Wide Band Gap Semiconductors

What was the first semiconductor in use?

Think a minute.

Was it Silicon? Germanium? The Transistor from Bell Labs?

Nope. It was the crystal diode in old crystal radio sets.


A semiconductor diode, the most common type today, is a crystalline piece of semiconductor material with a p–n junction connected to two electrical terminals. Semiconductor diodes were the first semiconductor electronic devices. The discovery of asymmetric electrical conduction across the contact between a crystalline mineral and a metal was made by German physicist Ferdinand Braun in 1874. Today, most diodes are made of silicon, but other materials such as gallium arsenide and germanium are used.

But the first ones were made of Galena


1874: Semiconductor Point-Contact Rectifier Effect is Discovered

In the first written description of a semiconductor diode, Ferdinand Braun notes that current flows freely in only one direction at the contact between a metal point and a galena crystal.

Ferdinand Braun shared the 1909 Nobel Prize in Physics with Guglielmo Marconi

German physicist Ferdinand Braun, a 24-year old graduate of the University of Berlin, studied the characteristics of electrolytes and crystals that conduct electricity at Würzburg University in 1874. When he probed a galena crystal (lead sulfide) with the point of a thin metal wire, Braun noted that current flowed freely in one direction only. He had discovered the rectification effect at the point of contact between metals and certain crystal materials.

Braun demonstrated this semiconductor device to an audience at Leipzig on November 14, 1876, but it found no useful application until the advent of radio in the early 1900s when it was used as the signal detector in a “crystal radio” set. (1901 Milestone) The common descriptive name “cat’s-whisker” detector is derived from the fine metallic probe used to make electrical contact with the crystal surface. Braun is better known for his development of the cathode ray tube (CRT) oscilloscope in 1897, known as the “Braun tube” (Braunsche Röhre) in German. He shared the 1909 Nobel Prize with Guglielmo Marconi for his “contributions to the development of wireless telegraphy,” mainly the development of tunable circuits for radio receivers.

So lead sulfide was first. Close behind it was another material. Silicon carbide.


Electronic circuit elements

Silicon carbide was the first commercially important semiconductor material. A crystal radio “carborundum” (synthetic silicon carbide) detector diode was patented by Henry Harrison Chase Dunwoody in 1906. It found much early use in shipboard receivers.

While not as sensitive as the Lead Sulfide detector, the Silicon Carbide “carborundum” detector was much more sturdy. It was used on ships as it stood up better to rough physical and electrical conditions. Still, ships carried spares just in case something like a lightning strike fried the primary one.

Then there’s another odd quirk of history. The LED has been around since before the transistor.


The phenomenon of electroluminescence was discovered in 1907 using silicon carbide and the first commercial LEDs were based on SiC. Yellow LEDs made from 3C-SiC were manufactured in the Soviet Union in the 1970s and blue LEDs (6H-SiC) worldwide in the 1980s. The production was soon stopped because gallium nitride showed 10–100 times brighter emission. This difference in efficiency is due to the unfavorable indirect bandgap of SiC, whereas GaN has a direct bandgap which favors light emission. However, SiC is still one of the important LED components – it is a popular substrate for growing GaN devices, and it also serves as a heat spreader in high-power LEDs

Seems that back when they were using those carborundum detectors, some of them glowed, and that lead to the discovery of the LED.


SiC LEDs came before transistors

Early in the 20th century, experimenters were finding that crystals of various substances such as germanium could give ‘unsymmetrical passage of current’ or rectification as we would know it, which found use in ‘crystal’ radios. When silicon carbide was tried, a strange phenomenon occurred; the crystal glowed yellow, sometimes green, orange or even blue. The first LED had been discovered, forty years before the transistor.

As an LED, SiC was soon superseded by gallium arsenide and gallium nitride with 10-100 times better emission but, as a material, SiC still generated interest in the electronics world; it has a thermal conductivity 3.5 times better than silicon and can be heavily doped for high conductivity while still maintaining high electric field breakdown. Mechanically, it is very hard, inert and has a very low coefficient of thermal expansion and high temperature rating. SiC does not even melt – it sublimates at about 2700⁰C.

Sturdy stuff indeed. Present at the start of the radio era. First LEDs, then set aside for more efficient materials. We’ll come back to those sturdy qualities a bit later.

If the walk was over a meteor impact crater you might find a few specks – the only naturally occurring SiC is in the form of Moissonite, debris from a supernova or ejecta from carbon-rich, red giant stars picked up in space and ending up as micron-sized particles in meteorites. Stardust indeed.

