Raw Materials of Solid-State Drives
(NAND Flash Memory)
Solid-state drives, and specifically the NAND flash memory at their core, are some of the most important components to modern technology. Although solid-state drives have only recently come into the mainstream tech vocabulary, chances are that almost all of the devices on store shelves today use solid-state drives to store their data, and they are only becoming more popular.
So what exactly are solid-state drives, and how will they affect the lives of an average user? This paper will explain solid-state drives, henceforth SSDs, from the perspective of the materials needed to construct them, with a focus on the flash memory at their cores. Although this technology is just coming into the mainstream, it has grown so fast that it is already squirming at the physical limits of the current materials. We will get an intimate view into the manufacturing materials and methods in modern SSDs, and glimpse the challenges and innovations on the horizon.
Before we dive into this world of acronyms and complex computer concepts, let’s lay a foundation for solid-state technology. For several decades now, the accepted standard for computer file storage has been the hard drive: a spinning disk that holds data as magnetic charges. It is inexpensive and stores a lot of data, but recently it has run into a rather hard physical limit: speed. A hard drive can only ever spin so fast, and so it can only read data sequentially. This is where SSDs come in, and where their title of solid-state really works to their advantage. SSDs are made up of trillions of tiny NAND transistors that store data as electrical charges, so there are no parts that need to spin up to read data. Additionally, they can read randomly from any point instantaneously unlike a hard drive. These forces collectively multiply the speed of an SSD significantly over that of a traditional hard drive, but that isn’t the whole story. SSDs suffer their own drawbacks, all tied in directly to the materials used within them. They are significantly more expensive and last significantly shorter than a standard hard drive, although neither of these drawbacks may be true for very long as manufacturers continue to innovate and push the envelope forward for faster and stronger memory in the future.
Solid-state drives begin their life as every other computer chip does—as a wafer of silicon. This silicon is derived from quartzite, sand that has been purified and fused by nature. It is mined from underground at purposefully private locations, and high-purity concentrations can fetch up to fifty-thousand dollars per ton (Nelson). Once extracted, the quartz is melted and then crystallized by a seed crystal and drawn out in a long cylinder. This cylinder is then sliced by a diamond saw to create silicon wafers. At this point, the wafers are distributed to more specialized plants for various computer components. SSD memory is created in dedicated factories that are referred to within the industry as fabs. These fabs are clean rooms that are 100 times cleaner than a hospital operating room to ensure that the silicon is not contaminated, as it is highly susceptible to contamination. (Micron).
Once the silicon disks reach the fab, they are put through numerous processes of chemical baths and platings. Just to prepare for the actual creation of the chips, the silicon wafers go through several cleaning baths. The first bath is sulfuric acid and hydrogen peroxide to clean the wafers. The second is deionized water, ammonium hydroxide and hydrogen peroxide to remove particles from the wafers. Hydrofluoric acid and deionized water is then used to remove the oxide layer, a specifically vital step since the oxide will later be generated as part of the circuitry. The final contamination removal bath is in deionized water, hydrochloric acid and hydrogen peroxide (Virginia Semiconductor). All of these chemicals are generally sourced either as byproducts of other chemical reactions or are synthesized. Hydrochloric acid is manufactured with hydrogen and chlorine, hydrofluoric acid is made from naturally occuring calcium fluoride and sulfuric acid which can be made from sulfur or pyrite.
Once the cleaning is through, the silicon wafer is ready for the actual circuitry. Unlike what one might picture as a circuit, NAND flash memory is not a bunch of components soldered together on a wafer. Because one square inch of circuitry can house hundreds of trillions of individual transistors, attaching each as an individual component would not be fast or affordable to manufacture. Already SSD manufacturers are reaching the physical limits of how many transistors silicon can hold, as they continue to shrink their manufacturing processes closer and closer to the atomic level. This has forced the development of stacked chip technologies, allowing for much more dense data storage. Whether a stacked or single-layer chip, the creation of the circuit board is the same. The process that is used is called etching, wherein the circuit design, along with all of the transistors, is etched or drawn onto the circuit board. A photoresist chemical is spread onto the wafer, and then a light is shined through a template to cure certain parts of the photoresist to form the circuit pattern. Then, an etching material is laid over the wafer, which will eat through in the places where the photoresist is cured by the light. The chemical used for etching is often potassium hydroxide (Virginia Semiconductor). This potassium hydroxide, commonly called caustic potash, is made through electrolysis of potassium chloride, which can be mined from the earth. This chemical causes the silicon chip to form an oxide layer of silicon dioxide. This silicon dioxide is what is called the dielectric, a non-conductive layer that helps isolate layers of the circuit. The chip is then coated with more layers of etched conductives and dielectric silicon dioxide. The conductive layers can be several different metals, but are commonly copper or aluminum. At this point, the chips are complete and are cut off of the wafer.
