the IR data links. RF plasma etchers and other fab tools can create RF interference at their fundamental and harmonic frequencies and sometimes create “birdies” (spurious emissions) all over the RF spectrum. In most cases FSO/IR communications effectively eliminates this problem. The infrared cellular transceivers were ceiling mounted and unobtrusive. Quick visual confirmation of the system's activity was accomplished by “eyeballing” the ceiling mounted IR transceivers with a special IR sensitive viewing lens. Today it's easy to confirm IR LED activity with a smart phone. Most smart phone cameras are sensitive to a wide range of IR LEDs. Point your smart phone camera at an active IR LED (your TV remote for example) and you can see a corresponding bright white spot on your phone's display which is otherwise invisible to the human eye.

More recently it's been possible to operate long distance point to point FSO computer networks using lasers. The GeoDesy [16] wireless, directed laser system was introduced in 1996 and has been improved to provide competitive error free network connectivity without wires or radio frequency devices providing gigabit speeds. Two stationary laser units target each other with narrow directed beams at distances up to five kilometers to create a network link.

The First Cloud Over the Moon

Although fiber optic cables connect the continents of our earth, there has been no such lunar connectivity. However, very recently an earth based laser system was utilized to establish long distance connectivity with a space craft orbiting the moon. In October 2013, NASA's LADEE Laser Communications Demonstration [17] (LLCD) experiment made history by pulsing laser data 239,000 miles from the moon to the earth at a rate of 622 megabits per second (Mbps). The LLCD is on aboard NASA's Lunar Atmosphere and Dust Environment Explorer (LADEE), launched in September from NASA's Wallops Flight Facility on Wallops Island, Virginia. The LLCD is NASA's first laser based communication system for use with space craft. A ground station in New Mexico also successfully uploaded data to the LLCD system on board the LADEE lunar orbiter at an error free rate of 20 Mbps. LLCD is the precursor to future laser communications in space. In 2017 NASA plans to launch the Laser Communications Relay Demonstration (LCRD) as part of an expansive effort to advance communications in the space program.

Closing Thoughts

The National Photonics Initiative has established a presence in Washington, DC and is engaging key political figures, government agencies and seeking alliances to pursue policies which can accelerate the funding and development of advanced photonic technologies. Although the NPI is a lobbying advocate for a wide range of industrial photonic applications, the development of fiber optic and laser based communications systems have probably made the most significant contribution to improving our daily lives.  A major subset of the NPI agenda might also include a National High Speed Internet Initiative similar to the government's subsidized interstate highway projects started in the 1960s. The world wide migration to the cloud has altered the landscape of the globe. The concept and coexistence of cloud based commerce and national sovereignty is continuously changing in complexity as borders can be breached electronically with no passport required. We are able to trade internationally while probing a potential adversary's information systems, currency exchanges, and telecommunications. Thus the current debate concerning computer network sovereignty at national and international levels are complicated by the security and intelligence gathering initiatives of the many nations populating the web in a dynamic global market place. The solution set to this concern is problematic, engaging heads of state, and established international law. As Americans we also value personal privacy. Our government security agencies and commercial entities must be vigilant and ensure a delicate balance of national security and personal privacy. In addition to NPI's opening white paper discussion points the National Photonics Initiative must also assist in speeding the development of complementary photonic technologies which will ensure the security and integrity of our communications infrastructure.

The speed of light is sufficient for most of today's technological endeavors.  As we extend our reach further from the Earth with instruments and manned space craft, interplanetary clock skew will become (is) the next challenge. Exceeding the speed of light as defined by Einstein's limit will soon become a priority as our need for speed accelerates.  A quantum level initiative may define our future. In the interim I invite you to join me in supporting the National Photonics Initiative.

Thomas D. Jay
Semiconductor Industry Consultant
Thomas Dale

Thomas D. Jay YouTube Channel 


Corporate or private entities mentioned in this article are the respective owners of their logos, trademarks, service marks and intellectual property. Unless otherwise disclosed, Thomas D. Jay has no financial interest in companies referenced in blog articles or other published media communications. No representation is made to either buy or sell securities. Opinions expressed by Thomas D. Jay are his own. Thomas D. Jay does not employ or otherwise utilize/authorize third party agents to express his opinions, represent his interests or conduct business on his behalf except where formally contractually designated.

Acknowledgments and Reference Links

[1] The National Photonics Initiative White Paper

[2] The history of the laser, Wikipedia

[3] The history of fiber optics, Wikipedia


[4] AT&T Tech Channel, "Lightwave Undersea Cable System"
Courtesy AT&T Archives and History Center, Warren, NJ

[5] Telegeography Submarine Cable Map

[6] Kuwait conducts business in exile 1990, New York Times

[7] Techspot reports world internet speed rankings

[8] Current World Internet Speed Rankings,

[9] Thomas D. Jay Commentary, on Seeking

[10] World's fastest internet arrives in Tokyo, Tech

[11] Quadrature Amplitude Modulation, Wikipedia

[12] "From Gigahertz to Systems to Solutions; Our Industry in Transition", Bernard S. Meyerson, Ph.D., IBM, .PDF Published 2010

