A-Level H2 Mathematics: Complex Numbers — A Systematic Framework for Argand Diagrams, De Moivre's Theorem and Loci

Why Complex Numbers Trip Up Even Strong H2 Math Students

Complex Numbers is unusual among H2 Mathematics topics. It is not conceptually difficult in the way that, say, Vectors or Differential Equations can be — the core idea, that we extend the real numbers by defining i² = −1, is introduced in a single lesson. What actually costs students marks is that the topic is tested in three very different registers within the same paper: algebraic manipulation (solving equations, conjugate pairs), geometric reasoning (the Argand diagram, loci), and a slightly mechanical but easy-to-fumble technique (De Moivre's Theorem for powers and roots).

Students who treat these as three unrelated skill sets end up memorising formulas in isolation and then freezing when a question blends them — for instance, asking for the roots of an equation, then plotting them, then describing a locus through one of the roots. The fix is to stop thinking of a complex number as "a formula" and start thinking of it as one object with three interchangeable representations. Once that shift happens, most exam questions become a matter of choosing the right representation for the job, not remembering a new rule.

The Three Faces of a Complex Number

Every complex number z can be written in three equivalent forms. The skill the H2 syllabus is really testing is knowing which form to use when, and being able to convert fluently between them.

1. Cartesian form: z = x + yi

Best for addition, subtraction, and any question that gives you real and imaginary parts directly (e.g. solving simultaneous equations, or comparing coefficients).

2. Polar (modulus-argument) form: z = r(cosθ + i sinθ)

Here r = |z| = √(x² + y²) is the modulus (distance from the origin), and θ = arg(z) is the argument (angle from the positive real axis, measured anticlockwise). This is best for multiplication, division, and describing angles — which is exactly why it is the natural form for loci questions.

3. Exponential form: z = re^iθ

Algebraically identical to polar form (Euler's formula: e^iθ = cosθ + i sinθ), but far faster to write and manipulate, especially inside De Moivre's Theorem. Once students are comfortable with it, most of our JC students switch to exponential form by default for any multiplication, division, power, or root question.

The rule we drill into students: if the question involves +/−, use Cartesian. If it involves ×, ÷, powers, or roots, use polar/exponential. Choosing the wrong form is the single biggest source of unnecessary working in this topic.

The Conjugate: Your Most Reliable Tool

The conjugate of z = x + yi is z* = x − yi. Geometrically, it is the reflection of z across the real axis. Two properties matter far more than students initially realise:

• z + z* = 2x (always real) and z × z* = x² + y² = |z|² (always real, always non-negative). This is the standard trick for rationalising a complex denominator: multiply numerator and denominator by the conjugate of the denominator.
• If a polynomial equation has real coefficients, then complex roots always occur in conjugate pairs. This single fact lets you find every root of a cubic or quartic once you've found one complex root — without doing any further complex algebra.

Worked Example 1: Solving a Cubic with Complex Roots

Question: Given that z = 1 + 2i is a root of z³ − 4z² + 9z − 10 = 0, find the other two roots.

Step 1 — Use the conjugate pair property. Since all coefficients are real, z = 1 − 2i must also be a root.

Step 2 — Build a quadratic factor from the known pair. The quadratic with roots 1 + 2i and 1 − 2i is:
(z − (1+2i))(z − (1−2i)) = z² − 2z + (1² + 2²) = z² − 2z + 5.

(Notice this uses the same z × z* = |z|² fact from above — the constant term is automatically the modulus squared, so there's no need to expand brackets from scratch.)

Step 3 — Divide out. Dividing z³ − 4z² + 9z − 10 by z² − 2z + 5 gives a linear factor z − 2, so the third root is z = 2.

Answer: the three roots are 1 + 2i, 1 − 2i, and 2.

This example illustrates the general framework for cubic/quartic problems: find or verify one root (often given), pair it using conjugates, construct the resulting quadratic using the modulus-squared shortcut, then use long division or comparison of coefficients to find any remaining real root.

De Moivre's Theorem — Turning Powers and Roots into Arithmetic

De Moivre's Theorem states that for any real n:

[r(cosθ + i sinθ)]ⁿ = rⁿ(cos nθ + i sin nθ)

The single biggest "aha moment" for students is realising this converts a repeated multiplication problem into simple arithmetic on the modulus and argument: raise r to the power n, multiply θ by n. That's it. No expansion of brackets, no binomial theorem on complex numbers required.

Worked Example 2: Evaluating a Power

Question: Express z = −1 + i√3 in exponential form, then find z&sup6;.

Step 1 — Convert to polar/exponential form.
r = |z| = √((−1)² + (√3)²) = √4 = 2.
θ = arg(z): since the point (−1, √3) lies in the second quadrant, θ = π − arctan(√3/1) = π − π/3 = 2π/3.
So z = 2e^i(2π/3).

Step 2 — Apply De Moivre's Theorem.
z&sup6; = 2&sup6; · e^{i(6 × 2π/3)} = 64e^i4π.

Step 3 — Reduce the argument to the principal range. Since e^i4π = eⁱ⁰ (a full 2π rotation twice), z&sup6; = 64eⁱ⁰ = 64.

Try expanding (−1 + i√3)&sup6; directly by binomial expansion and you'll see exactly why this shortcut matters — it turns a multi-line algebraic slog into three lines of arithmetic.

Worked Example 3: Finding the n-th Roots of a Complex Number

Question: Find all cube roots of z = 8i, giving your answers in exponential form with argument in the range (−π, π].