We might have never noticed SiC’s existence but in 1891 American inventor Edward G Acheson was trying to find a way to produce artificial diamonds, by heating clay (aluminium silicate) and carbon. He noticed shiny hexagonal crystals attached to the carbon arc light used for heating and called the compound carborundum, thinking it was a form of crystallized alumina like corundum. He might have thought he’d hit second best, as rubies and sapphires are types of corundum, but he realized he had something new, a compound nearly as hard as diamond which could be made as chips or powder on an industrial scale with application as an abrasive.

From that time forward we’ve used a lot of silicon carbide as industrial abrasives. Semiconductors not so much. Yet is is a semiconductor. (So I’m fighting back this urge to go get some Ultra Coarse carborundum paper and try making a diode with it ;-)

If we look at the current trend in semiconductors, the hot item is something called “Wide Band Gap” semiconductors. Why? I wondered. With a forward voltage of 0.2 VDC for Germanium, you get very little resistive heating. Germanium has been used for big power transistors for a long time and was the common material back when transitors were just replacing tubes. Silicon has a 0.7 VDC forward voltage. If you had a 2 VDC forward voltage, wouldn’t that make more heat and more problems? Well, it turns out there are other considerations.


Has an interesting chart with the band gap voltage for various materials. It is interesting to note that Diamond is listed with a band gap of 5.47 V. In theory one ought to be able to make a radio detector out of a diamond ring (details left as an exercise for the reader ;-) Silicon is shown as 1.12 V. This is different from the forward voltage of a junction device ( 0.7 V) due to doping and construction specifics. Germanium is 0.67 V.

Various SiC compounds with different ratios of C and Si give band gaps from 2.0 to 3.0 to 3.3 V. Relatively high. (Then there is Grey Tin at 0.0 V to 0.8 V – which gives me ideas but who knows what problems lurk…)

There’s a lot of other choices, from things like Aluminum Nitride or Boron Nitride both over 6 V to Zink Oxide or Zinc Sulfide at near 3.5 V. or Copper Sulfide at 1.2 V (first effective thin film solar cell). The point? While it would be necessary to take into account all the physical and chemical interactions of a given material, there are a lot of them to choose from and band gaps vary all over the place.

After a flush of activity in the Silicon technologies, folks have started to explore all these other system. Turns out that there are special properties for wide band gaps.


In the last century, silicon semiconductor-based power electronics – which control or convert electrical energy into usable power – transformed the computing, communication electric vehicle, and energy industries and gave consumers and businesses more powerful laptops, cell phones and motors. Over the coming decade, that era will begin to come to an end.
Today, Wide Band Gap (WBG) semiconductors offer new opportunities to achieve unprecedented performance while using less electricity. WBG semiconductors such as silicon carbide (SiC) and gallium nitride (GaN) can operate at higher temperatures, have greater durability and reliability at higher voltages and frequencies and be smaller, more efficient and cost less. A WBG semiconductor-based inverter, which switches electricity from direct current to alternating current, could be four times more powerful, half the cost and one-fourth the size and weight of a traditional inverter.

Indeed, adoption of WBG semiconductors is already happening today, notably in photovoltaic power generation – solar power replacing power stations – followed by more general power grid applications. Next, WBG semiconductors will transform the plug-in EV industry – making it easier and cheaper to own an EV and/or give it longer range. They may reduce the size of a vehicle cooling system by about 60% and cut the size of a fast DC charging station (eg 60kW) to the size of a kitchen microwave.

Heady stuff… The band gap determines the color of light emitted in LEDs as well as that absorbed in solar cells. This bit of physics matters as does the temperature limits of a given material and what voltage will break down the gap.


Wide-bandgap semiconductors (WBG or WBGS) are semiconductor materials which have a relatively large band gap compared to typical semiconductors. Typical semiconductors like silicon have a bandgap in the range of 1 – 1.5 electronvolt (eV), whereas wide-bandgap materials have bandgaps in the range of 2 – 4 eV. Generally, wide-bandgap semiconductors have electronic properties which fall in between those of typical semiconductors and insulators.

Wide-bandgap semiconductors permit devices to operate at much higher voltages, frequencies and temperatures than conventional semiconductor materials like silicon and gallium arsenide.
They are the key component used to make green and blue LEDs and lasers, and are also used in certain radio frequency applications, notably military radars. Their inherent qualities make them suitable for a wide range of roles, and they are one of the leading contenders for next-generation devices for general semiconductor use.

