Conductors, Insulators, and Semiconductors

Copper carries current, glass and plastic keep it where it belongs, and silicon can be made to do almost anything in between. This article explains why, starting from what the electrons inside each material are doing, and ends with how a few impurity atoms turn silicon into the raw material of every diode and transistor.

No background is assumed: every technical word is explained the first time it appears, and recap links point back to the first article where helpful.

About the analogies. Everyday comparisons appear only in shaded "Analogy" boxes beside the real physics, and each box says where the comparison stops working.

1. A short recap: valence electrons

From What is Electronics:

A conductor lets electric charge move easily, an insulator almost completely blocks it, and a semiconductor sits in between, with a conductivity very sensitive to temperature, light and tiny amounts of impurities.[3] Anything that carries charge through a material, such as a free electron, is a charge carrier.

2. Energy bands, step by step

From energy levels to energy bands

The energy of a single electron is measured in electron volts (eV): 1 eV = 1.60 × 10−19 joule.[4] An electron in a lone atom can only have certain fixed energies, called energy levels.

A crystal is different. A sugar-cube-sized piece of silicon (one cubic centimeter, cm3) holds about 5 × 1022 atoms.[5] Packed that close, the atoms disturb each other's electrons, and each sharp level spreads into a broad range of closely spaced levels called an energy band. Between the bands are energies no electron in the crystal can have.[3]

Why a full band carries no current

To carry current, an electron must pick up a little energy from the applied voltage and move into a slightly higher-energy state. The Pauli exclusion principle says no two electrons can occupy the same state,[7] so in a full band there is nowhere to go: every nearby state is taken.[8] Current needs a partly empty band, or electrons lifted into the empty conduction band.

How much energy does heat provide?

Temperature measures how hard the atoms are jiggling. The kelvin (K) scale starts at absolute zero, about −273 °C,[9] with steps the same size as Celsius degrees, so room temperature is about 300 K. At 300 K the typical thermal energy, written kT, is only about 0.026 eV.[8] Silicon's gap of about 1.12 eV[3] is roughly 43 times that, so only a rare, unusually large jolt frees an electron. Diamond's 5.5 eV gap[3] is about 210 times kT, and at room temperature essentially no electrons cross it.

The Fermi level, in plain terms

Fill the available electron states one at a time from the lowest energy up, as happens at absolute zero. The energy of the highest filled state is the Fermi level.[8] At room temperature the edge blurs only slightly: states well below it are almost certainly full and states well above it almost certainly empty.[8]

Energy bands and the Fermi level in a conductor, a semiconductor and an insulator Three band diagrams with energy increasing upward. Conductor (a metal such as copper): a single band is partly filled, and the dashed Fermi level line runs through the middle of the band, with empty states directly above the filled ones. Semiconductor (silicon): a filled valence band and an empty conduction band separated by a gap of about 1.1 electron volts; the dashed Fermi level line lies in the middle of the gap; a few electrons (dots) have been lifted into the conduction band, leaving holes (open circles) at the top of the valence band. Insulator (diamond): a much larger gap of about 5.5 electron volts with the Fermi level in the middle of it and no electrons in the conduction band. electron energy (higher is up) empty states just above filled states Fermi level one partly filled band Conductor e.g. copper conduction band valence band gap ≈ 1.1 eV Fermi level Semiconductor e.g. pure silicon conduction band valence band gap ≈ 5.5 eV Fermi level Insulator e.g. diamond free electron hole Fermi level filled empty
Figure 1. Energy bands and the Fermi level (dashed). Gaps are not to scale. Based on HyperPhysics and Britannica.[8]​[3]

Analogy: a water line. HyperPhysics calls the electrons a "Fermi sea" and the Fermi level its surface.[8] In a metal the surface is in the middle of an open tank, so the slightest nudge makes ripples. In a semiconductor the tank is full to the brim, with a separate empty tank on a shelf above.

Where it breaks down: each state holds only one electron of each spin, so electrons cannot pile up like water. And in a semiconductor the "water line" sits in the gap where no electrons can be; it is a marker, not a place.