Step 1 — Write z in exponential form. |8i| = 8, arg(8i) = π/2, so z = 8e^iπ/2.

Step 2 — Add multiples of 2π before taking the root. This is the step students most often skip, and it's the reason they only ever find one root instead of three. Since the argument is only defined up to a multiple of 2π, write:
z = 8e^{i(π/2 + 2kπ)}, for any integer k.

Step 3 — Take the cube root of both modulus and argument.
z^1/3 = 8^1/3 e^{i(π/6 + 2kπ/3)} = 2e^{i(π/6 + 2kπ/3)}.

Step 4 — Substitute k = 0, 1, 2 (three values, since it's a cube root) and reduce into the required range.
k = 0: 2e^iπ/6
k = 1: 2e^{i(π/6 + 2π/3)} = 2e^i5π/6
k = 2: 2e^{i(π/6 + 4π/3)} = 2e^i3π/2, which is outside (−π, π], so subtract 2π: 2e^{i(3π/2 − 2π)} = 2e^−iπ/2.

Answer: the three cube roots are 2e^iπ/6, 2e^i5π/6, and 2e^−iπ/2. Notice they are evenly spaced by 2π/3 around a circle of radius 2 — this is always true for n-th roots, and sketching that circle is a fast way to check your answers are correct on exam day.

Loci in the Argand Diagram — A Systematic Framework

Loci questions ask students to describe or sketch the set of points z that satisfy a given condition. There are really only three basic shapes tested at H2 level, and every exam question is a variation or combination of them.

Type 1: |z − a| = r → a circle

This says "the distance from z to the fixed point a is constant, equal to r." That is the definition of a circle of radius r centred at the point representing a.

Type 2: arg(z − a) = θ → a half-line

This says "the direction from a to z makes a fixed angle θ with the positive real axis." It is a ray starting at (and excluding) the point a, not a full line — a detail examiners test explicitly.

Type 3: |z − a| = |z − b| → a perpendicular bisector

This says "z is equidistant from a and b." That is the perpendicular bisector of the line segment joining a and b — a full straight line.

The framework we teach: before doing any algebra, identify which of these three shapes the equation matches, translate it into words, then sketch it. Almost every marks-losing error in this topic comes from jumping straight into x, y algebra instead of recognising the geometric shape first.

Worked Example 4: Combining Loci to Find a Range

Question: The locus of points representing z is given by |z − 3| = 2. Point P represents the complex number with the greatest argument on this locus, given arg(z) is measured in the usual way from the origin. Find arg(z) for point P.

Step 1 — Sketch first. |z − 3| = 2 is a circle centred at (3, 0) with radius 2. Since the radius (2) is less than the distance from the origin to the centre (3), the origin lies outside the circle — so there is a well-defined maximum argument, achieved where a line from the origin is tangent to the circle.

Step 2 — Use right-angle geometry, not complex algebra. At the point of tangency, the radius to that point is perpendicular to the line from the origin. This forms a right triangle with hypotenuse 3 (origin to centre) and opposite side 2 (the radius).

Step 3 — Solve. sin(α) = 2/3, where α is the angle between the line to the centre (which lies along the positive real axis) and the tangent line. So α = arcsin(2/3) ≈ 41.8°.

Answer: the maximum argument is approximately 41.8° (or 0.730 radians).

This example is typical of how H2 papers combine the loci framework with basic trigonometry — the complex numbers content is really just the setup; the actual mark-scoring step is recognising the tangent-line geometry.

Common Exam Traps

• Forgetting the argument is only defined up to a multiple of 2π. This is the single most common reason students miss roots in De Moivre's root-finding questions (Worked Example 3). Always add 2kπ before dividing by n.
• Giving the argument outside the required range. H2 papers almost always specify −π < θ ≤ π. An otherwise perfect answer loses marks if the final argument isn't adjusted into range.
• Treating arg(z − a) = θ as a full line instead of a half-line. Students often sketch the whole line through a, forgetting the locus only includes the ray in the direction of θ, excluding point a itself.
• Multiplying arguments instead of adding them. When multiplying two complex numbers in polar form, moduli multiply but arguments add (not multiply). This is a common slip under exam time pressure.
• Assuming complex roots come in pairs even when coefficients aren't real. The conjugate root theorem only applies to polynomials with real coefficients — always check this condition is actually given before using the shortcut.
• Skipping the sketch on loci questions. Many "describe the locus" or "find the range of values" questions become almost trivial once sketched, but very error-prone if attempted purely algebraically.

How to Practise Complex Numbers Systematically

We don't ask students to memorise a formula list for this topic — we ask them to build fluency in three specific skills, each with its own short, repeatable drill:

1. Form conversion drills. Given any z in Cartesian form, convert to exponential form (and back) in under 30 seconds, including correctly identifying the quadrant for the argument. This single skill underlies almost every other question type.
2. De Moivre's Theorem on both directions. Practise both "raise to a power" (Worked Example 2) and "find the n roots" (Worked Example 3) until the 2kπ step becomes automatic rather than something you have to remember to add.
3. Locus recognition before algebra. For every practice question, force yourself to write one sentence describing the shape ("circle centred at ___ with radius ___", "perpendicular bisector of ___ and ___", "half-line from ___ in direction ___") before attempting any calculation.

At Intuitional, we work through past-year A-Level questions using exactly this three-part structure, so students walk into the exam recognising which of the three "modes" a question is testing within the first few seconds of reading it — rather than trying to recall which formula applies. That recognition, more than any single formula, is what separates students who lose easy marks on Complex Numbers from those who don't.