But why? There’s clearly something other than resistive loss at a junction going on here. Seems that it is thermalized electrons and how much they can leak across a junction.

The wider bandgap is particularly important for allowing wide bandgap devices to operate at much higher temperatures, on the order of 300 °C. This makes them highly attractive for military applications, where they have seen a fair amount of use. The high temperature tolerance also means that these devices can be operated at much higher power levels under normal conditions. Additionally, most wide bandgap materials also have a much higher critical electrical field density, on the order of ten times that of conventional semiconductors. Combined, these properties allow them to operate at much higher voltages and currents, which makes them highly valuable in military, radio and energy conversion settings. The US Department of Energy believes they will be a foundational technology in new electrical grid and alternative energy devices, as well as the robust and efficient power components used in high energy vehicles from electric trains to plug-in electric vehicles. Most wide-bandgap materials also have high free-electron velocities, which allows them to work at higher switching speeds, which adds to their value in radio applications. A single WBG device can be used to make a complete radio system, eliminating the need for separate signal and radio frequency components, while operating at higher frequencies and power levels.

Currently, WBG materials are far behind the development level of conventional semiconductors, which has seen massive investment since the 1970s. However, their clear inherent advantages in many roles, combined with some unique features of the materials themselves, has led to increasing interest as an everyday material, replacing silicon in even more commonplace roles. Their higher energy density handling is particularly interesting in attempts to extend Moore’s law, where conventional technologies appear to be reaching a density plateau.

The band gap also determines the color of light made, so if you want out of the IR range, you must go wide band gap.

Use in devices

Wide bandgap materials have several characteristics that make them useful compared to lower bandgap materials. The higher energy gap gives devices the ability to operate at higher temperatures, and for some applications, allows devices to switch larger voltages. The wide bandgap also brings the electronic transition energy into the range of the energy of visible light, and hence light-emitting devices such as light-emitting diodes (LEDs) and semiconductor lasers can be made that emit in the visible spectrum, or even produce ultraviolet emission.

There’s a bunch more about the QM effects and electron behaviours in the gap in that article. Worth a read if you want to know the hows and whys.

Thermal properties

Silicon and other common materials have a bandgap on the order of 1 to 1.5 electronvolt (eV), which implies that such semiconductor devices can be controlled by relatively low voltages. However, it also implies that they are more readily activated by thermal energy, which interferes with their proper operation. This limits silicon based devices to operational temperatures below about 100 °C, beyond which the uncontrolled thermal activation of the devices makes it difficult for them to operate correctly. Wide-bandgap materials typically have bandgaps on the order of 2 to 4 eV, allowing them to operate at much higher temperatures on the order of 300 °C. This makes them highly attractive in military applications, where they have seen a fair amount of use.

Melting temperatures, thermal expansion coefficients, and thermal conductivity can be considered to be secondary properties that are essential in processing, and these properties are related to the bonding in wide bandgap materials. Strong bonds result in higher melting temperatures and lower thermal expansion coefficients. A high Debye temperature results in a high thermal conductivity. With such thermal properties, heat is easily removed.

So there’s the answer. It’s not about the creation of resistive heat, it is about the energetic electrons in a hot device not jumping the gap… This lets them work at higher voltages too. This is where are old friend Silicon Carbide returns to the stage:

High power applications

The high breakdown voltage of wide bandgap semiconductors is a useful property in high-power applications that require large electric fields.

Devices for high power and high temperature applications have been developed. Both gallium nitride and silicon carbide are robust materials well suited for such applications. Due to its robustness and ease of manufacture, semiconductors using silicon carbide are expected to be used widely, create simpler and higher efficiency charging for hybrid and all-electric vehicles, reduced energy loss and longer life solar and wind energy power converters, and elimination of bulky grid substation transformers. Cubic boron nitride is used as well. Most of these are for specialist applications in space programmes and military systems. They have not begun to displace silicon from its leading place in the general power semiconductor market.

Then there’s the cutting edge where folks have figured out how to make organic semiconductors. Who knows what that will lead to.


A Wide Band Gap Naphthalene Semiconductor for Thin-Film Transistors

Organic field transistors attracted considerable attention due to their excellent compatibility with flexible and biological substrates, low cost, easy processing and chemical tunability of properties that enable their wide application in electronics, including: large area flexible displays, smart cards and sensors, biomedical devices and radio frequency identification tags among many more applications. Many high mobility organic semiconductors have been developed based on π-extended acenes, oligothiophene and their derivatives. A majority of organic semiconductors have relatively low band gaps, typically in the visible region of the spectrum (~2-3 eV). On the other hand, large band-gap semiconductors would be beneficial due to their intrinsically higher stability and illumination-independent transport. They could likely find applications in transparent field-effect transistors that are currently fabricated with oxide semiconductors and carbon nanotubes.

