This page is the drill-hall for the parent proofs . The parent showed why d x d e x = e x and d x d a x = a x ln a . Here we hunt down every kind of exponential a problem can throw at you and solve one of each — including the sneaky degenerate cases (base = 1 , decaying bases, negative exponents) that people forget exist.
Definition Domain assumption (read this first)
Throughout this page the base a is a positive real number , a > 0 , and for a genuine exponential we also need a = 1 (base 1 gives the flat constant 1 x = 1 ). Why a > 0 ? Because a x must make sense for every real x : if a were negative, then something like ( − 2 ) 1/2 = − 2 is not a real number, so the smooth curve — and its derivative — would not exist over the reals. Positive base ⇒ a x > 0 everywhere ⇒ ln a exists, which is exactly the constant every rule below needs.
Intuition The only two facts you need on this page
Everything below is squeezed out of two rules plus the Chain Rule :
d x d e x = e x , d x d a x = a x ln a .
The chain rule says: if the exponent is not a bare x but some inside-function u ( x ) , you differentiate the outside as usual, then multiply by the derivative of the inside u ′ ( x ) . That single "multiply by the inside's slope" is where 90% of the cases differ from one another.
One shorthand used below: the slope-at-zero number of a x . Since a 0 = 1 , plugging x = 0 into a x ln a gives 1 ⋅ ln a = ln a . So "the slope of a x at x = 0 is ln a " — that single number is the whole personality of the base, and we will call it exactly that: the base's personality number ln a (no undefined symbols, just ln a ).
Every exponential-derivative problem falls into one of these boxes. We will solve at least one example per row.
#
Case class
What's tricky about it
Covered by
1
Base = e , exponent = x
the pure self-derivative
Ex 1
2
Base a > 1 , exponent = x
picks up an ln a factor
Ex 2
3
Base 0 < a < 1 (decay )
ln a < 0 → negative slope
Ex 3
4
Degenerate base a = 1
ln 1 = 0 → derivative is zero
Ex 4
5
Exponent is a linear inside k x
chain rule → extra constant k
Ex 5
6
Exponent is negative / non-linear
sign and shape of u ′ ( x )
Ex 6
7
General base with a function exponent
must convert to base e first
Ex 7
8
Slope at a point (geometry)
evaluate derivative, read a tangent
Ex 8
9
Real-world word problem (growth rate)
attach units, interpret the number
Ex 9
10
Exam twist : solve for the base
run the derivative rule backwards
Ex 10
Worked example Ex 1 (matrix cell 1). Differentiate
f ( x ) = e x and find its slope at x = 2 .
Forecast: before reading on, guess: is the slope at x = 2 bigger or smaller than the height there? (Trick: they're equal.)
f ′ ( x ) = e x . Why this step? e is defined as the base whose personality number is exactly 1 (i.e. ln e = 1 ), so the derivative is the function unchanged — no factor out front.
Slope at x = 2 is f ′ ( 2 ) = e 2 ≈ 7.389 . Why this step? The derivative is a function; you get a number by plugging in the point.
Verify: the height at x = 2 is e 2 ≈ 7.389 too. Height = slope — exactly the defining property of e x . ✓ See the height-equals-slope picture below.
Worked example Ex 2 (cell 2). Differentiate
g ( x ) = 3 x and find its slope at x = 0 .
Forecast: will 3 x climb faster or slower than e x at x = 0 ? (3 > e , so faster.)
g ′ ( x ) = 3 x ln 3 . Why this step? For any fixed base, d x d a x = a x ln a ; here a = 3 . The ln 3 is exactly the personality number of base 3 .
At x = 0 : g ′ ( 0 ) = 3 0 ln 3 = 1 ⋅ ln 3 = ln 3 ≈ 1.0986 . Why this step? 3 0 = 1 , so the slope at x = 0 is the personality number ln 3 .
Verify: ln 3 ≈ 1.0986 > 1 = (slope of e x at 0 ). Since 3 > e , a steeper start is exactly what we expect. ✓ Compare the two starting tangents in the figure below.
Worked example Ex 3 (cell 3,
decay ). Differentiate h ( x ) = ( 2 1 ) x and find its slope at x = 0 .
Forecast: this curve falls as x grows. So its slope should be… negative. Let's confirm the sign comes out right.
h ′ ( x ) = ( 2 1 ) x ln 2 1 . Why this step? Same rule a x ln a ; nothing about it required a > 1 — only a > 0 , and 2 1 > 0 . ✓
Note ln 2 1 = − ln 2 ≈ − 0.6931 . Why this step? ln of a number below 1 is negative — that negative is what makes the whole derivative negative, encoding "the curve goes down."