[13] Singapore SGX High Speed Computer Trading, Source:  Bloomberg News

[14] Skew Variability in 3-D ICs with Multiple Clock Domains” Hu Xu, Vasilis F. Pavlidis, and Giovanni De Micheli, LSI - EPFL, CH-1015, Switzerland Email: {hu.xu, vasileios.pavlidis, giovanni.demicheli}

[15] IBM Silicon Nanophotonic devices, IBM

[16] GeoDesy FSO PTP Laser Networking

[17] LADEE Laser Communications Demonstration, NASA

[18] Telegeography News Update 

Thursday, November 7, 2013

Florida's Governor Rick Scott and the Space Coast Economic Development Commission Announce Expansion of Embraer S.A.'s Business Jet Manufacturing in Melbourne

High High technology manufacturing is integral to Brevard County Florida's economy. On Friday October 29, 2013,     TDJ Technology Group provided special media coverage of an event hosted by the Space Coast Economic Development Commission. The EDC and regional political leadership welcomed Florida Governor Rick Scott and Embraer S.A. President and CEO Frederico Curado to make an important announcement. Ebraer S.A. is a Brazilian aircraft manufacturer with a major presence in the United States and intends to expand its facilities in Melbourne Florida where it currently employs 260 aerospace workers and customer service staff. The announcement celebrates Embraer's plan to expand its local manufacturing to include its newly developed Legacy 450 and 500 series business jets. The project will create 600 new jobs and add 250,000 square feet to Embraer’s facilities in Melbourne.  A new Engineering and Technology Center is planned to open in the second half of 2014.

The Space Coast Economic Development Commission is a Brevard County group of industry professionals who have teamed with community leaders and Florida's Governor Rick Scott to enhance the economic growth of the region. The commitment by Embraer exemplifies the confidence many have in the community, heavily weighted with aerospace and technology companies employing scientists, engineers and information technology professionals.  Brevard County Florida is a home to the NASA space program and many aerospace defense contractors.   Governor Scott acknowledged that the termination of the space shuttle program and more recent sequestration of federal funding has had a marked impact on the local aerospace and technology based economy.

I spoke with Lynda Weatherman, President and CEO of the Space Coast Economic Development Commission and Andy Anderson, Chairman of the Brevard County Board of Commissioners.  Both spoke to the urgency and commitment echoed by county political leadership to revitalize the local economy.  The economic recovery here has been slower than anticipated and challenging.  The announcement by Embraer S.A. is the result of community teamwork and a yet to be approved incentive package fomented by the EDC, local political leaders and Governor Scott.  The incentive package must be approved within 180 days but given the enthusiastic support by the region, expeditious approval is anticipated.   The special announcement joined by Governor Scott speaks to the confidence and broad based support for this project. Revitalization of the Brevard County economy is an on-going concern with Embraer S.A. marking an important inflection point in this effort.

Additional TDJ Technology Group media coverage of this Space Coast Economic Development Commission event can be found on the Thomas D. Jay YouTube Channel.

Thomas D. Jay
Semiconductor Industry Consultant
Thomas Dale Jay YouTube Channel


Special thanks to Julie Roslin, Business Manger at the Space Coast Economic Development Commission.

Referenced Web Links

Wednesday, August 7, 2013

The Continuing Evolution of Extreme Ultra Violet Lithography

On July 17, 2013 in a qualified safe harbor statement, an ASML press release (NASDAQ:ASML) announced its second quarter earnings and also indicated that its EUV platform, scanner imaging and overlay performance improved to levels enabling engagement with customers on a strategy targeting insertion at the 10nm logic node.  Additional discussions concerned strategies in which the first NXE:3300B systems running at currently available EUV power levels are utilized for production of future nodes at reduced throughput in place of mask strategies utilizing 193i multi-patterning techniques. Customers will be able to upgrade their systems as higher power levels become available and avoid a costly process transition from 193i to EUV HVM.  Combined with currently available multiple-patterning capabilities on existing 193i platforms, ASML could provide the best possible options for customers.  In the second quarter, ASML achieved good performance, stability and reliability of the pre-pulse source concept at power levels up to 55 Watts, and remains confident of targeted throughput of 70 wafers per hour in 2014.  ASML projects availability of upgraded systems with 125 wafer per hour throughput in 2015 provided further development of EUV power output progresses as planned. Additional strategies are being considered for continued funding of additional R&D on EUV.   450mm HVM insertion is anticipated in 2018.  News Source: Seeking 7/17/2013.

