Band gap and its meaning for conductivity: Worked examples

Before anything, let us fix every symbol we will reuse constantly, so no symbol is unearned.

Definition The symbols and constants we need

Boltzmann constant $k_{B} = 8.617 \times 1 0^{- 5} eV/K$ . Think of $k_{B} T$ as "how big a random thermal energy kick is at temperature $T$ ." It has units of energy (eV).
At room temperature $T = 300$ K: $k_{B} T \approx 0.0259 eV \approx \frac{1}{40} eV$ .
$E_{g}$ = band gap in eV (from the parent note).
$E_{c}$ = the conduction-band edge: the lowest energy a mobile conduction electron can have. An electron sitting at the mid-gap Fermi level must climb up to $E_{c}$ (a rise of $E_{g} /2$ ) before it can carry current. This is the energy in the exponent.
$n$ = intrinsic carrier concentration: the number of mobile electrons per cubic metre that have been thermally lifted into the conduction band (each also leaves a mobile "hole" behind). This is the quantity the master formula predicts.
$σ$ = electrical conductivity: how easily the material carries current for a given field (units $S/m$ , the inverse of resistivity). Because current needs carriers, $σ \propto n$ — so anything that multiplies $n$ multiplies $σ$ .
$e^{x}$ = the exponential. We use it because the chance of a random thermal kick being big enough to lift an electron over an energy barrier falls off exponentially with (barrier ÷ kick size) — that is the Boltzmann tail the parent derived. No other function does this.

The scenario matrix

Every worked example below is tagged with the cell of this table it covers. Together they fill the whole grid.

#	Cell class	What makes it different	Example
A	Zero gap (metal, degenerate)	$E_{g} = 0$ , formula gives $n \propto 1$ — the exponential switches off	Ex 1
B	Small gap (semiconductor, baseline)	plug a normal Si-scale number at 300 K	Ex 2
C	Large gap (insulator, limiting)	$E_{g}$ huge → answer underflows to "practically zero"	Ex 3
D	Temperature raised (ratio in $T$ )	same material, two temperatures	Ex 4
E	Temperature → limits ( $T \to 0, T \to \infty$ )	degenerate/limiting behaviour of the exponent	Ex 5
F	Two materials, same $T$ (ratio in $E_{g}$ )	compare Ge vs Si carrier counts	Ex 6
G	Real-world word problem	"sensor gets warmer, resistance drops by how much?"	Ex 7
H	Metal vs semiconductor sign twist	which way does $σ$ move with $T$ ? opposite signs	Ex 8
I	Exam trap: the missing 2	forgetting $E_{g} /2$ vs $E_{g}$	Ex 9
J	Back-solve (invert the exponential)	given a carrier ratio, find $E_{g}$	Ex 10

Ex 1 — Cell A: the zero-gap (metal) degenerate case

Forecast: guess before reading — is the exponential big, small, or exactly $1$ ?

Substitute $E_{g} = 0$ . The exponent becomes $- \frac{0}{2 k _{B} T} = 0$ . Why this step? Always test the degenerate input first — it usually collapses the formula and reveals its skeleton.
Evaluate. $e^{0} = 1$ , for any temperature $T$ . Why this step? $e^{0} = 1$ is the anchor value of the exponential; the temperature dependence has completely vanished.
Interpret. With no gap to climb, the "thermal filter" $e^{- E_{g} /2 k_{B} T}$ stops filtering. The number of conduction carriers is set purely by how many states overlap, not by temperature — exactly the metal behaviour from the parent's "bands merge" picture.

Verify: Units check — the exponent must be dimensionless: $\frac{eV}{eV/K \cdot K} = \frac{eV}{eV} = 1$ ✓. And $e^{0} = 1$ exactly. A metal's carrier count being $T$ -independent (the exponential $= 1$ ) matches Temperature dependence of resistivity in metals, where the $T$ effect comes from scattering, not carrier creation.

Ex 2 — Cell B: the small-gap semiconductor baseline

Forecast: the parent said "small but nonzero." Guess the order of magnitude: $1 0^{- 3}$ ? $1 0^{- 10}$ ? $1 0^{- 30}$ ?

