Radiation safety — units (Bq, Gy, Sv), shielding: Visual walkthrough

Step 1 — The one idea: "a fixed chance per step"

WHAT. Imagine a big crowd of unstable nuclei (or a swarm of photons flying into a lead wall). Each individual does not age or get tired. In any tiny slice — one tiny slice of time for nuclei, one thin slab of shield for photons — every survivor has the same small chance of being removed (decaying, or getting absorbed).

WHY this idea and not "they wear out"? If nuclei "wore out" with age, old ones would decay faster than fresh ones and the curve would be a different shape. Experiment says they don't: a nucleus that has survived a million years is exactly as likely to decay in the next second as a brand-new one. This "memoryless, constant chance" fact is the seed of everything.

PICTURE. Below, each column is one tiny step. A fixed fraction (here $\frac{1}{4}$ , shown in orange) is removed each step, from whatever is left. Notice the removals get smaller — not because the chance changed, but because there are fewer left to remove.

Step 2 — Turning "fixed fraction" into an equation

WHAT. Let $N$ = the number of nuclei present right now. In a tiny time step $d t$ , we lose a small number $d N$ . "Fixed chance per step" means the loss is proportional to both how many we have ( $N$ ) and how long the step is ( $d t$ ): $- d N = λ N d t$

Read every symbol, right where it sits:

$N$ — how many radioactive nuclei exist this instant. A plain count.
$d N$ — the tiny change in that count during the step. The " $d$ " means "a tiny bit of". It's negative because the count drops, so we write $- d N$ (a small positive amount lost).
$d t$ — the tiny duration of the step. Again " $d$ " = "a tiny bit of".
$λ$ (lambda) — the decay constant: the chance per nucleus, per second of decaying. This is the "fixed fraction per step" from Step 1, turned into a number. Big $λ$ = jittery, quick-to-decay; small $λ$ = sluggish.

WHY write it with $d t$ at all? Because "chance per second" only makes sense once we say how much second passed. Multiply the per-second chance $λ$ by the tiny time $d t$ to get the actual fraction removed in that step.

PICTURE. The thin blue slab is one time-step. Its height is $N$ (how many we have); its width is $d t$ ; the orange sliver peeled off the top is $d N = λ N d t$ .

Step 3 — Solving it: why $e$ shows up, and nothing else

WHAT. We have $\frac{d N}{N} = - λ d t$ . We want $N$ as a function of $t$ . We separate the variables (all the $N$ stuff on one side, all the $t$ stuff on the other) and add up every tiny step from start to now. Adding up tiny pieces is what the integral $\int$ does: $\int \frac{d N}{N} = - λ \int d t \Rightarrow ln N = - λ t + C$

Read the pieces:

$\int \frac{d N}{N} = ln N$ — the natural logarithm appears because the loss is proportional to $N$ itself. That's the mathematical fingerprint of "rate ∝ amount."
$C$ — a constant fixed by the starting condition (at $t = 0$ there are $N_{0}$ nuclei).

Undo the $ln$ by raising $e$ to both sides: $N = e^{- λ t + C} = = N_{0} e^{C} e^{- λ t} = N_{0} e^{- λ t}$

WHY $e$ specifically, not $2$ or $10$ ? Because $e^{x}$ is the one function that is its own rate of change: its slope at every point equals its own height. Our equation demanded exactly that ("the drop is proportional to the amount"), so $e$ is forced — no other base fits without an extra fudge factor.

PICTURE. The smooth blue curve is $N_{0} e^{- λ t}$ . The dashed orange tangent line at any point has a slope equal to $- λ \times$ (the height at that point) — steep where the curve is tall, gentle where it's low. That self-referential slope is the exponential.

Step 4 — Reading the half-life off the curve

WHAT. The Half-life $t_{1/2}$ is the time for the count to fall to half. Set $N = N_{0} /2$ : $\frac{1}{2} = e^{- λ t_{1/2}}$ Take $ln$ of both sides. Since $ln (1/2) = - ln 2$ : $- ln 2 = - λ t_{1/2} \Rightarrow t_{1/2} = \frac{ln 2}{λ}$

Read it: big $λ$ (jittery nuclei) ⇒ small $t_{1/2}$ (dies fast). They are inverses through the constant $ln 2 \approx 0.693$ .

WHY does the same half-life repeat forever? Because the curve is memoryless (Step 1). From any height, waiting one more $t_{1/2}$ halves it again — the drops from $N_{0} \to \frac{N _{0}}{2} \to \frac{N _{0}}{4} \to \frac{N _{0}}{8}$ take equal times. That equal-spacing is the visual signature of exponential decay.

PICTURE. Equal horizontal steps of width $t_{1/2}$ each chop the height exactly in half. Green markers sit at $N_{0}, \frac{1}{2} N_{0}, \frac{1}{4} N_{0}, \frac{1}{8} N_{0}$ — the "halving staircase".

