Choosing What to Multiply By Is the Whole Job of a Rationalize the Denominator Calculator
About the Author
Marko Šinko
Co-Founder & Lead Developer, AI Math Calculator
A rationalize the denominator calculator makes exactly one decision, and it is not the arithmetic. Multiplying out is mechanical — any student who can expand brackets can do it. The decision is what to multiply by, and that depends only on the shape of the bottom of the fraction, never on how big the numbers are. Get the shape right and the rest is bookkeeping. Get it wrong and you end up with a denominator that is messier than the one you started with.
The Four Shapes Any Rationalize the Denominator Calculator Must Tell Apart
Every rationalizing problem in a school or first-year syllabus falls into one of four patterns. Identify the pattern, read off the multiplier, and you are done deciding:
| Denominator | Multiply by | Why the root dies | Example |
|---|---|---|---|
| c√q | √q | √q × √q = q | 3/(2√5) → 3√5/10 |
| a ± √b | a ∓ √b | a² − b, no cross terms | 4/(3 − √5) → 3 + √5 |
| ⁿ√(pᵏ) | ⁿ√(pⁿ⁻ᵏ) | exponents add to n | 5/∛4 → 5∛2/2 |
| p ± qi | p ∓ qi | i² = −1 makes p² + q² | (3+2i)/(1−4i) → (−5+14i)/17 |
Three of those four rows are the same identity wearing different clothes: (x − y)(x + y) = x² − y². Squaring is what destroys a square root, and the conjugate is simply the partner that makes every leftover cross term cancel. Only the third row — higher-index roots — needs different thinking, and that is where most people get stuck.
When the Bottom Is a Single Square Root
Take 3/(2√5). Multiply top and bottom by √5, because √5 × √5 = 5 exactly. The bottom becomes 2 × 5 = 10 and the top becomes 3√5, so the answer is 3√5/10 ≈ 0.67082039. Check it against the original: 3 ÷ 4.47213595 = 0.67082039. Same number, different clothes.
The trap in this shape is not the multiplication — it is skipping the simplification that should come first. Try 7/√8. Multiply by √8 and you get 7√8/8, then you still have to notice that √8 = 2√2, giving 14√2/8, then reduce to 7√2/4. Simplify √8 to 2√2 before you multiply and the numbers stay small the whole way: 7/(2√2) × √2/√2 = 7√2/4 in a single step. Our square root calculator will break a radicand into its perfect-square factors if the factorisation is not obvious, and the simplify calculator handles the same job inside a longer expression.
Two Terms, and the Sign That Has to Flip
With two terms on the bottom, multiplying by the same two terms achieves nothing. Watch: (3 − √5)(3 − √5) = 9 − 6√5 + 5 = 14 − 6√5. Still irrational, and now uglier. Flip the middle sign instead. (3 − √5)(3 + √5) = 9 + 3√5 − 3√5 − 5 = 4. The two cross terms are equal in size and opposite in sign, so they annihilate — that is the entire mechanism.
So 4/(3 − √5) becomes 4(3 + √5)/4 = 3 + √5 ≈ 5.23606798. Both terms of the denominator can carry roots too: for 1/(√3 + √2) the conjugate is √3 − √2, the denominator becomes 3 − 2 = 1, and the answer is just √3 − √2 ≈ 0.31783725. A denominator of 1 looks like a trick, but it is common whenever the two radicands differ by 1.
One case fools people: 2√3 + 5√3 is not a conjugate problem. Both terms carry the same root, so they add first into 7√3, and you are back to a single radical. The calculator above detects this and switches methods rather than producing a pointless conjugate.
Cube Roots Don't Have Conjugates
Squaring kills a square root because the exponents 1/2 and 1/2 add to 1. For a cube root you need exponents that add to 1 in thirds, so ∛4 = ∛(2²) needs one more factor of ∛2, not another ∛4. Multiply 5/∛4 by ∛2/∛2 and the bottom becomes ∛8 = 2, giving 5∛2/2 ≈ 3.14980262.
Plenty of tutorials tell you to multiply by ∛(4²) = ∛16 instead. That works, but it is the long way round: you get 5∛16/∛64 = 10∛2/4, and then you still have to reduce to 5∛2/2. The rule that avoids the detour is worth memorising — for a denominator ⁿ√(pᵏ), multiply by ⁿ√(pⁿ⁻ᵏ), so the exponents inside sum to exactly n. Written with fractional exponents the whole thing is obvious: p2/3 × p1/3 = p1.
