Skip to main content

Roots of unity - TCE Mathematics Specialised (Tasmania)

Syllabus dot point

“Find the nth roots of unity and use their symmetry and algebraic properties.”

TCESpecialist MathematicsUnit 38 min read

Quick answer

The nnth roots of unity are the solutions of zn=1z^n = 1. They are zk=cis⁡2πknz_k = \operatorname{cis}\tfrac{2\pi k}{n} for k=0,1,…,n−1k = 0, 1, \dots, n-1, and they sit equally spaced around the unit circle, forming the vertices of a regular nn-sided polygon with one vertex at z=1z = 1.

What this dot point is asking

This dot point is a focused special case of finding roots of complex numbers, but it deserves its own treatment because the roots of unity have a beautiful structure that recurs throughout the course. Solving zn=1z^n = 1 is the cleanest possible application of De Moivre's theorem, and the answers carry symmetry that simplifies sums, products and factorisations.

Setting up the roots

Write 11 in polar form. Its modulus is 11 and its argument is 00, but we must allow all coterminal arguments 0+2πk0 + 2\pi k to capture every root. So 1=cis⁡(2πk)1 = \operatorname{cis}(2\pi k) for integer kk. Then

zn=cis⁡(2πk)⟹z=cis⁡ ⁣(2πkn),k=0,1,…,n−1. z^n = \operatorname{cis}(2\pi k) \quad\Longrightarrow\quad z = \operatorname{cis}\!\left(\frac{2\pi k}{n}\right), \qquad k = 0, 1, \dots, n-1.

Each root has modulus 11, so all nn roots lie on the unit circle. The arguments increase in equal steps of 2πn\tfrac{2\pi}{n}, so the roots are evenly spaced, like the numbers on a clock face.

Primitive root and the standard notation

Write ω=cis⁡2πn\omega = \operatorname{cis}\tfrac{2\pi}{n}. Then every nnth root of unity is a power of ω\omega, namely 1,ω,ω2,…,ωn−11, \omega, \omega^2, \dots, \omega^{n-1}. The number ω\omega is called a primitive nnth root of unity because its powers generate them all.

The geometry

Because the roots are vertices of a regular polygon centred at the origin, they have a strong rotational symmetry. Multiplying any root by ω\omega rotates it to the next root anticlockwise. This polygon picture is often the quickest way to write down all roots without separate calculation for each.

The fifth roots of unity

Worked example

Solve z5=1z^5 = 1 and describe the roots geometrically.

The roots are zk=cis⁡2πk5z_k = \operatorname{cis}\tfrac{2\pi k}{5} for k=0,1,2,3,4k = 0, 1, 2, 3, 4:

z0=cis⁡0=1,z1=cis⁡2π5,z2=cis⁡4π5,z3=cis⁡6π5,z4=cis⁡8π5. z_0 = \operatorname{cis}0 = 1, \quad z_1 = \operatorname{cis}\tfrac{2\pi}{5}, \quad z_2 = \operatorname{cis}\tfrac{4\pi}{5}, \quad z_3 = \operatorname{cis}\tfrac{6\pi}{5}, \quad z_4 = \operatorname{cis}\tfrac{8\pi}{5}.

All five have modulus 11, so they lie on the unit circle, spaced 2π5=72∘\tfrac{2\pi}{5} = 72^\circ apart. They form a regular pentagon with one vertex at z=1z = 1. The non-real roots occur in two conjugate pairs: z4=z1‾z_4 = \overline{z_1} and z3=z2‾z_3 = \overline{z_2}, since the polygon is symmetric about the real axis.

The sum and product of the roots

Sum of the roots of unity

For n≥2n \ge 2, the nnth roots of unity sum to zero:

1+ω+ω2+⋯+ωn−1=0. 1 + \omega + \omega^2 + \dots + \omega^{n-1} = 0.

This follows from the geometric series formula, since the sum equals ωn−1ω−1=1−1ω−1=0\dfrac{\omega^n - 1}{\omega - 1} = \dfrac{1 - 1}{\omega - 1} = 0. Geometrically the vectors to the vertices of a regular polygon cancel by symmetry.

This sum result is a favourite exam tool. For example, the real parts of the roots sum to zero, which gives identities such as cos⁡2π5+cos⁡4π5+cos⁡6π5+cos⁡8π5=−1\cos\tfrac{2\pi}{5} + \cos\tfrac{4\pi}{5} + \cos\tfrac{6\pi}{5} + \cos\tfrac{8\pi}{5} = -1 once you separate the root z0=1z_0 = 1.