The semiconductor equipment industry participants driving the evolution of EUV (Extreme Ultra Violet) and nanometer scale lithography might be characterized as "tag team" players as process enabling technologies provided by an interactive vendor base will often trigger a cascade of new concepts and innovation frequently spawning “disruptive technologies”  The semiconductor lithography market is technologically unique, supporting many diverse products demanding application specific process solutions.  Although this market place can seem extraordinarily complex it is commonly bound by the semiconductor industry phenomenon known as Moore's Law which states that the number of transistors on semiconductor devices doubles approximately every two years, reducing costs.  During the past several years semiconductor lithographers have been pursuing the means to create a sufficiently powerful source of EUV (Extreme Ultra Violet) light for use in next generation lithography stepper/scanners.  These complex machines are used to mass produce semiconductor “computer chips” utilizing a photolithography process similar to conventional film photography, the difference being the “photo negative” is actually comprised of many individual photo masks (as many as 50 or more), providing the discrete image negatives required for each process layer in a complex semiconductor device.  EUV light source technology is being pursued as the emitted 13.5nm light's wavelength is short enough to produce high resolution circuit imaging for present and future 14, 10 and 7 nanometer scale device structures. Although several well established companies in the field of semiconductor lithography have participated in the development of EUV technology the expense and diminishing returns on investment have narrowed the playing field to a few.  Lithography giant ASML based in the Netherlands recently purchased a key lithography light source supplier, Cymer, in an effort to consolidate EUV development efforts and secure the timely delivery of critical source components.  Underscoring the strategic value of EUV lithography are recent investments made in ASML by Intel, Samsung, and TSMC which total more than $6 Billion. To date, ASML and Cymer have produced excellent EUV lithography imaging with an available light source power output approximating 55 watts.  This is about half the power required to support the throughput of high volume manufacturing (HVM).  After years of industry anticipation, delayed development of EUV power output has put some semiconductor industry professionals in an uneasy holding pattern.  Although multiple patterning techniques for 193i lithography provide feasible lithography solutions in the interim, most would prefer a simplified EUV lithography process solution requiring fewer mask levels at reduced cost.  In light of ASML's recent announcements some are reconsidering the current status of EUV lithography development and the implementation of available alternatives such as double patterning and Directed Self Assembly (DSA) techniques.  In an effort to gain additional insight and inspiration from established experts in the EUV community my inquiries revealed that many are reluctant to speak to the current situation and perhaps for good reason. It is difficult to speculatively comment on the efforts of lithography industry experts who have diligently achieved the excellent results obtained thus far.  It appears we've hit a physical wall which has temporarily stalled EUV development and delayed the HVM insertion time line.  As we have learned over the years, difficult tasks sometimes take longer to complete.

In the interim 24 months ahead we must also realize that the semiconductor industry will assimilate a new crop of scientists, electrical engineers, sales, marketing and application specialists along with the traditional complement of Wall Street industry analysts and investors.  To the new and uninitiated intelligentsia, EUV technology might require some explanation.  Educating new entrants to our industry is always a challenge.  As a technologist you might have a friend or relative who've asked about your work in semiconductors and observed their eyes glaze over as you explain.  As we pause to await news of higher EUV output power, I will utilize the available “dwell time” to provide an EUV primer for new industry participants and observers and hopefully fill in the blanks for our current, friends and colleagues.  Where to begin?                    

Plasma 101

Unlike solid, liquid or gas, plasma, the fourth state of matter exists within a narrow, low pressure domain approximating 1/760th of our atmosphere at sea level.  Positively ionized plasma can be created in this low pressure regime by inducing high energy electron collisions with gas molecules displacing orbital electrons normally found in the stable atoms comprising the gas.  The imparted electrical imbalance causes the gas molecules to gain a net positive charge as there are more positive protons remaining in the atoms/molecules than negatively charged electrons.  The resulting plasma is a glowing cloud of charged particles and gas molecules having a positive electrical potential and the conductivity of copper while simultaneously emitting light comprised of photons with the characteristic color of its elemental spectra.  Utilizing Italian physicist Torricelli’s scale, the pressure of the air at sea level is 760 Torr.  More specifically, the pressure at which electrical gas plasma can be created and sustained approximates 5x10-4 Torr, a near vacuum at less than 1/760th of the pressure of our atmosphere.  At an altitude of 84 kilometers above the earth and well below the international space station, the pressure in near outer space falls within this range and is conducive to the natural creation of plasma known as the aurora borealis.  This phenomenon occurs when high energy “solar wind” from the sun ionizes oxygen, nitrogen and other gases in the upper atmosphere in the presence of earth's magnetic field.  Back on terra firma, high vacuum technologists design and maintain carefully engineered process chambers simulating pressures ranging from ambient atmosphere to the ultra high vacuum of deep outer space.  Scientists and electrical engineers utilize these physical phenomena and resources to generate, contain and control high energy plasma by electrically ionizing specially selected process gases at precise pressures in high vacuum systems. Plasma etchers utilized in semiconductor wafer fabs were among the first process tools to adapt this technology.  In this application, plasma is generated to produce ions for etching semiconductor wafers.  The incidental light emitted from the plasma can be utilized to monitor the process.  The primary methods for ionizing these gases include hot cathode/filament electron discharge, microwave excitation, and RF (Radio Frequency) energy excitation.  Interestingly the Federal Communications Commission has designated special radio frequencies for use in plasma etch systems.   A common frequency utilized for RF plasma etchers is 13.56 MHz.  As a ham radio operator I am more spectrum conscious than most and when near a wafer fab I sometimes tune in with my mobile transceiver to listen for “etchers on the air”.  The received signal is not particularly exciting and is usually a continuous uninterrupted carrier wave sometimes producing a loud heterodyne when tuned in the single side band mode.

