Electronics innovation is inching tantalizingly closer to the day when treating neurological disorders such as epilepsy and migraine could be as easy and convenient as dropping into a medical clinic for a minor medical procedure – brain surgery. What today is highly invasive surgery promises to be reduced to a doctor’s office visit as chip engineers work to tether the delicate, complex neurochemical workings of the human brain to the hard wiring of electronics. The goal is to use electrical stimulation to trigger the release of therapeutic doses of natural brain chemicals using small implantable devices in order to restore normal brain functioning, reduce human suffering and help slash the financial burden to economies around the world. The advances come as neurological disorders remain the leading cause of disability worldwide, afflicting up to 1 billion people, a number projected to rise sharply in the years to come, according to the World Health Organization. In 2015, conditions including dementia, epilepsy, multiple sclerosis, Parkinson’s disease and stroke accounted for more than 94 million disability-adjusted life years (DALYS), the number lost globally to ill-health, disability or early death – a total expected to swell to over 103 million by 2030. In the U.S. alone, brain diseases cost nearly $800 billion each year, according to a paper published in the Annals of Neurology in 2017. Bioelectronics Innovation Outpaces Drug Development The trendlines are heightening the urgency to develop new, effective medical treatments, yet traditional drug development alone may not be able to keep pace: The journey to create drugs ready for pick-up at your local pharmacy takes, on average, 10 years from the time they are hatched in the lab. “Unfortunately, pharma is unlikely to help address this problem because drug discovery is becoming slower and more expensive,” George Malliaras, Prince Philip Professor of Technology at the University of Cambridge, noted in his presentation, Electronics on the Brain, at last month’s virtual FLEX 2021 conference. In marked contrast, microelectronics are “becoming cheaper and faster every year.” Dating back to the 1950s with the development of implantable pacemakers to re-establish normal heart rhythms, bioelectronics medicine could help demystify how the brain processes information and lead to more effective treatments for neurological disorders. The field has come a long way since devising cochlear implants to treat hearing impairments in the 1970s, designing spinal cord stimulators to relieve chronic pain in the 1980s and targeting the brain with electrical impulses to help relieve Parkinson’s disease symptoms and neuropsychiatric disorders in the 2000s. ​Deep Brain Stimulation Implants Help Treat Neurological Disorders Deep brain stimulation involves implanting electrodes in the brain through small holes in the skull to send electrical impulses to specific target areas. Used in the U.S. since 1997 to treat Parkinson’s disease, deep brain stimulation can improve motor skills in patients suffering from other conditions too such as dystonia, tremors and epilepsy, enabling them to “function normally, with the flip of a switch,” Malliaras said. Researchers are even testing the technology to treat autoimmune and other disorders not originating in the brain. But the large, rigid electrodes used in the surgery are hostile to the soft, subtle confines of the brain. What’s more, implanting the devices is invasive, with multiple follow-up surgeries typically needed to replace batteries, reposition electrodes or replace deteriorating electrical leads. To overcome these drawbacks, engineers are now designing electronics that can process complex neurological signals to treat brain disorders while conforming to its soft tissue. Malliaras said that means developing electronics capable of interacting with the diverse chemicals the brain uses to bridge the tiny gaps between neurons, called synapses, in order to transmit the neurochemical impulses that give rise to thinking and behavior. Mixed Conductors Form Key Connection Between Electronics and Brain Mixed conductors, materials that can transmit brain signals both ionically and electrically, promise to form this key connection by enabling the development of high-resolution cortical electrodes that monitor neurons without penetrating the brain. They’re also a springboard to the development of flexible pin-sized electronic devices that make neurosurgery much less invasive. That brings new hope for more effective treatments of neurological disorders like epilepsy. Traditionally, the first line of defense against seizures has been antiepileptic drugs, an ineffective treatment since 30% of patients are resistant to the medications, Malliaras said. Another drawback are side effects that include short-term memory loss, fatigue, blurred vision, speech impairments dizziness, nausea and weight loss. Resective surgery – disconnecting the diseased portion of the brain that causes seizures – is often the next option, but is not possible in cases when the procedure would risk damaging circuitry that controls cognition and behavior. Flexible Substrates Fuel Development of Tiny, Expandable Bioelectronics Devices With recent advances, studies on lab rats show that the miniature electrodes designed using flexible substrates made possible by photolithography can conform to the brain’s curvatures and creases to measure the slight electrical signals emitted by individual neurons without penetrating brain tissue and deliver drugs to prevent seizures in animals. Measuring just micrometers in width, these horseshoe-shaped microfluidic devices can pump GABA, a natural neurotransmitter that acts as a brake against neuronal excitability throughout the nervous system, through their minute perforations into the ion exchange membrane of the brain to prevent epileptic seizures. “The data from the research is very exciting, but the path to the clinic is long,” Malliaras said. Still, the findings are a step forward in better understanding the brain and treating its pathologies. Today, microfluidic devices are under development to localize drug delivery in order to bypass the blood-brain barrier and destroy remaining brain cancer cells after a tumor is removed. The devices promise not only to improve cancer treatment since a broad array of cancer drugs can’t cross the protective barrier, but to enable doctors to administer cancer-fighting drugs in smaller doses to help reduce side effects. Implantable electronics today are used to bring relief to sufferers of chronic pain. However, the sizeable paddle-type electrodes involve invasive surgery under general anesthesia and a hospital stay of a few days. An alternative is to implant smaller flexible devices through an outpatient spinal tap with local anesthesia, an approach with its own disadvantages. The devices are less efficient than paddles in delivering electrical stimulation and tend to shift position as the body moves, so are seen as an unreliable solution. That leaves patients to choose between an effective treatment requiring invasive surgery and a less intrusive but less effective alternative. One promising solution combines bioelectronics with soft robotics to enable expandable implants containing microfluidic channels that can be activated mechanically. The device’s malleable paddle electrode can be rolled up inside a needle, inserted with a final tap and then pneumatically unrolled for treatment. While the device so far has been tested only on human cadavers, it could spur the design of a broader category of expandable microfluidics devices that minimize the invasiveness of neurosurgery and get patients back on their feet sooner. The tiny flexible electronics could be available to veterinarians to treat dogs in as soon as next year, Malliaras said, and “hopefully someday in the not-to-distant future they’ll be used to treat human patients.” Michael Hall is a marketing communications manager at SEMI.