3. Electron-hole pairs

When heat or light breaks a bond in pure silicon, the freed electron jumps to the conduction band and can move. The bond it left is one electron short, and that vacancy is a hole.[3]​[10] Created together, they form an electron-hole pair, and creating one is called generation.

An electron from a neighboring bond can hop into the hole, leaving a new hole behind. So the hole moves opposite to the electrons that fill it and behaves like a positive charge traveling through the crystal. Both free electrons and holes carry current.[3]​[10] Electrons are the faster of the two: in pure silicon they are about three times as mobile as holes.[3]

A flattened silicon crystal with one electron-hole pair A flat grid of fifteen silicon atoms. Each atom is joined to its four neighbors by covalent bonds, and each bond is drawn as a line holding two electrons (dots). In the top row, one bond between the second and third atoms has lost an electron: the empty spot is drawn as an open circle labeled hole, and the electron that escaped is drawn above the crystal, labeled free electron, with a dashed arrow showing where it came from. The free electron and the hole were created together as a pair. Si Si Si Si Si Si Si Si Si Si Si Si Si Si Si free electron (in the conduction band) hole bonding (valence) electron, two per bond hole free electron
Figure 2. A flattened sketch of a silicon crystal (the real one is three-dimensional). Heat has broken one bond, creating an electron-hole pair. Based on PVEducation and Britannica.[1]​[3]

The reverse also happens constantly. A free electron that meets a hole drops back into it and both disappear, which is called recombination. At a steady temperature, generation and recombination balance.[5]​[11] In silicon the energy released ends up mostly as heat.[12]

A pure semiconductor is called intrinsic. In it, free electrons and holes are equal in number, and that number is the intrinsic carrier concentration, ni.[13] At room temperature it is about 2 × 106 per cm3 in gallium arsenide, 1 × 1010 in silicon and 2 × 1013 in germanium.[6] The smaller the gap, the more carriers (germanium's gap is only about 0.66 eV[14]). Even so, silicon's 1010 means only about one broken bond per five trillion atoms,[5] which is why pure silicon conducts poorly.

Analogy: an empty seat in a full row. Empty one seat in a packed theater row and people can shift over one at a time: each person moves left, and the empty seat travels right.

Where it breaks down: nothing is missing from the crystal. The atom is still there, one bonding electron short, so the spot carries a net positive charge, which an empty seat does not. Real electrons and holes are also spread over many atoms rather than sitting in one seat.[11]

4. Conductors

In a metal the Fermi level lies inside a band, so free electrons with empty states beside them are always present, at any temperature.[8] Copper is modeled as having about one free electron per atom.[15] What limits the current is resistance, measured in ohms (Ω) (recap): moving electrons collide with the vibrating atoms and with impurities.[16] To compare materials, engineers use resistivity (ρ, "rho"), a property of the material itself. A wire's resistance is R = ρL/A, where L is its length and A its cross-sectional area. Resistivity is measured in ohm-meters (Ω·m), and lower means a better conductor.[16]

MaterialResistivity at 20 °C (Ω·m)Notes
Silver1.59 × 10−8Lowest in the table.[16]
Copper1.72 × 10−8Nearly as good as silver; the familiar metal of electrical wire.[16]
Gold2.44 × 10−871% of copper's conductivity, but it does not tarnish or corrode. Its biggest industrial use is plating contacts, terminals and printed circuits.[17]
Aluminum2.65 × 10−8Two-thirds of copper's conductivity at one-third the density, so per kilogram it conducts about twice as well. Widely used for electrical conductors.[18]
Graphite (carbon)about 3.5 × 10−5 and upAbout 2,000 times copper's resistivity.[16] Used for electric-motor brushes and furnace electrodes.[19]

Resistivities from OpenStax College Physics 2e, Table 20.1.[16]

Graphite is the odd one out. It is carbon arranged in flat sheets. Each atom bonds to three neighbors, and its fourth valence electron spreads across the whole sheet, free to move within it.[20] Diamond, by contrast, uses all four electrons in bonds and insulates.[20] Because the sheets do not share their mobile electrons, graphite conducts far better along the sheets than across them.[20] Its resistance also falls slightly as it warms, unlike a metal, which is why OpenStax files carbon under semiconductors.[16] In practice it is a usable but mediocre conductor.