Researchers led by professor Dmitrii Perepichka at McGill University in collaboration with professor Hong Meng at Nanjing Tech University in China developed a novel organic semiconductor 2,6-bis(4-methoxyphenyl)naphthalene (BOPNA) with unprecedentedly large band gap of 3.35 eV and an apparent hole mobility measured in thin-film organic field-effect transistors in a saturation regime. Their work, now published in Advanced Electronic Materials, presents a detailed analysis of photo-physical, electrical, and structural properties of BOPNA, showing its high thermal stability, facile morphology reorganization upon annealing at low temperatures, and band-like transport established through the temperature-dependent mobility measurements..

The researchers reported a multi-gram scale high-yield synthesis of BOPNA by a simple Suzuki coupling of commercially available 2,6-dibromonaphthalene with 4-methoxyphenylboronic acid. Photo-physical studies revealed the optical band gap of 372 nm demonstrating complete optical transparency of the film in the visible region. An excellent photo-stability of current response of BOPNA was also observed in the output characteristics of the organic field transistors.

In totality, a novel naphthalene semiconductor BOPNA with a wide band gap of 3.35 eV, high thermal, high photo-stability and complete optical transparency has been synthesized herein. A distinctive characteristic of BOPNA-based transistors is their excellent photo-stability and independence of the electrical characteristics of the illumination conditions; this is important for a wide-range practical application of low-cost organic electronics. Quoting Dr. Perepichka, “it is now clear that a notion of the “semiconducting region” in organic semiconductors being limited by 3 eV band-gap is just a misconception and that it is possible to design efficient field-effect transistors with large band-gap organic materials”.

So there you have it. A quick tour of the Wide Band Gap semiconductor and why you will be seeing more of them in the future. Literally, as they are the LED Light Bulbs in all the stores…

Subscribe to feed


About E.M.Smith

A technical managerial sort interested in things from Stonehenge to computer science. My present "hot buttons' are the mythology of Climate Change and ancient metrology; but things change...
This entry was posted in History, Tech Bits and tagged , , . Bookmark the permalink.

21 Responses to Wide Band Gap Semiconductors

  1. hillrj says:

    Hi EM, Not totally off-topic. Are you following the work of Bob Greenyer and the MFMP group?. What do think about the Russian “Strange Radiation” topic?

  2. E.M.Smith says:

    I’ve somewhat soured on the whole LENR thing since several “Shipping Now” dates have come and gone for Rossi. Yes, I’ve watched the MF Memorial Project on and off; and it’s nice they show some “color in the pan”, but there really ought to be a lot more than just suspicions by now.

    I’m not familiar with the “Strange Radiation” topic. Don’t know anything about it. Sounds interesting though. I do think any time there is light (photons of any energy really including radio waves) there’s a similar QM process at work. The “trick” is finding how to interconvert them all and make them productive. Sadly, while I have a conceptual grasp of QM, I don’t have a full enough grasp of it to do anything useful with it. Just marvel…

    Don’t know about Bob Greenyer at all. I’ve really not paid much attention for about the last 2 years. (Basically when wrapping up last Florida contract and coming back to Cali, then the couple of runs back and forth for various “Surprise!” events – took a lot out of my “fun time” allotment ;-) and LENR is very much a “fun time muse” thing… after other stuff is pretty much done / caught up.

    Basically where I left it was “I’ll now believe it is real when folks can buy one at Sears or Walmart”… and nothing has really changed from my POV since then. IFF there was a simple DIY kit you could knock together that would boil a cup of water in 2 minutes with 10 W input (absolutely clearly working) I’d buy the kit and make one… if under $200 or so for parts. Other than that, I’m waiting for the UL Lab Certification… Burned a few too many times with excited then not…

    But IF there is something interesting and absolutely clearly happening that’s not subject to slight of hand or ambiguity, it would be interesting to know about. If it is Yet Another 1 mW of suspected excess heat in a sea of 100 W of adjustments and allowances, er, not so much…

    Basically I want to see boiling water from a D cell and not a few mW unaccounted for in sloppy calorimetry. I’m past the point of accepting the computed results; it’s now got to be spectacularly obvious. That’s what happens when the suspicion of con-job hangs around in the air too long…

    But if you’ve got something interesting to point at, point away.

