Some metalloids are semiconductors, but not all of them. Of the seven widely recognized metalloids (boron, silicon, germanium, arsenic, antimony, tellurium, and polonium), four are well-characterized semiconductors: boron, silicon, germanium, and tellurium. The remaining metalloids, arsenic and antimony, are classified as semimetals, meaning they conduct electricity more readily and behave closer to metals. The overlap between “metalloid” and “semiconductor” is significant but not complete.
Which Metalloids Are Semiconductors
The distinction comes down to a property called the band gap, which is the energy barrier that electrons must overcome to move freely and conduct electricity. Semiconductors have a small but measurable band gap. Insulators have a large one. Metals and semimetals have essentially no gap at all.
Here’s how the metalloid semiconductors compare:
- Boron: 1.5 eV band gap (the widest among them, making it the least conductive)
- Silicon: 1.12 eV band gap (the standard material for computer chips and solar cells)
- Germanium: 0.67 eV band gap (used in early transistors and still used in specialized optics)
- Tellurium: 0.35 eV band gap (the narrowest, used in thermoelectric and infrared applications)
A smaller band gap means electrons can jump into the conducting state more easily, so tellurium conducts better than boron at room temperature. Arsenic and antimony, on the other hand, have their valence and conduction bands slightly overlapping. That overlap eliminates the gap entirely, which is why they’re called semimetals rather than semiconductors. Carbon in its graphite form also behaves as a semimetal, with high electrical conductivity that puts it in the same category as arsenic and antimony rather than with silicon.
How Semiconductors Behave Differently From Metals
The clearest way to tell a semiconductor from a metal is to heat it up. Metals become worse conductors as temperature rises because the vibrating atoms in the metal lattice scatter electrons and slow them down. Semiconductors do the opposite: heating them shakes more electrons free from their bonds, giving them more carriers to conduct electricity. Conductivity goes up with temperature, not down.
The numbers illustrate this starkly. Copper’s resistance increases by about 0.39% per degree Celsius. Silicon’s resistance drops by about 7.5% per degree, and germanium’s drops by about 4.8% per degree. That negative temperature coefficient is one of the defining fingerprints of a semiconductor, and it’s the reason silicon-based devices need careful thermal management. A chip that gets too hot doesn’t just degrade; it can enter a runaway loop where rising temperature increases conductivity, which increases current, which increases temperature further.
Why Silicon Dominates
Silicon’s 1.12 eV band gap sits in a sweet spot. It’s wide enough that silicon doesn’t conduct too much at room temperature (which would make transistors leak current), but narrow enough that small voltages can switch it on and off reliably. Silicon is also the second most abundant element in Earth’s crust, making it cheap to source. Decades of manufacturing refinement have made silicon fabrication extraordinarily precise, with chip features now measured in just a few nanometers.
Germanium was actually used first. The earliest transistors in the late 1940s and 1950s ran on germanium. But its narrower band gap meant more leakage current at room temperature, which caused reliability problems. Silicon took over by the 1960s and never looked back. Germanium still finds use in specialized roles like infrared optics and high-speed transistors where its higher electron mobility is an advantage.
How Doping Turns Pure Semiconductors Into Useful Ones
A pure semiconductor crystal doesn’t conduct very well on its own. To make it useful in electronics, manufacturers deliberately introduce tiny amounts of impurity atoms, a process called doping. The type of impurity determines whether the semiconductor carries current primarily through negatively charged electrons or through positively charged “holes” (spots where an electron is missing).
Adding an element with one extra electron in its outer shell (like phosphorus into silicon) creates an n-type semiconductor with extra free electrons. Adding an element with one fewer electron (like boron into silicon) creates a p-type semiconductor with extra holes. The junction between a p-type and n-type region is the basic building block of diodes, transistors, and solar cells. Boron is one of the most common p-type dopants for silicon, and at high doping concentrations it can reduce silicon’s electrical resistance to extremely low levels, essentially making it behave almost like a metal.
This same principle applies to other materials. Even diamond, which is normally an insulator with a very wide 5.5 eV band gap, can be turned into a semiconductor by doping it with boron or phosphorus. The ability to precisely control conductivity through doping is what makes semiconductors so versatile compared to materials that are permanently conducting or permanently insulating.
Beyond Silicon: Metalloid Compounds in Modern Tech
While pure metalloid elements get the textbook attention, compounds containing metalloids power many of today’s advanced electronics. Silicon carbide (a compound of silicon and carbon) handles high-voltage, high-power applications like electric vehicle inverters because it can withstand much higher temperatures and voltages than pure silicon. Gallium nitride, while not metalloid-based, competes in the same space and offers electron mobility more than five times that of silicon, enabling faster switching in 5G base stations, radar systems, and fast chargers.
Tellurium plays a growing role in thermoelectric devices, which convert heat directly into electricity. Compounds combining tellurium with bismuth or antimony are among the most efficient thermoelectric materials available. Germanium-antimony-tellurium alloys (known as GST) are phase-change materials used in rewritable optical discs and certain types of computer memory. These alloys can switch between amorphous and crystalline states, with the amorphous phase acting as an insulator and the crystalline phase behaving more like a metal, allowing them to store data based on which phase they’re in.
Graphene, a single-atom-thick sheet of carbon, represents another frontier. With carrier mobility up to 200,000 cm²/V·s (compared to about 1,400 for silicon), graphene could enable ultra-thin transistors and flexible electronics. Researchers are exploring hybrid designs that combine graphene or other two-dimensional materials with silicon carbide or gallium nitride to capture the advantages of each.
The Short Answer
Four of the seven standard metalloids (boron, silicon, germanium, and tellurium) are true semiconductors. The others (arsenic, antimony, and carbon in its graphite form) are semimetals. The categories overlap heavily, which is why metalloids and semiconductors are so often mentioned together, but they’re not synonyms. “Metalloid” describes a chemical classification based on a mix of metallic and nonmetallic properties. “Semiconductor” describes a specific electrical behavior defined by a measurable band gap. Most metalloids happen to be semiconductors, and the most important semiconductor (silicon) happens to be a metalloid, but neither category fully contains the other.