5. Insulators

In an insulator every valence electron is firmly held in a bond or a full shell, and the gap is large (diamond's is 5.5 eV).[3] In diamond and quartz all the electrons are held tightly between atoms.[20] Most solids are insulators.[21]

6. Semiconductors

Semiconductors have small gaps, between about 0.25 and 2.5 eV for most important ones.[3] At absolute zero every bond in pure silicon is intact and nothing conducts.[11] At room temperature a few bonds break, and conductivity then changes dramatically with temperature, light and impurities.[3] Some semiconductors are single elements from column IV of the periodic table, now numbered group 14 (Si, Ge). Others are compounds, such as gallium arsenide (GaAs), which pairs gallium from column III with arsenic from column V.[3]

MaterialBand gap (about 300 K)Where you meet it
Germanium (Ge)0.66 eV[14]The main semiconductor of the early 1950s. Largely dropped because its devices leaked current when only moderately warm.[3]
Silicon (Si)1.12 eV[3]​[26]Over 95% of semiconductor products, helped by low leakage and by silicon dioxide, a high-quality insulator easy to build into silicon devices.[3]
Gallium arsenide (GaAs)1.42 to 1.43 eV[3]​[26]Light-emitting and radio-frequency devices.[3] An early LED material.[12]
Silicon carbide (4H-SiC)3.26 eV[26]Power electronics that run at much higher temperatures, voltages and frequencies than silicon.[27]
Gallium nitride (GaN)3.4 to 3.45 eV[27]​[26]LEDs and power electronics.[27] The first bright blue LED (1994) was GaN.[12]

Published band gaps vary slightly with temperature and method. SiC comes in several crystal forms; ORNL lists 4H at 3.26 eV and 6H at 3.03 eV.[26]

SiC and GaN are wide-bandgap semiconductors: their gaps are far larger than silicon's,[26] so heat generates far fewer stray carriers and they keep working where silicon would fail.[27] The gap also sets the color of light a semiconductor can emit. When an electron falls back across the gap, the energy can leave as a photon (a particle of light). Red light needs about 1.77 eV per photon and violet about 3.1 eV,[12] which is why blue LEDs needed a wide-gap material like GaN.

7. Side by side

ConductorsSemiconductorsInsulators
Valence electronsShared by the whole crystal, free to moveAll in bonds when cold; a few freed by heat or lightTightly bound
Band pictureFermi level inside a band; no gapSmall gap: about 0.25 to 2.5 eV for most, up to about 3.4 eV for wide-gap typesLarge gap (diamond about 5.5 eV)
Charge carriersElectrons, about one per atom in copperElectrons and holes, set by temperature and dopingAlmost none
Resistivity at 20 °CAbout 10−8 to 10−6 Ω·m (silver to nichrome)About 10−3 to 103 Ω·m for Ge and Si, depending on impuritiesAbout 108 to 1018 Ω·m
When heatedResistance risesResistance falls (pure material)Resistance falls but stays huge
ExamplesSilver, copper, gold, aluminum; graphiteSi, Ge, GaAs, SiC, GaNGlass, rubber, plastics, ceramics, air, dry wood

Sources: OpenStax (resistivities),[16] Britannica and HyperPhysics (bands),[3]​[8] ORNL (wide-gap values).[26] The boundaries between classes are not sharp.