Connection to factorisation

Because the nnth roots of unity are exactly the solutions of zn−1=0z^n - 1 = 0, the polynomial factorises as

zn−1=(z−1)(z−ω)(z−ω2)⋯(z−ωn−1). z^n - 1 = (z - 1)(z - \omega)(z - \omega^2)\cdots(z - \omega^{n-1}).

Dividing by z−1z - 1 gives the useful identity zn−1+zn−2+⋯+z+1=(z−ω)⋯(z−ωn−1)z^{n-1} + z^{n-2} + \dots + z + 1 = (z - \omega)\cdots(z - \omega^{n-1}), which links the geometric series to the non-trivial roots.

Stopping after finding only the real roots

Students often solve zn=1z^n = 1 by writing z=1z = 1 (and z=−1z = -1 when nn is even) and stop, missing the complex roots. There are always exactly nn roots, counted with the full range k=0k = 0 to n−1n-1. Use De Moivre to list every argument, and check you have nn distinct points equally spaced on the unit circle before finishing.

Properties of the cube roots: a worked staple

The cube roots of unity 1,ω,ω21, \omega, \omega^2 appear so often that their two key identities are worth memorising: 1+ω+ω2=01 + \omega + \omega^2 = 0 and ω3=1\omega^3 = 1. These collapse otherwise messy expressions in a single line.

Simplifying with the cube roots of unity

If ω\omega is a non-real cube root of unity, evaluate (1+ω)(1+ω2)(1 + \omega)(1 + \omega^2) and (1−ω+ω2)(1 - \omega + \omega^2).

Step 1: Recall the two key identities

Because ω\omega is a cube root of unity, ω3=1\omega^3 = 1 and 1+ω+ω2=01 + \omega + \omega^2 = 0. The second identity rearranges to ω+ω2=−1\omega + \omega^2 = -1, which is the substitution that collapses products and sums involving all three cube roots.

Step 2: Expand and simplify the product (1+ω)(1+ω2)(1 + \omega)(1 + \omega^2)

Expand the brackets in the usual way:

(1+ω)(1+ω2)=1+ω2+ω+ω3. (1 + \omega)(1 + \omega^2) = 1 + \omega^2 + \omega + \omega^3.

Now substitute: ω+ω2=−1\omega + \omega^2 = -1 and ω3=1\omega^3 = 1:

=1+(−1)+1=1. = 1 + (-1) + 1 = 1.

Step 3: Simplify 1−ω+ω21 - \omega + \omega^2

Rewrite to reveal the sum-to-zero pattern. Add and subtract 2ω2\omega:

1−ω+ω2=(1+ω+ω2)−2ω=0−2ω=−2ω. 1 - \omega + \omega^2 = (1 + \omega + \omega^2) - 2\omega = 0 - 2\omega = -2\omega.

Final answer: (1+ω)(1+ω2)=1(1 + \omega)(1 + \omega^2) = 1 and 1−ω+ω2=−2ω1 - \omega + \omega^2 = -2\omega.

Why this matters

Roots of unity model anything periodic and symmetric, and their cancellation property turns awkward trigonometric sums into one-line answers. Recognising a hidden zn=1z^n = 1 inside a problem is a powerful shortcut that connects this topic to factorisation, geometric series, and the multiple-angle identities of complex powers.

Exam-style questions

Questions in the style of TASC exam questions on this dot point, each with a worked answer. They are written by ExamExplained unless tagged "Past paper"; the year shows the paper a question is modelled on.

Practice4 marks
(a) Write down the three roots of z3=1z^3 = 1, giving the non-real roots as eiθe^{i\theta} with −π<θ≤π-\pi < \theta \le \pi. (b) If one non-real root is ω\omega, show that the other non-real root is ω2\omega^2. (c) Verify that 1+ω+ω2=01 + \omega + \omega^2 = 0.
Show worked answer →

(a) The cube roots of unity are equally spaced around the unit circle at angles 00, 2π3\dfrac{2\pi}{3} and −2π3-\dfrac{2\pi}{3}. They are z=1z = 1, z=e2πi/3z = e^{2\pi i/3} and z=e−2πi/3z = e^{-2\pi i/3}. (2 marks)