The final element of an SSD is the printed circuit board, or PCB. This is the macro side of the SSD’s construction, and the board it is placed on is made up of a glass-reinforced epoxy polymer plate printed with copper wiring. The NAND flash chips are placed onto this board, and soldered in place with tin solder and gold wiring. The chips are covered in a plastic casing to protect them, and the whole assembly is placed into an aluminum case. This case is placed in a special antistatic bag that are made of plastic polyethylene terephthalate. This bag is placed in a cardboard box, and the SSD is ready to be shipped to a computer manufacturer or a consumer.
In the final product, silicon is hardly the dominant material. Instead, it is just an impossibly thin layer of a tiny part on the plastic PCB. Yet, the majority of the materials used in construction, as well as the majority of the manufacturing power, is concentrated on the silicon chips. Combined with their oxide and metal layers, these chips are what hold the trillions of transistors, in turn storing many hundreds of gigabytes of your computer files on them.
In the research of this paper, there is one element that came up consistently short—specific materials used within the silicon chip to create the NAND flash. This is for a fairly simple reason, and that is that within the flash memory manufacturing world, there are only a handful of active companies: Intel, Micron, SK Hynix, Samsung, SanDisk and Toshiba. Despite this, every phone and every computer uses flash memory, so this space is extremely competitive. This near-monopoly means that these companies can easily obscure the exact processes they use in order to retain their competitive edge. With only six companies in the world using any of these processes, it is fairly easy for them to control the data that gets to the public. They prefer to talk in vague acronyms than specific materials lists, but between all of them they do paint a patchwork picture of the manufacturing of an SSD.
With these solid-state drives rapidly becoming the new standard, there has been a huge boom in these markets. I bought my first SSD as a novelty while building a computer in high school, and it was very pricey and very small in storage size. In just a few years they have lowered their costs by hundreds of dollars and increased capacities tenfold. With 3D and stacked flash memory already pushing the boundaries of the storage capacity of a chip, it is certain that the future of and beyond SSDs with only continue to grow in power and potency.
Adams, Shawn. “3D TLC NAND the Next Dimension of Storage...” Micron, Micron
Technology, 16 Feb. 2016,
Dhavse. (n.d.). Fabrication and investigation of low-voltage programmable flash memory gate
stack. Lecture Notes in Electrical Engineering, 453, 35-49.
Kim, Areum, et al. “Development and Applications of 3-Dimensional Integration
Nanotechnologies.” Journal of Nanoscience and Nanotechnology, vol. 14, no. 2, Feb.
2014, pp. 2001–2011., doi:https://doi.org/10.1166/jnn.2014.8758.
Li, Yan, and Khandker N. Quader. “NAND Flash Memory: Challenges and
Opportunities.”Computer, vol. 46, no. 8, 2013, pp. 23–29., doi:10.1109/mc.2013.190.
Lu, Chih-Yuan. “ Future Prospects of NAND Flash Memory Technology—The Evolution from
Floating Gate to Charge Trapping to 3D Stacking.” Journal of Nanoscience and
Nanotechnology, vol. 12, no. 10, Oct. 2012, pp. 7604–7618.,
“Making Memory Chips - Process Steps.” YouTube, Micron Technology, 28 July 2017,
National Center for Biotechnology Information. PubChem Compound Database, U.S. National
Library of Medicine,
Nelson, Sue. “Silicon Valley's Secret Recipe.” BBC News, BBC, 2 Aug. 2009,
Wet-Chemical Etching and Cleaning of Silicon, Virginia Semiconductor, Inc, 2003.
