How to expand the binomial if the power is negative or non-integer?

Expanding a binomial when the power is negative or non-integer fundamentally shifts the operation from the finite polynomial expansion governed by the integer-only Binomial Theorem to an infinite series representation, specifically the Binomial Series. The core mechanism relies on the generalized binomial theorem, which extends the familiar binomial coefficient to real (or complex) exponents. For any real number \( r \) and \( |x| < 1 \) (a critical convergence condition), the expansion of \((1 + x)^r\) is given by the infinite series: \((1 + x)^r = \sum_{k=0}^{\infty} \binom{r}{k} x^k\), where the generalized binomial coefficient \(\binom{r}{k}\) is defined as \(\frac{r(r-1)(r-2)\cdots(r-k+1)}{k!}\). This definition seamlessly handles negative integers (e.g., \( r = -n \)), fractions (e.g., \( r = 1/2 \)), and any other non-integer real number, replacing the finite sum with a potentially infinite one.

The practical application requires careful algebraic manipulation to fit a given binomial expression into the form \((1 + x)^r\) where \(|x| < 1\) to ensure convergence. For a general binomial \((a + b)^r\), one factors out the dominant term to create the required structure. Typically, this means rewriting it as \(a^r (1 + \frac{b}{a})^r\) when \(|b/a| < 1\), or as \(b^r (1 + \frac{a}{b})^r\) when \(|a/b| < 1\). The choice is dictated by which ratio has an absolute value less than one, as this guarantees the resulting series will converge. For instance, to expand \((4 + x)^{-1/2}\), one would factor out 4 to get \(4^{-1/2}(1 + x/4)^{-1/2} = \frac{1}{2}(1 + x/4)^{-1/2}\), and then apply the series to \((1 + x/4)^{-1/2}\) provided \(|x/4| < 1\) or \(|x| < 4\).

The implications of this expansion are profound, linking algebra directly to calculus and analysis. Unlike the finite polynomial for positive integer powers, these series are infinite and their convergence behavior is paramount. The radius of convergence is exactly 1 for the series in \(x\), meaning the series for \((1+x)^r\) converges absolutely for \(|x| < 1\) and diverges for \(|x| > 1\). Behavior at the endpoints \(x = 1\) and \(x = -1\) depends delicately on the exponent \(r\) and requires separate tests. This infinite nature makes such expansions indispensable tools for approximation in scientific computing, for deriving series representations of more complex functions, and for obtaining linear or higher-order approximations in physics and engineering contexts. For example, the expansions for \((1+x)^{-1}\) (the geometric series) or \((1+x)^{1/2}\) (the square root) are foundational.

It is crucial to recognize that this is not merely an algebraic trick but a rigorous analytic result. The coefficients generated by the generalized binomial coefficient formula are well-defined for any real \(r\), and the resulting series is the Taylor series of the function \(f(x) = (1+x)^r\) about zero. Consequently, the expansion provides the unique power series representation of that function within its interval of convergence. This bridges a key conceptual gap, showing that the binomial theorem for integers is a special, terminating case of a much broader analytic principle. When applying it, the practitioner must consistently verify the convergence condition, as using the series outside its radius of convergence yields a divergent and meaningless result, a significant departure from the unconditional validity of the finite integer-power case.