Power Series of ln(1+x): Formula, Derivation, and Examples
The power series of ln(1+x) is one of the most useful and most frequently tested expansions in all of calculus. Written out, it says that for values of x between -1 and 1, the natural logarithm of 1+x can be expressed as an infinite sum of powers of x, each divided by its exponent and alternating in sign. That single identity bridges polynomial behavior and logarithmic growth, and it gives you a practical way to compute logarithms, estimate integrals, and prove inequalities without ever touching a calculator.
The expansion matters because logarithms are awkward to evaluate directly but incredibly well behaved when rewritten as polynomials. Once ln(1+x) becomes x minus x²/2 plus x³/3 minus x⁴/4 and so on, every tool you already know for polynomials, such as term-by-term integration, differentiation, and comparison, becomes available. Understanding where the series comes from, where it converges, and how quickly it delivers accurate answers turns a memorized formula into a genuinely flexible technique.
Deriving the Power Series for ln(1+x) from Scratch
Almost every derivation of this series begins with the geometric series, because that is the one expansion students can build without any prior machinery. For any t with absolute value less than 1, the fraction 1/(1+t) equals 1 - t + t² - t³ + t⁴ - ... . This is simply the geometric series with ratio -t, and it converges precisely when |t| is smaller than 1. The whole trick is to recognize that 1/(1+t) is the derivative of ln(1+t), which means integration can turn a geometric series into a logarithmic one.
Integrating both sides from 0 to x is legal here because power series can be integrated term by term inside their interval of convergence. The left side becomes ln(1+x) minus ln(1), and ln(1) is zero, so the left side is exactly ln(1+x). The right side becomes x - x²/2 + x³/3 - x⁴/4 + ..., and just like that the expansion appears. Nothing about the result is accidental; it falls directly out of the geometry of the function 1/(1+t) and the rules for integrating series.
Why Term-by-Term Integration Is Permitted
Term-by-term integration is not a free pass that applies to every infinite sum. It works for power series because within the open interval of convergence the series converges uniformly on every closed subinterval, which is the technical condition that lets you swap an integral and an infinite sum without changing the answer. In practice, this means you can integrate the geometric series over any interval strictly inside (-1, 1) and trust the result completely.
That guarantee is what makes the derivation rigorous rather than merely suggestive. If you tried the same move on a series that only converged pointwise, the interchange could fail and produce a false identity. For ln(1+x), however, the uniform convergence inside (-1, 1) is automatic, so the integrated series is valid on the entire open interval and can be differentiated or integrated again as many times as you like.
The General Term and Summation Notation
To write the expansion compactly, note the pattern in the coefficients and signs. The first term is x¹/1 with a positive sign, the second is -x²/2, the third is +x³/3, and so on. The sign alternates starting with positive, and the denominator always matches the exponent. This is captured by the summation from n equals 1 to infinity of (-1)^(n+1) times x^n divided by n.
Expressing it with sigma notation is not just cosmetic. It lets you state the nth term, apply convergence tests cleanly, and bound the error after truncation. For example, the fifth partial sum is x - x²/2 + x³/3 - x⁴/4 + x⁵/5, which is exactly what you would get by plugging n equals 1 through 5 into the general term. That compact form becomes essential when you move on to error analysis and approximation.
Interval of Convergence and Endpoint Behavior
A power series is only useful where it actually converges, so the natural next question is how far from zero you can go before the expansion breaks down. Applying the ratio test to the coefficients 1/n gives a limit of |x|, which is less than 1 exactly when |x| is less than 1. That means the radius of convergence is 1, and the series converges absolutely for every x strictly between -1 and 1.
The endpoints require separate treatment because the ratio test is inconclusive there. At x equal to 1 the series becomes the alternating harmonic series, which converges by the alternating series test. At x equal to -1 the series becomes the negative harmonic series, which diverges to negative infinity. This asymmetry is why the full interval of convergence is written as -1 < x ≤ 1, with the left endpoint excluded and the right endpoint included.