Yi-Hsuan Hsiao, Hang-Ting Lue, Wei-Chen Chen, Bing-Yue Tsui, Kuang-Yeu Hsieh, &
Chih-Yuan Lu. (2015). Ultra-High Bit Density 3D NAND Flash-Featuring-Assisted Gate
Operation. Electron Device Letters, IEEE,36(10), 1015-1017.
Professor Christina Cogdell
13, February 2018
Waste Emissions in the Production of the Solid-State Drive
Ever since the dawn of mankind, humanity has been progressing with the ideal of efficiency in their minds. Inanimate prime movers decreased in size but increased in their capacity and ability to produce work and energy. One such example is the evolution of data storage- more specifically, the solid-state drive. From humble beginnings in which humanity could store only megabytes using large, bulky devices the size of rooms, the solid-state drive is the latest, fastest and most efficient way of storing data inside of electronic devices. There are three parts of an SSD’s process of production: the creation of the wafer, the designing of circuitry, and the assembly. (geek.com) Even though almost all companies have strived to go greener in their production, each step of the process still emits waste, and that is what this paper will explain in more detail.
The wafer, also known as a slice, is a thin slice of semiconductor material that the circuitry for the SSDs are engraved on. (Geek.com) The wafer serves as a substrate for the fabrication of microelectronic circuitry, and undergoes many micro-transformations as needed by the final product such as ion implantation or oxidation. (Final Report: Recycling of Silicon-Wafers Production Wastes to SiAlON Based Ceramics with Improved Mechanical Properties) Being made of pure silicon, these wafers have to be manufactured in specifically designed cleanrooms so that the base material isn’t contaminated. (Geek.com) To create the wafers, semiconductor companies first export the silicon from an outside source, and then clean it. (Semiconductor Industry) The uncut silicon is then trimmed, with the leftovers being discarded as waste. (Final Report: Recycling of Silicon-Wafers Production Wastes to SiAlON Based Ceramics with Improved Mechanical Properties) After trimming, the silicon is then introduced into sealed quartz vessels and then heated and pressurized. (Final Report: Recycling of Silicon-Wafers Production Wastes to SiAlON Based Ceramics with Improved Mechanical Properties) More silicon is then gradually added to the vessels, and this process results in the formation of the polycrystalline silicon that is used in semiconductor manufacturing. (Final Report: Recycling of Silicon-Wafers Production Wastes to SiAlON Based Ceramics with Improved Mechanical Properties) These ingots are then cut into circular disks, resulting in the wafers, and this entire process of production is where the first large source of waste is introduced. (Final Report: Recycling of Silicon-Wafers Production Wastes to SiAlON Based Ceramics with Improved Mechanical Properties) Since the Solid-State Drive technology is on the rise, the output of pure semiconductor silicon for integrated circuitry and memory technology applications are increasing year by year. According to the United States Environmental Protection Agency (EPA), about 60% of the silicon, after being processed, is scrapped along with waste water. (Final Report: Recycling of Silicon-Wafers Production Wastes to SiAlON Based Ceramics with Improved Mechanical Properties) These wastes are transformed into a silicon sludge, which is very costly to recycle and poses a pollution based environmental risk if dried. (Final Report: Recycling of Silicon-Wafers Production Wastes to SiAlON Based Ceramics with Improved Mechanical Properties) The sludge contains materials like Al2O3, SiC, Si3N4, ZrO2, coagulants, polymers, grinding oils, and waste water, resulting in a costly mess. (Final Report: Recycling of Silicon-Wafers Production Wastes to SiAlON Based Ceramics with Improved Mechanical Properties)
The circuitry is first designed by a team of specialized engineers, and then measured in nanometers. (geek.com) Each path, or line, is 5000 times narrower then the width of a human hair. (geek.com) This is where the flash storage technology is developed. The team of specialized engineers design either single cell, double cell, or even triple cell circuitry depending on the customer’s order. The flash memory designed here is faster and more efficient than a normal, commercialized hard drive. With the latest technology being a triple level NAND flash memory, consumers can store up to trillions of bits of data in a nonvolatile environment that can even retain the data without a power source. (geek.com) All of this technology comes from the intricate circuitry designed by these specialized teams of engineers, but unfortunately, another large source of waste is also emitted at this step. Semiconductor manufacturers use a wide variety of GWP gases to create the circuitry on their wafers, with GWP fluorinated compounds such as perfluorocarbons (e.g., CF4, C2F6 and C3F8), hydrofluorocarbons (CHF3, CH3F and CH2F2), nitrogen trifluoride (NF3) and sulfur hexafluoride (SF6). (Semiconductor Industry) This manufacturing process also uses fluorinated heat conducting liquids as well as nitrous oxide. (Semiconductor Industry) Traces of these gases have been found to have been leaked into the atmosphere, adding to the pollution and threat of global warming that looms above the world. (Semiconductor Industry) According to the EPA, anywhere from 10% to as much as 80% of these fluorinated gases pass undetected through the manufacturing chambers and are released into the air. Furthermore, these gases also add to the silicon sludge that is produced earlier in this particular process of production.
