This site is a work in progress. New lessons are added regularly. Contact us

Imaginary Numbers

Learning goals

  • Define ii by i2=1i^2 = -1
  • Reduce ini^n using the remainder of nn divided by four
  • Simplify n\sqrt{-n} as ini\sqrt{n}
  • Avoid the product rule with two negative radicands
  • Solve x2=kx^2 = -k as x=±kix = \pm\sqrt{k}\,i

Why the real numbers are not enough

You can solve x2=9x^2 = 9 because 32=93^2 = 9, and (3)2=9(-3)^2 = 9 as well, so it has the two answers x=3x = 3 and x=3x = -3. You can even solve x2=7x^2 = 7, whose answers ±7\pm\sqrt{7} are irrational but still perfectly real. Try x2=1x^2 = -1, though, and the search fails. No real number works, and the reason is not that we have not looked hard enough. It is built into how signs behave under multiplication.

Why no real number squares to a negative number#

Take any real number aa and look at its square a2=a×aa^2 = a \times a. There are only three possibilities for the sign of aa. If aa is positive, then a×aa \times a is a positive times a positive, which is positive. If aa is negative, then a×aa \times a is a negative times a negative, and two negatives make a positive, so a2a^2 is again positive. If aa is zero, then a2=0a^2 = 0.

In every case a20a^2 \ge 0, so a real square is never negative. That rules out any real solution of a2=1a^2 = -1, or a2=9a^2 = -9, or any equation that asks for a negative square. The gap is genuine and permanent, and closing it needs a number that does not sit on the real number line at all.

Rather than stop there, we do what mathematics has done before. When subtraction outgrew the counting numbers, negative numbers were invented to fill the gap; when division outgrew the integers, fractions were invented. Here we invent a single new number, and that one addition turns out to be enough.

Defining the imaginary unit

Define the imaginary unit ii to be a number whose square is 1-1:

i2=1.i^2 = -1.

That equation is the entire definition, and everything else in this lesson follows from it. Because squaring ii gives 1-1, we can also write i=1i = \sqrt{-1}, since ii is by construction the number that squares to 1-1.

With ii available, x2=1x^2 = -1 finally has answers. One is x=ix = i, because i2=1i^2 = -1 by definition. The other is x=ix = -i, because

(i)2=(1)2i2=(1)(1)=1(-i)^2 = (-1)^2\, i^2 = (1)(-1) = -1

as well. So x2=1x^2 = -1 has the two solutions x=ix = i and x=ix = -i, written together as x=±ix = \pm i, matching the same two-answer pattern you already know from x2=9x^2 = 9.

A real number multiplied by ii, such as 3i3i, 7i-7i, or 12i\tfrac{1}{2}i, is called a pure imaginary number. These really are new numbers: 3i3i is not equal to any real number, because its square is (3i)2=9i2=9(3i)^2 = 9i^2 = -9, and no real number squares to 9-9. When you join a real number and a pure imaginary number into a sum like 2+3i2 + 3i, the result is called a complex number. This lesson stays with ii itself, its powers, and pure imaginary numbers. The arithmetic that combines the two parts of a complex number is the subject of the next lesson.

The powers of i

Since ii is a number, you can raise it to powers, and something useful happens: the powers repeat in a short cycle. Work them out one at a time, each from the one before, leaning on i2=1i^2 = -1 at every step.

i1=i,i2=1,i3=i,i4=1.i^1 = i, \qquad i^2 = -1, \qquad i^3 = -i, \qquad i^4 = 1.

Why the powers of ii repeat every four steps#

Start from the definition and climb. The first power is ii itself. The second is i2=1i^2 = -1, the definition. For the third, multiply the second by one more factor of ii:

i3=i2i=(1)i=i.i^3 = i^2 \cdot i = (-1)\,i = -i.

For the fourth, multiply again, which is the same as squaring the second:

i4=i2i2=(1)(1)=1.i^4 = i^2 \cdot i^2 = (-1)(-1) = 1.

Reaching 11 is the key event. Multiplying by ii once more returns to the start, since i5=i4i=1i=ii^5 = i^4 \cdot i = 1 \cdot i = i, and from there the four values i,1,i,1i, -1, -i, 1 repeat forever in the same order. To evaluate any power ini^n, split the exponent into a multiple of 44 plus a remainder. Write n=4q+rn = 4q + r, where rr is the remainder of nn divided by 44, one of 0,1,2,30, 1, 2, 3. Then

in=i4q+r=(i4)qir=1qir=ir.i^n = i^{4q + r} = (i^4)^q \cdot i^r = 1^q \cdot i^r = i^r.

