Numbers of the form a+bi with i2=−1: they give every polynomial equation a solution and turn multiplication into adding angles.
IntuitionWhy do we need a new kind of number?
The equation x2+1=0 has no solution among the real numbers, since x2≥0 for every real x. Instead of giving up, mathematicians extended the numbers: they introduced a new number i with i2=−1, and declared every combination a+bi, with a,b real, to also be a number. This is a complex number: a is its real part, b its imaginary part.
i2=−1
Definition: Complex number, conjugate, modulus
A complex number is z=a+bi with a,b∈R; write Re(z)=a and Im(z)=b. The set of all such z is denoted C. The conjugate of z is zˉ=a−bi, and its modulus is ∣z∣=a2+b2, the distance from z to the origin when it is plotted as the point (a,b) in the plane.
SchoolArithmetic with i2=−1
Adding and multiplying complex numbers follows the ordinary rules of algebra, with i2 replaced by −1 wherever it appears: (a+bi)+(c+di)=(a+c)+(b+d)i and (a+bi)(c+di)=(ac−bd)+(ad+bc)i.
Example: Multiplying two complex numbers
Compute (2+3i)(1−i).
Solution
(2+3i)(1−i)=2−2i+3i−3i2=2+i−3(−1)=5+i.
Dividing is multiplying by the conjugate: a+bi1=(a+bi)(a−bi)a−bi=a2+b2a−bi, since zzˉ=a2+b2=∣z∣2 is always real.
Powers of i repeat with period 4
i0
i1
i2
i3
i4
1
i
−1
−i
1
SchoolThe complex plane and polar form
Plotting z=a+bi as the point (a,b) turns C into a plane, called the complex plane. Writing a=rcosθ and b=rsinθ gives the polar form z=r(cosθ+isinθ), where r=∣z∣ is the modulus and θ, the argument, is the angle from the positive real axis.
A unit circle in the complex plane with a radius drawn to the point at angle 60 degrees, labelled with its coordinates cos 60° and sin 60°.
The point cosθ+isinθ on the unit circle, at θ=60°.
For every real θ and integer n, (r(cosθ+isinθ))n=rn(cosnθ+isinnθ).
Why is it true?
Multiplying two numbers in polar form multiplies their moduli and adds their arguments: r1(cosθ1+isinθ1)⋅r2(cosθ2+isinθ2)=r1r2(cos(θ1+θ2)+isin(θ1+θ2)), by the angle-sum identities for cosine and sine. Multiplying n copies of the same number adds its argument to itself n times.
Proof
We prove (cosθ+isinθ)n=cosnθ+isinnθ by induction on n≥0; the general modulus r=1 case and negative n=−m are then immediate.
**Base case (n=0).** (cosθ+isinθ)0=1=cos0+isin0. ✓
Inductive step. Assume (cosθ+isinθ)n=cosnθ+isinnθ. Multiply both sides by (cosθ+isinθ): (cosθ+isinθ)n+1=(cosnθ+isinnθ)(cosθ+isinθ). Expanding and applying the angle-addition identities cos(A+B)=cosAcosB−sinAsinB and sin(A+B)=sinAcosB+cosAsinB, the real part becomes cosnθcosθ−sinnθsinθ=cos(n+1)θ and the imaginary part becomes sinnθcosθ+cosnθsinθ=sin(n+1)θ. So the formula holds for n+1, completing the induction.
Extension to general modulus and negative exponent. For modulus r=1: (r(cosθ+isinθ))n=rn(cosnθ+isinnθ). For n=−m with m>0: since ∣cosmθ+isinmθ∣=1, the reciprocal is cosmθ+isinmθ=cos(−mθ)+isin(−mθ), giving (cosθ+isinθ)−m=cos(−mθ)+isin(−mθ).
Example: A power of a complex number
Compute (−1+i)8.
Solution
In polar form, −1+i has modulus r=2 and argument θ=135°, since it sits in the second quadrant. By de Moivre, (−1+i)8=(2)8(cos(8⋅135°)+isin(8⋅135°))=16(cos1080°+isin1080°)=16(cos0°+isin0°)=16, since 1080°=3⋅360°.
UndergraduateEuler's formula: a bridge to functions
Polar form hints that multiplying complex numbers adds angles — the same way exponents add when multiplying powers. Euler's formula makes this precise by extending the exponential function to imaginary exponents.
The Taylor series of ex, cosx and sinx still make sense when x is replaced by iθ: eiθ=∑n=0∞n!(iθ)n. Splitting the sum into even and odd n, and simplifying each power of i using i2=−1, gives exactly cosθ=∑(2k)!(−1)kθ2k as the real part and sinθ=∑(2k+1)!(−1)kθ2k+1 as the imaginary part. Making this fully rigorous — manipulating an infinite series of complex terms like a finite sum — belongs to the theory of functions of a complex variable.
Proof
Step 1 — Recall the Taylor series. For all real x: ex=∑n=0∞n!xn, cosx=∑k=0∞(2k)!(−1)kx2k, sinx=∑k=0∞(2k+1)!(−1)kx2k+1. All three series converge absolutely for every real (and complex) argument.
**Step 2 — Substitute x=iθ into the exponential series.** Absolute convergence justifies term-by-term substitution: eiθ=∑n=0∞n!(iθ)n=1+iθ+2!(iθ)2+3!(iθ)3+⋯.
**Step 3 — Simplify using i2=−1 and separate real and imaginary parts.** The powers of i cycle with period 4: i0=1,i1=i,i2=−1,i3=−i,i4=1,… Separating even-indexed terms (real) and odd-indexed terms (imaginary): eiθ=(1−2!θ2+4!θ4−⋯)+i(θ−3!θ3+5!θ5−⋯)=cosθ+isinθ. Making this fully rigorous requires absolute convergence of the complex exponential series, which holds because ∑∣an∣ converges.
eiπ+1=0
This single equation, Euler's identity, links five fundamental constants — 0, 1, e, i, π — through addition, multiplication and exponentiation. It is the case θ=π of Euler's formula, since cosπ=−1 and sinπ=0.
Every polynomial p(z)=anzn+⋯+a1z+a0 with complex coefficients and n≥1 has at least one root in C. Consequently p factors completely into n linear factors, so it has exactly n roots counted with multiplicity.
Why is it true?
Over R, a polynomial such as x2+1 can fail to have a root because its graph never crosses the x-axis. Complex numbers remove this obstruction: C contains a root for every non-constant polynomial, with no exceptions left. Once one root z1 is found, dividing p(z) by (z−z1) gives a polynomial of degree n−1, to which the same fact applies again.
Proof
There is no purely algebraic proof: every known proof borrows a tool from analysis or topology. The shortest uses Liouville's theorem from the theory of functions of a complex variable — see "Functions of a complex variable" for the full argument.
Complex numbers close algebra's oldest gap, but they open a much larger subject: what happens when a function itself takes a complex input, and its derivative is required to exist in every direction of the plane at once? That question — differentiating and integrating over the complex plane rather than the real line — is the subject of "Functions of a complex variable", where Euler's formula reappears as a special case of the complex exponential function.
A grid of the complex plane on the left and its image under z squared on the right; angles between grid lines are preserved at each point, but the grid is stretched more the farther a point is from the origin.
The map z↦z2: a preview of functions of a complex variable.