At the SEMI Foundation, we’re taking steps to support a big, audacious goal – achieving gender parity in the microelectronics industry. Dating to its roots at Bell Labs, Fairchild Semiconductor, and Intel in the late 1950s and 1960s, the semiconductor industry was pioneered by men at a time when far fewer women were in the workforce. While women have made major workforce gains since those early days, we’re still far from achieving anything close to an equitable representation of women. According to the U.S. Bureau of Labor, only 11.8% of electrical and electronics engineers – and just 8.7% of mechanical engineers – are women. What’s more, research from the American Association of University Women (AAUW), a non-profit that champions equity for women and girls through advocacy, education, and research, tells us that women drop out of engineering careers more steadily and quickly than men. According to AAUW research, just 30% of women working in engineering are still in the field after 20 years compared to 35% of men. By the time women have been in the field for 30-34 years, that number falls to 19% – while it increases to 39% of men among the same cohort. The small number of women in engineering careers and the fewer still who stay in engineering long term illustrate the troubling gender disparities in the industry. Even with these low numbers, however, there are still women who have managed to not just stay in the industry, but to thrive and lead within it. I talked with four of these women about their professional journeys and how they believe women can be best supported in careers in our industry. The AAUW research report Solving the Equation: The Variables for Women’s Success in Engineering and Computing shows that attrition in engineering is higher among women than men. Passion for math and scienceLam Research VP Gowri Kamarthy took her Ph.D. in chemical engineering from UC Berkeley directly to Lam Research, where she’s spent the past 22 years in technical positions. Today she heads the company’s conductor etch product line.Coming from a family of engineers, including her father and siblings, Dr. Kamarthy had a built-in support system that was essential to her success. She never felt intimidated by male peers after spending her formative years pursuing her passion for math and science.“I may have stood out as a minority in the field of engineering, but there was also a silver lining in standing out,” she said. “People notice you.”Kamarthy realizes that engineering careers are generally perceived as being less compatible with family life, for both women and men.“Anyone who wants work-life balance in an engineering career will have to navigate its special challenges, including the need to work long hours to match the rapid pace of innovation,” Kamarthy said.Drawing from her own experience, Kamarthy offers some career advice. “Perseverance and grit are key to success,” she said. “The other ingredient is luck. I was fortunate to have great bosses at Lam who didn’t see gender first and foremost. Instead, they recognized my ability to deliver on projects and encouraged me to perform at my best.” A love for math and science. The confidence to excel in those subjects. A support system to help her through the bumpy times. These were also truths for Sandy Vos, Ph.D., director of R D at NXP Semiconductors.“I was always good at figuring things out,” says Dr. Vos. “I remember feeling enthralled when I got my first internship because it combined engineering, math, science and manufacturing.” Like Kamarthy, Vos was aware of her status as a woman in a male-dominated field, but it didn’t stop her.“If anything, my gender drove me to prove myself,” Vos said. “And I’ve been fortunate because everywhere I’ve worked, I’ve been a part of a smart and collaborative team.”That doesn’t mean gender never came into play. Whenever it did become an issue, Vos didn’t shy away from hard conversations. She recalls having a conflict on the plant floor with two men who each stood over six feet and were about 100 pounds heavier.“I had a conversation with them, and we figured it out,” she said. “But for a while there, my heart was racing.”Gender felt like a bigger issue when Vos was younger. “Now that I have gray hair, it’s not much of a concern,” Vos said. “But earlier in my career, I started putting Ph.D. on my business card so people would know I could talk technical details.”Though just one of three women in an undergraduate class of 35 engineering students – and with a teaching cohort of all-male professors – Debbie Gustafson anticipated equitable treatment in her college engineering program. She had the same outlook when she began her career in semiconductor manufacturing. But the belief that she’d receive the same treatment as her male peers went largely unfulfilled. This didn’t slow her down. During her first year as CEO of Energetiq, she grew the company’s revenues and valuation. A year later, she steered the company through a successful acquisition by Hamamatsu Photonics. Today Gustafson continues to lead Energetiq as a wholly owned subsidiary, but the road to the top job wasn’t without hurdles. Gustafson muscled through the tough times.“When I started out, I traveled to Japan and Korea when there weren’t other women in technical roles,” she said. “My first meetings were extremely frustrating. I was the only woman in the room, and the men wouldn’t address me. This went on for a year, but I kept coming back and built the relationships.”Now a member of the SEMI Foundation Board of Trustees, Gustafson credits mentors with helping her navigate the nuances of doing business across cultures during those early years.A rocket scientist among usAlissa Fitzgerald might tell you that MEMS isn’t rocket science. But that’s only because she has a Ph.D. in Aeronautics and Astronautics, which actually is rocket science. Dr. Fitzgerald worked at a government laboratory and a large defense contractor before she got her Ph.D. and moved to a MEMS industry startup. Though gaining valuable experience, she found the environments too hierarchical and lacking in career development opportunities for young female engineers. As one of the few women engineers at these heavy-duty engineering firms where, in the 1990’s, there were no women in leadership roles, Dr. Fitzgerald sensed that opportunities for her to advance were remote. Fitzgerald started her own firm rather than climb up the ladder of another company, but it turns out, her motivation had nothing to do with gender.“It was the way engineers were treated like Dilbert,” she said. “I felt like a cog in the wheel, working for corporations that weren’t nurturing or appreciative of engineers.”After years of working for other companies, Fitzgerald founded the eponymous AMFitzgerald Associates, a developer of innovative MEMS and sensor solutions for specialty applications. When gender did come up for Fitzgerald, it manifested in men questioning her technical abilities.