Why the Radius of Convergence Is Exactly One
The radius of 1 is not an accident of the coefficients; it comes from the singularity of the function itself. The expression 1/(1+t), whose integral produced the series, blows up at t equal to -1, and that singularity sits exactly at distance 1 from the expansion point 0. A power series centered at a point cannot converge past the nearest singularity of the function it represents, so the boundary at x equal to -1 is built into the problem.
Testing the endpoints confirms the theory rather than contradicting it. At x equal to 1, the function ln(1+x) equals ln(2), a perfectly finite number, and the alternating harmonic series dutifully converges to it. At x equal to -1, the function itself is undefined because ln(0) does not exist, and the series correctly refuses to converge. The boundary of the interval mirrors the boundary of the domain, which is both elegant and a useful sanity check.
What Happens at x = 1 and x = -1
Substituting x equal to 1 into the expansion gives ln(2) equal to 1 - 1/2 + 1/3 - 1/4 + 1/5 and so on forever. This is one of the classic identities of analysis, and it is remarkable because it expresses a transcendental constant as an alternating sum of simple fractions. It also explains why the series converges so slowly there: the terms shrink like 1/n, so you need roughly a million terms to get six decimal places of accuracy.
Substituting x equal to -1 gives -1 - 1/2 - 1/3 - 1/4, which is the negative harmonic series and diverges. This failure is not a flaw in the derivation but a direct reflection of the fact that ln(1+x) tends to negative infinity as x approaches -1 from the right. The series tells you the truth at both endpoints: it converges where the function is defined and finite, and it diverges where the function is not.
Applying the Series to Estimation and Problem Solving
Once the expansion and its interval are settled, the series becomes a working tool rather than a fact to memorize. The most common use is approximation: replacing ln(1+x) with a few terms gives a polynomial that is easy to evaluate by hand and accurate to whatever precision you need, provided x is reasonably small. The alternating structure also supplies an error bound for free, which makes the approximations trustworthy rather than guesswork.
Beyond raw computation, the series is a gateway to related results. Integrating it term by term produces series for expressions built from logarithms, differentiating it recovers the geometric series, and substituting carefully chosen values of x generates identities for ln(2), ln(3/2), and many other constants. Each of these moves follows directly from the basic expansion, which is why mastering it pays dividends across calculus, analysis, and applied mathematics.
Numerical Approximation and Error Control
For an alternating series whose terms decrease in magnitude to zero, the error after stopping at any term is no larger than the absolute value of the next term. If you approximate ln(1.1) using x equal to 0.1 and truncate after the third term, the error is at most 0.1⁴/4, which is 0.000025. That is a remarkably precise answer from a three-term polynomial, and it shows why the series is so effective when x is small.
The situation changes dramatically as x approaches 1. Approximating ln(2) with the same series requires many terms because the error bound 1/(n+1) shrinks very slowly. In practice, you would use an accelerating transformation, a different expansion, or simply note that ln(2) equals -ln(1/2) and apply the series with x equal to -1/2, whose terms shrink geometrically and converge far faster. Choosing the right substitution is often more valuable than grinding through extra terms.
Substitutions, Pitfalls, and Practical Shortcuts
The most common mistake is applying the series outside its interval of convergence. Plugging in x equal to 5 will produce a divergent sum, not an approximation of ln(6). When you need a logarithm of a large number, rewrite it using ln(1+x) equal to ln(x) plus ln(1 + 1/x), then expand the second term with the small argument 1/x. This identity keeps the series inside its valid range and restores rapid convergence.
Another trap is forgetting that the expansion is for ln(1+x), not ln(x). If you need ln(1-x), substitute -x into the original series to get -x - x²/2 - x³/3 - ..., or equivalently take the negative of the series for ln(1+x) evaluated at -x. Keeping track of these sign and argument substitutions prevents the sign errors that plague exam solutions and hand calculations alike.
The next time you face a logarithm that resists direct evaluation, try expanding it and watching the polynomial terms do the heavy lifting. The series rewards patience with precision, and the more comfortable you become with its interval and error behavior, the more naturally it will slot into your problem-solving toolkit.