Computing (1+i)100 through repeated multiplication would fill pages with tedious algebra. Yet the answer emerges in seconds using De Moivre's theorem — a formula that transforms exponentiation into elementary arithmetic on angles and lengths. This same principle runs in reverse to extract roots, revealing that every nonzero complex number has exactly n distinct n-th roots, arranged in perfect geometric symmetry around the origin.
Multiplying out eight brackets of (1+i) is long and easy to get wrong. Watching the length and the angle of each power tells a much simpler story.
Each power multiplies the length by √2 and turns the angle by another 45°.
power
multiplied out
length (modulus)
angle (argument)
z
1 + i
√2
45°
z²
2i
2
90°
z³
−2 + 2i
2√2
135°
z⁴
−4
4
180°
z⁸
16
16
360°
The lengths go 2,2,22,4,… — powers of 2 — and the angles go 45°,90°,135°,180°,… — multiples of 45°. Eight steps reach length 16 at angle 360°, the number 16.
In algebraic form, powers drown in computation; in trigonometric form, they follow one simple rule.
Raising complex numbers to powers in algebraic form quickly becomes unmanageable. Squaring z=a+bi requires expanding (a+bi)2=a2+2abi+b2i2=(a2−b2)+2abi. Cubing demands multiplying this result by a+bi again. By the time we reach z10 or z100, the algebraic approach drowns in computation.
The trigonometric form offers rescue. When z=rcisθ, multiplication becomes elegant: multiply moduli, add arguments. Squaring gives z2=r2cis(2θ). Cubing gives z3=r3cis(3θ). The pattern suggests a general rule — and De Moivre's theorem confirms it.
The definition
De Moivre's theorem says that for a complex number in trigonometric form and any integer n, (rcisθ)n=rncis(nθ): raise the modulus to the power, multiply the angle by n.
The formula is set out in full, with its special cases, in De Moivre's formula.
So (1+i)8=(2)8cis(8⋅45°)=16cis360°=16.
The theorem turns a power into one power of a length and one multiple of an angle.
De Moivre's theorem stands among the most useful results in complex analysis. It reduces high powers to trivial calculations, enables extraction of roots, derives trigonometric identities, and appears throughout physics and engineering wherever oscillations and rotations arise.
The identities come from computing the same power in two ways. Expanding (cosθ+isinθ)3 term by term gives the real part cos3θ−3cosθsin2θ, while De Moivre's theorem says the real part is cos3θ. Replacing sin2θ by 1−cos2θ turns this into cos3θ=4cos3θ−3cosθ, and the same comparison for any n writes cosnθ and sinnθ as polynomials in cosθ and sinθ.
Abraham de Moivre, a French mathematician living in England, discovered this relationship in the early eighteenth century. Though he never stated the theorem in its modern form, his work on probability and trigonometric identities laid the foundation. The theorem bearing his name connects complex exponentiation to the simple operation of multiplying an angle by an integer.
Fact: A degree-45 equation solved by angles
In 1593 Adriaan van Roomen challenged the world's mathematicians with x45−45x43+945x41−⋯+45x=A. Viète saw that the left side is 2sin45θ written in terms of x=2sinθ, the identity this theorem gives for n=45, and listed 23 positive solutions.
The number must first be in trigonometric form, and the angle is multiplied by n, not raised to the power n.
(cis45°)3 is cis135°, not an angle of 453 degrees. Turning a+bi into rcisθ takes the modulus and the argument, as described in trigonometric form.
The rest of the page covers De Moivre's formula, its proof, applying it to powers and roots, and the trigonometric identities it produces.
De Moivre's Formula
The theorem states a precise relationship between powers and angles. For any integer n:
Raise cisθ to the n-th power, and the result is cis(nθ). The angle simply multiplies by n. No expansion, no collecting terms, no tracking powers of i — just multiply the argument.
In the condensed notation: (cisθ)n=cis(nθ).
The formula extends naturally to complex numbers with arbitrary modulus. If z=rcisθ, then:
De Moivre's Formula
zn=(rcisθ)n=rncis(nθ)
Pitfall: Forgetting to raise the modulus
Writing (2cis(45°))4 as 2cis(180°)=−2 multiplies the angle but leaves the modulus at 2. The modulus is raised to the fourth power too: 24cis(180°)=16cis(180°)=−16.