“Early in my career, I felt like I had to prove myself worthy, even though my degrees were from MIT and Stanford,” she said.Over 3,000 respondents to the Workplace Experiences Survey, sponsored by the Society of Women Engineers and the Center for WorkLife Law at UC Hastings Law, validate Fitzgerald’s experience. 61% of women vs. 35.1% of white men surveyed cited Prove-It-Again Bias – “having to prove themselves repeatedly to get the same levels of respect and recognition as their colleagues.” For engineers of color, that disparity was even worse. 68% of engineers of color (both women and men) reported Prove-It-Again Bias vs. 35% of white men.“For women and people of color, there’s rarely an assumption of competence,” Fitzgerald said.It’s sad but true that we can’t decouple the challenges women face from the challenges people of color face. Both are dramatically underrepresented as chip companies, and women of color represent the smallest percentage of the industry’s workforce and leadership.Inclusivity mattersWorking toward gender equity isn’t just a case of doing what’s right. It’s a case of doing what’s profitable. Research shows that companies with more women on the board perform better.“Given the pace of innovation in semiconductors, we need people from different backgrounds and perspectives to solve the hard problems challenging our industry,” Kamarthy said.Vos appreciates the fact that SEMI is creating a forum of inclusion.“Inclusion starts when you’re young,” she said. “School-aged kids are already making decisions about a future they see as exciting and possible. Our job is to make sure they have the opportunities to pursue what they envision.”Change won’t come magically, though. Fitzgerald believes companies need to make a concerted effort to attract a diverse population.“While I see a disproportionate number of female applicants, I’m more the exception than the rule,” she said. “When male executives call and ask, ‘How are you finding all these amazing female engineers?’ I say, ‘they’re finding me.’”Elevate the storyAchieving gender parity in microelectronics is a daunting task. Fortunately, access to SEMI’s global membership puts us in a unique position to make this deeply complex story clear and relevant to our members, so we can help support the shift.We’re looking at both the stark numbers of women working in microelectronics and at the lack of longevity of women in engineering. We’re elevating the conversation about childhood education. Why are girls passed over in math and science classes in early grade school, and what is the effect of teachers’ lowered expectations for girls taking these classes? What does it mean to be the only in the room? The only woman, or the only woman of color, on a team or in a meeting room. Feelings of isolation or disengagement – or frustration with Prove-It-Again bias – often lead to turnover in an industry that already struggles with retention.Reverse the trendThere’s much SEMI members can do to work toward gender parity in our industry. Look at recruitment, hiring, retention and promotion processes to see how women fare in them. Consider how to create a company culture of self-awareness and inclusion. Ensure equitable pay. Suggest and request women speakers for keynotes and panels at conferences. And offer workplace flexibility to allow women – who often bear most family responsibilities – to take time off or reconfigure schedules so they can help care for children or ailing parents.It’s time for our industry to reverse the trend of gender inequality. Research shows that companies with greater gender and racial parity are more productive, innovative, and profitable. If we welcome and support women in our companies, we will help women – and our industry – reach their full potential.Get involved with SEMIRegister for the Women in Semiconductors (May 3, 2021). This virtual event will include interactive exploration and discussion on strengthening the roles of women in hybrid and remote work environments. Everyone managing teams or experiencing the gender parity challenges and opportunities will benefit from the fresh thinking and best practices that the Women in Semiconductor program is known for.Participate in the SEMI Mentoring Program. By matching mentees with industry leaders and professionals, SEMI Foundation facilitates one-on-one mentoring relationships that benefit all participants. Whether you are a recent university graduate or growing in your microelectronics career and looking for support, participating in the SEMI Mentoring Program will put you on the right track.Participate in the McKinsey Company 2021 Women in the Workplace Study, which looks at representation and the experience of women in companies across the U.S. and offers recommendations on how to retain and support women. Email [email protected]. Shari Liss is executive director of SEMI Foundation. Connect with her on LinkedIn.

Connecting product development to manufacturing to the field in the semiconductor equipment industry by becoming a Model-Based Enterprise Manufacturers across many industries, including the semiconductor equipment industry, are facing pressure to dramatically reduce product development cycle and production ramp times while also enhancing product quality and reliability. This challenge is complicated when multiple configurations are deployed and then maintained, enhanced and upgraded in the field. Buzzwords like digital thread, digital twin, smart manufacturing, Industry 4.0, and digital supply networks point toward a fusion of process, technology, data, and talent that promise the game-changing outcomes needed to address these challenges. Yet, there remains uncertainty about how these various elements come together into a cohesive approach across the product lifecycle. One such approach involves establishing a Model-Based Enterprise (MBE), with alignment across organizational silos, processes, and technologies. Too often, however, an effort to transform to an MBE is frustrated by “random acts of digital” that pursue implementation of certain digital technologies but fall far short of delivering value. What is needed instead is a deeper understanding of the characteristics of a true MBE and the tools and approaches to transform the organization, processes, and data to an MBE and thrive in the marketplace. Elements of a Model-Based Enterprise Modeling in engineering is not new, but what’s emerging are MBEs that comprise digital models connected upstream and downstream over the entire product lifecycle, from a product’s conception and development through its production and end market installation and use. At the heart of such MBEs are digital threads—integrated sets of processes executed within an interconnected technology ecosystem that drive the end-to-end product lifecycle and provide MBE data traceability front to back. A digital thread in the MBE environment includes all of the process, data, and system capabilities that enable digital representations of the product