  3. tom0mason says:

    And there was the other awkward device of the 1930s-40s-50s—? the Selenium rectifier.
    Fragile, prone to failure even when new, couldn’t stand the heat, often fractured when too cold, couldn’t work at low voltages (1V or below), but they were better than some rectifier tubes. That red, pink, or gray finned thing that so often had sharp edges to it. Many an old TV had them in there.

  4. cdquarles says:

    So much of the really good stuff in chemistry simply isn’t known in the wider world. Surface effects, phosphorescence, sonoluminescence, chemiluminescence (fire flies), and lasing. IR is light. IR is not heat. Light can be converted to other forms of energy. Other forms of energy can be converted to light. Heat is the internal kinetic energy of a sample of matter and *only* its internal kinetic energy.

    Consider 18 grams of light water (about 1/2 fl oz). As a solid, at standard conditions, it occupies a bit more than 18 cubic centimeters (about a tablespoon) and would fit in many people’s palms. As a liquid, if pure, it occupies exactly 18 cubic centimeters at 4C/40F. As a gas, it occupies 22.4 liters (about 24 quarts) or 22,400 cubic centimeters or 2.24 cubic meters. Now that’s “thermal” expansion. Gases are mostly empty space. Only outer space has a greater percentage of emptiness to matter. At standard temperatures (thermodynamic temperature is not the same thing as a radiant color/brightness temperature) and pressures, the atoms and molecules that are what air is, are moving on the order of 1 km/sec or 0.609 mi/sec. Not per hour, mind you, per second! The net biased random walk that is wind is meters per second. Even sound carried in air is 0.3 km/sec, and that basically doesn’t ‘move’ the air carrying it.

    So, is it possible for LENR to occur? Absolutely, depending on conditions. Can this be made useful? Maybe, also depending on conditions. Conditions matter!

  5. cdquarles says:

    Oh, lest I forget, part of what makes biological membranes work is ‘doping’ them with ‘aromatics’, such as cholesterol. Then add surfaces such as porous polypeptides (plastics … ;p), with and without polyalcohols.

  6. jim2 says:

    Here are some WBGS articles from Mouser.


  7. Soronel Haetir says:

    I still find it amazing how transistors missed a couple decades, Lilienfeld had them in the late 20s but it took Bell Labs to bring them to the world. In Bell Lab’s defense they did have a proper explanation of what was going on but still … Lilienfeld’s devices worked (the Bell teams made some and they worked as disclosed).

    The only reason I know about it is I bought a batch of back issue Analog and the Science Fact article on was about it (I don’t recall the month, was 1961 and was the same issue as a Dune sand worm cover).

  8. E.M.Smith says:

    I vaguely remember some sort of transistor thing being before Bell, looking it up from the name given is interesting:


    1926: Field Effect Semiconductor Device Concepts Patented
    Julius Lilienfeld files a patent describing a three-electrode amplifying device based on the semiconducting properties of copper sulfide. Attempts to build such a device continue through the 1930s.

    Polish-American physicist and inventor Julius E. Lilienfeld filed a patent in 1926, “Method and Apparatus for Controlling Electric Currents,” in which he proposed a three-electrode structure using copper-sulfide semiconductor material. Today this device would be called a field-effect transistor. While working at Cambridge University in 1934, German electrical engineer and inventor Oskar Heil filed a patent on controlling current flow in a semiconductor via capacitive coupling at an electrode – essentially a field-effect transistor. Although both patents were granted, no records exist to prove that Heil or Lilienfeld actually constructed functioning devices.

    In 1938 Robert Pohl and Rudolf Hilsch experimented on potassium-bromide crystals with three electrodes at Gottingen University, Germany. They reported amplification of low-frequency (about 1 Hz) signals, but their research did not lead to any applications.

    Stimulated by research into copper-oxide rectifiers at Bell Telephone Laboratories and by explanations of semiconductor rectification by Mott and Schottky (1931 Milestone), William Shockley wrote in December 1939 that “It has today occurred to me that an amplifier using semi conductors rather than vacuum is in principle possible.” Under his direction, Walter Brattain and others performed experiments on such three-electrode devices but did not achieve amplification. On his return to Bell Labs after the war in 1945 Shockley resumed his work on semiconductor devices. Again he failed to achieve his predicted results. In 1946 physicist John Bardeen calculated that surface effects could account for the failure of these attempts to build working devices. (1947 Milestone)

    20 years from first patent mostly on an idea to good idea how to make one go…

  9. Bloke in Japan says:

    Like many things the timeline is complex and multi-faceted.

