Published: February 20, 2025
Silicon Is Everywhere
Semiconductors Before Transistors
From Natural Ore to Artificial Crystals (Germanium and Silicon)
Then, From Germanium to Silicon
In the series on miniaturization from the first to the previous volumes, I primarily discussed the “device” aspect of semiconductor devices. In the next three volumes, I will touch on devices to some extent, but I would like to talk mainly about “semiconductors” as a material.
Volume 10: Shift to Larger Diameter Silicon Wafers: How a Common Material, Silicon, Became a Main Player
I will talk about silicon (Si) in this volume, followed by a discussion on silicon wafers in the next volume, and then in the volume after that, I will address the “shift to larger diameter silicon wafers.” In this volume, I will begin by discussing the material of silicon and how silicon became mainstream in semiconductor devices.
Silicon is found everywhere in our daily lives. You might wonder, “Where exactly is silicon?” Silicon is a major component of the ground (or the Earth’s crust, in technical terms), including rocks, stones, sand, and soil. However, it does not exist in its pure form; instead, it is commonly found in compounds such as SiO2 (silicon dioxide).
Note: Soil consists not only of finely broken particles of rocks and sand but also organic matter, such as decayed animals and plant material.
For example, beach sand contains a significant amount of silicon. The photo below shows Shirarahama Beach in Shirahama Town, Wakayama Prefecture, Japan, which is famous for its white sand. This white sand is mostly quartz grains, and quartz is ore composed of SiO2. It is said that this sand was once used as a raw material for glass production. However, due to ongoing development, the beach has eroded, and sand from Perth, Australia, has been added to restore it.
Now let’s look at the numbers. The most abundant element in the Earth’s crust is oxygen, followed by silicon (Si). The reason for the high content of oxygen is that many of them are present in the form of compounds that contain oxygen, such as SiO2. Although the exact proportions by mass may vary slightly depending on the source, oxygen accounts for approximately 46 to 47%, silicon 27 to 28%, aluminum 8%, and iron 5%. Other elements contribute less than 5%. Oxygen and silicon make up three-quarters of the Earth’s crust by mass.
Because silicon does not exist in its pure form in nature, it must be extracted from SiO2-rich ores for industrial use. Artificially refined silicon, as shown in the photo below, is a gray material with a metallic luster. It is “hard but brittle,” meaning it cannot be bent or shaped like a metal. Excessive force can cause it to crack or break.
I mentioned in Volume 1, germanium (Ge) was initially used as the primary semiconductor material for a while after the invention of the transistor. The concentration of germanium in the Earth’s crust is extremely low, only about 1.5 to 2 ppm (0.00015 to 0.0002%). Unlike silicon, germanium is a very rare element.
From here, I would like to talk about the history of semiconductors until this common material called silicon became a main player in semiconductors.
Long before the invention of the transistor, the story goes back to the 19th century. Several discoveries about semiconductors were made in the 19th century.
One notable property of semiconductors, unlike metals, is that their electrical resistance decreases at high temperatures. In 1833 (some sources suggest 1839), Michael Faraday first discovered the phenomenon in a silver ore called silver sulfide (Ag2S). Although I am not sure since when the term “semiconductor” has been used, this discovery is considered to be the first case where the semiconductor specific properties were discovered.
Next, after a short period of time, in 1874, Karl Ferdinand Braun, a German scientist and inventor of the cathode-ray tube, sometimes called Braun tube, found that rectification, a phenomenon in which current flows more easily in one direction than the other, occurs when a metal needle is placed in contact with metal sulfides, compounds of metal and sulfur. Braun later won the Nobel Prize in Physics in 1909. Examples of naturally occurring metal sulfides include silver sulfide (Ag2S), pyrite (FeS2), and galena (PbS), all of which are semiconductors. In the same year, an English scientist named Arthur Schuster discovered rectifying properties at the interface between copper and copper oxide. Copper oxide is also classified as an oxide semiconductor.
Note: Some sources report that rectification was already observed in 1835.
Braun’s discovery eventually paved the way for crystal detectors, and Schuster’s work led to the development of rectifiers. Both devices use rectification properties: crystal detectors were used to extract signals from received radio waves in wireless communication devices such as radios and radar, while rectifiers were used to convert alternating current (AC) to direct current (DC). Crystal detectors also paved the way for the later invention of the transistor.