lifecycle stages, or digital twins, of which there are three primary types: The product digital twin is a virtual or simulated representation of the product and each of its components and configurations. While most manufacturers manage engineering models with CAD/CAE solutions, just 15% use product digital twins. Leaders in this space have seized a competitive advantage in product development and accelerated engineering. Yet, becoming a true MBE requires more than product digital twins. The process digital twin is a model of the manufacturing equipment, processes, and the workforce required to carry out related operations. The process digital twin represents the operation of the physical factory floor and its assets, complete with workflows and instructions that describe how the manufacturing processes are performed. A process digital twin relies on data from the product digital twin and allows the enterprise to build according to a product plan and predict what may happen on the factory floor. About 5% of enterprises use process digital twins. After production, a service digital twin represents the installation, use, maintenance, and repair of each product operating in the end market. The service digital twin is informed by product and process digital twins to facilitate adjustments and enhancements based on real-world data. Less than 5% of manufacturers use service digital twins, which is perhaps expected due to their reliance on the existence of the product and process digital twins. When underlying data (e.g., models, specifications, and configurations) are standardized and integrated across a digital thread, an enterprise has the capacity to monitor and refine a product over the span of the thread while also injecting insights and improvements back into the thread. The result is that engineering and manufacturing and end customer usage feedback is continuous and efficient, achieving in weeks what once took months. Visibility into materials, costs, suppliers, and more enable the enterprise to pivot and keep production moving, even when dealing with unforeseen challenges. Real-time monitoring that synthesizes live data also helps reveal performance insights and end market issues (e.g., installation issues, quality issues, etc.) that allow improvements to the offerings and operations of the enterprise. While the manufacturing and product development benefits of an MBE may be clear, the path to becoming an MBE and achieving these benefits can be challenging. Figure 1. Representative MBE end-to-end digital thread that connects product development, manufacturing, and the field (Source: Deloitte Development LLC, 2021) Accelerating Transformation New and emerging manufacturers have an opportunity to build toward their MBE vision without the historic data constraints legacy systems can impose. For more established manufacturers, such is more typical in the semiconductor equipment space, legacy technology and processes can present obstacles to an MBE transformation path. Stakeholders who have invested time and resources implementing certain enterprise platforms (e.g., CAD, PLM, ALM, MES, ERP, etc.) often look for how these can be used to enable a digital thread and digital twins over the enterprise. This limiting view frustrates a broader, more holistic opportunity to transform to an MBE and thrive in the marketplace. There is thereby a dual imperative to define a modernized and scalable end-to-end technology architecture for managing the MBE product data while also establishing a capabilities implementation roadmap that rapidly leads to MBE maturity. To be successful in your MBE transformation, four core functional capabilities that enable a digital thread must be considered: digital engineering, industrial simulation, manufacturing execution, and real-time monitoring. Addressing each with optimized tools allows a manufacturer to rapidly move from strategy to reality. Recognizing the complexity of addressing these functional needs, Deloitte has developed preconfigured solutions to help expedite and enhance transformation across each of these core areas. Design with D-PLM Simulate with D-Sim Execute with D-MES Monitor with D-IoT Accelerates product and application lifecycle management transformations with a multi-phased approach, including a phased, multi-year PLM/ALM roadmap and business case. Facilitates the ability to test and refine processes in a virtual environment, rapidly revealing the most efficient and effective industrial processes more quickly than is possible in a real-world environment. Integrates pre-defined processes related to production planning, execution, tracking and tracing, quality management, data collection, and visualization, with integrations to PLM and ERP. Delivers fast implementation of IoT capabilities that connect, collect, and analyze a broad scope of production data to drive quicker returns for high-impact areas while cultivating digital adoption. While many enterprises have various initiatives in model-based systems and manufacturing, few have tied them together with an end-to-end digital thread and set of data standards over the entire product life cycle. The foregoing preconfigured solutions can help enterprises transform to a true MBE that can typically achieve: 15% – 20% better development efficiency 30% – 50% faster time to market 8% – 20% product cost reduction 10% – 30% cost of quality reduction With such improvement potential, this could be the right time to map out and accelerate your MBE transformation to support your evolving business models and products. Deloitte Consulting LLP Co-Authors Kevin Prendeville Principal, Product Strategy Lifecycle Management [email protected] Vijay Santhanam Managing Director, Product Strategy Lifecycle Management [email protected] Kenneth Norton Senior Manager, Product Strategy Lifecycle Management [email protected] Dan Hamling Specialist Master, Technology Semiconductor [email protected] As used in this document, “Deloitte” means Deloitte Consulting LLP, a subsidiary of Deloitte LLP. Please see www.deloitte.com/us/about for a detailed description of the legal structure of Deloitte USA LLP, Deloitte LLP and their respective subsidiaries. Certain services may not be available to attest clients under the rules and regulations of public accounting. This publication contains general information only and Deloitte is not, by means of this publication, rendering accounting, business, financial, investment, legal, tax, or other professional advice or services. This publication is not a substitute for such professional advice or services, nor should it be used as a basis for any decision or action that may affect your business. Before making any decision or taking any action that may affect your business, you should consult a qualified professional advisor. Deloitte shall not be responsible for any loss sustained by any person who relies on this publication. Copyright © 2021 Deloitte Development LLC. All rights reserved.