In this semiconductor manufacturing application of physics, positively charged high energy ions are created in these special plasma systems and utilized to etch semiconductor circuits on silicon wafers.  The ions having mass and kinetic energy bombard the surface of product wafers and selectively etch surfaces to form intricate device structures and circuit patterns from the metal film and materials left behind.  A variety of gases are utilized to etch process specific materials on product wafers.  Some of the gases utilized are also chemically reactive and can selectively enhance the etch rate of specific materials on a wafer's surface.  Other noble gases are chemically inert (non-reactive) and advantageous for a wide range of etch applications.  Plasma etch chambers emit a characteristic light whose primary color/wavelength is determined by the elemental composition of the process gas.  Argon gas for example emits a purple glow which is pleasing to the eye. This spectral phenomenon is often utilized as a process indicator to determine when a critical etch step must stop. While etching a thin film metal, a faint light emitting plasma with the metal's signature color/wavelength is also produced. Using optical end point detectors, process engineers can sample the spectrum of light emitted from the plasma to detect the metal's specific signature wavelength.  When this spectral signature disappears the metal etch process is deemed complete as there is no remaining metal being ionized.  This “end point” signal is typically used to automatically stop the etch cycle providing highly accurate process control and negates any manual intervention by wafer fab engineers.  A key observation of these phenomena are that elementally specific positively charged particle ions (mass specific) and photons (wavelength specific light) are created during the ionization process.  The science of plasma physics for etching semiconductor circuits has been refined over the years by companies like Applied Materials, Lam Research and Veeco Instruments (there are many others) however, the plasma physics required to produce 13.5nm EUV for lithography are quite different and much more challenging.  

Producing Extreme Ultra Violet Light With Plasma

For several years the semiconductor industry has been engaged in the research and development of next generation nanometer scale lithography requiring a precision stepper/scanner equipped with a 13.5nm light source.  For lithography applications, plasma is created for the purpose of generating light at specific wavelengths.  The production of ions and other charged particles is incidental and sometimes suppressed to avoid etching and erosion of sensitive components in the source and wafer stage. Utilizing the most current methods for producing this light, lithography research engineers create a plasma which emits 13.5nm EUV from the resulting ionization of xenon gas (Xe), or the solid metal tin (Sn).  As a gas, xenon ionizes easily but has a lower energy Conversion Efficiency (CE) than tin.  ASML has chosen solid tin as the plasma feed material for its EUV light source because of its higher energy conversion efficiency.  Creating an ionized tin plasma is more difficult and also creates unwanted particulates which can potentially contaminate EUV masks and product wafers. Although the 13.5nm spectra has been dubbed EUV (Extreme Ultra Violet) it can also be described as “soft x-ray” radiation or “vacuum EUV” as it is absorbed by many gases and propagates most efficiently in a high vacuum environment.  Extremely short wavelength EUV light will enable semiconductor lithographers to print ever smaller transistors and associated circuitry on computer chips reducing CD features (Critical Dimensions) down to nanometer scales and below.  Utilizing a current 22nm fabrication process, a state of the art Intel Quad core i7 microprocessor introduced in the year 2012 contains 1.4 billion transistors.  Moore's law marches on and future process nodes will soon shrink to 14, 10 and 7nm.  Imagine your smart phone's chip sets with the processing power of another billion or so transistors.

A Best Of Breed EUV Source?

As there has been significant wide ranging research on the subject of EUV source technology, I will focus on the most current technology utilized by ASML and will discuss possible alternately efficient means of EUV production. Having conducted research on past experimentation and published papers, I will direct attention to recent prior work having potential to enhance EUV source performance.  By “borrowing” concepts from the most promising efforts I will propose a possible “best of breed” hybrid EUV source design.  I suspect similar dialogue and debate has been conducted previously.  My intention is to foment new discussion which might yield solutions providing the EUV power output and performance required for HVM lithography (High Volume Manufacturing).   

ASML's EUV Source Design

ASML/Cymer's current technique for production of EUV light utilizes Laser Produced Plasma (LPP).  In this different approach to producing plasma, small droplets of tin approximating 30 microns in diameter are injected into an ionization source and targeted by a solid state “pre-pulse” laser.  The imparted energy from the pre-pulse laser enlarges the tin droplet and raises its energy state, providing a larger cross sectional profile for a more powerful 20 kilowatt CO2 laser.  After pre-pulse conditioning, the tin droplet is targeted with the larger CO2 laser resulting in the high energy evaporation and ionization of the tin, releasing 13.5nm EUV light as a result.  A continuous stream of tin droplets and sustained laser interaction produces stable EUV light output commensurate with requirements for precision lithography and dosimetry.  Utilizing this technique, current EUV power output levels of 55 watts have been achieved, about half the power required for HVM (High Volume Manufacturing).  In the quest for more EUV power output, I suspect ASML and Cymer will experiment with larger, more powerful lasers while optimizing other coincidental parameters.  While successfully producing EUV light, an Sn/LPP source produces a large number of tin particulates as the targeted tin droplet residue collects on exposed surface areas in the ion source and wafer stage. Reflective EUV optical mirror surfaces can lose efficiency (reflectivity) when coated with tin, while EUV masks and product wafers can be compromised if contaminated with particles as small as one nanometer.  A more recent technique for keeping surfaces clean is to direct a beam of hydrogen ions from a small, integral ion source at critical optics and wafer stage components.  Experiments implementing a hydrogen plasma clean cycle as part of a periodic maintenance regimen have also been suggested. The challenges in the design and implementation of EUV technology for high volume manufacturing are formidable.  