So ini^n depends only on rr: the value is fixed entirely by the remainder of the exponent when it is divided by 44. When that remainder is 00, the power is i0=1i^0 = 1.

In short, reduce the exponent modulo 44. A remainder of 11 gives ii, a remainder of 22 gives 1-1, a remainder of 33 gives i-i, and a remainder of 00 gives 11.

The four-step cycle of the powers of iFour boxes showing i to the first equals i, i squared equals negative 1, i cubed equals negative i, and i to the fourth equals 1, joined by times i arrows, with a return arrow from the last box back to the first.i⁴i-1-i1× i× i× i× i (the cycle repeats)
Multiplying by i steps forward through a four-value cycle (i, then -1, then -i, then 1) and returns to i. Because the pattern repeats every four steps, a power of i is fixed by the remainder of its exponent divided by 4.

Worked example 1 Evaluate i50i^{50}, i23i^{23}, and i100i^{100}

Divide each exponent by 44 and keep only the remainder.

For i50i^{50}, since 50=4×12+250 = 4 \times 12 + 2, the remainder is 22:

i50=i2=1.i^{50} = i^2 = -1.

For i23i^{23}, since 23=4×5+323 = 4 \times 5 + 3, the remainder is 33:

i23=i3=i.i^{23} = i^3 = -i.

For i100i^{100}, since 100=4×25100 = 4 \times 25 exactly, the remainder is 00:

i100=i0=1.i^{100} = i^0 = 1.

A number is divisible by 44 exactly when its last two digits form a multiple of 44, so 100100 lands on remainder 00 without any long division. For the same reason, only the tail of a big exponent ever matters.

Check your understanding

Simplify i27i^{27}.

Answer choices

Square roots of negative numbers

The imaginary unit lets you take the square root of any negative number, not just 1-1. The rule is short:

n=infor n>0.\sqrt{-n} = i\sqrt{n} \quad \text{for } n > 0.

Here n\sqrt{n} is the ordinary real square root you already know, and the factor of ii carries the negative sign.

Why n=in\sqrt{-n} = i\sqrt{n}#

To call something the square root of n-n, it has to square to n-n. Test the candidate ini\sqrt{n} by squaring it:

(in)2=i2(n)2=(1)(n)=n.\left(i\sqrt{n}\right)^2 = i^2 \left(\sqrt{n}\right)^2 = (-1)(n) = -n.

It works: ini\sqrt{n} squares to n-n, so it is the square root we want. The step (n)2=n\left(\sqrt{n}\right)^2 = n is just the meaning of a real square root, and i2=1i^2 = -1 supplies the sign. Splitting the radical as n=1n=in\sqrt{-n} = \sqrt{-1}\cdot\sqrt{n} = i\sqrt{n} is allowed here because only one of the two factors under the roots is negative. The next section shows why that restriction is not optional.

Worked example 2 Simplify 9\sqrt{-9}, 12\sqrt{-12}, and 72\sqrt{-72}

Pull out the factor of ii first, then simplify the real square root that is left.

The root 9\sqrt{-9} has a perfect square inside:

9=i9=3i.\sqrt{-9} = i\sqrt{9} = 3i.

The root 12\sqrt{-12} needs the radical simplified, using 12=4×312 = 4 \times 3 and 4=2\sqrt{4} = 2:

12=i12=i4×3=2i3.\sqrt{-12} = i\sqrt{12} = i\sqrt{4 \times 3} = 2i\sqrt{3}.

The root 72\sqrt{-72} works the same way, with 72=36×272 = 36 \times 2:

72=i72=i36×2=6i2.\sqrt{-72} = i\sqrt{72} = i\sqrt{36 \times 2} = 6i\sqrt{2}.

In each case the answer is a real number (or a simplified radical) times ii, that is, a pure imaginary number.

Check your understanding

Simplify 50\sqrt{-50}.

Answer choices

The trap: two negative radicands

There is one place where square roots of negatives bite back, and nearly everyone is caught by it once. You have used the product rule for square roots, ab=ab\sqrt{a}\cdot\sqrt{b} = \sqrt{ab}, to simplify radicals. That rule is valid only when aa and bb are not both negative. Apply it blindly to two negative radicands and you get the wrong answer, including the wrong sign.

Watch it fail on 49\sqrt{-4}\cdot\sqrt{-9}. The correct method rewrites each square root using ii first, and only then multiplies:

49=(2i)(3i)=6i2=6.\sqrt{-4}\cdot\sqrt{-9} = (2i)(3i) = 6i^2 = -6.