MEMS actuators transform electronic signals into something that can be sensed or touched by the end user of an electronics device. A case in point: MEMS actuators such as print heads in inkjet printers transform electronic files into text or beautiful images. In 3D printers, actuators can produce real objects. Inside smart glasses, tiny MEMS mirrors can create virtual objects. Little surprise, then, that integrating these powerful devices into the end products is a multidisciplinary enterprise. STMicroelectronics has been successfully leading the deployment of dedicated MEMS actuator solutions with customer products in various market segments. SEMI spoke with Anton Hofmeister, group vice president and general manager of the MEMS Actuator Division at STMicroelectronics, about MEMS actuator trends. Hofmeister shared his views at the SEMI MEMS Imaging Sensors Forum as part of the virtual SEMI Technology Unites Global Summit. Watch the STMicroelectronics’ presentation on-demand until March 26, 2021. Registration is open. SEMI: What is the difference between MEMS devices that sense and MEMS devices that actuate? Hofmeister: MEMS sensors gather data from the world around us including motion, pressure and air temperature and transform them into an electrical signal. Actuators work the other way round. They receive an electrical signal and transform it into some well-controlled actuation such as ejecting a fluid, moving a membrane or deflecting a laser beam. SEMI: How can MEMS actuators’ integration be simplified to be embedded in new applications so they appeal to consumers? Hofmeister: The challenge of integrating MEMS sensors into devices has been simplified by demo kits and evaluation boards, which customers use to embed the sensor into a system. MEMS actuators are more difficult to integrate. They often power the core function of a system and therefore require deep system understanding. Reference designs are a big step forward in simplifying integration. My presentation at the SEMI MEMS Imaging Sensors Forum showcased some examples. MEMS micro-mirror projection for augmented reality (AR) glasses is an example of a complex system that requires multiple types of components to function. Together with several partners, STMicroelectronics recently announced the LaSAR Alliance, which will develop reference designs to enable the AR glasses market. SEMI: MEMS sensors and actuators are considered the backbone of many consumer products. Are MEMS actuators also mostly used in automotive? Hofmeister: The widest use of MEMS actuators has so far been in print heads for inkjet printers. In recent years, we have seen actuators adopted in emerging applications ranging from piezo heads for 3D printers to MEMS mirrors for laser beam scanning systems or 3D sensing solutions for consumer applications. The first high-volume application in automotive will likely be MEMS mirrors for LIDAR systems. SEMI: What market growth trends do you see for MEMS sensors and actuators? Hofmeister: The sensorization trend, which aims to collect data from homes, cities, factories, cars and personal devices, continues to drive the adoption of sensors and actuators for a wide variety of applications. While the last wave of MEMS growth was triggered by one end product – the smartphone – the next wave will be driven by multiple applications and use cases in industrial, medical, automotive and personal electronics. SEMI: How can technology unite us? Hofmeister: In recent months, we have all experienced vividly how vital technology has become. MEMS, and semiconductors in general, are an integral part of many products and services that make our lives easier. Communications technologies have been particularly important during this pandemic, whether using the personal devices as our interface to the digital world or the complex infrastructure that they operate through. I hope that my participation at the summit helped increase awareness of the new possibilities and opportunities that technologies like MEMS actuators have to offer to create products and services that further improve people’s lives. Anton Hofmeister is group vice president at STMicroelectronics, general manager of the company’s MEMS Actuator Division and managing director of its German subsidiaries. Hofmeister has been with STMicroelectronics for more than 30 years, working in Germany, France, the U.S. and Italy. He has held managerial positions in key account management, product and strategic marketing, advanced R D and general management. For the past 10 years, he has managed various product divisions in the MEMS sector. Hofmeister has also served as a board member of the Singapore-based molecular diagnostics company Veredus Laboratories. Serena Brischetto is senior manager of Marketing and Digital Engagement at SEMI Europe.

The pandemic unleashed by the coronavirus SARS-CoV-2 (which causes the disease COVID-19) has infected over 100 million and resulted in over 2.6 million deaths worldwide as of March 2021. It is well-established that this virus primarily spreads from person-to-person via respiratory droplets produced when an infected person coughs, sneezes or even breathes (see Ref. 1-3). Subsequently, the droplets meet the eyes, or enter nose or mouth of a nearby person, or transmit when a person touches an infected surface, then contacts their eyes, nose, or mouth. Since the virus is small, 0.06–0.14 microns in diameter, many copies can be contained in or attached to emitted respiratory droplets. Droplets as small as one micron can carry enough viral load to cause an infection. A particular concern is the interaction of droplets with ventilation systems, which potentially could enhance the propagation of pathogens. This has implications on situation-specific safe distancing and the design of building filtration systems, air distribution, heating, air-conditioning and decontamination systems. A particular instance of this is the semiconductor manufacturing cleanroom, where systems and protocols are specifically designed to minimize contamination. The $440 billion global semiconductor industry depends on these cleanrooms for integrated circuits (chips), and in turn, these chips form the lifeblood of the multi-trillion-dollar global electronic systems industry. Electronic systems are now critical for just about every aspect of human life, including health, work, finances, entertainment, transportation, power grids and many others. Thus, it is critical to understand how cleanrooms can operate more safely to ensure the health of workers while maintaining productivity levels to meet increasing global demand for semiconductors. In the work described here, we analyzed particle and droplet transport via modeling, simulation [Refs 1-3], and experimentation [Ref. 4] to help guide the industry. Modeling and Simulation In this part of the work, mathematical models were developed to simulate the progressive time-evolution of the distribution of locations of particles produced by a cough. Analytical and numerical studies were undertaken. The models ascertain the range, distribution and settling time of the particles under the influence of gravity and drag from the surrounding air. Beyond qualitative trends that illustrate that large particles travel far and settle quickly – versus small particles that do not travel far and settle slowly (yet can be carried far by ambient flow) – the models provide quantitative results for distances travelled and settling times, which are needed for constructing social distancing policies and workplace protocols. Figure 1 shows examples of the results of the modeling and simulation work. Figure 1: Model of particle spreading from a person coughing, with and without a mask. (Ref. 1) Following are key insights from the modeling and simulation work (Ref. 1): Large particles travel far (launched “ballistically”) and settle quickly, while small particles do not travel far and settle slowly (when there are no ambient externally-driven flow fields). Small particles do not settle even by the end of the simulation time (4 seconds in Ref. 1). Accordingly, the simulations were also run for extremely long periods to ascertain that the “mist” of small particles remained airborne for several minutes (as predicted by the theory). For strong opposing headwind, small particles move backwards, yet still remain airborne for extended periods of time. This is by far the most dangerous case since this will encounter other persons at the torso level. Ratio of the general drag to gravity indicates that at high velocities, the dynamics are dominated by drag. For general cough conditions, there can be cases where the change in the