The modulus raises to the n-th power while the argument multiplies by n. Both operations are elementary. The number 2cis(30°) raised to the sixth power becomes 26cis(180°)=64cis(180°)=−64. What would require extensive algebra in rectangular form reduces to mental arithmetic in trigonometric form.
The formula works for negative integers as well. Since z−n=1/zn, we have:
De Moivre's Formula
z−n=r−ncis(−nθ)
Negative exponents invert the modulus and negate the angle multiplier, maintaining the same structural simplicity.
z and z²: the argument doubles
Squaring the arrow turns it through its own angle a second time and squares its length: z² sits at angle 2θ and distance r², which for a point on the unit circle is the statement (cos θ + i sin θ)² = cos 2θ + i sin 2θ. Raise the power and watch the angle keep multiplying on the De Moivre visualizer.
Multiplying the angle is the entire content of the theorem; the modulus behaves as it does for any real power.
Proof of De Moivre's Theorem
Mathematical induction establishes De Moivre's theorem for positive integers. The proof builds from the multiplication rule for trigonometric form: when two complex numbers multiply, their moduli multiply and their arguments add.
The base case n=1 holds trivially: (cosθ+isinθ)1=cos(1⋅θ)+isin(1⋅θ).
For the inductive step, assume the formula holds for n=k:
Proof of De Moivre's Theorem
(cosθ+isinθ)k=cos(kθ)+isin(kθ)
We must prove it holds for n=k+1. Multiply both sides by cosθ+isinθ:
Proof of De Moivre's Theorem
(cosθ+isinθ)k+1=(cos(kθ)+isin(kθ))(cosθ+isinθ)
The right side is a product of two numbers in trigonometric form. By the multiplication rule, we add arguments:
Proof of De Moivre's Theorem
=cos(kθ+θ)+isin(kθ+θ)=cos((k+1)θ)+isin((k+1)θ)
This completes the induction. The theorem holds for all positive integers.
Extension to negative integers follows from the reciprocal. For n>0:
Proof of De Moivre's Theorem
z−n=zn1=rncis(nθ)1=rn1⋅cis(−nθ)=r−ncis(−nθ)
The formula zn=rncis(nθ) thus holds for all integers n, positive, negative, or zero.
Applying De Moivre's Theorem to Powers
Computing high powers follows a three-step procedure. First convert to trigonometric form, then apply De Moivre's formula, finally convert back to algebraic form if required.
Consider (1+i)10.
Step 1: Convert 1+i to trigonometric form. The modulus is ∣1+i∣=12+12=2. The argument satisfies tanθ=1/1=1 with the point in the first quadrant, giving θ=45°=π/4. Thus 1+i=2cis(45°).
Step 2: Apply De Moivre's theorem.
Applying De Moivre's Theorem to Powers
(1+i)10=(2)10cis(10×45°)=25cis(450°)=32cis(450°)
Since 450°=360°+90°, the angle reduces to 90°:
Applying De Moivre's Theorem to Powers
=32cis(90°)=32(cos90°+isin90°)=32(0+i)=32i
Step 3: The answer 32i is already in algebraic form.
Verification through direct calculation would require multiplying 1+i by itself ten times — tedious and error-prone. De Moivre's theorem completes the same computation in a few lines.
Pitfall: Taking the argument from the tangent alone
Taking θ=arctan−11=−45° for −1+i gives (−1+i)3=(2)3cis(−135°)=−2−2i. The point lies in the second quadrant, so θ=135° and (−1+i)3=22cis(405°)=2+2i.
