Visual walkthrough — Understand conductors, insulators, and semiconductors

---

Step 1 — What is "current," really?

WHAT. Current is just charge moving past a line, counted per second. Picture a wire as a pipe. Draw an imaginary flat window across it. Current $I$ = how much charge crosses that window each second.

WHY start here. Everything downstream — resistivity, conductivity, the whole classification — is only a careful bookkeeping of how much charge crosses, how fast. If we cannot count crossings, we cannot compare materials.

PICTURE. The window (dashed teal) sits across the wire. Little orange dots are charges. In one second, only the dots close enough to reach the window actually cross it.

---

Step 2 — Spread it out: current density $J$

WHAT. A fat wire carries more current than a thin one, even from the same material, just because it has more room. To compare materials we must divide that out. So we define current density $J$ = current per unit of window area $A$: $J=\frac{I}{A}$

Term by term:

• $I$ — the charge-per-second crossing the window (from Step 1).
• $A$ — the area of the window, in square metres (m²). A wider wire = bigger $A$.
• $J$ — current squeezed onto one square metre of window. Units: amperes per square metre (A/m²).

WHY. Dividing by $A$ removes "how thick is the wire" from the picture, leaving something that depends only on how the material behaves. That is the quantity worth studying.

PICTURE. Same charge-flow, two wires — one fat, one thin. Same $J$ (dots equally dense), but the fat wire's window catches more of them, so its $I$ is larger.

---

Step 3 — Count the crossings: $J=ne{v}_d$

WHAT. Now we actually count. Suppose the charges all drift rightward at speed $v_d$ (the "drift velocity" — Step 4 explains where it comes from). In one second, every charge moves a distance $v_d$. So only the charges sitting within a distance $v_d$ behind the window will cross it this second.

That defines a slab of length $v_d$ and area $A$ — a box of volume $A{v}_d$.

Let $n$ = number of free charges packed into each cubic metre. Then:

$$\underset{{\text{A/m}}^2}{\underbrace{J}}=\underset{\begin{array}{c}\text{charges}\\ {\text{per m}}^3\end{array}}{\underbrace{n}}\times \underset{\begin{array}{c}\text{charge}\\ \text{each}\end{array}}{\underbrace{e}}\times \underset{\begin{array}{c}\text{drift}\\ \text{speed}\end{array}}{\underbrace{v_d}}$$

Where each piece comes from: charges in the slab $=n\times (A{v}_d)$; multiply by $e$ for total charge; that crossed in one second so divide by 1 s; divide by $A$ to get density $\Rightarrow J=ne{v}_d$.

WHY this matters. Look at the three factors. $e$ is a fixed constant of nature. $v_d$ turns out to be similar across materials. The wild card is $n$ — and that is exactly the number that will separate our three classes.

PICTURE. The green slab of length $v_d$ behind the window; count the dots inside it — those are precisely the ones that cross in one second.

Recall Why is it

$v_d$ and not the electron's real speed? Real electrons zip around at ~$10^6$ m/s but in random directions — cancelling out. Do they average to zero motion? ::: Almost. The tiny leftover net drift caused by the field is $v_d$, often under a millimetre per second. Only that net drift carries current.

---

Step 4 — Where drift comes from: $v_d=\mu E$

WHAT. We plug in a battery. A battery sets up an electric field $E$ inside the wire — an invisible push felt by every charge, pointing along the wire. Field $E$ = force per unit charge, units volts per metre (V/m).

If nothing stopped them, charges would accelerate forever. But they keep colliding with the vibrating atoms of the material. The result is a steady average speed proportional to the push:

$$\underset{\text{drift speed}}{\underbrace{v_d}}=\underset{\text{mobility}}{\underbrace{\mu }}\times \underset{\text{field / push}}{\underbrace{E}}$$

Where does $\mu$ come from? Force on a charge is $F=eE$. Newton says acceleration $a=F/m=eE/m$ ($m$ = electron mass). Between collisions the charge accelerates for an average time $\tau$, reaching drift $v_d=a\tau =\frac{e\tau }{m}E$. So:

$$\mu =\frac{e\tau }{m}$$

• $\tau$ — average time between collisions (seconds). Smoother material = longer $\tau$ = faster drift.
• $\mu$ — mobility: how much drift speed you get per unit of push. Units m²/(V·s).

WHY the collisions matter. Without them, $v_d$ (and current) would run away to infinity — no material has infinite conductivity. The collisions are what make resistance a real, finite thing.

PICTURE. One charge: it accelerates (orange curve steepening) between collision dots, gets knocked back to random, accelerates again — a saw-tooth whose average rightward slope is $v_d$.

---

Step 5 — Assemble the master formula: $\sigma =ne\mu$

WHAT. Substitute Step 4 into Step 3:

$$J=ne{v}_d=ne(\mu E)=\underset{\text{call this }\sigma }{\underbrace{(ne\mu )}}E$$

The whole clump $ne\mu$ multiplies $E$ to give $J$. That clump is a single material property — we name it conductivity $\sigma$:

$$\boxed{\sigma =ne\mu }J=\sigma E$$

And its reciprocal is resistivity:

$$\rho =\frac{1}{\sigma }=\frac{1}{ne\mu }$$

• $\sigma$ big → charge flows easily for a small push → conductor.
• $\sigma$ small → you push hard, barely any flow → insulator.