(b) Let ω=e2πi/3\omega = e^{2\pi i/3}. Then ω2=e4πi/3=e−2πi/3\omega^2 = e^{4\pi i/3} = e^{-2\pi i/3} (subtracting 2π2\pi to bring the argument into (−π,π](-\pi, \pi]). This is exactly the other non-real root, so the two non-real roots are ω\omega and ω2\omega^2. (1 mark)

(c) The three roots 1,ω,ω21, \omega, \omega^2 are the roots of z3−1=(z−1)(z2+z+1)=0z^3 - 1 = (z - 1)(z^2 + z + 1) = 0. Since ω\omega and ω2\omega^2 are roots of z2+z+1=0z^2 + z + 1 = 0, the sum of all three roots equals the negative of the z2z^2 coefficient of z3−1z^3 - 1, which is 00. More directly, 1+ω+ω2=ω3−1ω−1=01 + \omega + \omega^2 = \dfrac{\omega^3 - 1}{\omega - 1} = 0 because ω3=1\omega^3 = 1 and ω≠1\omega \ne 1. (1 mark)

Practice6 marks
(a) Write down the sum of the geometric series ∑n=06zn\displaystyle\sum_{n=0}^{6} z^n. (b) Hence determine the solutions of z6+z5+z4+z3+z2+z+1=0z^6 + z^5 + z^4 + z^3 + z^2 + z + 1 = 0. (c) Write this polynomial as a product of three quadratic polynomials with real coefficients.
Show worked answer →

(a) The geometric series with first term 11, ratio zz and 77 terms sums to z7−1z−1\dfrac{z^7 - 1}{z - 1}, valid for z≠1z \ne 1. (1 mark)

(b) The polynomial equals z7−1z−1\dfrac{z^7 - 1}{z - 1}, so setting it to zero requires z7=1z^7 = 1 with z≠1z \ne 1. The solutions are the six non-trivial seventh roots of unity: z=e2πik/7z = e^{2\pi i k/7} for k=1,2,3,4,5,6k = 1, 2, 3, 4, 5, 6. (3 marks)

(c) These six roots form three conjugate pairs (kk with 7−k7 - k). Each pair gives a real quadratic z2−2cos⁡2πk7 z+1z^2 - 2\cos\dfrac{2\pi k}{7}\,z + 1. So the polynomial factors as (z2−2cos⁡2π7z+1)(z2−2cos⁡4π7z+1)(z2−2cos⁡6π7z+1)\left(z^2 - 2\cos\dfrac{2\pi}{7}z + 1\right)\left(z^2 - 2\cos\dfrac{4\pi}{7}z + 1\right)\left(z^2 - 2\cos\dfrac{6\pi}{7}z + 1\right). (2 marks)

Practice4 marks
Solve z5+1=0z^5 + 1 = 0. Give your answers in polar form reiθr e^{i\theta} with −π<θ≤π-\pi < \theta \le \pi.
Show worked answer →

Write −1-1 in polar form: −1=eiπ-1 = e^{i\pi}, and more generally −1=ei(π+2πk)-1 = e^{i(\pi + 2\pi k)} for integer kk. Then z5=ei(π+2πk)z^5 = e^{i(\pi + 2\pi k)}.

Taking fifth roots, z=ei(π+2πk)/5z = e^{i(\pi + 2\pi k)/5} for k=0,1,2,3,4k = 0, 1, 2, 3, 4, all with modulus r=1r = 1.

The arguments are π5\dfrac{\pi}{5}, 3π5\dfrac{3\pi}{5}, π\pi (from 5π5\dfrac{5\pi}{5}), and for the remaining two, 7π5\dfrac{7\pi}{5} and 9π5\dfrac{9\pi}{5}, which reduce into (−π,π](-\pi, \pi] as −3π5-\dfrac{3\pi}{5} and −π5-\dfrac{\pi}{5}.

So the five solutions are z=eiπ/5,e3iπ/5,eiπ=−1,e−3iπ/5z = e^{i\pi/5}, e^{3i\pi/5}, e^{i\pi} = -1, e^{-3i\pi/5} and e−iπ/5e^{-i\pi/5}. They are equally spaced on the unit circle. Markers reward writing −1-1 with the general argument and listing all five roots with arguments in the required range.

Practise this

Sources & how we know this

ExamExplained