Another example: find (3−i)6. The modulus is 3+1=2. The argument, with a=3>0 and b=−1<0, lies in the fourth quadrant: θ=−30°. So 3−i=2cis(−30°), and:
Applying De Moivre's Theorem to Powers
(3−i)6=26cis(−180°)=64cis(−180°)=64(−1)=−64
z
n
r, θ
rⁿ cis(nθ)
Result
1 + i
10
√2, 45°
(√2)¹⁰ cis(450°) = 32 cis(90°)
32i
√3 − i
6
2, −30°
2⁶ cis(−180°) = 64 · (−1)
−64
2 cis(30°)
6
2, 30°
2⁶ cis(180°)
−64
3 cis(60°)
−2
3, 60°
3⁻² cis(−120°) = (1⁄9) cis(−120°)
−1⁄18 − (√3⁄18) i
z through z¹⁰: a spiral of powers
The ten arrows step round by the same angle each time while their lengths grow geometrically, and the tenth lands where the theorem says: argument 10θ, modulus r¹⁰. The algebraic expansion of a tenth power would take dozens of terms; the picture takes ten turns. Follow the spiral for any base on the De Moivre visualizer.
Converting back to algebraic form is the only step with any arithmetic in it.
The section's first example, one move at a time: out of algebraic form, the power, and back.
The power itself is a single move in trigonometric form; the conversions at either end carry the work. The section's second example, (3−i)6=−64, follows the same three moves.
Introduction to n-th Roots
De Moivre's theorem runs in reverse to solve equations of the form zn=w. Given a complex number w, what values of z satisfy this equation? The answer reveals surprising multiplicity: every nonzero complex number has exactly n distinct n-th roots.
Real number experience suggests uniqueness. The equation x3=8 has one real solution: x=2. But in C, the equation z3=8 has three solutions. The cube root of 8 is not just 2 but three different complex numbers, each of which cubes to 8.
Where do these extra roots come from? The key lies in angle periodicity. Adding 360° (or 2π radians) to any argument produces the same complex number: cis(θ)=cis(θ+360°). When we write w=Rcisϕ, we could equally write w=Rcis(ϕ+360°) or w=Rcis(ϕ+720°).
Now suppose zn=w with z=rcisα. By De Moivre's theorem, rncis(nα)=Rcisϕ. Matching moduli gives r=R1/n. Matching arguments gives nα=ϕ+360°k for some integer k, so α=nϕ+360°k.
Different values of k yield different angles α. As k runs from 0 to n−1, we obtain n distinct arguments. Beyond k=n−1, the angles repeat modulo 360°. Exactly n roots exist, no more, no fewer.
The three cube roots of 8
Write 8 = 8 cis 0°. Every cube root has modulus 8^(1/3) = 2, so all three sit on the circle of radius 2. Their arguments are (0° + 360°k)/3 for k = 0, 1, 2: 0°, 120° and 240°. The real root 2 is the one real-number experience finds; the other two, −1 ± √3 i, are a conjugate pair, and each cubes to 8 because tripling 120° or 240° lands on a multiple of 360°.
Take k = 3 and the angle is 360°, the same point as k = 0: the list of roots closes after n steps.
The Formula for n-th Roots
Let w=Rcisϕ be any nonzero complex number. The n-th roots of w — all solutions to zn=w — are given by:
Each root has the same modulus: ∣zk∣=R1/n, the positive real n-th root of the original modulus. The roots differ only in their arguments.
Pitfall: Adding the full turn after dividing
For z3=8i, taking the angles 390°+360°k=30°+360°k gives the single root 2cis(30°) three times. Adding 360°k before dividing gives 390°+360°k=30°,150°,270°: the roots 3+i, −3+i and −2i.
The first root (k=0) has argument ϕ/n, the original angle divided by n. Each subsequent root adds 360°/n to the argument. The n roots space themselves evenly around the circle of radius R1/n.
Consider finding the cube roots of 8. Write 8=8cis(0°), so R=8 and ϕ=0°. The formula gives:
The Formula for n-th Roots
zk=81/3cis(30°+360°k)=2cis(120°k)
For k=0: z0=2cis(0°)=2. For k=1: z1=2cis(120°)=2(−21+23i)=−1+3i. For k=2: z2=2cis(240°)=2(−21−23i)=−1−3i.
Three roots: 2, −1+3i, and −1−3i. Each cubes to 8.