    Sony made it work commercially. (1 hour)

  10. E.M.Smith says:

    @Bloke in Japan:

    Ah, yes, the era of the “Transistor” (meaning small radio…) from Japan.

    While I was building tube radios at home and school, I bought my first Transistor Radio from some Japanese company ( likely Sony)…

    It sparked my interest in transistors and my first efforts at breadboard based devices.

  11. Tom Bakewell says:

    Somewhere in a lost copy of “Nuts and Volts” there is an article about the more obscure semiconductors. Included are comments about iron pyrite, galena, and rusty razor blades. The author was fascinated enough by the subject that he found a Tektronix 575 (?) Curve tracer and plotted some results. If I can find the reference I will send it your way.

  12. E.M.Smith says:

    Oh, and per the Selenium rectifier Tom0Mason mentioned:

    Yeah, had one on the Honda Trail 90 in the 6 VDC power system. A few inches on a side all fins. Very prone to failure and replacement. They likely ought to have stayed with a simple generator.

  13. p.g.sharrow says:

    IIRC my uncle started feeding me government briefing manuals in 1956 on the status of solid state development. He was an electronic engineer for the Air Force. I built my first solid state radio in 1956, a 1 transistor, 1 diode, battery powered 3″ speaker Heathkit that used a ferrite cored loopstick for tuning.
    The delay in solid state rollout was due to the difficulty in maintaining quality in large scale production. I worked with a guy that was involved in the creation of the solid state industry in California in the 1960s when they had to create this new industry from scratch. His first contribution was plumbing for Bell Labs that could withstand the chemistry involved. followed by the air handling equipment that could withstand the fumes and then the fume scrubbers for Fairchild that side stepped the efforts of CAB and EPA to outlaw all silicon work in California.
    There were a lot of things involve in the development of modern solid state electronics that had nothing to do with the science and had everything to do with political struggles to control the commerce of Electronics and the development of the infrastructure needed to support it …pg.

  14. Simon Derricutt says:

    For checking the bandgaps of various materials, https://www.materialsproject.org/ is very useful (needs sign-on). Somewhat interesting the number of different crystal forms of materials, and the difference that makes to the bands and the gaps. The complexity and variability of the same material, when you look at it from the electron’s point of view, makes it unsurprising how long it took to make working transistors let alone the more esoteric devices. Exactly how pure the material is, its stoichiometry, and the crystal from and face deposited or grown, makes a difference to how electrons travel in it.

    Basically, transistors are pretty difficult to make unless you get all the parameters right. Today’s successes are the result of a huge number of incremental improvements in techniques.

    Back around the turn of the century we were starting to see Silicon Carbide FETs and diodes available for making power supplies. Higher voltage drops, so they used more power when conducting, but on the other hand they could work happily when a lot hotter anyway so cooling became less of an issue. IIRC they also were somewhat faster at the higher voltages, so you got less switching-loss in an SMPS design than Silicon devices. At the time it was a toss-up whether to use the Silicon devices or the new Silicon Carbide ones, as the final designs were about the same efficiency, but I expect that by now the balance is in favour of SiC.

    If you spend a while looking at the band diagrams, you tend to stop thinking of materials as insulators – they’re all semiconductors with different band gaps. Could be some useful devices come out of making active devices using stuff that is normally thought of (and used) as an insulator. I have an idea for a cheap gamma-detector that could be much cheaper than Geiger-Müller tubes, and could maybe “see” down to very soft X-rays or very hard UV. Could be useful for LENR tests, where it is thought that X-rays of less than 5keV are produced but no-one has the kit to measure them.

    As regards Grey Tin, the problem is that it’s only semi-stable, and will gradually change to White Tin above (IIRC) 18°C. White tin (as we normally see it in tinplate, solder etc.), on the other hand, is also semi-stable and changes to Grey Tin below that temperature, which means that soldered joints using Tin don’t work too well in cold climates. May be one of the reasons it’s necessary to inspect all the water-pipes in an old house that’s been unoccupied for some years, where the pipes have been emptied and maybe have been cold for a long time. Brazed joints are better….