A crystal detector consists of a fine metal needle placed in contact with an ore, such as galena or pyrite. In about 1905, crystal detectors were independently invented and patented by several individuals around the world. It may be stated that these devices were the first semiconductor devices.
The crystal detector can be regarded as a Schottky barrier diode in modern terms, but the mechanism behind its rectification properties was not understood at the time. As I explained in Volume 6, the rapid development of solid-state physics based on quantum mechanics eventually clarified the nature of semiconductors and led to the invention of the transistor. Because it was 20 years before quantum mechanics was established, it is not surprising that the mechanism was not understood.
Although the crystal detector was later replaced by a vacuum tube diode, as the frequency of radio waves increased, vacuum tubes could no longer handle the higher frequencies, leading to the use of crystal detectors again.
During World War II, radar development became a critical subject around 1940, and crystal detectors emerged as essential devices in this field. However, the natural ore seemed to have unstable properties. Since it is natural, it is reasonable to have different properties from one object to another, and it is quite possible that even within a single clump there are subtle differences in properties depending on the location. It seems that they used to adjust the needle to locate areas with better characteristics. It is difficult to describe it as an industrial product.
Then, instead of conventional natural ores, artificial crystals such as silicon and germanium are used. It is finally time for the main player to appear. By the 1930s, advances in solid-state physics based on quantum mechanics had deepened the understanding of semiconductors, enabling scientific explanations for rectification properties. A framework was established to enable the research and development of materials and devices based on scientific understanding, leading to transistors and integrated circuits using silicon and germanium.
|
|
Point Contact Germanium Diodes (Advanced Form of Crystal Detector)
Incidentally, silicon was discovered in 1823, whereas germanium was discovered later, in 1885 (or 1886, depending on the source). The existence of germanium had been predicted earlier based on the periodic table of elements.
When I was a child, probably around 1970, I remember building a crystal radio using a germanium crystal detector. It might have been a supplement from a magazine.
While the crystal detectors using germanium consisted of a single metal needle placed on germanium, the first transistor, known as the point-contact transistor, had two needles placed on germanium. It could be considered as simply adding one more needle to the crystal detector. I believe that research on crystal detectors contributed to the invention of the transistor.
In my explanation of semiconductors at the beginning of Volume 1, I mentioned that when the transistor was invented, both germanium and silicon were used, but that the superiority of silicon in terms of practical use was established and silicon came to be used primarily. I think it was around the late 1950s. Let me explain this point in more detail.
Let’s start with the differences in the device performance. Germanium offers faster operation, but it is sensitive to heat and does not perform well at high temperatures. Silicon, on the other hand, remains stable even at high temperatures. Additionally, silicon is superior in terms of leakage current. For example, Bell Laboratories, which invented the transistor, eventually became convinced of the superiority of silicon and focused on silicon from a certain period.
I mentioned in Volume 1 that germanium transistors were used in transistor radios. Since transistor radios do not generate significant heat, the heat sensitivity of germanium was not an issue. However, when using germanium in televisions, the situation is different. There are components that generate heat, and with high current flowing through germanium, self-heating leads to high temperatures, making silicon indispensable. As discussed in Volume 4, calculators initially used germanium transistors, but due to their susceptibility to temperature changes, silicon transistors were adopted a year later.
Another important factor is the availability of a stable oxide layer, which directly relates to the ease of device fabrication. The availability of a stable oxide layer is so important that today's silicon integrated circuit processes would be unimaginable without it, but this is not a given. germanium cannot form a stable oxide layer.
As a preliminary step to silicon wafers, I talked about how silicon, a common material, came to play a key role in the industry. In the next volume, I would like to talk about how silicon wafers are made.
For more blog posts, please click the links below.
Volume 1: Semiconductor Miniaturization: What is Moore’s Law?
Volume 2: Semiconductor Miniaturization and Manufacturing Process
Volume 3: Semiconductor Miniaturization and International Technology Roadmap
Volume 4: Semiconductor Miniaturization and Semiconductor Business
Volume 5: Semiconductor Miniaturization and Semiconductor Business (Part 2)
Volume 6: Semiconductor Miniaturization and Semiconductor Devices
Volume 7: Semiconductor Miniaturization: What is MOSFET Scaling?
Volume 8: Semiconductor Miniaturization: Limitations of MOSFET Scaling
Volume 9: Semiconductor Miniaturization and Analog Circuits