Resistivity of common materials on a logarithmic scale A horizontal chart of resistivity at about 20 degrees Celsius, on a logarithmic scale running from 10 to the minus 8 to 10 to the 18 ohm-meters. Metals sit at the far left: silver, copper and aluminum near 2 times 10 to the minus 8, nichrome near 10 to the minus 6. Graphite (carbon) spans roughly 3.5 times 10 to the minus 5 to 6 times 10 to the minus 4. Germanium spans 10 to the minus 3 to 0.6 and silicon 0.1 to 2300 ohm-meters depending on impurities. Insulators sit far to the right: dry wood 10 to the 8 to 10 to the 11, glass 10 to the 9 to 10 to the 14, mica 10 to the 11 to 10 to the 15, hard rubber 10 to the 13 to 10 to the 16, and fused quartz about 7.5 times 10 to the 17. The whole range spans about 25 powers of ten. metals semiconductors insulators Silver Copper Aluminum Nichrome Graphite (carbon) Germanium Silicon Wood (dry) Glass Mica Hard rubber Fused quartz 10−8 10−6 10−4 10−2 100 102 104 106 108 1010 1012 1014 1016 1018 resistivity in ohm-meters (log scale: each tick is 100 times the one before)
Figure 3. Resistivity at about 20 °C on a logarithmic scale; each tick is 100 times the one before. Semiconductor bars run from heavily doped to pure. From silver to fused quartz is a factor of more than 1025. Data from OpenStax Table 20.1.[16]

8. Temperature changes everything

Metals: hotter means more resistance

For modest changes, R = R0(1 + αΔT). Here R0 is the starting resistance, ΔT the temperature change and α (alpha) the temperature coefficient of resistivity. For copper α is about +0.0039 per °C,[16] so a copper winding warming from 20 °C to 70 °C gains about 20% in resistance. The number of free electrons barely changes, but the atoms vibrate harder, so the electrons collide more often.[16]

Semiconductors: hotter means less resistance

In a pure semiconductor, heat breaks more bonds, and the extra carriers far outweigh the extra collisions.[16] The carrier count grows exponentially with temperature.[5] It roughly doubles for every 13 °C rise in germanium and every 8 °C in silicon.[8] OpenStax gives α as about −0.07 per °C for pure silicon and −0.05 for germanium, where the minus sign means resistance falls as temperature rises.[16]

Doped semiconductors (section 10) can behave differently. Their carrier count is set mainly by the impurities,[28] while mobility falls with temperature,[3] so their resistance can rise instead. Silicon PTC sensors called silistors rely on this.[29]

Insulators: same physics, tiny numbers

Heat lifts electrons across an insulator's gap too, more of them the hotter it gets,[8] but with a gap of several eV the number stays vanishingly small. In practice moisture is often the bigger threat, as damp wood shows.[25]

Putting it to work: thermistors

A thermistor is a resistor designed to change resistance strongly with temperature, so its resistance reveals its temperature.[16] There are two types:[29]

Superconductors: resistance that vanishes

In 1911 Heike Kamerlingh Onnes cooled mercury with liquid helium. Near 4.2 K (−269 °C) its resistance dropped abruptly to zero.[7] This is superconductivity, and it begins below a material's critical temperature. A superconductor is not just a very good conductor. It is a different state of the material, which also expels magnetic fields (the Meissner effect, first seen in 1933).[9] Currents in superconducting loops have kept flowing for years with no voltage source.[7] In the standard theory (BCS, 1957), electrons pair up through the vibrating crystal and move in a coordinated way that ordinary collisions cannot disrupt.[9]​[7] Oddly, copper, silver and gold never superconduct.[7]

In 1986 Georg Bednorz and Alex Müller found superconductivity in a copper-oxide ceramic. Within months others passed 90 K, cold enough to need only liquid nitrogen rather than liquid helium.[9] One example is YBCO (yttrium barium copper oxide), with a critical temperature of 92 K.[7] Room-temperature superconductivity remains a goal, not a reality. Today superconductors serve where the cooling pays off: MRI and particle-accelerator magnets, and ultra-sensitive magnetic sensors called SQUIDs.[7]

9. When insulators fail: breakdown

Push hard enough and any insulator conducts. The push that matters is the electric field, the force per unit of charge at a given point, measured in volts per meter (V/m). Across a gap it is roughly the voltage divided by the gap width. The largest field a material withstands before it "begins to break down and conduct" is its dielectric strength.[22] (A dielectric is an insulator used in an electric field, as in a capacitor.)