Step 5 — The SAME picture, now for shielding

WHAT. Swap the time-step for a thickness-step. A beam of intensity $I$ (photons per second per area) enters a shield. In a thin slab of thickness $d x$ , a fixed fraction is absorbed: $- d I = μ I d x \Rightarrow \frac{d I}{I} = - μ d x \Rightarrow I = I_{0} e^{- μx}$

Read the swap term-by-term:

$I$ ↔ $N$ — beam intensity plays the role of "how many are present".
$x$ (depth into the shield) ↔ $t$ (time). Distance replaces time.
$μ$ (mu), the linear attenuation coefficient (units $m^{- 1}$ ) ↔ $λ$ . It is the chance per metre of thickness that a photon gets removed. Dense, high- $Z$ lead ⇒ big $μ$ ⇒ fast falloff.

WHY is it literally the same maths? Both start from "loss ∝ amount present × size of the step." Same equation, same solution, same $e$ . Only the labels on the axes change.

PICTURE. Identical falling curve — but now the horizontal axis is millimetres of lead, and the orange slivers are photons absorbed in each successive slab.

Step 6 — Edge & degenerate cases (never hit a surprise)

WHAT & WHY — walk every corner of the curve:

At the start ( $t = 0$ or $x = 0$ ): $e^{0} = 1$ , so $N = N_{0}$ , $I = I_{0}$ . Nothing has decayed / nothing absorbed yet. The curve starts at full height. ✔
Zero rate ( $λ = 0$ or $μ = 0$ ): a perfectly stable nucleus, or a shield that absorbs nothing (e.g. a vacuum for a photon). Then $e^{0 \cdot t} = 1$ forever — a flat horizontal line, no decay. This is why stable isotopes have no half-life and why nothing shields a beam if $μ = 0$ .
Very long time / very thick shield ( $t \to \infty$ , $x \to \infty$ ): $e^{- (huge)} \to 0$ but never reaches exactly 0. Crucial for safety: gamma radiation is only ever attenuated, never fully stopped. There is no "safe wall thickness" that gives literal zero — only "small enough".
Negative axis is meaningless: $t < 0$ or $x < 0$ has no physical meaning (before the source existed / behind the shield's front face) — we only ever use the right half of the curve.
The 1/8 milestone: three halvings, $(\frac{1}{2})^{3} = \frac{1}{8}$ , i.e. 3 half-lives or 3 HVLs. This is the parent note's worked example, read straight off the halving staircase.

PICTURE. One frame overlaying all cases: the flat gray line ( $λ = 0$ ), the normal blue decay approaching but never touching zero (dashed red asymptote), and the green $1/8$ marker at three steps.

The one-picture summary

WHAT this final figure does. It stacks the whole derivation on one canvas: the "fixed chance per step" idea (Step 1) is the rate law (Step 2), which integrates to the $e$ -curve (Step 3), whose halvings define $t_{1/2}$ and $x_{1/2}$ (Step 4), and the exact same shape governs shielding (Step 5) with its never-reaching-zero tail (Step 6).

Recall Feynman retelling — say it to a 12-year-old

Picture a room full of popcorn kernels, each with the same tiny chance of popping every second. It doesn't matter how old a kernel is — the chance is always the same. So the number popping each second is just "that tiny chance" times "how many are still un-popped." Early on, lots of kernels ⇒ lots of pops. Later, few left ⇒ few pops. Draw how many are left over time and you get a curve that drops fast, then slower, then crawls toward the floor but never quite lands on it.

Now the magic: the exact same story is a beam of light hitting a lead wall. Slice the wall into thin sheets; each sheet has the same small chance of catching a photon. Same "fixed chance per step," so the same falling curve — just with "thickness of lead" on the bottom instead of "time." Cut it in half with one sheet, in half again with two, in half again with three — one-eighth left after three. And just like the popcorn, it never reaches zero: there's always a whisper getting through, which is why a radiation worker thinks in "make it small enough," not "block it all." That single memoryless coin-flip — decay or absorb — is the whole idea, and the number $e$ is just the honest bookkeeper that adds up all the tiny flips.

Flashcards

Why does radioactive decay follow an exponential curve?

Each nucleus has a fixed chance per second of decaying and doesn't age, so the number decaying is proportional to the number present (

- d N / d t = λ N

), whose solution is

N = N_{0} e^{- λ t}

Why does the base

e

appear (not 2 or 10)?

e^{x}

is the unique function equal to its own rate of change; our law "loss ∝ amount present" forces exactly that self-referential slope.

What does

λ

mean physically, and

μ

λ

= chance of decay per nucleus per second;

μ

= chance a photon is absorbed per metre of shield. They play identical roles in

e^{- λ t}

and

e^{- μx}

Derive

t_{1/2}

from the decay curve.

Set

N = N_{0} /2

N = N_{0} e^{- λ t}

; take ln:

- ln 2 = - λ t_{1/2}

, so

t_{1/2} = ln 2/ λ

As thickness or time → ∞, what does the curve do?

Approaches zero but never reaches it (horizontal asymptote) — gamma is only attenuated, never fully stopped.

What happens to the curve if

λ = 0

μ = 0

e^{0} = 1

for all steps — a flat horizontal line: no decay / no absorption (stable isotope, or a shield that absorbs nothing).

How many halvings give 1/8 the intensity?

Three, since

(1/2)^{3} = 1/8

— i.e. 3 half-lives or 3 half-value layers.