Taking the formula apart: the common modulus, the original argument, the full turns, and the division by n.
one modulus for all n roots; the arguments start at φ/n and step evenly around the circle
zk=R1/ncis(φ+2πkn)
1The k-th root
2Same modulus
3Direction
4Original argument
5Full turns
6Even spacing
1The k-th rootzₖ
Each value of k from 0 to n − 1 gives one root; together they are all solutions of zⁿ = w. Example: the cube roots of 8 are z₀ = 2, z₁ = −1 + √3 i and z₂ = −1 − √3 i. Each cubes to 8.
2Same modulusR¹ᐟⁿ
The positive real n-th root of the original modulus, shared by all n roots; the roots differ only in their arguments. Example: 8 = 8 cis 0°, so R = 8 and every cube root has modulus 8¹ᐟ³ = 2.
3Directioncis
It turns the angle in brackets into a point on the unit circle; the modulus in front scales it out to radius R¹ᐟⁿ. Example: 2 cis 120° = 2(−1/2 + (√3/2) i) = −1 + √3 i.
4Original argumentφ
For k = 0 the argument is φ/n: the original angle divided by n. That first root is where the evenly spaced set starts. Example: φ = 0° for w = 8, so z₀ = 2 cis 0° = 2.
5Full turnsk = 0 … n−1
Adding 2πk (360°k) leaves w unchanged, but after dividing by n each k lands somewhere new. Adding the turns after dividing gives the same root n times. Example: for z³ = 8i, (90° + 360°k)/3 gives 30°, 150°, 270°.
6Even spacingn
Dividing by n spreads the n roots evenly around the circle of radius R¹ᐟⁿ. Example: for the cube roots of 8 the step is 360°/3 = 120°, giving the angles 0°, 120° and 240°.
The part most often misread is the order of the full turns and the division: add 2πk first, then divide by n. The roots of unity are the case w=1.
Geometric Interpretation of Roots
The n-th roots of any nonzero complex number arrange themselves with perfect geometric regularity. Plotting them on the complex plane reveals a pattern of elegant symmetry.
All n roots share the same modulus R1/n, placing them on a circle centered at the origin with that radius. They do not cluster or scatter randomly — they distribute themselves at equal angular intervals around this circle.
The angular separation between consecutive roots is exactly n360° (or n2π radians). The cube roots of any nonzero number sit 120° apart. Fourth roots sit 90° apart. Fifth roots sit 72° apart. This uniform spacing holds regardless of which nonzero number we take roots of — only the radius and starting angle vary.
Fact: A polygon's chords multiply to n
In a regular n-gon drawn in a circle of radius 1, the n−1 segments from one corner to all the others multiply to exactly n: for a square, 2⋅2⋅2=4. The corners are the roots of zn−1; dividing out z−1 leaves 1+z+⋯+zn−1, the product of z minus each other corner, and at z=1 it equals n.
For n≥3, the roots form vertices of a regular n-gon inscribed in the circle. Cube roots give an equilateral triangle. Fourth roots give a square. Fifth roots give a regular pentagon. The polygon may appear rotated depending on the argument of the original number, but its shape is always perfectly regular.
This geometric picture aids computation and verification. If one root is known, the others follow by rotating 360°/n around the origin. If roots fail to form a regular polygon (for n≥3), an error has occurred. The visual pattern provides both a computational shortcut and a consistency check.
The conjugate relationship appears when the original number w is real. In that case ϕ=0° or 180°, and the roots distribute symmetrically about the real axis. Complex roots come in conjugate pairs, with the real axis serving as a line of mirror symmetry.
n
Angular separation
Polygon shape on the circle of radius R1/n
2
180°
two diametrically opposite points (a line segment through the origin)
3
120°
equilateral triangle
4
90°
square
5
72°
regular pentagon
6
60°
regular hexagon
general n
360°⁄n
regular n-gon centered at origin (n ≥ 3)
Whatever the nonzero number, its n-th roots cut one circle into n equal arcs.
Roots of Unity
The n-th roots of unity — solutions to zn=1 — form a special case of fundamental importance. Since 1=1cis(0°), the formula simplifies:
All roots of unity have modulus 1 — they lie on the unit circle.
Their arguments are multiples of n360°, dividing the circle into n equal arcs.