The tempting shortcut multiplies the radicands first, and it disagrees:

49(4)(9)=36=6.\sqrt{-4}\cdot\sqrt{-9} \ne \sqrt{(-4)(-9)} = \sqrt{36} = 6.

The two results, 6-6 and 66, are not equal, so the product rule genuinely breaks for two negatives. The fix is a firm habit: convert every square root of a negative into the form i  i\sqrt{\;} before doing anything else. Once the factors of ii are out in the open, they multiply like any other factors, and ii=i2=1i \cdot i = i^2 = -1 supplies the correct sign on its own.

Worked example 3 Evaluate 28\sqrt{-2}\cdot\sqrt{-8}

Convert each root to the i  i\sqrt{\;} form first. Here 2=i2\sqrt{-2} = i\sqrt{2}, and 8=i8=2i2\sqrt{-8} = i\sqrt{8} = 2i\sqrt{2}. Now multiply the two pure imaginary numbers:

(i2)(2i2)=2i2(22)=2(1)(2)=4.\left(i\sqrt{2}\right)\left(2i\sqrt{2}\right) = 2i^2\left(\sqrt{2}\cdot\sqrt{2}\right) = 2(-1)(2) = -4.

The wrong route, (2)(8)=16=4\sqrt{(-2)(-8)} = \sqrt{16} = 4, once again flips the sign. Convert first, and the answer 4-4 comes out right.

Check your understanding

Evaluate 425\sqrt{-4}\cdot\sqrt{-25}.

Answer choices

Solving x squared equals a negative

Putting the pieces together, you can now solve any equation of the form x2=kx^2 = -k with k>0k > 0, or equally x2+k=0x^2 + k = 0. Take the square root of both sides, and keep both signs.

Worked example 4 Solve x2+9=0x^2 + 9 = 0 and x2=20x^2 = -20

For x2+9=0x^2 + 9 = 0, move the constant across to isolate the square, then take the root of both sides:

x2=9x=±9=±3i.x^2 = -9 \quad\Rightarrow\quad x = \pm\sqrt{-9} = \pm 3i.

Both 3i3i and 3i-3i check out, since (±3i)2=9i2=9(\pm 3i)^2 = 9i^2 = -9. For x2=20x^2 = -20, the root needs simplifying, with 20=4×520 = 4 \times 5:

x=±20=±i20=±2i5.x = \pm\sqrt{-20} = \pm i\sqrt{20} = \pm 2i\sqrt{5}.

Every equation of this shape has two pure imaginary solutions, each the negative of the other.

A general quadratic can also lead to imaginary answers, but handling one that carries an xx term as well needs tools from the next chapters. For now, any equation that reduces to x2x^2 equal to a negative number is fully within reach.

Common mistakes

Practice

Multiple Choice Questions (MCQ)

Progressively harder sets of questions. Each opens on its own page.

Free Response Questions (FRQ)

Longer questions in parts, to be worked out on paper. Progressive hints, the answer on its own so you can check yourself and try again, then the full worked solution, plus a rubric to mark your own work against.

Free response Work it out on paper 5 questions Start →
More practice (optional)

Extra sets, as hard as the Challenge set. Each one opens on its own page.

More resources (optional)

Other explanations of this lesson, if you want a second take.

A bit of history (Optional)

The oldest square root of a negative number on record was quietly rubbed out.

It sits in a handbook on measuring solid shapes, written around the year 60. The author is usually taken to be Heron, an engineer in Alexandria, a city in Egypt. He is calculating the height of a pyramid with its top sliced off. The working reaches a subtraction. The larger of the two numbers is the one being taken away. So the height comes out as the square root of a negative.

Every surviving copy has that subtraction the other way round. The numbers were swapped, and the root came out positive and comfortable. Nobody knows whether Heron made the correction or a later scribe did. Either way, the negative was read as a slip of the pen rather than a result.

That reading held for another fifteen centuries. A minus sign under a square root meant you had blundered. It could also mean the question had no answer.

This lesson asks the opposite of you. The minus sign is information, not damage. It reports that the answer is off the real number line. It also says where the imaginary unit goes: the square root of a negative nn is ini\sqrt{n}. The scribe’s instinct survives in one place, and it still costs you. Letting the two minus signs cancel under the roots is exactly what turns 49\sqrt{-4}\cdot\sqrt{-9} into 66, when the honest answer is 6-6.