When the Denominator Ends in i
Clearing an imaginary unit uses the same conjugate, and i² = −1 does the work that squaring a root does elsewhere. For (3 + 2i)/(1 − 4i), multiply by (1 + 4i)/(1 + 4i). The bottom becomes 1 − (4i)² = 1 + 16 = 17. The top expands to 3 + 12i + 2i + 8i² = −5 + 14i, so the answer is (−5 + 14i)/17, or −0.29411765 + 0.82352941i.
Notice that the denominator came out as 1² + 4², a sum of squares, and sums of squares are never negative. That is a free error check: any complex rationalisation whose denominator comes out negative has a sign mistake in it. For arithmetic that goes beyond clearing the bottom, the complex number calculator handles full addition, multiplication and polar conversion, while the imaginary number calculator covers powers of i on their own.
The Reverse Trick Calculus Needs More Often
Sooner or later a course asks you to rationalize the numerator instead, and it is not a curiosity — it is how the derivative of √x gets derived. Start with the difference quotient (√(x+h) − √x)/h. Substituting h = 0 gives 0/0 and tells you nothing. Multiply top and bottom by (√(x+h) + √x) and the top collapses to (x + h) − x = h. Cancel the h and you are left with 1/(√(x+h) + √x), which at h = 0 is a perfectly ordinary 1/(2√x).
The same move rescues floating-point arithmetic, and this is the one place where rationalizing is not a matter of convention. Ask a computer for 1/(√(10⁸+1) − √(10⁸)) and the subtraction runs first: 10000.00005 minus 10000, two numbers that agree to nine significant digits. Eight of the sixteen digits a double carries are destroyed in that cancellation, and the result comes back as 19999.9997776 when the true value is 20000.00005. Rationalize the denominator and the same expression is simply √(10⁸+1) + √(10⁸) — an addition, no cancellation, every digit intact.
Four Places the Answer Goes Wrong
Almost every wrong answer in this topic is one of these, and none of them is a failure of understanding — they are bookkeeping slips:
- Reusing the denominator instead of its conjugate. Multiplying 1/(2 + √3) by (2 + √3) gives 7 + 4√3 on the bottom. The sign has to flip, always.
- Losing the minus while distributing. 4(3 + √5) is 12 + 4√5, not 12 + √5. The coefficient hits both terms.
- Leaving a negative denominator. 1/(√5 − 3) has denominator 5 − 9 = −4, so the answer is −(√5 + 3)/4. Conventional form keeps the minus sign on top.
- Cancelling into the radical. In 6√2/4 the common factor 2 divides the coefficient 6 and the denominator 4, giving 3√2/2. It does not touch the 2 under the root — that number is not a factor of anything outside it. The simplify fractions calculator is a quick sanity check when the coefficients get large.
Is Rationalizing Still Worth Doing?
The original reason was arithmetic by hand. Before calculators, a table gave you √2 = 1.41421356 and you needed 1/√2. Dividing 1 by an eight-digit number is a page of long division; halving 1.41421356 to get 0.70710678 takes about four seconds. Rationalizing converted a hard operation into an easy one, and that is why the convention exists at all.
That reason is gone, but three others are not. Answer keys and automatic graders compare against a canonical form, so 1/√2 is often marked wrong where √2/2 is marked right — the same reason every trig table prints sin 45° as √2/2. Equal expressions become recognisable: nothing about 1/(√3 + √2) suggests it equals √3 − √2 until you rationalize it. And as the floating-point example above shows, the rationalized twin of an expression is sometimes the numerically stable one. Wikipedia's entry on rationalisation in mathematics traces the convention back through the same table-based arithmetic.
Keep this short reference nearby — these eight cover most of what a homework set will throw at you:
| Expression | Rationalized | Decimal |
|---|---|---|
| 1/√2 | √2/2 | 0.70710678 |
| 1/√3 | √3/3 | 0.57735027 |
| 2/√3 | 2√3/3 | 1.15470054 |
| 1/√5 | √5/5 | 0.44721360 |
| 1/(2 + √3) | 2 − √3 | 0.26794919 |
| 1/(√3 + √2) | √3 − √2 | 0.31783725 |
| 1/(√5 − 2) | √5 + 2 | 4.23606798 |
| 1/∛2 | ∛4/2 | 0.79370053 |
Read down that middle column and the pattern of the whole topic shows up in miniature: the root never disappears, it just moves upstairs where it is easy to live with.