  15. E.M.Smith says:


    There’s always the “special uses” for the things that are bogus in our normal conditions. Like, oh, a semiconductor for use on a spacecraft in deep space. Having a whole different set of characteristics at near zero…

    Now realize I’m not advocating for grey tin. I only mentioned it as it’s an odd case of a metastable semiconductor. Just the trouble of needing to make one in a “cold clean room” and keeping it cold through launch makes it sort of silly. But I could see a hypothetical where you needed low temp-only semiconductors and it could be worth a look. I’d also wonder if you could alloy something else in with it to stabilize the grey tin structure at higher temperatures. So I could see a case where someone went off exploring the edge case of grey tin and found something interesting.

    Personally, I find it much more interesting that G.I.s in W.W.II figured out how to make “crystal” sets using a point junction on a razor blade. Discovering the diode hiding in the oxide of iron and ??? on the surface. Or perhaps the sulfide from shaving reactions with skin?

    Like you said, it looks a lot more like everything will work under some conditions as it’s all semiconductors… just most conditions are rather exotic.

    My muse at the moment is a “gas triode”. Seems to me one ought to be able to make a “cold cathode” electron emitter / gas charger and have it in a flat pack with an anode at the other end, then a FET like wrapping around the channel. Think of something about 1/10 the size of a mini-neon bulb, but as a flat channel. I’m pretty sure it could be made to work. Charge carriers being both free electrons and charged gas molecules.

    By making the channel small enough it ought to have high enough conductivity even without a full on discharge / arc. That would also let the Field Effect work through the channel. It ought to be extraordinarily robust to harsh radiation and heat environments… and could be made pretty darned shock resistant too. Basically what these folks are calling a “vacuum transistor” but it doesn’t have a vacuum in it:


    But even with a low probability of hitting, many electrons are still going to collide with gas molecules. If the impact knocks a bound electron from the gas molecule, it will become a positively charged ion, which means that the electric field will send it flying toward the cathode. Under the bombardment of all those positive ions, cathodes degrade. So you really want to avoid this as much as possible.

    Fortunately, if you keep the voltage low, the electrons will never acquire enough energy to ionize helium. So if the dimensions of the vacuum transistor are substantially smaller than the mean free path of electrons (which is not hard to arrange), and the working voltage is low enough (not difficult either), the device can operate just fine at atmospheric pressure. That is, you don’t, in fact, need to maintain any sort of vacuum at all for what is nominally a miniaturized piece of “vacuum” electronics!

    But asking the question of “Is there some way to make the ionized gas a feature instead of a problem?” Can a tough enough cathode and the proper choice of gas make a “gas transistor”?

    Perhaps by putting a screening grid in front of the cathode with a positive charge on it. Ionized gas from the “plate” side would be repelled and tend not to reach or corrode the cathode. It would also tend to accelerate negative charge ions and electrons away from the cathode.

    Essentially a miniaturized one of these:


    Electron tube filled with low-pressure helium gas, and containing a thermionic cathode, control grid, and anode. This tube is for quantitative investigations of the typical properties of a gas-immersed triode, recording the current-potential (IA-UA) characteristics of a thyratron, and observing discharges and discontinuous energy release of helium atoms during inelastic collisions with free electrons.

    Max. heater voltage: 7.5 V AC/DC
    Max. anode voltage: 500 V
    Anode current: approx. 10 mA at 200 V (anode)

    But with a screening grid as a potential improvement….

    Why? Nothing other than wondering if it could work…

  16. Simon Derricutt says:

    EM – I looked at Grey Tin as a low-bandgap semiconductor to convert IR to electricity, but I figured the phase-change problem would knock it out for that use. Puts a lifetime on the device, and I was looking for unlimited lifetime. It would however work, and in the right location might be a cheaper (though less efficient) substitute. I also couldn’t figure out how to get a good deposit, since the two phases have different volumes so a deposited layer of Tin would flake off. Technical problems that maybe someone with enough experience could solve. To get that sort of person involved, though, you first have to show that the principle works. Catch-22 – I need to solve the problem before I can get people in who could solve the problem.

    Interestingly, it’s not so much the rust on the razor blade, even though Fe3O4 is a semiconductor with a fairly-wide bandgap. The sharp point itself makes a diode, with electrons jumping from the point to a plane easier than from the plane to a point, and this is because of the shape of the field. As regards the rust, remember that blacksmiths used to sharpen a file by leaving it to rust in a bucket of water – rust makes the edges sharper. At THz frequencies in nantenna arrays, the diodes used are simply a very sharp point a small distance from a plane. Works a bit better with a thin layer of insulator to tunnel through (around 2nm thick) but also works with that size gap.