Z-Pinch Discharge Produced Plasma Source Technology

A Z-pinch source, consists of an insulated containment cylinder with electrodes attached on either end, mounted inside a vacuum chamber.  For the creation of EUV light emitting plasma, a gas such as SnH4 (Stannane) or xenon (Xe) is fed into the cylinder and pre-ionized with the resulting plasma having the electrical conductivity approximating that of copper.  A large pulse of electrical energy stored in a bank of capacitors or from a power supply is applied to the electrodes at the ends of the cylinder.  As a result, a large current flows through the electrically conductive plasma causing it to contract as a phenomenon known as the Lorentz force compels the mutual attraction of the ions flowing uniformly along the z-axis of the plasma. The term z-pinch is derived from the fact that the current flows along the z-axis of the plasma as it is being compressed and pinched. The contraction continues until the plasma becomes highly dense and further compression is resisted by the gas pressure comprising the plasma. Ultimately, the maximized “pinch” density releases a large burst of energy comprised of ions and characteristic spectral light.  In the case of SnH4 and Xe, Discharge Produced Plasma EUV light with a wavelength of 13.5 nanometers is emitted in the process.  A continuous, controlled flow of source gas and carefully timed high current pulses through the source cylinder repeat this cycle, sustaining the plasma and output of EUV light.  This is achieved by pulsing the power supply and the resulting high current flow through the conductive plasma between 6 and 8 thousand times a second.   A power supply pulsed at approximately 7 Khz usually delivers maximum output for this design.  Z-pinch operation can be further optimized by controlling the power supply frequency, pulse width and amplitude providing effective wave form control and regulation of power fed to the source.   With the power supply operating at a the relatively low frequency of 7 Khz, a computer operating at 1.5 to 2.0 GHz can be utilized to sample the sinusoidal/waveform input power and resulting EUV output at the wafer stage making it possible to close loop control associated lithography dosimetry and related functions.

Ushio's Sn/DPP EUV Source Design

At the SPIE 2007 International EUVL Symposium, Ushio published a paper on results obtained from experimentation conducted on a Discharge Produced Plasma EUV source utilizing SnH4 (Stannane) as the feed material: Development of Sn-fueled high-power DPP EUV Source Enabling HVM.  In this design a Z-pinch DPP source was fed with gaseous SnH4 providing the energy conversion efficiency of tin while affording the simplicity of source gas handling with conventional mass flow controllers and pressure measurement instruments.  Interestingly, Ushio's Discharge Produced Plasma (DPP) source produced 62 watts of EUV output utilizing gaseous SnH4 as compared with ASML's recently quoted EUV power output of 55 watts obtained with a Laser Produced Plasma sustained with source fed solid tin droplets.  A power output comparison of the two technologies builds a case for continued development of an SnH4 Discharge Produced Plasma EUV source with a more simplistic design.  An additional benefit of using SnH4 as a source feed material is it minimizes the erosion of electrodes and critical surfaces in the z-pinch tube.  

Zplasma's Xe/Stable DPP Z-Pinch Source With
Sheared Flow Stabilization (SFS)®

In my previous blog articles I've had the opportunity to interview Henry Berg, CEO of Zplasma, and discuss his company's Xenon based Discharge Produced Plasma EUV source.  Zplasma's Stable DPP source is quite different from the generic DPP description above.  Zplasma's stable pinch operation results from a patented, proprietary technique called Sheared Flow Stabilization (SFS)® which stabilizes the plasma and eliminates explosive pinch terminations. The stable pinch operation of an SFS design DPP source is required to produce the high power levels required for HVM lithography. As described by Henry Berg, there are six critical advantages to Zplasma's SFS source technology:

a) Longer Light Pulse: SFS pulses are 10 to 100 times longer due to their stable nature, allowing for more light collection.

b) High Power without High CE: Long SFS pulses and side-on optical collection access increase throughput and lower required CE, enabling HVM operation with xenon and eliminating the need for molten tin.

c) No Debris: SFS ends the plasma pinch without explosive termination, eliminating high-energy debris.

d) Low Instantaneous Power: SFS pulses allow for EUV light production without the high instantaneous power levels that cause electrode thermal stress and ablation.

e) Dose Uniformity: SFS allows the length of each EUV pulse to be adjusted under control of the power supply, allowing for extremely accurate dose uniformity.

f) Adjustable Geometry: SFS makes pinch geometry adjustable for optical matching to the stepper IF.
Given Discharge Produced Plasma (DPP) and Laser Produced Plasma (LPP) design concepts, how might we improve EUV power output and system performance?

