Rewriting Mixed Trigonometric Factors Before Integration

Why Rewrite a Trigonometric Product?

A product of tangent and secant factors, or cotangent and cosecant factors, may not initially display the derivative needed for a simple substitution. The factors are related, however, by squared identities. Replacing a selected squared factor can make the rest of the integrand a polynomial in one trigonometric function, while leaving its derivative factor in place.

The key is to make the rewrite serve a specific substitution. For example, the derivative of \(\sec x\) is \(\sec x\tan x\), and the derivative of \(\csc x\) is \(-\csc x\cot x\). When either of these factor pairs occurs, an appropriate squared identity can express the remaining factors using just \(\sec x\) or \(\csc x\).

The Identities and Derivative Factors

The Pythagorean identity \(\sin^2 x+\cos^2 x=1\), divided by \(\cos^2 x\) or by \(\sin^2 x\), gives the identities used here. They apply wherever the relevant trigonometric functions are defined.

The derivative factors to recognize are:

The negative signs in the last two derivatives matter. When using \(u=\csc x\) or \(u=\cot x\), the differential includes a minus sign. A valid rewrite must preserve the original integrand, including its sign, before that differential is substituted.

A Reliable Rewriting Process

Worked Example: Rewriting for a Secant Substitution

Worked Example: Rewriting for a Cosecant Substitution

Worked Example: Using the Cotangent Substitution

Why the Rewriting Works

The method is a direct consequence of substitution. If \(u=g(x)\), then \(du=g'(x)\,dx\). When the integrand can be written as \(F(g(x))g'(x)\,dx\), the substitution changes the integral to \(\int F(u)\,du\). The squared identities help produce this form: they express repeated tangent or cotangent factors in terms of the function selected for \(u\).

For instance, in the first example, the identity changes \(\tan^2 x\) into \(\sec^2 x-1\), while the remaining \(\sec x\tan x\,dx\) is exactly \(d(\sec x)\). In the second, \(\cot^2 x\) becomes \(\csc^2 x-1\), and the remaining \(\csc x\cot x\,dx\) is \(-d(\csc x)\). The substitution is useful because the rewritten expression depends only on \(u\), apart from the differential and its sign.

This reasoning also gives a practical test: after choosing a substitution and applying the identities, check that no unwanted \(x\)-dependent factor remains. If one does remain, that particular rewrite has not yet produced a single-variable integral.

Common Pitfall: Losing the Negative Sign

A frequent error is to substitute \(u=\csc x\) but use \(du=\csc x\cot x\,dx\). The correct derivative is \(du=-\csc x\cot x\,dx\). For the second example, omitting the minus sign would give the incorrect candidate \(\frac{\csc^4 x}{4}-\frac{\csc^2 x}{2}\). At \(x=\pi/4\), the original integrand is \(\cot^3(\pi/4)\csc^2(\pi/4)=2\). The derivative of that incorrect candidate at the same point is \(-2\), not \(2\).

Choosing a Useful Identity

The identities are equivalent, but they are not equally convenient for every substitution. If the intended variable is \(\sec x\), rewriting \(\tan^2 x\) as \(\sec^2 x-1\) usually makes the remaining expression a polynomial in \(\sec x\). If the intended variable is \(\csc x\), rewriting \(\cot^2 x\) as \(\csc^2 x-1\) serves the same role. For a tangent substitution, \(\sec^2 x=1+\tan^2 x\) can express a remaining secant square in tangent; for a cotangent substitution, use \(\csc^2 x=1+\cot^2 x\).

Not every mixed product has the right derivative factor or enough squared factors to make one of these rewrites useful. The identity does not guarantee that an integral becomes simple. It is a tool for reorganizing factors, not a reason to force a substitution that leaves the integrand more complicated.

These rewrites matter because they make the structure of a trigonometric product visible. A product that first appears to involve several functions can, after a careful identity and substitution, become an ordinary polynomial integral. A reliable solution records the factorization, the identity, and the differential explicitly; these steps make both the calculation and its sign easy to audit.