MaterialDielectric strength (V/m)Per millimeter
Air3 × 1063 kV/mm
Neoprene rubber12 × 10612 kV/mm
Pyrex glass14 × 10614 kV/mm
Paper16 × 10616 kV/mm
Teflon60 × 10660 kV/mm

From OpenStax College Physics 2e, Table 19.1.[22] Real values vary with thickness, purity, humidity and electrode shape.

A 1 mm air gap breaks down at about 3,000 V, while 1 mm of Teflon holds off about 60,000 V. Because the limit is a field, thinner insulation fails at a lower voltage.[22]

How air breaks down. Air always holds a few stray free electrons. A strong field accelerates them, and one that gains roughly 10 eV before hitting a molecule knocks another electron loose. This is impact ionization, and it leaves behind an ion, a molecule with a net charge.[24] Each freed electron is accelerated in turn and frees more, so their number multiplies. The air becomes a plasma, a glowing gas full of free charges, and you get a spark or arc, as in a static shock, a spark plug or an arc welder.[24]

Lightning is the same thing on a giant scale. Collisions between ice and hail in a storm separate charge. Air insulates between the charged regions until the difference grows too great, then breaks down in a rapid discharge that can heat the air to about 50,000 °F, and the expanding air is the thunder.[30] Dielectric strength is also why every capacitor and cable carries a voltage rating.

10. Doping: making n-type and p-type

Pure silicon is rarely used as it is. It is almost always doped,[13] meaning tiny, controlled amounts of chosen impurity atoms (dopants) are added. These take the place of silicon atoms in the crystal.[6] The effect is large: about 10 boron atoms per million silicon atoms can raise conductivity a thousandfold.[3] Doping is older than the transistor. The crystal rectifiers in World War II radar receivers were silicon or germanium doped with traces of boron or phosphorus.[31]

The choice depends on valence electrons. Periodic-table columns, called groups, are numbered 1 to 18. Older books use III, IV and V for groups 13, 14 and 15.[18]​[3] Silicon and germanium are in group 14.

n-type: donors add electrons

A group 15 atom, phosphorus (P), arsenic (As) or antimony (Sb), has five valence electrons. Four form the usual bonds, and the fifth is only weakly held.[32]​[3] Freeing it takes only about 0.045 eV for P, 0.054 for As and 0.039 for Sb, against 1.12 eV to break a silicon bond, so at room temperature practically all of them let go.[6] These atoms are donors, and the result is n-type silicon ("n" for the negative carriers).[3] Each donor adds a level just below the conduction band,[33] and the Fermi level moves up.[8]

p-type: acceptors create holes

A group 13 atom, boron (B), gallium (Ga) or indium (In) (aluminum also works), has only three valence electrons, so one bond is an electron short: a hole.[32]​[33] A neighboring electron moves in for about 0.045 eV (boron) to 0.072 eV (gallium). Indium needs about 0.16 eV,[6] so in silicon fewer indium atoms are active at room temperature. The hole then wanders off and carries current. These atoms are acceptors, giving p-type silicon ("p" for positive holes).[3] Each adds a level just above the valence band,[33] and the Fermi level moves down.[8] The same donors and acceptors work in germanium.[33]

n-type silicon doped with phosphorus and p-type silicon doped with boron Two small flattened silicon crystals. Left, n-type: the center atom is phosphorus, which has five valence electrons. Four of them form the usual four bonds with the neighboring silicon atoms, and the fifth, drawn as a larger dot just outside the bonds, is only weakly held and easily becomes a free electron, leaving the phosphorus as a fixed positive ion. Right, p-type: the center atom is boron, which has only three valence electrons, so one of its four bonds is missing an electron. That gap, drawn as an open circle, is a hole that neighboring electrons can hop into, leaving the boron as a fixed negative ion. n-type: phosphorus donor p-type: boron acceptor Si Si Si Si P Si Si Si Si Si Si Si Si B Si Si Si Si The 5th electron is held by only ≈0.045 eV, so heat frees it: it becomes a free electron and the P atom becomes a fixed + ion. One bond is an electron short: a hole. A neighbor electron fills it, the hole moves on, and the B atom becomes a fixed − ion.
Figure 4. Left: a phosphorus donor's fifth electron becomes free. Right: a boron acceptor leaves a bond one electron short, which is a hole. Based on Britannica, PVEducation and Stanford EE 116.[3]​[32]​[6]