surrounding fluid’s behavior, due to the motion of the particles and cough, may be important. One major implication of this work is that the challenge of infection must be addressed both spatially and temporally. In other words, it is necessary to maintain social distancing based on how far the virus travels, but it is also important to account for how long the virus stays at the location because of specific air patterns. On the positive side, understanding these spatio-temporal patterns accurately will enable companies to design (or re-design) ventilation and decontamination systems precisely to improve worker safety. Other aspects of this analysis entail contact tracing (Ref. 2) and decontamination (Ref. 3). Further details, including simulations, are available at https://msol.berkeley.edu/publications/. Experimentation The major vector of coronavirus spread is through respiratory droplets expelled when coughing, speaking, and breathing; and the efficacy of any safety measures depends on accurate characterization of the dispersal of these droplets. The term particle describes objects that begin their journey as a solid. The term droplet is reserved specifically for objects that are initially liquid, albeit it is important to note that droplets can evaporate and effectively transform into solid particles composed of non-evaporative material. A purpose-built room, the Cal Covid Cube, C3, was set up and utilized for this research [Thatcher et al. 4]. The C3 is a parallelepiped room that is 232 centimeters tall, 376 centimeters long and 284 centimeters wide on the inside. For experimental results to be meaningful and repeatable for scientific and practical purposes, it is essential that the experimental setup be carefully controlled and calibrated. The following precautions were taken to ensure this: Charge-free: When solid particles are released, it is critical to eliminate (or thoroughly know) static charge effects for obtaining accurate deposition patterns. Static charge effects can manifest through particle-particle interactions (affecting particle motion in flight) or particle-surface interaction (affecting deposition pattern). Two methods for the elimination of charge effects on the deposition surface were found to be effective: (1) ionized non-conductive adhesive sampling strips, and (2) grounded aluminum backed carbon sampling strips. Isothermal: The room is a converted walk-in freezer with 10.5-13 centimeter thermal insulation and located in the middle of a building, at least 5 meters away from all building walls. Temperature uniformity was checked and the C3 room temperatures were found to be isothermal within uncertainty of measurements. Quiescent: It was ensured that the room did not create uncontrolled thermal convection due to isothermal nature. Quiescence was verified with both hot-wire measurements and with free-falling particle drift observations. Isopotential: The outer and inner surfaces, including the door of the C3 were conductive aluminum and stainless steel, and copper tape were used to ensure reliable electrical connection of door, interior and exterior panels. Electric fields were surveyed and found to be negligible within precision of instruments. Other design elements: All interior surfaces were coated with black matte paint to reduce scattered light and provide uniform background for imaging measurements. The facility was located on ground floor to limit vibrations. Repeatable Launch: To emulate the release from a true cough or sneeze, and to better understand droplet motion in a canonical turbulent jet versus a cough type release, we studied different layers of complexity for the release geometry: (i) Straight round pipe (ii) Smooth 90-degree curved pipe, with a changing radius along the length of the pipe (iii) Intubation trainer doll, with realistic airways and mouth/tongue structure Figure 2 shows the experimental setup with the intubation doll in C3, with the particle/droplet release being measured after deposition on the sampling strips that appear green. Figure 2: Experimental Setup in C3 with both charge neutralized (white appearing green) and conductive (black) sampling strips placed on a conductive and grounded alignment grid [Ref. 4] We utilized both liquid droplets and solid particles. For droplets, we explored and found promise in a method of deposition analysis based on fluorescence. For particles, we explored many ways in which the smallest of thermal gradients or electrostatic charge issues can affect the data and developed practical methods to address these issues. For accurate measurement free of static charge effects even in environments where high ambient flow velocities may cause a nonconductive surface to rapidly acquire charge (e.g., clean room environment), we developed carbon-tape-based sampling strips that are cleanroom-compatible, conductive, and grounded. For analysis, we developed a cost-effective method utilizing a commercial photo negative scanner followed by image enhancement by blind deconvolution. Figure 3 shows sample results for particle deposition location along our centerline for particles in the ballistic, intermediate and aerosol regime. Figure 3: Experimental Results [Thatcher et al. 4] Following are key insights from experimental work: Significance of both static charge effects and thermal gradients in rooms for validation tests are more than usually appreciated. For modelling, accounting for the non-uniform initial particle velocity matters for the ballistic particles. For all sizes of particles, simulating the transient versus steady state significantly impacts predicted particle spread. Thermal plumes alone from humans along particle flight path can transport 50 micron particles across the room. In some situations, this was observed up to ~6 meters. There is a significant effect of Relative Humidity (RH) and temperature on droplet evaporation. The practical consequence is that, in low RH, particles spread further, with all other things being equal. (The reason is that particles shrunk more and entered the aerosol regime.) In summary, a systematic analysis of particle and droplet transport was conducted by simulation, modeling, and experimentation. We were able to develop robust, rigorous, and repeatable methodologies and draw meaningful insights that will support safer operation and productivity of semiconductor cleanrooms globally. Further, these studies will help with the design (or re-design) of ventilation and de-contamination systems that help protect both the health of humans and the economy from current and future pandemics. This article provides a high-level overview of the work, and further details will be available through a series of scientific papers that are in various phases of publication. We gratefully acknowledge the following support: Gift of the Lam Research Corporation Gifts coordinated through SEMI and provided by Advanced Energy Industries, Applied Materials, ASM, Entegris, JSR, KLA, TEL, and Wonik The 2020 Seed Fund Award from the Center for Information Technology Research in the Interest of Society (CITRIS) and the Banatao Institute at the University of California Vision Research for providing a v2640 camera to help quantify the particle velocities Graduate students Eric Thacher and Tvetene Carlson who conducted the experiments in C3 Valuable discussions with Brett Singer, Thomas Kirchstetter, Michael Sohn, Chelsea Preble of Lawrence Berkeley National Laboratory regarding droplet transport and COVID, and Keith Hansen on particle sampling and charge neutralization DOE Office of Science through the National Virtual Biotechnology Laboratory, a consortium of DOE national laboratories focused on response to COVID-19, with funding provided by the Coronavirus CARES Act Steven Ruzin and the Biological Imaging Facility for their assistance in obtaining the high-quality fluorescence microscopy scans to validate the particle counting methodology. References Zohdi, T.I. (2020) Modeling and simulation of the infection zone from a cough, Computational Mechanics. https://doi.org/10.1007/s00466-020-01875-5 Zohdi, T.I. (2020). An agent-based computational framework for simulation of global pandemic and social response on planet X, Computational Mechanics. https://doi.org/10.1007/s00466-020-01886-2 Zohdi, T.I. (2020) Rapid simulation of viral decontamination efficacy with UV irradiation. Computer Methods Appl. Mech. Eng. https://doi.org/10.1016/j.cma.2020.113216 Thatcher, E., Carlson, J., Castellini, J., Sohn, M.D., Variano, E. and Makiharju S.A. (2021) Droplet and Particle Methods to Investigate Turbulent Particle Laden Jets (in preparation) Authors Evan A. Variano, Professor, Environmental Engineering, UC Berkeley Simo Mäkiharju, Assistant Professor of Mechanical Engineering, UC Berkeley Tarek I. Zohdi, Will C. Hall Endowed Chair of the UCB Computational Data Science Engineering Program, Professor of Mechanical Engineering, UC Berkeley Pushkar P. Apte, Director of Strategic Initiatives, Center for Information Technology Research in the Interest of Society (CITRIS) and the Banatao Institute, UC Berkeley; and Strategic Technology Advisor, SEMI