At x = 0 : h ′ ( 0 ) = 1 ⋅ ( − ln 2 ) = − ln 2 ≈ − 0.6931 . Why this step? ( 2 1 ) 0 = 1 ; the slope at x = 0 equals the personality number, here negative.
Verify: the function decreases everywhere, so a negative slope is correct. Notice its size 0.6931 equals the slope of 2 x — mirror bases 2 and 2 1 have opposite-sign personality numbers because ln 2 1 = − ln 2 . ✓ The decay curve and its downward tangent are shown below. See Exponential Growth and Decay .
Worked example Ex 4 (cell 4,
degenerate ). Differentiate p ( x ) = 1 x .
Forecast: what is 1 x as a graph? It's the flat line y = 1 . A flat line has slope 0 . Does the rule agree?
p ′ ( x ) = 1 x ln 1 . Why this step? Apply a x ln a blindly with a = 1 — always start with the rule, then simplify.
ln 1 = 0 , so p ′ ( x ) = 1 x ⋅ 0 = 0 . Why this step? ln 1 = 0 because e 0 = 1 ; this zero kills the whole derivative.
Verify: 1 x = 1 for all x , a horizontal line — slope 0 everywhere. The rule handles the degenerate base perfectly. ✓ (This is why we exclude a = 1 as an "exponential" — it's just a constant.)
Worked example Ex 5 (cell 5,
chain rule / linear inside ). Differentiate e 5 x .
Forecast: the answer is not e 5 x . What extra factor sneaks in? (The slope of the inside, 5 .)
Identify inside u = 5 x , so u ′ = 5 . Why this step? The exponent isn't a bare x ; the Chain Rule needs the inside's slope.
Outside derivative of e u is e u ; multiply by u ′ : d x d e 5 x = e 5 x ⋅ 5 = 5 e 5 x . Why this step? Chain rule = (outside slope)× (inside slope). Here the outside is self-derivative, so only the 5 is added.
Verify (limit sanity): by the first-principles definition , the derivative is lim h → 0 h e 5 ( x + h ) − e 5 x = lim h → 0 h e 5 x ( e 5 h − 1 ) . Here we use the linearization e t ≈ 1 + t for tiny t (this is exactly the parent's defining fact lim t → 0 t e t − 1 = 1 , i.e. e t − 1 ≈ t ). With t = 5 h : e 5 h − 1 ≈ 5 h , so the quotient ≈ h e 5 x ⋅ 5 h = 5 e 5 x . ✓ Matches.
Worked example Ex 6 (cell 6). Differentiate
q ( x ) = e − x 2 (the bell-curve shape).
Forecast: the inside − x 2 has slope − 2 x . So the answer should carry a factor − 2 x — meaning the slope is 0 at x = 0 and negative for x > 0 . Predict the sign before computing.
Inside u = − x 2 , so u ′ = − 2 x . Why this step? Non-linear exponent → its derivative is itself a function of x , not a constant. Use the Power Rule on − x 2 .
Chain rule: q ′ ( x ) = e − x 2 ⋅ ( − 2 x ) = − 2 x e − x 2 . Why this step? Outside self-derivative e u times the inside slope − 2 x .
At x = 1 : q ′ ( 1 ) = − 2 ( 1 ) e − 1 = − 2/ e ≈ − 0.7358 . Why this step? Plug the point to get the actual tangent slope.
Verify: at x = 0 , q ′ ( 0 ) = − 2 ( 0 ) e 0 = 0 — the peak of the bell, where the tangent is flat. ✓ For x > 0 the factor − 2 x is negative, so the curve descends on the right — exactly the bell's right side, drawn below.
Worked example Ex 7 (cell 7). Differentiate
r ( x ) = 2 x 2 .
Forecast: two things fight here — a base that isn't e and a non-linear exponent. What's the safe move? Convert to base e first, then chain-rule.
Rewrite: 2 x 2 = e ( l n 2 ) x 2 . Why this step? The clean chain rule lives in base e ; any positive base becomes a = e l n a , so 2 x 2 = e x 2 l n 2 .
Inside u = ( ln 2 ) x 2 , slope u ′ = 2 x ln 2 . Why this step? ln 2 is a constant multiplier; differentiate x 2 by the power rule → 2 x , keep the constant.
r ′ ( x ) = e x 2 l n 2 ⋅ 2 x ln 2 = 2 x 2 ( 2 x ) ln 2 . Why this step? This is the chain rule again — outside derivative times inside derivative : the outside e u differentiates to itself e x 2 l n 2 , and we multiply by the inside slope u ′ = 2 x ln 2 from step 2. We then rewrite e x 2 l n 2 back as 2 x 2 for a tidy answer.