Compute the exponent numerator. $\frac{E _{g}}{2} = \frac{1.12}{2} = 0.56$ eV. Why this step? This is the actual climb: from the mid-gap Fermi level up to the conduction-band edge $E_{c}$ (defined above). Half the gap — the parent's "half the gap for the Fermi trap."
Compute $k_{B} T$ . $8.617 \times 1 0^{- 5} \times 300 = 0.02585$ eV. Why this step? This is the size of a typical thermal kick. We compare the climb to the kick.
Form the ratio. $\frac{0.56}{0.02585} \approx 21.66$ . Why this step? The climb is ~22× bigger than a typical kick — that's a steep hill.
Exponentiate. $e^{- 21.66} \approx 3.9 \times 1 0^{- 10}$ . Why this step? This is the fraction of the maximum possible carriers that actually make it across.

Verify: Matches the parent's stated " $\approx 5 \times 1 0^{- 10}$ " (same order of magnitude; the parent rounded $E_{g} /2 k_{B} T \approx 21.5$ ). Tiny but not zero → silicon conducts a little → it's a semiconductor, not an insulator. ✓

Ex 3 — Cell C: the large-gap insulator limit

Forecast: the gap is ~5× silicon's. Is the carrier factor 5× smaller? 100× smaller? Or unimaginably smaller?

Climb. $\frac{E _{g}}{2} = \frac{5.5}{2} = 2.75$ eV. Why this step? Same "half-gap" rule as Ex 2 — consistency across cells.
Ratio to $k_{B} T$ . $\frac{2.75}{0.02585} \approx 106.4$ . Why this step? Over 100 thermal kicks tall. The exponential punishes this savagely.
Exponentiate. $e^{- 106.4} \approx 6 \times 1 0^{- 47}$ . Why this step? Reveals why insulators are insulators — not "a bit fewer" carriers but astronomically fewer.
Ratio vs silicon. $\frac{e ^{- 106.4}}{e ^{- 21.66}} = e^{- (106.4 - 21.66)} = e^{- 84.7} \approx 1.6 \times 1 0^{- 37}$ . Why this step? A 4.9× larger gap gives ~ $1 0^{37}$ × fewer carriers — the exponential turns a modest input change into a colossal output change.

Verify: The parent said diamond gives $e^{- 106}$ , "utterly negligible." Our $106.4$ agrees ✓. The huge ratio explains the parent's "spectrum, not a wall": nothing qualitatively new happens at diamond, the number just falls off a cliff.

Figure — Band gap and its meaning for conductivity

Ex 4 — Cell D: same material, temperature raised

Forecast: the parent's temperature example is your hint — is the increase ~2×, ~20×, or ~200×?

Collapse constants into a temperature scale. Define $\frac{E _{g}}{2 k _{B}} = \frac{1.12}{2 \times 8.617 \times 1 0 ^{- 5}} \approx 6499$ K. Why this step? The exponent is $- \frac{E _{g}}{2 k _{B}} \cdot \frac{1}{T}$ ; pulling out the constant leaves one clean number in kelvin.
Write the ratio. $\frac{n ( 400 )}{n ( 300 )} = \frac{e ^{- 6499/400}}{e ^{- 6499/300}} = e^{6499 (\frac{1}{300} - \frac{1}{400})}$ . Why this step? Dividing two exponentials subtracts exponents — the messy absolute values cancel and only the change survives.
Evaluate the bracket. $\frac{1}{300} - \frac{1}{400} = \frac{400 - 300}{120000} = \frac{100}{120000} = 8.333 \times 1 0^{- 4} K^{- 1}$ . Why this step? This is the "temperature leverage" — how much room you gained.
Multiply and exponentiate. $6499 \times 8.333 \times 1 0^{- 4} = 5.416$ , so ratio $= e^{5.416} \approx 225$ .

Verify: Matches the parent's $e^{5.42} \approx 225$ ✓. A 33% temperature rise multiplies carriers ~225× — a hallmark of semiconductors. Connects to Conductivity, mobility and drift current and Doping and carrier concentration.

Ex 5 — Cell E: the temperature limits ( $T \to 0$ and $T \to \infty$ )

Forecast: guess both endpoints before computing.

Case $T \to 0^{+}$ . The exponent $- \frac{E _{g}}{2 k _{B} T} \to - \infty$ (dividing a positive number by a tiny positive $T$ blows up, negative sign flips it to $- \infty$ ). Why this step? Test the cold limit — the exponential's argument runs off to $- \infty$ .
Evaluate. $e^{- \infty} = 0$ . So $n \to 0$ . Why this step? At absolute zero there are no thermal kicks, so nobody crosses the gap — an intrinsic semiconductor becomes a perfect insulator. Correct physical limit.
Case $T \to \infty$ . The exponent $- \frac{E _{g}}{2 k _{B} T} \to 0^{-}$ . Why this step? A huge kick makes even a real gap look negligible.
Evaluate. $e^{0} = 1$ . So $n$ saturates at its maximum (every available state can be reached), behaving like the zero-gap metal of Ex 1.