Recent semiconductor supply chain constraints have drawn the attention of Washington policymakers at every level. Exasperated by the global pandemic, customers of semiconductor manufacturers have sounded the alarm about the chip shortage and the downstream consequences for end-user companies and consumers. Global automakers have suffered the brunt of the impact, shuttering factories and slashing vehicle production. Last month President Biden issued an Executive Order (EO) to review and secure America’s supply chains. The stated goals of this review are to revitalize and rebuild domestic manufacturing capacity, maintain America’s competitive edge in research and development, and create well-paying jobs. Under the EO, the U.S. will also work more closely with allies to strengthen supply chains. The EO directs supply chain reviews on several critical segments, including semiconductor manufacturing and advanced packaging. The Department of Commerce will identify risks throughout the U.S. semiconductor supply chain and make policy recommendations to address those risks within 100 days of the EO’s issuance. In coordination with the White House, Congress is contemplating a variety of measures to address supply chain issues. Recently, the Senate Finance Committee held a hearing on the effects of the U.S. tax code on domestic manufacturing. Both Chairman Ron Wyden (D-OR) and Ranking Member Mike Crapo (R-ID) highlighted their desire for bipartisan cooperation to use the economic tools within the jurisdiction of the committee to bolster domestic manufacturing. The committee discussed two pieces of legislation that would provide significant incentives to domestic manufacturing of semiconductors. The first was the investment tax credit (ITC) for semiconductor manufacturing that was included in last year’s CHIPS for America Act but not with the other semiconductor incentives in the FY2021 National Defense Authorization Act (NDAA). An ITC would provide predictability and stability in the U.S. tax code to promote large, long-term investments for the industry. The second was the American Innovation and Jobs Act, which repeals the R D amortization requirement set to go into effect in 2022 and expands the refundable tax credit for startups and small businesses. Enhancing domestic incentives for R D and manufacturing is an important step in putting the U.S. on equal footing with other countries and would promote its continued leadership in the chip industry. Senate Majority Leader Chuck Schumer (D-NY) has announced his intention to craft a package of measures to strengthen U.S. competitiveness vis-a-vis China. The package reportedly would include funding for the microelectronics R D and Commerce grant programs that were passed in the NDAA. The Senate plans to take up the legislation in April. SEMI applauds the renewed focus on incentivizing domestic manufacturing and R D for an industry that enables countless technologies, drives innovation in sectors throughout the U.S. economy, and powers the electronic systems essential to critical infrastructure and defense systems. We look forward to working with policymakers in Congress and the Administration to support the entire domestic semiconductor ecosystem. Kimberly Ekmark is director of Public Policy and Advocacy at SEMI

The global semiconductor industry is on track to register a rare three consecutive years of record highs in fab equipment spending with a 16% increase in 2020 followed by forecast gains of 15.5% this year and 12% in 2022, SEMI highlighted in its quarterly World Fab Forecast report.

For the first time in its 20-year history, the FLEX Conference dedicated an entire session to the important and timely twin topics of environmental sustainability and power consumption of electronic devices. The event planning committee recognized the urgent need to increase the awareness of how technology and electronics devices can help reduce greenhouse gas emissions (GGE) overall and meet aggressive targets to curb the impacts of climate change. Dr. Christine Ho, CEO of Imprint Energy, delivered the keynote for the session, focusing on the need for powering billions of sensors that will be deployed annually, and their role in reducing fossil fuel emissions through becoming aware of issues, monitoring our resources over time, and intervening early and often to combat waste in multiple sectors and industry. Quoting extensively from the organization Exponential Roadmap Initiative (ERI), Ho noted that “the digital sector has the potential to directly reduce fossil fuel emissions 15% by 2030 and indirectly support a further reduction of 35% by influencing consumer and business decisions and systems transformation.” The initiative’s playbook for reaching net zero carbon emissions by 2050 and limiting global warming to 1.5° Celsius outlines how the digital sector can help remove 13 of the 27 gigatons (GT) of CO2 needed to reach this goal. Ho stated that the rapidly emerging Internet of Things (IoT), devices, software systems, and data insights are the backbone of this digital transformation. The IoT's vast network of sensors can transform multiple sectors, such as the logistics industry, which on an annual basis moves and ships more than 10 billion tons of products worldwide by ships, airplanes, long haul trucks, and train - contributing 17% of GGE and more than 4 gigatons of CO2 annually. Always-connected IoT sensors used by the logistics industry can reduce waste and damage in the supply chain, which is especially problematic for temperature-sensitive and damage prone pharmaceutical and food products, mitigating the need for producing high volumes of buffer inventory to replace damaged goods Noting that the attendees of 20 Years of FLEX Conferences were a big part of the current advancements of low-cost printed, active, shipping tags, Ho said that Imprint Energy’s flexible and thin, Zinc based batteries are ideal for IoT devices, since they boast a significantly smaller carbon footprint than Lithium-Ion (Li-ion) batteries. Imprint Energy is working with systems designers and integrators to design the battery as an integral part of the device package and use low-power strategies to extend device lifetimes. Imprint recommends co-locating battery printing alongside the device integration to further minimize shipping and logistics. When manufactured separately, Imprint’s small footprint, low-operating temperature process line (less than 80°C) provides significant carbon footprint advantages over other technologies. Ho challenged the attendees, saying “we all need to participate in protecting our earth. We need to eliminate waste and contribute to reducing half of our current greenhouse gas emissions by 2030, and we can do that by deploying a global digital skin with more than 100 billion IoT devices in 2030 and up to 1 trillion by 2050. We can minimize the device carbon footprint and maximize its longevity by considering the power capability, as well as design for re-use and re-cycling of the critical materials.” Following Dr. Ho’s presentation, FLEX kicked off a spirited panel discussion with experts from PowerRox, ITN Energy Systems, Birla Carbon, and Auburn University and chaired by Bob Praino and Eric Forsythe, from Chasm Advanced Materials and the Army Research Labs, respectively. The speakers summarized their on-demand presentations and looked at what is being done today to recycle Lithium-Ion batteries, how IoT devices are currently being powered, and drew comparisons between the early days of the Internet and development of the IoT. The speakers generally agreed that the power requirements of wireless cellular and Blue-tooth devices were still too high and run times too short. FLEX 2021 was a virtual event in the 2021 SEMI Technology Series. It was organized by SEMI FlexTech, SEMI NBMC, and NextFlex. Major sponsors included E Ink and Novacentrix. The event covered technical developments in flexible, printed and hybrid electronics, featuring more than 100 presentations and networking opportunities. Technical proceedings are available until March 26 at http://flex.semi.org. Heidi Hoffman is senior director in Corporate Marketing at SEMI.