WHY this is the payoff. We started with "count crossings" and arrived at a formula where the three classes differ almost entirely in $n$. This is the parent note's boxed result, now built, not stated.

PICTURE. A funnel diagram: the three ingredients $n$, $e$, $\mu$ pour in and combine into the single output $\sigma$, with $e$ marked "fixed," $\mu$ "roughly steady," and $n$ "the wild card."

---

Step 6 — Case split: three classes from one knob ($n$)

WHAT. Now turn the single knob $n$ (free carriers per m³) and watch $\sigma =ne\mu$ move across the whole map, from copper to rubber.

Class	$n$ (per m³)	$\sigma =ne\mu$	Behaviour
Conductor (copper)	$\sim 10^{29}$	huge	flows freely
Semiconductor (silicon)	$\sim 10^{16}$	tiny but tunable	switchable
Insulator (rubber)	$\approx 0$	$\approx 0$	blocks

WHY show all three at once. So no case surprises you: the same formula covers a wire and a glove; only the exponent of $n$ changed. Nothing new was needed.

PICTURE. A single horizontal $n$-axis (log scale) with the three materials pinned at their positions, and $\sigma$ drawn rising with $n$ — one law, three regimes.

---

Step 7 — The degenerate & edge cases

Real derivations must survive the extremes. Here they are:

Case A — $n\to 0$ (perfect insulator). With zero free charges, $\sigma =0\cdot e\mu =0$, so $\rho =1/0\to \infty$: infinite resistance, no current for any push. The formula does not break — it correctly predicts a wall. (See Energy Bands in Solids: the band gap is so tall that $n$ stays essentially zero.)

Case B — heat a semiconductor. Here $n$ is not fixed. Thermal energy lifts electrons across the gap, and each one leaves a hole behind (a missing electron that also carries current). The count grows: $n\propto e^{-E_g/(2k_{B}T)}$

• $E_g$ — the band-gap energy (the "ladder height").
• $k_B$ — Boltzmann's constant, converting temperature to energy.
• $T$ — absolute temperature (kelvin).
• the 2 — because electron and hole are born together, the gap energy $E_g$ is split between the pair.

More $T$ → bigger $n$ → bigger $\sigma$ → resistivity falls. (Full treatment: Temperature Coefficient of Resistance.)

Case C — heat a metal. Here $n$ is already maxed out and barely changes. Instead heat makes atoms vibrate harder, so collisions come sooner: $\tau \downarrow \Rightarrow \mu \downarrow \Rightarrow \sigma \downarrow$. Resistivity rises — the opposite of the semiconductor. Same formula, different term moving.

WHY these matter. The naive reader thinks "hot = more conductive, always." Cases B and C show the formula predicts opposite signs depending on which factor temperature attacks — $n$ (semiconductor) or $\mu$ (metal).

PICTURE. Two $\rho$-versus-$T$ curves: teal metal sloping up, orange semiconductor sloping down, labelled with which term is moving.

---

The one-picture summary

Everything above, compressed: charges packed at density $n$ drift at $v_d=\mu E$ through a window of area $A$; counting crossings gives $J=ne{v}_d=(ne\mu )E$, so $\sigma =ne\mu$ — and turning only $n$ sweeps you from insulator to conductor.

Recall Feynman: the whole walkthrough in plain words

Imagine a hallway with a doorway. Kids (electrons) walk through, and you count how many pass the doorway each second — that's current. Divide by how wide the doorway is and you get current density, a fair measure that ignores door size.

How many pass per second? Just three things multiplied: how crowded the hallway is ($n$), how big each kid's ticket is ($e$, always the same), and how fast they shuffle forward ($v_d$). That shuffle speed is set by how hard you push them (the field $E$) times how slippery the hall is (mobility $\mu$) — a slippery hall lets them slide faster before bumping into a wall.

Multiply it all: crowd × ticket × (slipperiness × push). The "crowd × ticket × slipperiness" part is a property of the hallway itself — we call it conductivity $\sigma =ne\mu$.

Now the punchline. The ticket size never changes; slipperiness barely changes. The one thing that changes wildly is how crowded the hallway is. A copper wire is jam-packed (huge $n$, great conductor). Rubber is empty (no $n$, insulator). Silicon is nearly empty but you can invite kids in by warming it or sneaking in helpers (doping) — that's a semiconductor, and being able to open and close that crowd is exactly why we build chips from it. See Semiconductor Diodes and Transistors for what we do with that control.

Recall Quick self-test

In $\sigma =ne\mu$, which factor separates a conductor from an insulator? ::: $n$, the free-carrier density — it varies by ~20 orders of magnitude while $e$ is fixed and $\mu$ barely moves. Heating pure silicon: resistivity up or down, and why? ::: Down. $n\propto e^{-E_g/2k_{B}T}$ grows with $T$, so $\sigma =ne\mu$ rises and $\rho =1/\sigma$ falls. Heating copper: which term in $\sigma =ne\mu$ moves? ::: $\mu$ falls (shorter collision time $\tau$), so $\rho$ rises. $n$ stays essentially constant.

---

Connections: built on Electric Charge and Current and Ohm's Law and Resistance; deepened by Energy Bands in Solids and Temperature Coefficient of Resistance; applied in Semiconductor Diodes and Transistors. Hinglish version: 1.1.02 Understand conductors, insulators, and semiconductors (Hinglish).