The root z0=cis(0°)=1 always appears. Every integer power of 1 equals 1, making 1 an n-th root of unity for every n.
The fourth roots of unity illustrate the pattern: z0=1, z1=i, z2=−1, z3=−i. These four points mark the cardinal directions on the unit circle, forming a square. Each satisfies z4=1.
The sixth roots of unity are 1, cis(60°), cis(120°), −1, cis(240°), cis(300°). In algebraic form: 1, 21+23i, −21+23i, −1, −21−23i, 21−23i. They form a regular hexagon on the unit circle.
A primitive n-th root of unity, usually written ω=cis(n2π), generates all others through its powers. Since zk=ωk, the complete set of n-th roots of unity is {1,ω,ω2,…,ωn−1}. This generator perspective connects roots of unity to group theory and number theory.
One number, ω, generates every n-th root of unity: its powers walk the unit circle in equal steps.
Notation for n-th Roots
Notation
Notation for n-th Roots
Root-finding invents three marks of its own: the running label that names each answer, the Greek letter that generates them all, and a fractional exponent that means one thing on positive reals and n things on complex numbers.
The enumeration subscript with a job description: k runs 0,1,…,n−1, one label per root — see The Formula for n-th Roots above. Beyond k=n−1 the labels wrap (zn would repeat z0), which is exactly why the list stops.
Casesk=0 marks the principal root, argument ϕ/n; the target number wears capitals to stay distinct — w=Rcisϕ in, zk out. The same enumeration job as quadratic roots, here with a running range.
Do not confuseA power. zk is the k-th label, zk the k-th power — subscript versus superscript, and roots of unity put both in one equation: zk=ωk.
ω
omega — the primitive root of unity
One letter generates them all: ω=cis(2π/n), and the full set of n-th roots of unity is {1,ω,ω2,…,ωn−1} — Roots of Unity above. The convention travels intact into group theory, number theory and Fourier analysis.
CasesThe bare letter assumes context fixes n; texts juggling several orders subscript it — ωn=e2πi/n, the spelling signal processing uses for the DFT's twiddle factors.
Do not confuseThe letter w. In handwriting ω and w collapse into each other — and this very page uses w for the number whose roots are sought; keeping the Greek curl visible is not pedantry here.
Same glyph elsewherePhysics' angular frequency — the ω of Aeiωt in the exponential form's engineering costume — an unrelated job for the same letter.
R1/n · w1/n
R to the one over n
Two bases, two behaviors. On the positive real R, the fractional exponent means the single positive root — what makes R1/n safe in the roots formula. On a complex w, the same exponent is n-valued: w1/n without a chosen branch names all n roots at once.
CasesThe formula splits the jobs deliberately: single-valued R1/n for the modulus, the +2πk in the angle for the multiplicity — multivaluedness lives in the periodic exponent, never in the real root.
Do not confuseThe radical habit. 38=2 on the real line, but "the" cube root of 8i does not exist — three do; importing the single-answer instinct from real radicals is how the other n−1 roots vanish from solutions.
The n-th roots of unity satisfy algebraic identities that reflect their geometric symmetry. These properties appear throughout mathematics, from polynomial factorization to Fourier analysis.
Geometrically, the roots form a regular polygon centered at the origin. Their vector sum — placing all arrows tip to tail — returns to the starting point, yielding zero. Algebraically, this is the sum of a geometric series with ratio ω=1: 1−ω1−ωn=1−ω1−1=0.
The product of all n-th roots of unity follows a pattern:
Properties of Roots of Unity
k=0∏n−1zk=(−1)n+1
For odd n, the product is 1. For even n, the product is −1. This connects to the constant term in the polynomial zn−1, which factors as (z−z0)(z−z1)⋯(z−zn−1).
When n≥2, complex roots of unity come in conjugate pairs. If ωk is a root, so is ωk=ωn−k. The roots ω and ωn−1 are conjugates; ω2 and ωn−2 are conjugates; and so on. Only z=1 (always) and z=−1 (when n is even) sit on the real axis without conjugate partners among the roots.
These symmetries constrain the algebraic behavior of expressions involving roots of unity and simplify many calculations.