With each transition to a new technology node, fab requirements for metal and particle contamination become more stringent, posing challenges for existing coating methods such as anodization or plasma spray that may not provide complete protection against contamination especially on critical chamber components with complex geometry. SEMI spoke with Beneq business executive Sami Sneck about common metal and particle contamination issues with critical chamber components, coating methods to protect against corrosion, and properties to look for when selecting the optimal protective coating solution. Sneck discussed the unique benefits of atomic layer deposition (ALD)anti-corrosion coatings with Aluminiumoxide (Al2O3) and Yttrium Oxide (Y2O3) and offered recommendations on how to work with original equipment manufacturer (OEM) partners to design, test and implement an ALD coating solution for semiconductor equipment. To learn more, visit Beneq at its digital booth at SEMI Technology Unites Global Summit, available on-demand until March 26, 2021. Registration is open. SEMI: How does ALD compare with other coating methods such as anodization and plasma spray? Sneck: ALD enables conformal dense and pinhole-free coatings on complex shapes. We can deposit various ALD coating materials on parts made of various materials. All other coating techniques have limitations. For instance, anodization is conformal, but porous and is suitable for Al2O3 used for aluminum parts. Plasma Spray is a line-of-sight method and not conformal on complex shapes, such as holes in showerhead parts. SEMI: Which substrate materials work for ALD coatings? Sneck: In general, parts made of common metal materials, such as aluminum, stainless steel or titanium, all work well with ALD coatings. Commonly used ceramic materials work well with ALD too. Plastic materials need to be coated generally at a lower temperature, which limits the coating material selection, but materials such as Al2O3 can be applied as well. SEMI: What is the maximum coating thickness you can reach with ALD? Does this depend on the material? Sneck: Yes, indeed. The maximum coating thickness does depend on the material of the part that we are coating. Polymer materials for example, have a very large coefficient of thermal expansion, which limits the practical coating thickness to the 100-nanometer level. On metal and ceramic parts, coatings of several micrometers are possible too. Typically, ALD coating thickness on chamber components range from a few hundred nanometers to one micrometer. SEMI: Which aspect ratio can you coat with ALD? Sneck: Basically, ALD can coat aspect ratios of 1000:1, but this would be extremely slow. In practice, some of the most complex parts are showerhead parts with small holes. Typically, these have an aspect ratio of around 100:1, which is perfectly commercially feasible for ALD. An extreme example would be gas lines: In this case, the aspect ratio may be also around 100:1, but the physical distance from one end to the middle may be half a meter. In this respect, it is not practical to wait for gas diffusion to reach such a depth level. Instead, the gas lines can be coated by forcing the ALD precursor gas flow into the gas line parts. This works well but needs part-specific manifolds to guide the gases. SEMI: What is the lifetime of ALD coating compared to other coatings? Sneck: ALD coatings differ from other coatings a couple of ways. First of all, ALD coatings generate less particle contamination since they are non-porous. Secondly, and most importantly, ALD coatings can cover areas that other coatings cannot. What is considered the lifetime of a certain part depends on various factors. Ultimately, the lifetime needs to be confirmed by testing parts in actual process chambers by running a lot of wafers through the chamber and monitoring critical parameters such as particle level and yield. SEMI: If you have multiple shelves with parts in the reaction chamber, how does the shelf position affect the coating uniformity? Is center shelf better than top and bottom shelf? Sneck: Uniformity depends on many parameters, including the part geometry, part holder geometry, batch size and coating material. When the shelves supporting the parts are optimally designed and the gas flow is well-distributed to all shelves, all shelves from top to bottom show similar uniformity. SEMI: Is there any risk of cross-contamination? Sneck: Cross-contamination could potentially be caused by the parts themselves or by different coating materials. The batch setup is fixed in production use, which means the parts are the same in every batch. The only variation is that the batch may not be full in some cases, but then we do not fill the empty part of the batch with other parts that could cause contamination in order to prevent contamination from one part type to another. Cross-contamination from one coating material to another is not a usual concern but can be prevented by using dedicated reaction chambers for different coating materials. This is very easy to do with Beneq P800. Sami Sneck manages Beneq’s semiconductor part coating business. He joined Beneq in 2005 and since then has held various professional and management positions including product manager, application manager, director of ALD group, head of sales, and head of Asia. He earned his MSc degree in Chemical Engineering in 2001 from Helsinki University of Technology. Sneck has special expertise in Atomic Layer Deposition technology and business development. He has played a vital role in introducing various ALD production concepts and solutions to several industries ranging from jewelry to photovoltaics, electronics and semiconductors. Access the free webinar recording and discover the latest anti-corrosion coating solutions and the unique benefits of ALD (atomic layer deposition). This webinar is particularly helpful for process engineers, equipment engineers and others responsible for contamination control and equipment yield. Serena Brischetto is senior manager of Marketing and Digital Engagement at SEMI Europe.