Minimizing Source Plasma Opacity
Maximizing EUV light Output

EUV light emitted from a tin or xenon plasma discharge is partially absorbed by the opacity of source plasma.  More specifically, the 13.5nm light is readily absorbed by many gases and solid materials, inclusive of the low pressure vapor state of the tin or xenon utilized to create the plasma. This means that the intensity of the emitted EUV light diminishes as it is partially absorbed and diffused during its transit through the plasma cloud and transport to the stepper's intermediate focus (where the EUV light is collected).  A reduction in the rate of absorption and diffusion by the plasma can sometimes be achieved by reducing the pressure of the gas comprising the plasma. This has the effect of reducing the density of the plasma, lessening its opacity, and potentially increasing power output by increasing the transmission efficiency of the emitted EUV light.  This pressure adjustment is critical as the plasma will be extinguished if the pressure is too low (insufficient gas density/pressure to sustain the plasma) and will similarly cease to ionize if the pressure is too high. Experimentation is required to determine the optimal pressure and plasma density for maximum production and transmission of EUV light.  I suspect ASML has optimized this parameter on their Sn/LPP EUV source.  This technique is discussed in a 2005 paper:  Comparison of experimental and simulated extreme ultraviolet spectra of xenon and tin discharges by E.R. Keift, K. Garloff, J.J.A.M. van der Mullen and V. Banine.  

Maximizing the Transmission Efficiency
 of EUV Light to the Wafer Stage

To carry the EUV transmission efficiency discussion further, in some EUV system designs the wafer stage chamber is maintained at the same vacuum pressure (approximately 5x10-4 Torr), as the plasma source chamber and share the vacuum environment sustained by the pressure controllers and vacuum pumps supporting the EUV plasma. The pressure inside the wafer stage area is also critical as additional absorption and diffusion of the EUV light can occur there during transit from the intermediate focus coupling (IF) to the wafer stage.  Reductions of EUV energy propagation efficiency in the wafer stage chamber can be pressure related and contribute to path loss concerns exemplified in the source plasma opacity discussion above.  It is probable that higher levels of EUV dosing can be obtained by maintaining the wafer stage chamber under high vacuum pressure at <1x10-7 Torr. 

An excellent example of increased energy transfer efficiency under high vacuum conditions and the resulting effects on lithography are illustrated in a 2006 Nano Letters paper by Benjamin D. Myers and Vinayak P. Dravid, Variable Pressure Electron Beam Lithography (VP-eBL): A New Tool For Direct Patterning of Nanometer Scale Features on Substrates With Low Electrical Conductivity.  I found this paper quite interesting in that it addresses concerns common to both EUV and eBeam technologies.  The primary purpose of this paper was to illustrate a methodology for mitigating surface charging effects on electrically insulated wafer substrates patterned with eBeam lithography while simultaneously optimizing the resolution of the patterned images.  The paper also illustrates how differences in vacuum pressure effect eBeam (energy) propagation.  In the experiment, a differentially pumped vacuum system maintained an electron beam column under high vacuum while enabling operation of the wafer stage within pressures ranging from high vacuum to 3 Torr.  It was determined that a decrease in beam-gas path length (BGPL) and resulting electron beam scattering occurring at higher pressures mitigated surface charging and secondary electron proximity effects while enhancing image resolution.  However, the beam scattering also reduced the number of electrons available to dose the patterned resist requiring longer write times.  Longer write times or specialized dose requirements can now be optimized with computational lithography techniques inclusive of shot tasking and dose modulation.  To follow the discussion of exhibits in this paper, click on the link above. In Figure 1, exhibit (A) illustrates the insulated substrate surface charge induced displacement and distortion of eBeam imaging at high vacuum pressure.  Subsequent improvement in image resolution and accuracy are illustrated in exhibit (B) at .4 Torr, and in exhibit (C) at 1 Torr.  In Figure 2, Exhibits (A) through (D) and similarly Figure 3, Exhibits (A) through (C), illustrate eBeam induced surface trenching in the resist which decreases when the pressure is increased.  These examples illustrate the increase in energy transfer efficiency under high vacuum conditions and resulting decrease at higher pressures.  Figure 4, illustrates how electron beam dose requirements increase at higher gas pressures along with the interactive effect on minimum line width.  While there have been studies and process gas specific absorption coefficients established for electron beams and spectral light across a range of pressure regimes, it is well recognized that both forms of energy propagate most efficiently in a high vacuum environment, hence the term “vacuum EUV”. 