Concrete examples consolidate the root-finding technique. Each follows the same pattern: express the target in trigonometric form, apply the formula, enumerate all n roots.
Square roots of the imaginary unit: Solve z2=i.
Write i=1cis(90°). The roots are zk=11/2cis(290°+360°k) for k=0,1.
z0=cis(45°)=22+22i
z1=cis(225°)=−22−22i
Two roots, opposite each other on the unit circle.
Cube roots of negative eight: Solve z3=−8.
Write −8=8cis(180°). The roots are zk=2cis(3180°+360°k) for k=0,1,2.
z0=2cis(60°)=1+3i
z1=2cis(180°)=−2
z2=2cis(300°)=1−3i
Three roots forming an equilateral triangle with radius 2.
Fourth roots of unity: Solve z4=1.
Write 1=1cis(0°). The roots are zk=cis(90°k) for k=0,1,2,3.
z0=1, z1=i, z2=−1, z3=−i
A square on the unit circle at the four cardinal points.
Fifth roots of thirty-two: Solve z5=32.
Write 32=32cis(0°). The roots are zk=2cis(72°k) for k=0,1,2,3,4.
A regular pentagon with radius 2, one vertex on the positive real axis.
Summary: Forward and Reverse Directions of De Moivre's Theorem
De Moivre's theorem operates in two directions. Run forward, it turns a complex number into its nth power — a single answer obtained by raising the modulus and multiplying the argument. Run in reverse, it produces every nth root of a given number — n distinct values arranged as a regular polygon. The table below collects both directions side by side: formula, what happens to modulus and argument, how many results to expect, and the geometric picture each direction draws on the complex plane.
Aspect
Forward direction (powers)
Reverse direction (n-th roots)
Question
given z, what is zⁿ ?
given w, what are the z with zⁿ = w ?
Formula
zⁿ = rⁿ cis(nθ)
zₖ = R1/n cis((φ + 360°k) ⁄ n), k = 0…n−1
Modulus operation
r → rⁿ (raise to nth power)
R → R1/n(take positive nth root)
Argument operation
θ → nθ (multiply by n)
φ → (φ + 360°k) ⁄ n (divide, then shift by k turns)
Number of results
exactly 1 (a single value)
exactly n (distinct roots)
Geometric picture
a single point at distance rⁿ from the origin
vertices of a regular n-gon (n ≥ 3) on the circle of radius R1/n
Worked example
(1 + i)¹⁰ = 32i (one answer)
z⁴ = −16: four roots √2(±1 ± i) (a square)
Common Pitfalls
Four mistakes account for many of the wrong answers in work with De Moivre's theorem. Three are marked where they happen in the sections above; keeping only one root is the one the introduction to n-th roots already points out. Here each is worked through in full, with the wrong and the right working side by side.
Forgetting to raise the modulus
The angle is multiplied by n, but the modulus is left as it is, as if every number had modulus 1.
Right working: (2cis(45°))4=24cis(180°)=16cis(180°)=−16. In algebraic form 2cis(45°)=2+2i, its square is 2+4i+2i2=4i, and (4i)2=16i2=−16.
Why it happens: the theorem is first stated for cosθ+isinθ, whose modulus is 1, and it is remembered as "multiply the angle".
How to check: the modulus of zn is ∣z∣n. Here ∣z4∣=24=16, so −2 cannot be right.
Taking the argument from the tangent alone
The argument is found as arctanab without looking at the quadrant of the point, and De Moivre's theorem then multiplies the wrong angle.
Wrong working: for z=−1+i, θ=arctan−11=−45°, so (−1+i)3=(2)3cis(−135°)=22(−22−22i)=−2−2i.
Right working: the point (−1,1) lies in the second quadrant, so θ=180°−45°=135°. Then (−1+i)3=22cis(405°)=22cis(45°)=2+2i. Directly, (−1+i)2=1−2i+i2=−2i and −2i(−1+i)=2i−2i2=2+2i.
Why it happens: arctan only returns angles between −90° and 90°, so it never gives the direction of a point with a<0.