Misconception: n-type is not negatively charged. Each donor that gives up its electron is left as a fixed positive ion, so n-type silicon is neutral overall. Likewise every hole in p-type silicon is balanced by a fixed negative acceptor ion. "N" and "p" name the mobile carriers, not a net charge.[34]

How much dopant? Useful concentrations in silicon run from about 1015 to 1020 atoms per cm3, so even the heaviest doping replaces under 1% of the atoms.[6]​[11] Against 5 × 1022 silicon atoms per cm3,[5] that is about one dopant atom per 50 million (1015) down to one per 500 (1020). Even 1015 donors gives 100,000 times the free electrons of pure silicon, which is why silicon's resistivity spans 2,300 Ω·m pure to 0.1 Ω·m and below when doped.[16]

11. Majority and minority carriers

In doped material the more plentiful carrier is the majority carrier and the scarcer one the minority carrier.[32] In n-type, electrons are the majority and holes the minority. In p-type it is the reverse. Heat still makes pairs, so minority carriers never vanish, but they are rare. At a steady temperature the law of mass action holds: n·p = ni2, where n is the free-electron and p the hole concentration. Adding more of one carrier makes the other rarer, because the extras recombine with it.[28]

p-type silicon, 1017 acceptors per cm3:
holes p ≈ 1017; electrons n = (1010)2 ÷ 1017 = 103 per cm3

That MIT example[11] gives a ratio of a hundred trillion to one. n-type works the same way, with the roles swapped,[11] and heavier doping means fewer minority carriers.[28] Minority carriers matter a great deal in diodes, transistors and solar cells.

Intrinsic (pure)n-typep-type
Added atomsNoneDonors (group 15): P, As, SbAcceptors (group 13): B, Ga, In, also Al
Majority / minorityEqual numbersElectrons / holesHoles / electrons
Fixed ionsNonePositive donor ionsNegative acceptor ions
Fermi levelNear mid-gapToward the conduction bandToward the valence band

Sources: PVEducation, HyperPhysics and Stanford EE 116.[32]​[8]​[6]

12. Where this leads: the PN junction

Make p-type and n-type regions side by side in one crystal and the boundary is a PN junction, first identified in silicon in 1940.[35] Carriers spread from where they are crowded to where they are scarce, a process called diffusion. Electrons diffuse from the n side into the p side and recombine with holes. That leaves the fixed ions unbalanced: positive on the n side, negative on the p side. Their charge creates an electric field that sweeps free carriers out and stops further migration. This thin, carrier-free zone is the depletion region, and its field creates a built-in voltage.[34]​[36]

A PN junction and its depletion region A single crystal whose left half is p-type, full of holes (open circles), and whose right half is n-type, full of free electrons (dots). Along the boundary is a narrow shaded strip, the depletion region, that contains no free carriers, only fixed ions: negative acceptor ions on the p side and positive donor ions on the n side. An arrow beneath the strip points from the n side toward the p side, showing the direction of the built-in electric field created by those fixed charges. − + − + − + − + p-type: holes are the majority carriers n-type: electrons are the majority carriers built-in electric field depletion region: fixed ions, almost no free carriers hole free electron −fixed acceptor ion +fixed donor ion
Figure 5. A PN junction, with a depletion region of fixed ions and a built-in electric field. Based on PVEducation and HyperPhysics.[34]​[36]

Coming up in this section: 3 Resistors (resistivity and temperature coefficients at work), 4 Capacitors (insulators as dielectrics and voltage ratings), 5 Coils (Inductors), 6 Diodes (the PN junction in action), 7 Transistors, and 8 Transformers.