The assembly process is perhaps the most complex of all, and a plethora of steps is needed to produce the final product of a SSD. After the circuitry is designed and finished, it is cut into the silicon wafers and bonded together by gold wiring. (Geek.com) After inspection, each individual flash memory chip is encased in a protective shell of plastic, and then tested for errors. (Geek.com) They are then laser etched with their information, and with that the process of production for the components is finished. (Geek.com) With the components done, actual SSD assembly begins on a Printed Circuit Board (PCB), and the flash memory is mounted along with other wiring. (Geek.com) These SSDs are then inspected by both the human eye and the computer before being put into a shell of protective housing, which is what the consumer sees when buying. (Geek.com) The SSD is then labeled, and the firmware is installed to allow for memory storage. (Geek.com) They are then tested for up to 60 hours each to insure quality stability and performance with a variety of computers, each measuring differently on the power scale. After testing, the SSDs are then packaged and shipped to consumer demand. (Geek.com) With so many different parts and unique processes that needs to come together to create a functional, standardized SSD, the entire assembly process is the final large source of waste, larger than the other two individually.
The first large part of the waste comes from the PCBs that the chips are installed upon. PCBs are electroconductive boards that connect electronic components with one another and have their own unique lifecycles. They are generally made of sheets of copper laminated onto, or between sheet layers of non- conductive materials, and the maker generally solders, or electroplates the information needed onto the board. (Guides to Pollution Prevention) Like every lifecycle, the PCB comes with its own fair share of wastes and emissions. When the boards are first cut out and cleaned, they produce airborne particles, acid fume vapors, acid alkali solutions, spent halogenated solvents, and waste water. (Guides to Pollution Prevention) When they are electroplated, they produce electroless copper baths, spent catalyst solutions, acid solutions and waste water. (Guides to Pollution Prevention) When they are finally printed, they produce spent developing solution, spent resist removal solutions, spent acid solutions, and waste water. (Guides to Pollution Prevention) Finally, when they are etched with their specific information and firmware, they produce spent etchant and even more waste water. (Guides to Pollution Prevention)
The second part of the assembly process waste comes from the protective shell that the SSDs are incased in when they are finished putting the entire product together. This is the part that the consumer sees, and is usually slick and thin, with a futuristic kind of look to it. Since SSDs are so standardized these days, the protective shell is usually 2.5 inches in width, and made entirely of plastic and aluminum. (Geeks.com) Needless to say, manufacturing these shells also produce wastes and emissions. The plastic is first harvested from resins of derivatives of petroleum and natural gas, which results in the emissions of non- energy GHG gases through the material’s extraction and refinement. (Plastics) The plastic is then put through a process called “cracking”, which means that the hydrocarbons from the refined petroleum and natural gases are heated to extremely high temperatures to break down its larger molecules. (Plastics) It is then processed by connecting these hydrocarbons molecules into chains called polymers, which are then combined to make different variations of plastic as tailored to the consumer’s demand. (Plastics) These two processes also produce GHG emissions such as Carbon Dioxide and Nitrous Oxide. (Plastics)
The final large part comes from an unlikely source- the packaging and shipping of the SSDs. Since consumerism and the high demand of SDDs due to a technological boom is on the rise, more and more companies are shipping consumers their desired products from all around the world, packing them neatly in cardboard boxes and stuffing them with plastic foam. While most of these materials are recyclable, they still produce wastes and emissions that manages to translate to a landfill. According to the EPA, the wastes from packaging and shipping accounts for nearly 30% of the total waste generated in the United States in 2012. (Advancing Sustainable Materials Management: Facts and Figures.) That was almost 75.2 million tons of waste; today, due to many companies and entrepreneurs trying to tackle this problem, that number has been cut down to 36 million tons, nearly half of the original. (Advancing Sustainable Materials Management: Facts and Figures.)