How to check: the signs of cosθ and sinθ must match the signs of a and b. Here cos(−45°)>0 while a=−1<0.
Keeping only one root
An equation zn=w is solved with the single root that real-number experience suggests, and the other n−1 roots are never looked for.
Wrong working: z4=16, so z=2, or at most z=±2.
Right working: 16=16cis(0°), so zk=161/4cis(4360°k)=2cis(90°k) for k=0,1,2,3. The four roots are 2, 2i, −2 and −2i; for example (2i)4=16i4=16.
Why it happens: the radical sign and the exponent n1 name a single value on the positive reals, and that habit is carried over to C.
How to check: count the roots. A nonzero number has exactly n distinct n-th roots, evenly spaced n360° apart on one circle.
Adding the full turn after dividing
In the roots formula, the angle ϕ is divided by n first and the multiples of 360° are added afterwards. Every value of k then gives the same direction.
Wrong working: for z3=8i=8cis(90°), the angles 390°+360°k=30°,390°,750° all point the same way, so the only root found is 2cis(30°)=3+i.
Right working: the full turn is added before dividing, α=390°+360°k=30°,150°,270°. The roots are 2cis(30°)=3+i, 2cis(150°)=−3+i and 2cis(270°)=−2i. For example (−2i)3=−8i3=8i.
Why it happens: the identity cisθ=cis(θ+360°) is applied to the root's angle instead of to the angle of w, which is the one that can be written in many ways.
How to check: consecutive roots must differ by n360°, here 120°. If two of the listed roots coincide, the turns were added in the wrong place.
De Moivre's Theorem FAQ
How many nth roots does a complex number have?
+
Exactly n of them, all distinct, provided the number is not zero. The reason is that the argument is only determined up to full turns, so dividing it by n produces n genuinely different angles before the values start repeating. Real numbers are no exception: a nonzero real number has at most two real nth roots, and the rest are not real.Read more →
Why do the nth roots form a regular polygon?
+
Because all n share the same modulus, placing them on one circle, and their arguments differ by equal steps of a full turn divided by n. Equal radii and equal angular spacing is precisely the definition of a regular polygon inscribed in that circle.Read more →
What is the sum of all the nth roots of unity?
+
Zero, for every n greater than one. Geometrically the roots are symmetrically placed around the circle, so as vectors they cancel exactly. Algebraically it follows from the polynomial they satisfy, whose second coefficient is zero and therefore forces the sum of the roots to vanish.Read more →
How do you get the triple-angle formula from De Moivre's theorem?
+
Compute (cosθ+isinθ)3 in two ways. Expanding the cube gives the real part cos3θ−3cosθsin2θ, while De Moivre's theorem says the real part is cos3θ. Replacing sin2θ with 1−cos2θ gives cos3θ=4cos3θ−3cosθ, and the imaginary parts give sin3θ=3sinθ−4sin3θ the same way. Any n works, writing cosnθ and sinnθ as polynomials in cosθ and sinθ.Read more →
What is the product of all the nth roots of unity?
+
(−1)n+1: it is 1 when n is odd and −1 when n is even. The roots are exactly the roots of zn−1=(z−z0)(z−z1)⋯(z−zn−1), and comparing constant terms gives (−1)nz0z1⋯zn−1=−1. For n=4, 1⋅i⋅(−1)⋅(−i)=−1; for n=3, the two non-real roots are conjugates whose product is 1.Read more →
What do the distances from one corner of a regular polygon to all the others multiply to?
+
Exactly n, for a regular n-gon drawn in a circle of radius 1. Place the corners at the nth roots of unity, with the chosen corner at 1. Dividing zn−1 by z−1 leaves 1+z+⋯+zn−1, which is the product of z minus each other corner, and at z=1 it equals n. For a square the three distances are 2, 2 and 2, with product 4.Read more →
Does De Moivre's theorem work for fractional powers?
+
Not as a single formula. It holds for integer powers, where zn has one value. For the power n1, dividing the angle gives only one of the n roots, because θ can be replaced by θ+2πk before dividing. With −1=cisπ and n=2, the angle π gives cis2π=i, while the same number written as cis3π gives cis23π=−i; both square to −1.Read more →