The whole story in brief

Sources

All links were checked in October 2026. External links open in a new tab.

  1. C. Honsberg and S. Bowden, PVEducation, "Semiconductor Structure": https://www.pveducation.org/pvcdrom/pn-junctions/semiconductor-structure
  2. Chemistry LibreTexts, "Metallic Bonding" (Jim Clark and others): https://chem.libretexts.org/Bookshelves/Physical_and_Theoretical_Chemistry_Textbook_Maps/Supplemental_Modules_(Physical_and_Theoretical_Chemistry)/Chemical_Bonding/Fundamentals_of_Chemical_Bonding/Metallic_Bonding
  3. Encyclopaedia Britannica, "Semiconductor" (sections "Semiconductor materials" and "Electronic properties"): https://www.britannica.com/science/semiconductor
  4. OpenStax, College Physics 2e, section 19.1 "Electric Potential Energy: Potential Difference" (electron volt): https://openstax.org/books/college-physics-2e/pages/19-1-electric-potential-energy-potential-difference
  5. C. Fonstad, MIT OpenCourseWare 6.012 Microelectronic Devices and Circuits (Fall 2009), Lecture 1 slides (PDF): https://ocw.mit.edu/courses/6-012-microelectronic-devices-and-circuits-fall-2009/a37c73e7fcff4bcb5c5f1177e12d0891_MIT6_012F09_lec01.pdf
  6. E. Pop, Stanford EE 116, Lectures 6-8 "Doping, Fermi Level, Density of States" (PDF): https://poplab.stanford.edu/files/ee116/L6toL8_DopingFermi.pdf
  7. OpenStax, University Physics Volume 2, section 9.6 "Superconductors": https://openstax.org/books/university-physics-volume-2/pages/9-6-superconductors
  8. HyperPhysics (Georgia State University), "Fermi Level and Fermi Function": http://hyperphysics.phy-astr.gsu.edu/hbase/Solids/Fermi.html
  9. NobelPrize.org, "Press release: The 1987 Nobel Prize in Physics" (Bednorz and Müller, with background on superconductivity since 1911): https://www.nobelprize.org/prizes/physics/1987/press-release/
  10. HyperPhysics (Georgia State University), "Intrinsic Semiconductors": http://hyperphysics.phy-astr.gsu.edu/hbase/Solids/intrin.html
  11. MIT 6.012 Electronic Devices and Circuits (Spring 2007), Lecture 2 "Semiconductor physics" slides (PDF): https://web.mit.edu/6.012/www/SP07-L2.pdf
  12. HyperPhysics (Georgia State University), "Light Emitting Diodes": http://hyperphysics.phy-astr.gsu.edu/hbase/Electronic/led.html
  13. PVEducation, "Intrinsic Carrier Concentration": https://www.pveducation.org/pvcdrom/pn-junctions/intrinsic-carrier-concentration
  14. HyperPhysics (Georgia State University), "Semiconductor Band Gaps" (data from Kittel, Introduction to Solid State Physics): http://hyperphysics.phy-astr.gsu.edu/hbase/Tables/Semgap.html
  15. OpenStax, College Physics 2e, section 20.1 "Current": https://openstax.org/books/college-physics-2e/pages/20-1-current
  16. OpenStax, College Physics 2e, section 20.3 "Resistance and Resistivity" (Table 20.1, resistivities and temperature coefficients at 20 °C): https://openstax.org/books/college-physics-2e/pages/20-3-resistance-and-resistivity
  17. Encyclopaedia Britannica, "Gold" (section "Properties, occurrences, and uses"): https://www.britannica.com/science/gold-chemical-element/Properties-occurrences-and-uses
  18. Encyclopaedia Britannica, "Aluminum" (section "Uses and properties"): https://www.britannica.com/science/aluminum
  19. Encyclopaedia Britannica, "Graphite": https://www.britannica.com/science/graphite-carbon