In conclusion, the lifecycle of an SSD is a complicated one, and one that emits an almost unwelcome exchange of waste for the product it produces. However, in the wake of the 21st century and rise of the Solid-State Drive, wastes like these seem almost necessary if technology is to advance further. Unfortunately, this emission of waste is almost indefinite and also seems more and more absolute for the years to come. However, the EPA recognizes that in using various methods of recycling and chemical bonding, wastes like these can be reduced by almost 20% in case of the silicon sludge. Little by little, we can reduce waste and look for greener, healthier options to power a booming technological market, and in this new light, the wastes that are traded for better technologies doesn’t seem so bad after all.
1. “Advancing Sustainable Materials Management: Facts and Figures.” EPA, Environmental Protection Agency, 21 Nov. 2017, www.epa.gov/smm/advancing-sustainable-materials-management-facts-and-figures. “Semiconductor Industry.” EPA, Environmental Protection Agency, 2 May 2017, www.epa.gov/f-gas-partnership-programs/semiconductor-industry.
2. “Global Mitigation of Non-CO2 Greenhouse Gases: Semiconductor Manufacturing.” EPA, Environmental Protection Agency, 9 Aug. 2016, www.epa.gov/global-mitigation-non-co2-greenhouse-gases/global-mitigation-non-co2-greenhouse-gases-semiconductor.
3. EPA. “Guides to Pollution Prevention.” The Printed Circuit Board Manufacturing Industry, June 1990.
4. EPA. “Plastics.” Plastics, Mar. 2015, www3.epa.gov/epawaste/conserve/tools/warm/pdfs/Plastics.pdf.=
5. “Semiconductor Manufacturing: National Emission Standards for Hazardous Air Pollutants (NESHAP).” EPA, Environmental Protection Agency, 13 July 2016, www.epa.gov/stationary-sources-air-pollution/semiconductor-manufacturing-national-emission-standards-hazardous.
7. “Final Report: Recycling of Silicon-Wafers Production Wastes to SiAlON Based Ceramics with Improved Mechanical Properties.” EPA, Environmental Protection Agency, 6 Dec. 2007, cfpub.epa.gov/ncer_abstracts/index.cfm/fuseaction/display.highlight/abstract/8693/report/F.
8. Yoko, Akira, and Yoshito Oshima. “Recovery of Silicon from Silicon Sludge Using Supercritical Water.” The Journal of Supercritical Fluids, Elsevier, 31 Dec. 2012, www.sciencedirect.com/science/article/pii/S0896844612003907.
9. “Design for the Environment Alternatives Assessments.” EPA, Environmental Protection Agency, 14 Dec. 2017, www.epa.gov/saferchoice/design-environment-alternatives-assessments.
FMS 001 A06
19 March 2018
No More Funny Business
In the early era of film, laughter was elicited from audiences simply by placing two cats with miniature boxing gloves in front of a camera, shoving a pie in a policeman’s face, or dressing up a clown and having him slip on a banana peel in song and dance. Buster Keaton, famous director and actor, proclaimed “you cannot compare the successful comedy of today to the one of five years ago.” Keaton brought a unique approach to comedy that was ahead of his time, which made a lasting impression on the motion picture industry. Richard Koszarski’s An Evening's Entertainment: The Age of the Silent Feature Picture 1915-1928 describes the status of film at the time, stating, “by 1915 the motion-picture industry had achieved a certain sophistication in the way it was produced, marketed and exhibited its pictures.” With a more sophisticated product came a more sophisticated audience; people were no longer satisfied with silly caricatures, and demanded more out of their viewing experience. Nine years later, Keaton delivered with Sherlock, Jr., a satirical story about a young man working as an operator in a local theater. In this motion picture, he demonstrates his comprehension of the art of comedy, using relatable scenes such as the missing dollar, and the interaction at the end of the film between the “detective” and the girl he admires, including the subtle head scratch. While Keaton’s films do not rely on slapstick, absurdity is not absent in them in the slightest. Keaton makes sure the scenes which emphasize physical humor are outstanding, and doesn’t hold back. Between the exploding billiards ball and the makeshift sailboat-car, Sherlock, Jr. is littered with “simple” humor that amazes audiences and critics alike. Buster Keaton was a visionary for entertainment during his time, and his films should stand as a model for what contemporary comedies strive for. Keaton was able to tap into the emotions of his audiences through relatability and personality, evoking the most authentic human reaction, laughter.
