1. Newton–Leibniz Rule (Differentiation of Definite Integral w.r.t. a Limit) JEE Main & Advanced

When the limits of a definite integral are functions of x, we differentiate using the following rule:

ddx∫u(x)v(x)f(t)dt=f(v(x))⋅v′(x)−f(u(x))⋅u′(x)

Special Cases

Form Derivative Reason
ddx∫axf(t)dt f(x) Lower limit constant; v(x)=x, v′(x)=1
ddx∫xbf(t)dt −f(x) Upper limit constant; u(x)=x, minus sign
ddx∫ax2f(t)dt f(x2)⋅2x Chain rule: v(x)=x2, v′(x)=2x
ddx∫xx2f(t)dt f(x2)⋅2x−f(x) Both limits variable — apply full formula

Worked Example 1

Find F′(x) if F(x)=∫1x2ln⁡tdt.

F′(x)=ln⁡(x2)⋅ddx(x2)=2ln⁡x⋅2x=4xln⁡x

Worked Example 2

Find ddx∫xx2sin⁡tdt.

=sin⁡(x2)⋅2x−sin⁡(x)⋅1=2xsin⁡(x2)−sin⁡x

Application — Evaluating Limits via FTC + L'Hôpital

Many JEE limits of the form limx→a∫axf(t)dtg(x) are 00 forms. Apply L'Hôpital by differentiating numerator and denominator:

limx→a∫axf(t)dtg(x)=limx→af(x)g′(x)

Example: limx→0∫0xtsin⁡tdtx2

→L'Hlimx→0xsin⁡x2x=limx→0sin⁡x2=02=0

2. Differentiation Under the Integral Sign (DUIS) — Feynman's Technique JEE Advanced

If an integral contains a parameter (say a) inside the integrand, we can differentiate w.r.t. that parameter to create a simpler integral, solve it, then integrate back.

If I(a)=∫cdf(x,a)dx, then under suitable conditions:

I′(a)=dIda=∫cd∂f∂adx

The key idea: differentiating w.r.t. the parameter a is often much simpler than computing the original integral directly.

The Classic Example: ∫01xa−1ln⁡xdx

Let I(a)=∫01xa−1ln⁡xdx.

Step 1 — Differentiate w.r.t. a:

I′(a)=∫01∂∂a(xa−1ln⁡x)dx=∫01xaln⁡xln⁡xdx=∫01xadx=1a+1

Step 2 — Integrate back:

I(a)=∫1a+1da=ln⁡(a+1)+C

Step 3 — Determine constant using a known value:

I(0)=∫01x0−1ln⁡xdx=∫010dx=0. So C=0.

∴I(a)=ln⁡(a+1)

Now evaluate any integral of this family instantly:

  • ∫01x−1ln⁡xdx=I(1)=ln⁡2
  • ∫01x2−1ln⁡xdx=I(2)=ln⁡3
  • ∫01x3−x2ln⁡xdx=I(3)−I(2)=ln⁡4−ln⁡3=ln⁡43

When to Use DUIS

Look for these signals in the integrand:

  • xaln⁡x or xa−xbln⁡x — the ln⁡x in the denominator "cancels" on differentiation.
  • An integral you cannot evaluate by standard methods but could simplify by adding a parameter.
  • An integral family where one member is easy (e.g., a=0 gives a known value).

3. Integral Equations — Finding f(x) from a Functional Equation JEE Advanced

JEE Advanced frequently poses problems of the form: "Given that f(x) satisfies f(x)=g(x)+∫abh(t)⋅f(t)dt, find f(x)."

The key observation: ∫abh(t)⋅f(t)dt is a definite integral — just a number. So we call it c (a constant), solve for f(x) in terms of c, substitute back to find c.

Standard Method

  1. Let c=∫abh(t)⋅f(t)dt   (c is a constant).
  2. Then f(x)=g(x)+c — express f in terms of c.
  3. Substitute f(t)=g(t)+c into the definition of c:
  4. Solve the resulting algebraic equation for c.
  5. Substitute back to get explicit f(x).

Worked Example

Find f(x) if f(x)=x+∫01t⋅f(t)dt.

Step 1: Let c=∫01t⋅f(t)dt (a constant).

Step 2: Then f(x)=x+c, so f(t)=t+c.

Step 3: Substitute: c=∫01t(t+c)dt=∫01(t2+ct)dt=[t33+ct22]01=13+c2

Step 4: Solve: c−c2=13⇒c2=13⇒c=23.

Step 5: f(x)=x+23.

Verification: ∫01t(t+23)dt=∫01(t2+2t3)dt=13+13=23=c ✓

4. Summation Using Integration (Limit as a Riemann Sum) JEE Advanced

This technique (introduced in Topic 3) is extended here. The master formula is:

limn→∞1n∑r=1nf(rn)=∫01f(x)dx

More generally: limn→∞1n∑r=pq⋅nf(rn)=∫pqf(x)dx

Step-by-Step Identification

Signal in the sum Interpretation
1n multiplied overall This is dx (width of each strip)
rn appearing This is x (position of strip)
r runs from 1 to n Limits 0 to 1
r runs from 0 to n−1 Limits 0 to 1 (same integral)

Worked Example

Evaluate: limn→∞(1n2+12+1n2+22+⋯+1n2+n2)

=limn→∞∑r=1n1n2+r2=limn→∞1n∑r=1n11+(rn)2

=∫0111+x2dx=[ln(x+1+x2)]01=ln⁡(1+2)−ln⁡1=ln⁡(1+2)

5. Estimation and Comparison of Definite Integrals JEE Advanced

When an integral cannot be computed exactly, its value can be bounded using:

Bound Property

If m≤f(x)≤M for all x∈[a,b], then:

m(b−a)≤∫abf(x)dx≤M(b−a)

Comparison Property

If f(x)≥g(x) for all x∈[a,b], then:

∫abf(x)dx≥∫abg(x)dx

Example — Bounding ∫01e−x2dx

On [0,1]: e−1≤e−x2≤e0=1. So:

e−1⋅1≤∫01e−x2dx≤1, i.e., 1e≤∫01e−x2dx≤1

(This integral has no closed form — the bound is the best we can do analytically.)