    Boron-doped Diamond works as a point electron emitter without any heating needed – in fact any sharp-enough point will do that but if it isn’t tough enough it degrades too fast. The NASA link from 2014 seems like they’ve got all the ducks in a row, and I’d expect that by now they’ve probably built better ones. The ones shown are bidirectional, but of course a point and a plane with the field gate would work just as well for unidirectional current (which we normally use anyway). Note the dimensions, though – though you can make something that works easily enough, making something that works at low voltage, without a heated cathode, and without a vacuum isn’t easy without the right kit.

    There were rumours that a Japanese researcher had produced a material with a very low work-function, which thus makes possible room-temperature electron emission without needing to heat it. Generally the lowest workfunction is around 0.7eV, and the rumours were of around 0.1eV. Gone dark, and I don’t know if that’s because it doesn’t exist or that they’re developing some amazing device from it and are not publicising it yet. One thing with low workfunction surfaces is that they are photoelectric – can be a benefit (see photomultipliers) or a problem.

    As regards small “vacuum” tubes, the standard Vacuum Fluorescent Display are pretty small and rugged. They generally last many years, too. The only failure modes are from leaks (pretty-well non-existent) or problems on the fine electrode wires strung over the character electrodes, where I had one manufacturer who didn’t take the rough edges off the tensioning-springs – the wires heat up in use so expand a bit, and so with the wire bent over a sharp edge and being bent each time it powered up, the wires broke at the spring break-points. I think they subsequently rounded that edge, but I don’t know since I left FAL not too long after that. Getting a vacuum-based integrated circuit would be quite possible with current technology, with elements (tube equivalents) less than a millimetre. Multilayer PCBs can be made using sintered Al2O3 (used in hi-reliability situations, may be in cars by now) with some pretty tight tolerances. More a matter of finding the need to make them, therefore, and a niche where they do the job better and cheaper than the alternatives. Use in very high temperatures would be one of those – tubes can run red-hot providing the workfunctions of the surfaces are correctly designed.

  17. E.M.Smith says:

    Per “red hot tubes”: One of my early projects was a vacuum tube power supply. Used a big ‘ol transformer with a built in tube socket from some giant TV that had died. 5U4 tube. About 400 VDC depending on load…

    So one day I was doing something silly and the power lead I was moving self welded to the chassis… (ZZZZAPPPP! UUUNnnnnnnnng). (Zap being the discharge weld, Unngg being the transformer hum as it loaded up…)

    I looked over and watched as the tube grew a red spot middle of the plate… and then it grew… and then pretty much the entire plate was glowing cherry red… and I realized I could pull the plug out of the wall…

    Got the wire fatigued off the chassis… Fixed what I’d had started out to fix…

    From that time forward, every so often, just for fun, I’d short a (more sturdy not self welding) lead from the HV supply and enjoy watching the cheery glow for a minute… then un-short it…

    I now wish I’d not tossed that power supply when I ran off to college and other interests. It was quite a monster ;-) Ran many a tube experiment / radio / kit. Never failed. Made from “junk” and I just figured I could always make another one if I wanted… not so much now.

    I still get that warm cheery feeling just thinking about how un-Godly sturdy that thing was. If I had it now, I’d be trying my hand at DIY vacuum tubes in various sizes. ( I learned some limited glass blowing along the way too… but I’d likely use metal cans as self shielding.)

    ICs are fun and transistors are nice; but there’s something about vacuum tubes that’s just special ;-) In the corner I have a large (dresser sized) AM / FM / SW tube radio from about 1940. It hums like crazy (likely all the electrolytic caps in the heater / cathodes are leaking) but does get stations (so all tubes work to some degree). On my “someday” list is to go through that radio and fix it up. A “rescue” from a local store that clearly was as interested in throwing it out as selling it. Something like $50? Maybe less… Has one of those 2 foot diameter speakers in it ;-) and a wound antenna about 2.5 feet on a side on a frame that can rotate about 30 degrees around the inside edge of the speaker box. Wood is nice except on top that is a bit worn… Someday…

  18. Tom Bakewell says:

    Ah, the smell of hot transformers salvaged from early TV sets was one of the perfumes of my youth. Formvar wire insulation has a distinct aroma as it gets too hot. And there was the aroma of incandescing resistors pointing out another ‘more homework needed here’ locus. “Learn by burn” was my early mantra.

  19. jim2 says:

    Yep, electronic components cease to work when they lose their integral smoke.

Comments are closed.