Human or Reptilian?
Time and time again in the film industry, certain films rise above the competition and attract global levels of attention. These films are celebrated as art pieces, and even recognized by our government in the Hollywood hall of fame, the National Film Registry. Two films can be compared side-by-side and appear similar at first glance, with identical characters, same settings, and comparable “unexpected twists,” but there are other factors which distinguish successful films from weeknight rentals. The two largest determinants of a film’s success are sound and acting quality, and furthermore, their ability to evoke either empathy or sympathy from the viewer. The more subtle of the two factors, sound, impacts the reptilian part of the human brain, specifically the subconscious “motivated by visual images, sounds, touch, smell, and taste, … [which] helps us to create a vivid mental image.” (Alex 2010) This mental image, stimulated by dramatic sound effects such as music, laugh tracks, and various onomatopeia, can have implications such as “evoking emotions, triggering the release of stress chemicals and impacting the development of new neural pathways in the brain,” (Woods 2017) meaning sounds change the way we perceive a situation, for better or for worse. Donald Crafton, in “Introduction: the Uncertainty of Sound,” claims “few demarcations are so sharply drawn, so elegantly opposed, so pristinely binary. In the movies, sound is either off or on.” This may be true concerning talkies, but sound is commonly conflated with audible dialogue in this discussion. Sound in film is not exclusive to dialogue, and, in fact, may incite a stronger response in silent films due to its isolation from the acting. Even the earliest directors in Hollywood noticed this and incorporated music and effects as accessories to the production. In silent films, both elements are independent of each other, implying that each needs to complement the other without becoming dependent. In modern motion pictures, it is a common occurrence that a film is carried solely by famous acting talent, or vice versa- poor acting compensated by incredible soundtracks, sound effects, and timing.
Parallel to the auditory experience in filmare the personalities presenting the narrative, the actors. This aspect of entertainment held true long before motion pictures were brought into reality. A person’s ability to emulate a character and capture their emotion through genuine facial expression and body movement is an ability unattainable to most; the intangible element, authenticity, is a quality that cannot be taught. The majority of actors lead entire lives without being able to completely immerse themselves into their character, acting requires more than a simple understanding of their character’s strife. The few who accomplish this feat are regarded as “stars” in Hollywood, powerful bodies of energy illuminating the darkness. Daisuke Miyao acknowledges one of these stars who shined during the infancy of motion picture (1915) in his book, Sessue Hayakawa: Silent Cinema and Transnational Stardom. Sessue Hayakawa’s success is remarkable because of the adversity he faced as an Asian immigrant in a close-minded America, especially being hired by and performing for Americans while working with other American actors. It is nigh impossible that Hayakawa avoided conflict due to the intense political climate, the writhing “yellow fever”, and the beginnings of the war soon to be known as World War I, which pitted the United States and Hayakawa’s home country, Japan, against each other. American directors auditioned Hayakawa for stereotypical, villainous roles in their films, which made him infamous in Japan. Despite surviving a suicide attempt, due to an unhealthy relationship with his father, and being surrounded by consistent hatred, Hayakawa succeeded. His raw passion and fervor were undeniable, as every second of screen time was dominated by the Japanese star. Critics, audiences, Americans, Japanese, children, and elders were all captivated by the storytelling capability of this young actor, and his paycheck reflected it. Hayakawa’s legacy serves as evidence that appearance and personal history are thrown out the window when the director calls action; all that matters is makinge the person on the other side of the screen buy into the entertainment phenomenon of a motion picture.
Woods, Gae-Lynn. “The Effects of Sound in the Human Brain.” Livestrong.com, 14 Aug. 2017,
Alex. “Triggers of the Reptilian Brain.” Feed The Right Wolf, 3 Nov. 2010,