It follows that an EUV source and wafer stage chamber maintained at separate optimal pressure levels, approximately 5x10-4 Torr for the source (plasma opacity optimized),  and <10-7 Torr for the wafer stage chamber, more efficient transfer of EUV light can likely be obtained. This scenario requires a means of isolating the two vacuum chambers, perhaps by differential pumping, and transferring the EUV light through a transparent window or aperture at the intermediate focus (IF).  There are candidates for low loss EUV window material which could facilitate such a design.  In addition, there is a patented EUV source design featuring a differentially pumped source and wafer stage chamber coupled by a gas flow/conduction limiting optical aperture which efficiently conducts the EUV light across the two pressure regimes.

Dueling Etendue – A Hybrid Dual DPP Source Design
 to Double Power Output

The search for high power EUV production techniques has been primarily focused on identification of energy conversion efficient source materials (solid and gaseous) and the most efficient means of ionizing them.  Recent research has determined that the best two ionization techniques are Discharge Produced Plasma (DPP) and Laser Produced Plasma (LPP).  As I point out in this article, the two techniques seem to be tied at maximum EUV power levels of 62 watts for DPP and 55 watts for LPP.  An approach to achieving higher power levels might be to operate two DPP sources simultaneously within one vacuum system, theoretically generating twice the power output.  The challenge would be optically coupling two sources to the stepper IF to truly double the power.  This might not be easily achieved.   Simultaneously capturing the maximum brilliance of two plasma sources could be challenging as the hardware interface to the IF must accommodate the optimal plasma profile angles for two discrete light sources to effect the most efficient etendue.  It's possible Bragg Cell EUV mirrors could be utilized to assist in effecting dual source DPP etendue.  Any concerns with absorptive losses from the Bragg Cell mirrors would be negated by the additional power afforded by the dual source design.

In my opinion a good candidate design technology for a hybrid EUV source resides with Zplasma.  The original system built by Zplasma utilized Xe as its source gas feed which produced a 2% bandwidth energy conversion efficiency (CE) of 0.5% which is low, however the efficiency will increase to 1.5% with a power supply optimized for the source.  By substituting Xe with SnH4, the energy conversion efficiency could be increased further to become competitive with Sn Laser Produced Plasma efforts to date. In addition to the increase in CE, higher EUV output power levels might be obtained if Zplasma's patented Sheared Flow Stabilization Z-pinch performs as well with SnH4 as it does with Xe.  Zplasma's system design might also accommodate the concept of Dual DPP sources as etendue match concerns might be minimized by its z-pinch source characteristics affording side on collection of light from the plasma.  In addition, the DPP configuration could accommodate an optional H2 plasma clean maintenance regimen.  A successful technique utilized by Gigaphoton in its source development program minimizes tin (Sn) particulate contamination by optimal positioning of a superconducting magnet in the source which deflects tin particles away from critical wafer stage and optical components.  Assuming a viable hybrid EUV source design is identified, how might further R&D funding be obtained given the current economic climate?

A Proposal for a New National Photonics Initiative Agenda Funding High Priority Challenges
 to Manufacturing

An important SPIE co-sponsored initiative took place on February 28, 2013 in Washington, DC, titled, “Optics and Photonics: Lighting a Path to the Future.”  Other organizations co-sponsoring the event included the IEEE Photonics Society, Optical Society of America (OSA), IEEE Photonics Society,  American Physical Society, and the Laser Institute of America.  The event was attended by many government agencies who traditionally sponsor research. The goal of the conference was to foster better government collaboration with American optics and photonics industries and forge a National Photonics Initiative (NPI).  The value of this proposed initiative is best exemplified by recent (non-government) semiconductor industry manufacturers' investments in ASML and Cymer.  The collaboration of industry and government in key photonic and optical science endeavors could distribute R&D expenditures among participants and reduce costs across many disciplines. While I was researching and writing this blog article, the benefits of a coordinated NPI program became more apparent.  After reviewing the many technical papers written on the subject of EUV and related technologies, I believe it is accurate to say that the level of participation by private industry, universities and government sponsored funding is probably unprecedented, and has created a reciprocally proportionate wealth of intellectual property and patent filings spanning many entities and corporations.  While we all support the concept of independent research, competitive concerns with ownership of applicable intellectual property may have slowed investment in the development of the EUV program. It is my opinion that a hybrid EUV source with superior power output and HVM performance might be assembled from the assortment of promising technologies and IP developed thus far.  Should a hybrid EUV concept be successfully proven, ensuing IP cross licensing concerns could potentially inhibit cooperation and progress.  This concern has been addressed previously by SEMATECH and other groups cooperatively sponsoring other past and present research programs.  The current pause in the pace of EUV development should foment renewed enthusiasm for jointly developed technology, intellectual property and reduction of R&D costs outside the domain of the foundries in order to prevent further stratification and/or dissolution of the traditional semiconductor equipment vendor base.  