  20. Chemistry LibreTexts, "Graphite and Diamond: Structure and Properties" (Map: Inorganic Chemistry, Housecroft, 14.4A): https://chem.libretexts.org/Bookshelves/Inorganic_Chemistry/Map%3A_Inorganic_Chemistry_(Housecroft)/14%3A_The_Group_14_Elements/14.04%3A_Allotropes_of_Carbon/14.4A%3A_Graphite_and_Diamond_-_Structure_and_Properties
  21. HyperPhysics (R. Nave, Georgia State University), "Band Theory of Solids": http://hyperphysics.phy-astr.gsu.edu/hbase/Solids/band.html
  22. OpenStax, College Physics 2e, section 19.5 "Capacitors and Dielectrics" (Table 19.1, dielectric strengths): https://openstax.org/books/college-physics-2e/pages/19-5-capacitors-and-dielectrics
  23. The American Ceramic Society, "Ceramics and Glass in Electrical and Electronic Applications": https://ceramics.org/about/what-are-ceramics/electrical-and-electronic-appliances/
  24. MIT OpenCourseWare 8.02X Physics II: Electricity and Magnetism (Spring 2005), "Experiment EB: Electrical Breakdown" (PDF): https://ocw.mit.edu/courses/8-02x-physics-ii-electricity-magnetism-with-an-experimental-focus-spring-2005/ecb0db9176aee59873ea4e5d0acfd96f_experiment_eb.pdf
  25. USDA Forest Service, Dry Kiln Operator's Manual (Agriculture Handbook 188), chapter 1 "Properties of Wood Related to Drying", section "Electrical Properties" (PDF; copy hosted by the National Wooden Pallet & Container Association): https://www.woodencrates.org/standards/AH188-KD-chapter-01.pdf
  26. B. Ozpineci and L. M. Tolbert, "Comparison of Wide-Bandgap Semiconductors for Power Electronics Applications", Oak Ridge National Laboratory report ORNL/TM-2003/257 (PDF): https://info.ornl.gov/sites/publications/Files/Pub57423.pdf
  27. U.S. Department of Energy, Advanced Manufacturing Office, "Wide Bandgap Semiconductors: Pursuing the Promise" (fact sheet, PDF): https://www.energy.gov/sites/default/files/2013/12/f5/wide_bandgap_semiconductors_factsheet.pdf
  28. PVEducation, "Equilibrium Carrier Concentration": https://www.pveducation.org/pvcdrom/pn-junctions/equilibrium-carrier-concentration
  29. Texas Instruments, application report SNOAA12, "Methods to Reduce Thermistor Linearization Error, Memory, and Power Requirements Over Wide Operating Temperature Ranges" (2018; introduction covers NTC and PTC types; PDF): https://www.ti.com/lit/an/snoaa12/snoaa12.pdf
  30. NOAA National Weather Service, "Understanding Lightning Science": https://www.weather.gov/safety/lightning-science-overview
  31. Encyclopaedia Britannica, "Transistor" (Michael Riordan): https://www.britannica.com/technology/transistor
  32. PVEducation, "Doping": https://www.pveducation.org/pvcdrom/pn-junctions/doping
  33. HyperPhysics (Georgia State University), "The Doping of Semiconductors": http://hyperphysics.phy-astr.gsu.edu/hbase/Solids/dope.html
  34. PVEducation, "Formation of a PN Junction": https://www.pveducation.org/pvcdrom/pn-junctions/formation-of-a-pn-junction
  35. Computer History Museum, The Silicon Engine timeline (1940: Discovery of the p-n Junction): https://www.computerhistory.org/siliconengine/timeline/
  36. HyperPhysics (Georgia State University), "P-N Junction": http://hyperphysics.phy-astr.gsu.edu/hbase/Solids/pnjun.html
  37. PVEducation, "Solar Cell Structure": https://www.pveducation.org/pvcdrom/solar-cell-operation/solar-cell-structure
  38. PVEducation, "Collection Probability": https://www.pveducation.org/pvcdrom/solar-cell-operation/collection-probability