How to raise a complex number in polar form to any integer power, how to find all n of its nth roots, and why those roots sit evenly spaced at the corners of a regular polygon.
In the real numbers the equation x3=8 has one solution, the number 2. In the complex numbers it has three, and they sit at the corners of an equilateral triangle centered at the origin. Polar form shows why. A power multiplies the angle, so a root divides it, and dividing a full turn by three leaves room for three different answers.
Key terms
De Moivre’s theorem
Definition Let z=rcisθ with r>0 and θ real. For every integer n,
zn=rncis(nθ)=rn(cosnθ+isinnθ).
For n≥1 this follows from the rule for products in polar form, since each new factor multiplies the modulus by r and adds θ to the argument. For n=0 both sides equal 1. For a negative integer n=−m the power means z−m=1/zm, and the quotient rule gives
rmcis(mθ)1=r−mcis(−mθ).
The theorem is about integer exponents. A fractional exponent such as z1/3 has several candidate values, and those are the roots in the sections below.
In plain words To raise a complex number to the nth power, raise its length to the nth power and multiply its angle by n. For a negative power −m, raise the reciprocal of the length to the power m and multiply the angle by −m.
Example The number 1+i has modulus 2 and argument 4π, so
(1+i)10=(2)10cis410π=32cis25π=32i,
because 25π is coterminal with 2π. Multiplying out ten factors would take much longer. A negative power works the same way. With 1−i=2cis(−4π),
(1−i)−4=(2)−4cisπ=−41.
As a check, (1−i)2=−2i, so (1−i)4=(−2i)2=−4, and its reciprocal is −41.
The nth roots of a complex number
Definition Let n be a positive integer and w a complex number. An nth root of w is a complex number z with zn=w. If w=Rcisφ with R>0, the nth roots of w are exactly the numbers
zk=R1/ncisnφ+2kπ,k=0,1,…,n−1,
where R1/n is the positive real nth root of R. Here is why. Write a candidate as z=rcisθ with r>0. By De Moivre’s theorem zn=rncis(nθ), and this equals Rcisφ exactly when rn=R and nθ=φ+2kπ for some integer k. The first condition forces r=R1/n and the second gives θ=(φ+2kπ)/n. Integers k outside 0,1,…,n−1 give no new numbers, as the next section shows. The only nth root of 0 is 0 itself.
In plain words Take the ordinary positive nth root of the length. For the angle, divide φ by n and then keep adding n2π. Each angle you get, multiplied by n, lands on φ plus some whole number of turns.
Example Find the cube roots of 8i. Since 8i=8cis2π, the roots have modulus 38=2 and arguments 6π+32kπ.
k
Argument
Root
0
6π
3+i
1
65π
−3+i
2
23π
−2i
The last root is the quickest to check, since (−2i)3=−8i3=8i.
Exactly n distinct roots
Definition Let w=0 and let n be a positive integer. Then w has exactly n distinct nth roots, the numbers z0,z1,…,zn−1 above. They all have the modulus R1/n, and their arguments
nφ,nφ+n2π,…,nφ+n2(n−1)π
differ from each other by more than 0 and less than 2π, so no two of them are coterminal. Any other integer k repeats one of these, because replacing k by k+n adds exactly 2π to the argument. For n≥2 a nonreal w has n equally good roots, so the symbol nw is ambiguous for it unless a particular root is singled out.
In plain words A nonzero number always has two square roots, three cube roots, four fourth roots, and so on. After n steps of n2π you have gone once around the circle and the list starts over.
Example Take the fourth roots of −16=16cisπ. They have modulus 2 and arguments 4π+2kπ for k=0,1,2,3, which gives
2+2i,−2+2i,−2−2i,2−2i.
With k=4 the argument is 49π, coterminal with 4π, so the list repeats. None of the four roots is real, which fits the fact that a real number raised to the fourth power is never negative.
Roots of unity
Definition For a positive integer n, the nth roots of unity are the solutions of zn=1. Since 1=cis0, they are
ωk=cisn2kπ,k=0,1,…,n−1.
With ω=cisn2π, De Moivre’s theorem gives ωk=ωk, so every nth root of unity is a power of ω. All of them have modulus 1 and lie on the unit circle. For n≥2 their sum is 0, because
(ω−1)(1+ω+ω2+⋯+ωn−1)=ωn−1=0
and ω=1.
In plain words Start at 1 on the unit circle and step around it by the angle n2π. After n steps you are back at 1, and the points you stood on are the nth roots of unity.
Example For n=3 the step is 32π, and the cube roots of unity are
1,−21+23i,−21−23i.
Their sum is 0, and the last two are conjugates. Squaring the second one gives
41−43−23i=−21−23i,
the third one, just as ω2 should.
Angular spacing
Definition Let w=0 and let n be a positive integer. Consecutive nth roots zk and zk+1 of w have the same modulus and arguments that differ by exactly n2π, so zk+1=ωzk with ω=cisn2π. Consequently, if u is any one nth root of w, the full list of roots is
u,uω,uω2,…,uωn−1.
Each of these is a root, because (uωk)n=un(ωn)k=w. They are n different numbers, because u=0 and the powers ω0,…,ωn−1 are different, so no root is missing.
In plain words Once you know one root, you know them all. Turn it about the origin by n2π again and again, which is the same as multiplying by the root of unity ω.
Example Since (1+i)2=2i, the fourth power is (1+i)4=(2i)2=−4, so 1+i is a fourth root of −4. Here ω=cis2π=i, and the other fourth roots are
(1+i)i(1+i)i2(1+i)i3=−1+i,=−1−i,=1−i.
They are a quarter turn apart, all at distance 2 from the origin.
Regular polygons
Definition Let w=0 and let n≥3 be an integer. The nth roots of w are the vertices of a regular n-gon centered at 0 and inscribed in the circle of radius ρ=∣w∣1/n. All vertices are at distance ρ from the center and neighboring vertices are n2π apart, so every side is the base of an isosceles triangle with legs ρ and apex angle n2π, and its length is
s=2ρsinnπ.
For n=2 the two square roots are the endpoints of a diameter of that circle, and each is the negative of the other.
In plain words Equal distances from the center and equal angles between neighbors are exactly what make a polygon regular. The roots have both, so connecting them in order gives a regular polygon.
Example The sixth roots of unity are cis3kπ for k=0,1,…,5.
k
Argument
Root
0
0
1
1
3π
21+23i
2
32π
−21+23i
3
π
−1
4
34π
−21−23i
5
35π
21−23i
They form a regular hexagon inscribed in the unit circle, with side
s=2sin6π=1.
Directly, the distance from 1 to 21+23i is 41+43=1.
Check yourself
Compute (1−3i)6.
Show the answer
The number 1−3i has modulus 2 and argument −3π, so
(1−3i)6=26cis(−2π)=64.
Find the three cube roots of −8 in rectangular form.
Show the answer
Since −8=8cisπ, the roots are 2cis3π+2kπ for k=0,1,2, which gives
2cis3π2cisπ2cis35π=1+3i,=−2,=1−3i.
The real cube root −2 is one of the three.
The number 2 is a sixth root of 64. Use the angular spacing to list the other five sixth roots.
Show the answer
Multiply 2 by the powers of ω=cis3π=21+23i. The roots 2ωk for k=1,…,5 are
1+3i,−1+3i,−2,−1−3i,1−3i.
Together with 2 they are the corners of a regular hexagon with side 2.
May we count page visits to understand which courses and learning features are useful? Cloudflare collects visit and performance statistics, without your account details or answers. Your choice does not affect access. Details