Verify: The two limits bracket everything: cold ⇒ factor $0$ (insulating), hot ⇒ factor $1$ (metal-like). Ex 1 (metal) is literally the $T \to \infty$ shape at any $T$ . Consistent ✓.

Ex 6 — Cell F: two materials, same temperature

Forecast: Ge has the smaller gap — so more or fewer carriers? By roughly how much?

Difference of the half-gaps. $\frac{E _{g, Si} - E _{g, Ge}}{2} = \frac{1.12 - 0.66}{2} = \frac{0.46}{2} = 0.23$ eV. Why this step? The ratio of two Boltzmann factors depends only on the difference of climbs (exponents subtract).
Divide by $k_{B} T$ . $\frac{0.23}{0.02585} \approx 8.90$ . Why this step? Converts the energy difference into "how many thermal kicks apart" the two materials sit.
Exponentiate. $\frac{n _{Ge}}{n _{Si}} = e^{+ 8.90} \approx 7.3 \times 1 0^{3}$ . Why this step? Ge's smaller gap → the positive exponent → thousands of times more carriers.

Verify: Sign check: smaller gap ⇒ more carriers ⇒ ratio $> 1$ ✓. A 0.46 eV smaller gap buys ~ $7000 \times$ the carriers — this is exactly why early transistors used germanium. Ties to Direct vs indirect band gap and Intrinsic vs extrinsic semiconductors.

Ex 7 — Cell G: real-world word problem (a thermistor)

Forecast: a mere 10 K rise — does resistance change by ~1%, ~10%, or ~50%?

Use the same temperature scale. $\frac{E _{g}}{2 k _{B}} = 6499$ K (from Ex 4). Why this step? Reuse the collapsed constant — no need to recompute.
Resistance ratio. $\frac{R ( 310 )}{R ( 300 )} = \frac{e ^{6499/310}}{e ^{6499/300}} = e^{6499 (\frac{1}{310} - \frac{1}{300})}$ . Why this step? $R \propto e^{+ E_{g} /2 k_{B} T}$ (inverse of $n$ ), so the sign in the exponent is $+$ ; dividing again cancels the huge absolute value.
Bracket. $\frac{1}{310} - \frac{1}{300} = \frac{300 - 310}{93000} = \frac{- 10}{93000} = - 1.0753 \times 1 0^{- 4}$ . Why this step? Negative because we raised $T$ → $1/ T$ dropped → resistance drops.
Evaluate. $6499 \times (- 1.0753 \times 1 0^{- 4}) = - 0.6989$ , so $\frac{R ( 310 )}{R ( 300 )} = e^{- 0.6989} \approx 0.497$ . Why this step? Resistance falls to about half.
Percentage drop. $1 - 0.497 = 0.503 \Rightarrow\approx 50.3%$ .

Verify: Units of exponent dimensionless ✓. A 10 K warming halving the resistance is the extreme sensitivity that makes semiconductors excellent thermometers — and dangerous if not compensated (thermal runaway). Reasonableness ✓.

Ex 8 — Cell H: metal vs semiconductor, the sign twist

Forecast: do both go up, both down, or opposite?

Silicon (semiconductor). $σ \propto n \propto e^{- E_{g} / (2 k_{B} T)}$ ; raising $T$ raises the exponent toward $0$ ⇒ $n ↑$ ⇒ $σ ↑$ . Why this step? Here heat creates carriers — the dominant effect.
Copper (metal). $E_{g} = 0$ , so the carrier count is $T$ -independent (Ex 1). But higher $T$ ⇒ more lattice vibration ⇒ more electron scattering ⇒ mobility $μ ↓$ ⇒ $σ ↓$ . Why this step? With no carriers to gain, the only lever left is scattering, which hurts.
Conclusion. Semiconductor: $σ$ increases. Metal: $σ$ decreases. Opposite signs.

Verify: This is the parent's mistake #4 fixed. No formula contradiction: the metal's $e^{- 0/2 k_{B} T} = 1$ removes the carrier-creation term, leaving only the scattering term from Temperature dependence of resistivity in metals. Sign logic consistent ✓.