At x = 1 : r ′ ( 1 ) = 2 1 ⋅ 2 ⋅ ln 2 = 4 ln 2 ≈ 2.7726 . Why this step? Evaluate to a number for a concrete check.
Verify: at x = 0 the factor 2 x is 0 , so r ′ ( 0 ) = 0 — and 2 x 2 has a flat minimum at x = 0 (its lowest point, value 1 ). ✓
Worked example Ex 8 (cell 8,
geometry ). Find the equation of the tangent line to y = e x at x = 0 .
Forecast: the tangent touches at the point ( 0 , 1 ) and has slope equal to the height there. What's the height at x = 0 ? (e 0 = 1 .) So slope = 1 .
Point of contact: ( 0 , e 0 ) = ( 0 , 1 ) . Why this step? A tangent line needs a point; use the curve's value at x = 0 .
Slope there: f ′ ( 0 ) = e 0 = 1 . Why this step? Self-derivative → slope equals height = 1 .
Tangent line: y − 1 = 1 ( x − 0 ) , i.e. y = x + 1 . Why this step? Point–slope form of a straight line.
Verify: y = x + 1 passes through ( 0 , 1 ) ✓ and has slope 1 ✓. Near x = 0 the tangent hugs the curve, and e x ≈ 1 + x for small x — the same approximation the parent used to link e to ( 1 + 1/ n ) n .
Figure (Ex 8): the blue curve y = e x and its orange tangent y = x + 1 meet at the red point ( 0 , 1 ) . The tangent's slope (1 ) equals the curve's height there (e 0 = 1 ) — the self-derivative property made visible. Away from x = 0 the straight line peels off, since the true curve keeps steepening.
Worked example Ex 9 (cell 9,
word problem ). A bacteria population is P ( t ) = 100 e 0.4 t , where t is in hours. How fast is it growing at t = 3 hours?
Forecast: "how fast" means a derivative . The units of the answer will be bacteria per hour , not bacteria. Guess: will the rate be bigger than the current count or smaller? (Rate = 0.4 × count, so smaller.)
P ′ ( t ) = 100 e 0.4 t ⋅ 0.4 = 40 e 0.4 t . Why this step? Inside u = 0.4 t has slope 0.4 ; chain rule multiplies the self-derivative by 0.4 . The constant 100 rides along untouched.
At t = 3 : P ′ ( 3 ) = 40 e 1.2 ≈ 40 × 3.3201 ≈ 132.8 . Why this step? Evaluate at the requested time to get the instantaneous rate.
Verify (units + structure): P ′ ( 3 ) ≈ 132.8 bacteria per hour . Cross-check: the current count is P ( 3 ) = 100 e 1.2 ≈ 332.0 , and P ′ ( 3 ) = 0.4 × P ( 3 ) = 0.4 × 332.0 = 132.8 . ✓ For any self-scaling exponential, growth rate = (growth constant)× (current size) — the signature of Exponential Growth and Decay .
Worked example Ex 10 (cell 10,
twist / solve for the base ). For which base a > 0 does y = a x have slope exactly 2 at x = 0 ?
Forecast: slope-at-0 of a x is the personality number ln a . So we need ln a = 2 . Solve for a .
Slope at x = 0 : d x d a x x = 0 = a 0 ln a = ln a . Why this step? a 0 = 1 ; the slope at x = 0 is precisely the personality number ln a .
Set ln a = 2 . Why this step? We want that slope to equal the required 2 .
Solve: a = e 2 ≈ 7.389 . Why this step? ln and e ( ⋅ ) are inverses (see Natural Logarithm ln x ); apply e to both sides to undo ln .
Verify: with a = e 2 , d x d a x 0 = ln ( e 2 ) = 2 . ✓ Sanity: e 2 > e , so a steeper-than-e base — matches "we wanted a steeper start (2 > 1 )."
Recall Case-checklist for exams
When you see an exponential to differentiate, ask in order:
Is the base e or some other positive number? ::: If not e , pull out the personality number ln a (or convert to e ( l n a ) u first).
Is the exponent a bare x or an inside function u ( x ) ? ::: If an inside function, multiply by u ′ ( x ) — the chain rule.
Is 0 < a < 1 ? ::: Then ln a < 0 , so the slope is negative — this is decay.
Is a = 1 ? ::: Then ln 1 = 0 , so the derivative is 0 (it's just a constant).
Do they want a number (slope at a point)? ::: Differentiate first, THEN substitute the point.
Mnemonic The universal move
"Copy the exp, times the log of the base, times the slope of the inside."
d x d a u ( x ) = a u ( x ) ⋅ ln a ⋅ u ′ ( x ) . Set a = e → ln e = 1 drops out; set u = x → u ′ = 1 drops out.