On July 25, 2013 I received the following communication from NIST from which I have excerpted the following:

MEMORANDUM FOR Advanced Manufacturing Distribution 
From: The NIST Advanced Manufacturing Technology Consortia (AMTech) Program Team

“It is my pleasure to inform you that the National Institute of Standards and Technology (NIST) has released a Federal Funding Opportunity (FFO) for the Advanced Manufacturing Technology Consortia (AMTech) Program.  This is a new competition for awards to establish new or strengthen existing industry-led consortia in planning research that addresses high-priority challenges impeding the growth of advanced manufacturing in the United States.  NIST anticipates awarding $4 million in early 2014 as a result of this AMTech announcement.  Awards are expected to be up to two-years in duration and range between about $250,000 and $500,000.  As a leading voice in advanced manufacturing, we believe your members or constituents may have interest in AMTech and this newly announced funding opportunity.  AMTech-supported consortia will identify and prioritize long-term, pre-competitive industrial research needs; enable technology development; and create the infrastructure necessary for more efficient transfer of technology.  Teaming and partnerships are strongly encouraged including participation by the full value chain, including small-and mid-sized firms.  By convening key players across the entire innovation life cycle, the objectives of the AMTech Program are to eliminate critical barriers to innovation; increase the efficiency of domestic innovation efforts; and collapse the time scale to deliver new products and services based on scientific and technological advances. The end goal will be a growth of advanced manufacturing in the U.S. and an increase in the global competitiveness of U.S. Companies.  The AMTech Program will host two webinars on August 15, 2013, and August 20, 2013, at 2:00 p.m. Eastern time.  Participants are required to register in advance.  The events will offer guidance on the AMTech Program and preparing proposals, and will provide an opportunity to answer questions from the public about the program.  Participation in the free event is not required to submit an application.  Information on and registration for the event is available at  To assist in identifying potential collaborators for a consortium, the AMTech Program is creating a LinkedIn group where individuals can provide their area(s) of interest and communicate with each other.  Visit for more information about joining.”

In considering this proactive initiative by NIST, I believe it is accurate to state that timely availability of EUV lithography technology is one of the high-priority challenges impeding the growth of more advanced manufacturing in the United States.

I propose that SPIE and the industry groups comprising the National Photonics Initiative consider establishing an additional agenda coordinated with SEMATECH to pursue government funding of advanced research in EUV development.  Further, there should be emphasis on directing available grant funding to the many promising start up companies whose entrepreneurial potential for technological contribution have been precluded by the prevailing economic climate in Silicon Valley, Wall Street and the larger U.S. economy.  We all applaud the cooperation of major players such as Intel, Samsung, and TSMC as they collectively support the efforts of ASML, but we must also ensure proactive support for promising, small business based entrepreneurial start up companies that have traditionally made the best and brightest contributions to Silicon Valley and the semiconductor industry.

Thomas D. Jay
Semiconductor Industry Consultant
Thomas D. Jay  Investment Commentary
Thomas D. Jay  YouTube Channel

Click here for a PDF download of this article.

Corporate or private entities mentioned in this article are the respective owners of their logos, trademarks, service marks and intellectual property. Unless otherwise disclosed, Thomas D. Jay has no financial interest in companies referenced in blog articles or other published media communications. No representation is made to either buy or sell securities. Opinions expressed by Thomas D. Jay are his own. Thomas D. Jay does not employ or otherwise utilize/authorize third party agents to express his opinions, represent his interests or conduct business on his behalf except where formally contractually designated.

Correction:  A previous reference in this article was made crediting Ushio for a technique utilizing a superconducting magnetic field to mitigate tin particulate debris.  This statement was made in error and corrected on August 8, 2013.  Gigaphoton developed the superconducting magnet particulate mitigation technique as reported by Business Wire on July 1, 2013.  My apologies to Gigaphoton and Ushio for this unintentional error.  
Thomas D. Jay, August 8, 2013

Acknowledgments and Links

ASML News Release Source: Seeking 7/17/2013

Intel's Quad Core i7 Microprocessor
Intel Corporation, You Tube

ASML/Cymer's current technique for production of EUV light, ASML, You Tube

The challenges in the design and implementation of EUV technology for high volume manufacturing are formidable,
Jos Benschop, ASML, You Tube

Z-Pinch DPP Source Reference,

Development of Sn-fueled high-power DPP EUV Source Enabling HVM, Yusuke Teramoto, Zenzo Narihiro, Daiki Yamatani, Takuma Yokoyama, Kazunori Bessho,
Yuki Joshima, Takahiro Shirai, Shinsuke Mouri, Takahiro Inoue, Hiroshi Mizokoshi, Gohta Niimi*, Tomonao Hosokai*, Hironobu Yabuta, Kohkan C. Paul, Tetsu Takemura, Toshio Yokota, Kiyoyuki Kabuki, Koji Miyauchi, Kazuaki Hotta, and Hiroto Sato Gotenba Branch, Hiratsuka Research and Development Center, Extreme Ultraviolet Lithography System Development Association  *Ushio Inc., Henry Berg, CEO, Zplasma, Inc.

Comparison of experimental and simulated extreme ultraviolet spectra of xenon and tin discharges by E.R. Keift, K. Garloff, J.J.A.M. van der Mullen and V. Banine

Variable Pressure Electron Beam Lithography (VP-eBL): A New Tool For Direct Patterning of Nanometer Scale Features on Substrates With Low Electrical Conductivity, Benjamin D. Myers and Vinayak P. Dravid

National Photonics Initiative

NPI Agenda