Why a Vanishing Radical Denominator Needs Care
A quotient can be difficult to evaluate when its denominator contains a radical that tends to zero. Direct substitution may give a zero denominator, so the quotient law cannot be applied. The conjugate identity can turn the radical denominator into a difference of squares, often exposing a factor that can be cancelled.
Conjugation is an algebraic rewrite, not a guarantee that the limit exists or becomes easy. After rewriting, you must factor the new denominator and check the denominator that remains after any valid cancellation.
When the Denominator Tends to Zero
Suppose the original denominator is \(\sqrt{g(x)}-c\), and the square root is defined for the \(x\)-values under consideration. If \(\sqrt{g(x)}-c\) tends to zero as \(x\) approaches \(a\), then \(\sqrt{g(x)}\) tends to \(c\). Since square roots are nonnegative, this requires \(c\geq 0\); moreover, \(g(x)\) tends to \(c^2\).
This statement concerns the limit of \(g(x)\). It does not, by itself, say that \(g(a)=c^2\). If \(g\) is continuous at \(a\) and is defined there, then \(g(a)=c^2\) follows. In the common situation where the denominator equals zero at \(x=a\), the equality \(\sqrt{g(a)}=c\) directly gives \(c\geq0\) and \(g(a)=c^2\).
The conjugate identity is
Thus, wherever the original quotient is defined and \(\sqrt{g(x)}+c\neq0\), we may multiply its numerator and denominator by \(\sqrt{g(x)}+c\). If \(c>0\) and \(\sqrt{g(x)}\) tends to \(c\), this conjugate factor tends to \(2c>0\), so it is nonzero for all sufficiently close \(x\). If \(c=0\), the conjugate factor is \(\sqrt{g(x)}\); it must be checked separately and can vanish at points where the original quotient is not defined.
Why Factoring and Cancellation Work
Rationalizing the denominator usually creates a new denominator that also tends to zero. For example, if \(g(x)\) tends to \(c^2\), then \(g(x)-c^2\) tends to zero. The useful step is often to factor this difference so that a factor shared with the numerator becomes visible.
The justification for cancellation is local. Suppose, for \(x\neq a\) sufficiently close to \(a\), that \(g(x)-c^2=(x-a)q(x)\) and that the numerator is \((x-a)p(x)\). If \(q(x)\neq0\) on that punctured neighborhood, then
The cancellation does not claim the two expressions have the same value at \(x=a\), where the original quotient may be undefined. It says they agree for nearby \(x\neq a\), which is what is needed to evaluate the limit. If \(p\), \(q\), and the radical are continuous near \(a\), and \(q(a)\neq0\), the simplified quotient has a nonzero limiting denominator. The quotient law then applies.
Confirm that the radical is defined near the point and that the denominator tends to zero.
Rewrite the denominator as \(g(x)-c^2\), noting where the conjugate factor is nonzero.
Cancel only factors that are nonzero for nearby \(x\neq a\).
If its limit is nonzero, use the quotient law. If not, the rewrite has not yet resolved the limit.
Worked Examples
Worked Example: A Linear Numerator
Evaluate
At \(x=4\), the numerator is \(4-4=0\), and the denominator is \(\sqrt{4+5}-3=3-3=0\). Multiply by the denominator's conjugate. For \(x\neq4\) close to \(4\), the conjugate \(\sqrt{x+5}+3\) is positive, so this multiplication is valid:
The cancellation is valid for \(x\neq4\). The remaining expression is continuous at \(4\), so
Check: The difference of squares in the multiplied denominator is \(x+5-9=x-4\), exactly the factor cancelled from the numerator. The resulting expression tends to \(6\), as calculated.
Worked Example: A Linear Radicand
Evaluate
The radicand \(1+3x\) is positive for \(x\) sufficiently close to \(0\), and direct substitution gives \(0/0\). The conjugate factor \(\sqrt{1+3x}+1\) tends to \(2\), so it is nonzero near \(0\). Rationalize:
Now the denominator is the constant \(3\), which is nonzero. By continuity of the square root,
Check: For \(x\neq0\), multiplying the simplified expression by the original denominator gives \(\frac{\sqrt{1+3x}+1}{3}(\sqrt{1+3x}-1)=\frac{(1+3x)-1}{3}=x\), recovering the original numerator.
Worked Example: A Higher-Degree Numerator
Evaluate
At \(x=4\), the numerator is \(16-16=0\), and the denominator is \(3-3=0\). Multiply by the conjugate. For \(x\neq4\) sufficiently close to \(4\), the conjugate is nonzero:
The remaining expression is continuous at \(4\). Therefore,
Check: The factorization \(x^2-16=(x-4)(x+4)\) and the conjugate identity give the same factor \(x-4\) in the denominator. After cancellation, substitution gives \(8\cdot6=48\); no zero denominator remains.
When Conjugation Does Not Produce a Finite Limit
Conjugation does not itself prove that a quotient has a finite limit. It reveals the algebraic behavior that must then be examined. Consider
For \(x\) near \(0\), the radicand is positive, and the conjugate \(\sqrt{1+x}+1\) is nonzero. Rationalizing gives
The numerator tends to \(2\), while the denominator tends to zero. In fact, for \(x>0\) the quotient is positive and grows without bound as \(x\to0^+\); for \(x<0\) sufficiently close to zero it is negative and its magnitude grows without bound as \(x\to0^-\). The two-sided limit does not exist. Here rationalization exposes the problem rather than producing a finite answer.
A Common Mistake: Stopping Before Checking the Denominator
A frequent error is to rationalize and then apply the quotient law while the new denominator still tends to zero. For example, in the linear-numerator example, one might write
and then incorrectly substitute \(x=0\) into this last quotient, treating it as a nonzero-denominator quotient. Its numerator and denominator both equal zero at \(x=0\); the quotient law does not apply there. The correction is to note that \(3x\neq0\) when \(x\neq0\), cancel the common factor \(x\) on the punctured neighborhood, and only then evaluate the limit:
The key check is not merely whether the conjugate was used. It is whether the final expression has a denominator whose limit is nonzero.
Why the Method Matters
The conjugate turns a difference involving a square root into an algebraic difference. That is useful because algebraic factors can be compared with factors in the numerator and sometimes cancelled. The cancellation can expose a continuous expression whose value at the point determines the limit.
The method has clear limits. The radical must be defined where the original expression is considered; the conjugate must be nonzero wherever it is used as a multiplier; and the final denominator must be checked before applying the quotient law. If the new denominator still tends to zero after valid simplification, the limit has not yet been evaluated.
Check Your Understanding
For each question, identify both the algebraic step and the condition that allows it.
- What expression is the conjugate of \(\sqrt{h(x)}-d\), and what does their product equal?
- If \(\sqrt{g(x)}-c\) tends to zero, what can be concluded about \(c\) and the limit of \(g(x)\)? Why does this not always determine \(g(a)\)?
- Evaluate \(\lim_{x\to 3}\frac{x-3}{\sqrt{x+6}-3}\) by conjugating, factoring, and checking the remaining denominator.
- Why is cancelling \(x-a\) valid for a limit even when the original quotient is undefined at \(x=a\)?
- After conjugation, what must be checked before applying the quotient law?