Ex 9 — Cell I: the exam trap (the missing 2)

Forecast: will forgetting the 2 make them a little off or wildly off?

Wrong exponent. $\frac{E _{g}}{k _{B} T} = \frac{1.12}{0.02585} \approx 43.33$ ⇒ wrong factor $e^{- 43.33} \approx 1.5 \times 1 0^{- 19}$ . Why this step? Show exactly the number the mistake produces.
Correct factor (Ex 2). $e^{- 21.66} \approx 3.9 \times 1 0^{- 10}$ . Why this step? Direct comparison target.
Ratio of error. $\frac{e ^{- 21.66}}{e ^{- 43.33}} = e^{43.33 - 21.66} = e^{21.67} \approx 2.6 \times 1 0^{9}$ . Why this step? Reveals the mistake understates carriers by ~2.6 billion times — a catastrophic, not cosmetic, error.

Verify: The correct factor is the square root of the wrong one: $(e^{- 43.33})^{1/2} = e^{- 21.66}$ ✓. That's the mathematical fingerprint of the missing 2 — halving the exponent square-roots the result. Reinforces "Half the gap for the Fermi trap." See Fermi level and Fermi-Dirac distribution for why $E_{F}$ sits mid-gap.

Ex 10 — Cell J: back-solve (invert the exponential)

Forecast: we run every previous example backwards. Expect a gap somewhere in the semiconductor range (0.1–3 eV)?

Write the measured ratio. $\frac{n ( 350 )}{n ( 300 )} = e^{\frac{E _{g}}{2 k _{B}} (\frac{1}{300} - \frac{1}{350})} = 100$ . Why this step? Same ratio structure as Ex 4, but now $E_{g}$ is the unknown.
Take natural log of both sides. $\frac{E _{g}}{2 k _{B}} (\frac{1}{300} - \frac{1}{350}) = ln 100 = 4.6052$ . Why this step? $ln$ is the inverse of $e^{x}$ — it's the tool that "undoes" the exponential to expose $E_{g}$ . That's precisely why we reach for it here.
Evaluate the bracket. $\frac{1}{300} - \frac{1}{350} = \frac{350 - 300}{105000} = \frac{50}{105000} = 4.7619 \times 1 0^{- 4} K^{- 1}$ . Why this step? The temperature leverage, as in Ex 4.
Solve for $E_{g}$ . $\frac{E _{g}}{2 k _{B}} = \frac{4.6052}{4.7619 \times 1 0 ^{- 4}} = 9671 K$ , so $E_{g} = 2 k_{B} \times 9671 = 2 \times 8.617 \times 1 0^{- 5} \times 9671 \approx 1.667$ eV.

Verify: $E_{g} \approx 1.67$ eV lands squarely in the semiconductor band (0.1–3 eV) ✓. Plug back: $e^{9671 \times 4.7619 \times 1 0^{- 4}} = e^{4.605} = 100.0$ ✓. Self-consistent inversion of the master formula.

Recall One-line summary of the whole matrix

The exponential $e^{- E_{g} / (2 k_{B} T)}$ has two knobs: the gap (Ex 1,2,3,6,9 — bigger gap, fewer carriers, catastrophically) and the temperature (Ex 4,5,7,10 — hotter, more carriers, exponentially). Metals (Ex 1,8) switch the gap knob off, so only scattering — the opposite sign — remains.

Answer these fast:

Zero gap (

E_{g} = 0

): what is

e^{- E_{g} /2 k_{B} T}

Exactly

1

, at any temperature — the metal case.

T \to 0^{+}

with a real gap: carrier factor →?

0

(perfect insulator; no thermal kicks).

Silicon carrier factor at 300 K (order of magnitude)?

About

4 \times 1 0^{- 10}

Ge vs Si carrier ratio at 300 K?

About

7 \times 1 0^{3}

(Ge has ~~7000× more, smaller gap). Forgetting the "2" in $E_{g} /2 k_{B} T$ under- or over-states carriers? ::: Understates them enormously (~~

1 0^{9} \times

); the true value is the square root of the wrong one.

Which tool inverts the exponential to back-solve for

E_{g}

The natural logarithm

ln

Connections

Band gap and its meaning for conductivity (parent)
Fermi level and Fermi-Dirac distribution
Intrinsic vs extrinsic semiconductors
Doping and carrier concentration
Direct vs indirect band gap
Conductivity, mobility and drift current
Temperature dependence of resistivity in metals